linux-stable/mm/page_isolation.c
Vlastimil Babka ec6e8c7e03 mm, page_alloc: disable pcplists during memory offline
Memory offlining relies on page isolation to guarantee a forward progress
because pages cannot be reused while they are isolated.  But the page
isolation itself doesn't prevent from races while freed pages are stored
on pcp lists and thus can be reused.  This can be worked around by
repeated draining of pcplists, as done by commit 9683182612
("mm/memory_hotplug: drain per-cpu pages again during memory offline").

David and Michal would prefer that this race was closed in a way that
callers of page isolation who need stronger guarantees don't need to
repeatedly drain.  David suggested disabling pcplists usage completely
during page isolation, instead of repeatedly draining them.

To achieve this without adding special cases in alloc/free fastpath, we
can use the same approach as boot pagesets - when pcp->high is 0, any
pcplist addition will be immediately flushed.

The race can thus be closed by setting pcp->high to 0 and draining
pcplists once, before calling start_isolate_page_range().  The draining
will serialize after processes that already disabled interrupts and read
the old value of pcp->high in free_unref_page_commit(), and processes that
have not yet disabled interrupts, will observe pcp->high == 0 when they
are rescheduled, and skip pcplists.  This guarantees no stray pages on
pcplists in zones where isolation happens.

This patch thus adds zone_pcp_disable() and zone_pcp_enable() functions
that page isolation users can call before start_isolate_page_range() and
after unisolating (or offlining) the isolated pages.

Also, drain_all_pages() is optimized to only execute on cpus where
pcplists are not empty.  The check can however race with a free to pcplist
that has not yet increased the pcp->count from 0 to 1.  Thus make the
drain optionally skip the racy check and drain on all cpus, and use this
option in zone_pcp_disable().

As we have to avoid external updates to high and batch while pcplists are
disabled, we take pcp_batch_high_lock in zone_pcp_disable() and release it
in zone_pcp_enable().  This also synchronizes multiple users of
zone_pcp_disable()/enable().

Currently the only user of this functionality is offline_pages().

[vbabka@suse.cz: add comment, per David]
  Link: https://lkml.kernel.org/r/527480ef-ed72-e1c1-52a0-1c5b0113df45@suse.cz

Link: https://lkml.kernel.org/r/20201111092812.11329-8-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Suggested-by: David Hildenbrand <david@redhat.com>
Suggested-by: Michal Hocko <mhocko@suse.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: David Hildenbrand <david@redhat.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2020-12-15 12:13:43 -08:00

313 lines
9.2 KiB
C

// SPDX-License-Identifier: GPL-2.0
/*
* linux/mm/page_isolation.c
*/
#include <linux/mm.h>
#include <linux/page-isolation.h>
#include <linux/pageblock-flags.h>
#include <linux/memory.h>
#include <linux/hugetlb.h>
#include <linux/page_owner.h>
#include <linux/migrate.h>
#include "internal.h"
#define CREATE_TRACE_POINTS
#include <trace/events/page_isolation.h>
static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
{
struct zone *zone = page_zone(page);
struct page *unmovable;
unsigned long flags;
spin_lock_irqsave(&zone->lock, flags);
/*
* We assume the caller intended to SET migrate type to isolate.
* If it is already set, then someone else must have raced and
* set it before us.
*/
if (is_migrate_isolate_page(page)) {
spin_unlock_irqrestore(&zone->lock, flags);
return -EBUSY;
}
/*
* FIXME: Now, memory hotplug doesn't call shrink_slab() by itself.
* We just check MOVABLE pages.
*/
unmovable = has_unmovable_pages(zone, page, migratetype, isol_flags);
if (!unmovable) {
unsigned long nr_pages;
int mt = get_pageblock_migratetype(page);
set_pageblock_migratetype(page, MIGRATE_ISOLATE);
zone->nr_isolate_pageblock++;
nr_pages = move_freepages_block(zone, page, MIGRATE_ISOLATE,
NULL);
__mod_zone_freepage_state(zone, -nr_pages, mt);
spin_unlock_irqrestore(&zone->lock, flags);
return 0;
}
spin_unlock_irqrestore(&zone->lock, flags);
if (isol_flags & REPORT_FAILURE) {
/*
* printk() with zone->lock held will likely trigger a
* lockdep splat, so defer it here.
*/
dump_page(unmovable, "unmovable page");
}
return -EBUSY;
}
static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
{
struct zone *zone;
unsigned long flags, nr_pages;
bool isolated_page = false;
unsigned int order;
unsigned long pfn, buddy_pfn;
struct page *buddy;
zone = page_zone(page);
spin_lock_irqsave(&zone->lock, flags);
if (!is_migrate_isolate_page(page))
goto out;
/*
* Because freepage with more than pageblock_order on isolated
* pageblock is restricted to merge due to freepage counting problem,
* it is possible that there is free buddy page.
* move_freepages_block() doesn't care of merge so we need other
* approach in order to merge them. Isolation and free will make
* these pages to be merged.
*/
if (PageBuddy(page)) {
order = buddy_order(page);
if (order >= pageblock_order) {
pfn = page_to_pfn(page);
buddy_pfn = __find_buddy_pfn(pfn, order);
buddy = page + (buddy_pfn - pfn);
if (pfn_valid_within(buddy_pfn) &&
!is_migrate_isolate_page(buddy)) {
__isolate_free_page(page, order);
isolated_page = true;
}
}
}
/*
* If we isolate freepage with more than pageblock_order, there
* should be no freepage in the range, so we could avoid costly
* pageblock scanning for freepage moving.
*
* We didn't actually touch any of the isolated pages, so place them
* to the tail of the freelist. This is an optimization for memory
* onlining - just onlined memory won't immediately be considered for
* allocation.
*/
if (!isolated_page) {
nr_pages = move_freepages_block(zone, page, migratetype, NULL);
__mod_zone_freepage_state(zone, nr_pages, migratetype);
}
set_pageblock_migratetype(page, migratetype);
if (isolated_page)
__putback_isolated_page(page, order, migratetype);
zone->nr_isolate_pageblock--;
out:
spin_unlock_irqrestore(&zone->lock, flags);
}
static inline struct page *
__first_valid_page(unsigned long pfn, unsigned long nr_pages)
{
int i;
for (i = 0; i < nr_pages; i++) {
struct page *page;
page = pfn_to_online_page(pfn + i);
if (!page)
continue;
return page;
}
return NULL;
}
/**
* start_isolate_page_range() - make page-allocation-type of range of pages to
* be MIGRATE_ISOLATE.
* @start_pfn: The lower PFN of the range to be isolated.
* @end_pfn: The upper PFN of the range to be isolated.
* start_pfn/end_pfn must be aligned to pageblock_order.
* @migratetype: Migrate type to set in error recovery.
* @flags: The following flags are allowed (they can be combined in
* a bit mask)
* MEMORY_OFFLINE - isolate to offline (!allocate) memory
* e.g., skip over PageHWPoison() pages
* and PageOffline() pages.
* REPORT_FAILURE - report details about the failure to
* isolate the range
*
* Making page-allocation-type to be MIGRATE_ISOLATE means free pages in
* the range will never be allocated. Any free pages and pages freed in the
* future will not be allocated again. If specified range includes migrate types
* other than MOVABLE or CMA, this will fail with -EBUSY. For isolating all
* pages in the range finally, the caller have to free all pages in the range.
* test_page_isolated() can be used for test it.
*
* There is no high level synchronization mechanism that prevents two threads
* from trying to isolate overlapping ranges. If this happens, one thread
* will notice pageblocks in the overlapping range already set to isolate.
* This happens in set_migratetype_isolate, and set_migratetype_isolate
* returns an error. We then clean up by restoring the migration type on
* pageblocks we may have modified and return -EBUSY to caller. This
* prevents two threads from simultaneously working on overlapping ranges.
*
* Please note that there is no strong synchronization with the page allocator
* either. Pages might be freed while their page blocks are marked ISOLATED.
* A call to drain_all_pages() after isolation can flush most of them. However
* in some cases pages might still end up on pcp lists and that would allow
* for their allocation even when they are in fact isolated already. Depending
* on how strong of a guarantee the caller needs, zone_pcp_disable/enable()
* might be used to flush and disable pcplist before isolation and enable after
* unisolation.
*
* Return: 0 on success and -EBUSY if any part of range cannot be isolated.
*/
int start_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
unsigned migratetype, int flags)
{
unsigned long pfn;
unsigned long undo_pfn;
struct page *page;
BUG_ON(!IS_ALIGNED(start_pfn, pageblock_nr_pages));
BUG_ON(!IS_ALIGNED(end_pfn, pageblock_nr_pages));
for (pfn = start_pfn;
pfn < end_pfn;
pfn += pageblock_nr_pages) {
page = __first_valid_page(pfn, pageblock_nr_pages);
if (page) {
if (set_migratetype_isolate(page, migratetype, flags)) {
undo_pfn = pfn;
goto undo;
}
}
}
return 0;
undo:
for (pfn = start_pfn;
pfn < undo_pfn;
pfn += pageblock_nr_pages) {
struct page *page = pfn_to_online_page(pfn);
if (!page)
continue;
unset_migratetype_isolate(page, migratetype);
}
return -EBUSY;
}
/*
* Make isolated pages available again.
*/
void undo_isolate_page_range(unsigned long start_pfn, unsigned long end_pfn,
unsigned migratetype)
{
unsigned long pfn;
struct page *page;
BUG_ON(!IS_ALIGNED(start_pfn, pageblock_nr_pages));
BUG_ON(!IS_ALIGNED(end_pfn, pageblock_nr_pages));
for (pfn = start_pfn;
pfn < end_pfn;
pfn += pageblock_nr_pages) {
page = __first_valid_page(pfn, pageblock_nr_pages);
if (!page || !is_migrate_isolate_page(page))
continue;
unset_migratetype_isolate(page, migratetype);
}
}
/*
* Test all pages in the range is free(means isolated) or not.
* all pages in [start_pfn...end_pfn) must be in the same zone.
* zone->lock must be held before call this.
*
* Returns the last tested pfn.
*/
static unsigned long
__test_page_isolated_in_pageblock(unsigned long pfn, unsigned long end_pfn,
int flags)
{
struct page *page;
while (pfn < end_pfn) {
if (!pfn_valid_within(pfn)) {
pfn++;
continue;
}
page = pfn_to_page(pfn);
if (PageBuddy(page))
/*
* If the page is on a free list, it has to be on
* the correct MIGRATE_ISOLATE freelist. There is no
* simple way to verify that as VM_BUG_ON(), though.
*/
pfn += 1 << buddy_order(page);
else if ((flags & MEMORY_OFFLINE) && PageHWPoison(page))
/* A HWPoisoned page cannot be also PageBuddy */
pfn++;
else if ((flags & MEMORY_OFFLINE) && PageOffline(page) &&
!page_count(page))
/*
* The responsible driver agreed to skip PageOffline()
* pages when offlining memory by dropping its
* reference in MEM_GOING_OFFLINE.
*/
pfn++;
else
break;
}
return pfn;
}
/* Caller should ensure that requested range is in a single zone */
int test_pages_isolated(unsigned long start_pfn, unsigned long end_pfn,
int isol_flags)
{
unsigned long pfn, flags;
struct page *page;
struct zone *zone;
/*
* Note: pageblock_nr_pages != MAX_ORDER. Then, chunks of free pages
* are not aligned to pageblock_nr_pages.
* Then we just check migratetype first.
*/
for (pfn = start_pfn; pfn < end_pfn; pfn += pageblock_nr_pages) {
page = __first_valid_page(pfn, pageblock_nr_pages);
if (page && !is_migrate_isolate_page(page))
break;
}
page = __first_valid_page(start_pfn, end_pfn - start_pfn);
if ((pfn < end_pfn) || !page)
return -EBUSY;
/* Check all pages are free or marked as ISOLATED */
zone = page_zone(page);
spin_lock_irqsave(&zone->lock, flags);
pfn = __test_page_isolated_in_pageblock(start_pfn, end_pfn, isol_flags);
spin_unlock_irqrestore(&zone->lock, flags);
trace_test_pages_isolated(start_pfn, end_pfn, pfn);
return pfn < end_pfn ? -EBUSY : 0;
}