linux-stable/fs/nilfs2/page.c

562 lines
14 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
/*
* page.c - buffer/page management specific to NILFS
*
* Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation.
*
* Written by Ryusuke Konishi and Seiji Kihara.
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
*/
#include <linux/pagemap.h>
#include <linux/writeback.h>
#include <linux/swap.h>
#include <linux/bitops.h>
#include <linux/page-flags.h>
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/pagevec.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/gfp.h>
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
#include "nilfs.h"
#include "page.h"
#include "mdt.h"
#define NILFS_BUFFER_INHERENT_BITS \
(BIT(BH_Uptodate) | BIT(BH_Mapped) | BIT(BH_NILFS_Node) | \
BIT(BH_NILFS_Volatile) | BIT(BH_NILFS_Checked))
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
static struct buffer_head *
__nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index,
int blkbits, unsigned long b_state)
{
unsigned long first_block;
struct buffer_head *bh;
if (!page_has_buffers(page))
create_empty_buffers(page, 1 << blkbits, b_state);
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
first_block = (unsigned long)index << (PAGE_SHIFT - blkbits);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
bh = nilfs_page_get_nth_block(page, block - first_block);
touch_buffer(bh);
wait_on_buffer(bh);
return bh;
}
struct buffer_head *nilfs_grab_buffer(struct inode *inode,
struct address_space *mapping,
unsigned long blkoff,
unsigned long b_state)
{
int blkbits = inode->i_blkbits;
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
pgoff_t index = blkoff >> (PAGE_SHIFT - blkbits);
struct page *page;
struct buffer_head *bh;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
page = grab_cache_page(mapping, index);
if (unlikely(!page))
return NULL;
bh = __nilfs_get_page_block(page, blkoff, index, blkbits, b_state);
if (unlikely(!bh)) {
unlock_page(page);
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
put_page(page);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
return NULL;
}
return bh;
}
/**
* nilfs_forget_buffer - discard dirty state
* @inode: owner inode of the buffer
* @bh: buffer head of the buffer to be discarded
*/
void nilfs_forget_buffer(struct buffer_head *bh)
{
struct page *page = bh->b_page;
const unsigned long clear_bits =
(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
lock_buffer(bh);
set_mask_bits(&bh->b_state, clear_bits, 0);
if (nilfs_page_buffers_clean(page))
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
__nilfs_clear_page_dirty(page);
bh->b_blocknr = -1;
ClearPageUptodate(page);
ClearPageMappedToDisk(page);
unlock_buffer(bh);
brelse(bh);
}
/**
* nilfs_copy_buffer -- copy buffer data and flags
* @dbh: destination buffer
* @sbh: source buffer
*/
void nilfs_copy_buffer(struct buffer_head *dbh, struct buffer_head *sbh)
{
void *kaddr0, *kaddr1;
unsigned long bits;
struct page *spage = sbh->b_page, *dpage = dbh->b_page;
struct buffer_head *bh;
kaddr0 = kmap_atomic(spage);
kaddr1 = kmap_atomic(dpage);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
memcpy(kaddr1 + bh_offset(dbh), kaddr0 + bh_offset(sbh), sbh->b_size);
kunmap_atomic(kaddr1);
kunmap_atomic(kaddr0);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
dbh->b_state = sbh->b_state & NILFS_BUFFER_INHERENT_BITS;
dbh->b_blocknr = sbh->b_blocknr;
dbh->b_bdev = sbh->b_bdev;
bh = dbh;
bits = sbh->b_state & (BIT(BH_Uptodate) | BIT(BH_Mapped));
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
while ((bh = bh->b_this_page) != dbh) {
lock_buffer(bh);
bits &= bh->b_state;
unlock_buffer(bh);
}
if (bits & BIT(BH_Uptodate))
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
SetPageUptodate(dpage);
else
ClearPageUptodate(dpage);
if (bits & BIT(BH_Mapped))
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
SetPageMappedToDisk(dpage);
else
ClearPageMappedToDisk(dpage);
}
/**
* nilfs_page_buffers_clean - check if a page has dirty buffers or not.
* @page: page to be checked
*
* nilfs_page_buffers_clean() returns zero if the page has dirty buffers.
* Otherwise, it returns non-zero value.
*/
int nilfs_page_buffers_clean(struct page *page)
{
struct buffer_head *bh, *head;
bh = head = page_buffers(page);
do {
if (buffer_dirty(bh))
return 0;
bh = bh->b_this_page;
} while (bh != head);
return 1;
}
void nilfs_page_bug(struct page *page)
{
struct address_space *m;
unsigned long ino;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
if (unlikely(!page)) {
printk(KERN_CRIT "NILFS_PAGE_BUG(NULL)\n");
return;
}
m = page->mapping;
ino = m ? m->host->i_ino : 0;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
printk(KERN_CRIT "NILFS_PAGE_BUG(%p): cnt=%d index#=%llu flags=0x%lx "
"mapping=%p ino=%lu\n",
2016-03-17 21:19:26 +00:00
page, page_ref_count(page),
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
(unsigned long long)page->index, page->flags, m, ino);
if (page_has_buffers(page)) {
struct buffer_head *bh, *head;
int i = 0;
bh = head = page_buffers(page);
do {
printk(KERN_CRIT
" BH[%d] %p: cnt=%d block#=%llu state=0x%lx\n",
i++, bh, atomic_read(&bh->b_count),
(unsigned long long)bh->b_blocknr, bh->b_state);
bh = bh->b_this_page;
} while (bh != head);
}
}
/**
* nilfs_copy_page -- copy the page with buffers
* @dst: destination page
* @src: source page
* @copy_dirty: flag whether to copy dirty states on the page's buffer heads.
*
* This function is for both data pages and btnode pages. The dirty flag
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
* should be treated by caller. The page must not be under i/o.
* Both src and dst page must be locked
*/
static void nilfs_copy_page(struct page *dst, struct page *src, int copy_dirty)
{
struct buffer_head *dbh, *dbufs, *sbh, *sbufs;
unsigned long mask = NILFS_BUFFER_INHERENT_BITS;
BUG_ON(PageWriteback(dst));
sbh = sbufs = page_buffers(src);
if (!page_has_buffers(dst))
create_empty_buffers(dst, sbh->b_size, 0);
if (copy_dirty)
mask |= BIT(BH_Dirty);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
dbh = dbufs = page_buffers(dst);
do {
lock_buffer(sbh);
lock_buffer(dbh);
dbh->b_state = sbh->b_state & mask;
dbh->b_blocknr = sbh->b_blocknr;
dbh->b_bdev = sbh->b_bdev;
sbh = sbh->b_this_page;
dbh = dbh->b_this_page;
} while (dbh != dbufs);
copy_highpage(dst, src);
if (PageUptodate(src) && !PageUptodate(dst))
SetPageUptodate(dst);
else if (!PageUptodate(src) && PageUptodate(dst))
ClearPageUptodate(dst);
if (PageMappedToDisk(src) && !PageMappedToDisk(dst))
SetPageMappedToDisk(dst);
else if (!PageMappedToDisk(src) && PageMappedToDisk(dst))
ClearPageMappedToDisk(dst);
do {
unlock_buffer(sbh);
unlock_buffer(dbh);
sbh = sbh->b_this_page;
dbh = dbh->b_this_page;
} while (dbh != dbufs);
}
int nilfs_copy_dirty_pages(struct address_space *dmap,
struct address_space *smap)
{
struct pagevec pvec;
unsigned int i;
pgoff_t index = 0;
int err = 0;
pagevec_init(&pvec);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
repeat:
if (!pagevec_lookup_tag(&pvec, smap, &index, PAGECACHE_TAG_DIRTY))
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
return 0;
for (i = 0; i < pagevec_count(&pvec); i++) {
struct page *page = pvec.pages[i], *dpage;
lock_page(page);
if (unlikely(!PageDirty(page)))
NILFS_PAGE_BUG(page, "inconsistent dirty state");
dpage = grab_cache_page(dmap, page->index);
if (unlikely(!dpage)) {
/* No empty page is added to the page cache */
err = -ENOMEM;
unlock_page(page);
break;
}
if (unlikely(!page_has_buffers(page)))
NILFS_PAGE_BUG(page,
"found empty page in dat page cache");
nilfs_copy_page(dpage, page, 1);
__set_page_dirty_nobuffers(dpage);
unlock_page(dpage);
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
put_page(dpage);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
unlock_page(page);
}
pagevec_release(&pvec);
cond_resched();
if (likely(!err))
goto repeat;
return err;
}
/**
* nilfs_copy_back_pages -- copy back pages to original cache from shadow cache
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
* @dmap: destination page cache
* @smap: source page cache
*
* No pages must be added to the cache during this process.
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
* This must be ensured by the caller.
*/
void nilfs_copy_back_pages(struct address_space *dmap,
struct address_space *smap)
{
struct pagevec pvec;
unsigned int i, n;
pgoff_t index = 0;
pagevec_init(&pvec);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
repeat:
n = pagevec_lookup(&pvec, smap, &index);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
if (!n)
return;
for (i = 0; i < pagevec_count(&pvec); i++) {
struct page *page = pvec.pages[i], *dpage;
pgoff_t offset = page->index;
lock_page(page);
dpage = find_lock_page(dmap, offset);
if (dpage) {
/* overwrite existing page in the destination cache */
WARN_ON(PageDirty(dpage));
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
nilfs_copy_page(dpage, page, 0);
unlock_page(dpage);
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
put_page(dpage);
/* Do we not need to remove page from smap here? */
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
} else {
struct page *p;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
/* move the page to the destination cache */
xa_lock_irq(&smap->i_pages);
p = __xa_erase(&smap->i_pages, offset);
WARN_ON(page != p);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
smap->nrpages--;
xa_unlock_irq(&smap->i_pages);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
xa_lock_irq(&dmap->i_pages);
p = __xa_store(&dmap->i_pages, offset, page, GFP_NOFS);
if (unlikely(p)) {
/* Probably -ENOMEM */
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
page->mapping = NULL;
put_page(page);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
} else {
page->mapping = dmap;
dmap->nrpages++;
if (PageDirty(page))
__xa_set_mark(&dmap->i_pages, offset,
PAGECACHE_TAG_DIRTY);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
}
xa_unlock_irq(&dmap->i_pages);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
}
unlock_page(page);
}
pagevec_release(&pvec);
cond_resched();
goto repeat;
}
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
/**
* nilfs_clear_dirty_pages - discard dirty pages in address space
* @mapping: address space with dirty pages for discarding
* @silent: suppress [true] or print [false] warning messages
*/
void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
{
struct pagevec pvec;
unsigned int i;
pgoff_t index = 0;
pagevec_init(&pvec);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
while (pagevec_lookup_tag(&pvec, mapping, &index,
PAGECACHE_TAG_DIRTY)) {
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
for (i = 0; i < pagevec_count(&pvec); i++) {
struct page *page = pvec.pages[i];
lock_page(page);
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
nilfs_clear_dirty_page(page, silent);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
unlock_page(page);
}
pagevec_release(&pvec);
cond_resched();
}
}
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
/**
* nilfs_clear_dirty_page - discard dirty page
* @page: dirty page that will be discarded
* @silent: suppress [true] or print [false] warning messages
*/
void nilfs_clear_dirty_page(struct page *page, bool silent)
{
struct inode *inode = page->mapping->host;
struct super_block *sb = inode->i_sb;
BUG_ON(!PageLocked(page));
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
if (!silent)
nilfs_msg(sb, KERN_WARNING,
"discard dirty page: offset=%lld, ino=%lu",
page_offset(page), inode->i_ino);
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
ClearPageUptodate(page);
ClearPageMappedToDisk(page);
if (page_has_buffers(page)) {
struct buffer_head *bh, *head;
const unsigned long clear_bits =
(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
bh = head = page_buffers(page);
do {
lock_buffer(bh);
if (!silent)
nilfs_msg(sb, KERN_WARNING,
"discard dirty block: blocknr=%llu, size=%zu",
(u64)bh->b_blocknr, bh->b_size);
set_mask_bits(&bh->b_state, clear_bits, 0);
nilfs2: fix issue with flush kernel thread after remount in RO mode because of driver's internal error or metadata corruption The NILFS2 driver remounts itself in RO mode in the case of discovering metadata corruption (for example, discovering a broken bmap). But usually, this takes place when there have been file system operations before remounting in RO mode. Thereby, NILFS2 driver can be in RO mode with presence of dirty pages in modified inodes' address spaces. It results in flush kernel thread's infinite trying to flush dirty pages in RO mode. As a result, it is possible to see such side effects as: (1) flush kernel thread occupies 50% - 99% of CPU time; (2) system can't be shutdowned without manual power switch off. SYMPTOMS: (1) System log contains error message: "Remounting filesystem read-only". (2) The flush kernel thread occupies 50% - 99% of CPU time. (3) The system can't be shutdowned without manual power switch off. REPRODUCTION PATH: (1) Create volume group with name "unencrypted" by means of vgcreate utility. (2) Run script (prepared by Anthony Doggett <Anthony2486@interfaces.org.uk>): ----------------[BEGIN SCRIPT]-------------------- #!/bin/bash VG=unencrypted #apt-get install nilfs-tools darcs lvcreate --size 2G --name ntest $VG mkfs.nilfs2 -b 1024 -B 8192 /dev/mapper/$VG-ntest mkdir /var/tmp/n mkdir /var/tmp/n/ntest mount /dev/mapper/$VG-ntest /var/tmp/n/ntest mkdir /var/tmp/n/ntest/thedir cd /var/tmp/n/ntest/thedir sleep 2 date darcs init sleep 2 dmesg|tail -n 5 date darcs whatsnew || true date sleep 2 dmesg|tail -n 5 ----------------[END SCRIPT]-------------------- (3) Try to shutdown the system. REPRODUCIBILITY: 100% FIX: This patch implements checking mount state of NILFS2 driver in nilfs_writepage(), nilfs_writepages() and nilfs_mdt_write_page() methods. If it is detected the RO mount state then all dirty pages are simply discarded with warning messages is written in system log. [akpm@linux-foundation.org: fix printk warning] Signed-off-by: Vyacheslav Dubeyko <slava@dubeyko.com> Acked-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Cc: Anthony Doggett <Anthony2486@interfaces.org.uk> Cc: ARAI Shun-ichi <hermes@ceres.dti.ne.jp> Cc: Piotr Szymaniak <szarpaj@grubelek.pl> Cc: Zahid Chowdhury <zahid.chowdhury@starsolutions.com> Cc: Elmer Zhang <freeboy6716@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-04-30 22:27:48 +00:00
unlock_buffer(bh);
} while (bh = bh->b_this_page, bh != head);
}
__nilfs_clear_page_dirty(page);
}
unsigned int nilfs_page_count_clean_buffers(struct page *page,
unsigned int from, unsigned int to)
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
{
unsigned int block_start, block_end;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
struct buffer_head *bh, *head;
unsigned int nc = 0;
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
for (bh = head = page_buffers(page), block_start = 0;
bh != head || !block_start;
block_start = block_end, bh = bh->b_this_page) {
block_end = block_start + bh->b_size;
if (block_end > from && block_start < to && !buffer_dirty(bh))
nc++;
}
return nc;
}
void nilfs_mapping_init(struct address_space *mapping, struct inode *inode)
{
mapping->host = inode;
mapping->flags = 0;
mapping_set_gfp_mask(mapping, GFP_NOFS);
mapping->private_data = NULL;
nilfs2: fix oops due to a bad aops initialization Nilfs in 2.6.39-rc1 hit the following oops: BUG: unable to handle kernel NULL pointer dereference at 0000000000000048 IP: [<ffffffff810ac235>] try_to_release_page+0x2a/0x3d PGD 234cb6067 PUD 234c72067 PMD 0 Oops: 0000 [#1] SMP <snip> Process truncate (pid: 10995, threadinfo ffff8802353c2000, task ffff880234cfa000) Stack: ffff8802333c77b8 ffffffff810b64b0 0000000000003802 ffffffffa0052cca 0000000000000000 ffff8802353c3b58 0000000000000000 ffff8802353c3b58 0000000000000001 0000000000000000 ffffea0007b92308 ffffea0007b92308 Call Trace: [<ffffffff810b64b0>] ? invalidate_inode_pages2_range+0x15f/0x273 [<ffffffffa0052cca>] ? nilfs_palloc_get_block+0x2d/0xaf [nilfs2] [<ffffffff810589e7>] ? bit_waitqueue+0x14/0xa1 [<ffffffff81058ab1>] ? wake_up_bit+0x10/0x20 [<ffffffffa00433fd>] ? nilfs_forget_buffer+0x66/0x7a [nilfs2] [<ffffffffa00467b8>] ? nilfs_btree_concat_left+0x5c/0x77 [nilfs2] [<ffffffffa00471fc>] ? nilfs_btree_delete+0x395/0x3cf [nilfs2] [<ffffffffa00449a3>] ? nilfs_bmap_do_delete+0x6e/0x79 [nilfs2] [<ffffffffa0045845>] ? nilfs_btree_last_key+0x14b/0x15e [nilfs2] [<ffffffffa00449dd>] ? nilfs_bmap_truncate+0x2f/0x83 [nilfs2] [<ffffffffa0044ab2>] ? nilfs_bmap_last_key+0x35/0x62 [nilfs2] [<ffffffffa003e99b>] ? nilfs_truncate_bmap+0x6b/0xc7 [nilfs2] [<ffffffffa003ee4a>] ? nilfs_truncate+0x79/0xe4 [nilfs2] [<ffffffff810b6c00>] ? vmtruncate+0x33/0x3b [<ffffffffa003e8f1>] ? nilfs_setattr+0x4d/0x8c [nilfs2] [<ffffffff81026106>] ? do_page_fault+0x31b/0x356 [<ffffffff810f9d61>] ? notify_change+0x17d/0x262 [<ffffffff810e5046>] ? do_truncate+0x65/0x80 [<ffffffff810e52af>] ? sys_ftruncate+0xf1/0xf6 [<ffffffff8132c012>] ? system_call_fastpath+0x16/0x1b Code: c3 48 83 ec 08 48 8b 17 48 8b 47 18 80 e2 01 75 04 0f 0b eb fe 48 8b 17 80 e6 20 74 05 31 c0 41 59 c3 48 85 c0 74 11 48 8b 40 58 8b 40 48 48 85 c0 74 04 41 58 ff e0 59 e9 b1 b5 05 00 41 54 RIP [<ffffffff810ac235>] try_to_release_page+0x2a/0x3d RSP <ffff8802353c3b08> CR2: 0000000000000048 This oops was brought in by the change "block: remove per-queue plugging" (commit: 7eaceaccab5f40bb). It initializes mapping->a_ops with a NULL pointer for some pages in nilfs (e.g. btree node pages), but mm code doesn't NULL pointer checks against mapping->a_ops. (the check is done for each callback function) This corrects the aops initialization and fixes the oops. Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Acked-by: Jens Axboe <jaxboe@fusionio.com>
2011-03-30 02:49:20 +00:00
mapping->a_ops = &empty_aops;
}
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
/*
* NILFS2 needs clear_page_dirty() in the following two cases:
*
* 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears
* page dirty flags when it copies back pages from the shadow cache
* (gcdat->{i_mapping,i_btnode_cache}) to its original cache
* (dat->{i_mapping,i_btnode_cache}).
*
* 2) Some B-tree operations like insertion or deletion may dispose buffers
* in dirty state, and this needs to cancel the dirty state of their pages.
*/
int __nilfs_clear_page_dirty(struct page *page)
{
struct address_space *mapping = page->mapping;
if (mapping) {
xa_lock_irq(&mapping->i_pages);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
if (test_bit(PG_dirty, &page->flags)) {
__xa_clear_mark(&mapping->i_pages, page_index(page),
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
PAGECACHE_TAG_DIRTY);
xa_unlock_irq(&mapping->i_pages);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
return clear_page_dirty_for_io(page);
}
xa_unlock_irq(&mapping->i_pages);
nilfs2: buffer and page operations This adds common routines for buffer/page operations used in B-tree node caches, meta data files, or segment constructor (log writer). NILFS uses copy functions for buffers and pages due to the following reasons: 1) Relocation required for COW Since NILFS changes address of on-disk blocks, moving buffers in page cache is needed for the buffers which are not addressed by a file offset. If buffer size is smaller than page size, this involves partial copy of pages. 2) Freezing mmapped pages NILFS calculates checksums for each log to ensure its validity. If page data changes after the checksum calculation, this validity check will not work correctly. To avoid this failure for mmaped pages, NILFS freezes their data by copying. 3) Copy-on-write for DAT pages NILFS makes clones of DAT page caches in a copy-on-write manner during GC processes, and this ensures atomicity and consistency of the DAT in the transient state. In addition, NILFS uses two obsolete functions, nilfs_mark_buffer_dirty() and nilfs_clear_page_dirty() respectively. * nilfs_mark_buffer_dirty() was required to avoid NULL pointer dereference faults: Since the page cache of B-tree node pages or data page cache of pseudo inodes does not have a valid mapping->host, calling mark_buffer_dirty() for their buffers causes the fault; it calls __mark_inode_dirty(NULL) through __set_page_dirty(). * nilfs_clear_page_dirty() was needed in the two cases: 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears page dirty flags when it copies back pages from the cloned cache (gcdat->{i_mapping,i_btnode_cache}) to its original cache (dat->{i_mapping,i_btnode_cache}). 2) Some B-tree operations like insertion or deletion may dispose buffers in dirty state, and this needs to cancel the dirty state of their pages. clear_page_dirty_for_io() caused faults because it does not clear the dirty tag on the page cache. Signed-off-by: Seiji Kihara <kihara.seiji@lab.ntt.co.jp> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-07 02:01:27 +00:00
return 0;
}
return TestClearPageDirty(page);
}
/**
* nilfs_find_uncommitted_extent - find extent of uncommitted data
* @inode: inode
* @start_blk: start block offset (in)
* @blkoff: start offset of the found extent (out)
*
* This function searches an extent of buffers marked "delayed" which
* starts from a block offset equal to or larger than @start_blk. If
* such an extent was found, this will store the start offset in
* @blkoff and return its length in blocks. Otherwise, zero is
* returned.
*/
unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
sector_t start_blk,
sector_t *blkoff)
{
unsigned int i;
pgoff_t index;
unsigned int nblocks_in_page;
unsigned long length = 0;
sector_t b;
struct pagevec pvec;
struct page *page;
if (inode->i_mapping->nrpages == 0)
return 0;
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
index = start_blk >> (PAGE_SHIFT - inode->i_blkbits);
nblocks_in_page = 1U << (PAGE_SHIFT - inode->i_blkbits);
pagevec_init(&pvec);
repeat:
pvec.nr = find_get_pages_contig(inode->i_mapping, index, PAGEVEC_SIZE,
pvec.pages);
if (pvec.nr == 0)
return length;
if (length > 0 && pvec.pages[0]->index > index)
goto out;
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 12:29:47 +00:00
b = pvec.pages[0]->index << (PAGE_SHIFT - inode->i_blkbits);
i = 0;
do {
page = pvec.pages[i];
lock_page(page);
if (page_has_buffers(page)) {
struct buffer_head *bh, *head;
bh = head = page_buffers(page);
do {
if (b < start_blk)
continue;
if (buffer_delay(bh)) {
if (length == 0)
*blkoff = b;
length++;
} else if (length > 0) {
goto out_locked;
}
} while (++b, bh = bh->b_this_page, bh != head);
} else {
if (length > 0)
goto out_locked;
b += nblocks_in_page;
}
unlock_page(page);
} while (++i < pagevec_count(&pvec));
index = page->index + 1;
pagevec_release(&pvec);
cond_resched();
goto repeat;
out_locked:
unlock_page(page);
out:
pagevec_release(&pvec);
return length;
}