2019-05-19 12:08:55 +00:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2014-06-04 23:08:10 +00:00
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/spinlock.h>
|
|
|
|
|
|
|
|
#include <linux/mm.h>
|
2016-01-16 00:56:55 +00:00
|
|
|
#include <linux/memremap.h>
|
2014-06-04 23:08:10 +00:00
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/rmap.h>
|
|
|
|
#include <linux/swap.h>
|
|
|
|
#include <linux/swapops.h>
|
mm: introduce memfd_secret system call to create "secret" memory areas
Introduce "memfd_secret" system call with the ability to create memory
areas visible only in the context of the owning process and not mapped not
only to other processes but in the kernel page tables as well.
The secretmem feature is off by default and the user must explicitly
enable it at the boot time.
Once secretmem is enabled, the user will be able to create a file
descriptor using the memfd_secret() system call. The memory areas created
by mmap() calls from this file descriptor will be unmapped from the kernel
direct map and they will be only mapped in the page table of the processes
that have access to the file descriptor.
Secretmem is designed to provide the following protections:
* Enhanced protection (in conjunction with all the other in-kernel
attack prevention systems) against ROP attacks. Seceretmem makes
"simple" ROP insufficient to perform exfiltration, which increases the
required complexity of the attack. Along with other protections like
the kernel stack size limit and address space layout randomization which
make finding gadgets is really hard, absence of any in-kernel primitive
for accessing secret memory means the one gadget ROP attack can't work.
Since the only way to access secret memory is to reconstruct the missing
mapping entry, the attacker has to recover the physical page and insert
a PTE pointing to it in the kernel and then retrieve the contents. That
takes at least three gadgets which is a level of difficulty beyond most
standard attacks.
* Prevent cross-process secret userspace memory exposures. Once the
secret memory is allocated, the user can't accidentally pass it into the
kernel to be transmitted somewhere. The secreremem pages cannot be
accessed via the direct map and they are disallowed in GUP.
* Harden against exploited kernel flaws. In order to access secretmem,
a kernel-side attack would need to either walk the page tables and
create new ones, or spawn a new privileged uiserspace process to perform
secrets exfiltration using ptrace.
The file descriptor based memory has several advantages over the
"traditional" mm interfaces, such as mlock(), mprotect(), madvise(). File
descriptor approach allows explicit and controlled sharing of the memory
areas, it allows to seal the operations. Besides, file descriptor based
memory paves the way for VMMs to remove the secret memory range from the
userspace hipervisor process, for instance QEMU. Andy Lutomirski says:
"Getting fd-backed memory into a guest will take some possibly major
work in the kernel, but getting vma-backed memory into a guest without
mapping it in the host user address space seems much, much worse."
memfd_secret() is made a dedicated system call rather than an extension to
memfd_create() because it's purpose is to allow the user to create more
secure memory mappings rather than to simply allow file based access to
the memory. Nowadays a new system call cost is negligible while it is way
simpler for userspace to deal with a clear-cut system calls than with a
multiplexer or an overloaded syscall. Moreover, the initial
implementation of memfd_secret() is completely distinct from
memfd_create() so there is no much sense in overloading memfd_create() to
begin with. If there will be a need for code sharing between these
implementation it can be easily achieved without a need to adjust user
visible APIs.
The secret memory remains accessible in the process context using uaccess
primitives, but it is not exposed to the kernel otherwise; secret memory
areas are removed from the direct map and functions in the
follow_page()/get_user_page() family will refuse to return a page that
belongs to the secret memory area.
Once there will be a use case that will require exposing secretmem to the
kernel it will be an opt-in request in the system call flags so that user
would have to decide what data can be exposed to the kernel.
Removing of the pages from the direct map may cause its fragmentation on
architectures that use large pages to map the physical memory which
affects the system performance. However, the original Kconfig text for
CONFIG_DIRECT_GBPAGES said that gigabyte pages in the direct map "... can
improve the kernel's performance a tiny bit ..." (commit 00d1c5e05736
("x86: add gbpages switches")) and the recent report [1] showed that "...
although 1G mappings are a good default choice, there is no compelling
evidence that it must be the only choice". Hence, it is sufficient to
have secretmem disabled by default with the ability of a system
administrator to enable it at boot time.
Pages in the secretmem regions are unevictable and unmovable to avoid
accidental exposure of the sensitive data via swap or during page
migration.
Since the secretmem mappings are locked in memory they cannot exceed
RLIMIT_MEMLOCK. Since these mappings are already locked independently
from mlock(), an attempt to mlock()/munlock() secretmem range would fail
and mlockall()/munlockall() will ignore secretmem mappings.
However, unlike mlock()ed memory, secretmem currently behaves more like
long-term GUP: secretmem mappings are unmovable mappings directly consumed
by user space. With default limits, there is no excessive use of
secretmem and it poses no real problem in combination with
ZONE_MOVABLE/CMA, but in the future this should be addressed to allow
balanced use of large amounts of secretmem along with ZONE_MOVABLE/CMA.
A page that was a part of the secret memory area is cleared when it is
freed to ensure the data is not exposed to the next user of that page.
The following example demonstrates creation of a secret mapping (error
handling is omitted):
fd = memfd_secret(0);
ftruncate(fd, MAP_SIZE);
ptr = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);
[1] https://lore.kernel.org/linux-mm/213b4567-46ce-f116-9cdf-bbd0c884eb3c@linux.intel.com/
[akpm@linux-foundation.org: suppress Kconfig whine]
Link: https://lkml.kernel.org/r/20210518072034.31572-5-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Elena Reshetova <elena.reshetova@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tycho Andersen <tycho@tycho.ws>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-08 01:08:03 +00:00
|
|
|
#include <linux/secretmem.h>
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2017-02-02 18:15:33 +00:00
|
|
|
#include <linux/sched/signal.h>
|
2014-10-09 22:29:14 +00:00
|
|
|
#include <linux/rwsem.h>
|
2014-11-05 16:27:40 +00:00
|
|
|
#include <linux/hugetlb.h>
|
2019-03-05 23:47:44 +00:00
|
|
|
#include <linux/migrate.h>
|
|
|
|
#include <linux/mm_inline.h>
|
|
|
|
#include <linux/sched/mm.h>
|
2015-09-04 22:47:55 +00:00
|
|
|
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
#include <asm/mmu_context.h>
|
2015-09-04 22:47:55 +00:00
|
|
|
#include <asm/tlbflush.h>
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2014-06-04 23:08:10 +00:00
|
|
|
#include "internal.h"
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
struct follow_page_context {
|
|
|
|
struct dev_pagemap *pgmap;
|
|
|
|
unsigned int page_mask;
|
|
|
|
};
|
|
|
|
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
static inline void sanity_check_pinned_pages(struct page **pages,
|
|
|
|
unsigned long npages)
|
|
|
|
{
|
|
|
|
if (!IS_ENABLED(CONFIG_DEBUG_VM))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We only pin anonymous pages if they are exclusive. Once pinned, we
|
|
|
|
* can no longer turn them possibly shared and PageAnonExclusive() will
|
|
|
|
* stick around until the page is freed.
|
|
|
|
*
|
|
|
|
* We'd like to verify that our pinned anonymous pages are still mapped
|
|
|
|
* exclusively. The issue with anon THP is that we don't know how
|
|
|
|
* they are/were mapped when pinning them. However, for anon
|
|
|
|
* THP we can assume that either the given page (PTE-mapped THP) or
|
|
|
|
* the head page (PMD-mapped THP) should be PageAnonExclusive(). If
|
|
|
|
* neither is the case, there is certainly something wrong.
|
|
|
|
*/
|
|
|
|
for (; npages; npages--, pages++) {
|
|
|
|
struct page *page = *pages;
|
|
|
|
struct folio *folio = page_folio(page);
|
|
|
|
|
|
|
|
if (!folio_test_anon(folio))
|
|
|
|
continue;
|
|
|
|
if (!folio_test_large(folio) || folio_test_hugetlb(folio))
|
|
|
|
VM_BUG_ON_PAGE(!PageAnonExclusive(&folio->page), page);
|
|
|
|
else
|
|
|
|
/* Either a PTE-mapped or a PMD-mapped THP. */
|
|
|
|
VM_BUG_ON_PAGE(!PageAnonExclusive(&folio->page) &&
|
|
|
|
!PageAnonExclusive(page), page);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Revert "mm/gup: remove try_get_page(), call try_get_compound_head() directly"
This reverts commit 9857a17f206ff374aea78bccfb687f145368be2e.
That commit was completely broken, and I should have caught on to it
earlier. But happily, the kernel test robot noticed the breakage fairly
quickly.
The breakage is because "try_get_page()" is about avoiding the page
reference count overflow case, but is otherwise the exact same as a
plain "get_page()".
In contrast, "try_get_compound_head()" is an entirely different beast,
and uses __page_cache_add_speculative() because it's not just about the
page reference count, but also about possibly racing with the underlying
page going away.
So all the commentary about how
"try_get_page() has fallen a little behind in terms of maintenance,
try_get_compound_head() handles speculative page references more
thoroughly"
was just completely wrong: yes, try_get_compound_head() handles
speculative page references, but the point is that try_get_page() does
not, and must not.
So there's no lack of maintainance - there are fundamentally different
semantics.
A speculative page reference would be entirely wrong in "get_page()",
and it's entirely wrong in "try_get_page()". It's not about
speculation, it's purely about "uhhuh, you can't get this page because
you've tried to increment the reference count too much already".
The reason the kernel test robot noticed this bug was that it hit the
VM_BUG_ON() in __page_cache_add_speculative(), which is all about
verifying that the context of any speculative page access is correct.
But since that isn't what try_get_page() is all about, the VM_BUG_ON()
tests things that are not correct to test for try_get_page().
Reported-by: kernel test robot <oliver.sang@intel.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-09-07 18:03:45 +00:00
|
|
|
/*
|
2022-02-04 15:27:40 +00:00
|
|
|
* Return the folio with ref appropriately incremented,
|
Revert "mm/gup: remove try_get_page(), call try_get_compound_head() directly"
This reverts commit 9857a17f206ff374aea78bccfb687f145368be2e.
That commit was completely broken, and I should have caught on to it
earlier. But happily, the kernel test robot noticed the breakage fairly
quickly.
The breakage is because "try_get_page()" is about avoiding the page
reference count overflow case, but is otherwise the exact same as a
plain "get_page()".
In contrast, "try_get_compound_head()" is an entirely different beast,
and uses __page_cache_add_speculative() because it's not just about the
page reference count, but also about possibly racing with the underlying
page going away.
So all the commentary about how
"try_get_page() has fallen a little behind in terms of maintenance,
try_get_compound_head() handles speculative page references more
thoroughly"
was just completely wrong: yes, try_get_compound_head() handles
speculative page references, but the point is that try_get_page() does
not, and must not.
So there's no lack of maintainance - there are fundamentally different
semantics.
A speculative page reference would be entirely wrong in "get_page()",
and it's entirely wrong in "try_get_page()". It's not about
speculation, it's purely about "uhhuh, you can't get this page because
you've tried to increment the reference count too much already".
The reason the kernel test robot noticed this bug was that it hit the
VM_BUG_ON() in __page_cache_add_speculative(), which is all about
verifying that the context of any speculative page access is correct.
But since that isn't what try_get_page() is all about, the VM_BUG_ON()
tests things that are not correct to test for try_get_page().
Reported-by: kernel test robot <oliver.sang@intel.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-09-07 18:03:45 +00:00
|
|
|
* or NULL if that failed.
|
2020-01-31 06:12:21 +00:00
|
|
|
*/
|
2022-02-04 15:27:40 +00:00
|
|
|
static inline struct folio *try_get_folio(struct page *page, int refs)
|
2020-01-31 06:12:21 +00:00
|
|
|
{
|
2022-02-04 15:27:40 +00:00
|
|
|
struct folio *folio;
|
2020-01-31 06:12:21 +00:00
|
|
|
|
2022-01-07 19:04:55 +00:00
|
|
|
retry:
|
2022-02-04 15:27:40 +00:00
|
|
|
folio = page_folio(page);
|
|
|
|
if (WARN_ON_ONCE(folio_ref_count(folio) < 0))
|
2020-01-31 06:12:21 +00:00
|
|
|
return NULL;
|
2022-02-04 15:27:40 +00:00
|
|
|
if (unlikely(!folio_ref_try_add_rcu(folio, refs)))
|
2020-01-31 06:12:21 +00:00
|
|
|
return NULL;
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
|
|
|
|
/*
|
2022-02-04 15:27:40 +00:00
|
|
|
* At this point we have a stable reference to the folio; but it
|
|
|
|
* could be that between calling page_folio() and the refcount
|
|
|
|
* increment, the folio was split, in which case we'd end up
|
|
|
|
* holding a reference on a folio that has nothing to do with the page
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
* we were given anymore.
|
2022-02-04 15:27:40 +00:00
|
|
|
* So now that the folio is stable, recheck that the page still
|
|
|
|
* belongs to this folio.
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
*/
|
2022-02-04 15:27:40 +00:00
|
|
|
if (unlikely(page_folio(page) != folio)) {
|
2022-07-05 12:35:32 +00:00
|
|
|
if (!put_devmap_managed_page_refs(&folio->page, refs))
|
|
|
|
folio_put_refs(folio, refs);
|
2022-01-07 19:04:55 +00:00
|
|
|
goto retry;
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
}
|
|
|
|
|
2022-02-04 15:27:40 +00:00
|
|
|
return folio;
|
2020-01-31 06:12:21 +00:00
|
|
|
}
|
|
|
|
|
2021-09-02 21:53:48 +00:00
|
|
|
/**
|
2022-02-04 15:27:40 +00:00
|
|
|
* try_grab_folio() - Attempt to get or pin a folio.
|
2021-09-02 21:53:48 +00:00
|
|
|
* @page: pointer to page to be grabbed
|
2022-02-04 15:27:40 +00:00
|
|
|
* @refs: the value to (effectively) add to the folio's refcount
|
2021-09-02 21:53:48 +00:00
|
|
|
* @flags: gup flags: these are the FOLL_* flag values.
|
|
|
|
*
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* "grab" names in this file mean, "look at flags to decide whether to use
|
2022-02-04 15:27:40 +00:00
|
|
|
* FOLL_PIN or FOLL_GET behavior, when incrementing the folio's refcount.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*
|
|
|
|
* Either FOLL_PIN or FOLL_GET (or neither) must be set, but not both at the
|
|
|
|
* same time. (That's true throughout the get_user_pages*() and
|
|
|
|
* pin_user_pages*() APIs.) Cases:
|
|
|
|
*
|
2022-02-04 15:27:40 +00:00
|
|
|
* FOLL_GET: folio's refcount will be incremented by @refs.
|
2021-09-02 21:53:48 +00:00
|
|
|
*
|
2022-02-04 15:27:40 +00:00
|
|
|
* FOLL_PIN on large folios: folio's refcount will be incremented by
|
2023-01-11 14:28:47 +00:00
|
|
|
* @refs, and its pincount will be incremented by @refs.
|
2021-09-02 21:53:48 +00:00
|
|
|
*
|
2022-02-04 15:27:40 +00:00
|
|
|
* FOLL_PIN on single-page folios: folio's refcount will be incremented by
|
2022-01-06 21:46:43 +00:00
|
|
|
* @refs * GUP_PIN_COUNTING_BIAS.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*
|
2022-02-04 15:27:40 +00:00
|
|
|
* Return: The folio containing @page (with refcount appropriately
|
|
|
|
* incremented) for success, or NULL upon failure. If neither FOLL_GET
|
|
|
|
* nor FOLL_PIN was set, that's considered failure, and furthermore,
|
|
|
|
* a likely bug in the caller, so a warning is also emitted.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*/
|
2022-02-04 15:27:40 +00:00
|
|
|
struct folio *try_grab_folio(struct page *page, int refs, unsigned int flags)
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
{
|
2022-10-21 17:41:09 +00:00
|
|
|
if (unlikely(!(flags & FOLL_PCI_P2PDMA) && is_pci_p2pdma_page(page)))
|
|
|
|
return NULL;
|
|
|
|
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
if (flags & FOLL_GET)
|
2022-02-04 15:27:40 +00:00
|
|
|
return try_get_folio(page, refs);
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
else if (flags & FOLL_PIN) {
|
2022-02-04 15:27:40 +00:00
|
|
|
struct folio *folio;
|
|
|
|
|
2020-04-02 04:06:04 +00:00
|
|
|
/*
|
2021-05-05 01:39:08 +00:00
|
|
|
* Can't do FOLL_LONGTERM + FOLL_PIN gup fast path if not in a
|
|
|
|
* right zone, so fail and let the caller fall back to the slow
|
|
|
|
* path.
|
2020-04-02 04:06:04 +00:00
|
|
|
*/
|
2021-05-05 01:39:08 +00:00
|
|
|
if (unlikely((flags & FOLL_LONGTERM) &&
|
mm: rename is_pinnable_page() to is_longterm_pinnable_page()
Patch series "Add MEMORY_DEVICE_COHERENT for coherent device memory
mapping", v9.
This patch series introduces MEMORY_DEVICE_COHERENT, a type of memory
owned by a device that can be mapped into CPU page tables like
MEMORY_DEVICE_GENERIC and can also be migrated like MEMORY_DEVICE_PRIVATE.
This patch series is mostly self-contained except for a few places where
it needs to update other subsystems to handle the new memory type.
System stability and performance are not affected according to our ongoing
testing, including xfstests.
How it works: The system BIOS advertises the GPU device memory (aka VRAM)
as SPM (special purpose memory) in the UEFI system address map.
The amdgpu driver registers the memory with devmap as
MEMORY_DEVICE_COHERENT using devm_memremap_pages. The initial user for
this hardware page migration capability is the Frontier supercomputer
project. This functionality is not AMD-specific. We expect other GPU
vendors to find this functionality useful, and possibly other hardware
types in the future.
Our test nodes in the lab are similar to the Frontier configuration, with
.5 TB of system memory plus 256 GB of device memory split across 4 GPUs,
all in a single coherent address space. Page migration is expected to
improve application efficiency significantly. We will report empirical
results as they become available.
Coherent device type pages at gup are now migrated back to system memory
if they are being pinned long-term (FOLL_LONGTERM). The reason is, that
long-term pinning would interfere with the device memory manager owning
the device-coherent pages (e.g. evictions in TTM). These series
incorporate Alistair Popple patches to do this migration from
pin_user_pages() calls. hmm_gup_test has been added to hmm-test to test
different get user pages calls.
This series includes handling of device-managed anonymous pages returned
by vm_normal_pages. Although they behave like normal pages for purposes
of mapping in CPU page tables and for COW, they do not support LRU lists,
NUMA migration or THP.
We also introduced a FOLL_LRU flag that adds the same behaviour to
follow_page and related APIs, to allow callers to specify that they expect
to put pages on an LRU list.
This patch (of 14):
is_pinnable_page() and folio_is_pinnable() are renamed to
is_longterm_pinnable_page() and folio_is_longterm_pinnable() respectively.
These functions are used in the FOLL_LONGTERM flag context.
Link: https://lkml.kernel.org/r/20220715150521.18165-1-alex.sierra@amd.com
Link: https://lkml.kernel.org/r/20220715150521.18165-2-alex.sierra@amd.com
Signed-off-by: Alex Sierra <alex.sierra@amd.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-07-15 15:05:08 +00:00
|
|
|
!is_longterm_pinnable_page(page)))
|
2020-04-02 04:06:04 +00:00
|
|
|
return NULL;
|
|
|
|
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
/*
|
|
|
|
* CAUTION: Don't use compound_head() on the page before this
|
|
|
|
* point, the result won't be stable.
|
|
|
|
*/
|
2022-02-04 15:27:40 +00:00
|
|
|
folio = try_get_folio(page, refs);
|
|
|
|
if (!folio)
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
return NULL;
|
|
|
|
|
mm/gup: page->hpage_pinned_refcount: exact pin counts for huge pages
For huge pages (and in fact, any compound page), the GUP_PIN_COUNTING_BIAS
scheme tends to overflow too easily, each tail page increments the head
page->_refcount by GUP_PIN_COUNTING_BIAS (1024). That limits the number
of huge pages that can be pinned.
This patch removes that limitation, by using an exact form of pin counting
for compound pages of order > 1. The "order > 1" is required because this
approach uses the 3rd struct page in the compound page, and order 1
compound pages only have two pages, so that won't work there.
A new struct page field, hpage_pinned_refcount, has been added, replacing
a padding field in the union (so no new space is used).
This enhancement also has a useful side effect: huge pages and compound
pages (of order > 1) do not suffer from the "potential false positives"
problem that is discussed in the page_dma_pinned() comment block. That is
because these compound pages have extra space for tracking things, so they
get exact pin counts instead of overloading page->_refcount.
Documentation/core-api/pin_user_pages.rst is updated accordingly.
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-8-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:33 +00:00
|
|
|
/*
|
2022-02-04 15:27:40 +00:00
|
|
|
* When pinning a large folio, use an exact count to track it.
|
mm/gup: page->hpage_pinned_refcount: exact pin counts for huge pages
For huge pages (and in fact, any compound page), the GUP_PIN_COUNTING_BIAS
scheme tends to overflow too easily, each tail page increments the head
page->_refcount by GUP_PIN_COUNTING_BIAS (1024). That limits the number
of huge pages that can be pinned.
This patch removes that limitation, by using an exact form of pin counting
for compound pages of order > 1. The "order > 1" is required because this
approach uses the 3rd struct page in the compound page, and order 1
compound pages only have two pages, so that won't work there.
A new struct page field, hpage_pinned_refcount, has been added, replacing
a padding field in the union (so no new space is used).
This enhancement also has a useful side effect: huge pages and compound
pages (of order > 1) do not suffer from the "potential false positives"
problem that is discussed in the page_dma_pinned() comment block. That is
because these compound pages have extra space for tracking things, so they
get exact pin counts instead of overloading page->_refcount.
Documentation/core-api/pin_user_pages.rst is updated accordingly.
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-8-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:33 +00:00
|
|
|
*
|
2022-02-04 15:27:40 +00:00
|
|
|
* However, be sure to *also* increment the normal folio
|
|
|
|
* refcount field at least once, so that the folio really
|
2022-01-07 19:15:11 +00:00
|
|
|
* is pinned. That's why the refcount from the earlier
|
2022-02-04 15:27:40 +00:00
|
|
|
* try_get_folio() is left intact.
|
mm/gup: page->hpage_pinned_refcount: exact pin counts for huge pages
For huge pages (and in fact, any compound page), the GUP_PIN_COUNTING_BIAS
scheme tends to overflow too easily, each tail page increments the head
page->_refcount by GUP_PIN_COUNTING_BIAS (1024). That limits the number
of huge pages that can be pinned.
This patch removes that limitation, by using an exact form of pin counting
for compound pages of order > 1. The "order > 1" is required because this
approach uses the 3rd struct page in the compound page, and order 1
compound pages only have two pages, so that won't work there.
A new struct page field, hpage_pinned_refcount, has been added, replacing
a padding field in the union (so no new space is used).
This enhancement also has a useful side effect: huge pages and compound
pages (of order > 1) do not suffer from the "potential false positives"
problem that is discussed in the page_dma_pinned() comment block. That is
because these compound pages have extra space for tracking things, so they
get exact pin counts instead of overloading page->_refcount.
Documentation/core-api/pin_user_pages.rst is updated accordingly.
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-8-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:33 +00:00
|
|
|
*/
|
2022-02-04 15:27:40 +00:00
|
|
|
if (folio_test_large(folio))
|
2023-01-11 14:28:47 +00:00
|
|
|
atomic_add(refs, &folio->_pincount);
|
mm/gup: fix try_grab_compound_head() race with split_huge_page()
try_grab_compound_head() is used to grab a reference to a page from
get_user_pages_fast(), which is only protected against concurrent freeing
of page tables (via local_irq_save()), but not against concurrent TLB
flushes, freeing of data pages, or splitting of compound pages.
Because no reference is held to the page when try_grab_compound_head() is
called, the page may have been freed and reallocated by the time its
refcount has been elevated; therefore, once we're holding a stable
reference to the page, the caller re-checks whether the PTE still points
to the same page (with the same access rights).
The problem is that try_grab_compound_head() has to grab a reference on
the head page; but between the time we look up what the head page is and
the time we actually grab a reference on the head page, the compound page
may have been split up (either explicitly through split_huge_page() or by
freeing the compound page to the buddy allocator and then allocating its
individual order-0 pages). If that happens, get_user_pages_fast() may end
up returning the right page but lifting the refcount on a now-unrelated
page, leading to use-after-free of pages.
To fix it: Re-check whether the pages still belong together after lifting
the refcount on the head page. Move anything else that checks
compound_head(page) below the refcount increment.
This can't actually happen on bare-metal x86 (because there, disabling
IRQs locks out remote TLB flushes), but it can happen on virtualized x86
(e.g. under KVM) and probably also on arm64. The race window is pretty
narrow, and constantly allocating and shattering hugepages isn't exactly
fast; for now I've only managed to reproduce this in an x86 KVM guest with
an artificially widened timing window (by adding a loop that repeatedly
calls `inl(0x3f8 + 5)` in `try_get_compound_head()` to force VM exits, so
that PV TLB flushes are used instead of IPIs).
As requested on the list, also replace the existing VM_BUG_ON_PAGE() with
a warning and bailout. Since the existing code only performed the BUG_ON
check on DEBUG_VM kernels, ensure that the new code also only performs the
check under that configuration - I don't want to mix two logically
separate changes together too much. The macro VM_WARN_ON_ONCE_PAGE()
doesn't return a value on !DEBUG_VM, so wrap the whole check in an #ifdef
block. An alternative would be to change the VM_WARN_ON_ONCE_PAGE()
definition for !DEBUG_VM such that it always returns false, but since that
would differ from the behavior of the normal WARN macros, it might be too
confusing for readers.
Link: https://lkml.kernel.org/r/20210615012014.1100672-1-jannh@google.com
Fixes: 7aef4172c795 ("mm: handle PTE-mapped tail pages in gerneric fast gup implementaiton")
Signed-off-by: Jann Horn <jannh@google.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jan Kara <jack@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-06-29 02:33:23 +00:00
|
|
|
else
|
2022-02-04 15:27:40 +00:00
|
|
|
folio_ref_add(folio,
|
|
|
|
refs * (GUP_PIN_COUNTING_BIAS - 1));
|
mm: fix PageAnonExclusive clearing racing with concurrent RCU GUP-fast
commit 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with
PG_anon_exclusive") made sure that when PageAnonExclusive() has to be
cleared during temporary unmapping of a page, that the PTE is
cleared/invalidated and that the TLB is flushed.
What we want to achieve in all cases is that we cannot end up with a pin on
an anonymous page that may be shared, because such pins would be
unreliable and could result in memory corruptions when the mapped page
and the pin go out of sync due to a write fault.
That TLB flush handling was inspired by an outdated comment in
mm/ksm.c:write_protect_page(), which similarly required the TLB flush in
the past to synchronize with GUP-fast. However, ever since general RCU GUP
fast was introduced in commit 2667f50e8b81 ("mm: introduce a general RCU
get_user_pages_fast()"), a TLB flush is no longer sufficient to handle
concurrent GUP-fast in all cases -- it only handles traditional IPI-based
GUP-fast correctly.
Peter Xu (thankfully) questioned whether that TLB flush is really
required. On architectures that send an IPI broadcast on TLB flush,
it works as expected. To synchronize with RCU GUP-fast properly, we're
conceptually fine, however, we have to enforce a certain memory order and
are missing memory barriers.
Let's document that, avoid the TLB flush where possible and use proper
explicit memory barriers where required. We shouldn't really care about the
additional memory barriers here, as we're not on extremely hot paths --
and we're getting rid of some TLB flushes.
We use a smp_mb() pair for handling concurrent pinning and a
smp_rmb()/smp_wmb() pair for handling the corner case of only temporary
PTE changes but permanent PageAnonExclusive changes.
One extreme example, whereby GUP-fast takes a R/O pin and KSM wants to
convert an exclusive anonymous page to a KSM page, and that page is already
mapped write-protected (-> no PTE change) would be:
Thread 0 (KSM) Thread 1 (GUP-fast)
(B1) Read the PTE
# (B2) skipped without FOLL_WRITE
(A1) Clear PTE
smp_mb()
(A2) Check pinned
(B3) Pin the mapped page
smp_mb()
(A3) Clear PageAnonExclusive
smp_wmb()
(A4) Restore PTE
(B4) Check if the PTE changed
smp_rmb()
(B5) Check PageAnonExclusive
Thread 1 will properly detect that PageAnonExclusive was cleared and
back off.
Note that we don't need a memory barrier between checking if the page is
pinned and clearing PageAnonExclusive, because stores are not
speculated.
The possible issues due to reordering are of theoretical nature so far
and attempts to reproduce the race failed.
Especially the "no PTE change" case isn't the common case, because we'd
need an exclusive anonymous page that's mapped R/O and the PTE is clean
in KSM code -- and using KSM with page pinning isn't extremely common.
Further, the clear+TLB flush we used for now implies a memory barrier.
So the problematic missing part should be the missing memory barrier
after pinning but before checking if the PTE changed.
Link: https://lkml.kernel.org/r/20220901083559.67446-1-david@redhat.com
Fixes: 6c287605fd56 ("mm: remember exclusively mapped anonymous pages with PG_anon_exclusive")
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Andrea Parri <parri.andrea@gmail.com>
Cc: Will Deacon <will@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: "Paul E. McKenney" <paulmck@kernel.org>
Cc: Christoph von Recklinghausen <crecklin@redhat.com>
Cc: Don Dutile <ddutile@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-09-01 08:35:59 +00:00
|
|
|
/*
|
|
|
|
* Adjust the pincount before re-checking the PTE for changes.
|
|
|
|
* This is essentially a smp_mb() and is paired with a memory
|
|
|
|
* barrier in page_try_share_anon_rmap().
|
|
|
|
*/
|
|
|
|
smp_mb__after_atomic();
|
|
|
|
|
2022-02-04 15:27:40 +00:00
|
|
|
node_stat_mod_folio(folio, NR_FOLL_PIN_ACQUIRED, refs);
|
mm/gup: page->hpage_pinned_refcount: exact pin counts for huge pages
For huge pages (and in fact, any compound page), the GUP_PIN_COUNTING_BIAS
scheme tends to overflow too easily, each tail page increments the head
page->_refcount by GUP_PIN_COUNTING_BIAS (1024). That limits the number
of huge pages that can be pinned.
This patch removes that limitation, by using an exact form of pin counting
for compound pages of order > 1. The "order > 1" is required because this
approach uses the 3rd struct page in the compound page, and order 1
compound pages only have two pages, so that won't work there.
A new struct page field, hpage_pinned_refcount, has been added, replacing
a padding field in the union (so no new space is used).
This enhancement also has a useful side effect: huge pages and compound
pages (of order > 1) do not suffer from the "potential false positives"
problem that is discussed in the page_dma_pinned() comment block. That is
because these compound pages have extra space for tracking things, so they
get exact pin counts instead of overloading page->_refcount.
Documentation/core-api/pin_user_pages.rst is updated accordingly.
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-8-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:33 +00:00
|
|
|
|
2022-02-04 15:27:40 +00:00
|
|
|
return folio;
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
WARN_ON_ONCE(1);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2021-12-10 20:39:04 +00:00
|
|
|
static void gup_put_folio(struct folio *folio, int refs, unsigned int flags)
|
2020-12-15 03:05:51 +00:00
|
|
|
{
|
|
|
|
if (flags & FOLL_PIN) {
|
2021-12-10 20:39:04 +00:00
|
|
|
node_stat_mod_folio(folio, NR_FOLL_PIN_RELEASED, refs);
|
|
|
|
if (folio_test_large(folio))
|
2023-01-11 14:28:47 +00:00
|
|
|
atomic_sub(refs, &folio->_pincount);
|
2020-12-15 03:05:51 +00:00
|
|
|
else
|
|
|
|
refs *= GUP_PIN_COUNTING_BIAS;
|
|
|
|
}
|
|
|
|
|
2022-07-05 12:35:32 +00:00
|
|
|
if (!put_devmap_managed_page_refs(&folio->page, refs))
|
|
|
|
folio_put_refs(folio, refs);
|
2020-12-15 03:05:51 +00:00
|
|
|
}
|
|
|
|
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
/**
|
|
|
|
* try_grab_page() - elevate a page's refcount by a flag-dependent amount
|
2022-02-04 15:32:01 +00:00
|
|
|
* @page: pointer to page to be grabbed
|
|
|
|
* @flags: gup flags: these are the FOLL_* flag values.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*
|
|
|
|
* This might not do anything at all, depending on the flags argument.
|
|
|
|
*
|
|
|
|
* "grab" names in this file mean, "look at flags to decide whether to use
|
|
|
|
* FOLL_PIN or FOLL_GET behavior, when incrementing the page's refcount.
|
|
|
|
*
|
|
|
|
* Either FOLL_PIN or FOLL_GET (or neither) may be set, but not both at the same
|
2022-02-04 15:27:40 +00:00
|
|
|
* time. Cases: please see the try_grab_folio() documentation, with
|
2021-09-02 21:53:48 +00:00
|
|
|
* "refs=1".
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*
|
2022-10-21 17:41:08 +00:00
|
|
|
* Return: 0 for success, or if no action was required (if neither FOLL_PIN
|
|
|
|
* nor FOLL_GET was set, nothing is done). A negative error code for failure:
|
|
|
|
*
|
|
|
|
* -ENOMEM FOLL_GET or FOLL_PIN was set, but the page could not
|
|
|
|
* be grabbed.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
*/
|
2022-10-21 17:41:08 +00:00
|
|
|
int __must_check try_grab_page(struct page *page, unsigned int flags)
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
{
|
2022-02-04 15:32:01 +00:00
|
|
|
struct folio *folio = page_folio(page);
|
|
|
|
|
|
|
|
if (WARN_ON_ONCE(folio_ref_count(folio) <= 0))
|
2022-10-21 17:41:08 +00:00
|
|
|
return -ENOMEM;
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
|
2022-10-21 17:41:09 +00:00
|
|
|
if (unlikely(!(flags & FOLL_PCI_P2PDMA) && is_pci_p2pdma_page(page)))
|
|
|
|
return -EREMOTEIO;
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
|
2022-02-02 03:23:17 +00:00
|
|
|
if (flags & FOLL_GET)
|
2022-02-04 15:32:01 +00:00
|
|
|
folio_ref_inc(folio);
|
2022-02-02 03:23:17 +00:00
|
|
|
else if (flags & FOLL_PIN) {
|
|
|
|
/*
|
2022-02-04 15:32:01 +00:00
|
|
|
* Similar to try_grab_folio(): be sure to *also*
|
2022-01-07 19:15:11 +00:00
|
|
|
* increment the normal page refcount field at least once,
|
|
|
|
* so that the page really is pinned.
|
2022-02-02 03:23:17 +00:00
|
|
|
*/
|
2022-02-04 15:32:01 +00:00
|
|
|
if (folio_test_large(folio)) {
|
|
|
|
folio_ref_add(folio, 1);
|
2023-01-11 14:28:47 +00:00
|
|
|
atomic_add(1, &folio->_pincount);
|
2022-02-04 14:24:26 +00:00
|
|
|
} else {
|
2022-02-04 15:32:01 +00:00
|
|
|
folio_ref_add(folio, GUP_PIN_COUNTING_BIAS);
|
2022-02-04 14:24:26 +00:00
|
|
|
}
|
2022-02-02 03:23:17 +00:00
|
|
|
|
2022-02-04 15:32:01 +00:00
|
|
|
node_stat_mod_folio(folio, NR_FOLL_PIN_ACQUIRED, 1);
|
2022-02-02 03:23:17 +00:00
|
|
|
}
|
|
|
|
|
2022-10-21 17:41:08 +00:00
|
|
|
return 0;
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* unpin_user_page() - release a dma-pinned page
|
|
|
|
* @page: pointer to page to be released
|
|
|
|
*
|
|
|
|
* Pages that were pinned via pin_user_pages*() must be released via either
|
|
|
|
* unpin_user_page(), or one of the unpin_user_pages*() routines. This is so
|
|
|
|
* that such pages can be separately tracked and uniquely handled. In
|
|
|
|
* particular, interactions with RDMA and filesystems need special handling.
|
|
|
|
*/
|
|
|
|
void unpin_user_page(struct page *page)
|
|
|
|
{
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
sanity_check_pinned_pages(&page, 1);
|
2021-12-10 20:39:04 +00:00
|
|
|
gup_put_folio(page_folio(page), 1, FOLL_PIN);
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(unpin_user_page);
|
|
|
|
|
2021-12-23 15:20:12 +00:00
|
|
|
static inline struct folio *gup_folio_range_next(struct page *start,
|
2022-01-09 21:05:11 +00:00
|
|
|
unsigned long npages, unsigned long i, unsigned int *ntails)
|
2021-04-30 05:55:50 +00:00
|
|
|
{
|
2021-12-23 15:20:12 +00:00
|
|
|
struct page *next = nth_page(start, i);
|
|
|
|
struct folio *folio = page_folio(next);
|
2021-04-30 05:55:50 +00:00
|
|
|
unsigned int nr = 1;
|
|
|
|
|
2021-12-23 15:20:12 +00:00
|
|
|
if (folio_test_large(folio))
|
2022-01-07 18:45:25 +00:00
|
|
|
nr = min_t(unsigned int, npages - i,
|
2021-12-23 15:20:12 +00:00
|
|
|
folio_nr_pages(folio) - folio_page_idx(folio, next));
|
2021-04-30 05:55:50 +00:00
|
|
|
|
|
|
|
*ntails = nr;
|
2021-12-23 15:20:12 +00:00
|
|
|
return folio;
|
2021-04-30 05:55:50 +00:00
|
|
|
}
|
|
|
|
|
2021-12-23 04:43:16 +00:00
|
|
|
static inline struct folio *gup_folio_next(struct page **list,
|
2022-01-10 02:03:47 +00:00
|
|
|
unsigned long npages, unsigned long i, unsigned int *ntails)
|
2021-04-30 05:55:44 +00:00
|
|
|
{
|
2021-12-23 04:43:16 +00:00
|
|
|
struct folio *folio = page_folio(list[i]);
|
2021-04-30 05:55:44 +00:00
|
|
|
unsigned int nr;
|
|
|
|
|
|
|
|
for (nr = i + 1; nr < npages; nr++) {
|
2021-12-23 04:43:16 +00:00
|
|
|
if (page_folio(list[nr]) != folio)
|
2021-04-30 05:55:44 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
*ntails = nr - i;
|
2021-12-23 04:43:16 +00:00
|
|
|
return folio;
|
2021-04-30 05:55:44 +00:00
|
|
|
}
|
|
|
|
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
/**
|
2020-01-31 06:13:35 +00:00
|
|
|
* unpin_user_pages_dirty_lock() - release and optionally dirty gup-pinned pages
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
* @pages: array of pages to be maybe marked dirty, and definitely released.
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
* @npages: number of pages in the @pages array.
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
* @make_dirty: whether to mark the pages dirty
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*
|
|
|
|
* "gup-pinned page" refers to a page that has had one of the get_user_pages()
|
|
|
|
* variants called on that page.
|
|
|
|
*
|
|
|
|
* For each page in the @pages array, make that page (or its head page, if a
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
* compound page) dirty, if @make_dirty is true, and if the page was previously
|
2020-01-31 06:13:35 +00:00
|
|
|
* listed as clean. In any case, releases all pages using unpin_user_page(),
|
|
|
|
* possibly via unpin_user_pages(), for the non-dirty case.
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*
|
2020-01-31 06:13:35 +00:00
|
|
|
* Please see the unpin_user_page() documentation for details.
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
* set_page_dirty_lock() is used internally. If instead, set_page_dirty() is
|
|
|
|
* required, then the caller should a) verify that this is really correct,
|
|
|
|
* because _lock() is usually required, and b) hand code it:
|
2020-01-31 06:13:35 +00:00
|
|
|
* set_page_dirty_lock(), unpin_user_page().
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*
|
|
|
|
*/
|
2020-01-31 06:13:35 +00:00
|
|
|
void unpin_user_pages_dirty_lock(struct page **pages, unsigned long npages,
|
|
|
|
bool make_dirty)
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
{
|
2021-12-23 04:43:16 +00:00
|
|
|
unsigned long i;
|
|
|
|
struct folio *folio;
|
|
|
|
unsigned int nr;
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
|
|
|
|
if (!make_dirty) {
|
2020-01-31 06:13:35 +00:00
|
|
|
unpin_user_pages(pages, npages);
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
sanity_check_pinned_pages(pages, npages);
|
2021-12-23 04:43:16 +00:00
|
|
|
for (i = 0; i < npages; i += nr) {
|
|
|
|
folio = gup_folio_next(pages, npages, i, &nr);
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
/*
|
|
|
|
* Checking PageDirty at this point may race with
|
|
|
|
* clear_page_dirty_for_io(), but that's OK. Two key
|
|
|
|
* cases:
|
|
|
|
*
|
|
|
|
* 1) This code sees the page as already dirty, so it
|
|
|
|
* skips the call to set_page_dirty(). That could happen
|
|
|
|
* because clear_page_dirty_for_io() called
|
|
|
|
* page_mkclean(), followed by set_page_dirty().
|
|
|
|
* However, now the page is going to get written back,
|
|
|
|
* which meets the original intention of setting it
|
|
|
|
* dirty, so all is well: clear_page_dirty_for_io() goes
|
|
|
|
* on to call TestClearPageDirty(), and write the page
|
|
|
|
* back.
|
|
|
|
*
|
|
|
|
* 2) This code sees the page as clean, so it calls
|
|
|
|
* set_page_dirty(). The page stays dirty, despite being
|
|
|
|
* written back, so it gets written back again in the
|
|
|
|
* next writeback cycle. This is harmless.
|
|
|
|
*/
|
2021-12-23 04:43:16 +00:00
|
|
|
if (!folio_test_dirty(folio)) {
|
|
|
|
folio_lock(folio);
|
|
|
|
folio_mark_dirty(folio);
|
|
|
|
folio_unlock(folio);
|
|
|
|
}
|
|
|
|
gup_put_folio(folio, nr, FOLL_PIN);
|
mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
[11~From: John Hubbard <jhubbard@nvidia.com>
Subject: mm/gup: add make_dirty arg to put_user_pages_dirty_lock()
Patch series "mm/gup: add make_dirty arg to put_user_pages_dirty_lock()",
v3.
There are about 50+ patches in my tree [2], and I'll be sending out the
remaining ones in a few more groups:
* The block/bio related changes (Jerome mostly wrote those, but I've had
to move stuff around extensively, and add a little code)
* mm/ changes
* other subsystem patches
* an RFC that shows the current state of the tracking patch set. That
can only be applied after all call sites are converted, but it's good to
get an early look at it.
This is part a tree-wide conversion, as described in fc1d8e7cca2d ("mm:
introduce put_user_page*(), placeholder versions").
This patch (of 3):
Provide more capable variation of put_user_pages_dirty_lock(), and delete
put_user_pages_dirty(). This is based on the following:
1. Lots of call sites become simpler if a bool is passed into
put_user_page*(), instead of making the call site choose which
put_user_page*() variant to call.
2. Christoph Hellwig's observation that set_page_dirty_lock() is
usually correct, and set_page_dirty() is usually a bug, or at least
questionable, within a put_user_page*() calling chain.
This leads to the following API choices:
* put_user_pages_dirty_lock(page, npages, make_dirty)
* There is no put_user_pages_dirty(). You have to
hand code that, in the rare case that it's
required.
[jhubbard@nvidia.com: remove unused variable in siw_free_plist()]
Link: http://lkml.kernel.org/r/20190729074306.10368-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20190724044537.10458-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-09-23 22:35:04 +00:00
|
|
|
}
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
}
|
2020-01-31 06:13:35 +00:00
|
|
|
EXPORT_SYMBOL(unpin_user_pages_dirty_lock);
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
|
2021-04-30 05:55:50 +00:00
|
|
|
/**
|
|
|
|
* unpin_user_page_range_dirty_lock() - release and optionally dirty
|
|
|
|
* gup-pinned page range
|
|
|
|
*
|
|
|
|
* @page: the starting page of a range maybe marked dirty, and definitely released.
|
|
|
|
* @npages: number of consecutive pages to release.
|
|
|
|
* @make_dirty: whether to mark the pages dirty
|
|
|
|
*
|
|
|
|
* "gup-pinned page range" refers to a range of pages that has had one of the
|
|
|
|
* pin_user_pages() variants called on that page.
|
|
|
|
*
|
|
|
|
* For the page ranges defined by [page .. page+npages], make that range (or
|
|
|
|
* its head pages, if a compound page) dirty, if @make_dirty is true, and if the
|
|
|
|
* page range was previously listed as clean.
|
|
|
|
*
|
|
|
|
* set_page_dirty_lock() is used internally. If instead, set_page_dirty() is
|
|
|
|
* required, then the caller should a) verify that this is really correct,
|
|
|
|
* because _lock() is usually required, and b) hand code it:
|
|
|
|
* set_page_dirty_lock(), unpin_user_page().
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
void unpin_user_page_range_dirty_lock(struct page *page, unsigned long npages,
|
|
|
|
bool make_dirty)
|
|
|
|
{
|
2021-12-23 15:20:12 +00:00
|
|
|
unsigned long i;
|
|
|
|
struct folio *folio;
|
|
|
|
unsigned int nr;
|
|
|
|
|
|
|
|
for (i = 0; i < npages; i += nr) {
|
|
|
|
folio = gup_folio_range_next(page, npages, i, &nr);
|
|
|
|
if (make_dirty && !folio_test_dirty(folio)) {
|
|
|
|
folio_lock(folio);
|
|
|
|
folio_mark_dirty(folio);
|
|
|
|
folio_unlock(folio);
|
|
|
|
}
|
|
|
|
gup_put_folio(folio, nr, FOLL_PIN);
|
2021-04-30 05:55:50 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(unpin_user_page_range_dirty_lock);
|
|
|
|
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
static void unpin_user_pages_lockless(struct page **pages, unsigned long npages)
|
|
|
|
{
|
|
|
|
unsigned long i;
|
|
|
|
struct folio *folio;
|
|
|
|
unsigned int nr;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Don't perform any sanity checks because we might have raced with
|
|
|
|
* fork() and some anonymous pages might now actually be shared --
|
|
|
|
* which is why we're unpinning after all.
|
|
|
|
*/
|
|
|
|
for (i = 0; i < npages; i += nr) {
|
|
|
|
folio = gup_folio_next(pages, npages, i, &nr);
|
|
|
|
gup_put_folio(folio, nr, FOLL_PIN);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
/**
|
2020-01-31 06:13:35 +00:00
|
|
|
* unpin_user_pages() - release an array of gup-pinned pages.
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
* @pages: array of pages to be marked dirty and released.
|
|
|
|
* @npages: number of pages in the @pages array.
|
|
|
|
*
|
2020-01-31 06:13:35 +00:00
|
|
|
* For each page in the @pages array, release the page using unpin_user_page().
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*
|
2020-01-31 06:13:35 +00:00
|
|
|
* Please see the unpin_user_page() documentation for details.
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
*/
|
2020-01-31 06:13:35 +00:00
|
|
|
void unpin_user_pages(struct page **pages, unsigned long npages)
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
{
|
2021-12-23 04:43:16 +00:00
|
|
|
unsigned long i;
|
|
|
|
struct folio *folio;
|
|
|
|
unsigned int nr;
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
|
2020-10-13 23:52:01 +00:00
|
|
|
/*
|
|
|
|
* If this WARN_ON() fires, then the system *might* be leaking pages (by
|
|
|
|
* leaving them pinned), but probably not. More likely, gup/pup returned
|
|
|
|
* a hard -ERRNO error to the caller, who erroneously passed it here.
|
|
|
|
*/
|
|
|
|
if (WARN_ON(IS_ERR_VALUE(npages)))
|
|
|
|
return;
|
2021-04-30 05:55:47 +00:00
|
|
|
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
sanity_check_pinned_pages(pages, npages);
|
2021-12-23 04:43:16 +00:00
|
|
|
for (i = 0; i < npages; i += nr) {
|
|
|
|
folio = gup_folio_next(pages, npages, i, &nr);
|
|
|
|
gup_put_folio(folio, nr, FOLL_PIN);
|
2022-01-09 01:23:46 +00:00
|
|
|
}
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
}
|
2020-01-31 06:13:35 +00:00
|
|
|
EXPORT_SYMBOL(unpin_user_pages);
|
mm: introduce put_user_page*(), placeholder versions
A discussion of the overall problem is below.
As mentioned in patch 0001, the steps are to fix the problem are:
1) Provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
Overview
========
Some kernel components (file systems, device drivers) need to access
memory that is specified via process virtual address. For a long time,
the API to achieve that was get_user_pages ("GUP") and its variations.
However, GUP has critical limitations that have been overlooked; in
particular, GUP does not interact correctly with filesystems in all
situations. That means that file-backed memory + GUP is a recipe for
potential problems, some of which have already occurred in the field.
GUP was first introduced for Direct IO (O_DIRECT), allowing filesystem
code to get the struct page behind a virtual address and to let storage
hardware perform a direct copy to or from that page. This is a
short-lived access pattern, and as such, the window for a concurrent
writeback of GUP'd page was small enough that there were not (we think)
any reported problems. Also, userspace was expected to understand and
accept that Direct IO was not synchronized with memory-mapped access to
that data, nor with any process address space changes such as munmap(),
mremap(), etc.
Over the years, more GUP uses have appeared (virtualization, device
drivers, RDMA) that can keep the pages they get via GUP for a long period
of time (seconds, minutes, hours, days, ...). This long-term pinning
makes an underlying design problem more obvious.
In fact, there are a number of key problems inherent to GUP:
Interactions with file systems
==============================
File systems expect to be able to write back data, both to reclaim pages,
and for data integrity. Allowing other hardware (NICs, GPUs, etc) to gain
write access to the file memory pages means that such hardware can dirty
the pages, without the filesystem being aware. This can, in some cases
(depending on filesystem, filesystem options, block device, block device
options, and other variables), lead to data corruption, and also to kernel
bugs of the form:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on every
write access to a clean file backed page, not just the first one.
How long the GUP reference lasts is irrelevant, if the page is clean
and you need to dirty it, you must call ->page_mkwrite before it is
marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
Long term GUP
=============
Long term GUP is an issue when FOLL_WRITE is specified to GUP (so, a
writeable mapping is created), and the pages are file-backed. That can
lead to filesystem corruption. What happens is that when a file-backed
page is being written back, it is first mapped read-only in all of the CPU
page tables; the file system then assumes that nobody can write to the
page, and that the page content is therefore stable. Unfortunately, the
GUP callers generally do not monitor changes to the CPU pages tables; they
instead assume that the following pattern is safe (it's not):
get_user_pages()
Hardware can keep a reference to those pages for a very long time,
and write to it at any time. Because "hardware" here means "devices
that are not a CPU", this activity occurs without any interaction with
the kernel's file system code.
for each page
set_page_dirty
put_page()
In fact, the GUP documentation even recommends that pattern.
Anyway, the file system assumes that the page is stable (nothing is
writing to the page), and that is a problem: stable page content is
necessary for many filesystem actions during writeback, such as checksum,
encryption, RAID striping, etc. Furthermore, filesystem features like COW
(copy on write) or snapshot also rely on being able to use a new page for
as memory for that memory range inside the file.
Corruption during write back is clearly possible here. To solve that, one
idea is to identify pages that have active GUP, so that we can use a
bounce page to write stable data to the filesystem. The filesystem would
work on the bounce page, while any of the active GUP might write to the
original page. This would avoid the stable page violation problem, but
note that it is only part of the overall solution, because other problems
remain.
Other filesystem features that need to replace the page with a new one can
be inhibited for pages that are GUP-pinned. This will, however, alter and
limit some of those filesystem features. The only fix for that would be
to require GUP users to monitor and respond to CPU page table updates.
Subsystems such as ODP and HMM do this, for example. This aspect of the
problem is still under discussion.
Direct IO
=========
Direct IO can cause corruption, if userspace does Direct-IO that writes to
a range of virtual addresses that are mmap'd to a file. The pages written
to are file-backed pages that can be under write back, while the Direct IO
is taking place. Here, Direct IO races with a write back: it calls GUP
before page_mkclean() has replaced the CPU pte with a read-only entry.
The race window is pretty small, which is probably why years have gone by
before we noticed this problem: Direct IO is generally very quick, and
tends to finish up before the filesystem gets around to do anything with
the page contents. However, it's still a real problem. The solution is
to never let GUP return pages that are under write back, but instead,
force GUP to take a write fault on those pages. That way, GUP will
properly synchronize with the active write back. This does not change the
required GUP behavior, it just avoids that race.
Details
=======
Introduces put_user_page(), which simply calls put_page(). This provides
a way to update all get_user_pages*() callers, so that they call
put_user_page(), instead of put_page().
Also introduces put_user_pages(), and a few dirty/locked variations, as a
replacement for release_pages(), and also as a replacement for open-coded
loops that release multiple pages. These may be used for subsequent
performance improvements, via batching of pages to be released.
This is the first step of fixing a problem (also described in [1] and [2])
with interactions between get_user_pages ("gup") and filesystems.
Problem description: let's start with a bug report. Below, is what
happens sometimes, under memory pressure, when a driver pins some pages
via gup, and then marks those pages dirty, and releases them. Note that
the gup documentation actually recommends that pattern. The problem is
that the filesystem may do a writeback while the pages were gup-pinned,
and then the filesystem believes that the pages are clean. So, when the
driver later marks the pages as dirty, that conflicts with the
filesystem's page tracking and results in a BUG(), like this one that I
experienced:
kernel BUG at /build/linux-fQ94TU/linux-4.4.0/fs/ext4/inode.c:1899!
backtrace:
ext4_writepage
__writepage
write_cache_pages
ext4_writepages
do_writepages
__writeback_single_inode
writeback_sb_inodes
__writeback_inodes_wb
wb_writeback
wb_workfn
process_one_work
worker_thread
kthread
ret_from_fork
...which is due to the file system asserting that there are still buffer
heads attached:
({ \
BUG_ON(!PagePrivate(page)); \
((struct buffer_head *)page_private(page)); \
})
Dave Chinner's description of this is very clear:
"The fundamental issue is that ->page_mkwrite must be called on
every write access to a clean file backed page, not just the first
one. How long the GUP reference lasts is irrelevant, if the page is
clean and you need to dirty it, you must call ->page_mkwrite before it
is marked writeable and dirtied. Every. Time."
This is just one symptom of the larger design problem: real filesystems
that actually write to a backing device, do not actually support
get_user_pages() being called on their pages, and letting hardware write
directly to those pages--even though that pattern has been going on since
about 2005 or so.
The steps are to fix it are:
1) (This patch): provide put_user_page*() routines, intended to be used
for releasing pages that were pinned via get_user_pages*().
2) Convert all of the call sites for get_user_pages*(), to
invoke put_user_page*(), instead of put_page(). This involves dozens of
call sites, and will take some time.
3) After (2) is complete, use get_user_pages*() and put_user_page*() to
implement tracking of these pages. This tracking will be separate from
the existing struct page refcounting.
4) Use the tracking and identification of these pages, to implement
special handling (especially in writeback paths) when the pages are
backed by a filesystem.
[1] https://lwn.net/Articles/774411/ : "DMA and get_user_pages()"
[2] https://lwn.net/Articles/753027/ : "The Trouble with get_user_pages()"
Link: http://lkml.kernel.org/r/20190327023632.13307-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> [docs]
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Tested-by: Ira Weiny <ira.weiny@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:19:08 +00:00
|
|
|
|
2021-06-29 02:36:40 +00:00
|
|
|
/*
|
|
|
|
* Set the MMF_HAS_PINNED if not set yet; after set it'll be there for the mm's
|
|
|
|
* lifecycle. Avoid setting the bit unless necessary, or it might cause write
|
|
|
|
* cache bouncing on large SMP machines for concurrent pinned gups.
|
|
|
|
*/
|
|
|
|
static inline void mm_set_has_pinned_flag(unsigned long *mm_flags)
|
|
|
|
{
|
|
|
|
if (!test_bit(MMF_HAS_PINNED, mm_flags))
|
|
|
|
set_bit(MMF_HAS_PINNED, mm_flags);
|
|
|
|
}
|
|
|
|
|
2019-07-12 03:57:21 +00:00
|
|
|
#ifdef CONFIG_MMU
|
2014-06-04 23:08:11 +00:00
|
|
|
static struct page *no_page_table(struct vm_area_struct *vma,
|
|
|
|
unsigned int flags)
|
2014-06-04 23:08:10 +00:00
|
|
|
{
|
2014-06-04 23:08:11 +00:00
|
|
|
/*
|
|
|
|
* When core dumping an enormous anonymous area that nobody
|
|
|
|
* has touched so far, we don't want to allocate unnecessary pages or
|
|
|
|
* page tables. Return error instead of NULL to skip handle_mm_fault,
|
|
|
|
* then get_dump_page() will return NULL to leave a hole in the dump.
|
|
|
|
* But we can only make this optimization where a hole would surely
|
|
|
|
* be zero-filled if handle_mm_fault() actually did handle it.
|
|
|
|
*/
|
2020-04-07 03:03:55 +00:00
|
|
|
if ((flags & FOLL_DUMP) &&
|
|
|
|
(vma_is_anonymous(vma) || !vma->vm_ops->fault))
|
2014-06-04 23:08:11 +00:00
|
|
|
return ERR_PTR(-EFAULT);
|
|
|
|
return NULL;
|
|
|
|
}
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2015-09-04 22:47:55 +00:00
|
|
|
static int follow_pfn_pte(struct vm_area_struct *vma, unsigned long address,
|
|
|
|
pte_t *pte, unsigned int flags)
|
|
|
|
{
|
|
|
|
if (flags & FOLL_TOUCH) {
|
|
|
|
pte_t entry = *pte;
|
|
|
|
|
|
|
|
if (flags & FOLL_WRITE)
|
|
|
|
entry = pte_mkdirty(entry);
|
|
|
|
entry = pte_mkyoung(entry);
|
|
|
|
|
|
|
|
if (!pte_same(*pte, entry)) {
|
|
|
|
set_pte_at(vma->vm_mm, address, pte, entry);
|
|
|
|
update_mmu_cache(vma, address, pte);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Proper page table entry exists, but no corresponding struct page */
|
|
|
|
return -EEXIST;
|
|
|
|
}
|
|
|
|
|
mm/gup: fix FOLL_FORCE COW security issue and remove FOLL_COW
Ever since the Dirty COW (CVE-2016-5195) security issue happened, we know
that FOLL_FORCE can be possibly dangerous, especially if there are races
that can be exploited by user space.
Right now, it would be sufficient to have some code that sets a PTE of a
R/O-mapped shared page dirty, in order for it to erroneously become
writable by FOLL_FORCE. The implications of setting a write-protected PTE
dirty might not be immediately obvious to everyone.
And in fact ever since commit 9ae0f87d009c ("mm/shmem: unconditionally set
pte dirty in mfill_atomic_install_pte"), we can use UFFDIO_CONTINUE to map
a shmem page R/O while marking the pte dirty. This can be used by
unprivileged user space to modify tmpfs/shmem file content even if the
user does not have write permissions to the file, and to bypass memfd
write sealing -- Dirty COW restricted to tmpfs/shmem (CVE-2022-2590).
To fix such security issues for good, the insight is that we really only
need that fancy retry logic (FOLL_COW) for COW mappings that are not
writable (!VM_WRITE). And in a COW mapping, we really only broke COW if
we have an exclusive anonymous page mapped. If we have something else
mapped, or the mapped anonymous page might be shared (!PageAnonExclusive),
we have to trigger a write fault to break COW. If we don't find an
exclusive anonymous page when we retry, we have to trigger COW breaking
once again because something intervened.
Let's move away from this mandatory-retry + dirty handling and rely on our
PageAnonExclusive() flag for making a similar decision, to use the same
COW logic as in other kernel parts here as well. In case we stumble over
a PTE in a COW mapping that does not map an exclusive anonymous page, COW
was not properly broken and we have to trigger a fake write-fault to break
COW.
Just like we do in can_change_pte_writable() added via commit 64fe24a3e05e
("mm/mprotect: try avoiding write faults for exclusive anonymous pages
when changing protection") and commit 76aefad628aa ("mm/mprotect: fix
soft-dirty check in can_change_pte_writable()"), take care of softdirty
and uffd-wp manually.
For example, a write() via /proc/self/mem to a uffd-wp-protected range has
to fail instead of silently granting write access and bypassing the
userspace fault handler. Note that FOLL_FORCE is not only used for debug
access, but also triggered by applications without debug intentions, for
example, when pinning pages via RDMA.
This fixes CVE-2022-2590. Note that only x86_64 and aarch64 are
affected, because only those support CONFIG_HAVE_ARCH_USERFAULTFD_MINOR.
Fortunately, FOLL_COW is no longer required to handle FOLL_FORCE. So
let's just get rid of it.
Thanks to Nadav Amit for pointing out that the pte_dirty() check in
FOLL_FORCE code is problematic and might be exploitable.
Note 1: We don't check for the PTE being dirty because it doesn't matter
for making a "was COWed" decision anymore, and whoever modifies the
page has to set the page dirty either way.
Note 2: Kernels before extended uffd-wp support and before
PageAnonExclusive (< 5.19) can simply revert the problematic
commit instead and be safe regarding UFFDIO_CONTINUE. A backport to
v5.19 requires minor adjustments due to lack of
vma_soft_dirty_enabled().
Link: https://lkml.kernel.org/r/20220809205640.70916-1-david@redhat.com
Fixes: 9ae0f87d009c ("mm/shmem: unconditionally set pte dirty in mfill_atomic_install_pte")
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: David Laight <David.Laight@ACULAB.COM>
Cc: <stable@vger.kernel.org> [5.16]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-08-09 20:56:40 +00:00
|
|
|
/* FOLL_FORCE can write to even unwritable PTEs in COW mappings. */
|
|
|
|
static inline bool can_follow_write_pte(pte_t pte, struct page *page,
|
|
|
|
struct vm_area_struct *vma,
|
|
|
|
unsigned int flags)
|
2016-10-13 20:07:36 +00:00
|
|
|
{
|
mm/gup: fix FOLL_FORCE COW security issue and remove FOLL_COW
Ever since the Dirty COW (CVE-2016-5195) security issue happened, we know
that FOLL_FORCE can be possibly dangerous, especially if there are races
that can be exploited by user space.
Right now, it would be sufficient to have some code that sets a PTE of a
R/O-mapped shared page dirty, in order for it to erroneously become
writable by FOLL_FORCE. The implications of setting a write-protected PTE
dirty might not be immediately obvious to everyone.
And in fact ever since commit 9ae0f87d009c ("mm/shmem: unconditionally set
pte dirty in mfill_atomic_install_pte"), we can use UFFDIO_CONTINUE to map
a shmem page R/O while marking the pte dirty. This can be used by
unprivileged user space to modify tmpfs/shmem file content even if the
user does not have write permissions to the file, and to bypass memfd
write sealing -- Dirty COW restricted to tmpfs/shmem (CVE-2022-2590).
To fix such security issues for good, the insight is that we really only
need that fancy retry logic (FOLL_COW) for COW mappings that are not
writable (!VM_WRITE). And in a COW mapping, we really only broke COW if
we have an exclusive anonymous page mapped. If we have something else
mapped, or the mapped anonymous page might be shared (!PageAnonExclusive),
we have to trigger a write fault to break COW. If we don't find an
exclusive anonymous page when we retry, we have to trigger COW breaking
once again because something intervened.
Let's move away from this mandatory-retry + dirty handling and rely on our
PageAnonExclusive() flag for making a similar decision, to use the same
COW logic as in other kernel parts here as well. In case we stumble over
a PTE in a COW mapping that does not map an exclusive anonymous page, COW
was not properly broken and we have to trigger a fake write-fault to break
COW.
Just like we do in can_change_pte_writable() added via commit 64fe24a3e05e
("mm/mprotect: try avoiding write faults for exclusive anonymous pages
when changing protection") and commit 76aefad628aa ("mm/mprotect: fix
soft-dirty check in can_change_pte_writable()"), take care of softdirty
and uffd-wp manually.
For example, a write() via /proc/self/mem to a uffd-wp-protected range has
to fail instead of silently granting write access and bypassing the
userspace fault handler. Note that FOLL_FORCE is not only used for debug
access, but also triggered by applications without debug intentions, for
example, when pinning pages via RDMA.
This fixes CVE-2022-2590. Note that only x86_64 and aarch64 are
affected, because only those support CONFIG_HAVE_ARCH_USERFAULTFD_MINOR.
Fortunately, FOLL_COW is no longer required to handle FOLL_FORCE. So
let's just get rid of it.
Thanks to Nadav Amit for pointing out that the pte_dirty() check in
FOLL_FORCE code is problematic and might be exploitable.
Note 1: We don't check for the PTE being dirty because it doesn't matter
for making a "was COWed" decision anymore, and whoever modifies the
page has to set the page dirty either way.
Note 2: Kernels before extended uffd-wp support and before
PageAnonExclusive (< 5.19) can simply revert the problematic
commit instead and be safe regarding UFFDIO_CONTINUE. A backport to
v5.19 requires minor adjustments due to lack of
vma_soft_dirty_enabled().
Link: https://lkml.kernel.org/r/20220809205640.70916-1-david@redhat.com
Fixes: 9ae0f87d009c ("mm/shmem: unconditionally set pte dirty in mfill_atomic_install_pte")
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: David Laight <David.Laight@ACULAB.COM>
Cc: <stable@vger.kernel.org> [5.16]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-08-09 20:56:40 +00:00
|
|
|
/* If the pte is writable, we can write to the page. */
|
|
|
|
if (pte_write(pte))
|
|
|
|
return true;
|
|
|
|
|
|
|
|
/* Maybe FOLL_FORCE is set to override it? */
|
|
|
|
if (!(flags & FOLL_FORCE))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* But FOLL_FORCE has no effect on shared mappings */
|
|
|
|
if (vma->vm_flags & (VM_MAYSHARE | VM_SHARED))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* ... or read-only private ones */
|
|
|
|
if (!(vma->vm_flags & VM_MAYWRITE))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* ... or already writable ones that just need to take a write fault */
|
|
|
|
if (vma->vm_flags & VM_WRITE)
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* See can_change_pte_writable(): we broke COW and could map the page
|
|
|
|
* writable if we have an exclusive anonymous page ...
|
|
|
|
*/
|
|
|
|
if (!page || !PageAnon(page) || !PageAnonExclusive(page))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* ... and a write-fault isn't required for other reasons. */
|
|
|
|
if (vma_soft_dirty_enabled(vma) && !pte_soft_dirty(pte))
|
|
|
|
return false;
|
|
|
|
return !userfaultfd_pte_wp(vma, pte);
|
2016-10-13 20:07:36 +00:00
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:11 +00:00
|
|
|
static struct page *follow_page_pte(struct vm_area_struct *vma,
|
2018-10-26 22:10:28 +00:00
|
|
|
unsigned long address, pmd_t *pmd, unsigned int flags,
|
|
|
|
struct dev_pagemap **pgmap)
|
2014-06-04 23:08:11 +00:00
|
|
|
{
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
struct page *page;
|
|
|
|
spinlock_t *ptl;
|
|
|
|
pte_t *ptep, pte;
|
2020-04-02 04:05:56 +00:00
|
|
|
int ret;
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2020-01-31 06:12:54 +00:00
|
|
|
/* FOLL_GET and FOLL_PIN are mutually exclusive. */
|
|
|
|
if (WARN_ON_ONCE((flags & (FOLL_PIN | FOLL_GET)) ==
|
|
|
|
(FOLL_PIN | FOLL_GET)))
|
|
|
|
return ERR_PTR(-EINVAL);
|
2014-06-04 23:08:10 +00:00
|
|
|
if (unlikely(pmd_bad(*pmd)))
|
2014-06-04 23:08:11 +00:00
|
|
|
return no_page_table(vma, flags);
|
2014-06-04 23:08:10 +00:00
|
|
|
|
|
|
|
ptep = pte_offset_map_lock(mm, pmd, address, &ptl);
|
|
|
|
pte = *ptep;
|
2022-10-21 10:11:41 +00:00
|
|
|
if (!pte_present(pte))
|
|
|
|
goto no_page;
|
2022-08-25 16:46:57 +00:00
|
|
|
if (pte_protnone(pte) && !gup_can_follow_protnone(flags))
|
2014-06-04 23:08:10 +00:00
|
|
|
goto no_page;
|
|
|
|
|
|
|
|
page = vm_normal_page(vma, address, pte);
|
mm/gup: fix FOLL_FORCE COW security issue and remove FOLL_COW
Ever since the Dirty COW (CVE-2016-5195) security issue happened, we know
that FOLL_FORCE can be possibly dangerous, especially if there are races
that can be exploited by user space.
Right now, it would be sufficient to have some code that sets a PTE of a
R/O-mapped shared page dirty, in order for it to erroneously become
writable by FOLL_FORCE. The implications of setting a write-protected PTE
dirty might not be immediately obvious to everyone.
And in fact ever since commit 9ae0f87d009c ("mm/shmem: unconditionally set
pte dirty in mfill_atomic_install_pte"), we can use UFFDIO_CONTINUE to map
a shmem page R/O while marking the pte dirty. This can be used by
unprivileged user space to modify tmpfs/shmem file content even if the
user does not have write permissions to the file, and to bypass memfd
write sealing -- Dirty COW restricted to tmpfs/shmem (CVE-2022-2590).
To fix such security issues for good, the insight is that we really only
need that fancy retry logic (FOLL_COW) for COW mappings that are not
writable (!VM_WRITE). And in a COW mapping, we really only broke COW if
we have an exclusive anonymous page mapped. If we have something else
mapped, or the mapped anonymous page might be shared (!PageAnonExclusive),
we have to trigger a write fault to break COW. If we don't find an
exclusive anonymous page when we retry, we have to trigger COW breaking
once again because something intervened.
Let's move away from this mandatory-retry + dirty handling and rely on our
PageAnonExclusive() flag for making a similar decision, to use the same
COW logic as in other kernel parts here as well. In case we stumble over
a PTE in a COW mapping that does not map an exclusive anonymous page, COW
was not properly broken and we have to trigger a fake write-fault to break
COW.
Just like we do in can_change_pte_writable() added via commit 64fe24a3e05e
("mm/mprotect: try avoiding write faults for exclusive anonymous pages
when changing protection") and commit 76aefad628aa ("mm/mprotect: fix
soft-dirty check in can_change_pte_writable()"), take care of softdirty
and uffd-wp manually.
For example, a write() via /proc/self/mem to a uffd-wp-protected range has
to fail instead of silently granting write access and bypassing the
userspace fault handler. Note that FOLL_FORCE is not only used for debug
access, but also triggered by applications without debug intentions, for
example, when pinning pages via RDMA.
This fixes CVE-2022-2590. Note that only x86_64 and aarch64 are
affected, because only those support CONFIG_HAVE_ARCH_USERFAULTFD_MINOR.
Fortunately, FOLL_COW is no longer required to handle FOLL_FORCE. So
let's just get rid of it.
Thanks to Nadav Amit for pointing out that the pte_dirty() check in
FOLL_FORCE code is problematic and might be exploitable.
Note 1: We don't check for the PTE being dirty because it doesn't matter
for making a "was COWed" decision anymore, and whoever modifies the
page has to set the page dirty either way.
Note 2: Kernels before extended uffd-wp support and before
PageAnonExclusive (< 5.19) can simply revert the problematic
commit instead and be safe regarding UFFDIO_CONTINUE. A backport to
v5.19 requires minor adjustments due to lack of
vma_soft_dirty_enabled().
Link: https://lkml.kernel.org/r/20220809205640.70916-1-david@redhat.com
Fixes: 9ae0f87d009c ("mm/shmem: unconditionally set pte dirty in mfill_atomic_install_pte")
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: David Laight <David.Laight@ACULAB.COM>
Cc: <stable@vger.kernel.org> [5.16]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-08-09 20:56:40 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* We only care about anon pages in can_follow_write_pte() and don't
|
|
|
|
* have to worry about pte_devmap() because they are never anon.
|
|
|
|
*/
|
|
|
|
if ((flags & FOLL_WRITE) &&
|
|
|
|
!can_follow_write_pte(pte, page, vma, flags)) {
|
|
|
|
page = NULL;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
if (!page && pte_devmap(pte) && (flags & (FOLL_GET | FOLL_PIN))) {
|
2016-01-16 00:56:55 +00:00
|
|
|
/*
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* Only return device mapping pages in the FOLL_GET or FOLL_PIN
|
|
|
|
* case since they are only valid while holding the pgmap
|
|
|
|
* reference.
|
2016-01-16 00:56:55 +00:00
|
|
|
*/
|
2018-10-26 22:10:28 +00:00
|
|
|
*pgmap = get_dev_pagemap(pte_pfn(pte), *pgmap);
|
|
|
|
if (*pgmap)
|
2016-01-16 00:56:55 +00:00
|
|
|
page = pte_page(pte);
|
|
|
|
else
|
|
|
|
goto no_page;
|
|
|
|
} else if (unlikely(!page)) {
|
2015-09-04 22:47:55 +00:00
|
|
|
if (flags & FOLL_DUMP) {
|
|
|
|
/* Avoid special (like zero) pages in core dumps */
|
|
|
|
page = ERR_PTR(-EFAULT);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (is_zero_pfn(pte_pfn(pte))) {
|
|
|
|
page = pte_page(pte);
|
|
|
|
} else {
|
|
|
|
ret = follow_pfn_pte(vma, address, ptep, flags);
|
|
|
|
page = ERR_PTR(ret);
|
|
|
|
goto out;
|
|
|
|
}
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
|
|
|
|
mm/gup: reliable R/O long-term pinning in COW mappings
We already support reliable R/O pinning of anonymous memory. However,
assume we end up pinning (R/O long-term) a pagecache page or the shared
zeropage inside a writable private ("COW") mapping. The next write access
will trigger a write-fault and replace the pinned page by an exclusive
anonymous page in the process page tables to break COW: the pinned page no
longer corresponds to the page mapped into the process' page table.
Now that FAULT_FLAG_UNSHARE can break COW on anything mapped into a
COW mapping, let's properly break COW first before R/O long-term
pinning something that's not an exclusive anon page inside a COW
mapping. FAULT_FLAG_UNSHARE will break COW and map an exclusive anon page
instead that can get pinned safely.
With this change, we can stop using FOLL_FORCE|FOLL_WRITE for reliable
R/O long-term pinning in COW mappings.
With this change, the new R/O long-term pinning tests for non-anonymous
memory succeed:
# [RUN] R/O longterm GUP pin ... with shared zeropage
ok 151 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd
ok 152 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with tmpfile
ok 153 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with huge zeropage
ok 154 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (2048 kB)
ok 155 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (1048576 kB)
ok 156 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with shared zeropage
ok 157 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd
ok 158 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with tmpfile
ok 159 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with huge zeropage
ok 160 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (2048 kB)
ok 161 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (1048576 kB)
ok 162 Longterm R/O pin is reliable
Note 1: We don't care about short-term R/O-pinning, because they have
snapshot semantics: they are not supposed to observe modifications that
happen after pinning.
As one example, assume we start direct I/O to read from a page and store
page content into a file: modifications to page content after starting
direct I/O are not guaranteed to end up in the file. So even if we'd pin
the shared zeropage, the end result would be as expected -- getting zeroes
stored to the file.
Note 2: For shared mappings we'll now always fallback to the slow path to
lookup the VMA when R/O long-term pining. While that's the necessary price
we have to pay right now, it's actually not that bad in practice: most
FOLL_LONGTERM users already specify FOLL_WRITE, for example, along with
FOLL_FORCE because they tried dealing with COW mappings correctly ...
Note 3: For users that use FOLL_LONGTERM right now without FOLL_WRITE,
such as VFIO, we'd now no longer pin the shared zeropage. Instead, we'd
populate exclusive anon pages that we can pin. There was a concern that
this could affect the memlock limit of existing setups.
For example, a VM running with VFIO could run into the memlock limit and
fail to run. However, we essentially had the same behavior already in
commit 17839856fd58 ("gup: document and work around "COW can break either
way" issue") which got merged into some enterprise distros, and there were
not any such complaints. So most probably, we're fine.
Link: https://lkml.kernel.org/r/20221116102659.70287-10-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-16 10:26:48 +00:00
|
|
|
if (!pte_write(pte) && gup_must_unshare(vma, flags, page)) {
|
2022-05-10 01:20:45 +00:00
|
|
|
page = ERR_PTR(-EMLINK);
|
|
|
|
goto out;
|
|
|
|
}
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
|
|
|
|
VM_BUG_ON_PAGE((flags & FOLL_PIN) && PageAnon(page) &&
|
|
|
|
!PageAnonExclusive(page), page);
|
|
|
|
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
/* try_grab_page() does nothing unless FOLL_GET or FOLL_PIN is set. */
|
2022-10-21 17:41:08 +00:00
|
|
|
ret = try_grab_page(page, flags);
|
|
|
|
if (unlikely(ret)) {
|
|
|
|
page = ERR_PTR(ret);
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
goto out;
|
2019-04-11 17:49:19 +00:00
|
|
|
}
|
2022-10-21 17:41:09 +00:00
|
|
|
|
2020-04-02 04:05:56 +00:00
|
|
|
/*
|
|
|
|
* We need to make the page accessible if and only if we are going
|
|
|
|
* to access its content (the FOLL_PIN case). Please see
|
|
|
|
* Documentation/core-api/pin_user_pages.rst for details.
|
|
|
|
*/
|
|
|
|
if (flags & FOLL_PIN) {
|
|
|
|
ret = arch_make_page_accessible(page);
|
|
|
|
if (ret) {
|
|
|
|
unpin_user_page(page);
|
|
|
|
page = ERR_PTR(ret);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
2014-06-04 23:08:10 +00:00
|
|
|
if (flags & FOLL_TOUCH) {
|
|
|
|
if ((flags & FOLL_WRITE) &&
|
|
|
|
!pte_dirty(pte) && !PageDirty(page))
|
|
|
|
set_page_dirty(page);
|
|
|
|
/*
|
|
|
|
* pte_mkyoung() would be more correct here, but atomic care
|
|
|
|
* is needed to avoid losing the dirty bit: it is easier to use
|
|
|
|
* mark_page_accessed().
|
|
|
|
*/
|
|
|
|
mark_page_accessed(page);
|
|
|
|
}
|
2015-09-04 22:47:55 +00:00
|
|
|
out:
|
2014-06-04 23:08:10 +00:00
|
|
|
pte_unmap_unlock(ptep, ptl);
|
|
|
|
return page;
|
|
|
|
no_page:
|
|
|
|
pte_unmap_unlock(ptep, ptl);
|
|
|
|
if (!pte_none(pte))
|
2014-06-04 23:08:11 +00:00
|
|
|
return NULL;
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
}
|
|
|
|
|
2017-07-06 22:38:44 +00:00
|
|
|
static struct page *follow_pmd_mask(struct vm_area_struct *vma,
|
|
|
|
unsigned long address, pud_t *pudp,
|
2018-10-26 22:10:28 +00:00
|
|
|
unsigned int flags,
|
|
|
|
struct follow_page_context *ctx)
|
2014-06-04 23:08:11 +00:00
|
|
|
{
|
2018-06-08 00:06:34 +00:00
|
|
|
pmd_t *pmd, pmdval;
|
2014-06-04 23:08:11 +00:00
|
|
|
spinlock_t *ptl;
|
|
|
|
struct page *page;
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
|
2017-07-06 22:38:44 +00:00
|
|
|
pmd = pmd_offset(pudp, address);
|
2018-06-08 00:06:34 +00:00
|
|
|
/*
|
|
|
|
* The READ_ONCE() will stabilize the pmdval in a register or
|
|
|
|
* on the stack so that it will stop changing under the code.
|
|
|
|
*/
|
|
|
|
pmdval = READ_ONCE(*pmd);
|
|
|
|
if (pmd_none(pmdval))
|
2014-06-04 23:08:11 +00:00
|
|
|
return no_page_table(vma, flags);
|
2022-10-21 10:11:41 +00:00
|
|
|
if (!pmd_present(pmdval))
|
mm/hugetlb: take page table lock in follow_huge_pmd()
We have a race condition between move_pages() and freeing hugepages, where
move_pages() calls follow_page(FOLL_GET) for hugepages internally and
tries to get its refcount without preventing concurrent freeing. This
race crashes the kernel, so this patch fixes it by moving FOLL_GET code
for hugepages into follow_huge_pmd() with taking the page table lock.
This patch intentionally removes page==NULL check after pte_page.
This is justified because pte_page() never returns NULL for any
architectures or configurations.
This patch changes the behavior of follow_huge_pmd() for tail pages and
then tail pages can be pinned/returned. So the caller must be changed to
properly handle the returned tail pages.
We could have a choice to add the similar locking to
follow_huge_(addr|pud) for consistency, but it's not necessary because
currently these functions don't support FOLL_GET flag, so let's leave it
for future development.
Here is the reproducer:
$ cat movepages.c
#include <stdio.h>
#include <stdlib.h>
#include <numaif.h>
#define ADDR_INPUT 0x700000000000UL
#define HPS 0x200000
#define PS 0x1000
int main(int argc, char *argv[]) {
int i;
int nr_hp = strtol(argv[1], NULL, 0);
int nr_p = nr_hp * HPS / PS;
int ret;
void **addrs;
int *status;
int *nodes;
pid_t pid;
pid = strtol(argv[2], NULL, 0);
addrs = malloc(sizeof(char *) * nr_p + 1);
status = malloc(sizeof(char *) * nr_p + 1);
nodes = malloc(sizeof(char *) * nr_p + 1);
while (1) {
for (i = 0; i < nr_p; i++) {
addrs[i] = (void *)ADDR_INPUT + i * PS;
nodes[i] = 1;
status[i] = 0;
}
ret = numa_move_pages(pid, nr_p, addrs, nodes, status,
MPOL_MF_MOVE_ALL);
if (ret == -1)
err("move_pages");
for (i = 0; i < nr_p; i++) {
addrs[i] = (void *)ADDR_INPUT + i * PS;
nodes[i] = 0;
status[i] = 0;
}
ret = numa_move_pages(pid, nr_p, addrs, nodes, status,
MPOL_MF_MOVE_ALL);
if (ret == -1)
err("move_pages");
}
return 0;
}
$ cat hugepage.c
#include <stdio.h>
#include <sys/mman.h>
#include <string.h>
#define ADDR_INPUT 0x700000000000UL
#define HPS 0x200000
int main(int argc, char *argv[]) {
int nr_hp = strtol(argv[1], NULL, 0);
char *p;
while (1) {
p = mmap((void *)ADDR_INPUT, nr_hp * HPS, PROT_READ | PROT_WRITE,
MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB, -1, 0);
if (p != (void *)ADDR_INPUT) {
perror("mmap");
break;
}
memset(p, 0, nr_hp * HPS);
munmap(p, nr_hp * HPS);
}
}
$ sysctl vm.nr_hugepages=40
$ ./hugepage 10 &
$ ./movepages 10 $(pgrep -f hugepage)
Fixes: e632a938d914 ("mm: migrate: add hugepage migration code to move_pages()")
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Reported-by: Hugh Dickins <hughd@google.com>
Cc: James Hogan <james.hogan@imgtec.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Luiz Capitulino <lcapitulino@redhat.com>
Cc: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Steve Capper <steve.capper@linaro.org>
Cc: <stable@vger.kernel.org> [3.12+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:25:22 +00:00
|
|
|
return no_page_table(vma, flags);
|
2018-06-08 00:06:34 +00:00
|
|
|
if (pmd_devmap(pmdval)) {
|
2016-01-16 00:56:55 +00:00
|
|
|
ptl = pmd_lock(mm, pmd);
|
2018-10-26 22:10:28 +00:00
|
|
|
page = follow_devmap_pmd(vma, address, pmd, flags, &ctx->pgmap);
|
2016-01-16 00:56:55 +00:00
|
|
|
spin_unlock(ptl);
|
|
|
|
if (page)
|
|
|
|
return page;
|
|
|
|
}
|
2018-06-08 00:06:34 +00:00
|
|
|
if (likely(!pmd_trans_huge(pmdval)))
|
2018-10-26 22:10:28 +00:00
|
|
|
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
|
2016-01-16 00:52:28 +00:00
|
|
|
|
2022-08-25 16:46:57 +00:00
|
|
|
if (pmd_protnone(pmdval) && !gup_can_follow_protnone(flags))
|
2017-02-24 22:59:53 +00:00
|
|
|
return no_page_table(vma, flags);
|
|
|
|
|
2016-01-16 00:52:28 +00:00
|
|
|
ptl = pmd_lock(mm, pmd);
|
mm: thp: check pmd migration entry in common path
When THP migration is being used, memory management code needs to handle
pmd migration entries properly. This patch uses !pmd_present() or
is_swap_pmd() (depending on whether pmd_none() needs separate code or
not) to check pmd migration entries at the places where a pmd entry is
present.
Since pmd-related code uses split_huge_page(), split_huge_pmd(),
pmd_trans_huge(), pmd_trans_unstable(), or
pmd_none_or_trans_huge_or_clear_bad(), this patch:
1. adds pmd migration entry split code in split_huge_pmd(),
2. takes care of pmd migration entries whenever pmd_trans_huge() is present,
3. makes pmd_none_or_trans_huge_or_clear_bad() pmd migration entry aware.
Since split_huge_page() uses split_huge_pmd() and pmd_trans_unstable()
is equivalent to pmd_none_or_trans_huge_or_clear_bad(), we do not change
them.
Until this commit, a pmd entry should be:
1. pointing to a pte page,
2. is_swap_pmd(),
3. pmd_trans_huge(),
4. pmd_devmap(), or
5. pmd_none().
Signed-off-by: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Nellans <dnellans@nvidia.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-08 23:11:01 +00:00
|
|
|
if (unlikely(!pmd_present(*pmd))) {
|
|
|
|
spin_unlock(ptl);
|
2022-10-21 10:11:41 +00:00
|
|
|
return no_page_table(vma, flags);
|
mm: thp: check pmd migration entry in common path
When THP migration is being used, memory management code needs to handle
pmd migration entries properly. This patch uses !pmd_present() or
is_swap_pmd() (depending on whether pmd_none() needs separate code or
not) to check pmd migration entries at the places where a pmd entry is
present.
Since pmd-related code uses split_huge_page(), split_huge_pmd(),
pmd_trans_huge(), pmd_trans_unstable(), or
pmd_none_or_trans_huge_or_clear_bad(), this patch:
1. adds pmd migration entry split code in split_huge_pmd(),
2. takes care of pmd migration entries whenever pmd_trans_huge() is present,
3. makes pmd_none_or_trans_huge_or_clear_bad() pmd migration entry aware.
Since split_huge_page() uses split_huge_pmd() and pmd_trans_unstable()
is equivalent to pmd_none_or_trans_huge_or_clear_bad(), we do not change
them.
Until this commit, a pmd entry should be:
1. pointing to a pte page,
2. is_swap_pmd(),
3. pmd_trans_huge(),
4. pmd_devmap(), or
5. pmd_none().
Signed-off-by: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Nellans <dnellans@nvidia.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-08 23:11:01 +00:00
|
|
|
}
|
2016-01-16 00:52:28 +00:00
|
|
|
if (unlikely(!pmd_trans_huge(*pmd))) {
|
|
|
|
spin_unlock(ptl);
|
2018-10-26 22:10:28 +00:00
|
|
|
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
|
2016-01-16 00:52:28 +00:00
|
|
|
}
|
2021-04-30 05:55:56 +00:00
|
|
|
if (flags & FOLL_SPLIT_PMD) {
|
2016-01-16 00:52:28 +00:00
|
|
|
int ret;
|
|
|
|
page = pmd_page(*pmd);
|
|
|
|
if (is_huge_zero_page(page)) {
|
|
|
|
spin_unlock(ptl);
|
|
|
|
ret = 0;
|
2016-01-16 00:52:42 +00:00
|
|
|
split_huge_pmd(vma, pmd, address);
|
2016-07-26 22:24:03 +00:00
|
|
|
if (pmd_trans_unstable(pmd))
|
|
|
|
ret = -EBUSY;
|
2021-04-30 05:55:56 +00:00
|
|
|
} else {
|
2019-09-23 22:38:25 +00:00
|
|
|
spin_unlock(ptl);
|
|
|
|
split_huge_pmd(vma, pmd, address);
|
|
|
|
ret = pte_alloc(mm, pmd) ? -ENOMEM : 0;
|
2016-01-16 00:52:28 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
return ret ? ERR_PTR(ret) :
|
2018-10-26 22:10:28 +00:00
|
|
|
follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
|
2014-06-04 23:08:11 +00:00
|
|
|
}
|
2016-01-16 00:52:28 +00:00
|
|
|
page = follow_trans_huge_pmd(vma, address, pmd, flags);
|
|
|
|
spin_unlock(ptl);
|
2018-10-26 22:10:28 +00:00
|
|
|
ctx->page_mask = HPAGE_PMD_NR - 1;
|
2016-01-16 00:52:28 +00:00
|
|
|
return page;
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
|
|
|
|
2017-07-06 22:38:44 +00:00
|
|
|
static struct page *follow_pud_mask(struct vm_area_struct *vma,
|
|
|
|
unsigned long address, p4d_t *p4dp,
|
2018-10-26 22:10:28 +00:00
|
|
|
unsigned int flags,
|
|
|
|
struct follow_page_context *ctx)
|
2017-07-06 22:38:44 +00:00
|
|
|
{
|
|
|
|
pud_t *pud;
|
|
|
|
spinlock_t *ptl;
|
|
|
|
struct page *page;
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
|
|
|
|
pud = pud_offset(p4dp, address);
|
|
|
|
if (pud_none(*pud))
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
if (pud_devmap(*pud)) {
|
|
|
|
ptl = pud_lock(mm, pud);
|
2018-10-26 22:10:28 +00:00
|
|
|
page = follow_devmap_pud(vma, address, pud, flags, &ctx->pgmap);
|
2017-07-06 22:38:44 +00:00
|
|
|
spin_unlock(ptl);
|
|
|
|
if (page)
|
|
|
|
return page;
|
|
|
|
}
|
|
|
|
if (unlikely(pud_bad(*pud)))
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
return follow_pmd_mask(vma, address, pud, flags, ctx);
|
2017-07-06 22:38:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct page *follow_p4d_mask(struct vm_area_struct *vma,
|
|
|
|
unsigned long address, pgd_t *pgdp,
|
2018-10-26 22:10:28 +00:00
|
|
|
unsigned int flags,
|
|
|
|
struct follow_page_context *ctx)
|
2017-07-06 22:38:44 +00:00
|
|
|
{
|
|
|
|
p4d_t *p4d;
|
|
|
|
|
|
|
|
p4d = p4d_offset(pgdp, address);
|
|
|
|
if (p4d_none(*p4d))
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
BUILD_BUG_ON(p4d_huge(*p4d));
|
|
|
|
if (unlikely(p4d_bad(*p4d)))
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
return follow_pud_mask(vma, address, p4d, flags, ctx);
|
2017-07-06 22:38:44 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* follow_page_mask - look up a page descriptor from a user-virtual address
|
|
|
|
* @vma: vm_area_struct mapping @address
|
|
|
|
* @address: virtual address to look up
|
|
|
|
* @flags: flags modifying lookup behaviour
|
2018-11-16 23:08:29 +00:00
|
|
|
* @ctx: contains dev_pagemap for %ZONE_DEVICE memory pinning and a
|
|
|
|
* pointer to output page_mask
|
2017-07-06 22:38:44 +00:00
|
|
|
*
|
|
|
|
* @flags can have FOLL_ flags set, defined in <linux/mm.h>
|
|
|
|
*
|
2018-11-16 23:08:29 +00:00
|
|
|
* When getting pages from ZONE_DEVICE memory, the @ctx->pgmap caches
|
|
|
|
* the device's dev_pagemap metadata to avoid repeating expensive lookups.
|
|
|
|
*
|
2022-05-10 01:20:45 +00:00
|
|
|
* When getting an anonymous page and the caller has to trigger unsharing
|
|
|
|
* of a shared anonymous page first, -EMLINK is returned. The caller should
|
|
|
|
* trigger a fault with FAULT_FLAG_UNSHARE set. Note that unsharing is only
|
|
|
|
* relevant with FOLL_PIN and !FOLL_WRITE.
|
|
|
|
*
|
2018-11-16 23:08:29 +00:00
|
|
|
* On output, the @ctx->page_mask is set according to the size of the page.
|
|
|
|
*
|
|
|
|
* Return: the mapped (struct page *), %NULL if no mapping exists, or
|
2017-07-06 22:38:44 +00:00
|
|
|
* an error pointer if there is a mapping to something not represented
|
|
|
|
* by a page descriptor (see also vm_normal_page()).
|
|
|
|
*/
|
2019-07-12 03:54:34 +00:00
|
|
|
static struct page *follow_page_mask(struct vm_area_struct *vma,
|
2017-07-06 22:38:44 +00:00
|
|
|
unsigned long address, unsigned int flags,
|
2018-10-26 22:10:28 +00:00
|
|
|
struct follow_page_context *ctx)
|
2017-07-06 22:38:44 +00:00
|
|
|
{
|
|
|
|
pgd_t *pgd;
|
|
|
|
struct page *page;
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
ctx->page_mask = 0;
|
2017-07-06 22:38:44 +00:00
|
|
|
|
2022-09-19 02:13:48 +00:00
|
|
|
/*
|
|
|
|
* Call hugetlb_follow_page_mask for hugetlb vmas as it will use
|
|
|
|
* special hugetlb page table walking code. This eliminates the
|
|
|
|
* need to check for hugetlb entries in the general walking code.
|
|
|
|
*
|
|
|
|
* hugetlb_follow_page_mask is only for follow_page() handling here.
|
|
|
|
* Ordinary GUP uses follow_hugetlb_page for hugetlb processing.
|
|
|
|
*/
|
|
|
|
if (is_vm_hugetlb_page(vma)) {
|
|
|
|
page = hugetlb_follow_page_mask(vma, address, flags);
|
|
|
|
if (!page)
|
|
|
|
page = no_page_table(vma, flags);
|
2017-07-06 22:38:44 +00:00
|
|
|
return page;
|
|
|
|
}
|
|
|
|
|
|
|
|
pgd = pgd_offset(mm, address);
|
|
|
|
|
|
|
|
if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
|
|
|
|
return no_page_table(vma, flags);
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
return follow_p4d_mask(vma, address, pgd, flags, ctx);
|
|
|
|
}
|
|
|
|
|
|
|
|
struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
|
|
|
|
unsigned int foll_flags)
|
|
|
|
{
|
|
|
|
struct follow_page_context ctx = { NULL };
|
|
|
|
struct page *page;
|
|
|
|
|
mm: introduce memfd_secret system call to create "secret" memory areas
Introduce "memfd_secret" system call with the ability to create memory
areas visible only in the context of the owning process and not mapped not
only to other processes but in the kernel page tables as well.
The secretmem feature is off by default and the user must explicitly
enable it at the boot time.
Once secretmem is enabled, the user will be able to create a file
descriptor using the memfd_secret() system call. The memory areas created
by mmap() calls from this file descriptor will be unmapped from the kernel
direct map and they will be only mapped in the page table of the processes
that have access to the file descriptor.
Secretmem is designed to provide the following protections:
* Enhanced protection (in conjunction with all the other in-kernel
attack prevention systems) against ROP attacks. Seceretmem makes
"simple" ROP insufficient to perform exfiltration, which increases the
required complexity of the attack. Along with other protections like
the kernel stack size limit and address space layout randomization which
make finding gadgets is really hard, absence of any in-kernel primitive
for accessing secret memory means the one gadget ROP attack can't work.
Since the only way to access secret memory is to reconstruct the missing
mapping entry, the attacker has to recover the physical page and insert
a PTE pointing to it in the kernel and then retrieve the contents. That
takes at least three gadgets which is a level of difficulty beyond most
standard attacks.
* Prevent cross-process secret userspace memory exposures. Once the
secret memory is allocated, the user can't accidentally pass it into the
kernel to be transmitted somewhere. The secreremem pages cannot be
accessed via the direct map and they are disallowed in GUP.
* Harden against exploited kernel flaws. In order to access secretmem,
a kernel-side attack would need to either walk the page tables and
create new ones, or spawn a new privileged uiserspace process to perform
secrets exfiltration using ptrace.
The file descriptor based memory has several advantages over the
"traditional" mm interfaces, such as mlock(), mprotect(), madvise(). File
descriptor approach allows explicit and controlled sharing of the memory
areas, it allows to seal the operations. Besides, file descriptor based
memory paves the way for VMMs to remove the secret memory range from the
userspace hipervisor process, for instance QEMU. Andy Lutomirski says:
"Getting fd-backed memory into a guest will take some possibly major
work in the kernel, but getting vma-backed memory into a guest without
mapping it in the host user address space seems much, much worse."
memfd_secret() is made a dedicated system call rather than an extension to
memfd_create() because it's purpose is to allow the user to create more
secure memory mappings rather than to simply allow file based access to
the memory. Nowadays a new system call cost is negligible while it is way
simpler for userspace to deal with a clear-cut system calls than with a
multiplexer or an overloaded syscall. Moreover, the initial
implementation of memfd_secret() is completely distinct from
memfd_create() so there is no much sense in overloading memfd_create() to
begin with. If there will be a need for code sharing between these
implementation it can be easily achieved without a need to adjust user
visible APIs.
The secret memory remains accessible in the process context using uaccess
primitives, but it is not exposed to the kernel otherwise; secret memory
areas are removed from the direct map and functions in the
follow_page()/get_user_page() family will refuse to return a page that
belongs to the secret memory area.
Once there will be a use case that will require exposing secretmem to the
kernel it will be an opt-in request in the system call flags so that user
would have to decide what data can be exposed to the kernel.
Removing of the pages from the direct map may cause its fragmentation on
architectures that use large pages to map the physical memory which
affects the system performance. However, the original Kconfig text for
CONFIG_DIRECT_GBPAGES said that gigabyte pages in the direct map "... can
improve the kernel's performance a tiny bit ..." (commit 00d1c5e05736
("x86: add gbpages switches")) and the recent report [1] showed that "...
although 1G mappings are a good default choice, there is no compelling
evidence that it must be the only choice". Hence, it is sufficient to
have secretmem disabled by default with the ability of a system
administrator to enable it at boot time.
Pages in the secretmem regions are unevictable and unmovable to avoid
accidental exposure of the sensitive data via swap or during page
migration.
Since the secretmem mappings are locked in memory they cannot exceed
RLIMIT_MEMLOCK. Since these mappings are already locked independently
from mlock(), an attempt to mlock()/munlock() secretmem range would fail
and mlockall()/munlockall() will ignore secretmem mappings.
However, unlike mlock()ed memory, secretmem currently behaves more like
long-term GUP: secretmem mappings are unmovable mappings directly consumed
by user space. With default limits, there is no excessive use of
secretmem and it poses no real problem in combination with
ZONE_MOVABLE/CMA, but in the future this should be addressed to allow
balanced use of large amounts of secretmem along with ZONE_MOVABLE/CMA.
A page that was a part of the secret memory area is cleared when it is
freed to ensure the data is not exposed to the next user of that page.
The following example demonstrates creation of a secret mapping (error
handling is omitted):
fd = memfd_secret(0);
ftruncate(fd, MAP_SIZE);
ptr = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);
[1] https://lore.kernel.org/linux-mm/213b4567-46ce-f116-9cdf-bbd0c884eb3c@linux.intel.com/
[akpm@linux-foundation.org: suppress Kconfig whine]
Link: https://lkml.kernel.org/r/20210518072034.31572-5-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Elena Reshetova <elena.reshetova@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tycho Andersen <tycho@tycho.ws>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-08 01:08:03 +00:00
|
|
|
if (vma_is_secretmem(vma))
|
|
|
|
return NULL;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (WARN_ON_ONCE(foll_flags & FOLL_PIN))
|
2022-05-10 01:20:44 +00:00
|
|
|
return NULL;
|
|
|
|
|
2018-10-26 22:10:28 +00:00
|
|
|
page = follow_page_mask(vma, address, foll_flags, &ctx);
|
|
|
|
if (ctx.pgmap)
|
|
|
|
put_dev_pagemap(ctx.pgmap);
|
|
|
|
return page;
|
2017-07-06 22:38:44 +00:00
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:11 +00:00
|
|
|
static int get_gate_page(struct mm_struct *mm, unsigned long address,
|
|
|
|
unsigned int gup_flags, struct vm_area_struct **vma,
|
|
|
|
struct page **page)
|
|
|
|
{
|
|
|
|
pgd_t *pgd;
|
2017-03-09 14:24:07 +00:00
|
|
|
p4d_t *p4d;
|
2014-06-04 23:08:11 +00:00
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
pte_t *pte;
|
|
|
|
int ret = -EFAULT;
|
|
|
|
|
|
|
|
/* user gate pages are read-only */
|
|
|
|
if (gup_flags & FOLL_WRITE)
|
|
|
|
return -EFAULT;
|
|
|
|
if (address > TASK_SIZE)
|
|
|
|
pgd = pgd_offset_k(address);
|
|
|
|
else
|
|
|
|
pgd = pgd_offset_gate(mm, address);
|
2019-07-12 03:57:43 +00:00
|
|
|
if (pgd_none(*pgd))
|
|
|
|
return -EFAULT;
|
2017-03-09 14:24:07 +00:00
|
|
|
p4d = p4d_offset(pgd, address);
|
2019-07-12 03:57:43 +00:00
|
|
|
if (p4d_none(*p4d))
|
|
|
|
return -EFAULT;
|
2017-03-09 14:24:07 +00:00
|
|
|
pud = pud_offset(p4d, address);
|
2019-07-12 03:57:43 +00:00
|
|
|
if (pud_none(*pud))
|
|
|
|
return -EFAULT;
|
2014-06-04 23:08:11 +00:00
|
|
|
pmd = pmd_offset(pud, address);
|
mm: thp: check pmd migration entry in common path
When THP migration is being used, memory management code needs to handle
pmd migration entries properly. This patch uses !pmd_present() or
is_swap_pmd() (depending on whether pmd_none() needs separate code or
not) to check pmd migration entries at the places where a pmd entry is
present.
Since pmd-related code uses split_huge_page(), split_huge_pmd(),
pmd_trans_huge(), pmd_trans_unstable(), or
pmd_none_or_trans_huge_or_clear_bad(), this patch:
1. adds pmd migration entry split code in split_huge_pmd(),
2. takes care of pmd migration entries whenever pmd_trans_huge() is present,
3. makes pmd_none_or_trans_huge_or_clear_bad() pmd migration entry aware.
Since split_huge_page() uses split_huge_pmd() and pmd_trans_unstable()
is equivalent to pmd_none_or_trans_huge_or_clear_bad(), we do not change
them.
Until this commit, a pmd entry should be:
1. pointing to a pte page,
2. is_swap_pmd(),
3. pmd_trans_huge(),
4. pmd_devmap(), or
5. pmd_none().
Signed-off-by: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Nellans <dnellans@nvidia.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-08 23:11:01 +00:00
|
|
|
if (!pmd_present(*pmd))
|
2014-06-04 23:08:11 +00:00
|
|
|
return -EFAULT;
|
|
|
|
VM_BUG_ON(pmd_trans_huge(*pmd));
|
|
|
|
pte = pte_offset_map(pmd, address);
|
|
|
|
if (pte_none(*pte))
|
|
|
|
goto unmap;
|
|
|
|
*vma = get_gate_vma(mm);
|
|
|
|
if (!page)
|
|
|
|
goto out;
|
|
|
|
*page = vm_normal_page(*vma, address, *pte);
|
|
|
|
if (!*page) {
|
|
|
|
if ((gup_flags & FOLL_DUMP) || !is_zero_pfn(pte_pfn(*pte)))
|
|
|
|
goto unmap;
|
|
|
|
*page = pte_page(*pte);
|
|
|
|
}
|
2022-10-21 17:41:08 +00:00
|
|
|
ret = try_grab_page(*page, gup_flags);
|
|
|
|
if (unlikely(ret))
|
2019-04-11 17:49:19 +00:00
|
|
|
goto unmap;
|
2014-06-04 23:08:11 +00:00
|
|
|
out:
|
|
|
|
ret = 0;
|
|
|
|
unmap:
|
|
|
|
pte_unmap(pte);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2014-08-06 23:07:24 +00:00
|
|
|
/*
|
2023-01-24 20:34:30 +00:00
|
|
|
* mmap_lock must be held on entry. If @flags has FOLL_UNLOCKABLE but not
|
|
|
|
* FOLL_NOWAIT, the mmap_lock may be released. If it is, *@locked will be set
|
|
|
|
* to 0 and -EBUSY returned.
|
2014-08-06 23:07:24 +00:00
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
static int faultin_page(struct vm_area_struct *vma,
|
2022-05-10 01:20:45 +00:00
|
|
|
unsigned long address, unsigned int *flags, bool unshare,
|
|
|
|
int *locked)
|
2014-06-04 23:08:12 +00:00
|
|
|
{
|
|
|
|
unsigned int fault_flags = 0;
|
2018-08-24 00:01:36 +00:00
|
|
|
vm_fault_t ret;
|
2014-06-04 23:08:12 +00:00
|
|
|
|
2021-08-17 20:52:08 +00:00
|
|
|
if (*flags & FOLL_NOFAULT)
|
|
|
|
return -EFAULT;
|
2014-06-04 23:08:12 +00:00
|
|
|
if (*flags & FOLL_WRITE)
|
|
|
|
fault_flags |= FAULT_FLAG_WRITE;
|
2016-02-12 21:02:21 +00:00
|
|
|
if (*flags & FOLL_REMOTE)
|
|
|
|
fault_flags |= FAULT_FLAG_REMOTE;
|
2023-01-24 20:34:29 +00:00
|
|
|
if (*flags & FOLL_UNLOCKABLE) {
|
2020-04-02 04:08:53 +00:00
|
|
|
fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
|
2022-10-11 19:58:06 +00:00
|
|
|
/*
|
|
|
|
* FAULT_FLAG_INTERRUPTIBLE is opt-in. GUP callers must set
|
|
|
|
* FOLL_INTERRUPTIBLE to enable FAULT_FLAG_INTERRUPTIBLE.
|
|
|
|
* That's because some callers may not be prepared to
|
|
|
|
* handle early exits caused by non-fatal signals.
|
|
|
|
*/
|
|
|
|
if (*flags & FOLL_INTERRUPTIBLE)
|
|
|
|
fault_flags |= FAULT_FLAG_INTERRUPTIBLE;
|
|
|
|
}
|
2014-06-04 23:08:12 +00:00
|
|
|
if (*flags & FOLL_NOWAIT)
|
|
|
|
fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_RETRY_NOWAIT;
|
2014-09-17 17:51:48 +00:00
|
|
|
if (*flags & FOLL_TRIED) {
|
2020-04-02 04:08:49 +00:00
|
|
|
/*
|
|
|
|
* Note: FAULT_FLAG_ALLOW_RETRY and FAULT_FLAG_TRIED
|
|
|
|
* can co-exist
|
|
|
|
*/
|
2014-09-17 17:51:48 +00:00
|
|
|
fault_flags |= FAULT_FLAG_TRIED;
|
|
|
|
}
|
2022-05-10 01:20:45 +00:00
|
|
|
if (unshare) {
|
|
|
|
fault_flags |= FAULT_FLAG_UNSHARE;
|
|
|
|
/* FAULT_FLAG_WRITE and FAULT_FLAG_UNSHARE are incompatible */
|
|
|
|
VM_BUG_ON(fault_flags & FAULT_FLAG_WRITE);
|
|
|
|
}
|
2014-06-04 23:08:12 +00:00
|
|
|
|
mm: do page fault accounting in handle_mm_fault
Patch series "mm: Page fault accounting cleanups", v5.
This is v5 of the pf accounting cleanup series. It originates from Gerald
Schaefer's report on an issue a week ago regarding to incorrect page fault
accountings for retried page fault after commit 4064b9827063 ("mm: allow
VM_FAULT_RETRY for multiple times"):
https://lore.kernel.org/lkml/20200610174811.44b94525@thinkpad/
What this series did:
- Correct page fault accounting: we do accounting for a page fault
(no matter whether it's from #PF handling, or gup, or anything else)
only with the one that completed the fault. For example, page fault
retries should not be counted in page fault counters. Same to the
perf events.
- Unify definition of PERF_COUNT_SW_PAGE_FAULTS: currently this perf
event is used in an adhoc way across different archs.
Case (1): for many archs it's done at the entry of a page fault
handler, so that it will also cover e.g. errornous faults.
Case (2): for some other archs, it is only accounted when the page
fault is resolved successfully.
Case (3): there're still quite some archs that have not enabled
this perf event.
Since this series will touch merely all the archs, we unify this
perf event to always follow case (1), which is the one that makes most
sense. And since we moved the accounting into handle_mm_fault, the
other two MAJ/MIN perf events are well taken care of naturally.
- Unify definition of "major faults": the definition of "major
fault" is slightly changed when used in accounting (not
VM_FAULT_MAJOR). More information in patch 1.
- Always account the page fault onto the one that triggered the page
fault. This does not matter much for #PF handlings, but mostly for
gup. More information on this in patch 25.
Patchset layout:
Patch 1: Introduced the accounting in handle_mm_fault(), not enabled.
Patch 2-23: Enable the new accounting for arch #PF handlers one by one.
Patch 24: Enable the new accounting for the rest outliers (gup, iommu, etc.)
Patch 25: Cleanup GUP task_struct pointer since it's not needed any more
This patch (of 25):
This is a preparation patch to move page fault accountings into the
general code in handle_mm_fault(). This includes both the per task
flt_maj/flt_min counters, and the major/minor page fault perf events. To
do this, the pt_regs pointer is passed into handle_mm_fault().
PERF_COUNT_SW_PAGE_FAULTS should still be kept in per-arch page fault
handlers.
So far, all the pt_regs pointer that passed into handle_mm_fault() is
NULL, which means this patch should have no intented functional change.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Cain <bcain@codeaurora.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Chris Zankel <chris@zankel.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Cc: Greentime Hu <green.hu@gmail.com>
Cc: Guo Ren <guoren@kernel.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Nick Hu <nickhu@andestech.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Rich Felker <dalias@libc.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Vincent Chen <deanbo422@gmail.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: http://lkml.kernel.org/r/20200707225021.200906-1-peterx@redhat.com
Link: http://lkml.kernel.org/r/20200707225021.200906-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-08-12 01:37:44 +00:00
|
|
|
ret = handle_mm_fault(vma, address, fault_flags, NULL);
|
mm: avoid unnecessary page fault retires on shared memory types
I observed that for each of the shared file-backed page faults, we're very
likely to retry one more time for the 1st write fault upon no page. It's
because we'll need to release the mmap lock for dirty rate limit purpose
with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
Then after that throttling we return VM_FAULT_RETRY.
We did that probably because VM_FAULT_RETRY is the only way we can return
to the fault handler at that time telling it we've released the mmap lock.
However that's not ideal because it's very likely the fault does not need
to be retried at all since the pgtable was well installed before the
throttling, so the next continuous fault (including taking mmap read lock,
walk the pgtable, etc.) could be in most cases unnecessary.
It's not only slowing down page faults for shared file-backed, but also add
more mmap lock contention which is in most cases not needed at all.
To observe this, one could try to write to some shmem page and look at
"pgfault" value in /proc/vmstat, then we should expect 2 counts for each
shmem write simply because we retried, and vm event "pgfault" will capture
that.
To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
show that we've completed the whole fault and released the lock. It's also
a hint that we should very possibly not need another fault immediately on
this page because we've just completed it.
This patch provides a ~12% perf boost on my aarch64 test VM with a simple
program sequentially dirtying 400MB shmem file being mmap()ed and these are
the time it needs:
Before: 650.980 ms (+-1.94%)
After: 569.396 ms (+-1.38%)
I believe it could help more than that.
We need some special care on GUP and the s390 pgfault handler (for gmap
code before returning from pgfault), the rest changes in the page fault
handlers should be relatively straightforward.
Another thing to mention is that mm_account_fault() does take this new
fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
them as-is.
Link: https://lkml.kernel.org/r/20220530183450.42886-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vineet Gupta <vgupta@kernel.org>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> [arm part]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Stafford Horne <shorne@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Brian Cain <bcain@quicinc.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Weinberger <richard@nod.at>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Anton Ivanov <anton.ivanov@cambridgegreys.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Will Deacon <will@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Chris Zankel <chris@zankel.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Rich Felker <dalias@libc.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Helge Deller <deller@gmx.de>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-30 18:34:50 +00:00
|
|
|
|
|
|
|
if (ret & VM_FAULT_COMPLETED) {
|
|
|
|
/*
|
|
|
|
* With FAULT_FLAG_RETRY_NOWAIT we'll never release the
|
|
|
|
* mmap lock in the page fault handler. Sanity check this.
|
|
|
|
*/
|
|
|
|
WARN_ON_ONCE(fault_flags & FAULT_FLAG_RETRY_NOWAIT);
|
2023-01-24 20:34:30 +00:00
|
|
|
*locked = 0;
|
|
|
|
|
mm: avoid unnecessary page fault retires on shared memory types
I observed that for each of the shared file-backed page faults, we're very
likely to retry one more time for the 1st write fault upon no page. It's
because we'll need to release the mmap lock for dirty rate limit purpose
with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
Then after that throttling we return VM_FAULT_RETRY.
We did that probably because VM_FAULT_RETRY is the only way we can return
to the fault handler at that time telling it we've released the mmap lock.
However that's not ideal because it's very likely the fault does not need
to be retried at all since the pgtable was well installed before the
throttling, so the next continuous fault (including taking mmap read lock,
walk the pgtable, etc.) could be in most cases unnecessary.
It's not only slowing down page faults for shared file-backed, but also add
more mmap lock contention which is in most cases not needed at all.
To observe this, one could try to write to some shmem page and look at
"pgfault" value in /proc/vmstat, then we should expect 2 counts for each
shmem write simply because we retried, and vm event "pgfault" will capture
that.
To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
show that we've completed the whole fault and released the lock. It's also
a hint that we should very possibly not need another fault immediately on
this page because we've just completed it.
This patch provides a ~12% perf boost on my aarch64 test VM with a simple
program sequentially dirtying 400MB shmem file being mmap()ed and these are
the time it needs:
Before: 650.980 ms (+-1.94%)
After: 569.396 ms (+-1.38%)
I believe it could help more than that.
We need some special care on GUP and the s390 pgfault handler (for gmap
code before returning from pgfault), the rest changes in the page fault
handlers should be relatively straightforward.
Another thing to mention is that mm_account_fault() does take this new
fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
them as-is.
Link: https://lkml.kernel.org/r/20220530183450.42886-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vineet Gupta <vgupta@kernel.org>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> [arm part]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Stafford Horne <shorne@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Brian Cain <bcain@quicinc.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Weinberger <richard@nod.at>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Anton Ivanov <anton.ivanov@cambridgegreys.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Will Deacon <will@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Chris Zankel <chris@zankel.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Rich Felker <dalias@libc.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Helge Deller <deller@gmx.de>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-30 18:34:50 +00:00
|
|
|
/*
|
|
|
|
* We should do the same as VM_FAULT_RETRY, but let's not
|
|
|
|
* return -EBUSY since that's not reflecting the reality of
|
|
|
|
* what has happened - we've just fully completed a page
|
|
|
|
* fault, with the mmap lock released. Use -EAGAIN to show
|
|
|
|
* that we want to take the mmap lock _again_.
|
|
|
|
*/
|
|
|
|
return -EAGAIN;
|
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:12 +00:00
|
|
|
if (ret & VM_FAULT_ERROR) {
|
2017-06-02 21:46:46 +00:00
|
|
|
int err = vm_fault_to_errno(ret, *flags);
|
|
|
|
|
|
|
|
if (err)
|
|
|
|
return err;
|
2014-06-04 23:08:12 +00:00
|
|
|
BUG();
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ret & VM_FAULT_RETRY) {
|
2023-01-24 20:34:30 +00:00
|
|
|
if (!(fault_flags & FAULT_FLAG_RETRY_NOWAIT))
|
mm/gup: rename "nonblocking" to "locked" where proper
Patch series "mm: Page fault enhancements", v6.
This series contains cleanups and enhancements to current page fault
logic. The whole idea comes from the discussion between Andrea and Linus
on the bug reported by syzbot here:
https://lkml.org/lkml/2017/11/2/833
Basically it does two things:
(a) Allows the page fault logic to be more interactive on not only
SIGKILL, but also the rest of userspace signals, and,
(b) Allows the page fault retry (VM_FAULT_RETRY) to happen for more
than once.
For (a): with the changes we should be able to react faster when page
faults are working in parallel with userspace signals like SIGSTOP and
SIGCONT (and more), and with that we can remove the buggy part in
userfaultfd and benefit the whole page fault mechanism on faster signal
processing to reach the userspace.
For (b), we should be able to allow the page fault handler to loop for
even more than twice. Some context: for now since we have
FAULT_FLAG_ALLOW_RETRY we can allow to retry the page fault once with the
same interrupt context, however never more than twice. This can be not
only a potential cleanup to remove this assumption since AFAIU the code
itself doesn't really have this twice-only limitation (though that should
be a protective approach in the past), at the same time it'll greatly
simplify future works like userfaultfd write-protect where it's possible
to retry for more than twice (please have a look at [1] below for a
possible user that might require the page fault to be handled for a third
time; if we can remove the retry limitation we can simply drop that patch
and those complexity).
This patch (of 16):
There's plenty of places around __get_user_pages() that has a parameter
"nonblocking" which does not really mean that "it won't block" (because it
can really block) but instead it shows whether the mmap_sem is released by
up_read() during the page fault handling mostly when VM_FAULT_RETRY is
returned.
We have the correct naming in e.g. get_user_pages_locked() or
get_user_pages_remote() as "locked", however there're still many places
that are using the "nonblocking" as name.
Renaming the places to "locked" where proper to better suite the
functionality of the variable. While at it, fixing up some of the
comments accordingly.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Brian Geffon <bgeffon@google.com>
Reviewed-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Martin Cracauer <cracauer@cons.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Bobby Powers <bobbypowers@gmail.com>
Cc: Maya Gokhale <gokhale2@llnl.gov>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Marty McFadden <mcfadden8@llnl.gov>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Link: http://lkml.kernel.org/r/20200220155353.8676-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:07:58 +00:00
|
|
|
*locked = 0;
|
2014-06-04 23:08:12 +00:00
|
|
|
return -EBUSY;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:13 +00:00
|
|
|
static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags)
|
|
|
|
{
|
|
|
|
vm_flags_t vm_flags = vma->vm_flags;
|
2016-02-12 21:02:21 +00:00
|
|
|
int write = (gup_flags & FOLL_WRITE);
|
|
|
|
int foreign = (gup_flags & FOLL_REMOTE);
|
2014-06-04 23:08:13 +00:00
|
|
|
|
|
|
|
if (vm_flags & (VM_IO | VM_PFNMAP))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2018-05-11 06:11:44 +00:00
|
|
|
if (gup_flags & FOLL_ANON && !vma_is_anonymous(vma))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2020-12-15 03:05:48 +00:00
|
|
|
if ((gup_flags & FOLL_LONGTERM) && vma_is_fsdax(vma))
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
mm: introduce memfd_secret system call to create "secret" memory areas
Introduce "memfd_secret" system call with the ability to create memory
areas visible only in the context of the owning process and not mapped not
only to other processes but in the kernel page tables as well.
The secretmem feature is off by default and the user must explicitly
enable it at the boot time.
Once secretmem is enabled, the user will be able to create a file
descriptor using the memfd_secret() system call. The memory areas created
by mmap() calls from this file descriptor will be unmapped from the kernel
direct map and they will be only mapped in the page table of the processes
that have access to the file descriptor.
Secretmem is designed to provide the following protections:
* Enhanced protection (in conjunction with all the other in-kernel
attack prevention systems) against ROP attacks. Seceretmem makes
"simple" ROP insufficient to perform exfiltration, which increases the
required complexity of the attack. Along with other protections like
the kernel stack size limit and address space layout randomization which
make finding gadgets is really hard, absence of any in-kernel primitive
for accessing secret memory means the one gadget ROP attack can't work.
Since the only way to access secret memory is to reconstruct the missing
mapping entry, the attacker has to recover the physical page and insert
a PTE pointing to it in the kernel and then retrieve the contents. That
takes at least three gadgets which is a level of difficulty beyond most
standard attacks.
* Prevent cross-process secret userspace memory exposures. Once the
secret memory is allocated, the user can't accidentally pass it into the
kernel to be transmitted somewhere. The secreremem pages cannot be
accessed via the direct map and they are disallowed in GUP.
* Harden against exploited kernel flaws. In order to access secretmem,
a kernel-side attack would need to either walk the page tables and
create new ones, or spawn a new privileged uiserspace process to perform
secrets exfiltration using ptrace.
The file descriptor based memory has several advantages over the
"traditional" mm interfaces, such as mlock(), mprotect(), madvise(). File
descriptor approach allows explicit and controlled sharing of the memory
areas, it allows to seal the operations. Besides, file descriptor based
memory paves the way for VMMs to remove the secret memory range from the
userspace hipervisor process, for instance QEMU. Andy Lutomirski says:
"Getting fd-backed memory into a guest will take some possibly major
work in the kernel, but getting vma-backed memory into a guest without
mapping it in the host user address space seems much, much worse."
memfd_secret() is made a dedicated system call rather than an extension to
memfd_create() because it's purpose is to allow the user to create more
secure memory mappings rather than to simply allow file based access to
the memory. Nowadays a new system call cost is negligible while it is way
simpler for userspace to deal with a clear-cut system calls than with a
multiplexer or an overloaded syscall. Moreover, the initial
implementation of memfd_secret() is completely distinct from
memfd_create() so there is no much sense in overloading memfd_create() to
begin with. If there will be a need for code sharing between these
implementation it can be easily achieved without a need to adjust user
visible APIs.
The secret memory remains accessible in the process context using uaccess
primitives, but it is not exposed to the kernel otherwise; secret memory
areas are removed from the direct map and functions in the
follow_page()/get_user_page() family will refuse to return a page that
belongs to the secret memory area.
Once there will be a use case that will require exposing secretmem to the
kernel it will be an opt-in request in the system call flags so that user
would have to decide what data can be exposed to the kernel.
Removing of the pages from the direct map may cause its fragmentation on
architectures that use large pages to map the physical memory which
affects the system performance. However, the original Kconfig text for
CONFIG_DIRECT_GBPAGES said that gigabyte pages in the direct map "... can
improve the kernel's performance a tiny bit ..." (commit 00d1c5e05736
("x86: add gbpages switches")) and the recent report [1] showed that "...
although 1G mappings are a good default choice, there is no compelling
evidence that it must be the only choice". Hence, it is sufficient to
have secretmem disabled by default with the ability of a system
administrator to enable it at boot time.
Pages in the secretmem regions are unevictable and unmovable to avoid
accidental exposure of the sensitive data via swap or during page
migration.
Since the secretmem mappings are locked in memory they cannot exceed
RLIMIT_MEMLOCK. Since these mappings are already locked independently
from mlock(), an attempt to mlock()/munlock() secretmem range would fail
and mlockall()/munlockall() will ignore secretmem mappings.
However, unlike mlock()ed memory, secretmem currently behaves more like
long-term GUP: secretmem mappings are unmovable mappings directly consumed
by user space. With default limits, there is no excessive use of
secretmem and it poses no real problem in combination with
ZONE_MOVABLE/CMA, but in the future this should be addressed to allow
balanced use of large amounts of secretmem along with ZONE_MOVABLE/CMA.
A page that was a part of the secret memory area is cleared when it is
freed to ensure the data is not exposed to the next user of that page.
The following example demonstrates creation of a secret mapping (error
handling is omitted):
fd = memfd_secret(0);
ftruncate(fd, MAP_SIZE);
ptr = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);
[1] https://lore.kernel.org/linux-mm/213b4567-46ce-f116-9cdf-bbd0c884eb3c@linux.intel.com/
[akpm@linux-foundation.org: suppress Kconfig whine]
Link: https://lkml.kernel.org/r/20210518072034.31572-5-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Elena Reshetova <elena.reshetova@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tycho Andersen <tycho@tycho.ws>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-08 01:08:03 +00:00
|
|
|
if (vma_is_secretmem(vma))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2016-02-12 21:02:21 +00:00
|
|
|
if (write) {
|
2014-06-04 23:08:13 +00:00
|
|
|
if (!(vm_flags & VM_WRITE)) {
|
|
|
|
if (!(gup_flags & FOLL_FORCE))
|
|
|
|
return -EFAULT;
|
mm/gup: disallow FOLL_FORCE|FOLL_WRITE on hugetlb mappings
hugetlb does not support fake write-faults (write faults without write
permissions). However, we are currently able to trigger a
FAULT_FLAG_WRITE fault on a VMA without VM_WRITE.
If we'd ever want to support FOLL_FORCE|FOLL_WRITE, we'd have to teach
hugetlb to:
(1) Leave the page mapped R/O after the fake write-fault, like
maybe_mkwrite() does.
(2) Allow writing to an exclusive anon page that's mapped R/O when
FOLL_FORCE is set, like can_follow_write_pte(). E.g.,
__follow_hugetlb_must_fault() needs adjustment.
For now, it's not clear if that added complexity is really required.
History tolds us that FOLL_FORCE is dangerous and that we better limit its
use to a bare minimum.
--------------------------------------------------------------------------
#include <stdio.h>
#include <stdlib.h>
#include <fcntl.h>
#include <unistd.h>
#include <errno.h>
#include <stdint.h>
#include <sys/mman.h>
#include <linux/mman.h>
int main(int argc, char **argv)
{
char *map;
int mem_fd;
map = mmap(NULL, 2 * 1024 * 1024u, PROT_READ,
MAP_PRIVATE|MAP_ANON|MAP_HUGETLB|MAP_HUGE_2MB, -1, 0);
if (map == MAP_FAILED) {
fprintf(stderr, "mmap() failed: %d\n", errno);
return 1;
}
mem_fd = open("/proc/self/mem", O_RDWR);
if (mem_fd < 0) {
fprintf(stderr, "open(/proc/self/mem) failed: %d\n", errno);
return 1;
}
if (pwrite(mem_fd, "0", 1, (uintptr_t) map) == 1) {
fprintf(stderr, "write() succeeded, which is unexpected\n");
return 1;
}
printf("write() failed as expected: %d\n", errno);
return 0;
}
--------------------------------------------------------------------------
Fortunately, we have a sanity check in hugetlb_wp() in place ever since
commit 1d8d14641fd9 ("mm/hugetlb: support write-faults in shared
mappings"), that bails out instead of silently mapping a page writable in
a !PROT_WRITE VMA.
Consequently, above reproducer triggers a warning, similar to the one
reported by szsbot:
------------[ cut here ]------------
WARNING: CPU: 1 PID: 3612 at mm/hugetlb.c:5313 hugetlb_wp+0x20a/0x1af0 mm/hugetlb.c:5313
Modules linked in:
CPU: 1 PID: 3612 Comm: syz-executor250 Not tainted 6.1.0-rc2-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022
RIP: 0010:hugetlb_wp+0x20a/0x1af0 mm/hugetlb.c:5313
Code: ea 03 80 3c 02 00 0f 85 31 14 00 00 49 8b 5f 20 31 ff 48 89 dd 83 e5 02 48 89 ee e8 70 ab b7 ff 48 85 ed 75 5b e8 76 ae b7 ff <0f> 0b 41 bd 40 00 00 00 e8 69 ae b7 ff 48 b8 00 00 00 00 00 fc ff
RSP: 0018:ffffc90003caf620 EFLAGS: 00010293
RAX: 0000000000000000 RBX: 0000000008640070 RCX: 0000000000000000
RDX: ffff88807b963a80 RSI: ffffffff81c4ed2a RDI: 0000000000000007
RBP: 0000000000000000 R08: 0000000000000007 R09: 0000000000000000
R10: 0000000000000000 R11: 000000000008c07e R12: ffff888023805800
R13: 0000000000000000 R14: ffffffff91217f38 R15: ffff88801d4b0360
FS: 0000555555bba300(0000) GS:ffff8880b9b00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fff7a47a1b8 CR3: 000000002378d000 CR4: 00000000003506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
hugetlb_no_page mm/hugetlb.c:5755 [inline]
hugetlb_fault+0x19cc/0x2060 mm/hugetlb.c:5874
follow_hugetlb_page+0x3f3/0x1850 mm/hugetlb.c:6301
__get_user_pages+0x2cb/0xf10 mm/gup.c:1202
__get_user_pages_locked mm/gup.c:1434 [inline]
__get_user_pages_remote+0x18f/0x830 mm/gup.c:2187
get_user_pages_remote+0x84/0xc0 mm/gup.c:2260
__access_remote_vm+0x287/0x6b0 mm/memory.c:5517
ptrace_access_vm+0x181/0x1d0 kernel/ptrace.c:61
generic_ptrace_pokedata kernel/ptrace.c:1323 [inline]
ptrace_request+0xb46/0x10c0 kernel/ptrace.c:1046
arch_ptrace+0x36/0x510 arch/x86/kernel/ptrace.c:828
__do_sys_ptrace kernel/ptrace.c:1296 [inline]
__se_sys_ptrace kernel/ptrace.c:1269 [inline]
__x64_sys_ptrace+0x178/0x2a0 kernel/ptrace.c:1269
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
[...]
So let's silence that warning by teaching GUP code that FOLL_FORCE -- so
far -- does not apply to hugetlb.
Note that FOLL_FORCE for read-access seems to be working as expected. The
assumption is that this has been broken forever, only ever since above
commit, we actually detect the wrong handling and WARN_ON_ONCE().
I assume this has been broken at least since 2014, when mm/gup.c came to
life. I failed to come up with a suitable Fixes tag quickly.
Link: https://lkml.kernel.org/r/20221031152524.173644-1-david@redhat.com
Fixes: 1d8d14641fd9 ("mm/hugetlb: support write-faults in shared mappings")
Signed-off-by: David Hildenbrand <david@redhat.com>
Reported-by: <syzbot+f0b97304ef90f0d0b1dc@syzkaller.appspotmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-10-31 15:25:24 +00:00
|
|
|
/* hugetlb does not support FOLL_FORCE|FOLL_WRITE. */
|
|
|
|
if (is_vm_hugetlb_page(vma))
|
|
|
|
return -EFAULT;
|
2014-06-04 23:08:13 +00:00
|
|
|
/*
|
|
|
|
* We used to let the write,force case do COW in a
|
|
|
|
* VM_MAYWRITE VM_SHARED !VM_WRITE vma, so ptrace could
|
|
|
|
* set a breakpoint in a read-only mapping of an
|
|
|
|
* executable, without corrupting the file (yet only
|
|
|
|
* when that file had been opened for writing!).
|
|
|
|
* Anon pages in shared mappings are surprising: now
|
|
|
|
* just reject it.
|
|
|
|
*/
|
2016-01-31 02:03:16 +00:00
|
|
|
if (!is_cow_mapping(vm_flags))
|
2014-06-04 23:08:13 +00:00
|
|
|
return -EFAULT;
|
|
|
|
}
|
|
|
|
} else if (!(vm_flags & VM_READ)) {
|
|
|
|
if (!(gup_flags & FOLL_FORCE))
|
|
|
|
return -EFAULT;
|
|
|
|
/*
|
|
|
|
* Is there actually any vma we can reach here which does not
|
|
|
|
* have VM_MAYREAD set?
|
|
|
|
*/
|
|
|
|
if (!(vm_flags & VM_MAYREAD))
|
|
|
|
return -EFAULT;
|
|
|
|
}
|
2016-02-12 21:02:24 +00:00
|
|
|
/*
|
|
|
|
* gups are always data accesses, not instruction
|
|
|
|
* fetches, so execute=false here
|
|
|
|
*/
|
|
|
|
if (!arch_vma_access_permitted(vma, write, false, foreign))
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
return -EFAULT;
|
2014-06-04 23:08:13 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:10 +00:00
|
|
|
/**
|
|
|
|
* __get_user_pages() - pin user pages in memory
|
|
|
|
* @mm: mm_struct of target mm
|
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying pin behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long. Or NULL, if caller
|
|
|
|
* only intends to ensure the pages are faulted in.
|
|
|
|
* @vmas: array of pointers to vmas corresponding to each page.
|
|
|
|
* Or NULL if the caller does not require them.
|
2020-06-09 04:33:54 +00:00
|
|
|
* @locked: whether we're still with the mmap_lock held
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
2019-12-01 01:49:53 +00:00
|
|
|
* Returns either number of pages pinned (which may be less than the
|
|
|
|
* number requested), or an error. Details about the return value:
|
|
|
|
*
|
|
|
|
* -- If nr_pages is 0, returns 0.
|
|
|
|
* -- If nr_pages is >0, but no pages were pinned, returns -errno.
|
|
|
|
* -- If nr_pages is >0, and some pages were pinned, returns the number of
|
|
|
|
* pages pinned. Again, this may be less than nr_pages.
|
2020-06-03 23:03:25 +00:00
|
|
|
* -- 0 return value is possible when the fault would need to be retried.
|
2019-12-01 01:49:53 +00:00
|
|
|
*
|
|
|
|
* The caller is responsible for releasing returned @pages, via put_page().
|
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* @vmas are valid only as long as mmap_lock is held.
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* Must be called with mmap_lock held. It may be released. See below.
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
|
|
|
* __get_user_pages walks a process's page tables and takes a reference to
|
|
|
|
* each struct page that each user address corresponds to at a given
|
|
|
|
* instant. That is, it takes the page that would be accessed if a user
|
|
|
|
* thread accesses the given user virtual address at that instant.
|
|
|
|
*
|
|
|
|
* This does not guarantee that the page exists in the user mappings when
|
|
|
|
* __get_user_pages returns, and there may even be a completely different
|
|
|
|
* page there in some cases (eg. if mmapped pagecache has been invalidated
|
2023-01-25 18:08:47 +00:00
|
|
|
* and subsequently re-faulted). However it does guarantee that the page
|
2014-06-04 23:08:10 +00:00
|
|
|
* won't be freed completely. And mostly callers simply care that the page
|
|
|
|
* contains data that was valid *at some point in time*. Typically, an IO
|
|
|
|
* or similar operation cannot guarantee anything stronger anyway because
|
|
|
|
* locks can't be held over the syscall boundary.
|
|
|
|
*
|
|
|
|
* If @gup_flags & FOLL_WRITE == 0, the page must not be written to. If
|
|
|
|
* the page is written to, set_page_dirty (or set_page_dirty_lock, as
|
|
|
|
* appropriate) must be called after the page is finished with, and
|
|
|
|
* before put_page is called.
|
|
|
|
*
|
2023-01-24 20:34:30 +00:00
|
|
|
* If FOLL_UNLOCKABLE is set without FOLL_NOWAIT then the mmap_lock may
|
|
|
|
* be released. If this happens *@locked will be set to 0 on return.
|
2014-08-06 23:07:24 +00:00
|
|
|
*
|
2023-01-24 20:34:30 +00:00
|
|
|
* A caller using such a combination of @gup_flags must therefore hold the
|
|
|
|
* mmap_lock for reading only, and recognize when it's been released. Otherwise,
|
|
|
|
* it must be held for either reading or writing and will not be released.
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
|
|
|
* In most cases, get_user_pages or get_user_pages_fast should be used
|
|
|
|
* instead of __get_user_pages. __get_user_pages should be used only if
|
|
|
|
* you need some special @gup_flags.
|
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
static long __get_user_pages(struct mm_struct *mm,
|
2014-06-04 23:08:10 +00:00
|
|
|
unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
mm/gup: rename "nonblocking" to "locked" where proper
Patch series "mm: Page fault enhancements", v6.
This series contains cleanups and enhancements to current page fault
logic. The whole idea comes from the discussion between Andrea and Linus
on the bug reported by syzbot here:
https://lkml.org/lkml/2017/11/2/833
Basically it does two things:
(a) Allows the page fault logic to be more interactive on not only
SIGKILL, but also the rest of userspace signals, and,
(b) Allows the page fault retry (VM_FAULT_RETRY) to happen for more
than once.
For (a): with the changes we should be able to react faster when page
faults are working in parallel with userspace signals like SIGSTOP and
SIGCONT (and more), and with that we can remove the buggy part in
userfaultfd and benefit the whole page fault mechanism on faster signal
processing to reach the userspace.
For (b), we should be able to allow the page fault handler to loop for
even more than twice. Some context: for now since we have
FAULT_FLAG_ALLOW_RETRY we can allow to retry the page fault once with the
same interrupt context, however never more than twice. This can be not
only a potential cleanup to remove this assumption since AFAIU the code
itself doesn't really have this twice-only limitation (though that should
be a protective approach in the past), at the same time it'll greatly
simplify future works like userfaultfd write-protect where it's possible
to retry for more than twice (please have a look at [1] below for a
possible user that might require the page fault to be handled for a third
time; if we can remove the retry limitation we can simply drop that patch
and those complexity).
This patch (of 16):
There's plenty of places around __get_user_pages() that has a parameter
"nonblocking" which does not really mean that "it won't block" (because it
can really block) but instead it shows whether the mmap_sem is released by
up_read() during the page fault handling mostly when VM_FAULT_RETRY is
returned.
We have the correct naming in e.g. get_user_pages_locked() or
get_user_pages_remote() as "locked", however there're still many places
that are using the "nonblocking" as name.
Renaming the places to "locked" where proper to better suite the
functionality of the variable. While at it, fixing up some of the
comments accordingly.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Brian Geffon <bgeffon@google.com>
Reviewed-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Martin Cracauer <cracauer@cons.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Bobby Powers <bobbypowers@gmail.com>
Cc: Maya Gokhale <gokhale2@llnl.gov>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Marty McFadden <mcfadden8@llnl.gov>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Link: http://lkml.kernel.org/r/20200220155353.8676-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:07:58 +00:00
|
|
|
struct vm_area_struct **vmas, int *locked)
|
2014-06-04 23:08:10 +00:00
|
|
|
{
|
2018-10-26 22:10:28 +00:00
|
|
|
long ret = 0, i = 0;
|
2014-06-04 23:08:13 +00:00
|
|
|
struct vm_area_struct *vma = NULL;
|
2018-10-26 22:10:28 +00:00
|
|
|
struct follow_page_context ctx = { NULL };
|
2014-06-04 23:08:10 +00:00
|
|
|
|
|
|
|
if (!nr_pages)
|
|
|
|
return 0;
|
|
|
|
|
2023-03-12 11:26:00 +00:00
|
|
|
start = untagged_addr_remote(mm, start);
|
2019-09-25 23:48:34 +00:00
|
|
|
|
2020-01-31 06:12:54 +00:00
|
|
|
VM_BUG_ON(!!pages != !!(gup_flags & (FOLL_GET | FOLL_PIN)));
|
2014-06-04 23:08:10 +00:00
|
|
|
|
|
|
|
do {
|
2014-06-04 23:08:13 +00:00
|
|
|
struct page *page;
|
|
|
|
unsigned int foll_flags = gup_flags;
|
|
|
|
unsigned int page_increm;
|
|
|
|
|
|
|
|
/* first iteration or cross vma bound */
|
|
|
|
if (!vma || start >= vma->vm_end) {
|
2023-06-25 21:02:25 +00:00
|
|
|
vma = find_vma(mm, start);
|
|
|
|
if (vma && (start < vma->vm_start)) {
|
|
|
|
WARN_ON_ONCE(vma->vm_flags & VM_GROWSDOWN);
|
|
|
|
vma = NULL;
|
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
if (!vma && in_gate_area(mm, start)) {
|
|
|
|
ret = get_gate_page(mm, start & PAGE_MASK,
|
|
|
|
gup_flags, &vma,
|
|
|
|
pages ? &pages[i] : NULL);
|
|
|
|
if (ret)
|
mm/gup: finish consolidating error handling
Commit df06b37ffe5a ("mm/gup: cache dev_pagemap while pinning pages")
attempted to operate on each page that get_user_pages had retrieved. In
order to do that, it created a common exit point from the routine.
However, one case was missed, which this patch fixes up.
Also, there was still an unnecessary shadow declaration (with a
different type) of the "ret" variable, which this patch removes.
Keith's description of the situation is:
This also fixes a potentially leaked dev_pagemap reference count if a
failure occurs when an iteration crosses a vma boundary. I don't think
it's normal to have different vma's on a users mapped zone device
memory, but good to fix anyway.
I actually thought that this code:
/* first iteration or cross vma bound */
if (!vma || start >= vma->vm_end) {
vma = find_extend_vma(mm, start);
if (!vma && in_gate_area(mm, start)) {
ret = get_gate_page(mm, start & PAGE_MASK,
gup_flags, &vma,
pages ? &pages[i] : NULL);
if (ret)
goto out;
dealt with the "you're trying to pin the gate page, as part of this
call", rather than the generic case of crossing a vma boundary. (I
think there's a fine point that I must be overlooking.) But it's still a
valid case, either way.
Link: http://lkml.kernel.org/r/20181121081402.29641-2-jhubbard@nvidia.com
Fixes: df06b37ffe5a4 ("mm/gup: cache dev_pagemap while pinning pages")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Keith Busch <keith.busch@intel.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-11-30 22:08:53 +00:00
|
|
|
goto out;
|
2018-10-26 22:10:28 +00:00
|
|
|
ctx.page_mask = 0;
|
2014-06-04 23:08:13 +00:00
|
|
|
goto next_page;
|
|
|
|
}
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2020-12-15 03:05:48 +00:00
|
|
|
if (!vma) {
|
2018-10-26 22:10:28 +00:00
|
|
|
ret = -EFAULT;
|
|
|
|
goto out;
|
|
|
|
}
|
2020-12-15 03:05:48 +00:00
|
|
|
ret = check_vma_flags(vma, gup_flags);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
|
2014-06-04 23:08:13 +00:00
|
|
|
if (is_vm_hugetlb_page(vma)) {
|
|
|
|
i = follow_hugetlb_page(mm, vma, pages, vmas,
|
|
|
|
&start, &nr_pages, i,
|
2020-08-21 23:49:57 +00:00
|
|
|
gup_flags, locked);
|
2023-01-24 20:34:30 +00:00
|
|
|
if (!*locked) {
|
2020-04-02 04:08:02 +00:00
|
|
|
/*
|
|
|
|
* We've got a VM_FAULT_RETRY
|
2020-06-09 04:33:54 +00:00
|
|
|
* and we've lost mmap_lock.
|
2020-04-02 04:08:02 +00:00
|
|
|
* We must stop here.
|
|
|
|
*/
|
|
|
|
BUG_ON(gup_flags & FOLL_NOWAIT);
|
|
|
|
goto out;
|
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
continue;
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
}
|
|
|
|
retry:
|
|
|
|
/*
|
|
|
|
* If we have a pending SIGKILL, don't keep faulting pages and
|
|
|
|
* potentially allocating memory.
|
|
|
|
*/
|
2019-01-03 23:28:55 +00:00
|
|
|
if (fatal_signal_pending(current)) {
|
2020-04-21 01:13:55 +00:00
|
|
|
ret = -EINTR;
|
2018-10-26 22:10:28 +00:00
|
|
|
goto out;
|
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
cond_resched();
|
2018-10-26 22:10:28 +00:00
|
|
|
|
|
|
|
page = follow_page_mask(vma, start, foll_flags, &ctx);
|
2022-05-10 01:20:45 +00:00
|
|
|
if (!page || PTR_ERR(page) == -EMLINK) {
|
|
|
|
ret = faultin_page(vma, start, &foll_flags,
|
|
|
|
PTR_ERR(page) == -EMLINK, locked);
|
2014-06-04 23:08:13 +00:00
|
|
|
switch (ret) {
|
|
|
|
case 0:
|
|
|
|
goto retry;
|
2018-10-26 22:10:28 +00:00
|
|
|
case -EBUSY:
|
mm: avoid unnecessary page fault retires on shared memory types
I observed that for each of the shared file-backed page faults, we're very
likely to retry one more time for the 1st write fault upon no page. It's
because we'll need to release the mmap lock for dirty rate limit purpose
with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
Then after that throttling we return VM_FAULT_RETRY.
We did that probably because VM_FAULT_RETRY is the only way we can return
to the fault handler at that time telling it we've released the mmap lock.
However that's not ideal because it's very likely the fault does not need
to be retried at all since the pgtable was well installed before the
throttling, so the next continuous fault (including taking mmap read lock,
walk the pgtable, etc.) could be in most cases unnecessary.
It's not only slowing down page faults for shared file-backed, but also add
more mmap lock contention which is in most cases not needed at all.
To observe this, one could try to write to some shmem page and look at
"pgfault" value in /proc/vmstat, then we should expect 2 counts for each
shmem write simply because we retried, and vm event "pgfault" will capture
that.
To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
show that we've completed the whole fault and released the lock. It's also
a hint that we should very possibly not need another fault immediately on
this page because we've just completed it.
This patch provides a ~12% perf boost on my aarch64 test VM with a simple
program sequentially dirtying 400MB shmem file being mmap()ed and these are
the time it needs:
Before: 650.980 ms (+-1.94%)
After: 569.396 ms (+-1.38%)
I believe it could help more than that.
We need some special care on GUP and the s390 pgfault handler (for gmap
code before returning from pgfault), the rest changes in the page fault
handlers should be relatively straightforward.
Another thing to mention is that mm_account_fault() does take this new
fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
them as-is.
Link: https://lkml.kernel.org/r/20220530183450.42886-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vineet Gupta <vgupta@kernel.org>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> [arm part]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Stafford Horne <shorne@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Brian Cain <bcain@quicinc.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Weinberger <richard@nod.at>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Anton Ivanov <anton.ivanov@cambridgegreys.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Will Deacon <will@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Chris Zankel <chris@zankel.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Rich Felker <dalias@libc.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Helge Deller <deller@gmx.de>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-30 18:34:50 +00:00
|
|
|
case -EAGAIN:
|
2018-10-26 22:10:28 +00:00
|
|
|
ret = 0;
|
2020-04-07 03:08:39 +00:00
|
|
|
fallthrough;
|
2014-06-04 23:08:13 +00:00
|
|
|
case -EFAULT:
|
|
|
|
case -ENOMEM:
|
|
|
|
case -EHWPOISON:
|
2018-10-26 22:10:28 +00:00
|
|
|
goto out;
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
BUG();
|
2015-09-04 22:47:55 +00:00
|
|
|
} else if (PTR_ERR(page) == -EEXIST) {
|
|
|
|
/*
|
|
|
|
* Proper page table entry exists, but no corresponding
|
2022-03-22 21:39:40 +00:00
|
|
|
* struct page. If the caller expects **pages to be
|
|
|
|
* filled in, bail out now, because that can't be done
|
|
|
|
* for this page.
|
2015-09-04 22:47:55 +00:00
|
|
|
*/
|
2022-03-22 21:39:40 +00:00
|
|
|
if (pages) {
|
|
|
|
ret = PTR_ERR(page);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2015-09-04 22:47:55 +00:00
|
|
|
goto next_page;
|
|
|
|
} else if (IS_ERR(page)) {
|
2018-10-26 22:10:28 +00:00
|
|
|
ret = PTR_ERR(page);
|
|
|
|
goto out;
|
2015-09-04 22:47:55 +00:00
|
|
|
}
|
2014-06-04 23:08:13 +00:00
|
|
|
if (pages) {
|
|
|
|
pages[i] = page;
|
|
|
|
flush_anon_page(vma, page, start);
|
|
|
|
flush_dcache_page(page);
|
2018-10-26 22:10:28 +00:00
|
|
|
ctx.page_mask = 0;
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
|
|
|
next_page:
|
2014-06-04 23:08:13 +00:00
|
|
|
if (vmas) {
|
|
|
|
vmas[i] = vma;
|
2018-10-26 22:10:28 +00:00
|
|
|
ctx.page_mask = 0;
|
2014-06-04 23:08:13 +00:00
|
|
|
}
|
2018-10-26 22:10:28 +00:00
|
|
|
page_increm = 1 + (~(start >> PAGE_SHIFT) & ctx.page_mask);
|
2014-06-04 23:08:13 +00:00
|
|
|
if (page_increm > nr_pages)
|
|
|
|
page_increm = nr_pages;
|
|
|
|
i += page_increm;
|
|
|
|
start += page_increm * PAGE_SIZE;
|
|
|
|
nr_pages -= page_increm;
|
2014-06-04 23:08:10 +00:00
|
|
|
} while (nr_pages);
|
2018-10-26 22:10:28 +00:00
|
|
|
out:
|
|
|
|
if (ctx.pgmap)
|
|
|
|
put_dev_pagemap(ctx.pgmap);
|
|
|
|
return i ? i : ret;
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
|
|
|
|
2016-12-13 00:41:53 +00:00
|
|
|
static bool vma_permits_fault(struct vm_area_struct *vma,
|
|
|
|
unsigned int fault_flags)
|
2016-02-12 21:02:16 +00:00
|
|
|
{
|
2016-02-12 21:02:21 +00:00
|
|
|
bool write = !!(fault_flags & FAULT_FLAG_WRITE);
|
|
|
|
bool foreign = !!(fault_flags & FAULT_FLAG_REMOTE);
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
vm_flags_t vm_flags = write ? VM_WRITE : VM_READ;
|
2016-02-12 21:02:16 +00:00
|
|
|
|
|
|
|
if (!(vm_flags & vma->vm_flags))
|
|
|
|
return false;
|
|
|
|
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
/*
|
|
|
|
* The architecture might have a hardware protection
|
2016-02-12 21:02:21 +00:00
|
|
|
* mechanism other than read/write that can deny access.
|
2016-02-12 21:02:24 +00:00
|
|
|
*
|
|
|
|
* gup always represents data access, not instruction
|
|
|
|
* fetches, so execute=false here:
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
*/
|
2016-02-12 21:02:24 +00:00
|
|
|
if (!arch_vma_access_permitted(vma, write, false, foreign))
|
mm/gup, x86/mm/pkeys: Check VMAs and PTEs for protection keys
Today, for normal faults and page table walks, we check the VMA
and/or PTE to ensure that it is compatible with the action. For
instance, if we get a write fault on a non-writeable VMA, we
SIGSEGV.
We try to do the same thing for protection keys. Basically, we
try to make sure that if a user does this:
mprotect(ptr, size, PROT_NONE);
*ptr = foo;
they see the same effects with protection keys when they do this:
mprotect(ptr, size, PROT_READ|PROT_WRITE);
set_pkey(ptr, size, 4);
wrpkru(0xffffff3f); // access disable pkey 4
*ptr = foo;
The state to do that checking is in the VMA, but we also
sometimes have to do it on the page tables only, like when doing
a get_user_pages_fast() where we have no VMA.
We add two functions and expose them to generic code:
arch_pte_access_permitted(pte_flags, write)
arch_vma_access_permitted(vma, write)
These are, of course, backed up in x86 arch code with checks
against the PTE or VMA's protection key.
But, there are also cases where we do not want to respect
protection keys. When we ptrace(), for instance, we do not want
to apply the tracer's PKRU permissions to the PTEs from the
process being traced.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Boaz Harrosh <boaz@plexistor.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: David Hildenbrand <dahi@linux.vnet.ibm.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Dominik Vogt <vogt@linux.vnet.ibm.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matthew Wilcox <willy@linux.intel.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mikulas Patocka <mpatocka@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Shachar Raindel <raindel@mellanox.com>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Toshi Kani <toshi.kani@hpe.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-arch@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-s390@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Link: http://lkml.kernel.org/r/20160212210219.14D5D715@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-02-12 21:02:19 +00:00
|
|
|
return false;
|
|
|
|
|
2016-02-12 21:02:16 +00:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2020-06-02 04:48:24 +00:00
|
|
|
/**
|
2014-06-04 23:08:10 +00:00
|
|
|
* fixup_user_fault() - manually resolve a user page fault
|
|
|
|
* @mm: mm_struct of target mm
|
|
|
|
* @address: user address
|
|
|
|
* @fault_flags:flags to pass down to handle_mm_fault()
|
2020-06-09 04:33:54 +00:00
|
|
|
* @unlocked: did we unlock the mmap_lock while retrying, maybe NULL if caller
|
2020-06-02 04:48:33 +00:00
|
|
|
* does not allow retry. If NULL, the caller must guarantee
|
|
|
|
* that fault_flags does not contain FAULT_FLAG_ALLOW_RETRY.
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
|
|
|
* This is meant to be called in the specific scenario where for locking reasons
|
|
|
|
* we try to access user memory in atomic context (within a pagefault_disable()
|
|
|
|
* section), this returns -EFAULT, and we want to resolve the user fault before
|
|
|
|
* trying again.
|
|
|
|
*
|
|
|
|
* Typically this is meant to be used by the futex code.
|
|
|
|
*
|
|
|
|
* The main difference with get_user_pages() is that this function will
|
|
|
|
* unconditionally call handle_mm_fault() which will in turn perform all the
|
|
|
|
* necessary SW fixup of the dirty and young bits in the PTE, while
|
2016-01-16 00:57:04 +00:00
|
|
|
* get_user_pages() only guarantees to update these in the struct page.
|
2014-06-04 23:08:10 +00:00
|
|
|
*
|
|
|
|
* This is important for some architectures where those bits also gate the
|
|
|
|
* access permission to the page because they are maintained in software. On
|
|
|
|
* such architectures, gup() will not be enough to make a subsequent access
|
|
|
|
* succeed.
|
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* This function will not return with an unlocked mmap_lock. So it has not the
|
|
|
|
* same semantics wrt the @mm->mmap_lock as does filemap_fault().
|
2014-06-04 23:08:10 +00:00
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
int fixup_user_fault(struct mm_struct *mm,
|
2016-01-16 00:57:04 +00:00
|
|
|
unsigned long address, unsigned int fault_flags,
|
|
|
|
bool *unlocked)
|
2014-06-04 23:08:10 +00:00
|
|
|
{
|
|
|
|
struct vm_area_struct *vma;
|
2021-09-02 21:53:33 +00:00
|
|
|
vm_fault_t ret;
|
2016-01-16 00:57:04 +00:00
|
|
|
|
2023-03-12 11:26:00 +00:00
|
|
|
address = untagged_addr_remote(mm, address);
|
2019-09-25 23:48:34 +00:00
|
|
|
|
2016-01-16 00:57:04 +00:00
|
|
|
if (unlocked)
|
2020-04-02 04:08:53 +00:00
|
|
|
fault_flags |= FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2016-01-16 00:57:04 +00:00
|
|
|
retry:
|
2023-06-25 21:02:25 +00:00
|
|
|
vma = find_vma(mm, address);
|
2023-06-24 20:45:51 +00:00
|
|
|
if (!vma)
|
2014-06-04 23:08:10 +00:00
|
|
|
return -EFAULT;
|
2023-06-25 21:02:25 +00:00
|
|
|
if (address < vma->vm_start ) {
|
|
|
|
WARN_ON_ONCE(vma->vm_flags & VM_GROWSDOWN);
|
|
|
|
return -EFAULT;
|
|
|
|
}
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2016-02-12 21:02:16 +00:00
|
|
|
if (!vma_permits_fault(vma, fault_flags))
|
2014-06-04 23:08:10 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
2020-05-14 00:50:41 +00:00
|
|
|
if ((fault_flags & FAULT_FLAG_KILLABLE) &&
|
|
|
|
fatal_signal_pending(current))
|
|
|
|
return -EINTR;
|
|
|
|
|
mm: do page fault accounting in handle_mm_fault
Patch series "mm: Page fault accounting cleanups", v5.
This is v5 of the pf accounting cleanup series. It originates from Gerald
Schaefer's report on an issue a week ago regarding to incorrect page fault
accountings for retried page fault after commit 4064b9827063 ("mm: allow
VM_FAULT_RETRY for multiple times"):
https://lore.kernel.org/lkml/20200610174811.44b94525@thinkpad/
What this series did:
- Correct page fault accounting: we do accounting for a page fault
(no matter whether it's from #PF handling, or gup, or anything else)
only with the one that completed the fault. For example, page fault
retries should not be counted in page fault counters. Same to the
perf events.
- Unify definition of PERF_COUNT_SW_PAGE_FAULTS: currently this perf
event is used in an adhoc way across different archs.
Case (1): for many archs it's done at the entry of a page fault
handler, so that it will also cover e.g. errornous faults.
Case (2): for some other archs, it is only accounted when the page
fault is resolved successfully.
Case (3): there're still quite some archs that have not enabled
this perf event.
Since this series will touch merely all the archs, we unify this
perf event to always follow case (1), which is the one that makes most
sense. And since we moved the accounting into handle_mm_fault, the
other two MAJ/MIN perf events are well taken care of naturally.
- Unify definition of "major faults": the definition of "major
fault" is slightly changed when used in accounting (not
VM_FAULT_MAJOR). More information in patch 1.
- Always account the page fault onto the one that triggered the page
fault. This does not matter much for #PF handlings, but mostly for
gup. More information on this in patch 25.
Patchset layout:
Patch 1: Introduced the accounting in handle_mm_fault(), not enabled.
Patch 2-23: Enable the new accounting for arch #PF handlers one by one.
Patch 24: Enable the new accounting for the rest outliers (gup, iommu, etc.)
Patch 25: Cleanup GUP task_struct pointer since it's not needed any more
This patch (of 25):
This is a preparation patch to move page fault accountings into the
general code in handle_mm_fault(). This includes both the per task
flt_maj/flt_min counters, and the major/minor page fault perf events. To
do this, the pt_regs pointer is passed into handle_mm_fault().
PERF_COUNT_SW_PAGE_FAULTS should still be kept in per-arch page fault
handlers.
So far, all the pt_regs pointer that passed into handle_mm_fault() is
NULL, which means this patch should have no intented functional change.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Cain <bcain@codeaurora.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Chris Zankel <chris@zankel.net>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Cc: Greentime Hu <green.hu@gmail.com>
Cc: Guo Ren <guoren@kernel.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Nick Hu <nickhu@andestech.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Rich Felker <dalias@libc.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Vincent Chen <deanbo422@gmail.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: http://lkml.kernel.org/r/20200707225021.200906-1-peterx@redhat.com
Link: http://lkml.kernel.org/r/20200707225021.200906-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-08-12 01:37:44 +00:00
|
|
|
ret = handle_mm_fault(vma, address, fault_flags, NULL);
|
mm: avoid unnecessary page fault retires on shared memory types
I observed that for each of the shared file-backed page faults, we're very
likely to retry one more time for the 1st write fault upon no page. It's
because we'll need to release the mmap lock for dirty rate limit purpose
with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
Then after that throttling we return VM_FAULT_RETRY.
We did that probably because VM_FAULT_RETRY is the only way we can return
to the fault handler at that time telling it we've released the mmap lock.
However that's not ideal because it's very likely the fault does not need
to be retried at all since the pgtable was well installed before the
throttling, so the next continuous fault (including taking mmap read lock,
walk the pgtable, etc.) could be in most cases unnecessary.
It's not only slowing down page faults for shared file-backed, but also add
more mmap lock contention which is in most cases not needed at all.
To observe this, one could try to write to some shmem page and look at
"pgfault" value in /proc/vmstat, then we should expect 2 counts for each
shmem write simply because we retried, and vm event "pgfault" will capture
that.
To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
show that we've completed the whole fault and released the lock. It's also
a hint that we should very possibly not need another fault immediately on
this page because we've just completed it.
This patch provides a ~12% perf boost on my aarch64 test VM with a simple
program sequentially dirtying 400MB shmem file being mmap()ed and these are
the time it needs:
Before: 650.980 ms (+-1.94%)
After: 569.396 ms (+-1.38%)
I believe it could help more than that.
We need some special care on GUP and the s390 pgfault handler (for gmap
code before returning from pgfault), the rest changes in the page fault
handlers should be relatively straightforward.
Another thing to mention is that mm_account_fault() does take this new
fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
them as-is.
Link: https://lkml.kernel.org/r/20220530183450.42886-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vineet Gupta <vgupta@kernel.org>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> [arm part]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Stafford Horne <shorne@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Brian Cain <bcain@quicinc.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Weinberger <richard@nod.at>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Anton Ivanov <anton.ivanov@cambridgegreys.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Will Deacon <will@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Chris Zankel <chris@zankel.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Rich Felker <dalias@libc.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Helge Deller <deller@gmx.de>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-30 18:34:50 +00:00
|
|
|
|
|
|
|
if (ret & VM_FAULT_COMPLETED) {
|
|
|
|
/*
|
|
|
|
* NOTE: it's a pity that we need to retake the lock here
|
|
|
|
* to pair with the unlock() in the callers. Ideally we
|
|
|
|
* could tell the callers so they do not need to unlock.
|
|
|
|
*/
|
|
|
|
mmap_read_lock(mm);
|
|
|
|
*unlocked = true;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:10 +00:00
|
|
|
if (ret & VM_FAULT_ERROR) {
|
2017-06-02 21:46:46 +00:00
|
|
|
int err = vm_fault_to_errno(ret, 0);
|
|
|
|
|
|
|
|
if (err)
|
|
|
|
return err;
|
2014-06-04 23:08:10 +00:00
|
|
|
BUG();
|
|
|
|
}
|
2016-01-16 00:57:04 +00:00
|
|
|
|
|
|
|
if (ret & VM_FAULT_RETRY) {
|
2020-06-09 04:33:25 +00:00
|
|
|
mmap_read_lock(mm);
|
2020-05-14 00:50:41 +00:00
|
|
|
*unlocked = true;
|
|
|
|
fault_flags |= FAULT_FLAG_TRIED;
|
|
|
|
goto retry;
|
2016-01-16 00:57:04 +00:00
|
|
|
}
|
|
|
|
|
2014-06-04 23:08:10 +00:00
|
|
|
return 0;
|
|
|
|
}
|
2016-06-07 15:51:18 +00:00
|
|
|
EXPORT_SYMBOL_GPL(fixup_user_fault);
|
2014-06-04 23:08:10 +00:00
|
|
|
|
2022-10-11 19:58:06 +00:00
|
|
|
/*
|
|
|
|
* GUP always responds to fatal signals. When FOLL_INTERRUPTIBLE is
|
|
|
|
* specified, it'll also respond to generic signals. The caller of GUP
|
|
|
|
* that has FOLL_INTERRUPTIBLE should take care of the GUP interruption.
|
|
|
|
*/
|
|
|
|
static bool gup_signal_pending(unsigned int flags)
|
|
|
|
{
|
|
|
|
if (fatal_signal_pending(current))
|
|
|
|
return true;
|
|
|
|
|
|
|
|
if (!(flags & FOLL_INTERRUPTIBLE))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
return signal_pending(current);
|
|
|
|
}
|
|
|
|
|
2020-06-03 23:03:25 +00:00
|
|
|
/*
|
2023-01-24 20:34:22 +00:00
|
|
|
* Locking: (*locked == 1) means that the mmap_lock has already been acquired by
|
|
|
|
* the caller. This function may drop the mmap_lock. If it does so, then it will
|
|
|
|
* set (*locked = 0).
|
|
|
|
*
|
|
|
|
* (*locked == 0) means that the caller expects this function to acquire and
|
|
|
|
* drop the mmap_lock. Therefore, the value of *locked will still be zero when
|
|
|
|
* the function returns, even though it may have changed temporarily during
|
|
|
|
* function execution.
|
|
|
|
*
|
|
|
|
* Please note that this function, unlike __get_user_pages(), will not return 0
|
|
|
|
* for nr_pages > 0, unless FOLL_NOWAIT is used.
|
2020-06-03 23:03:25 +00:00
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
static __always_inline long __get_user_pages_locked(struct mm_struct *mm,
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
unsigned long start,
|
|
|
|
unsigned long nr_pages,
|
|
|
|
struct page **pages,
|
|
|
|
struct vm_area_struct **vmas,
|
2017-11-19 16:32:05 +00:00
|
|
|
int *locked,
|
2015-02-11 23:27:20 +00:00
|
|
|
unsigned int flags)
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
{
|
|
|
|
long ret, pages_done;
|
2023-01-24 20:34:22 +00:00
|
|
|
bool must_unlock = false;
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
|
2023-01-24 20:34:22 +00:00
|
|
|
/*
|
|
|
|
* The internal caller expects GUP to manage the lock internally and the
|
|
|
|
* lock must be released when this returns.
|
|
|
|
*/
|
2023-01-24 20:34:30 +00:00
|
|
|
if (!*locked) {
|
2023-01-24 20:34:22 +00:00
|
|
|
if (mmap_read_lock_killable(mm))
|
|
|
|
return -EAGAIN;
|
|
|
|
must_unlock = true;
|
|
|
|
*locked = 1;
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
}
|
2023-01-24 20:34:27 +00:00
|
|
|
else
|
|
|
|
mmap_assert_locked(mm);
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
|
2021-06-29 02:36:40 +00:00
|
|
|
if (flags & FOLL_PIN)
|
|
|
|
mm_set_has_pinned_flag(&mm->flags);
|
2020-09-25 22:25:57 +00:00
|
|
|
|
2020-01-31 06:12:54 +00:00
|
|
|
/*
|
|
|
|
* FOLL_PIN and FOLL_GET are mutually exclusive. Traditional behavior
|
|
|
|
* is to set FOLL_GET if the caller wants pages[] filled in (but has
|
|
|
|
* carelessly failed to specify FOLL_GET), so keep doing that, but only
|
|
|
|
* for FOLL_GET, not for the newer FOLL_PIN.
|
|
|
|
*
|
|
|
|
* FOLL_PIN always expects pages to be non-null, but no need to assert
|
|
|
|
* that here, as any failures will be obvious enough.
|
|
|
|
*/
|
|
|
|
if (pages && !(flags & FOLL_PIN))
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
flags |= FOLL_GET;
|
|
|
|
|
|
|
|
pages_done = 0;
|
|
|
|
for (;;) {
|
2020-08-12 01:39:01 +00:00
|
|
|
ret = __get_user_pages(mm, start, nr_pages, flags, pages,
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
vmas, locked);
|
2023-01-24 20:34:29 +00:00
|
|
|
if (!(flags & FOLL_UNLOCKABLE)) {
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
/* VM_FAULT_RETRY couldn't trigger, bypass */
|
2023-01-24 20:34:29 +00:00
|
|
|
pages_done = ret;
|
|
|
|
break;
|
|
|
|
}
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
|
mm: avoid unnecessary page fault retires on shared memory types
I observed that for each of the shared file-backed page faults, we're very
likely to retry one more time for the 1st write fault upon no page. It's
because we'll need to release the mmap lock for dirty rate limit purpose
with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
Then after that throttling we return VM_FAULT_RETRY.
We did that probably because VM_FAULT_RETRY is the only way we can return
to the fault handler at that time telling it we've released the mmap lock.
However that's not ideal because it's very likely the fault does not need
to be retried at all since the pgtable was well installed before the
throttling, so the next continuous fault (including taking mmap read lock,
walk the pgtable, etc.) could be in most cases unnecessary.
It's not only slowing down page faults for shared file-backed, but also add
more mmap lock contention which is in most cases not needed at all.
To observe this, one could try to write to some shmem page and look at
"pgfault" value in /proc/vmstat, then we should expect 2 counts for each
shmem write simply because we retried, and vm event "pgfault" will capture
that.
To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
show that we've completed the whole fault and released the lock. It's also
a hint that we should very possibly not need another fault immediately on
this page because we've just completed it.
This patch provides a ~12% perf boost on my aarch64 test VM with a simple
program sequentially dirtying 400MB shmem file being mmap()ed and these are
the time it needs:
Before: 650.980 ms (+-1.94%)
After: 569.396 ms (+-1.38%)
I believe it could help more than that.
We need some special care on GUP and the s390 pgfault handler (for gmap
code before returning from pgfault), the rest changes in the page fault
handlers should be relatively straightforward.
Another thing to mention is that mm_account_fault() does take this new
fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
them as-is.
Link: https://lkml.kernel.org/r/20220530183450.42886-1-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Vineet Gupta <vgupta@kernel.org>
Acked-by: Guo Ren <guoren@kernel.org>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> [arm part]
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Stafford Horne <shorne@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: Brian Cain <bcain@quicinc.com>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Weinberger <richard@nod.at>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Janosch Frank <frankja@linux.ibm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Anton Ivanov <anton.ivanov@cambridgegreys.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Will Deacon <will@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Chris Zankel <chris@zankel.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Rich Felker <dalias@libc.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Helge Deller <deller@gmx.de>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-30 18:34:50 +00:00
|
|
|
/* VM_FAULT_RETRY or VM_FAULT_COMPLETED cannot return errors */
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
if (!*locked) {
|
|
|
|
BUG_ON(ret < 0);
|
|
|
|
BUG_ON(ret >= nr_pages);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ret > 0) {
|
|
|
|
nr_pages -= ret;
|
|
|
|
pages_done += ret;
|
|
|
|
if (!nr_pages)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (*locked) {
|
2018-03-09 23:51:06 +00:00
|
|
|
/*
|
|
|
|
* VM_FAULT_RETRY didn't trigger or it was a
|
|
|
|
* FOLL_NOWAIT.
|
|
|
|
*/
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
if (!pages_done)
|
|
|
|
pages_done = ret;
|
|
|
|
break;
|
|
|
|
}
|
2019-06-01 05:30:33 +00:00
|
|
|
/*
|
|
|
|
* VM_FAULT_RETRY triggered, so seek to the faulting offset.
|
|
|
|
* For the prefault case (!pages) we only update counts.
|
|
|
|
*/
|
|
|
|
if (likely(pages))
|
|
|
|
pages += ret;
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
start += ret << PAGE_SHIFT;
|
2023-01-24 20:34:22 +00:00
|
|
|
|
|
|
|
/* The lock was temporarily dropped, so we must unlock later */
|
|
|
|
must_unlock = true;
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
|
2020-04-02 04:08:49 +00:00
|
|
|
retry:
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
/*
|
|
|
|
* Repeat on the address that fired VM_FAULT_RETRY
|
2020-04-02 04:08:49 +00:00
|
|
|
* with both FAULT_FLAG_ALLOW_RETRY and
|
|
|
|
* FAULT_FLAG_TRIED. Note that GUP can be interrupted
|
2022-10-11 19:58:06 +00:00
|
|
|
* by fatal signals of even common signals, depending on
|
|
|
|
* the caller's request. So we need to check it before we
|
2020-04-02 04:08:49 +00:00
|
|
|
* start trying again otherwise it can loop forever.
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
*/
|
2022-10-11 19:58:06 +00:00
|
|
|
if (gup_signal_pending(flags)) {
|
2020-04-08 15:59:24 +00:00
|
|
|
if (!pages_done)
|
|
|
|
pages_done = -EINTR;
|
2020-04-02 04:08:49 +00:00
|
|
|
break;
|
2020-04-08 15:59:24 +00:00
|
|
|
}
|
2020-04-02 04:08:49 +00:00
|
|
|
|
2020-06-09 04:33:25 +00:00
|
|
|
ret = mmap_read_lock_killable(mm);
|
2020-04-02 04:08:53 +00:00
|
|
|
if (ret) {
|
|
|
|
BUG_ON(ret > 0);
|
|
|
|
if (!pages_done)
|
|
|
|
pages_done = ret;
|
|
|
|
break;
|
|
|
|
}
|
2020-04-02 04:08:49 +00:00
|
|
|
|
2020-04-08 01:40:10 +00:00
|
|
|
*locked = 1;
|
2020-08-12 01:39:01 +00:00
|
|
|
ret = __get_user_pages(mm, start, 1, flags | FOLL_TRIED,
|
2020-04-02 04:08:49 +00:00
|
|
|
pages, NULL, locked);
|
|
|
|
if (!*locked) {
|
|
|
|
/* Continue to retry until we succeeded */
|
|
|
|
BUG_ON(ret != 0);
|
|
|
|
goto retry;
|
|
|
|
}
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
if (ret != 1) {
|
|
|
|
BUG_ON(ret > 1);
|
|
|
|
if (!pages_done)
|
|
|
|
pages_done = ret;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
nr_pages--;
|
|
|
|
pages_done++;
|
|
|
|
if (!nr_pages)
|
|
|
|
break;
|
2019-06-01 05:30:33 +00:00
|
|
|
if (likely(pages))
|
|
|
|
pages++;
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
start += PAGE_SIZE;
|
|
|
|
}
|
2023-01-24 20:34:22 +00:00
|
|
|
if (must_unlock && *locked) {
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
/*
|
2023-01-24 20:34:22 +00:00
|
|
|
* We either temporarily dropped the lock, or the caller
|
|
|
|
* requested that we both acquire and drop the lock. Either way,
|
|
|
|
* we must now unlock, and notify the caller of that state.
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
*/
|
2020-06-09 04:33:25 +00:00
|
|
|
mmap_read_unlock(mm);
|
mm: gup: add get_user_pages_locked and get_user_pages_unlocked
FAULT_FOLL_ALLOW_RETRY allows the page fault to drop the mmap_sem for
reading to reduce the mmap_sem contention (for writing), like while
waiting for I/O completion. The problem is that right now practically no
get_user_pages call uses FAULT_FOLL_ALLOW_RETRY, so we're not leveraging
that nifty feature.
Andres fixed it for the KVM page fault. However get_user_pages_fast
remains uncovered, and 99% of other get_user_pages aren't using it either
(the only exception being FOLL_NOWAIT in KVM which is really nonblocking
and in fact it doesn't even release the mmap_sem).
So this patchsets extends the optimization Andres did in the KVM page
fault to the whole kernel. It makes most important places (including
gup_fast) to use FAULT_FOLL_ALLOW_RETRY to reduce the mmap_sem hold times
during I/O.
The only few places that remains uncovered are drivers like v4l and other
exceptions that tends to work on their own memory and they're not working
on random user memory (for example like O_DIRECT that uses gup_fast and is
fully covered by this patch).
A follow up patch should probably also add a printk_once warning to
get_user_pages that should go obsolete and be phased out eventually. The
"vmas" parameter of get_user_pages makes it fundamentally incompatible
with FAULT_FOLL_ALLOW_RETRY (vmas array becomes meaningless the moment the
mmap_sem is released).
While this is just an optimization, this becomes an absolute requirement
for the userfaultfd feature http://lwn.net/Articles/615086/ .
The userfaultfd allows to block the page fault, and in order to do so I
need to drop the mmap_sem first. So this patch also ensures that all
memory where userfaultfd could be registered by KVM, the very first fault
(no matter if it is a regular page fault, or a get_user_pages) always has
FAULT_FOLL_ALLOW_RETRY set. Then the userfaultfd blocks and it is waken
only when the pagetable is already mapped. The second fault attempt after
the wakeup doesn't need FAULT_FOLL_ALLOW_RETRY, so it's ok to retry
without it.
This patch (of 5):
We can leverage the VM_FAULT_RETRY functionality in the page fault paths
better by using either get_user_pages_locked or get_user_pages_unlocked.
The former allows conversion of get_user_pages invocations that will have
to pass a "&locked" parameter to know if the mmap_sem was dropped during
the call. Example from:
down_read(&mm->mmap_sem);
do_something()
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
to:
int locked = 1;
down_read(&mm->mmap_sem);
do_something()
get_user_pages_locked(tsk, mm, ..., pages, &locked);
if (locked)
up_read(&mm->mmap_sem);
The latter is suitable only as a drop in replacement of the form:
down_read(&mm->mmap_sem);
get_user_pages(tsk, mm, ..., pages, NULL);
up_read(&mm->mmap_sem);
into:
get_user_pages_unlocked(tsk, mm, ..., pages);
Where tsk, mm, the intermediate "..." paramters and "pages" can be any
value as before. Just the last parameter of get_user_pages (vmas) must be
NULL for get_user_pages_locked|unlocked to be usable (the latter original
form wouldn't have been safe anyway if vmas wasn't null, for the former we
just make it explicit by dropping the parameter).
If vmas is not NULL these two methods cannot be used.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Andres Lagar-Cavilla <andreslc@google.com>
Reviewed-by: Peter Feiner <pfeiner@google.com>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-11 23:27:17 +00:00
|
|
|
*locked = 0;
|
|
|
|
}
|
|
|
|
return pages_done;
|
|
|
|
}
|
|
|
|
|
2019-07-12 03:57:18 +00:00
|
|
|
/**
|
|
|
|
* populate_vma_page_range() - populate a range of pages in the vma.
|
|
|
|
* @vma: target vma
|
|
|
|
* @start: start address
|
|
|
|
* @end: end address
|
2020-06-09 04:33:54 +00:00
|
|
|
* @locked: whether the mmap_lock is still held
|
2019-07-12 03:57:18 +00:00
|
|
|
*
|
|
|
|
* This takes care of mlocking the pages too if VM_LOCKED is set.
|
|
|
|
*
|
2020-08-07 06:20:01 +00:00
|
|
|
* Return either number of pages pinned in the vma, or a negative error
|
|
|
|
* code on error.
|
2019-07-12 03:57:18 +00:00
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* vma->vm_mm->mmap_lock must be held.
|
2019-07-12 03:57:18 +00:00
|
|
|
*
|
mm/gup: rename "nonblocking" to "locked" where proper
Patch series "mm: Page fault enhancements", v6.
This series contains cleanups and enhancements to current page fault
logic. The whole idea comes from the discussion between Andrea and Linus
on the bug reported by syzbot here:
https://lkml.org/lkml/2017/11/2/833
Basically it does two things:
(a) Allows the page fault logic to be more interactive on not only
SIGKILL, but also the rest of userspace signals, and,
(b) Allows the page fault retry (VM_FAULT_RETRY) to happen for more
than once.
For (a): with the changes we should be able to react faster when page
faults are working in parallel with userspace signals like SIGSTOP and
SIGCONT (and more), and with that we can remove the buggy part in
userfaultfd and benefit the whole page fault mechanism on faster signal
processing to reach the userspace.
For (b), we should be able to allow the page fault handler to loop for
even more than twice. Some context: for now since we have
FAULT_FLAG_ALLOW_RETRY we can allow to retry the page fault once with the
same interrupt context, however never more than twice. This can be not
only a potential cleanup to remove this assumption since AFAIU the code
itself doesn't really have this twice-only limitation (though that should
be a protective approach in the past), at the same time it'll greatly
simplify future works like userfaultfd write-protect where it's possible
to retry for more than twice (please have a look at [1] below for a
possible user that might require the page fault to be handled for a third
time; if we can remove the retry limitation we can simply drop that patch
and those complexity).
This patch (of 16):
There's plenty of places around __get_user_pages() that has a parameter
"nonblocking" which does not really mean that "it won't block" (because it
can really block) but instead it shows whether the mmap_sem is released by
up_read() during the page fault handling mostly when VM_FAULT_RETRY is
returned.
We have the correct naming in e.g. get_user_pages_locked() or
get_user_pages_remote() as "locked", however there're still many places
that are using the "nonblocking" as name.
Renaming the places to "locked" where proper to better suite the
functionality of the variable. While at it, fixing up some of the
comments accordingly.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Brian Geffon <bgeffon@google.com>
Reviewed-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Martin Cracauer <cracauer@cons.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Bobby Powers <bobbypowers@gmail.com>
Cc: Maya Gokhale <gokhale2@llnl.gov>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Marty McFadden <mcfadden8@llnl.gov>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Link: http://lkml.kernel.org/r/20200220155353.8676-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:07:58 +00:00
|
|
|
* If @locked is NULL, it may be held for read or write and will
|
2019-07-12 03:57:18 +00:00
|
|
|
* be unperturbed.
|
|
|
|
*
|
mm/gup: rename "nonblocking" to "locked" where proper
Patch series "mm: Page fault enhancements", v6.
This series contains cleanups and enhancements to current page fault
logic. The whole idea comes from the discussion between Andrea and Linus
on the bug reported by syzbot here:
https://lkml.org/lkml/2017/11/2/833
Basically it does two things:
(a) Allows the page fault logic to be more interactive on not only
SIGKILL, but also the rest of userspace signals, and,
(b) Allows the page fault retry (VM_FAULT_RETRY) to happen for more
than once.
For (a): with the changes we should be able to react faster when page
faults are working in parallel with userspace signals like SIGSTOP and
SIGCONT (and more), and with that we can remove the buggy part in
userfaultfd and benefit the whole page fault mechanism on faster signal
processing to reach the userspace.
For (b), we should be able to allow the page fault handler to loop for
even more than twice. Some context: for now since we have
FAULT_FLAG_ALLOW_RETRY we can allow to retry the page fault once with the
same interrupt context, however never more than twice. This can be not
only a potential cleanup to remove this assumption since AFAIU the code
itself doesn't really have this twice-only limitation (though that should
be a protective approach in the past), at the same time it'll greatly
simplify future works like userfaultfd write-protect where it's possible
to retry for more than twice (please have a look at [1] below for a
possible user that might require the page fault to be handled for a third
time; if we can remove the retry limitation we can simply drop that patch
and those complexity).
This patch (of 16):
There's plenty of places around __get_user_pages() that has a parameter
"nonblocking" which does not really mean that "it won't block" (because it
can really block) but instead it shows whether the mmap_sem is released by
up_read() during the page fault handling mostly when VM_FAULT_RETRY is
returned.
We have the correct naming in e.g. get_user_pages_locked() or
get_user_pages_remote() as "locked", however there're still many places
that are using the "nonblocking" as name.
Renaming the places to "locked" where proper to better suite the
functionality of the variable. While at it, fixing up some of the
comments accordingly.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Brian Geffon <bgeffon@google.com>
Reviewed-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Martin Cracauer <cracauer@cons.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Bobby Powers <bobbypowers@gmail.com>
Cc: Maya Gokhale <gokhale2@llnl.gov>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Marty McFadden <mcfadden8@llnl.gov>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Link: http://lkml.kernel.org/r/20200220155353.8676-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:07:58 +00:00
|
|
|
* If @locked is non-NULL, it must held for read only and may be
|
|
|
|
* released. If it's released, *@locked will be set to 0.
|
2019-07-12 03:57:18 +00:00
|
|
|
*/
|
|
|
|
long populate_vma_page_range(struct vm_area_struct *vma,
|
mm/gup: rename "nonblocking" to "locked" where proper
Patch series "mm: Page fault enhancements", v6.
This series contains cleanups and enhancements to current page fault
logic. The whole idea comes from the discussion between Andrea and Linus
on the bug reported by syzbot here:
https://lkml.org/lkml/2017/11/2/833
Basically it does two things:
(a) Allows the page fault logic to be more interactive on not only
SIGKILL, but also the rest of userspace signals, and,
(b) Allows the page fault retry (VM_FAULT_RETRY) to happen for more
than once.
For (a): with the changes we should be able to react faster when page
faults are working in parallel with userspace signals like SIGSTOP and
SIGCONT (and more), and with that we can remove the buggy part in
userfaultfd and benefit the whole page fault mechanism on faster signal
processing to reach the userspace.
For (b), we should be able to allow the page fault handler to loop for
even more than twice. Some context: for now since we have
FAULT_FLAG_ALLOW_RETRY we can allow to retry the page fault once with the
same interrupt context, however never more than twice. This can be not
only a potential cleanup to remove this assumption since AFAIU the code
itself doesn't really have this twice-only limitation (though that should
be a protective approach in the past), at the same time it'll greatly
simplify future works like userfaultfd write-protect where it's possible
to retry for more than twice (please have a look at [1] below for a
possible user that might require the page fault to be handled for a third
time; if we can remove the retry limitation we can simply drop that patch
and those complexity).
This patch (of 16):
There's plenty of places around __get_user_pages() that has a parameter
"nonblocking" which does not really mean that "it won't block" (because it
can really block) but instead it shows whether the mmap_sem is released by
up_read() during the page fault handling mostly when VM_FAULT_RETRY is
returned.
We have the correct naming in e.g. get_user_pages_locked() or
get_user_pages_remote() as "locked", however there're still many places
that are using the "nonblocking" as name.
Renaming the places to "locked" where proper to better suite the
functionality of the variable. While at it, fixing up some of the
comments accordingly.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Brian Geffon <bgeffon@google.com>
Reviewed-by: Mike Rapoport <rppt@linux.vnet.ibm.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Martin Cracauer <cracauer@cons.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Bobby Powers <bobbypowers@gmail.com>
Cc: Maya Gokhale <gokhale2@llnl.gov>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Marty McFadden <mcfadden8@llnl.gov>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Link: http://lkml.kernel.org/r/20200220155353.8676-2-peterx@redhat.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:07:58 +00:00
|
|
|
unsigned long start, unsigned long end, int *locked)
|
2019-07-12 03:57:18 +00:00
|
|
|
{
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
unsigned long nr_pages = (end - start) / PAGE_SIZE;
|
2023-01-24 20:34:30 +00:00
|
|
|
int local_locked = 1;
|
2019-07-12 03:57:18 +00:00
|
|
|
int gup_flags;
|
2022-04-01 18:28:27 +00:00
|
|
|
long ret;
|
2019-07-12 03:57:18 +00:00
|
|
|
|
2021-09-02 21:53:45 +00:00
|
|
|
VM_BUG_ON(!PAGE_ALIGNED(start));
|
|
|
|
VM_BUG_ON(!PAGE_ALIGNED(end));
|
2019-07-12 03:57:18 +00:00
|
|
|
VM_BUG_ON_VMA(start < vma->vm_start, vma);
|
|
|
|
VM_BUG_ON_VMA(end > vma->vm_end, vma);
|
2020-06-09 04:33:44 +00:00
|
|
|
mmap_assert_locked(mm);
|
2019-07-12 03:57:18 +00:00
|
|
|
|
2022-02-15 02:21:52 +00:00
|
|
|
/*
|
|
|
|
* Rightly or wrongly, the VM_LOCKONFAULT case has never used
|
|
|
|
* faultin_page() to break COW, so it has no work to do here.
|
|
|
|
*/
|
2019-07-12 03:57:18 +00:00
|
|
|
if (vma->vm_flags & VM_LOCKONFAULT)
|
2022-02-15 02:21:52 +00:00
|
|
|
return nr_pages;
|
|
|
|
|
|
|
|
gup_flags = FOLL_TOUCH;
|
2019-07-12 03:57:18 +00:00
|
|
|
/*
|
|
|
|
* We want to touch writable mappings with a write fault in order
|
|
|
|
* to break COW, except for shared mappings because these don't COW
|
|
|
|
* and we would not want to dirty them for nothing.
|
|
|
|
*/
|
|
|
|
if ((vma->vm_flags & (VM_WRITE | VM_SHARED)) == VM_WRITE)
|
|
|
|
gup_flags |= FOLL_WRITE;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We want mlock to succeed for regions that have any permissions
|
|
|
|
* other than PROT_NONE.
|
|
|
|
*/
|
2020-04-07 03:03:47 +00:00
|
|
|
if (vma_is_accessible(vma))
|
2019-07-12 03:57:18 +00:00
|
|
|
gup_flags |= FOLL_FORCE;
|
|
|
|
|
2023-01-24 20:34:29 +00:00
|
|
|
if (locked)
|
|
|
|
gup_flags |= FOLL_UNLOCKABLE;
|
|
|
|
|
2019-07-12 03:57:18 +00:00
|
|
|
/*
|
|
|
|
* We made sure addr is within a VMA, so the following will
|
|
|
|
* not result in a stack expansion that recurses back here.
|
|
|
|
*/
|
2022-04-01 18:28:27 +00:00
|
|
|
ret = __get_user_pages(mm, start, nr_pages, gup_flags,
|
2023-01-24 20:34:30 +00:00
|
|
|
NULL, NULL, locked ? locked : &local_locked);
|
2022-04-01 18:28:27 +00:00
|
|
|
lru_add_drain();
|
|
|
|
return ret;
|
2019-07-12 03:57:18 +00:00
|
|
|
}
|
|
|
|
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
/*
|
|
|
|
* faultin_vma_page_range() - populate (prefault) page tables inside the
|
|
|
|
* given VMA range readable/writable
|
|
|
|
*
|
|
|
|
* This takes care of mlocking the pages, too, if VM_LOCKED is set.
|
|
|
|
*
|
|
|
|
* @vma: target vma
|
|
|
|
* @start: start address
|
|
|
|
* @end: end address
|
|
|
|
* @write: whether to prefault readable or writable
|
|
|
|
* @locked: whether the mmap_lock is still held
|
|
|
|
*
|
|
|
|
* Returns either number of processed pages in the vma, or a negative error
|
|
|
|
* code on error (see __get_user_pages()).
|
|
|
|
*
|
|
|
|
* vma->vm_mm->mmap_lock must be held. The range must be page-aligned and
|
2023-01-24 20:34:28 +00:00
|
|
|
* covered by the VMA. If it's released, *@locked will be set to 0.
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
*/
|
|
|
|
long faultin_vma_page_range(struct vm_area_struct *vma, unsigned long start,
|
|
|
|
unsigned long end, bool write, int *locked)
|
|
|
|
{
|
|
|
|
struct mm_struct *mm = vma->vm_mm;
|
|
|
|
unsigned long nr_pages = (end - start) / PAGE_SIZE;
|
|
|
|
int gup_flags;
|
2022-04-01 18:28:27 +00:00
|
|
|
long ret;
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
|
|
|
|
VM_BUG_ON(!PAGE_ALIGNED(start));
|
|
|
|
VM_BUG_ON(!PAGE_ALIGNED(end));
|
|
|
|
VM_BUG_ON_VMA(start < vma->vm_start, vma);
|
|
|
|
VM_BUG_ON_VMA(end > vma->vm_end, vma);
|
|
|
|
mmap_assert_locked(mm);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* FOLL_TOUCH: Mark page accessed and thereby young; will also mark
|
|
|
|
* the page dirty with FOLL_WRITE -- which doesn't make a
|
|
|
|
* difference with !FOLL_FORCE, because the page is writable
|
|
|
|
* in the page table.
|
|
|
|
* FOLL_HWPOISON: Return -EHWPOISON instead of -EFAULT when we hit
|
|
|
|
* a poisoned page.
|
|
|
|
* !FOLL_FORCE: Require proper access permissions.
|
|
|
|
*/
|
2023-01-24 20:34:29 +00:00
|
|
|
gup_flags = FOLL_TOUCH | FOLL_HWPOISON | FOLL_UNLOCKABLE;
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
if (write)
|
|
|
|
gup_flags |= FOLL_WRITE;
|
|
|
|
|
|
|
|
/*
|
2021-08-13 23:54:37 +00:00
|
|
|
* We want to report -EINVAL instead of -EFAULT for any permission
|
|
|
|
* problems or incompatible mappings.
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
*/
|
2021-08-13 23:54:37 +00:00
|
|
|
if (check_vma_flags(vma, gup_flags))
|
|
|
|
return -EINVAL;
|
|
|
|
|
2022-04-01 18:28:27 +00:00
|
|
|
ret = __get_user_pages(mm, start, nr_pages, gup_flags,
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
NULL, NULL, locked);
|
2022-04-01 18:28:27 +00:00
|
|
|
lru_add_drain();
|
|
|
|
return ret;
|
mm/madvise: introduce MADV_POPULATE_(READ|WRITE) to prefault page tables
I. Background: Sparse Memory Mappings
When we manage sparse memory mappings dynamically in user space - also
sometimes involving MAP_NORESERVE - we want to dynamically populate/
discard memory inside such a sparse memory region. Example users are
hypervisors (especially implementing memory ballooning or similar
technologies like virtio-mem) and memory allocators. In addition, we want
to fail in a nice way (instead of generating SIGBUS) if populating does
not succeed because we are out of backend memory (which can happen easily
with file-based mappings, especially tmpfs and hugetlbfs).
While MADV_DONTNEED, MADV_REMOVE and FALLOC_FL_PUNCH_HOLE allow for
reliably discarding memory for most mapping types, there is no generic
approach to populate page tables and preallocate memory.
Although mmap() supports MAP_POPULATE, it is not applicable to the concept
of sparse memory mappings, where we want to populate/discard dynamically
and avoid expensive/problematic remappings. In addition, we never
actually report errors during the final populate phase - it is best-effort
only.
fallocate() can be used to preallocate file-based memory and fail in a
safe way. However, it cannot really be used for any private mappings on
anonymous files via memfd due to COW semantics. In addition, fallocate()
does not actually populate page tables, so we still always get pagefaults
on first access - which is sometimes undesired (i.e., real-time workloads)
and requires real prefaulting of page tables, not just a preallocation of
backend storage. There might be interesting use cases for sparse memory
regions along with mlockall(MCL_ONFAULT) which fallocate() cannot satisfy
as it does not prefault page tables.
II. On preallcoation/prefaulting from user space
Because we don't have a proper interface, what applications (like QEMU and
databases) end up doing is touching (i.e., reading+writing one byte to not
overwrite existing data) all individual pages.
However, that approach
1) Can result in wear on storage backing, because we end up reading/writing
each page; this is especially a problem for dax/pmem.
2) Can result in mmap_sem contention when prefaulting via multiple
threads.
3) Requires expensive signal handling, especially to catch SIGBUS in case
of hugetlbfs/shmem/file-backed memory. For example, this is
problematic in hypervisors like QEMU where SIGBUS handlers might already
be used by other subsystems concurrently to e.g, handle hardware errors.
"Simply" doing preallocation concurrently from other thread is not that
easy.
III. On MADV_WILLNEED
Extending MADV_WILLNEED is not an option because
1. It would change the semantics: "Expect access in the near future." and
"might be a good idea to read some pages" vs. "Definitely populate/
preallocate all memory and definitely fail on errors.".
2. Existing users (like virtio-balloon in QEMU when deflating the balloon)
don't want populate/prealloc semantics. They treat this rather as a hint
to give a little performance boost without too much overhead - and don't
expect that a lot of memory might get consumed or a lot of time
might be spent.
IV. MADV_POPULATE_READ and MADV_POPULATE_WRITE
Let's introduce MADV_POPULATE_READ and MADV_POPULATE_WRITE, inspired by
MAP_POPULATE, with the following semantics:
1. MADV_POPULATE_READ can be used to prefault page tables just like
manually reading each individual page. This will not break any COW
mappings. The shared zero page might get mapped and no backend storage
might get preallocated -- allocation might be deferred to
write-fault time. Especially shared file mappings require an explicit
fallocate() upfront to actually preallocate backend memory (blocks in
the file system) in case the file might have holes.
2. If MADV_POPULATE_READ succeeds, all page tables have been populated
(prefaulted) readable once.
3. MADV_POPULATE_WRITE can be used to preallocate backend memory and
prefault page tables just like manually writing (or
reading+writing) each individual page. This will break any COW
mappings -- e.g., the shared zeropage is never populated.
4. If MADV_POPULATE_WRITE succeeds, all page tables have been populated
(prefaulted) writable once.
5. MADV_POPULATE_READ and MADV_POPULATE_WRITE cannot be applied to special
mappings marked with VM_PFNMAP and VM_IO. Also, proper access
permissions (e.g., PROT_READ, PROT_WRITE) are required. If any such
mapping is encountered, madvise() fails with -EINVAL.
6. If MADV_POPULATE_READ or MADV_POPULATE_WRITE fails, some page tables
might have been populated.
7. MADV_POPULATE_READ and MADV_POPULATE_WRITE will return -EHWPOISON
when encountering a HW poisoned page in the range.
8. Similar to MAP_POPULATE, MADV_POPULATE_READ and MADV_POPULATE_WRITE
cannot protect from the OOM (Out Of Memory) handler killing the
process.
While the use case for MADV_POPULATE_WRITE is fairly obvious (i.e.,
preallocate memory and prefault page tables for VMs), one issue is that
whenever we prefault pages writable, the pages have to be marked dirty,
because the CPU could dirty them any time. while not a real problem for
hugetlbfs or dax/pmem, it can be a problem for shared file mappings: each
page will be marked dirty and has to be written back later when evicting.
MADV_POPULATE_READ allows for optimizing this scenario: Pre-read a whole
mapping from backend storage without marking it dirty, such that eviction
won't have to write it back. As discussed above, shared file mappings
might require an explciit fallocate() upfront to achieve
preallcoation+prepopulation.
Although sparse memory mappings are the primary use case, this will also
be useful for other preallocate/prefault use cases where MAP_POPULATE is
not desired or the semantics of MAP_POPULATE are not sufficient: as one
example, QEMU users can trigger preallocation/prefaulting of guest RAM
after the mapping was created -- and don't want errors to be silently
suppressed.
Looking at the history, MADV_POPULATE was already proposed in 2013 [1],
however, the main motivation back than was performance improvements --
which should also still be the case.
V. Single-threaded performance comparison
I did a short experiment, prefaulting page tables on completely *empty
mappings/files* and repeated the experiment 10 times. The results
correspond to the shortest execution time. In general, the performance
benefit for huge pages is negligible with small mappings.
V.1: Private mappings
POPULATE_READ and POPULATE_WRITE is fastest. Note that
Reading/POPULATE_READ will populate the shared zeropage where applicable
-- which result in short population times.
The fastest way to allocate backend storage (here: swap or huge pages) and
prefault page tables is POPULATE_WRITE.
V.2: Shared mappings
fallocate() is fastest, however, doesn't prefault page tables.
POPULATE_WRITE is faster than simple writes and read/writes.
POPULATE_READ is faster than simple reads.
Without a fd, the fastest way to allocate backend storage and prefault
page tables is POPULATE_WRITE. With an fd, the fastest way is usually
FALLOCATE+POPULATE_READ or FALLOCATE+POPULATE_WRITE respectively; one
exception are actual files: FALLOCATE+Read is slightly faster than
FALLOCATE+POPULATE_READ.
The fastest way to allocate backend storage prefault page tables is
FALLOCATE+POPULATE_WRITE -- except when dealing with actual files; then,
FALLOCATE+POPULATE_READ is fastest and won't directly mark all pages as
dirty.
v.3: Detailed results
==================================================
2 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 0.119 ms
Anon 4 KiB : Write : 0.222 ms
Anon 4 KiB : Read/Write : 0.380 ms
Anon 4 KiB : POPULATE_READ : 0.060 ms
Anon 4 KiB : POPULATE_WRITE : 0.158 ms
Memfd 4 KiB : Read : 0.034 ms
Memfd 4 KiB : Write : 0.310 ms
Memfd 4 KiB : Read/Write : 0.362 ms
Memfd 4 KiB : POPULATE_READ : 0.039 ms
Memfd 4 KiB : POPULATE_WRITE : 0.229 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.033 ms
tmpfs : Write : 0.313 ms
tmpfs : Read/Write : 0.406 ms
tmpfs : POPULATE_READ : 0.039 ms
tmpfs : POPULATE_WRITE : 0.285 ms
file : Read : 0.033 ms
file : Write : 0.351 ms
file : Read/Write : 0.408 ms
file : POPULATE_READ : 0.039 ms
file : POPULATE_WRITE : 0.290 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_PRIVATE:
**************************************************
Anon 4 KiB : Read : 237.940 ms
Anon 4 KiB : Write : 708.409 ms
Anon 4 KiB : Read/Write : 1054.041 ms
Anon 4 KiB : POPULATE_READ : 124.310 ms
Anon 4 KiB : POPULATE_WRITE : 572.582 ms
Memfd 4 KiB : Read : 136.928 ms
Memfd 4 KiB : Write : 963.898 ms
Memfd 4 KiB : Read/Write : 1106.561 ms
Memfd 4 KiB : POPULATE_READ : 78.450 ms
Memfd 4 KiB : POPULATE_WRITE : 805.881 ms
Memfd 2 MiB : Read : 357.116 ms
Memfd 2 MiB : Write : 357.210 ms
Memfd 2 MiB : Read/Write : 357.606 ms
Memfd 2 MiB : POPULATE_READ : 356.094 ms
Memfd 2 MiB : POPULATE_WRITE : 356.937 ms
tmpfs : Read : 137.536 ms
tmpfs : Write : 954.362 ms
tmpfs : Read/Write : 1105.954 ms
tmpfs : POPULATE_READ : 80.289 ms
tmpfs : POPULATE_WRITE : 822.826 ms
file : Read : 137.874 ms
file : Write : 987.025 ms
file : Read/Write : 1107.439 ms
file : POPULATE_READ : 80.413 ms
file : POPULATE_WRITE : 857.622 ms
hugetlbfs : Read : 355.607 ms
hugetlbfs : Write : 355.729 ms
hugetlbfs : Read/Write : 356.127 ms
hugetlbfs : POPULATE_READ : 354.585 ms
hugetlbfs : POPULATE_WRITE : 355.138 ms
**************************************************
2 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 0.394 ms
Anon 4 KiB : Write : 0.348 ms
Anon 4 KiB : Read/Write : 0.400 ms
Anon 4 KiB : POPULATE_READ : 0.326 ms
Anon 4 KiB : POPULATE_WRITE : 0.273 ms
Anon 2 MiB : Read : 0.030 ms
Anon 2 MiB : Write : 0.030 ms
Anon 2 MiB : Read/Write : 0.030 ms
Anon 2 MiB : POPULATE_READ : 0.030 ms
Anon 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 4 KiB : Read : 0.412 ms
Memfd 4 KiB : Write : 0.372 ms
Memfd 4 KiB : Read/Write : 0.419 ms
Memfd 4 KiB : POPULATE_READ : 0.343 ms
Memfd 4 KiB : POPULATE_WRITE : 0.288 ms
Memfd 4 KiB : FALLOCATE : 0.137 ms
Memfd 4 KiB : FALLOCATE+Read : 0.446 ms
Memfd 4 KiB : FALLOCATE+Write : 0.330 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 0.454 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 0.379 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 0.268 ms
Memfd 2 MiB : Read : 0.030 ms
Memfd 2 MiB : Write : 0.030 ms
Memfd 2 MiB : Read/Write : 0.030 ms
Memfd 2 MiB : POPULATE_READ : 0.030 ms
Memfd 2 MiB : POPULATE_WRITE : 0.030 ms
Memfd 2 MiB : FALLOCATE : 0.030 ms
Memfd 2 MiB : FALLOCATE+Read : 0.031 ms
Memfd 2 MiB : FALLOCATE+Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 0.031 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 0.030 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 0.030 ms
tmpfs : Read : 0.416 ms
tmpfs : Write : 0.369 ms
tmpfs : Read/Write : 0.425 ms
tmpfs : POPULATE_READ : 0.346 ms
tmpfs : POPULATE_WRITE : 0.295 ms
tmpfs : FALLOCATE : 0.139 ms
tmpfs : FALLOCATE+Read : 0.447 ms
tmpfs : FALLOCATE+Write : 0.333 ms
tmpfs : FALLOCATE+Read/Write : 0.454 ms
tmpfs : FALLOCATE+POPULATE_READ : 0.380 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 0.272 ms
file : Read : 0.191 ms
file : Write : 0.511 ms
file : Read/Write : 0.524 ms
file : POPULATE_READ : 0.196 ms
file : POPULATE_WRITE : 0.434 ms
file : FALLOCATE : 0.004 ms
file : FALLOCATE+Read : 0.197 ms
file : FALLOCATE+Write : 0.554 ms
file : FALLOCATE+Read/Write : 0.480 ms
file : FALLOCATE+POPULATE_READ : 0.201 ms
file : FALLOCATE+POPULATE_WRITE : 0.381 ms
hugetlbfs : Read : 0.030 ms
hugetlbfs : Write : 0.030 ms
hugetlbfs : Read/Write : 0.030 ms
hugetlbfs : POPULATE_READ : 0.030 ms
hugetlbfs : POPULATE_WRITE : 0.030 ms
hugetlbfs : FALLOCATE : 0.030 ms
hugetlbfs : FALLOCATE+Read : 0.031 ms
hugetlbfs : FALLOCATE+Write : 0.031 ms
hugetlbfs : FALLOCATE+Read/Write : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 0.030 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 0.030 ms
**************************************************
4096 MiB MAP_SHARED:
**************************************************
Anon 4 KiB : Read : 1053.090 ms
Anon 4 KiB : Write : 913.642 ms
Anon 4 KiB : Read/Write : 1060.350 ms
Anon 4 KiB : POPULATE_READ : 893.691 ms
Anon 4 KiB : POPULATE_WRITE : 782.885 ms
Anon 2 MiB : Read : 358.553 ms
Anon 2 MiB : Write : 358.419 ms
Anon 2 MiB : Read/Write : 357.992 ms
Anon 2 MiB : POPULATE_READ : 357.533 ms
Anon 2 MiB : POPULATE_WRITE : 357.808 ms
Memfd 4 KiB : Read : 1078.144 ms
Memfd 4 KiB : Write : 942.036 ms
Memfd 4 KiB : Read/Write : 1100.391 ms
Memfd 4 KiB : POPULATE_READ : 925.829 ms
Memfd 4 KiB : POPULATE_WRITE : 804.394 ms
Memfd 4 KiB : FALLOCATE : 304.632 ms
Memfd 4 KiB : FALLOCATE+Read : 1163.359 ms
Memfd 4 KiB : FALLOCATE+Write : 933.186 ms
Memfd 4 KiB : FALLOCATE+Read/Write : 1187.304 ms
Memfd 4 KiB : FALLOCATE+POPULATE_READ : 1013.660 ms
Memfd 4 KiB : FALLOCATE+POPULATE_WRITE : 794.560 ms
Memfd 2 MiB : Read : 358.131 ms
Memfd 2 MiB : Write : 358.099 ms
Memfd 2 MiB : Read/Write : 358.250 ms
Memfd 2 MiB : POPULATE_READ : 357.563 ms
Memfd 2 MiB : POPULATE_WRITE : 357.334 ms
Memfd 2 MiB : FALLOCATE : 356.735 ms
Memfd 2 MiB : FALLOCATE+Read : 358.152 ms
Memfd 2 MiB : FALLOCATE+Write : 358.331 ms
Memfd 2 MiB : FALLOCATE+Read/Write : 358.018 ms
Memfd 2 MiB : FALLOCATE+POPULATE_READ : 357.286 ms
Memfd 2 MiB : FALLOCATE+POPULATE_WRITE : 357.523 ms
tmpfs : Read : 1087.265 ms
tmpfs : Write : 950.840 ms
tmpfs : Read/Write : 1107.567 ms
tmpfs : POPULATE_READ : 922.605 ms
tmpfs : POPULATE_WRITE : 810.094 ms
tmpfs : FALLOCATE : 306.320 ms
tmpfs : FALLOCATE+Read : 1169.796 ms
tmpfs : FALLOCATE+Write : 933.730 ms
tmpfs : FALLOCATE+Read/Write : 1191.610 ms
tmpfs : FALLOCATE+POPULATE_READ : 1020.474 ms
tmpfs : FALLOCATE+POPULATE_WRITE : 798.945 ms
file : Read : 654.101 ms
file : Write : 1259.142 ms
file : Read/Write : 1289.509 ms
file : POPULATE_READ : 661.642 ms
file : POPULATE_WRITE : 1106.816 ms
file : FALLOCATE : 1.864 ms
file : FALLOCATE+Read : 656.328 ms
file : FALLOCATE+Write : 1153.300 ms
file : FALLOCATE+Read/Write : 1180.613 ms
file : FALLOCATE+POPULATE_READ : 668.347 ms
file : FALLOCATE+POPULATE_WRITE : 996.143 ms
hugetlbfs : Read : 357.245 ms
hugetlbfs : Write : 357.413 ms
hugetlbfs : Read/Write : 357.120 ms
hugetlbfs : POPULATE_READ : 356.321 ms
hugetlbfs : POPULATE_WRITE : 356.693 ms
hugetlbfs : FALLOCATE : 355.927 ms
hugetlbfs : FALLOCATE+Read : 357.074 ms
hugetlbfs : FALLOCATE+Write : 357.120 ms
hugetlbfs : FALLOCATE+Read/Write : 356.983 ms
hugetlbfs : FALLOCATE+POPULATE_READ : 356.413 ms
hugetlbfs : FALLOCATE+POPULATE_WRITE : 356.266 ms
**************************************************
[1] https://lkml.org/lkml/2013/6/27/698
[akpm@linux-foundation.org: coding style fixes]
Link: https://lkml.kernel.org/r/20210419135443.12822-3-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rolf Eike Beer <eike-kernel@sf-tec.de>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-01 01:52:28 +00:00
|
|
|
}
|
|
|
|
|
2019-07-12 03:57:18 +00:00
|
|
|
/*
|
|
|
|
* __mm_populate - populate and/or mlock pages within a range of address space.
|
|
|
|
*
|
|
|
|
* This is used to implement mlock() and the MAP_POPULATE / MAP_LOCKED mmap
|
|
|
|
* flags. VMAs must be already marked with the desired vm_flags, and
|
2020-06-09 04:33:54 +00:00
|
|
|
* mmap_lock must not be held.
|
2019-07-12 03:57:18 +00:00
|
|
|
*/
|
|
|
|
int __mm_populate(unsigned long start, unsigned long len, int ignore_errors)
|
|
|
|
{
|
|
|
|
struct mm_struct *mm = current->mm;
|
|
|
|
unsigned long end, nstart, nend;
|
|
|
|
struct vm_area_struct *vma = NULL;
|
|
|
|
int locked = 0;
|
|
|
|
long ret = 0;
|
|
|
|
|
|
|
|
end = start + len;
|
|
|
|
|
|
|
|
for (nstart = start; nstart < end; nstart = nend) {
|
|
|
|
/*
|
|
|
|
* We want to fault in pages for [nstart; end) address range.
|
|
|
|
* Find first corresponding VMA.
|
|
|
|
*/
|
|
|
|
if (!locked) {
|
|
|
|
locked = 1;
|
2020-06-09 04:33:25 +00:00
|
|
|
mmap_read_lock(mm);
|
2022-09-06 19:49:00 +00:00
|
|
|
vma = find_vma_intersection(mm, nstart, end);
|
2019-07-12 03:57:18 +00:00
|
|
|
} else if (nstart >= vma->vm_end)
|
2022-09-06 19:49:00 +00:00
|
|
|
vma = find_vma_intersection(mm, vma->vm_end, end);
|
|
|
|
|
|
|
|
if (!vma)
|
2019-07-12 03:57:18 +00:00
|
|
|
break;
|
|
|
|
/*
|
|
|
|
* Set [nstart; nend) to intersection of desired address
|
|
|
|
* range with the first VMA. Also, skip undesirable VMA types.
|
|
|
|
*/
|
|
|
|
nend = min(end, vma->vm_end);
|
|
|
|
if (vma->vm_flags & (VM_IO | VM_PFNMAP))
|
|
|
|
continue;
|
|
|
|
if (nstart < vma->vm_start)
|
|
|
|
nstart = vma->vm_start;
|
|
|
|
/*
|
|
|
|
* Now fault in a range of pages. populate_vma_page_range()
|
|
|
|
* double checks the vma flags, so that it won't mlock pages
|
|
|
|
* if the vma was already munlocked.
|
|
|
|
*/
|
|
|
|
ret = populate_vma_page_range(vma, nstart, nend, &locked);
|
|
|
|
if (ret < 0) {
|
|
|
|
if (ignore_errors) {
|
|
|
|
ret = 0;
|
|
|
|
continue; /* continue at next VMA */
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
nend = nstart + ret * PAGE_SIZE;
|
|
|
|
ret = 0;
|
|
|
|
}
|
|
|
|
if (locked)
|
2020-06-09 04:33:25 +00:00
|
|
|
mmap_read_unlock(mm);
|
2019-07-12 03:57:18 +00:00
|
|
|
return ret; /* 0 or negative error code */
|
|
|
|
}
|
2019-07-12 03:57:21 +00:00
|
|
|
#else /* CONFIG_MMU */
|
2020-08-12 01:39:01 +00:00
|
|
|
static long __get_user_pages_locked(struct mm_struct *mm, unsigned long start,
|
2019-07-12 03:57:21 +00:00
|
|
|
unsigned long nr_pages, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas, int *locked,
|
|
|
|
unsigned int foll_flags)
|
|
|
|
{
|
|
|
|
struct vm_area_struct *vma;
|
2023-01-24 20:34:22 +00:00
|
|
|
bool must_unlock = false;
|
2019-07-12 03:57:21 +00:00
|
|
|
unsigned long vm_flags;
|
2021-05-05 01:39:15 +00:00
|
|
|
long i;
|
2019-07-12 03:57:21 +00:00
|
|
|
|
2023-01-24 20:34:22 +00:00
|
|
|
if (!nr_pages)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The internal caller expects GUP to manage the lock internally and the
|
|
|
|
* lock must be released when this returns.
|
|
|
|
*/
|
2023-01-24 20:34:30 +00:00
|
|
|
if (!*locked) {
|
2023-01-24 20:34:22 +00:00
|
|
|
if (mmap_read_lock_killable(mm))
|
|
|
|
return -EAGAIN;
|
|
|
|
must_unlock = true;
|
|
|
|
*locked = 1;
|
|
|
|
}
|
|
|
|
|
2019-07-12 03:57:21 +00:00
|
|
|
/* calculate required read or write permissions.
|
|
|
|
* If FOLL_FORCE is set, we only require the "MAY" flags.
|
|
|
|
*/
|
|
|
|
vm_flags = (foll_flags & FOLL_WRITE) ?
|
|
|
|
(VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD);
|
|
|
|
vm_flags &= (foll_flags & FOLL_FORCE) ?
|
|
|
|
(VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE);
|
|
|
|
|
|
|
|
for (i = 0; i < nr_pages; i++) {
|
|
|
|
vma = find_vma(mm, start);
|
|
|
|
if (!vma)
|
2023-01-24 20:34:22 +00:00
|
|
|
break;
|
2019-07-12 03:57:21 +00:00
|
|
|
|
|
|
|
/* protect what we can, including chardevs */
|
|
|
|
if ((vma->vm_flags & (VM_IO | VM_PFNMAP)) ||
|
|
|
|
!(vm_flags & vma->vm_flags))
|
2023-01-24 20:34:22 +00:00
|
|
|
break;
|
2019-07-12 03:57:21 +00:00
|
|
|
|
|
|
|
if (pages) {
|
2022-06-30 08:41:23 +00:00
|
|
|
pages[i] = virt_to_page((void *)start);
|
2019-07-12 03:57:21 +00:00
|
|
|
if (pages[i])
|
|
|
|
get_page(pages[i]);
|
|
|
|
}
|
|
|
|
if (vmas)
|
|
|
|
vmas[i] = vma;
|
|
|
|
start = (start + PAGE_SIZE) & PAGE_MASK;
|
|
|
|
}
|
|
|
|
|
2023-01-24 20:34:22 +00:00
|
|
|
if (must_unlock && *locked) {
|
|
|
|
mmap_read_unlock(mm);
|
|
|
|
*locked = 0;
|
|
|
|
}
|
2019-07-12 03:57:21 +00:00
|
|
|
|
|
|
|
return i ? : -EFAULT;
|
|
|
|
}
|
|
|
|
#endif /* !CONFIG_MMU */
|
2019-07-12 03:57:18 +00:00
|
|
|
|
2021-08-02 11:44:20 +00:00
|
|
|
/**
|
|
|
|
* fault_in_writeable - fault in userspace address range for writing
|
|
|
|
* @uaddr: start of address range
|
|
|
|
* @size: size of address range
|
|
|
|
*
|
|
|
|
* Returns the number of bytes not faulted in (like copy_to_user() and
|
|
|
|
* copy_from_user()).
|
|
|
|
*/
|
|
|
|
size_t fault_in_writeable(char __user *uaddr, size_t size)
|
|
|
|
{
|
|
|
|
char __user *start = uaddr, *end;
|
|
|
|
|
|
|
|
if (unlikely(size == 0))
|
|
|
|
return 0;
|
2022-01-14 22:05:13 +00:00
|
|
|
if (!user_write_access_begin(uaddr, size))
|
|
|
|
return size;
|
2021-08-02 11:44:20 +00:00
|
|
|
if (!PAGE_ALIGNED(uaddr)) {
|
2022-01-14 22:05:13 +00:00
|
|
|
unsafe_put_user(0, uaddr, out);
|
2021-08-02 11:44:20 +00:00
|
|
|
uaddr = (char __user *)PAGE_ALIGN((unsigned long)uaddr);
|
|
|
|
}
|
|
|
|
end = (char __user *)PAGE_ALIGN((unsigned long)start + size);
|
|
|
|
if (unlikely(end < start))
|
|
|
|
end = NULL;
|
|
|
|
while (uaddr != end) {
|
2022-01-14 22:05:13 +00:00
|
|
|
unsafe_put_user(0, uaddr, out);
|
2021-08-02 11:44:20 +00:00
|
|
|
uaddr += PAGE_SIZE;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
2022-01-14 22:05:13 +00:00
|
|
|
user_write_access_end();
|
2021-08-02 11:44:20 +00:00
|
|
|
if (size > uaddr - start)
|
|
|
|
return size - (uaddr - start);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(fault_in_writeable);
|
|
|
|
|
2022-04-23 10:07:49 +00:00
|
|
|
/**
|
|
|
|
* fault_in_subpage_writeable - fault in an address range for writing
|
|
|
|
* @uaddr: start of address range
|
|
|
|
* @size: size of address range
|
|
|
|
*
|
|
|
|
* Fault in a user address range for writing while checking for permissions at
|
|
|
|
* sub-page granularity (e.g. arm64 MTE). This function should be used when
|
|
|
|
* the caller cannot guarantee forward progress of a copy_to_user() loop.
|
|
|
|
*
|
|
|
|
* Returns the number of bytes not faulted in (like copy_to_user() and
|
|
|
|
* copy_from_user()).
|
|
|
|
*/
|
|
|
|
size_t fault_in_subpage_writeable(char __user *uaddr, size_t size)
|
|
|
|
{
|
|
|
|
size_t faulted_in;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Attempt faulting in at page granularity first for page table
|
|
|
|
* permission checking. The arch-specific probe_subpage_writeable()
|
|
|
|
* functions may not check for this.
|
|
|
|
*/
|
|
|
|
faulted_in = size - fault_in_writeable(uaddr, size);
|
|
|
|
if (faulted_in)
|
|
|
|
faulted_in -= probe_subpage_writeable(uaddr, faulted_in);
|
|
|
|
|
|
|
|
return size - faulted_in;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(fault_in_subpage_writeable);
|
|
|
|
|
2021-07-05 15:26:28 +00:00
|
|
|
/*
|
|
|
|
* fault_in_safe_writeable - fault in an address range for writing
|
|
|
|
* @uaddr: start of address range
|
|
|
|
* @size: length of address range
|
|
|
|
*
|
2022-03-08 19:55:48 +00:00
|
|
|
* Faults in an address range for writing. This is primarily useful when we
|
|
|
|
* already know that some or all of the pages in the address range aren't in
|
|
|
|
* memory.
|
2021-07-05 15:26:28 +00:00
|
|
|
*
|
2022-03-08 19:55:48 +00:00
|
|
|
* Unlike fault_in_writeable(), this function is non-destructive.
|
2021-07-05 15:26:28 +00:00
|
|
|
*
|
|
|
|
* Note that we don't pin or otherwise hold the pages referenced that we fault
|
|
|
|
* in. There's no guarantee that they'll stay in memory for any duration of
|
|
|
|
* time.
|
|
|
|
*
|
|
|
|
* Returns the number of bytes not faulted in, like copy_to_user() and
|
|
|
|
* copy_from_user().
|
|
|
|
*/
|
|
|
|
size_t fault_in_safe_writeable(const char __user *uaddr, size_t size)
|
|
|
|
{
|
2022-03-08 19:55:48 +00:00
|
|
|
unsigned long start = (unsigned long)uaddr, end;
|
2021-07-05 15:26:28 +00:00
|
|
|
struct mm_struct *mm = current->mm;
|
2022-03-08 19:55:48 +00:00
|
|
|
bool unlocked = false;
|
2021-07-05 15:26:28 +00:00
|
|
|
|
2022-03-08 19:55:48 +00:00
|
|
|
if (unlikely(size == 0))
|
|
|
|
return 0;
|
2021-07-05 15:26:28 +00:00
|
|
|
end = PAGE_ALIGN(start + size);
|
2022-03-08 19:55:48 +00:00
|
|
|
if (end < start)
|
2021-07-05 15:26:28 +00:00
|
|
|
end = 0;
|
|
|
|
|
2022-03-08 19:55:48 +00:00
|
|
|
mmap_read_lock(mm);
|
|
|
|
do {
|
|
|
|
if (fixup_user_fault(mm, start, FAULT_FLAG_WRITE, &unlocked))
|
2021-07-05 15:26:28 +00:00
|
|
|
break;
|
2022-03-08 19:55:48 +00:00
|
|
|
start = (start + PAGE_SIZE) & PAGE_MASK;
|
|
|
|
} while (start != end);
|
|
|
|
mmap_read_unlock(mm);
|
|
|
|
|
|
|
|
if (size > (unsigned long)uaddr - start)
|
|
|
|
return size - ((unsigned long)uaddr - start);
|
|
|
|
return 0;
|
2021-07-05 15:26:28 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(fault_in_safe_writeable);
|
|
|
|
|
2021-08-02 11:44:20 +00:00
|
|
|
/**
|
|
|
|
* fault_in_readable - fault in userspace address range for reading
|
|
|
|
* @uaddr: start of user address range
|
|
|
|
* @size: size of user address range
|
|
|
|
*
|
|
|
|
* Returns the number of bytes not faulted in (like copy_to_user() and
|
|
|
|
* copy_from_user()).
|
|
|
|
*/
|
|
|
|
size_t fault_in_readable(const char __user *uaddr, size_t size)
|
|
|
|
{
|
|
|
|
const char __user *start = uaddr, *end;
|
|
|
|
volatile char c;
|
|
|
|
|
|
|
|
if (unlikely(size == 0))
|
|
|
|
return 0;
|
2022-01-14 22:05:13 +00:00
|
|
|
if (!user_read_access_begin(uaddr, size))
|
|
|
|
return size;
|
2021-08-02 11:44:20 +00:00
|
|
|
if (!PAGE_ALIGNED(uaddr)) {
|
2022-01-14 22:05:13 +00:00
|
|
|
unsafe_get_user(c, uaddr, out);
|
2021-08-02 11:44:20 +00:00
|
|
|
uaddr = (const char __user *)PAGE_ALIGN((unsigned long)uaddr);
|
|
|
|
}
|
|
|
|
end = (const char __user *)PAGE_ALIGN((unsigned long)start + size);
|
|
|
|
if (unlikely(end < start))
|
|
|
|
end = NULL;
|
|
|
|
while (uaddr != end) {
|
2022-01-14 22:05:13 +00:00
|
|
|
unsafe_get_user(c, uaddr, out);
|
2021-08-02 11:44:20 +00:00
|
|
|
uaddr += PAGE_SIZE;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
2022-01-14 22:05:13 +00:00
|
|
|
user_read_access_end();
|
2021-08-02 11:44:20 +00:00
|
|
|
(void)c;
|
|
|
|
if (size > uaddr - start)
|
|
|
|
return size - (uaddr - start);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(fault_in_readable);
|
|
|
|
|
binfmt_elf_fdpic: stop using dump_emit() on user pointers on !MMU
Patch series "Fix ELF / FDPIC ELF core dumping, and use mmap_lock properly in there", v5.
At the moment, we have that rather ugly mmget_still_valid() helper to work
around <https://crbug.com/project-zero/1790>: ELF core dumping doesn't
take the mmap_sem while traversing the task's VMAs, and if anything (like
userfaultfd) then remotely messes with the VMA tree, fireworks ensue. So
at the moment we use mmget_still_valid() to bail out in any writers that
might be operating on a remote mm's VMAs.
With this series, I'm trying to get rid of the need for that as cleanly as
possible. ("cleanly" meaning "avoid holding the mmap_lock across
unbounded sleeps".)
Patches 1, 2, 3 and 4 are relatively unrelated cleanups in the core
dumping code.
Patches 5 and 6 implement the main change: Instead of repeatedly accessing
the VMA list with sleeps in between, we snapshot it at the start with
proper locking, and then later we just use our copy of the VMA list. This
ensures that the kernel won't crash, that VMA metadata in the coredump is
consistent even in the presence of concurrent modifications, and that any
virtual addresses that aren't being concurrently modified have their
contents show up in the core dump properly.
The disadvantage of this approach is that we need a bit more memory during
core dumping for storing metadata about all VMAs.
At the end of the series, patch 7 removes the old workaround for this
issue (mmget_still_valid()).
I have tested:
- Creating a simple core dump on X86-64 still works.
- The created coredump on X86-64 opens in GDB and looks plausible.
- X86-64 core dumps contain the first page for executable mappings at
offset 0, and don't contain the first page for non-executable file
mappings or executable mappings at offset !=0.
- NOMMU 32-bit ARM can still generate plausible-looking core dumps
through the FDPIC implementation. (I can't test this with GDB because
GDB is missing some structure definition for nommu ARM, but I've
poked around in the hexdump and it looked decent.)
This patch (of 7):
dump_emit() is for kernel pointers, and VMAs describe userspace memory.
Let's be tidy here and avoid accessing userspace pointers under KERNEL_DS,
even if it probably doesn't matter much on !MMU systems - especially given
that it looks like we can just use the same get_dump_page() as on MMU if
we move it out of the CONFIG_MMU block.
One small change we have to make in get_dump_page() is to use
__get_user_pages_locked() instead of __get_user_pages(), since the latter
doesn't exist on nommu. On mmu builds, __get_user_pages_locked() will
just call __get_user_pages() for us.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Link: http://lkml.kernel.org/r/20200827114932.3572699-1-jannh@google.com
Link: http://lkml.kernel.org/r/20200827114932.3572699-2-jannh@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-10-16 03:12:40 +00:00
|
|
|
/**
|
|
|
|
* get_dump_page() - pin user page in memory while writing it to core dump
|
|
|
|
* @addr: user address
|
|
|
|
*
|
|
|
|
* Returns struct page pointer of user page pinned for dump,
|
|
|
|
* to be freed afterwards by put_page().
|
|
|
|
*
|
|
|
|
* Returns NULL on any kind of failure - a hole must then be inserted into
|
|
|
|
* the corefile, to preserve alignment with its headers; and also returns
|
|
|
|
* NULL wherever the ZERO_PAGE, or an anonymous pte_none, has been found -
|
2021-05-07 01:06:47 +00:00
|
|
|
* allowing a hole to be left in the corefile to save disk space.
|
binfmt_elf_fdpic: stop using dump_emit() on user pointers on !MMU
Patch series "Fix ELF / FDPIC ELF core dumping, and use mmap_lock properly in there", v5.
At the moment, we have that rather ugly mmget_still_valid() helper to work
around <https://crbug.com/project-zero/1790>: ELF core dumping doesn't
take the mmap_sem while traversing the task's VMAs, and if anything (like
userfaultfd) then remotely messes with the VMA tree, fireworks ensue. So
at the moment we use mmget_still_valid() to bail out in any writers that
might be operating on a remote mm's VMAs.
With this series, I'm trying to get rid of the need for that as cleanly as
possible. ("cleanly" meaning "avoid holding the mmap_lock across
unbounded sleeps".)
Patches 1, 2, 3 and 4 are relatively unrelated cleanups in the core
dumping code.
Patches 5 and 6 implement the main change: Instead of repeatedly accessing
the VMA list with sleeps in between, we snapshot it at the start with
proper locking, and then later we just use our copy of the VMA list. This
ensures that the kernel won't crash, that VMA metadata in the coredump is
consistent even in the presence of concurrent modifications, and that any
virtual addresses that aren't being concurrently modified have their
contents show up in the core dump properly.
The disadvantage of this approach is that we need a bit more memory during
core dumping for storing metadata about all VMAs.
At the end of the series, patch 7 removes the old workaround for this
issue (mmget_still_valid()).
I have tested:
- Creating a simple core dump on X86-64 still works.
- The created coredump on X86-64 opens in GDB and looks plausible.
- X86-64 core dumps contain the first page for executable mappings at
offset 0, and don't contain the first page for non-executable file
mappings or executable mappings at offset !=0.
- NOMMU 32-bit ARM can still generate plausible-looking core dumps
through the FDPIC implementation. (I can't test this with GDB because
GDB is missing some structure definition for nommu ARM, but I've
poked around in the hexdump and it looked decent.)
This patch (of 7):
dump_emit() is for kernel pointers, and VMAs describe userspace memory.
Let's be tidy here and avoid accessing userspace pointers under KERNEL_DS,
even if it probably doesn't matter much on !MMU systems - especially given
that it looks like we can just use the same get_dump_page() as on MMU if
we move it out of the CONFIG_MMU block.
One small change we have to make in get_dump_page() is to use
__get_user_pages_locked() instead of __get_user_pages(), since the latter
doesn't exist on nommu. On mmu builds, __get_user_pages_locked() will
just call __get_user_pages() for us.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Link: http://lkml.kernel.org/r/20200827114932.3572699-1-jannh@google.com
Link: http://lkml.kernel.org/r/20200827114932.3572699-2-jannh@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-10-16 03:12:40 +00:00
|
|
|
*
|
2020-10-16 03:12:57 +00:00
|
|
|
* Called without mmap_lock (takes and releases the mmap_lock by itself).
|
binfmt_elf_fdpic: stop using dump_emit() on user pointers on !MMU
Patch series "Fix ELF / FDPIC ELF core dumping, and use mmap_lock properly in there", v5.
At the moment, we have that rather ugly mmget_still_valid() helper to work
around <https://crbug.com/project-zero/1790>: ELF core dumping doesn't
take the mmap_sem while traversing the task's VMAs, and if anything (like
userfaultfd) then remotely messes with the VMA tree, fireworks ensue. So
at the moment we use mmget_still_valid() to bail out in any writers that
might be operating on a remote mm's VMAs.
With this series, I'm trying to get rid of the need for that as cleanly as
possible. ("cleanly" meaning "avoid holding the mmap_lock across
unbounded sleeps".)
Patches 1, 2, 3 and 4 are relatively unrelated cleanups in the core
dumping code.
Patches 5 and 6 implement the main change: Instead of repeatedly accessing
the VMA list with sleeps in between, we snapshot it at the start with
proper locking, and then later we just use our copy of the VMA list. This
ensures that the kernel won't crash, that VMA metadata in the coredump is
consistent even in the presence of concurrent modifications, and that any
virtual addresses that aren't being concurrently modified have their
contents show up in the core dump properly.
The disadvantage of this approach is that we need a bit more memory during
core dumping for storing metadata about all VMAs.
At the end of the series, patch 7 removes the old workaround for this
issue (mmget_still_valid()).
I have tested:
- Creating a simple core dump on X86-64 still works.
- The created coredump on X86-64 opens in GDB and looks plausible.
- X86-64 core dumps contain the first page for executable mappings at
offset 0, and don't contain the first page for non-executable file
mappings or executable mappings at offset !=0.
- NOMMU 32-bit ARM can still generate plausible-looking core dumps
through the FDPIC implementation. (I can't test this with GDB because
GDB is missing some structure definition for nommu ARM, but I've
poked around in the hexdump and it looked decent.)
This patch (of 7):
dump_emit() is for kernel pointers, and VMAs describe userspace memory.
Let's be tidy here and avoid accessing userspace pointers under KERNEL_DS,
even if it probably doesn't matter much on !MMU systems - especially given
that it looks like we can just use the same get_dump_page() as on MMU if
we move it out of the CONFIG_MMU block.
One small change we have to make in get_dump_page() is to use
__get_user_pages_locked() instead of __get_user_pages(), since the latter
doesn't exist on nommu. On mmu builds, __get_user_pages_locked() will
just call __get_user_pages() for us.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Link: http://lkml.kernel.org/r/20200827114932.3572699-1-jannh@google.com
Link: http://lkml.kernel.org/r/20200827114932.3572699-2-jannh@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-10-16 03:12:40 +00:00
|
|
|
*/
|
|
|
|
#ifdef CONFIG_ELF_CORE
|
|
|
|
struct page *get_dump_page(unsigned long addr)
|
|
|
|
{
|
|
|
|
struct page *page;
|
2023-01-24 20:34:22 +00:00
|
|
|
int locked = 0;
|
2020-10-16 03:12:57 +00:00
|
|
|
int ret;
|
binfmt_elf_fdpic: stop using dump_emit() on user pointers on !MMU
Patch series "Fix ELF / FDPIC ELF core dumping, and use mmap_lock properly in there", v5.
At the moment, we have that rather ugly mmget_still_valid() helper to work
around <https://crbug.com/project-zero/1790>: ELF core dumping doesn't
take the mmap_sem while traversing the task's VMAs, and if anything (like
userfaultfd) then remotely messes with the VMA tree, fireworks ensue. So
at the moment we use mmget_still_valid() to bail out in any writers that
might be operating on a remote mm's VMAs.
With this series, I'm trying to get rid of the need for that as cleanly as
possible. ("cleanly" meaning "avoid holding the mmap_lock across
unbounded sleeps".)
Patches 1, 2, 3 and 4 are relatively unrelated cleanups in the core
dumping code.
Patches 5 and 6 implement the main change: Instead of repeatedly accessing
the VMA list with sleeps in between, we snapshot it at the start with
proper locking, and then later we just use our copy of the VMA list. This
ensures that the kernel won't crash, that VMA metadata in the coredump is
consistent even in the presence of concurrent modifications, and that any
virtual addresses that aren't being concurrently modified have their
contents show up in the core dump properly.
The disadvantage of this approach is that we need a bit more memory during
core dumping for storing metadata about all VMAs.
At the end of the series, patch 7 removes the old workaround for this
issue (mmget_still_valid()).
I have tested:
- Creating a simple core dump on X86-64 still works.
- The created coredump on X86-64 opens in GDB and looks plausible.
- X86-64 core dumps contain the first page for executable mappings at
offset 0, and don't contain the first page for non-executable file
mappings or executable mappings at offset !=0.
- NOMMU 32-bit ARM can still generate plausible-looking core dumps
through the FDPIC implementation. (I can't test this with GDB because
GDB is missing some structure definition for nommu ARM, but I've
poked around in the hexdump and it looked decent.)
This patch (of 7):
dump_emit() is for kernel pointers, and VMAs describe userspace memory.
Let's be tidy here and avoid accessing userspace pointers under KERNEL_DS,
even if it probably doesn't matter much on !MMU systems - especially given
that it looks like we can just use the same get_dump_page() as on MMU if
we move it out of the CONFIG_MMU block.
One small change we have to make in get_dump_page() is to use
__get_user_pages_locked() instead of __get_user_pages(), since the latter
doesn't exist on nommu. On mmu builds, __get_user_pages_locked() will
just call __get_user_pages() for us.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Link: http://lkml.kernel.org/r/20200827114932.3572699-1-jannh@google.com
Link: http://lkml.kernel.org/r/20200827114932.3572699-2-jannh@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-10-16 03:12:40 +00:00
|
|
|
|
2023-01-24 20:34:22 +00:00
|
|
|
ret = __get_user_pages_locked(current->mm, addr, 1, &page, NULL,
|
|
|
|
&locked,
|
2020-10-16 03:12:57 +00:00
|
|
|
FOLL_FORCE | FOLL_DUMP | FOLL_GET);
|
|
|
|
return (ret == 1) ? page : NULL;
|
binfmt_elf_fdpic: stop using dump_emit() on user pointers on !MMU
Patch series "Fix ELF / FDPIC ELF core dumping, and use mmap_lock properly in there", v5.
At the moment, we have that rather ugly mmget_still_valid() helper to work
around <https://crbug.com/project-zero/1790>: ELF core dumping doesn't
take the mmap_sem while traversing the task's VMAs, and if anything (like
userfaultfd) then remotely messes with the VMA tree, fireworks ensue. So
at the moment we use mmget_still_valid() to bail out in any writers that
might be operating on a remote mm's VMAs.
With this series, I'm trying to get rid of the need for that as cleanly as
possible. ("cleanly" meaning "avoid holding the mmap_lock across
unbounded sleeps".)
Patches 1, 2, 3 and 4 are relatively unrelated cleanups in the core
dumping code.
Patches 5 and 6 implement the main change: Instead of repeatedly accessing
the VMA list with sleeps in between, we snapshot it at the start with
proper locking, and then later we just use our copy of the VMA list. This
ensures that the kernel won't crash, that VMA metadata in the coredump is
consistent even in the presence of concurrent modifications, and that any
virtual addresses that aren't being concurrently modified have their
contents show up in the core dump properly.
The disadvantage of this approach is that we need a bit more memory during
core dumping for storing metadata about all VMAs.
At the end of the series, patch 7 removes the old workaround for this
issue (mmget_still_valid()).
I have tested:
- Creating a simple core dump on X86-64 still works.
- The created coredump on X86-64 opens in GDB and looks plausible.
- X86-64 core dumps contain the first page for executable mappings at
offset 0, and don't contain the first page for non-executable file
mappings or executable mappings at offset !=0.
- NOMMU 32-bit ARM can still generate plausible-looking core dumps
through the FDPIC implementation. (I can't test this with GDB because
GDB is missing some structure definition for nommu ARM, but I've
poked around in the hexdump and it looked decent.)
This patch (of 7):
dump_emit() is for kernel pointers, and VMAs describe userspace memory.
Let's be tidy here and avoid accessing userspace pointers under KERNEL_DS,
even if it probably doesn't matter much on !MMU systems - especially given
that it looks like we can just use the same get_dump_page() as on MMU if
we move it out of the CONFIG_MMU block.
One small change we have to make in get_dump_page() is to use
__get_user_pages_locked() instead of __get_user_pages(), since the latter
doesn't exist on nommu. On mmu builds, __get_user_pages_locked() will
just call __get_user_pages() for us.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: "Eric W . Biederman" <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Link: http://lkml.kernel.org/r/20200827114932.3572699-1-jannh@google.com
Link: http://lkml.kernel.org/r/20200827114932.3572699-2-jannh@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-10-16 03:12:40 +00:00
|
|
|
}
|
|
|
|
#endif /* CONFIG_ELF_CORE */
|
|
|
|
|
2021-05-05 01:39:08 +00:00
|
|
|
#ifdef CONFIG_MIGRATION
|
2021-05-05 01:39:19 +00:00
|
|
|
/*
|
2022-08-24 05:09:52 +00:00
|
|
|
* Returns the number of collected pages. Return value is always >= 0.
|
2021-05-05 01:39:19 +00:00
|
|
|
*/
|
2022-08-24 05:09:52 +00:00
|
|
|
static unsigned long collect_longterm_unpinnable_pages(
|
|
|
|
struct list_head *movable_page_list,
|
|
|
|
unsigned long nr_pages,
|
|
|
|
struct page **pages)
|
2019-03-05 23:47:44 +00:00
|
|
|
{
|
2022-08-24 05:09:52 +00:00
|
|
|
unsigned long i, collected = 0;
|
2022-02-17 17:46:35 +00:00
|
|
|
struct folio *prev_folio = NULL;
|
2022-08-24 05:09:52 +00:00
|
|
|
bool drain_allow = true;
|
2019-03-05 23:47:44 +00:00
|
|
|
|
2021-05-05 01:38:42 +00:00
|
|
|
for (i = 0; i < nr_pages; i++) {
|
2022-02-17 17:46:35 +00:00
|
|
|
struct folio *folio = page_folio(pages[i]);
|
2022-02-16 04:31:37 +00:00
|
|
|
|
2022-02-17 17:46:35 +00:00
|
|
|
if (folio == prev_folio)
|
2021-05-05 01:38:42 +00:00
|
|
|
continue;
|
2022-02-17 17:46:35 +00:00
|
|
|
prev_folio = folio;
|
2022-02-16 04:31:37 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
if (folio_is_longterm_pinnable(folio))
|
|
|
|
continue;
|
2022-07-15 15:05:13 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
collected++;
|
2022-07-15 15:05:13 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
if (folio_is_device_coherent(folio))
|
2022-02-16 04:31:37 +00:00
|
|
|
continue;
|
|
|
|
|
2022-02-17 17:46:35 +00:00
|
|
|
if (folio_test_hugetlb(folio)) {
|
2023-01-13 22:30:50 +00:00
|
|
|
isolate_hugetlb(folio, movable_page_list);
|
2022-02-16 04:31:37 +00:00
|
|
|
continue;
|
|
|
|
}
|
2019-03-05 23:47:44 +00:00
|
|
|
|
2022-02-17 17:46:35 +00:00
|
|
|
if (!folio_test_lru(folio) && drain_allow) {
|
2022-02-16 04:31:37 +00:00
|
|
|
lru_add_drain_all();
|
|
|
|
drain_allow = false;
|
|
|
|
}
|
|
|
|
|
mm: change to return bool for folio_isolate_lru()
Patch series "Change the return value for page isolation functions", v3.
Now the page isolation functions did not return a boolean to indicate
success or not, instead it will return a negative error when failed
to isolate a page. So below code used in most places seem a boolean
success/failure thing, which can confuse people whether the isolation
is successful.
if (folio_isolate_lru(folio))
continue;
Moreover the page isolation functions only return 0 or -EBUSY, and
most users did not care about the negative error except for few users,
thus we can convert all page isolation functions to return a boolean
value, which can remove the confusion to make code more clear.
No functional changes intended in this patch series.
This patch (of 4):
Now the folio_isolate_lru() did not return a boolean value to indicate
isolation success or not, however below code checking the return value can
make people think that it was a boolean success/failure thing, which makes
people easy to make mistakes (see the fix patch[1]).
if (folio_isolate_lru(folio))
continue;
Thus it's better to check the negative error value expilictly returned by
folio_isolate_lru(), which makes code more clear per Linus's
suggestion[2]. Moreover Matthew suggested we can convert the isolation
functions to return a boolean[3], since most users did not care about the
negative error value, and can also remove the confusing of checking return
value.
So this patch converts the folio_isolate_lru() to return a boolean value,
which means return 'true' to indicate the folio isolation is successful,
and 'false' means a failure to isolation. Meanwhile changing all users'
logic of checking the isolation state.
No functional changes intended.
[1] https://lore.kernel.org/all/20230131063206.28820-1-Kuan-Ying.Lee@mediatek.com/T/#u
[2] https://lore.kernel.org/all/CAHk-=wiBrY+O-4=2mrbVyxR+hOqfdJ=Do6xoucfJ9_5az01L4Q@mail.gmail.com/
[3] https://lore.kernel.org/all/Y+sTFqwMNAjDvxw3@casper.infradead.org/
Link: https://lkml.kernel.org/r/cover.1676424378.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/8a4e3679ed4196168efadf7ea36c038f2f7d5aa9.1676424378.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: SeongJae Park <sj@kernel.org>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2023-02-15 10:39:34 +00:00
|
|
|
if (!folio_isolate_lru(folio))
|
2022-02-16 04:31:37 +00:00
|
|
|
continue;
|
2022-08-24 05:09:52 +00:00
|
|
|
|
|
|
|
list_add_tail(&folio->lru, movable_page_list);
|
2022-02-17 17:46:35 +00:00
|
|
|
node_stat_mod_folio(folio,
|
|
|
|
NR_ISOLATED_ANON + folio_is_file_lru(folio),
|
|
|
|
folio_nr_pages(folio));
|
2019-03-05 23:47:44 +00:00
|
|
|
}
|
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
return collected;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Unpins all pages and migrates device coherent pages and movable_page_list.
|
|
|
|
* Returns -EAGAIN if all pages were successfully migrated or -errno for failure
|
|
|
|
* (or partial success).
|
|
|
|
*/
|
|
|
|
static int migrate_longterm_unpinnable_pages(
|
|
|
|
struct list_head *movable_page_list,
|
|
|
|
unsigned long nr_pages,
|
|
|
|
struct page **pages)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
unsigned long i;
|
2021-05-05 01:38:49 +00:00
|
|
|
|
2022-07-15 15:05:13 +00:00
|
|
|
for (i = 0; i < nr_pages; i++) {
|
2022-08-24 05:09:52 +00:00
|
|
|
struct folio *folio = page_folio(pages[i]);
|
|
|
|
|
|
|
|
if (folio_is_device_coherent(folio)) {
|
|
|
|
/*
|
|
|
|
* Migration will fail if the page is pinned, so convert
|
|
|
|
* the pin on the source page to a normal reference.
|
|
|
|
*/
|
|
|
|
pages[i] = NULL;
|
|
|
|
folio_get(folio);
|
|
|
|
gup_put_folio(folio, 1, FOLL_PIN);
|
|
|
|
|
|
|
|
if (migrate_device_coherent_page(&folio->page)) {
|
|
|
|
ret = -EBUSY;
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
2022-07-15 15:05:13 +00:00
|
|
|
continue;
|
2022-08-24 05:09:52 +00:00
|
|
|
}
|
2022-07-15 15:05:13 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
/*
|
|
|
|
* We can't migrate pages with unexpected references, so drop
|
|
|
|
* the reference obtained by __get_user_pages_locked().
|
|
|
|
* Migrating pages have been added to movable_page_list after
|
|
|
|
* calling folio_isolate_lru() which takes a reference so the
|
|
|
|
* page won't be freed if it's migrating.
|
|
|
|
*/
|
2022-08-24 05:09:51 +00:00
|
|
|
unpin_user_page(pages[i]);
|
2022-08-24 05:09:52 +00:00
|
|
|
pages[i] = NULL;
|
2021-05-05 01:39:19 +00:00
|
|
|
}
|
2022-02-16 04:31:37 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
if (!list_empty(movable_page_list)) {
|
2022-02-16 04:31:37 +00:00
|
|
|
struct migration_target_control mtc = {
|
|
|
|
.nid = NUMA_NO_NODE,
|
|
|
|
.gfp_mask = GFP_USER | __GFP_NOWARN,
|
|
|
|
};
|
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
if (migrate_pages(movable_page_list, alloc_migration_target,
|
|
|
|
NULL, (unsigned long)&mtc, MIGRATE_SYNC,
|
|
|
|
MR_LONGTERM_PIN, NULL)) {
|
2022-02-16 04:31:37 +00:00
|
|
|
ret = -ENOMEM;
|
2022-08-24 05:09:52 +00:00
|
|
|
goto err;
|
|
|
|
}
|
2019-03-05 23:47:44 +00:00
|
|
|
}
|
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
putback_movable_pages(movable_page_list);
|
|
|
|
|
|
|
|
return -EAGAIN;
|
|
|
|
|
|
|
|
err:
|
|
|
|
for (i = 0; i < nr_pages; i++)
|
|
|
|
if (pages[i])
|
|
|
|
unpin_user_page(pages[i]);
|
|
|
|
putback_movable_pages(movable_page_list);
|
2022-07-29 02:46:45 +00:00
|
|
|
|
2022-08-24 05:09:52 +00:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check whether all pages are *allowed* to be pinned. Rather confusingly, all
|
|
|
|
* pages in the range are required to be pinned via FOLL_PIN, before calling
|
|
|
|
* this routine.
|
|
|
|
*
|
|
|
|
* If any pages in the range are not allowed to be pinned, then this routine
|
|
|
|
* will migrate those pages away, unpin all the pages in the range and return
|
|
|
|
* -EAGAIN. The caller should re-pin the entire range with FOLL_PIN and then
|
|
|
|
* call this routine again.
|
|
|
|
*
|
|
|
|
* If an error other than -EAGAIN occurs, this indicates a migration failure.
|
|
|
|
* The caller should give up, and propagate the error back up the call stack.
|
|
|
|
*
|
|
|
|
* If everything is OK and all pages in the range are allowed to be pinned, then
|
|
|
|
* this routine leaves all pages pinned and returns zero for success.
|
|
|
|
*/
|
|
|
|
static long check_and_migrate_movable_pages(unsigned long nr_pages,
|
|
|
|
struct page **pages)
|
|
|
|
{
|
|
|
|
unsigned long collected;
|
|
|
|
LIST_HEAD(movable_page_list);
|
|
|
|
|
|
|
|
collected = collect_longterm_unpinnable_pages(&movable_page_list,
|
|
|
|
nr_pages, pages);
|
|
|
|
if (!collected)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
return migrate_longterm_unpinnable_pages(&movable_page_list, nr_pages,
|
|
|
|
pages);
|
2019-03-05 23:47:44 +00:00
|
|
|
}
|
|
|
|
#else
|
2021-05-05 01:39:19 +00:00
|
|
|
static long check_and_migrate_movable_pages(unsigned long nr_pages,
|
2022-08-24 05:09:51 +00:00
|
|
|
struct page **pages)
|
2019-03-05 23:47:44 +00:00
|
|
|
{
|
2022-07-29 02:46:45 +00:00
|
|
|
return 0;
|
2019-03-05 23:47:44 +00:00
|
|
|
}
|
2021-05-05 01:39:08 +00:00
|
|
|
#endif /* CONFIG_MIGRATION */
|
2019-03-05 23:47:44 +00:00
|
|
|
|
2017-11-30 00:10:35 +00:00
|
|
|
/*
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
* __gup_longterm_locked() is a wrapper for __get_user_pages_locked which
|
|
|
|
* allows us to process the FOLL_LONGTERM flag.
|
2017-11-30 00:10:35 +00:00
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
static long __gup_longterm_locked(struct mm_struct *mm,
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
unsigned long start,
|
|
|
|
unsigned long nr_pages,
|
|
|
|
struct page **pages,
|
|
|
|
struct vm_area_struct **vmas,
|
2022-11-16 20:07:09 +00:00
|
|
|
int *locked,
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
unsigned int gup_flags)
|
2017-11-30 00:10:35 +00:00
|
|
|
{
|
2021-05-05 01:39:19 +00:00
|
|
|
unsigned int flags;
|
2022-07-29 02:46:45 +00:00
|
|
|
long rc, nr_pinned_pages;
|
2017-11-30 00:10:35 +00:00
|
|
|
|
2021-05-05 01:39:19 +00:00
|
|
|
if (!(gup_flags & FOLL_LONGTERM))
|
|
|
|
return __get_user_pages_locked(mm, start, nr_pages, pages, vmas,
|
2022-11-16 20:07:09 +00:00
|
|
|
locked, gup_flags);
|
2022-08-24 05:09:52 +00:00
|
|
|
|
2021-05-05 01:39:19 +00:00
|
|
|
flags = memalloc_pin_save();
|
|
|
|
do {
|
2022-07-29 02:46:45 +00:00
|
|
|
nr_pinned_pages = __get_user_pages_locked(mm, start, nr_pages,
|
2022-11-16 20:07:09 +00:00
|
|
|
pages, vmas, locked,
|
2022-07-29 02:46:45 +00:00
|
|
|
gup_flags);
|
|
|
|
if (nr_pinned_pages <= 0) {
|
|
|
|
rc = nr_pinned_pages;
|
2021-05-05 01:39:19 +00:00
|
|
|
break;
|
2022-07-29 02:46:45 +00:00
|
|
|
}
|
2023-01-24 20:34:26 +00:00
|
|
|
|
|
|
|
/* FOLL_LONGTERM implies FOLL_PIN */
|
2022-08-24 05:09:51 +00:00
|
|
|
rc = check_and_migrate_movable_pages(nr_pinned_pages, pages);
|
2022-07-29 02:46:45 +00:00
|
|
|
} while (rc == -EAGAIN);
|
2021-05-05 01:39:19 +00:00
|
|
|
memalloc_pin_restore(flags);
|
2022-07-29 02:46:45 +00:00
|
|
|
return rc ? rc : nr_pinned_pages;
|
2017-11-30 00:10:35 +00:00
|
|
|
}
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
/*
|
|
|
|
* Check that the given flags are valid for the exported gup/pup interface, and
|
|
|
|
* update them with the required flags that the caller must have set.
|
|
|
|
*/
|
|
|
|
static bool is_valid_gup_args(struct page **pages, struct vm_area_struct **vmas,
|
|
|
|
int *locked, unsigned int *gup_flags_p,
|
|
|
|
unsigned int to_set)
|
2020-10-13 23:51:58 +00:00
|
|
|
{
|
2023-01-24 20:34:26 +00:00
|
|
|
unsigned int gup_flags = *gup_flags_p;
|
|
|
|
|
2020-10-13 23:51:58 +00:00
|
|
|
/*
|
2023-01-24 20:34:26 +00:00
|
|
|
* These flags not allowed to be specified externally to the gup
|
|
|
|
* interfaces:
|
|
|
|
* - FOLL_PIN/FOLL_TRIED/FOLL_FAST_ONLY are internal only
|
|
|
|
* - FOLL_REMOTE is internal only and used on follow_page()
|
2023-01-24 20:34:29 +00:00
|
|
|
* - FOLL_UNLOCKABLE is internal only and used if locked is !NULL
|
2020-10-13 23:51:58 +00:00
|
|
|
*/
|
2023-01-24 20:34:29 +00:00
|
|
|
if (WARN_ON_ONCE(gup_flags & (FOLL_PIN | FOLL_TRIED | FOLL_UNLOCKABLE |
|
2023-01-24 20:34:26 +00:00
|
|
|
FOLL_REMOTE | FOLL_FAST_ONLY)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
gup_flags |= to_set;
|
2023-01-24 20:34:29 +00:00
|
|
|
if (locked) {
|
|
|
|
/* At the external interface locked must be set */
|
|
|
|
if (WARN_ON_ONCE(*locked != 1))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
gup_flags |= FOLL_UNLOCKABLE;
|
|
|
|
}
|
2023-01-24 20:34:26 +00:00
|
|
|
|
|
|
|
/* FOLL_GET and FOLL_PIN are mutually exclusive. */
|
|
|
|
if (WARN_ON_ONCE((gup_flags & (FOLL_PIN | FOLL_GET)) ==
|
|
|
|
(FOLL_PIN | FOLL_GET)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* LONGTERM can only be specified when pinning */
|
|
|
|
if (WARN_ON_ONCE(!(gup_flags & FOLL_PIN) && (gup_flags & FOLL_LONGTERM)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
/* Pages input must be given if using GET/PIN */
|
|
|
|
if (WARN_ON_ONCE((gup_flags & (FOLL_GET | FOLL_PIN)) && !pages))
|
2020-10-13 23:51:58 +00:00
|
|
|
return false;
|
2023-01-24 20:34:26 +00:00
|
|
|
|
|
|
|
/* We want to allow the pgmap to be hot-unplugged at all times */
|
|
|
|
if (WARN_ON_ONCE((gup_flags & FOLL_LONGTERM) &&
|
|
|
|
(gup_flags & FOLL_PCI_P2PDMA)))
|
|
|
|
return false;
|
|
|
|
|
2020-10-13 23:51:58 +00:00
|
|
|
/*
|
2023-01-24 20:34:26 +00:00
|
|
|
* Can't use VMAs with locked, as locked allows GUP to unlock
|
|
|
|
* which invalidates the vmas array
|
2020-10-13 23:51:58 +00:00
|
|
|
*/
|
2023-01-24 20:34:29 +00:00
|
|
|
if (WARN_ON_ONCE(vmas && (gup_flags & FOLL_UNLOCKABLE)))
|
2020-10-13 23:51:58 +00:00
|
|
|
return false;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
*gup_flags_p = gup_flags;
|
2020-10-13 23:51:58 +00:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
mm/gup: split get_user_pages_remote() into two routines
Patch series "mm/gup: track FOLL_PIN pages", v6.
This activates tracking of FOLL_PIN pages. This is in support of fixing
the get_user_pages()+DMA problem described in [1]-[4].
FOLL_PIN support is now in the main linux tree. However, the patch to use
FOLL_PIN to track pages was *not* submitted, because Leon saw an RDMA test
suite failure that involved (I think) page refcount overflows when huge
pages were used.
This patch definitively solves that kind of overflow problem, by adding an
exact pincount, for compound pages (of order > 1), in the 3rd struct page
of a compound page. If available, that form of pincounting is used,
instead of the GUP_PIN_COUNTING_BIAS approach. Thanks again to Jan Kara
for that idea.
Other interesting changes:
* dump_page(): added one, or two new things to report for compound
pages: head refcount (for all compound pages), and map_pincount (for
compound pages of order > 1).
* Documentation/core-api/pin_user_pages.rst: removed the "TODO" for the
huge page refcount upper limit problems, and added notes about how it
works now. Also added a note about the dump_page() enhancements.
* Added some comments in gup.c and mm.h, to explain that there are two
ways to count pinned pages: exact (for compound pages of order > 1) and
fuzzy (GUP_PIN_COUNTING_BIAS: for all other pages).
============================================================
General notes about the tracking patch:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], [4] and in a remarkable number of email threads since about
2017. :)
In contrast to earlier approaches, the page tracking can be incrementally
applied to the kernel call sites that, until now, have been simply calling
get_user_pages() ("gup"). In other words, opt-in by changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
============================================================
Future steps:
* Convert more subsystems from get_user_pages() to pin_user_pages().
The first probably needs to be bio/biovecs, because any filesystem
testing is too difficult without those in place.
* Change VFS and filesystems to respond appropriately when encountering
dma-pinned pages.
* Work with Ira and others to connect this all up with file system
leases.
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages()
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
This patch (of 12):
An upcoming patch requires reusing the implementation of
get_user_pages_remote(). Split up get_user_pages_remote() into an outer
routine that checks flags, and an implementation routine that will be
reused. This makes subsequent changes much easier to understand.
There should be no change in behavior due to this patch.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:10 +00:00
|
|
|
#ifdef CONFIG_MMU
|
2020-06-02 04:48:24 +00:00
|
|
|
/**
|
2020-01-31 06:12:36 +00:00
|
|
|
* get_user_pages_remote() - pin user pages in memory
|
|
|
|
* @mm: mm_struct of target mm
|
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying lookup behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long. Or NULL, if caller
|
|
|
|
* only intends to ensure the pages are faulted in.
|
|
|
|
* @vmas: array of pointers to vmas corresponding to each page.
|
|
|
|
* Or NULL if the caller does not require them.
|
|
|
|
* @locked: pointer to lock flag indicating whether lock is held and
|
|
|
|
* subsequently whether VM_FAULT_RETRY functionality can be
|
|
|
|
* utilised. Lock must initially be held.
|
|
|
|
*
|
|
|
|
* Returns either number of pages pinned (which may be less than the
|
|
|
|
* number requested), or an error. Details about the return value:
|
|
|
|
*
|
|
|
|
* -- If nr_pages is 0, returns 0.
|
|
|
|
* -- If nr_pages is >0, but no pages were pinned, returns -errno.
|
|
|
|
* -- If nr_pages is >0, and some pages were pinned, returns the number of
|
|
|
|
* pages pinned. Again, this may be less than nr_pages.
|
|
|
|
*
|
|
|
|
* The caller is responsible for releasing returned @pages, via put_page().
|
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* @vmas are valid only as long as mmap_lock is held.
|
2020-01-31 06:12:36 +00:00
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* Must be called with mmap_lock held for read or write.
|
2020-01-31 06:12:36 +00:00
|
|
|
*
|
2020-06-02 04:48:24 +00:00
|
|
|
* get_user_pages_remote walks a process's page tables and takes a reference
|
|
|
|
* to each struct page that each user address corresponds to at a given
|
2020-01-31 06:12:36 +00:00
|
|
|
* instant. That is, it takes the page that would be accessed if a user
|
|
|
|
* thread accesses the given user virtual address at that instant.
|
|
|
|
*
|
|
|
|
* This does not guarantee that the page exists in the user mappings when
|
2020-06-02 04:48:24 +00:00
|
|
|
* get_user_pages_remote returns, and there may even be a completely different
|
2020-01-31 06:12:36 +00:00
|
|
|
* page there in some cases (eg. if mmapped pagecache has been invalidated
|
2023-03-09 10:48:13 +00:00
|
|
|
* and subsequently re-faulted). However it does guarantee that the page
|
2020-01-31 06:12:36 +00:00
|
|
|
* won't be freed completely. And mostly callers simply care that the page
|
|
|
|
* contains data that was valid *at some point in time*. Typically, an IO
|
|
|
|
* or similar operation cannot guarantee anything stronger anyway because
|
|
|
|
* locks can't be held over the syscall boundary.
|
|
|
|
*
|
|
|
|
* If gup_flags & FOLL_WRITE == 0, the page must not be written to. If the page
|
|
|
|
* is written to, set_page_dirty (or set_page_dirty_lock, as appropriate) must
|
|
|
|
* be called after the page is finished with, and before put_page is called.
|
|
|
|
*
|
2020-06-02 04:48:24 +00:00
|
|
|
* get_user_pages_remote is typically used for fewer-copy IO operations,
|
|
|
|
* to get a handle on the memory by some means other than accesses
|
|
|
|
* via the user virtual addresses. The pages may be submitted for
|
|
|
|
* DMA to devices or accessed via their kernel linear mapping (via the
|
|
|
|
* kmap APIs). Care should be taken to use the correct cache flushing APIs.
|
2020-01-31 06:12:36 +00:00
|
|
|
*
|
|
|
|
* See also get_user_pages_fast, for performance critical applications.
|
|
|
|
*
|
2020-06-02 04:48:24 +00:00
|
|
|
* get_user_pages_remote should be phased out in favor of
|
2020-01-31 06:12:36 +00:00
|
|
|
* get_user_pages_locked|unlocked or get_user_pages_fast. Nothing
|
2020-06-02 04:48:24 +00:00
|
|
|
* should use get_user_pages_remote because it cannot pass
|
2020-01-31 06:12:36 +00:00
|
|
|
* FAULT_FLAG_ALLOW_RETRY to handle_mm_fault.
|
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
long get_user_pages_remote(struct mm_struct *mm,
|
2020-01-31 06:12:36 +00:00
|
|
|
unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas, int *locked)
|
|
|
|
{
|
2023-01-24 20:34:30 +00:00
|
|
|
int local_locked = 1;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, vmas, locked, &gup_flags,
|
|
|
|
FOLL_TOUCH | FOLL_REMOTE))
|
2020-01-31 06:12:54 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2023-01-24 20:34:30 +00:00
|
|
|
return __get_user_pages_locked(mm, start, nr_pages, pages, vmas,
|
|
|
|
locked ? locked : &local_locked,
|
2023-01-24 20:34:26 +00:00
|
|
|
gup_flags);
|
2020-01-31 06:12:36 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(get_user_pages_remote);
|
|
|
|
|
2020-01-31 06:12:54 +00:00
|
|
|
#else /* CONFIG_MMU */
|
2020-08-12 01:39:01 +00:00
|
|
|
long get_user_pages_remote(struct mm_struct *mm,
|
2020-01-31 06:12:54 +00:00
|
|
|
unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas, int *locked)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif /* !CONFIG_MMU */
|
|
|
|
|
2020-06-02 04:48:24 +00:00
|
|
|
/**
|
|
|
|
* get_user_pages() - pin user pages in memory
|
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying lookup behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long. Or NULL, if caller
|
|
|
|
* only intends to ensure the pages are faulted in.
|
|
|
|
* @vmas: array of pointers to vmas corresponding to each page.
|
|
|
|
* Or NULL if the caller does not require them.
|
|
|
|
*
|
2020-08-12 01:39:01 +00:00
|
|
|
* This is the same as get_user_pages_remote(), just with a less-flexible
|
|
|
|
* calling convention where we assume that the mm being operated on belongs to
|
|
|
|
* the current task, and doesn't allow passing of a locked parameter. We also
|
|
|
|
* obviously don't pass FOLL_REMOTE in here.
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
*/
|
|
|
|
long get_user_pages(unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas)
|
|
|
|
{
|
2023-01-24 20:34:30 +00:00
|
|
|
int locked = 1;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, vmas, NULL, &gup_flags, FOLL_TOUCH))
|
2020-01-31 06:12:54 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2023-01-24 20:34:24 +00:00
|
|
|
return __get_user_pages_locked(current->mm, start, nr_pages, pages,
|
2023-01-24 20:34:30 +00:00
|
|
|
vmas, &locked, gup_flags);
|
mm/gup: replace get_user_pages_longterm() with FOLL_LONGTERM
Pach series "Add FOLL_LONGTERM to GUP fast and use it".
HFI1, qib, and mthca, use get_user_pages_fast() due to its performance
advantages. These pages can be held for a significant time. But
get_user_pages_fast() does not protect against mapping FS DAX pages.
Introduce FOLL_LONGTERM and use this flag in get_user_pages_fast() which
retains the performance while also adding the FS DAX checks. XDP has also
shown interest in using this functionality.[1]
In addition we change get_user_pages() to use the new FOLL_LONGTERM flag
and remove the specialized get_user_pages_longterm call.
[1] https://lkml.org/lkml/2019/3/19/939
"longterm" is a relative thing and at this point is probably a misnomer.
This is really flagging a pin which is going to be given to hardware and
can't move. I've thought of a couple of alternative names but I think we
have to settle on if we are going to use FL_LAYOUT or something else to
solve the "longterm" problem. Then I think we can change the flag to a
better name.
Secondly, it depends on how often you are registering memory. I have
spoken with some RDMA users who consider MR in the performance path...
For the overall application performance. I don't have the numbers as the
tests for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an aside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
This patch (of 7):
This patch starts a series which aims to support FOLL_LONGTERM in
get_user_pages_fast(). Some callers who would like to do a longterm (user
controlled pin) of pages with the fast variant of GUP for performance
purposes.
Rather than have a separate get_user_pages_longterm() call, introduce
FOLL_LONGTERM and change the longterm callers to use it.
This patch does not change any functionality. In the short term
"longterm" or user controlled pins are unsafe for Filesystems and FS DAX
in particular has been blocked. However, callers of get_user_pages_fast()
were not "protected".
FOLL_LONGTERM can _only_ be supported with get_user_pages[_fast]() as it
requires vmas to determine if DAX is in use.
NOTE: In merging with the CMA changes we opt to change the
get_user_pages() call in check_and_migrate_cma_pages() to a call of
__get_user_pages_locked() on the newly migrated pages. This makes the
code read better in that we are calling __get_user_pages_locked() on the
pages before and after a potential migration.
As a side affect some of the interfaces are cleaned up but this is not the
primary purpose of the series.
In review[1] it was asked:
<quote>
> This I don't get - if you do lock down long term mappings performance
> of the actual get_user_pages call shouldn't matter to start with.
>
> What do I miss?
A couple of points.
First "longterm" is a relative thing and at this point is probably a
misnomer. This is really flagging a pin which is going to be given to
hardware and can't move. I've thought of a couple of alternative names
but I think we have to settle on if we are going to use FL_LAYOUT or
something else to solve the "longterm" problem. Then I think we can
change the flag to a better name.
Second, It depends on how often you are registering memory. I have spoken
with some RDMA users who consider MR in the performance path... For the
overall application performance. I don't have the numbers as the tests
for HFI1 were done a long time ago. But there was a significant
advantage. Some of which is probably due to the fact that you don't have
to hold mmap_sem.
Finally, architecturally I think it would be good for everyone to use
*_fast. There are patches submitted to the RDMA list which would allow
the use of *_fast (they reworking the use of mmap_sem) and as soon as they
are accepted I'll submit a patch to convert the RDMA core as well. Also
to this point others are looking to use *_fast.
As an asside, Jasons pointed out in my previous submission that *_fast and
*_unlocked look very much the same. I agree and I think further cleanup
will be coming. But I'm focused on getting the final solution for DAX at
the moment.
</quote>
[1] https://lore.kernel.org/lkml/20190220180255.GA12020@iweiny-DESK2.sc.intel.com/T/#md6abad2569f3bf6c1f03686c8097ab6563e94965
[ira.weiny@intel.com: v3]
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190328084422.29911-2-ira.weiny@intel.com
Link: http://lkml.kernel.org/r/20190317183438.2057-2-ira.weiny@intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Rich Felker <dalias@libc.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: James Hogan <jhogan@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Mike Marshall <hubcap@omnibond.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-05-14 00:17:03 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(get_user_pages);
|
2017-11-30 00:10:35 +00:00
|
|
|
|
2015-04-14 22:44:45 +00:00
|
|
|
/*
|
2019-07-12 03:57:18 +00:00
|
|
|
* get_user_pages_unlocked() is suitable to replace the form:
|
2015-04-14 22:44:45 +00:00
|
|
|
*
|
2020-06-09 04:33:51 +00:00
|
|
|
* mmap_read_lock(mm);
|
2020-08-12 01:39:01 +00:00
|
|
|
* get_user_pages(mm, ..., pages, NULL);
|
2020-06-09 04:33:51 +00:00
|
|
|
* mmap_read_unlock(mm);
|
2019-07-12 03:57:18 +00:00
|
|
|
*
|
|
|
|
* with:
|
|
|
|
*
|
2020-08-12 01:39:01 +00:00
|
|
|
* get_user_pages_unlocked(mm, ..., pages);
|
2019-07-12 03:57:18 +00:00
|
|
|
*
|
|
|
|
* It is functionally equivalent to get_user_pages_fast so
|
|
|
|
* get_user_pages_fast should be used instead if specific gup_flags
|
|
|
|
* (e.g. FOLL_FORCE) are not required.
|
2015-04-14 22:44:45 +00:00
|
|
|
*/
|
2019-07-12 03:57:18 +00:00
|
|
|
long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
|
|
|
|
struct page **pages, unsigned int gup_flags)
|
2015-04-14 22:44:45 +00:00
|
|
|
{
|
2023-01-24 20:34:22 +00:00
|
|
|
int locked = 0;
|
2015-04-14 22:44:45 +00:00
|
|
|
|
2023-01-24 20:34:29 +00:00
|
|
|
if (!is_valid_gup_args(pages, NULL, NULL, &gup_flags,
|
|
|
|
FOLL_TOUCH | FOLL_UNLOCKABLE))
|
2023-01-24 20:34:26 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2023-01-24 20:34:24 +00:00
|
|
|
return __get_user_pages_locked(current->mm, start, nr_pages, pages,
|
2023-01-24 20:34:26 +00:00
|
|
|
NULL, &locked, gup_flags);
|
2014-06-04 23:08:10 +00:00
|
|
|
}
|
2019-07-12 03:57:18 +00:00
|
|
|
EXPORT_SYMBOL(get_user_pages_unlocked);
|
2014-10-09 22:29:14 +00:00
|
|
|
|
|
|
|
/*
|
2019-07-12 03:57:14 +00:00
|
|
|
* Fast GUP
|
2014-10-09 22:29:14 +00:00
|
|
|
*
|
|
|
|
* get_user_pages_fast attempts to pin user pages by walking the page
|
|
|
|
* tables directly and avoids taking locks. Thus the walker needs to be
|
|
|
|
* protected from page table pages being freed from under it, and should
|
|
|
|
* block any THP splits.
|
|
|
|
*
|
|
|
|
* One way to achieve this is to have the walker disable interrupts, and
|
|
|
|
* rely on IPIs from the TLB flushing code blocking before the page table
|
|
|
|
* pages are freed. This is unsuitable for architectures that do not need
|
|
|
|
* to broadcast an IPI when invalidating TLBs.
|
|
|
|
*
|
|
|
|
* Another way to achieve this is to batch up page table containing pages
|
|
|
|
* belonging to more than one mm_user, then rcu_sched a callback to free those
|
|
|
|
* pages. Disabling interrupts will allow the fast_gup walker to both block
|
|
|
|
* the rcu_sched callback, and an IPI that we broadcast for splitting THPs
|
|
|
|
* (which is a relatively rare event). The code below adopts this strategy.
|
|
|
|
*
|
|
|
|
* Before activating this code, please be aware that the following assumptions
|
|
|
|
* are currently made:
|
|
|
|
*
|
2020-02-04 01:37:02 +00:00
|
|
|
* *) Either MMU_GATHER_RCU_TABLE_FREE is enabled, and tlb_remove_table() is used to
|
2017-06-06 11:31:20 +00:00
|
|
|
* free pages containing page tables or TLB flushing requires IPI broadcast.
|
2014-10-09 22:29:14 +00:00
|
|
|
*
|
|
|
|
* *) ptes can be read atomically by the architecture.
|
|
|
|
*
|
|
|
|
* *) access_ok is sufficient to validate userspace address ranges.
|
|
|
|
*
|
|
|
|
* The last two assumptions can be relaxed by the addition of helper functions.
|
|
|
|
*
|
|
|
|
* This code is based heavily on the PowerPC implementation by Nick Piggin.
|
|
|
|
*/
|
2019-07-12 03:57:14 +00:00
|
|
|
#ifdef CONFIG_HAVE_FAST_GUP
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
|
2019-07-12 03:57:46 +00:00
|
|
|
static void __maybe_unused undo_dev_pagemap(int *nr, int nr_start,
|
2020-04-02 04:05:22 +00:00
|
|
|
unsigned int flags,
|
2019-07-12 03:57:46 +00:00
|
|
|
struct page **pages)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
while ((*nr) - nr_start) {
|
|
|
|
struct page *page = pages[--(*nr)];
|
|
|
|
|
|
|
|
ClearPageReferenced(page);
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
if (flags & FOLL_PIN)
|
|
|
|
unpin_user_page(page);
|
|
|
|
else
|
|
|
|
put_page(page);
|
2017-03-16 15:26:53 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-06-08 00:06:08 +00:00
|
|
|
#ifdef CONFIG_ARCH_HAS_PTE_SPECIAL
|
2022-09-07 18:01:43 +00:00
|
|
|
/*
|
|
|
|
* Fast-gup relies on pte change detection to avoid concurrent pgtable
|
|
|
|
* operations.
|
|
|
|
*
|
|
|
|
* To pin the page, fast-gup needs to do below in order:
|
|
|
|
* (1) pin the page (by prefetching pte), then (2) check pte not changed.
|
|
|
|
*
|
|
|
|
* For the rest of pgtable operations where pgtable updates can be racy
|
|
|
|
* with fast-gup, we need to do (1) clear pte, then (2) check whether page
|
|
|
|
* is pinned.
|
|
|
|
*
|
|
|
|
* Above will work for all pte-level operations, including THP split.
|
|
|
|
*
|
|
|
|
* For THP collapse, it's a bit more complicated because fast-gup may be
|
|
|
|
* walking a pgtable page that is being freed (pte is still valid but pmd
|
|
|
|
* can be cleared already). To avoid race in such condition, we need to
|
|
|
|
* also check pmd here to make sure pmd doesn't change (corresponds to
|
|
|
|
* pmdp_collapse_flush() in the THP collapse code path).
|
|
|
|
*/
|
|
|
|
static int gup_pte_range(pmd_t pmd, pmd_t *pmdp, unsigned long addr,
|
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
2017-03-16 15:26:53 +00:00
|
|
|
struct dev_pagemap *pgmap = NULL;
|
|
|
|
int nr_start = *nr, ret = 0;
|
2014-10-09 22:29:14 +00:00
|
|
|
pte_t *ptep, *ptem;
|
|
|
|
|
|
|
|
ptem = ptep = pte_offset_map(&pmd, addr);
|
|
|
|
do {
|
2020-11-13 10:41:40 +00:00
|
|
|
pte_t pte = ptep_get_lockless(ptep);
|
2021-12-10 20:54:11 +00:00
|
|
|
struct page *page;
|
|
|
|
struct folio *folio;
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2022-08-25 16:46:58 +00:00
|
|
|
if (pte_protnone(pte) && !gup_can_follow_protnone(flags))
|
2017-03-16 15:26:50 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!pte_access_permitted(pte, flags & FOLL_WRITE))
|
2017-03-16 15:26:50 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
|
2017-03-16 15:26:53 +00:00
|
|
|
if (pte_devmap(pte)) {
|
2019-05-14 00:17:14 +00:00
|
|
|
if (unlikely(flags & FOLL_LONGTERM))
|
|
|
|
goto pte_unmap;
|
|
|
|
|
2017-03-16 15:26:53 +00:00
|
|
|
pgmap = get_dev_pagemap(pte_pfn(pte), pgmap);
|
|
|
|
if (unlikely(!pgmap)) {
|
2020-04-02 04:05:22 +00:00
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
2017-03-16 15:26:53 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
}
|
|
|
|
} else if (pte_special(pte))
|
2014-10-09 22:29:14 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
|
|
|
|
VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
|
|
|
|
page = pte_page(pte);
|
|
|
|
|
2021-12-10 20:54:11 +00:00
|
|
|
folio = try_grab_folio(page, 1, flags);
|
|
|
|
if (!folio)
|
2014-10-09 22:29:14 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
|
mm: introduce memfd_secret system call to create "secret" memory areas
Introduce "memfd_secret" system call with the ability to create memory
areas visible only in the context of the owning process and not mapped not
only to other processes but in the kernel page tables as well.
The secretmem feature is off by default and the user must explicitly
enable it at the boot time.
Once secretmem is enabled, the user will be able to create a file
descriptor using the memfd_secret() system call. The memory areas created
by mmap() calls from this file descriptor will be unmapped from the kernel
direct map and they will be only mapped in the page table of the processes
that have access to the file descriptor.
Secretmem is designed to provide the following protections:
* Enhanced protection (in conjunction with all the other in-kernel
attack prevention systems) against ROP attacks. Seceretmem makes
"simple" ROP insufficient to perform exfiltration, which increases the
required complexity of the attack. Along with other protections like
the kernel stack size limit and address space layout randomization which
make finding gadgets is really hard, absence of any in-kernel primitive
for accessing secret memory means the one gadget ROP attack can't work.
Since the only way to access secret memory is to reconstruct the missing
mapping entry, the attacker has to recover the physical page and insert
a PTE pointing to it in the kernel and then retrieve the contents. That
takes at least three gadgets which is a level of difficulty beyond most
standard attacks.
* Prevent cross-process secret userspace memory exposures. Once the
secret memory is allocated, the user can't accidentally pass it into the
kernel to be transmitted somewhere. The secreremem pages cannot be
accessed via the direct map and they are disallowed in GUP.
* Harden against exploited kernel flaws. In order to access secretmem,
a kernel-side attack would need to either walk the page tables and
create new ones, or spawn a new privileged uiserspace process to perform
secrets exfiltration using ptrace.
The file descriptor based memory has several advantages over the
"traditional" mm interfaces, such as mlock(), mprotect(), madvise(). File
descriptor approach allows explicit and controlled sharing of the memory
areas, it allows to seal the operations. Besides, file descriptor based
memory paves the way for VMMs to remove the secret memory range from the
userspace hipervisor process, for instance QEMU. Andy Lutomirski says:
"Getting fd-backed memory into a guest will take some possibly major
work in the kernel, but getting vma-backed memory into a guest without
mapping it in the host user address space seems much, much worse."
memfd_secret() is made a dedicated system call rather than an extension to
memfd_create() because it's purpose is to allow the user to create more
secure memory mappings rather than to simply allow file based access to
the memory. Nowadays a new system call cost is negligible while it is way
simpler for userspace to deal with a clear-cut system calls than with a
multiplexer or an overloaded syscall. Moreover, the initial
implementation of memfd_secret() is completely distinct from
memfd_create() so there is no much sense in overloading memfd_create() to
begin with. If there will be a need for code sharing between these
implementation it can be easily achieved without a need to adjust user
visible APIs.
The secret memory remains accessible in the process context using uaccess
primitives, but it is not exposed to the kernel otherwise; secret memory
areas are removed from the direct map and functions in the
follow_page()/get_user_page() family will refuse to return a page that
belongs to the secret memory area.
Once there will be a use case that will require exposing secretmem to the
kernel it will be an opt-in request in the system call flags so that user
would have to decide what data can be exposed to the kernel.
Removing of the pages from the direct map may cause its fragmentation on
architectures that use large pages to map the physical memory which
affects the system performance. However, the original Kconfig text for
CONFIG_DIRECT_GBPAGES said that gigabyte pages in the direct map "... can
improve the kernel's performance a tiny bit ..." (commit 00d1c5e05736
("x86: add gbpages switches")) and the recent report [1] showed that "...
although 1G mappings are a good default choice, there is no compelling
evidence that it must be the only choice". Hence, it is sufficient to
have secretmem disabled by default with the ability of a system
administrator to enable it at boot time.
Pages in the secretmem regions are unevictable and unmovable to avoid
accidental exposure of the sensitive data via swap or during page
migration.
Since the secretmem mappings are locked in memory they cannot exceed
RLIMIT_MEMLOCK. Since these mappings are already locked independently
from mlock(), an attempt to mlock()/munlock() secretmem range would fail
and mlockall()/munlockall() will ignore secretmem mappings.
However, unlike mlock()ed memory, secretmem currently behaves more like
long-term GUP: secretmem mappings are unmovable mappings directly consumed
by user space. With default limits, there is no excessive use of
secretmem and it poses no real problem in combination with
ZONE_MOVABLE/CMA, but in the future this should be addressed to allow
balanced use of large amounts of secretmem along with ZONE_MOVABLE/CMA.
A page that was a part of the secret memory area is cleared when it is
freed to ensure the data is not exposed to the next user of that page.
The following example demonstrates creation of a secret mapping (error
handling is omitted):
fd = memfd_secret(0);
ftruncate(fd, MAP_SIZE);
ptr = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);
[1] https://lore.kernel.org/linux-mm/213b4567-46ce-f116-9cdf-bbd0c884eb3c@linux.intel.com/
[akpm@linux-foundation.org: suppress Kconfig whine]
Link: https://lkml.kernel.org/r/20210518072034.31572-5-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Elena Reshetova <elena.reshetova@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tycho Andersen <tycho@tycho.ws>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-08 01:08:03 +00:00
|
|
|
if (unlikely(page_is_secretmem(page))) {
|
2021-12-10 20:54:11 +00:00
|
|
|
gup_put_folio(folio, 1, flags);
|
mm: introduce memfd_secret system call to create "secret" memory areas
Introduce "memfd_secret" system call with the ability to create memory
areas visible only in the context of the owning process and not mapped not
only to other processes but in the kernel page tables as well.
The secretmem feature is off by default and the user must explicitly
enable it at the boot time.
Once secretmem is enabled, the user will be able to create a file
descriptor using the memfd_secret() system call. The memory areas created
by mmap() calls from this file descriptor will be unmapped from the kernel
direct map and they will be only mapped in the page table of the processes
that have access to the file descriptor.
Secretmem is designed to provide the following protections:
* Enhanced protection (in conjunction with all the other in-kernel
attack prevention systems) against ROP attacks. Seceretmem makes
"simple" ROP insufficient to perform exfiltration, which increases the
required complexity of the attack. Along with other protections like
the kernel stack size limit and address space layout randomization which
make finding gadgets is really hard, absence of any in-kernel primitive
for accessing secret memory means the one gadget ROP attack can't work.
Since the only way to access secret memory is to reconstruct the missing
mapping entry, the attacker has to recover the physical page and insert
a PTE pointing to it in the kernel and then retrieve the contents. That
takes at least three gadgets which is a level of difficulty beyond most
standard attacks.
* Prevent cross-process secret userspace memory exposures. Once the
secret memory is allocated, the user can't accidentally pass it into the
kernel to be transmitted somewhere. The secreremem pages cannot be
accessed via the direct map and they are disallowed in GUP.
* Harden against exploited kernel flaws. In order to access secretmem,
a kernel-side attack would need to either walk the page tables and
create new ones, or spawn a new privileged uiserspace process to perform
secrets exfiltration using ptrace.
The file descriptor based memory has several advantages over the
"traditional" mm interfaces, such as mlock(), mprotect(), madvise(). File
descriptor approach allows explicit and controlled sharing of the memory
areas, it allows to seal the operations. Besides, file descriptor based
memory paves the way for VMMs to remove the secret memory range from the
userspace hipervisor process, for instance QEMU. Andy Lutomirski says:
"Getting fd-backed memory into a guest will take some possibly major
work in the kernel, but getting vma-backed memory into a guest without
mapping it in the host user address space seems much, much worse."
memfd_secret() is made a dedicated system call rather than an extension to
memfd_create() because it's purpose is to allow the user to create more
secure memory mappings rather than to simply allow file based access to
the memory. Nowadays a new system call cost is negligible while it is way
simpler for userspace to deal with a clear-cut system calls than with a
multiplexer or an overloaded syscall. Moreover, the initial
implementation of memfd_secret() is completely distinct from
memfd_create() so there is no much sense in overloading memfd_create() to
begin with. If there will be a need for code sharing between these
implementation it can be easily achieved without a need to adjust user
visible APIs.
The secret memory remains accessible in the process context using uaccess
primitives, but it is not exposed to the kernel otherwise; secret memory
areas are removed from the direct map and functions in the
follow_page()/get_user_page() family will refuse to return a page that
belongs to the secret memory area.
Once there will be a use case that will require exposing secretmem to the
kernel it will be an opt-in request in the system call flags so that user
would have to decide what data can be exposed to the kernel.
Removing of the pages from the direct map may cause its fragmentation on
architectures that use large pages to map the physical memory which
affects the system performance. However, the original Kconfig text for
CONFIG_DIRECT_GBPAGES said that gigabyte pages in the direct map "... can
improve the kernel's performance a tiny bit ..." (commit 00d1c5e05736
("x86: add gbpages switches")) and the recent report [1] showed that "...
although 1G mappings are a good default choice, there is no compelling
evidence that it must be the only choice". Hence, it is sufficient to
have secretmem disabled by default with the ability of a system
administrator to enable it at boot time.
Pages in the secretmem regions are unevictable and unmovable to avoid
accidental exposure of the sensitive data via swap or during page
migration.
Since the secretmem mappings are locked in memory they cannot exceed
RLIMIT_MEMLOCK. Since these mappings are already locked independently
from mlock(), an attempt to mlock()/munlock() secretmem range would fail
and mlockall()/munlockall() will ignore secretmem mappings.
However, unlike mlock()ed memory, secretmem currently behaves more like
long-term GUP: secretmem mappings are unmovable mappings directly consumed
by user space. With default limits, there is no excessive use of
secretmem and it poses no real problem in combination with
ZONE_MOVABLE/CMA, but in the future this should be addressed to allow
balanced use of large amounts of secretmem along with ZONE_MOVABLE/CMA.
A page that was a part of the secret memory area is cleared when it is
freed to ensure the data is not exposed to the next user of that page.
The following example demonstrates creation of a secret mapping (error
handling is omitted):
fd = memfd_secret(0);
ftruncate(fd, MAP_SIZE);
ptr = mmap(NULL, MAP_SIZE, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);
[1] https://lore.kernel.org/linux-mm/213b4567-46ce-f116-9cdf-bbd0c884eb3c@linux.intel.com/
[akpm@linux-foundation.org: suppress Kconfig whine]
Link: https://lkml.kernel.org/r/20210518072034.31572-5-rppt@kernel.org
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Acked-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Elena Reshetova <elena.reshetova@intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Palmer Dabbelt <palmerdabbelt@google.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tycho Andersen <tycho@tycho.ws>
Cc: Will Deacon <will@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-07-08 01:08:03 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
}
|
|
|
|
|
2022-09-07 18:01:43 +00:00
|
|
|
if (unlikely(pmd_val(pmd) != pmd_val(*pmdp)) ||
|
|
|
|
unlikely(pte_val(pte) != pte_val(*ptep))) {
|
2021-12-10 20:54:11 +00:00
|
|
|
gup_put_folio(folio, 1, flags);
|
2014-10-09 22:29:14 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
}
|
|
|
|
|
mm/gup: reliable R/O long-term pinning in COW mappings
We already support reliable R/O pinning of anonymous memory. However,
assume we end up pinning (R/O long-term) a pagecache page or the shared
zeropage inside a writable private ("COW") mapping. The next write access
will trigger a write-fault and replace the pinned page by an exclusive
anonymous page in the process page tables to break COW: the pinned page no
longer corresponds to the page mapped into the process' page table.
Now that FAULT_FLAG_UNSHARE can break COW on anything mapped into a
COW mapping, let's properly break COW first before R/O long-term
pinning something that's not an exclusive anon page inside a COW
mapping. FAULT_FLAG_UNSHARE will break COW and map an exclusive anon page
instead that can get pinned safely.
With this change, we can stop using FOLL_FORCE|FOLL_WRITE for reliable
R/O long-term pinning in COW mappings.
With this change, the new R/O long-term pinning tests for non-anonymous
memory succeed:
# [RUN] R/O longterm GUP pin ... with shared zeropage
ok 151 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd
ok 152 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with tmpfile
ok 153 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with huge zeropage
ok 154 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (2048 kB)
ok 155 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (1048576 kB)
ok 156 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with shared zeropage
ok 157 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd
ok 158 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with tmpfile
ok 159 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with huge zeropage
ok 160 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (2048 kB)
ok 161 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (1048576 kB)
ok 162 Longterm R/O pin is reliable
Note 1: We don't care about short-term R/O-pinning, because they have
snapshot semantics: they are not supposed to observe modifications that
happen after pinning.
As one example, assume we start direct I/O to read from a page and store
page content into a file: modifications to page content after starting
direct I/O are not guaranteed to end up in the file. So even if we'd pin
the shared zeropage, the end result would be as expected -- getting zeroes
stored to the file.
Note 2: For shared mappings we'll now always fallback to the slow path to
lookup the VMA when R/O long-term pining. While that's the necessary price
we have to pay right now, it's actually not that bad in practice: most
FOLL_LONGTERM users already specify FOLL_WRITE, for example, along with
FOLL_FORCE because they tried dealing with COW mappings correctly ...
Note 3: For users that use FOLL_LONGTERM right now without FOLL_WRITE,
such as VFIO, we'd now no longer pin the shared zeropage. Instead, we'd
populate exclusive anon pages that we can pin. There was a concern that
this could affect the memlock limit of existing setups.
For example, a VM running with VFIO could run into the memlock limit and
fail to run. However, we essentially had the same behavior already in
commit 17839856fd58 ("gup: document and work around "COW can break either
way" issue") which got merged into some enterprise distros, and there were
not any such complaints. So most probably, we're fine.
Link: https://lkml.kernel.org/r/20221116102659.70287-10-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-16 10:26:48 +00:00
|
|
|
if (!pte_write(pte) && gup_must_unshare(NULL, flags, page)) {
|
2022-05-10 01:20:45 +00:00
|
|
|
gup_put_folio(folio, 1, flags);
|
|
|
|
goto pte_unmap;
|
|
|
|
}
|
|
|
|
|
2020-04-02 04:05:56 +00:00
|
|
|
/*
|
|
|
|
* We need to make the page accessible if and only if we are
|
|
|
|
* going to access its content (the FOLL_PIN case). Please
|
|
|
|
* see Documentation/core-api/pin_user_pages.rst for
|
|
|
|
* details.
|
|
|
|
*/
|
|
|
|
if (flags & FOLL_PIN) {
|
|
|
|
ret = arch_make_page_accessible(page);
|
|
|
|
if (ret) {
|
2021-12-10 20:54:11 +00:00
|
|
|
gup_put_folio(folio, 1, flags);
|
2020-04-02 04:05:56 +00:00
|
|
|
goto pte_unmap;
|
|
|
|
}
|
|
|
|
}
|
2021-12-10 20:54:11 +00:00
|
|
|
folio_set_referenced(folio);
|
2014-10-09 22:29:14 +00:00
|
|
|
pages[*nr] = page;
|
|
|
|
(*nr)++;
|
|
|
|
} while (ptep++, addr += PAGE_SIZE, addr != end);
|
|
|
|
|
|
|
|
ret = 1;
|
|
|
|
|
|
|
|
pte_unmap:
|
2017-12-29 07:54:01 +00:00
|
|
|
if (pgmap)
|
|
|
|
put_dev_pagemap(pgmap);
|
2014-10-09 22:29:14 +00:00
|
|
|
pte_unmap(ptem);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we can't determine whether or not a pte is special, then fail immediately
|
|
|
|
* for ptes. Note, we can still pin HugeTLB and THP as these are guaranteed not
|
|
|
|
* to be special.
|
|
|
|
*
|
|
|
|
* For a futex to be placed on a THP tail page, get_futex_key requires a
|
2020-06-08 04:40:55 +00:00
|
|
|
* get_user_pages_fast_only implementation that can pin pages. Thus it's still
|
2014-10-09 22:29:14 +00:00
|
|
|
* useful to have gup_huge_pmd even if we can't operate on ptes.
|
|
|
|
*/
|
2022-09-07 18:01:43 +00:00
|
|
|
static int gup_pte_range(pmd_t pmd, pmd_t *pmdp, unsigned long addr,
|
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
2018-06-08 00:06:08 +00:00
|
|
|
#endif /* CONFIG_ARCH_HAS_PTE_SPECIAL */
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2019-07-16 23:30:47 +00:00
|
|
|
#if defined(CONFIG_ARCH_HAS_PTE_DEVMAP) && defined(CONFIG_TRANSPARENT_HUGEPAGE)
|
2017-03-16 15:26:53 +00:00
|
|
|
static int __gup_device_huge(unsigned long pfn, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
int nr_start = *nr;
|
|
|
|
struct dev_pagemap *pgmap = NULL;
|
|
|
|
|
|
|
|
do {
|
|
|
|
struct page *page = pfn_to_page(pfn);
|
|
|
|
|
|
|
|
pgmap = get_dev_pagemap(pfn, pgmap);
|
|
|
|
if (unlikely(!pgmap)) {
|
2020-04-02 04:05:22 +00:00
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
2021-09-02 21:53:42 +00:00
|
|
|
break;
|
2017-03-16 15:26:53 +00:00
|
|
|
}
|
2022-10-21 17:41:09 +00:00
|
|
|
|
|
|
|
if (!(flags & FOLL_PCI_P2PDMA) && is_pci_p2pdma_page(page)) {
|
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2017-03-16 15:26:53 +00:00
|
|
|
SetPageReferenced(page);
|
|
|
|
pages[*nr] = page;
|
2022-10-21 17:41:08 +00:00
|
|
|
if (unlikely(try_grab_page(page, flags))) {
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
2021-09-02 21:53:42 +00:00
|
|
|
break;
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
}
|
2017-03-16 15:26:53 +00:00
|
|
|
(*nr)++;
|
|
|
|
pfn++;
|
|
|
|
} while (addr += PAGE_SIZE, addr != end);
|
2017-12-29 07:54:01 +00:00
|
|
|
|
2021-09-02 21:53:42 +00:00
|
|
|
put_dev_pagemap(pgmap);
|
2021-11-05 20:37:16 +00:00
|
|
|
return addr == end;
|
2017-03-16 15:26:53 +00:00
|
|
|
}
|
|
|
|
|
2018-04-20 04:32:19 +00:00
|
|
|
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
unsigned long fault_pfn;
|
2018-04-20 04:32:19 +00:00
|
|
|
int nr_start = *nr;
|
|
|
|
|
|
|
|
fault_pfn = pmd_pfn(orig) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
|
2020-04-02 04:05:14 +00:00
|
|
|
if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
|
2018-04-20 04:32:19 +00:00
|
|
|
return 0;
|
2017-03-16 15:26:53 +00:00
|
|
|
|
2018-04-20 04:32:19 +00:00
|
|
|
if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
|
2020-04-02 04:05:22 +00:00
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
2018-04-20 04:32:19 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
return 1;
|
2017-03-16 15:26:53 +00:00
|
|
|
}
|
|
|
|
|
2018-04-20 04:32:19 +00:00
|
|
|
static int __gup_device_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
unsigned long fault_pfn;
|
2018-04-20 04:32:19 +00:00
|
|
|
int nr_start = *nr;
|
|
|
|
|
|
|
|
fault_pfn = pud_pfn(orig) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
|
2020-04-02 04:05:14 +00:00
|
|
|
if (!__gup_device_huge(fault_pfn, addr, end, flags, pages, nr))
|
2018-04-20 04:32:19 +00:00
|
|
|
return 0;
|
2017-03-16 15:26:53 +00:00
|
|
|
|
2018-04-20 04:32:19 +00:00
|
|
|
if (unlikely(pud_val(orig) != pud_val(*pudp))) {
|
2020-04-02 04:05:22 +00:00
|
|
|
undo_dev_pagemap(nr, nr_start, flags, pages);
|
2018-04-20 04:32:19 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
return 1;
|
2017-03-16 15:26:53 +00:00
|
|
|
}
|
|
|
|
#else
|
2018-04-20 04:32:19 +00:00
|
|
|
static int __gup_device_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
BUILD_BUG();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2018-04-20 04:32:19 +00:00
|
|
|
static int __gup_device_huge_pud(pud_t pud, pud_t *pudp, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2017-03-16 15:26:53 +00:00
|
|
|
{
|
|
|
|
BUILD_BUG();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
static int record_subpages(struct page *page, unsigned long addr,
|
|
|
|
unsigned long end, struct page **pages)
|
|
|
|
{
|
|
|
|
int nr;
|
|
|
|
|
2022-01-07 18:25:55 +00:00
|
|
|
for (nr = 0; addr != end; nr++, addr += PAGE_SIZE)
|
|
|
|
pages[nr] = nth_page(page, nr);
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
|
|
|
|
return nr;
|
|
|
|
}
|
|
|
|
|
2019-07-12 03:57:28 +00:00
|
|
|
#ifdef CONFIG_ARCH_HAS_HUGEPD
|
|
|
|
static unsigned long hugepte_addr_end(unsigned long addr, unsigned long end,
|
|
|
|
unsigned long sz)
|
|
|
|
{
|
|
|
|
unsigned long __boundary = (addr + sz) & ~(sz-1);
|
|
|
|
return (__boundary - 1 < end - 1) ? __boundary : end;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int gup_hugepte(pte_t *ptep, unsigned long sz, unsigned long addr,
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2019-07-12 03:57:28 +00:00
|
|
|
{
|
|
|
|
unsigned long pte_end;
|
2021-12-22 21:38:30 +00:00
|
|
|
struct page *page;
|
|
|
|
struct folio *folio;
|
2019-07-12 03:57:28 +00:00
|
|
|
pte_t pte;
|
|
|
|
int refs;
|
|
|
|
|
|
|
|
pte_end = (addr + sz) & ~(sz-1);
|
|
|
|
if (pte_end < end)
|
|
|
|
end = pte_end;
|
|
|
|
|
2020-06-15 12:57:57 +00:00
|
|
|
pte = huge_ptep_get(ptep);
|
2019-07-12 03:57:28 +00:00
|
|
|
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
if (!pte_access_permitted(pte, flags & FOLL_WRITE))
|
2019-07-12 03:57:28 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* hugepages are never "special" */
|
|
|
|
VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
|
|
|
|
|
2021-12-22 21:38:30 +00:00
|
|
|
page = nth_page(pte_page(pte), (addr & (sz - 1)) >> PAGE_SHIFT);
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
refs = record_subpages(page, addr, end, pages + *nr);
|
2019-07-12 03:57:28 +00:00
|
|
|
|
2021-12-22 21:38:30 +00:00
|
|
|
folio = try_grab_folio(page, refs, flags);
|
|
|
|
if (!folio)
|
2019-07-12 03:57:28 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (unlikely(pte_val(pte) != pte_val(*ptep))) {
|
2021-12-22 21:38:30 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
2019-07-12 03:57:28 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: reliable R/O long-term pinning in COW mappings
We already support reliable R/O pinning of anonymous memory. However,
assume we end up pinning (R/O long-term) a pagecache page or the shared
zeropage inside a writable private ("COW") mapping. The next write access
will trigger a write-fault and replace the pinned page by an exclusive
anonymous page in the process page tables to break COW: the pinned page no
longer corresponds to the page mapped into the process' page table.
Now that FAULT_FLAG_UNSHARE can break COW on anything mapped into a
COW mapping, let's properly break COW first before R/O long-term
pinning something that's not an exclusive anon page inside a COW
mapping. FAULT_FLAG_UNSHARE will break COW and map an exclusive anon page
instead that can get pinned safely.
With this change, we can stop using FOLL_FORCE|FOLL_WRITE for reliable
R/O long-term pinning in COW mappings.
With this change, the new R/O long-term pinning tests for non-anonymous
memory succeed:
# [RUN] R/O longterm GUP pin ... with shared zeropage
ok 151 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd
ok 152 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with tmpfile
ok 153 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with huge zeropage
ok 154 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (2048 kB)
ok 155 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (1048576 kB)
ok 156 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with shared zeropage
ok 157 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd
ok 158 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with tmpfile
ok 159 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with huge zeropage
ok 160 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (2048 kB)
ok 161 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (1048576 kB)
ok 162 Longterm R/O pin is reliable
Note 1: We don't care about short-term R/O-pinning, because they have
snapshot semantics: they are not supposed to observe modifications that
happen after pinning.
As one example, assume we start direct I/O to read from a page and store
page content into a file: modifications to page content after starting
direct I/O are not guaranteed to end up in the file. So even if we'd pin
the shared zeropage, the end result would be as expected -- getting zeroes
stored to the file.
Note 2: For shared mappings we'll now always fallback to the slow path to
lookup the VMA when R/O long-term pining. While that's the necessary price
we have to pay right now, it's actually not that bad in practice: most
FOLL_LONGTERM users already specify FOLL_WRITE, for example, along with
FOLL_FORCE because they tried dealing with COW mappings correctly ...
Note 3: For users that use FOLL_LONGTERM right now without FOLL_WRITE,
such as VFIO, we'd now no longer pin the shared zeropage. Instead, we'd
populate exclusive anon pages that we can pin. There was a concern that
this could affect the memlock limit of existing setups.
For example, a VM running with VFIO could run into the memlock limit and
fail to run. However, we essentially had the same behavior already in
commit 17839856fd58 ("gup: document and work around "COW can break either
way" issue") which got merged into some enterprise distros, and there were
not any such complaints. So most probably, we're fine.
Link: https://lkml.kernel.org/r/20221116102659.70287-10-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-16 10:26:48 +00:00
|
|
|
if (!pte_write(pte) && gup_must_unshare(NULL, flags, &folio->page)) {
|
2022-05-10 01:20:45 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
*nr += refs;
|
2021-12-22 21:38:30 +00:00
|
|
|
folio_set_referenced(folio);
|
2019-07-12 03:57:28 +00:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
unsigned int pdshift, unsigned long end, unsigned int flags,
|
2019-07-12 03:57:28 +00:00
|
|
|
struct page **pages, int *nr)
|
|
|
|
{
|
|
|
|
pte_t *ptep;
|
|
|
|
unsigned long sz = 1UL << hugepd_shift(hugepd);
|
|
|
|
unsigned long next;
|
|
|
|
|
|
|
|
ptep = hugepte_offset(hugepd, addr, pdshift);
|
|
|
|
do {
|
|
|
|
next = hugepte_addr_end(addr, end, sz);
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
if (!gup_hugepte(ptep, sz, addr, end, flags, pages, nr))
|
2019-07-12 03:57:28 +00:00
|
|
|
return 0;
|
|
|
|
} while (ptep++, addr = next, addr != end);
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
static inline int gup_huge_pd(hugepd_t hugepd, unsigned long addr,
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
unsigned int pdshift, unsigned long end, unsigned int flags,
|
2019-07-12 03:57:28 +00:00
|
|
|
struct page **pages, int *nr)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_ARCH_HAS_HUGEPD */
|
|
|
|
|
2014-10-09 22:29:14 +00:00
|
|
|
static int gup_huge_pmd(pmd_t orig, pmd_t *pmdp, unsigned long addr,
|
mm/gup: fix a misnamed "write" argument, and a related bug
In several routines, the "flags" argument is incorrectly named "write".
Change it to "flags".
Also, in one place, the misnaming led to an actual bug:
"flags & FOLL_WRITE" is required, rather than just "flags".
(That problem was flagged by krobot, in v1 of this patch.)
Also, change the flags argument from int, to unsigned int.
You can see that this was a simple oversight, because the
calling code passes "flags" to the fifth argument:
gup_pgd_range():
...
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
PGDIR_SHIFT, next, flags, pages, nr))
...which, until this patch, the callees referred to as "write".
Also, change two lines to avoid checkpatch line length
complaints, and another line to fix another oversight
that checkpatch called out: missing "int" on pdshift.
Link: http://lkml.kernel.org/r/20191014184639.1512873-3-jhubbard@nvidia.com
Fixes: b798bec4741b ("mm/gup: change write parameter to flags in fast walk")
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Kirill A. Shutemov <kirill@shutemov.name>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-10-19 03:19:53 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
2021-12-22 21:57:23 +00:00
|
|
|
struct page *page;
|
|
|
|
struct folio *folio;
|
2014-10-09 22:29:14 +00:00
|
|
|
int refs;
|
|
|
|
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!pmd_access_permitted(orig, flags & FOLL_WRITE))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
2019-05-14 00:17:14 +00:00
|
|
|
if (pmd_devmap(orig)) {
|
|
|
|
if (unlikely(flags & FOLL_LONGTERM))
|
|
|
|
return 0;
|
2020-04-02 04:05:14 +00:00
|
|
|
return __gup_device_huge_pmd(orig, pmdp, addr, end, flags,
|
|
|
|
pages, nr);
|
2019-05-14 00:17:14 +00:00
|
|
|
}
|
2017-03-16 15:26:53 +00:00
|
|
|
|
2022-01-07 18:25:55 +00:00
|
|
|
page = nth_page(pmd_page(orig), (addr & ~PMD_MASK) >> PAGE_SHIFT);
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
refs = record_subpages(page, addr, end, pages + *nr);
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2021-12-22 21:57:23 +00:00
|
|
|
folio = try_grab_folio(page, refs, flags);
|
|
|
|
if (!folio)
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (unlikely(pmd_val(orig) != pmd_val(*pmdp))) {
|
2021-12-22 21:57:23 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: reliable R/O long-term pinning in COW mappings
We already support reliable R/O pinning of anonymous memory. However,
assume we end up pinning (R/O long-term) a pagecache page or the shared
zeropage inside a writable private ("COW") mapping. The next write access
will trigger a write-fault and replace the pinned page by an exclusive
anonymous page in the process page tables to break COW: the pinned page no
longer corresponds to the page mapped into the process' page table.
Now that FAULT_FLAG_UNSHARE can break COW on anything mapped into a
COW mapping, let's properly break COW first before R/O long-term
pinning something that's not an exclusive anon page inside a COW
mapping. FAULT_FLAG_UNSHARE will break COW and map an exclusive anon page
instead that can get pinned safely.
With this change, we can stop using FOLL_FORCE|FOLL_WRITE for reliable
R/O long-term pinning in COW mappings.
With this change, the new R/O long-term pinning tests for non-anonymous
memory succeed:
# [RUN] R/O longterm GUP pin ... with shared zeropage
ok 151 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd
ok 152 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with tmpfile
ok 153 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with huge zeropage
ok 154 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (2048 kB)
ok 155 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (1048576 kB)
ok 156 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with shared zeropage
ok 157 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd
ok 158 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with tmpfile
ok 159 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with huge zeropage
ok 160 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (2048 kB)
ok 161 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (1048576 kB)
ok 162 Longterm R/O pin is reliable
Note 1: We don't care about short-term R/O-pinning, because they have
snapshot semantics: they are not supposed to observe modifications that
happen after pinning.
As one example, assume we start direct I/O to read from a page and store
page content into a file: modifications to page content after starting
direct I/O are not guaranteed to end up in the file. So even if we'd pin
the shared zeropage, the end result would be as expected -- getting zeroes
stored to the file.
Note 2: For shared mappings we'll now always fallback to the slow path to
lookup the VMA when R/O long-term pining. While that's the necessary price
we have to pay right now, it's actually not that bad in practice: most
FOLL_LONGTERM users already specify FOLL_WRITE, for example, along with
FOLL_FORCE because they tried dealing with COW mappings correctly ...
Note 3: For users that use FOLL_LONGTERM right now without FOLL_WRITE,
such as VFIO, we'd now no longer pin the shared zeropage. Instead, we'd
populate exclusive anon pages that we can pin. There was a concern that
this could affect the memlock limit of existing setups.
For example, a VM running with VFIO could run into the memlock limit and
fail to run. However, we essentially had the same behavior already in
commit 17839856fd58 ("gup: document and work around "COW can break either
way" issue") which got merged into some enterprise distros, and there were
not any such complaints. So most probably, we're fine.
Link: https://lkml.kernel.org/r/20221116102659.70287-10-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-16 10:26:48 +00:00
|
|
|
if (!pmd_write(orig) && gup_must_unshare(NULL, flags, &folio->page)) {
|
2022-05-10 01:20:45 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
*nr += refs;
|
2021-12-22 21:57:23 +00:00
|
|
|
folio_set_referenced(folio);
|
2014-10-09 22:29:14 +00:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int gup_huge_pud(pud_t orig, pud_t *pudp, unsigned long addr,
|
2020-04-02 04:05:14 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
|
|
|
struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
2021-12-22 23:07:47 +00:00
|
|
|
struct page *page;
|
|
|
|
struct folio *folio;
|
2014-10-09 22:29:14 +00:00
|
|
|
int refs;
|
|
|
|
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!pud_access_permitted(orig, flags & FOLL_WRITE))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
2019-05-14 00:17:14 +00:00
|
|
|
if (pud_devmap(orig)) {
|
|
|
|
if (unlikely(flags & FOLL_LONGTERM))
|
|
|
|
return 0;
|
2020-04-02 04:05:14 +00:00
|
|
|
return __gup_device_huge_pud(orig, pudp, addr, end, flags,
|
|
|
|
pages, nr);
|
2019-05-14 00:17:14 +00:00
|
|
|
}
|
2017-03-16 15:26:53 +00:00
|
|
|
|
2022-01-07 18:25:55 +00:00
|
|
|
page = nth_page(pud_page(orig), (addr & ~PUD_MASK) >> PAGE_SHIFT);
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
refs = record_subpages(page, addr, end, pages + *nr);
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2021-12-22 23:07:47 +00:00
|
|
|
folio = try_grab_folio(page, refs, flags);
|
|
|
|
if (!folio)
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (unlikely(pud_val(orig) != pud_val(*pudp))) {
|
2021-12-22 23:07:47 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: reliable R/O long-term pinning in COW mappings
We already support reliable R/O pinning of anonymous memory. However,
assume we end up pinning (R/O long-term) a pagecache page or the shared
zeropage inside a writable private ("COW") mapping. The next write access
will trigger a write-fault and replace the pinned page by an exclusive
anonymous page in the process page tables to break COW: the pinned page no
longer corresponds to the page mapped into the process' page table.
Now that FAULT_FLAG_UNSHARE can break COW on anything mapped into a
COW mapping, let's properly break COW first before R/O long-term
pinning something that's not an exclusive anon page inside a COW
mapping. FAULT_FLAG_UNSHARE will break COW and map an exclusive anon page
instead that can get pinned safely.
With this change, we can stop using FOLL_FORCE|FOLL_WRITE for reliable
R/O long-term pinning in COW mappings.
With this change, the new R/O long-term pinning tests for non-anonymous
memory succeed:
# [RUN] R/O longterm GUP pin ... with shared zeropage
ok 151 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd
ok 152 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with tmpfile
ok 153 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with huge zeropage
ok 154 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (2048 kB)
ok 155 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP pin ... with memfd hugetlb (1048576 kB)
ok 156 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with shared zeropage
ok 157 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd
ok 158 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with tmpfile
ok 159 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with huge zeropage
ok 160 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (2048 kB)
ok 161 Longterm R/O pin is reliable
# [RUN] R/O longterm GUP-fast pin ... with memfd hugetlb (1048576 kB)
ok 162 Longterm R/O pin is reliable
Note 1: We don't care about short-term R/O-pinning, because they have
snapshot semantics: they are not supposed to observe modifications that
happen after pinning.
As one example, assume we start direct I/O to read from a page and store
page content into a file: modifications to page content after starting
direct I/O are not guaranteed to end up in the file. So even if we'd pin
the shared zeropage, the end result would be as expected -- getting zeroes
stored to the file.
Note 2: For shared mappings we'll now always fallback to the slow path to
lookup the VMA when R/O long-term pining. While that's the necessary price
we have to pay right now, it's actually not that bad in practice: most
FOLL_LONGTERM users already specify FOLL_WRITE, for example, along with
FOLL_FORCE because they tried dealing with COW mappings correctly ...
Note 3: For users that use FOLL_LONGTERM right now without FOLL_WRITE,
such as VFIO, we'd now no longer pin the shared zeropage. Instead, we'd
populate exclusive anon pages that we can pin. There was a concern that
this could affect the memlock limit of existing setups.
For example, a VM running with VFIO could run into the memlock limit and
fail to run. However, we essentially had the same behavior already in
commit 17839856fd58 ("gup: document and work around "COW can break either
way" issue") which got merged into some enterprise distros, and there were
not any such complaints. So most probably, we're fine.
Link: https://lkml.kernel.org/r/20221116102659.70287-10-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-16 10:26:48 +00:00
|
|
|
if (!pud_write(orig) && gup_must_unshare(NULL, flags, &folio->page)) {
|
2022-05-10 01:20:45 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
*nr += refs;
|
2021-12-22 23:07:47 +00:00
|
|
|
folio_set_referenced(folio);
|
2014-10-09 22:29:14 +00:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2014-11-05 16:27:40 +00:00
|
|
|
static int gup_huge_pgd(pgd_t orig, pgd_t *pgdp, unsigned long addr,
|
2019-05-14 00:17:07 +00:00
|
|
|
unsigned long end, unsigned int flags,
|
2014-11-05 16:27:40 +00:00
|
|
|
struct page **pages, int *nr)
|
|
|
|
{
|
|
|
|
int refs;
|
2021-12-23 03:30:29 +00:00
|
|
|
struct page *page;
|
|
|
|
struct folio *folio;
|
2014-11-05 16:27:40 +00:00
|
|
|
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!pgd_access_permitted(orig, flags & FOLL_WRITE))
|
2014-11-05 16:27:40 +00:00
|
|
|
return 0;
|
|
|
|
|
2017-03-16 15:26:53 +00:00
|
|
|
BUILD_BUG_ON(pgd_devmap(orig));
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
|
2022-01-07 18:25:55 +00:00
|
|
|
page = nth_page(pgd_page(orig), (addr & ~PGDIR_MASK) >> PAGE_SHIFT);
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
refs = record_subpages(page, addr, end, pages + *nr);
|
2014-11-05 16:27:40 +00:00
|
|
|
|
2021-12-23 03:30:29 +00:00
|
|
|
folio = try_grab_folio(page, refs, flags);
|
|
|
|
if (!folio)
|
2014-11-05 16:27:40 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (unlikely(pgd_val(orig) != pgd_val(*pgdp))) {
|
2021-12-23 03:30:29 +00:00
|
|
|
gup_put_folio(folio, refs, flags);
|
2014-11-05 16:27:40 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
mm/gup: factor out duplicate code from four routines
Patch series "mm/gup: prereqs to track dma-pinned pages: FOLL_PIN", v12.
Overview:
This is a prerequisite to solving the problem of proper interactions
between file-backed pages, and [R]DMA activities, as discussed in [1],
[2], [3], and in a remarkable number of email threads since about
2017. :)
A new internal gup flag, FOLL_PIN is introduced, and thoroughly
documented in the last patch's Documentation/vm/pin_user_pages.rst.
I believe that this will provide a good starting point for doing the
layout lease work that Ira Weiny has been working on. That's because
these new wrapper functions provide a clean, constrained, systematically
named set of functionality that, again, is required in order to even
know if a page is "dma-pinned".
In contrast to earlier approaches, the page tracking can be
incrementally applied to the kernel call sites that, until now, have
been simply calling get_user_pages() ("gup"). In other words, opt-in by
changing from this:
get_user_pages() (sets FOLL_GET)
put_page()
to this:
pin_user_pages() (sets FOLL_PIN)
unpin_user_page()
Testing:
* I've done some overall kernel testing (LTP, and a few other goodies),
and some directed testing to exercise some of the changes. And as you
can see, gup_benchmark is enhanced to exercise this. Basically, I've
been able to runtime test the core get_user_pages() and
pin_user_pages() and related routines, but not so much on several of
the call sites--but those are generally just a couple of lines
changed, each.
Not much of the kernel is actually using this, which on one hand
reduces risk quite a lot. But on the other hand, testing coverage
is low. So I'd love it if, in particular, the Infiniband and PowerPC
folks could do a smoke test of this series for me.
Runtime testing for the call sites so far is pretty light:
* io_uring: Some directed tests from liburing exercise this, and
they pass.
* process_vm_access.c: A small directed test passes.
* gup_benchmark: the enhanced version hits the new gup.c code, and
passes.
* infiniband: Ran rdma-core tests: rdma-core/build/bin/run_tests.py
* VFIO: compiles (I'm vowing to set up a run time test soon, but it's
not ready just yet)
* powerpc: it compiles...
* drm/via: compiles...
* goldfish: compiles...
* net/xdp: compiles...
* media/v4l2: compiles...
[1] Some slow progress on get_user_pages() (Apr 2, 2019): https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018): https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018): https://lwn.net/Articles/753027/
This patch (of 22):
There are four locations in gup.c that have a fair amount of code
duplication. This means that changing one requires making the same
changes in four places, not to mention reading the same code four times,
and wondering if there are subtle differences.
Factor out the common code into static functions, thus reducing the
overall line count and the code's complexity.
Also, take the opportunity to slightly improve the efficiency of the
error cases, by doing a mass subtraction of the refcount, surrounded by
get_page()/put_page().
Also, further simplify (slightly), by waiting until the the successful
end of each routine, to increment *nr.
Link: http://lkml.kernel.org/r/20200107224558.2362728-2-jhubbard@nvidia.com
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jérôme Glisse <jglisse@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Björn Töpel <bjorn.topel@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Leon Romanovsky <leonro@mellanox.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-01-31 06:12:17 +00:00
|
|
|
*nr += refs;
|
2021-12-23 03:30:29 +00:00
|
|
|
folio_set_referenced(folio);
|
2014-11-05 16:27:40 +00:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
static int gup_pmd_range(pud_t *pudp, pud_t pud, unsigned long addr, unsigned long end,
|
2019-05-14 00:17:07 +00:00
|
|
|
unsigned int flags, struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
|
|
|
unsigned long next;
|
|
|
|
pmd_t *pmdp;
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
pmdp = pmd_offset_lockless(pudp, pud, addr);
|
2014-10-09 22:29:14 +00:00
|
|
|
do {
|
2020-11-26 16:21:30 +00:00
|
|
|
pmd_t pmd = pmdp_get_lockless(pmdp);
|
2014-10-09 22:29:14 +00:00
|
|
|
|
|
|
|
next = pmd_addr_end(addr, end);
|
mm: thp: check pmd migration entry in common path
When THP migration is being used, memory management code needs to handle
pmd migration entries properly. This patch uses !pmd_present() or
is_swap_pmd() (depending on whether pmd_none() needs separate code or
not) to check pmd migration entries at the places where a pmd entry is
present.
Since pmd-related code uses split_huge_page(), split_huge_pmd(),
pmd_trans_huge(), pmd_trans_unstable(), or
pmd_none_or_trans_huge_or_clear_bad(), this patch:
1. adds pmd migration entry split code in split_huge_pmd(),
2. takes care of pmd migration entries whenever pmd_trans_huge() is present,
3. makes pmd_none_or_trans_huge_or_clear_bad() pmd migration entry aware.
Since split_huge_page() uses split_huge_pmd() and pmd_trans_unstable()
is equivalent to pmd_none_or_trans_huge_or_clear_bad(), we do not change
them.
Until this commit, a pmd entry should be:
1. pointing to a pte page,
2. is_swap_pmd(),
3. pmd_trans_huge(),
4. pmd_devmap(), or
5. pmd_none().
Signed-off-by: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: David Nellans <dnellans@nvidia.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-09-08 23:11:01 +00:00
|
|
|
if (!pmd_present(pmd))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
2019-02-12 23:35:58 +00:00
|
|
|
if (unlikely(pmd_trans_huge(pmd) || pmd_huge(pmd) ||
|
|
|
|
pmd_devmap(pmd))) {
|
2022-08-25 16:46:58 +00:00
|
|
|
if (pmd_protnone(pmd) &&
|
|
|
|
!gup_can_follow_protnone(flags))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!gup_huge_pmd(pmd, pmdp, addr, next, flags,
|
2014-10-09 22:29:14 +00:00
|
|
|
pages, nr))
|
|
|
|
return 0;
|
|
|
|
|
2014-11-05 16:27:40 +00:00
|
|
|
} else if (unlikely(is_hugepd(__hugepd(pmd_val(pmd))))) {
|
|
|
|
/*
|
|
|
|
* architecture have different format for hugetlbfs
|
|
|
|
* pmd format and THP pmd format
|
|
|
|
*/
|
|
|
|
if (!gup_huge_pd(__hugepd(pmd_val(pmd)), addr,
|
2019-05-14 00:17:07 +00:00
|
|
|
PMD_SHIFT, next, flags, pages, nr))
|
2014-11-05 16:27:40 +00:00
|
|
|
return 0;
|
2022-09-07 18:01:43 +00:00
|
|
|
} else if (!gup_pte_range(pmd, pmdp, addr, next, flags, pages, nr))
|
2018-04-05 23:24:18 +00:00
|
|
|
return 0;
|
2014-10-09 22:29:14 +00:00
|
|
|
} while (pmdp++, addr = next, addr != end);
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
static int gup_pud_range(p4d_t *p4dp, p4d_t p4d, unsigned long addr, unsigned long end,
|
2019-05-14 00:17:07 +00:00
|
|
|
unsigned int flags, struct page **pages, int *nr)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
|
|
|
unsigned long next;
|
|
|
|
pud_t *pudp;
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
pudp = pud_offset_lockless(p4dp, p4d, addr);
|
2014-10-09 22:29:14 +00:00
|
|
|
do {
|
2014-12-07 20:41:33 +00:00
|
|
|
pud_t pud = READ_ONCE(*pudp);
|
2014-10-09 22:29:14 +00:00
|
|
|
|
|
|
|
next = pud_addr_end(addr, end);
|
2020-01-31 06:12:10 +00:00
|
|
|
if (unlikely(!pud_present(pud)))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
2022-12-07 06:00:53 +00:00
|
|
|
if (unlikely(pud_huge(pud) || pud_devmap(pud))) {
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!gup_huge_pud(pud, pudp, addr, next, flags,
|
2014-11-05 16:27:40 +00:00
|
|
|
pages, nr))
|
|
|
|
return 0;
|
|
|
|
} else if (unlikely(is_hugepd(__hugepd(pud_val(pud))))) {
|
|
|
|
if (!gup_huge_pd(__hugepd(pud_val(pud)), addr,
|
2019-05-14 00:17:07 +00:00
|
|
|
PUD_SHIFT, next, flags, pages, nr))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
} else if (!gup_pmd_range(pudp, pud, addr, next, flags, pages, nr))
|
2014-10-09 22:29:14 +00:00
|
|
|
return 0;
|
|
|
|
} while (pudp++, addr = next, addr != end);
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
static int gup_p4d_range(pgd_t *pgdp, pgd_t pgd, unsigned long addr, unsigned long end,
|
2019-05-14 00:17:07 +00:00
|
|
|
unsigned int flags, struct page **pages, int *nr)
|
2017-03-09 14:24:07 +00:00
|
|
|
{
|
|
|
|
unsigned long next;
|
|
|
|
p4d_t *p4dp;
|
|
|
|
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
p4dp = p4d_offset_lockless(pgdp, pgd, addr);
|
2017-03-09 14:24:07 +00:00
|
|
|
do {
|
|
|
|
p4d_t p4d = READ_ONCE(*p4dp);
|
|
|
|
|
|
|
|
next = p4d_addr_end(addr, end);
|
|
|
|
if (p4d_none(p4d))
|
|
|
|
return 0;
|
|
|
|
BUILD_BUG_ON(p4d_huge(p4d));
|
|
|
|
if (unlikely(is_hugepd(__hugepd(p4d_val(p4d))))) {
|
|
|
|
if (!gup_huge_pd(__hugepd(p4d_val(p4d)), addr,
|
2019-05-14 00:17:07 +00:00
|
|
|
P4D_SHIFT, next, flags, pages, nr))
|
2017-03-09 14:24:07 +00:00
|
|
|
return 0;
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
} else if (!gup_pud_range(p4dp, p4d, addr, next, flags, pages, nr))
|
2017-03-09 14:24:07 +00:00
|
|
|
return 0;
|
|
|
|
} while (p4dp++, addr = next, addr != end);
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2017-09-08 21:56:03 +00:00
|
|
|
static void gup_pgd_range(unsigned long addr, unsigned long end,
|
2019-05-14 00:17:07 +00:00
|
|
|
unsigned int flags, struct page **pages, int *nr)
|
2017-09-08 21:56:03 +00:00
|
|
|
{
|
|
|
|
unsigned long next;
|
|
|
|
pgd_t *pgdp;
|
|
|
|
|
|
|
|
pgdp = pgd_offset(current->mm, addr);
|
|
|
|
do {
|
|
|
|
pgd_t pgd = READ_ONCE(*pgdp);
|
|
|
|
|
|
|
|
next = pgd_addr_end(addr, end);
|
|
|
|
if (pgd_none(pgd))
|
|
|
|
return;
|
|
|
|
if (unlikely(pgd_huge(pgd))) {
|
2019-05-14 00:17:07 +00:00
|
|
|
if (!gup_huge_pgd(pgd, pgdp, addr, next, flags,
|
2017-09-08 21:56:03 +00:00
|
|
|
pages, nr))
|
|
|
|
return;
|
|
|
|
} else if (unlikely(is_hugepd(__hugepd(pgd_val(pgd))))) {
|
|
|
|
if (!gup_huge_pd(__hugepd(pgd_val(pgd)), addr,
|
2019-05-14 00:17:07 +00:00
|
|
|
PGDIR_SHIFT, next, flags, pages, nr))
|
2017-09-08 21:56:03 +00:00
|
|
|
return;
|
mm/gup: fix gup_fast with dynamic page table folding
Currently to make sure that every page table entry is read just once
gup_fast walks perform READ_ONCE and pass pXd value down to the next
gup_pXd_range function by value e.g.:
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
...
pudp = pud_offset(&p4d, addr);
This function passes a reference on that local value copy to pXd_offset,
and might get the very same pointer in return. This happens when the
level is folded (on most arches), and that pointer should not be
iterated.
On s390 due to the fact that each task might have different 5,4 or
3-level address translation and hence different levels folded the logic
is more complex and non-iteratable pointer to a local copy leads to
severe problems.
Here is an example of what happens with gup_fast on s390, for a task
with 3-level paging, crossing a 2 GB pud boundary:
// addr = 0x1007ffff000, end = 0x10080001000
static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
unsigned int flags, struct page **pages, int *nr)
{
unsigned long next;
pud_t *pudp;
// pud_offset returns &p4d itself (a pointer to a value on stack)
pudp = pud_offset(&p4d, addr);
do {
// on second iteratation reading "random" stack value
pud_t pud = READ_ONCE(*pudp);
// next = 0x10080000000, due to PUD_SIZE/MASK != PGDIR_SIZE/MASK on s390
next = pud_addr_end(addr, end);
...
} while (pudp++, addr = next, addr != end); // pudp++ iterating over stack
return 1;
}
This happens since s390 moved to common gup code with commit
d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") and
commit 1a42010cdc26 ("s390/mm: convert to the generic
get_user_pages_fast code").
s390 tried to mimic static level folding by changing pXd_offset
primitives to always calculate top level page table offset in pgd_offset
and just return the value passed when pXd_offset has to act as folded.
What is crucial for gup_fast and what has been overlooked is that
PxD_SIZE/MASK and thus pXd_addr_end should also change correspondingly.
And the latter is not possible with dynamic folding.
To fix the issue in addition to pXd values pass original pXdp pointers
down to gup_pXd_range functions. And introduce pXd_offset_lockless
helpers, which take an additional pXd entry value parameter. This has
already been discussed in
https://lkml.kernel.org/r/20190418100218.0a4afd51@mschwideX1
Fixes: 1a42010cdc26 ("s390/mm: convert to the generic get_user_pages_fast code")
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Claudio Imbrenda <imbrenda@linux.ibm.com>
Cc: <stable@vger.kernel.org> [5.2+]
Link: https://lkml.kernel.org/r/patch.git-943f1e5dcff2.your-ad-here.call-01599856292-ext-8676@work.hours
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-09-26 04:19:10 +00:00
|
|
|
} else if (!gup_p4d_range(pgdp, pgd, addr, next, flags, pages, nr))
|
2017-09-08 21:56:03 +00:00
|
|
|
return;
|
|
|
|
} while (pgdp++, addr = next, addr != end);
|
|
|
|
}
|
2019-07-12 03:57:21 +00:00
|
|
|
#else
|
|
|
|
static inline void gup_pgd_range(unsigned long addr, unsigned long end,
|
|
|
|
unsigned int flags, struct page **pages, int *nr)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_HAVE_FAST_GUP */
|
2017-09-08 21:56:03 +00:00
|
|
|
|
|
|
|
#ifndef gup_fast_permitted
|
|
|
|
/*
|
2020-06-08 04:40:55 +00:00
|
|
|
* Check if it's allowed to use get_user_pages_fast_only() for the range, or
|
2017-09-08 21:56:03 +00:00
|
|
|
* we need to fall back to the slow version:
|
|
|
|
*/
|
2019-07-12 03:56:45 +00:00
|
|
|
static bool gup_fast_permitted(unsigned long start, unsigned long end)
|
2017-09-08 21:56:03 +00:00
|
|
|
{
|
2019-07-12 03:56:45 +00:00
|
|
|
return true;
|
2017-09-08 21:56:03 +00:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2020-12-15 03:05:41 +00:00
|
|
|
static unsigned long lockless_pages_from_mm(unsigned long start,
|
|
|
|
unsigned long end,
|
|
|
|
unsigned int gup_flags,
|
|
|
|
struct page **pages)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
int nr_pinned = 0;
|
2020-12-15 03:05:44 +00:00
|
|
|
unsigned seq;
|
2020-12-15 03:05:41 +00:00
|
|
|
|
|
|
|
if (!IS_ENABLED(CONFIG_HAVE_FAST_GUP) ||
|
|
|
|
!gup_fast_permitted(start, end))
|
|
|
|
return 0;
|
|
|
|
|
2020-12-15 03:05:44 +00:00
|
|
|
if (gup_flags & FOLL_PIN) {
|
|
|
|
seq = raw_read_seqcount(¤t->mm->write_protect_seq);
|
|
|
|
if (seq & 1)
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2020-12-15 03:05:41 +00:00
|
|
|
/*
|
|
|
|
* Disable interrupts. The nested form is used, in order to allow full,
|
|
|
|
* general purpose use of this routine.
|
|
|
|
*
|
|
|
|
* With interrupts disabled, we block page table pages from being freed
|
|
|
|
* from under us. See struct mmu_table_batch comments in
|
|
|
|
* include/asm-generic/tlb.h for more details.
|
|
|
|
*
|
|
|
|
* We do not adopt an rcu_read_lock() here as we also want to block IPIs
|
|
|
|
* that come from THPs splitting.
|
|
|
|
*/
|
|
|
|
local_irq_save(flags);
|
|
|
|
gup_pgd_range(start, end, gup_flags, pages, &nr_pinned);
|
|
|
|
local_irq_restore(flags);
|
2020-12-15 03:05:44 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* When pinning pages for DMA there could be a concurrent write protect
|
|
|
|
* from fork() via copy_page_range(), in this case always fail fast GUP.
|
|
|
|
*/
|
|
|
|
if (gup_flags & FOLL_PIN) {
|
|
|
|
if (read_seqcount_retry(¤t->mm->write_protect_seq, seq)) {
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
unpin_user_pages_lockless(pages, nr_pinned);
|
2020-12-15 03:05:44 +00:00
|
|
|
return 0;
|
mm/gup: sanity-check with CONFIG_DEBUG_VM that anonymous pages are exclusive when (un)pinning
Let's verify when (un)pinning anonymous pages that we always deal with
exclusive anonymous pages, which guarantees that we'll have a reliable
PIN, meaning that we cannot end up with the GUP pin being inconsistent
with he pages mapped into the page tables due to a COW triggered by a
write fault.
When pinning pages, after conditionally triggering GUP unsharing of
possibly shared anonymous pages, we should always only see exclusive
anonymous pages. Note that anonymous pages that are mapped writable must
be marked exclusive, otherwise we'd have a BUG.
When pinning during ordinary GUP, simply add a check after our conditional
GUP-triggered unsharing checks. As we know exactly how the page is
mapped, we know exactly in which page we have to check for
PageAnonExclusive().
When pinning via GUP-fast we have to be careful, because we can race with
fork(): verify only after we made sure via the seqcount that we didn't
race with concurrent fork() that we didn't end up pinning a possibly
shared anonymous page.
Similarly, when unpinning, verify that the pages are still marked as
exclusive: otherwise something turned the pages possibly shared, which can
result in random memory corruptions, which we really want to catch.
With only the pinned pages at hand and not the actual page table entries
we have to be a bit careful: hugetlb pages are always mapped via a single
logical page table entry referencing the head page and PG_anon_exclusive
of the head page applies. Anon THP are a bit more complicated, because we
might have obtained the page reference either via a PMD or a PTE --
depending on the mapping type we either have to check PageAnonExclusive of
the head page (PMD-mapped THP) or the tail page (PTE-mapped THP) applies:
as we don't know and to make our life easier, check that either is set.
Take care to not verify in case we're unpinning during GUP-fast because we
detected concurrent fork(): we might stumble over an anonymous page that
is now shared.
Link: https://lkml.kernel.org/r/20220428083441.37290-18-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Liang Zhang <zhangliang5@huawei.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Oded Gabbay <oded.gabbay@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Pedro Demarchi Gomes <pedrodemargomes@gmail.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-05-10 01:20:45 +00:00
|
|
|
} else {
|
|
|
|
sanity_check_pinned_pages(pages, nr_pinned);
|
2020-12-15 03:05:44 +00:00
|
|
|
}
|
|
|
|
}
|
2020-12-15 03:05:41 +00:00
|
|
|
return nr_pinned;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int internal_get_user_pages_fast(unsigned long start,
|
|
|
|
unsigned long nr_pages,
|
2020-01-31 06:12:54 +00:00
|
|
|
unsigned int gup_flags,
|
|
|
|
struct page **pages)
|
2014-10-09 22:29:14 +00:00
|
|
|
{
|
2020-12-15 03:05:41 +00:00
|
|
|
unsigned long len, end;
|
|
|
|
unsigned long nr_pinned;
|
2023-01-24 20:34:22 +00:00
|
|
|
int locked = 0;
|
2020-12-15 03:05:41 +00:00
|
|
|
int ret;
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2020-01-31 06:12:43 +00:00
|
|
|
if (WARN_ON_ONCE(gup_flags & ~(FOLL_WRITE | FOLL_LONGTERM |
|
2020-06-03 22:56:30 +00:00
|
|
|
FOLL_FORCE | FOLL_PIN | FOLL_GET |
|
2022-10-21 17:41:09 +00:00
|
|
|
FOLL_FAST_ONLY | FOLL_NOFAULT |
|
|
|
|
FOLL_PCI_P2PDMA)))
|
2019-07-12 03:57:25 +00:00
|
|
|
return -EINVAL;
|
|
|
|
|
2021-06-29 02:36:40 +00:00
|
|
|
if (gup_flags & FOLL_PIN)
|
|
|
|
mm_set_has_pinned_flag(¤t->mm->flags);
|
2020-09-25 22:25:57 +00:00
|
|
|
|
2020-06-03 22:56:40 +00:00
|
|
|
if (!(gup_flags & FOLL_FAST_ONLY))
|
2020-06-09 04:33:47 +00:00
|
|
|
might_lock_read(¤t->mm->mmap_lock);
|
2020-06-03 22:56:40 +00:00
|
|
|
|
2019-07-12 03:56:41 +00:00
|
|
|
start = untagged_addr(start) & PAGE_MASK;
|
2020-12-15 03:05:41 +00:00
|
|
|
len = nr_pages << PAGE_SHIFT;
|
|
|
|
if (check_add_overflow(start, len, &end))
|
2023-06-19 18:27:25 +00:00
|
|
|
return -EOVERFLOW;
|
x86-64: make access_ok() independent of LAM
The linear address masking (LAM) code made access_ok() more complicated,
in that it now needs to untag the address in order to verify the access
range. See commit 74c228d20a51 ("x86/uaccess: Provide untagged_addr()
and remove tags before address check").
We were able to avoid that overhead in the get_user/put_user code paths
by simply using the sign bit for the address check, and depending on the
GP fault if the address was non-canonical, which made it all independent
of LAM.
And we can do the same thing for access_ok(): simply check that the user
pointer range has the high bit clear. No need to bother with any
address bit masking.
In fact, we can go a bit further, and just check the starting address
for known small accesses ranges: any accesses that overflow will still
be in the non-canonical area and will still GP fault.
To still make syzkaller catch any potentially unchecked user addresses,
we'll continue to warn about GP faults that are caused by accesses in
the non-canonical range. But we'll limit that to purely "high bit set
and past the one-page 'slop' area".
We could probably just do that "check only starting address" for any
arbitrary range size: realistically all kernel accesses to user space
will be done starting at the low address. But let's leave that kind of
optimization for later. As it is, this already allows us to generate
simpler code and not worry about any tag bits in the address.
The one thing to look out for is the GUP address check: instead of
actually copying data in the virtual address range (and thus bad
addresses being caught by the GP fault), GUP will look up the page
tables manually. As a result, the page table limits need to be checked,
and that was previously implicitly done by the access_ok().
With the relaxed access_ok() check, we need to just do an explicit check
for TASK_SIZE_MAX in the GUP code instead. The GUP code already needs
to do the tag bit unmasking anyway, so there this is all very
straightforward, and there are no LAM issues.
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2023-04-28 19:55:10 +00:00
|
|
|
if (end > TASK_SIZE_MAX)
|
|
|
|
return -EFAULT;
|
Remove 'type' argument from access_ok() function
Nobody has actually used the type (VERIFY_READ vs VERIFY_WRITE) argument
of the user address range verification function since we got rid of the
old racy i386-only code to walk page tables by hand.
It existed because the original 80386 would not honor the write protect
bit when in kernel mode, so you had to do COW by hand before doing any
user access. But we haven't supported that in a long time, and these
days the 'type' argument is a purely historical artifact.
A discussion about extending 'user_access_begin()' to do the range
checking resulted this patch, because there is no way we're going to
move the old VERIFY_xyz interface to that model. And it's best done at
the end of the merge window when I've done most of my merges, so let's
just get this done once and for all.
This patch was mostly done with a sed-script, with manual fix-ups for
the cases that weren't of the trivial 'access_ok(VERIFY_xyz' form.
There were a couple of notable cases:
- csky still had the old "verify_area()" name as an alias.
- the iter_iov code had magical hardcoded knowledge of the actual
values of VERIFY_{READ,WRITE} (not that they mattered, since nothing
really used it)
- microblaze used the type argument for a debug printout
but other than those oddities this should be a total no-op patch.
I tried to fix up all architectures, did fairly extensive grepping for
access_ok() uses, and the changes are trivial, but I may have missed
something. Any missed conversion should be trivially fixable, though.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2019-01-04 02:57:57 +00:00
|
|
|
if (unlikely(!access_ok((void __user *)start, len)))
|
get_user_pages_fast(): return -EFAULT on access_ok failure
get_user_pages_fast is supposed to be a faster drop-in equivalent of
get_user_pages. As such, callers expect it to return a negative return
code when passed an invalid address, and never expect it to return 0
when passed a positive number of pages, since its documentation says:
* Returns number of pages pinned. This may be fewer than the number
* requested. If nr_pages is 0 or negative, returns 0. If no pages
* were pinned, returns -errno.
When get_user_pages_fast fall back on get_user_pages this is exactly
what happens. Unfortunately the implementation is inconsistent: it
returns 0 if passed a kernel address, confusing callers: for example,
the following is pretty common but does not appear to do the right thing
with a kernel address:
ret = get_user_pages_fast(addr, 1, writeable, &page);
if (ret < 0)
return ret;
Change get_user_pages_fast to return -EFAULT when supplied a kernel
address to make it match expectations.
All callers have been audited for consistency with the documented
semantics.
Link: http://lkml.kernel.org/r/1522962072-182137-4-git-send-email-mst@redhat.com
Fixes: 5b65c4677a57 ("mm, x86/mm: Fix performance regression in get_user_pages_fast()")
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reported-by: syzbot+6304bf97ef436580fede@syzkaller.appspotmail.com
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Thorsten Leemhuis <regressions@leemhuis.info>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-04-13 22:35:20 +00:00
|
|
|
return -EFAULT;
|
2017-03-16 15:26:54 +00:00
|
|
|
|
2020-12-15 03:05:41 +00:00
|
|
|
nr_pinned = lockless_pages_from_mm(start, end, gup_flags, pages);
|
|
|
|
if (nr_pinned == nr_pages || gup_flags & FOLL_FAST_ONLY)
|
|
|
|
return nr_pinned;
|
2014-10-09 22:29:14 +00:00
|
|
|
|
2020-12-15 03:05:41 +00:00
|
|
|
/* Slow path: try to get the remaining pages with get_user_pages */
|
|
|
|
start += nr_pinned << PAGE_SHIFT;
|
|
|
|
pages += nr_pinned;
|
2023-01-24 20:34:22 +00:00
|
|
|
ret = __gup_longterm_locked(current->mm, start, nr_pages - nr_pinned,
|
|
|
|
pages, NULL, &locked,
|
2023-01-24 20:34:29 +00:00
|
|
|
gup_flags | FOLL_TOUCH | FOLL_UNLOCKABLE);
|
2020-12-15 03:05:41 +00:00
|
|
|
if (ret < 0) {
|
|
|
|
/*
|
|
|
|
* The caller has to unpin the pages we already pinned so
|
|
|
|
* returning -errno is not an option
|
|
|
|
*/
|
|
|
|
if (nr_pinned)
|
|
|
|
return nr_pinned;
|
|
|
|
return ret;
|
2014-10-09 22:29:14 +00:00
|
|
|
}
|
2020-12-15 03:05:41 +00:00
|
|
|
return ret + nr_pinned;
|
2014-10-09 22:29:14 +00:00
|
|
|
}
|
2020-12-15 03:05:41 +00:00
|
|
|
|
2020-06-08 04:40:55 +00:00
|
|
|
/**
|
|
|
|
* get_user_pages_fast_only() - pin user pages in memory
|
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying pin behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long.
|
|
|
|
*
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
* Like get_user_pages_fast() except it's IRQ-safe in that it won't fall back to
|
|
|
|
* the regular GUP.
|
|
|
|
*
|
|
|
|
* If the architecture does not support this function, simply return with no
|
|
|
|
* pages pinned.
|
|
|
|
*
|
|
|
|
* Careful, careful! COW breaking can go either way, so a non-write
|
|
|
|
* access can get ambiguous page results. If you call this function without
|
|
|
|
* 'write' set, you'd better be sure that you're ok with that ambiguity.
|
|
|
|
*/
|
2020-06-08 04:40:55 +00:00
|
|
|
int get_user_pages_fast_only(unsigned long start, int nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages)
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
{
|
|
|
|
/*
|
|
|
|
* Internally (within mm/gup.c), gup fast variants must set FOLL_GET,
|
|
|
|
* because gup fast is always a "pin with a +1 page refcount" request.
|
2020-06-03 22:56:30 +00:00
|
|
|
*
|
|
|
|
* FOLL_FAST_ONLY is required in order to match the API description of
|
|
|
|
* this routine: no fall back to regular ("slow") GUP.
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
*/
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, NULL, NULL, &gup_flags,
|
|
|
|
FOLL_GET | FOLL_FAST_ONLY))
|
|
|
|
return -EINVAL;
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
|
2023-01-24 20:34:32 +00:00
|
|
|
return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
}
|
2020-06-08 04:40:55 +00:00
|
|
|
EXPORT_SYMBOL_GPL(get_user_pages_fast_only);
|
mm/gup: move __get_user_pages_fast() down a few lines in gup.c
Patch series "mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()", v2.
In order to convert the drm/i915 driver from get_user_pages() to
pin_user_pages(), a FOLL_PIN equivalent of __get_user_pages_fast() was
required. That led to refactoring __get_user_pages_fast(), with the
following goals:
1) As above: provide a pin_user_pages*() routine for drm/i915 to call,
in place of __get_user_pages_fast(),
2) Get rid of the gup.c duplicate code for walking page tables with
interrupts disabled. This duplicate code is a minor maintenance
problem anyway.
3) Make it easy for an upcoming patch from Souptick, which aims to
convert __get_user_pages_fast() to use a gup_flags argument, instead
of a bool writeable arg. Also, if this series looks good, we can
ask Souptick to change the name as well, to whatever the consensus
is. My initial recommendation is: get_user_pages_fast_only(), to
match the new pin_user_pages_only().
This patch (of 4):
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should be identical.
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: "Joonas Lahtinen" <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Souptick Joarder <jrdr.linux@gmail.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://lkml.kernel.org/r/20200522051931.54191-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-1-jhubbard@nvidia.com
Link: http://lkml.kernel.org/r/20200519002124.2025955-2-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-06-03 22:56:27 +00:00
|
|
|
|
2020-01-31 06:12:54 +00:00
|
|
|
/**
|
|
|
|
* get_user_pages_fast() - pin user pages in memory
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying pin behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long.
|
2020-01-31 06:12:54 +00:00
|
|
|
*
|
2020-06-09 04:33:54 +00:00
|
|
|
* Attempt to pin user pages in memory without taking mm->mmap_lock.
|
2020-01-31 06:12:54 +00:00
|
|
|
* If not successful, it will fall back to taking the lock and
|
|
|
|
* calling get_user_pages().
|
|
|
|
*
|
|
|
|
* Returns number of pages pinned. This may be fewer than the number requested.
|
|
|
|
* If nr_pages is 0 or negative, returns 0. If no pages were pinned, returns
|
|
|
|
* -errno.
|
|
|
|
*/
|
|
|
|
int get_user_pages_fast(unsigned long start, int nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages)
|
|
|
|
{
|
2020-04-02 04:05:25 +00:00
|
|
|
/*
|
|
|
|
* The caller may or may not have explicitly set FOLL_GET; either way is
|
|
|
|
* OK. However, internally (within mm/gup.c), gup fast variants must set
|
|
|
|
* FOLL_GET, because gup fast is always a "pin with a +1 page refcount"
|
|
|
|
* request.
|
|
|
|
*/
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, NULL, NULL, &gup_flags, FOLL_GET))
|
|
|
|
return -EINVAL;
|
2020-01-31 06:12:54 +00:00
|
|
|
return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
|
|
|
|
}
|
2019-07-12 03:57:21 +00:00
|
|
|
EXPORT_SYMBOL_GPL(get_user_pages_fast);
|
2020-01-31 06:12:54 +00:00
|
|
|
|
|
|
|
/**
|
|
|
|
* pin_user_pages_fast() - pin user pages in memory without taking locks
|
|
|
|
*
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying pin behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
|
|
|
* Should be at least nr_pages long.
|
|
|
|
*
|
|
|
|
* Nearly the same as get_user_pages_fast(), except that FOLL_PIN is set. See
|
|
|
|
* get_user_pages_fast() for documentation on the function arguments, because
|
|
|
|
* the arguments here are identical.
|
|
|
|
*
|
|
|
|
* FOLL_PIN means that the pages must be released via unpin_user_page(). Please
|
2020-04-14 16:48:35 +00:00
|
|
|
* see Documentation/core-api/pin_user_pages.rst for further details.
|
2020-01-31 06:12:54 +00:00
|
|
|
*/
|
|
|
|
int pin_user_pages_fast(unsigned long start, int nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages)
|
|
|
|
{
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, NULL, NULL, &gup_flags, FOLL_PIN))
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
return -EINVAL;
|
|
|
|
return internal_get_user_pages_fast(start, nr_pages, gup_flags, pages);
|
2020-01-31 06:12:54 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(pin_user_pages_fast);
|
|
|
|
|
|
|
|
/**
|
2020-08-12 01:39:01 +00:00
|
|
|
* pin_user_pages_remote() - pin pages of a remote process
|
2020-01-31 06:12:54 +00:00
|
|
|
*
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @mm: mm_struct of target mm
|
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying lookup behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
2022-05-10 01:20:47 +00:00
|
|
|
* Should be at least nr_pages long.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @vmas: array of pointers to vmas corresponding to each page.
|
|
|
|
* Or NULL if the caller does not require them.
|
|
|
|
* @locked: pointer to lock flag indicating whether lock is held and
|
|
|
|
* subsequently whether VM_FAULT_RETRY functionality can be
|
|
|
|
* utilised. Lock must initially be held.
|
|
|
|
*
|
|
|
|
* Nearly the same as get_user_pages_remote(), except that FOLL_PIN is set. See
|
|
|
|
* get_user_pages_remote() for documentation on the function arguments, because
|
|
|
|
* the arguments here are identical.
|
|
|
|
*
|
|
|
|
* FOLL_PIN means that the pages must be released via unpin_user_page(). Please
|
2020-04-14 16:48:35 +00:00
|
|
|
* see Documentation/core-api/pin_user_pages.rst for details.
|
2020-01-31 06:12:54 +00:00
|
|
|
*/
|
2020-08-12 01:39:01 +00:00
|
|
|
long pin_user_pages_remote(struct mm_struct *mm,
|
2020-01-31 06:12:54 +00:00
|
|
|
unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas, int *locked)
|
|
|
|
{
|
2023-01-24 20:34:30 +00:00
|
|
|
int local_locked = 1;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, vmas, locked, &gup_flags,
|
|
|
|
FOLL_PIN | FOLL_TOUCH | FOLL_REMOTE))
|
|
|
|
return 0;
|
2023-01-24 20:34:30 +00:00
|
|
|
return __gup_longterm_locked(mm, start, nr_pages, pages, vmas,
|
|
|
|
locked ? locked : &local_locked,
|
2023-01-24 20:34:26 +00:00
|
|
|
gup_flags);
|
2020-01-31 06:12:54 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(pin_user_pages_remote);
|
|
|
|
|
|
|
|
/**
|
|
|
|
* pin_user_pages() - pin user pages in memory for use by other devices
|
|
|
|
*
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @start: starting user address
|
|
|
|
* @nr_pages: number of pages from start to pin
|
|
|
|
* @gup_flags: flags modifying lookup behaviour
|
|
|
|
* @pages: array that receives pointers to the pages pinned.
|
2022-05-10 01:20:47 +00:00
|
|
|
* Should be at least nr_pages long.
|
mm/gup: track FOLL_PIN pages
Add tracking of pages that were pinned via FOLL_PIN. This tracking is
implemented via overloading of page->_refcount: pins are added by adding
GUP_PIN_COUNTING_BIAS (1024) to the refcount. This provides a fuzzy
indication of pinning, and it can have false positives (and that's OK).
Please see the pre-existing Documentation/core-api/pin_user_pages.rst for
details.
As mentioned in pin_user_pages.rst, callers who effectively set FOLL_PIN
(typically via pin_user_pages*()) are required to ultimately free such
pages via unpin_user_page().
Please also note the limitation, discussed in pin_user_pages.rst under the
"TODO: for 1GB and larger huge pages" section. (That limitation will be
removed in a following patch.)
The effect of a FOLL_PIN flag is similar to that of FOLL_GET, and may be
thought of as "FOLL_GET for DIO and/or RDMA use".
Pages that have been pinned via FOLL_PIN are identifiable via a new
function call:
bool page_maybe_dma_pinned(struct page *page);
What to do in response to encountering such a page, is left to later
patchsets. There is discussion about this in [1], [2], [3], and [4].
This also changes a BUG_ON(), to a WARN_ON(), in follow_page_mask().
[1] Some slow progress on get_user_pages() (Apr 2, 2019):
https://lwn.net/Articles/784574/
[2] DMA and get_user_pages() (LPC: Dec 12, 2018):
https://lwn.net/Articles/774411/
[3] The trouble with get_user_pages() (Apr 30, 2018):
https://lwn.net/Articles/753027/
[4] LWN kernel index: get_user_pages():
https://lwn.net/Kernel/Index/#Memory_management-get_user_pages
[jhubbard@nvidia.com: add kerneldoc]
Link: http://lkml.kernel.org/r/20200307021157.235726-1-jhubbard@nvidia.com
[imbrenda@linux.ibm.com: if pin fails, we need to unpin, a simple put_page will not be enough]
Link: http://lkml.kernel.org/r/20200306132537.783769-2-imbrenda@linux.ibm.com
[akpm@linux-foundation.org: fix put_compound_head defined but not used]
Suggested-by: Jan Kara <jack@suse.cz>
Suggested-by: Jérôme Glisse <jglisse@redhat.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-7-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-04-02 04:05:29 +00:00
|
|
|
* @vmas: array of pointers to vmas corresponding to each page.
|
|
|
|
* Or NULL if the caller does not require them.
|
|
|
|
*
|
|
|
|
* Nearly the same as get_user_pages(), except that FOLL_TOUCH is not set, and
|
|
|
|
* FOLL_PIN is set.
|
|
|
|
*
|
|
|
|
* FOLL_PIN means that the pages must be released via unpin_user_page(). Please
|
2020-04-14 16:48:35 +00:00
|
|
|
* see Documentation/core-api/pin_user_pages.rst for details.
|
2020-01-31 06:12:54 +00:00
|
|
|
*/
|
|
|
|
long pin_user_pages(unsigned long start, unsigned long nr_pages,
|
|
|
|
unsigned int gup_flags, struct page **pages,
|
|
|
|
struct vm_area_struct **vmas)
|
|
|
|
{
|
2023-01-24 20:34:30 +00:00
|
|
|
int locked = 1;
|
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, vmas, NULL, &gup_flags, FOLL_PIN))
|
|
|
|
return 0;
|
2020-08-12 01:39:01 +00:00
|
|
|
return __gup_longterm_locked(current->mm, start, nr_pages,
|
2023-01-24 20:34:30 +00:00
|
|
|
pages, vmas, &locked, gup_flags);
|
2020-01-31 06:12:54 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(pin_user_pages);
|
2020-06-02 04:48:27 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* pin_user_pages_unlocked() is the FOLL_PIN variant of
|
|
|
|
* get_user_pages_unlocked(). Behavior is the same, except that this one sets
|
|
|
|
* FOLL_PIN and rejects FOLL_GET.
|
|
|
|
*/
|
|
|
|
long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
|
|
|
|
struct page **pages, unsigned int gup_flags)
|
|
|
|
{
|
2023-01-24 20:34:22 +00:00
|
|
|
int locked = 0;
|
2020-06-02 04:48:27 +00:00
|
|
|
|
2023-01-24 20:34:26 +00:00
|
|
|
if (!is_valid_gup_args(pages, NULL, NULL, &gup_flags,
|
2023-01-24 20:34:29 +00:00
|
|
|
FOLL_PIN | FOLL_TOUCH | FOLL_UNLOCKABLE))
|
2023-01-24 20:34:26 +00:00
|
|
|
return 0;
|
2022-05-10 01:20:47 +00:00
|
|
|
|
2023-01-24 20:34:22 +00:00
|
|
|
return __gup_longterm_locked(current->mm, start, nr_pages, pages, NULL,
|
|
|
|
&locked, gup_flags);
|
2020-06-02 04:48:27 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(pin_user_pages_unlocked);
|