bdev: Avoid discarding buffers under a filesystem
Subsystems (3)
Jan Kara
reviewerv2 cover bdev: Avoid discarding buffers under a filesystem
Hello,
this patch set fixes problems when buffer heads are discarded under a live filesystem (which can lead to all sorts of issues like crashes in case of ext4). Patch 1 drops some stale buffer invalidation code, patch 2 temporarily gets exclusive access to the block device for the duration of buffer cache handling to avoid interfering with other exclusive bdev user. The patch fixes the problems for me and pass xfstests for ext4.
Changes since v1:
* Check for exclusive access to the bdev instead of for the presence of superblock
Honza
Jan Kara
reviewerPATCH v2 1/2 fs: Don't invalidate page buffers in block_write_full_page()
If block_write_full_page() is called for a page that is beyond current inode size, it will truncate page buffers for the page and return 0. This logic has been added in 2.5.62 in commit 81eb69062588 ("fix ext3 BUG due to race with truncate") in history.git tree to fix a problem with ext3 in data=ordered mode. This particular problem doesn't exist anymore because ext3 is long gone and ext4 handles ordered data differently. Also normally buffers are invalidated by truncate code and there's no need to specially handle this in ->writepage() code.
This invalidation of page buffers in block_write_full_page() is causing issues to filesystems (e.g. ext4 or ocfs2) when block device is shrunk under filesystem's hands and metadata buffers get discarded while being tracked by the journalling layer. Although it is obviously "not supported" it can cause kernel crashes like:
[ 7986.689400] BUG: unable to handle kernel NULL pointer dereference at +0000000000000008 [ 7986.697197] PGD 0 P4D 0 [ 7986.699724] Oops: 0002 [#1] SMP PTI [ 7986.703200] CPU: 4 PID: 203778 Comm: jbd2/dm-3-8 Kdump: loaded Tainted: G +O --------- - - 4.18.0-147.5.0.5.h126.eulerosv2r9.x86_64 #1 [ 7986.716438] Hardware name: Huawei RH2288H V3/BC11HGSA0, BIOS 1.57 08/11/2015 [ 7986.723462] RIP: 0010:jbd2_journal_grab_journal_head+0x1b/0x40 [jbd2] ... [ 7986.810150] Call Trace: [ 7986.812595] __jbd2_journal_insert_checkpoint+0x23/0x70 [jbd2] [ 7986.818408] jbd2_journal_commit_transaction+0x155f/0x1b60 [jbd2] [ 7986.836467] kjournald2+0xbd/0x270 [jbd2]
which is not great. The crash happens because bh->b_private is suddently NULL although BH_JBD flag is still set (this is because block_invalidatepage() cleared BH_Mapped flag and subsequent bh lookup found buffer without BH_Mapped set, called init_page_buffers() which has rewritten bh->b_private). So just remove the invalidation in block_write_full_page().
Note that the buffer cache invalidation when block device changes size is already careful to avoid similar problems by using invalidate_mapping_pages() which skips busy buffers so it was only this odd block_write_full_page() behavior that could tear down bdev buffers under filesystem's hands.
--- fs/buffer.c | 16 ---------------- 1 file changed, 16 deletions(-)
fs/buffer.c +0 -16
@@ -2771,16 +2771,6 @@ int nobh_writepage(struct page *page, get_block_t *get_block,/* Is the page fully outside i_size? (truncate in progress) */ offset = i_size & (PAGE_SIZE -1 );if (page->index >= end_index+1 || !offset) { -/* -* The page may have dirty, unmapped buffers. For example, -* they may have been added in ext3_writepage(). Make them -* freeable here, so the page does not leak. -*/ -#if 0 -/* Not really sure about this - do we need this ? */ -if (page->mapping->a_ops->invalidatepage) - page->mapping->a_ops->invalidatepage (page, offset); -#endif unlock_page (page);return 0 ;/* don't care */ }
@@ -2975,12 +2965,6 @@ int block_write_full_page(struct page *page, get_block_t *get_block,/* Is the page fully outside i_size? (truncate in progress) */ offset = i_size & (PAGE_SIZE -1 );if (page->index >= end_index+1 || !offset) { -/* -* The page may have dirty, unmapped buffers. For example, -* they may have been added in ext3_writepage(). Make them -* freeable here, so the page does not leak. -*/ -do_invalidatepage (page,0 ,PAGE_SIZE );unlock_page (page);return 0 ;/* don't care */ }
Signature · 1 line
2.16.4
Re: [PATCH 1/2] fs: Don't invalidate page buffers in block_write_full_page()
Quoted 9 lines · Jan Kara
If
block_write_full_page()is called for a page that is beyond current inode size, it will truncate page buffers for the page and return 0. This logic has been added in 2.5.62 in commit81eb69062588("fix ext3 BUG due to race with truncate") in history.git tree to fix a problem with ext3 in data=ordered mode. This particular problem doesn't exist anymore because ext3 is long gone and ext4 handles ordered data differently. Also normally buffers are invalidated by truncate code and there's no need to specially handle this in ->writepage() code.
Quoted 5 lines
This invalidation of page buffers in
block_write_full_page()is causing issues to filesystems (e.g. ext4 or ocfs2) when block device is shrunk under filesystem's hands and metadata buffers get discarded while being tracked by the journalling layer. Although it is obviously "not supported" it can cause kernel crashes like:
Btw, while looking over the block device revalidation code I think all the magic we do on shrinking block devices actually is a bad idea - potentially very harmful, but without much real benefit. And it only is run on file systems directly created on the whole device, meaning it isn't even used at all with the typical setups that use partitions.
Anyway, this patch looks good:
Jan Kara
reviewerPATCH v2 2/2 block: Do not discard buffers under a mounted filesystem
In reply to Jan KaraDiscarding blocks and buffers under a mounted filesystem is hardly anything admin wants to do. Usually it will confuse the filesystem and sometimes the loss of buffer_head state (including b_private field) can even cause crashes like:
BUG: unable to handle kernel NULL pointer dereference at 0000000000000008 PGD 0 P4D 0 Oops: 0002 [#1] SMP PTI CPU: 4 PID: 203778 Comm: jbd2/dm-3-8 Kdump: loaded Tainted: G O --------- - - 4.18.0-147.5.0.5.h126.eulerosv2r9.x86_64 #1 Hardware name: Huawei RH2288H V3/BC11HGSA0, BIOS 1.57 08/11/2015 RIP: 0010:jbd2_journal_grab_journal_head+0x1b/0x40 [jbd2] ... Call Trace: __jbd2_journal_insert_checkpoint+0x23/0x70 [jbd2] jbd2_journal_commit_transaction+0x155f/0x1b60 [jbd2] kjournald2+0xbd/0x270 [jbd2]
So if we don't have block device open with O_EXCL already, claim the block device while we truncate buffer cache. This makes sure any exclusive block device user (such as filesystem) cannot operate on the device while we are discarding buffer cache.
--- block/ioctl.c | 16 ++++++++++------ fs/block_dev.c | 37 +++++++++++++++++++++++++++++++++---- include/linux/blkdev.h | 7 +++++++ 3 files changed, 50 insertions(+), 10 deletions(-)
block/ioctl.c +10 -6
@@ -112,8 +112,7 @@ static int blk_ioctl_discard(struct block_device *bdev, fmode_t mode,uint64_t range[2 ];uint64_t start, len;struct request_queue *q =bdev_get_queue (bdev); -struct address_space *mapping = bdev->bd_inode->i_mapping; - +int err;if (!(mode &FMODE_WRITE ))return -EBADF ;
@@ -134,7 +133,11 @@ static int blk_ioctl_discard(struct block_device *bdev, fmode_t mode,if (start + len >i_size_read (bdev->bd_inode))return -EINVAL ; -truncate_inode_pages_range (mapping, start, start + len -1 ); + + err =truncate_bdev_range (bdev, mode, start, start + len -1 ); +if (err) +return err; +return blkdev_issue_discard (bdev, start >>9 , len >>9 ,GFP_KERNEL , flags); }
@@ -143,8 +146,8 @@ static int blk_ioctl_zeroout(struct block_device *bdev, fmode_t mode, unsigned long arg) {uint64_t range[2 ]; -struct address_space *mapping;uint64_t start, end, len; +int err;if (!(mode &FMODE_WRITE ))return -EBADF ;
@@ -166,8 +169,9 @@ static int blk_ioctl_zeroout(struct block_device *bdev, fmode_t mode,return -EINVAL ;/* Invalidate the page cache, including dirty pages */ - mapping = bdev->bd_inode->i_mapping; -truncate_inode_pages_range (mapping, start, end); + err =truncate_bdev_range (bdev, mode, start, end); +if (err) +return err;return blkdev_issue_zeroout (bdev, start >>9 , len >>9 ,GFP_KERNEL ,BLKDEV_ZERO_NOUNMAP );
fs/block_dev.c +33 -4
@@ -103,6 +103,35 @@ void invalidate_bdev(struct block_device *bdev) }EXPORT_SYMBOL (invalidate_bdev); +/* +* Drop all buffers & page cache for given bdev range. This function bails +* with error if bdev has other exclusive owner (such as filesystem). +*/ +int truncate_bdev_range (struct block_device *bdev,fmode_t mode, +loff_t lstart,loff_t lend) +{ +struct block_device *claimed_bdev =NULL ; +int err; + +/* +* If we don't hold exclusive handle for the device, upgrade to it +* while we discard the buffer cache to avoid discarding buffers +* under live filesystem. +*/ +if (!(mode &FMODE_EXCL )) { + claimed_bdev = bdev->bd_contains; + err =bd_prepare_to_claim (bdev, claimed_bdev, + truncate_bdev_range); +if (err) +return err; + } +truncate_inode_pages_range (bdev->bd_inode->i_mapping, lstart, lend); +if (claimed_bdev) +bd_abort_claiming (bdev, claimed_bdev, truncate_bdev_range); +return 0 ; +} +EXPORT_SYMBOL (truncate_bdev_range); +static void set_init_blocksize (struct block_device *bdev) { bdev->bd_inode->i_blkbits =blksize_bits (bdev_logical_block_size (bdev));
@@ -1969,7 +1998,6 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start, loff_t len) {struct block_device *bdev =I_BDEV (bdev_file_inode (file)); -struct address_space *mapping;loff_t end = start + len -1 ;loff_t isize;int error;
@@ -1997,8 +2025,9 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,return -EINVAL ;/* Invalidate the page cache, including dirty pages. */ - mapping = bdev->bd_inode->i_mapping; -truncate_inode_pages_range (mapping, start, end); + error =truncate_bdev_range (bdev, file->f_mode, start, end); +if (error) +return error;switch (mode) {case FALLOC_FL_ZERO_RANGE :
@@ -2025,7 +2054,7 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,* the caller will be given -EBUSY. The third argument is * inclusive, so the rounding here is safe. */ -return invalidate_inode_pages2_range (mapping, +return invalidate_inode_pages2_range (bdev->bd_inode->i_mapping, start >>PAGE_SHIFT , end >>PAGE_SHIFT ); }
include/linux/blkdev.h +7 -0
@@ -1984,11 +1984,18 @@ void bdput(struct block_device *);#ifdef CONFIG_BLOCK void invalidate_bdev (struct block_device *bdev); +int truncate_bdev_range (struct block_device *bdev,fmode_t mode,loff_t lstart, +loff_t lend);int sync_blockdev (struct block_device *bdev);#else static inline void invalidate_bdev (struct block_device *bdev) { } +int truncate_bdev_range (struct block_device *bdev,fmode_t mode,loff_t lstart, +loff_t lend) +{ +return 0 ; +}static inline int sync_blockdev (struct block_device *bdev) {return 0 ;
Signature · 1 line
2.16.4
Re: [PATCH 2/2] block: Do not discard buffers under a mounted filesystem
Looks good,
Jan Kara
reviewerHello!
Quoted 10 lines · Jan Kara
this patch set fixes problems when buffer heads are discarded under a live filesystem (which can lead to all sorts of issues like crashes in case of ext4). Patch 1 drops some stale buffer invalidation code, patch 2 temporarily gets exclusive access to the block device for the duration of buffer cache handling to avoid interfering with other exclusive bdev user. The patch fixes the problems for me and pass xfstests for ext4.
Changes since v1:
* Check for exclusive access to the bdev instead of for the presence of superblock
Jens, now that Christoph has reviewed the patches (thanks Christoph!), can you pick up the patches to your tree please? Thanks!
Honza
Signature · 2 lines
Jan Kara <[email protected]> SUSE Labs, CR
Jens Axboe
maintainerQuoted 10 lines · Jan Kara
Hello!
Quoted 7 lines · Jan Kara
this patch set fixes problems when buffer heads are discarded under a live filesystem (which can lead to all sorts of issues like crashes in case of ext4). Patch 1 drops some stale buffer invalidation code, patch 2 temporarily gets exclusive access to the block device for the duration of buffer cache handling to avoid interfering with other exclusive bdev user. The patch fixes the problems for me and pass xfstests for ext4.
Quoted 6 lines
Quoted 3 lines
Changes since v1:
* Check for exclusive access to the bdev instead of for the presence of superblockJens, now that Christoph has reviewed the patches (thanks Christoph!), can you pick up the patches to your tree please? Thanks!
Yep, I applied them for 5.10. Thanks!
Signature · 1 line
Jens Axboe