While we are finishing a device replace operation, we can make a discard
operation (fs mounted with -o discard) do an invalid memory access like
the one reported by the following trace:
[ 3206.384654] general protection fault: 0000 [#1] PREEMPT SMP
[ 3206.387520] Modules linked in: dm_mod btrfs crc32c_generic xor raid6_pq acpi_cpufreq tpm_tis psmouse tpm ppdev sg parport_pc evdev i2c_piix4 parport
processor serio_raw i2c_core pcspkr button loop autofs4 ext4 crc16 jbd2 mbcache sr_mod cdrom ata_generic sd_mod virtio_scsi ata_piix libata virtio_pci
virtio_ring scsi_mod e1000 virtio floppy [last unloaded: btrfs]
[ 3206.388595] CPU: 14 PID: 29194 Comm: fsstress Not tainted 4.6.0-rc7-btrfs-next-29+ #1
[ 3206.388595] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS by qemu-project.org 04/01/2014
[ 3206.388595] task: ffff88017ace0100 ti: ffff880171b98000 task.ti: ffff880171b98000
[ 3206.388595] RIP: 0010:[<ffffffff8124d233>] [<ffffffff8124d233>] blkdev_issue_discard+0x5c/0x2a7
[ 3206.388595] RSP: 0018:ffff880171b9bb80 EFLAGS: 00010246
[ 3206.388595] RAX: ffff880171b9bc28 RBX: 000000000090d000 RCX: 0000000000000000
[ 3206.388595] RDX: ffffffff82fa1b48 RSI: ffffffff8179f46c RDI: ffffffff82fa1b48
[ 3206.388595] RBP: ffff880171b9bcc0 R08: 0000000000000000 R09: 0000000000000001
[ 3206.388595] R10: ffff880171b9bce0 R11: 000000000090f000 R12: ffff880171b9bbe8
[ 3206.388595] R13: 0000000000000010 R14: 0000000000004868 R15: 6b6b6b6b6b6b6b6b
[ 3206.388595] FS: 00007f6182e4e700(0000) GS:ffff88023fdc0000(0000) knlGS:0000000000000000
[ 3206.388595] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 3206.388595] CR2: 00007f617c2bbb18 CR3: 000000017ad9c000 CR4: 00000000000006e0
[ 3206.388595] Stack:
[ 3206.388595] 0000000000004878 0000000000000000 0000000002400040 0000000000000000
[ 3206.388595] 0000000000000000 ffff880171b9bbe8 ffff880171b9bbb0 ffff880171b9bbb0
[ 3206.388595] ffff880171b9bbc0 ffff880171b9bbc0 ffff880171b9bbd0 ffff880171b9bbd0
[ 3206.388595] Call Trace:
[ 3206.388595] [<ffffffffa042899e>] btrfs_issue_discard+0x12f/0x143 [btrfs]
[ 3206.388595] [<ffffffffa042899e>] ? btrfs_issue_discard+0x12f/0x143 [btrfs]
[ 3206.388595] [<ffffffffa042e862>] btrfs_discard_extent+0x87/0xde [btrfs]
[ 3206.388595] [<ffffffffa04303b5>] btrfs_finish_extent_commit+0xb2/0x1df [btrfs]
[ 3206.388595] [<ffffffff8149c246>] ? __mutex_unlock_slowpath+0x150/0x15b
[ 3206.388595] [<ffffffffa04464c4>] btrfs_commit_transaction+0x7fc/0x980 [btrfs]
[ 3206.388595] [<ffffffff8149c246>] ? __mutex_unlock_slowpath+0x150/0x15b
[ 3206.388595] [<ffffffffa0459af6>] btrfs_sync_file+0x38f/0x428 [btrfs]
[ 3206.388595] [<ffffffff811a8292>] vfs_fsync_range+0x8c/0x9e
[ 3206.388595] [<ffffffff811a82c0>] vfs_fsync+0x1c/0x1e
[ 3206.388595] [<ffffffff811a8417>] do_fsync+0x31/0x4a
[ 3206.388595] [<ffffffff811a8637>] SyS_fsync+0x10/0x14
[ 3206.388595] [<ffffffff8149e025>] entry_SYSCALL_64_fastpath+0x18/0xa8
[ 3206.388595] [<ffffffff81100c6b>] ? time_hardirqs_off+0x9/0x14
[ 3206.388595] [<ffffffff8108e87d>] ? trace_hardirqs_off_caller+0x1f/0xaa
This happens because when we call btrfs_map_block() from
btrfs_discard_extent() to get a btrfs_bio structure, the device replace
operation has not finished yet, but before we use the device of one of the
stripes from the returned btrfs_bio structure, the device object is freed.
This is illustrated by the following diagram.
CPU 1 CPU 2
btrfs_dev_replace_start()
(...)
btrfs_dev_replace_finishing()
btrfs_start_transaction()
btrfs_commit_transaction()
(...)
btrfs_sync_file()
btrfs_start_transaction()
(...)
btrfs_commit_transaction()
btrfs_finish_extent_commit()
btrfs_discard_extent()
btrfs_map_block()
--> returns a struct btrfs_bio
with a stripe that has a
device field pointing to
source device of the replace
operation (the device that
is being replaced)
mutex_lock(&uuid_mutex)
mutex_lock(&fs_info->fs_devices->device_list_mutex)
mutex_lock(&fs_info->chunk_mutex)
btrfs_dev_replace_update_device_in_mapping_tree()
--> iterates the mapping tree and for each
extent map that has a stripe pointing to
the source device, it updates the stripe
to point to the target device instead
btrfs_rm_dev_replace_blocked()
--> waits for fs_info->bio_counter to go down to 0
btrfs_rm_dev_replace_remove_srcdev()
--> removes source device from the list of devices
mutex_unlock(&fs_info->chunk_mutex)
mutex_unlock(&fs_info->fs_devices->device_list_mutex)
mutex_unlock(&uuid_mutex)
btrfs_rm_dev_replace_free_srcdev()
--> frees the source device
--> iterates over all stripes
of the returned struct
btrfs_bio
--> for each stripe it
dereferences its device
pointer
--> it ends up finding a
pointer to the device
used as the source
device for the replace
operation and that was
already freed
So fix this by surrounding the call to btrfs_map_block(), and the code
that uses the returned struct btrfs_bio, with calls to
btrfs_bio_counter_inc_blocked() and btrfs_bio_counter_dec(), so that
the finishing phase of the device replace operation blocks until the
the bio counter decreases to zero before it frees the source device.
This is the same approach we do at btrfs_map_bio() for example.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Relocation of a block group waits for all existing tasks flushing
dellaloc, starting direct IO writes and any ordered extents before
starting the relocation process. However for direct IO writes that end
up doing nocow (inode either has the flag nodatacow set or the write is
against a prealloc extent) we have a short time window that allows for a
race that makes relocation proceed without waiting for the direct IO
write to complete first, resulting in data loss after the relocation
finishes. This is illustrated by the following diagram:
CPU 1 CPU 2
btrfs_relocate_block_group(bg X)
direct IO write starts against
an extent in block group X
using nocow mode (inode has the
nodatacow flag or the write is
for a prealloc extent)
btrfs_direct_IO()
btrfs_get_blocks_direct()
--> can_nocow_extent() returns 1
btrfs_inc_block_group_ro(bg X)
--> turns block group into RO mode
btrfs_wait_ordered_roots()
--> returns and does not know about
the DIO write happening at CPU 2
(the task there has not created
yet an ordered extent)
relocate_block_group(bg X)
--> rc->stage == MOVE_DATA_EXTENTS
find_next_extent()
--> returns extent that the DIO
write is going to write to
relocate_data_extent()
relocate_file_extent_cluster()
--> reads the extent from disk into
pages belonging to the relocation
inode and dirties them
--> creates DIO ordered extent
btrfs_submit_direct()
--> submits bio against a location
on disk obtained from an extent
map before the relocation started
btrfs_wait_ordered_range()
--> writes all the pages read before
to disk (belonging to the
relocation inode)
relocation finishes
bio completes and wrote new data
to the old location of the block
group
So fix this by tracking the number of nocow writers for a block group and
make sure relocation waits for that number to go down to 0 before starting
to move the extents.
The same race can also happen with buffered writes in nocow mode since the
patch I recently made titled "Btrfs: don't do unnecessary delalloc flushes
when relocating", because we are no longer flushing all delalloc which
served as a synchonization mechanism (due to page locking) and ensured
the ordered extents for nocow buffered writes were created before we
called btrfs_wait_ordered_roots(). The race with direct IO writes in nocow
mode existed before that patch (no pages are locked or used during direct
IO) and that fixed only races with direct IO writes that do cow.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Before we start the actual relocation process of a block group, we do
calls to flush delalloc of all inodes and then wait for ordered extents
to complete. However we do these flush calls just to make sure we don't
race with concurrent tasks that have actually already started to run
delalloc and have allocated an extent from the block group we want to
relocate, right before we set it to readonly mode, but have not yet
created the respective ordered extents. The flush calls make us wait
for such concurrent tasks because they end up calling
filemap_fdatawrite_range() (through btrfs_start_delalloc_roots() ->
__start_delalloc_inodes() -> btrfs_alloc_delalloc_work() ->
btrfs_run_delalloc_work()) which ends up serializing us with those tasks
due to attempts to lock the same pages (and the delalloc flush procedure
calls the allocator and creates the ordered extents before unlocking the
pages).
These flushing calls not only make us waste time (cpu, IO) but also reduce
the chances of writing larger extents (applications might be writing to
contiguous ranges and we flush before they finish dirtying the whole
ranges).
So make sure we don't flush delalloc and just wait for concurrent tasks
that have already started flushing delalloc and have allocated an extent
from the block group we are about to relocate.
This change also ends up fixing a race with direct IO writes that makes
relocation not wait for direct IO ordered extents. This race is
illustrated by the following diagram:
CPU 1 CPU 2
btrfs_relocate_block_group(bg X)
starts direct IO write,
target inode currently has no
ordered extents ongoing nor
dirty pages (delalloc regions),
therefore the root for our inode
is not in the list
fs_info->ordered_roots
btrfs_direct_IO()
__blockdev_direct_IO()
btrfs_get_blocks_direct()
btrfs_lock_extent_direct()
locks range in the io tree
btrfs_new_extent_direct()
btrfs_reserve_extent()
--> extent allocated
from bg X
btrfs_inc_block_group_ro(bg X)
btrfs_start_delalloc_roots()
__start_delalloc_inodes()
--> does nothing, no dealloc ranges
in the inode's io tree so the
inode's root is not in the list
fs_info->delalloc_roots
btrfs_wait_ordered_roots()
--> does not find the inode's root in the
list fs_info->ordered_roots
--> ends up not waiting for the direct IO
write started by the task at CPU 2
relocate_block_group(rc->stage ==
MOVE_DATA_EXTENTS)
prepare_to_relocate()
btrfs_commit_transaction()
iterates the extent tree, using its
commit root and moves extents into new
locations
btrfs_add_ordered_extent_dio()
--> now a ordered extent is
created and added to the
list root->ordered_extents
and the root added to the
list fs_info->ordered_roots
--> this is too late and the
task at CPU 1 already
started the relocation
btrfs_commit_transaction()
btrfs_finish_ordered_io()
btrfs_alloc_reserved_file_extent()
--> adds delayed data reference
for the extent allocated
from bg X
relocate_block_group(rc->stage ==
UPDATE_DATA_PTRS)
prepare_to_relocate()
btrfs_commit_transaction()
--> delayed refs are run, so an extent
item for the allocated extent from
bg X is added to extent tree
--> commit roots are switched, so the
next scan in the extent tree will
see the extent item
sees the extent in the extent tree
When this happens the relocation produces the following warning when it
finishes:
[ 7260.832836] ------------[ cut here ]------------
[ 7260.834653] WARNING: CPU: 5 PID: 6765 at fs/btrfs/relocation.c:4318 btrfs_relocate_block_group+0x245/0x2a1 [btrfs]()
[ 7260.838268] Modules linked in: btrfs crc32c_generic xor ppdev raid6_pq psmouse sg acpi_cpufreq evdev i2c_piix4 tpm_tis serio_raw tpm i2c_core pcspkr parport_pc
[ 7260.850935] CPU: 5 PID: 6765 Comm: btrfs Not tainted 4.5.0-rc6-btrfs-next-28+ #1
[ 7260.852998] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS by qemu-project.org 04/01/2014
[ 7260.852998] 0000000000000000 ffff88020bf57bc0 ffffffff812648b3 0000000000000000
[ 7260.852998] 0000000000000009 ffff88020bf57bf8 ffffffff81051608 ffffffffa03c1b2d
[ 7260.852998] ffff8800b2bbb800 0000000000000000 ffff8800b17bcc58 ffff8800399dd000
[ 7260.852998] Call Trace:
[ 7260.852998] [<ffffffff812648b3>] dump_stack+0x67/0x90
[ 7260.852998] [<ffffffff81051608>] warn_slowpath_common+0x99/0xb2
[ 7260.852998] [<ffffffffa03c1b2d>] ? btrfs_relocate_block_group+0x245/0x2a1 [btrfs]
[ 7260.852998] [<ffffffff810516d4>] warn_slowpath_null+0x1a/0x1c
[ 7260.852998] [<ffffffffa03c1b2d>] btrfs_relocate_block_group+0x245/0x2a1 [btrfs]
[ 7260.852998] [<ffffffffa039d9de>] btrfs_relocate_chunk.isra.29+0x66/0xdb [btrfs]
[ 7260.852998] [<ffffffffa039f314>] btrfs_balance+0xde1/0xe4e [btrfs]
[ 7260.852998] [<ffffffff8127d671>] ? debug_smp_processor_id+0x17/0x19
[ 7260.852998] [<ffffffffa03a9583>] btrfs_ioctl_balance+0x255/0x2d3 [btrfs]
[ 7260.852998] [<ffffffffa03ac96a>] btrfs_ioctl+0x11e0/0x1dff [btrfs]
[ 7260.852998] [<ffffffff811451df>] ? handle_mm_fault+0x443/0xd63
[ 7260.852998] [<ffffffff81491817>] ? _raw_spin_unlock+0x31/0x44
[ 7260.852998] [<ffffffff8108b36a>] ? arch_local_irq_save+0x9/0xc
[ 7260.852998] [<ffffffff811876ab>] vfs_ioctl+0x18/0x34
[ 7260.852998] [<ffffffff81187cb2>] do_vfs_ioctl+0x550/0x5be
[ 7260.852998] [<ffffffff81190c30>] ? __fget_light+0x4d/0x71
[ 7260.852998] [<ffffffff81187d77>] SyS_ioctl+0x57/0x79
[ 7260.852998] [<ffffffff81492017>] entry_SYSCALL_64_fastpath+0x12/0x6b
[ 7260.893268] ---[ end trace eb7803b24ebab8ad ]---
This is because at the end of the first stage, in relocate_block_group(),
we commit the current transaction, which makes delayed refs run, the
commit roots are switched and so the second stage will find the extent
item that the ordered extent added to the delayed refs. But this extent
was not moved (ordered extent completed after first stage finished), so
at the end of the relocation our block group item still has a positive
used bytes counter, triggering a warning at the end of
btrfs_relocate_block_group(). Later on when trying to read the extent
contents from disk we hit a BUG_ON() due to the inability to map a block
with a logical address that belongs to the block group we relocated and
is no longer valid, resulting in the following trace:
[ 7344.885290] BTRFS critical (device sdi): unable to find logical 12845056 len 4096
[ 7344.887518] ------------[ cut here ]------------
[ 7344.888431] kernel BUG at fs/btrfs/inode.c:1833!
[ 7344.888431] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
[ 7344.888431] Modules linked in: btrfs crc32c_generic xor ppdev raid6_pq psmouse sg acpi_cpufreq evdev i2c_piix4 tpm_tis serio_raw tpm i2c_core pcspkr parport_pc
[ 7344.888431] CPU: 0 PID: 6831 Comm: od Tainted: G W 4.5.0-rc6-btrfs-next-28+ #1
[ 7344.888431] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS by qemu-project.org 04/01/2014
[ 7344.888431] task: ffff880215818600 ti: ffff880204684000 task.ti: ffff880204684000
[ 7344.888431] RIP: 0010:[<ffffffffa037c88c>] [<ffffffffa037c88c>] btrfs_merge_bio_hook+0x54/0x6b [btrfs]
[ 7344.888431] RSP: 0018:ffff8802046878f0 EFLAGS: 00010282
[ 7344.888431] RAX: 00000000ffffffea RBX: 0000000000001000 RCX: 0000000000000001
[ 7344.888431] RDX: ffff88023ec0f950 RSI: ffffffff8183b638 RDI: 00000000ffffffff
[ 7344.888431] RBP: ffff880204687908 R08: 0000000000000001 R09: 0000000000000000
[ 7344.888431] R10: ffff880204687770 R11: ffffffff82f2d52d R12: 0000000000001000
[ 7344.888431] R13: ffff88021afbfee8 R14: 0000000000006208 R15: ffff88006cd199b0
[ 7344.888431] FS: 00007f1f9e1d6700(0000) GS:ffff88023ec00000(0000) knlGS:0000000000000000
[ 7344.888431] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 7344.888431] CR2: 00007f1f9dc8cb60 CR3: 000000023e3b6000 CR4: 00000000000006f0
[ 7344.888431] Stack:
[ 7344.888431] 0000000000001000 0000000000001000 ffff880204687b98 ffff880204687950
[ 7344.888431] ffffffffa0395c8f ffffea0004d64d48 0000000000000000 0000000000001000
[ 7344.888431] ffffea0004d64d48 0000000000001000 0000000000000000 0000000000000000
[ 7344.888431] Call Trace:
[ 7344.888431] [<ffffffffa0395c8f>] submit_extent_page+0xf5/0x16f [btrfs]
[ 7344.888431] [<ffffffffa03970ac>] __do_readpage+0x4a0/0x4f1 [btrfs]
[ 7344.888431] [<ffffffffa039680d>] ? btrfs_create_repair_bio+0xcb/0xcb [btrfs]
[ 7344.888431] [<ffffffffa037eeb4>] ? btrfs_writepage_start_hook+0xbc/0xbc [btrfs]
[ 7344.888431] [<ffffffff8108df55>] ? trace_hardirqs_on+0xd/0xf
[ 7344.888431] [<ffffffffa039728c>] __do_contiguous_readpages.constprop.26+0xc2/0xe4 [btrfs]
[ 7344.888431] [<ffffffffa037eeb4>] ? btrfs_writepage_start_hook+0xbc/0xbc [btrfs]
[ 7344.888431] [<ffffffffa039739b>] __extent_readpages.constprop.25+0xed/0x100 [btrfs]
[ 7344.888431] [<ffffffff81129d24>] ? lru_cache_add+0xe/0x10
[ 7344.888431] [<ffffffffa0397ea8>] extent_readpages+0x160/0x1aa [btrfs]
[ 7344.888431] [<ffffffffa037eeb4>] ? btrfs_writepage_start_hook+0xbc/0xbc [btrfs]
[ 7344.888431] [<ffffffff8115daad>] ? alloc_pages_current+0xa9/0xcd
[ 7344.888431] [<ffffffffa037cdc9>] btrfs_readpages+0x1f/0x21 [btrfs]
[ 7344.888431] [<ffffffff81128316>] __do_page_cache_readahead+0x168/0x1fc
[ 7344.888431] [<ffffffff811285a0>] ondemand_readahead+0x1f6/0x207
[ 7344.888431] [<ffffffff811285a0>] ? ondemand_readahead+0x1f6/0x207
[ 7344.888431] [<ffffffff8111cf34>] ? pagecache_get_page+0x2b/0x154
[ 7344.888431] [<ffffffff8112870e>] page_cache_sync_readahead+0x3d/0x3f
[ 7344.888431] [<ffffffff8111dbf7>] generic_file_read_iter+0x197/0x4e1
[ 7344.888431] [<ffffffff8117773a>] __vfs_read+0x79/0x9d
[ 7344.888431] [<ffffffff81178050>] vfs_read+0x8f/0xd2
[ 7344.888431] [<ffffffff81178a38>] SyS_read+0x50/0x7e
[ 7344.888431] [<ffffffff81492017>] entry_SYSCALL_64_fastpath+0x12/0x6b
[ 7344.888431] Code: 8d 4d e8 45 31 c9 45 31 c0 48 8b 00 48 c1 e2 09 48 8b 80 80 fc ff ff 4c 89 65 e8 48 8b b8 f0 01 00 00 e8 1d 42 02 00 85 c0 79 02 <0f> 0b 4c 0
[ 7344.888431] RIP [<ffffffffa037c88c>] btrfs_merge_bio_hook+0x54/0x6b [btrfs]
[ 7344.888431] RSP <ffff8802046878f0>
[ 7344.970544] ---[ end trace eb7803b24ebab8ae ]---
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Before the relocation process of a block group starts, it sets the block
group to readonly mode, then flushes all delalloc writes and then finally
it waits for all ordered extents to complete. This last step includes
waiting for ordered extents destinated at extents allocated in other block
groups, making us waste unecessary time.
So improve this by waiting only for ordered extents that fall into the
block group's range.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
UBSAN: Undefined behaviour in fs/btrfs/extent-tree.c:4623:21
signed integer overflow:
10808 * 262144 cannot be represented in type 'int [8]'
If 8192<=items<16384, we request a writeback of an insane number of pages
which is benign (everything will be written). But if items>=16384, the
space reservation won't be enough.
Signed-off-by: Adam Borowski <kilobyte@angband.pl>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/extent-tree.c: In function ‘btrfs_lock_cluster’:
fs/btrfs/extent-tree.c:6399: warning: ‘used_bg’ may be used uninitialized in this function
- Replace "again: ... goto again;" by standard C "while (1) { ... }",
- Move block not processed during the first iteration of the loop to the
end of the loop, which allows to kill the "locked" variable,
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Reviewed-and-Tested-by: Miao Xie <miaox@cn.fujitsu.com>
[ the compilation warning has been fixed by other patch, now we want to
clean up the function ]
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_std_error() handles errors, puts FS into readonly mode
(as of now). So its good idea to rename it to btrfs_handle_fs_error().
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: David Sterba <dsterba@suse.com>
[ edit changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
Pull btrfs fixes from Chris Mason:
"These are bug fixes, including a really old fsync bug, and a few trace
points to help us track down problems in the quota code"
* 'for-linus-4.6' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
Btrfs: fix file/data loss caused by fsync after rename and new inode
btrfs: Reset IO error counters before start of device replacing
btrfs: Add qgroup tracing
Btrfs: don't use src fd for printk
btrfs: fallback to vmalloc in btrfs_compare_tree
btrfs: handle non-fatal errors in btrfs_qgroup_inherit()
btrfs: Output more info for enospc_debug mount option
Btrfs: fix invalid reference in replace_path
Btrfs: Improve FL_KEEP_SIZE handling in fallocate
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>
As one user in mail list report reproducible balance ENOSPC error, it's
better to add more debug info for enospc_debug mount option.
Reported-by: Marc Haber <mh+linux-btrfs@zugschlus.de>
Signed-off-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: David Sterba <dsterba@suse.com>
We were getting build warning about:
fs/btrfs/extent-tree.c:7021:34: warning: ‘used_bg’ may be used
uninitialized in this function
It is not a valid warning as used_bg is never used uninitilized since
locked is initially false so we can never be in the section where
'used_bg' is used. But gcc is not able to understand that and we can
initialize it while declaring to silence the warning.
Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org>
Signed-off-by: David Sterba <dsterba@suse.com>
We will sometimes start background flushing the various enospc related things
(delayed nodes, delalloc, etc) if we are getting close to reserving all of our
available space. We don't want to do this however when we are actually using
this space as it causes unneeded thrashing. We currently try to do this by
checking bytes_used >= thresh, but bytes_used is only part of the equation, we
need to use bytes_reserved as well as this represents space that is very likely
to become bytes_used in the future.
My tracing tool will keep count of the number of times we kick off the async
flusher, the following are counts for the entire run of generic/027
No Patch Patch
avg: 5385 5009
median: 5500 4916
We skewed lower than the average with my patch and higher than the average with
the patch, overall it cuts the flushing from anywhere from 5-10%, which in the
case of actual ENOSPC is quite helpful. Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Reviewed-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
I'm writing a tool to visualize the enospc system in order to help debug enospc
bugs and I found weird data and ran it down to when we update the global block
rsv. We add all of the remaining free space to the block rsv, do a trace event,
then remove the extra and do another trace event. This makes my visualization
look silly and is unintuitive code as well. Fix this stuff to only add the
amount we are missing, or free the amount we are missing. This is less clean to
read but more explicit in what it is doing, as well as only emitting events for
values that make sense. Thanks,
Signed-off-by: Josef Bacik <jbacik@fb.com>
Signed-off-by: David Sterba <dsterba@suse.com>
I see no_space in v4.4-rc1 again in xfstests generic/102.
It happened randomly in some node only.
(one of 4 phy-node, and a kvm with non-virtio block driver)
By bisect, we can found the first-bad is:
commit bdced438ac ("block: setup bi_phys_segments after splitting")'
But above patch only triggered the bug by making bio operation
faster(or slower).
Main reason is in our space_allocating code, we need to commit
page writeback before wait it complish, this patch fixed above
bug.
BTW, there is another reason for generic/102 fail, caused by
disable default mixed-blockgroup, I'll fix it in xfstests.
Signed-off-by: Zhao Lei <zhaolei@cn.fujitsu.com>
Signed-off-by: Chris Mason <clm@fb.com>
wait_for_snapshot_creation() is in same group with oher two:
btrfs_start_write_no_snapshoting()
btrfs_end_write_no_snapshoting()
Rename wait_for_snapshot_creation() and move it into same place
with other two.
Signed-off-by: Zhao Lei <zhaolei@cn.fujitsu.com>
Signed-off-by: Chris Mason <clm@fb.com>
If we return 1 here, then the caller treats it as an error and returns
-EINVAL. It causes a static checker warning to treat positive returns
as an error.
Fixes: 1aba86d67f ('Btrfs: fix easily get into ENOSPC in mixed case')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Overloading extent_map->bdev to struct map_lookup * might have started out
as a means to an end, but it's a pattern that's used all over the place
now. Let's get rid of the casting and just add a union instead.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Replace the integers by enums for better readability. The value 2 does
not have any meaning since a717531942
"Btrfs: do less aggressive btree readahead" (2009-01-22).
Signed-off-by: David Sterba <dsterba@suse.com>
We use many constants to represent size and offset value. And to make
code readable we use '256 * 1024 * 1024' instead of '268435456' to
represent '256MB'. However we can make far more readable with 'SZ_256MB'
which is defined in the 'linux/sizes.h'.
So this patch replaces 'xxx * 1024 * 1024' kind of expression with
single 'SZ_xxxMB' if 'xxx' is a power of 2 then 'xxx * SZ_1M' if 'xxx' is
not a power of 2. And I haven't touched to '4096' & '8192' because it's
more intuitive than 'SZ_4KB' & 'SZ_8KB'.
Signed-off-by: Byongho Lee <bhlee.kernel@gmail.com>
Signed-off-by: David Sterba <dsterba@suse.com>
btrfs_delayed_extent_op can be packed in a better way, it's 40 bytes now
and has 8 unused bytes. Reducing the level type to u8 makes it possible
to squeeze it to the padding byte after key. The bitfields were switched
to bool as there's space to store the full byte without increasing the
whole structure, besides that the generated assembly is smaller.
struct btrfs_delayed_extent_op {
struct btrfs_disk_key key; /* 0 17 */
u8 level; /* 17 1 */
bool update_key; /* 18 1 */
bool update_flags; /* 19 1 */
bool is_data; /* 20 1 */
/* XXX 3 bytes hole, try to pack */
u64 flags_to_set; /* 24 8 */
/* size: 32, cachelines: 1, members: 6 */
/* sum members: 29, holes: 1, sum holes: 3 */
/* last cacheline: 32 bytes */
};
The final size is 32 bytes which gives +26 object per slab page.
text data bss dec hex filename
938811 43670 23144 1005625 f5839 fs/btrfs/btrfs.ko.before
938747 43670 23144 1005561 f57f9 fs/btrfs/btrfs.ko.after
Signed-off-by: David Sterba <dsterba@suse.com>
While running a stress test I ran into the following trace/transaction
abort:
[471626.672243] ------------[ cut here ]------------
[471626.673322] WARNING: CPU: 9 PID: 19107 at fs/btrfs/extent-tree.c:3740 btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]()
[471626.675492] BTRFS: Transaction aborted (error -2)
[471626.676748] Modules linked in: btrfs dm_flakey dm_mod crc32c_generic xor raid6_pq nfsd auth_rpcgss oid_registry nfs_acl nfs lockd grace fscache sunrpc loop fuse parport_pc i2c_piix
[471626.688802] CPU: 14 PID: 19107 Comm: fsstress Tainted: G W 4.3.0-rc5-btrfs-next-17+ #1
[471626.690148] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.1-0-g4adadbd-20150316_085822-nilsson.home.kraxel.org 04/01/2014
[471626.691901] 0000000000000000 ffff880016037cf0 ffffffff812566f4 ffff880016037d38
[471626.695009] ffff880016037d28 ffffffff8104d0a6 ffffffffa040c84e 00000000fffffffe
[471626.697490] ffff88011fe855f8 ffff88000c484cb0 ffff88000d195000 ffff880016037d90
[471626.699201] Call Trace:
[471626.699804] [<ffffffff812566f4>] dump_stack+0x4e/0x79
[471626.701049] [<ffffffff8104d0a6>] warn_slowpath_common+0x9f/0xb8
[471626.702542] [<ffffffffa040c84e>] ? btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]
[471626.704326] [<ffffffff8104d107>] warn_slowpath_fmt+0x48/0x50
[471626.705636] [<ffffffffa0403717>] ? write_one_cache_group.isra.32+0x77/0x82 [btrfs]
[471626.707048] [<ffffffffa040c84e>] btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]
[471626.708616] [<ffffffffa048a50a>] commit_cowonly_roots+0x1d7/0x25a [btrfs]
[471626.709950] [<ffffffffa041e34a>] btrfs_commit_transaction+0x4c4/0x991 [btrfs]
[471626.711286] [<ffffffff81081c61>] ? signal_pending_state+0x31/0x31
[471626.712611] [<ffffffffa03f6df4>] btrfs_sync_fs+0x145/0x1ad [btrfs]
[471626.715610] [<ffffffff811962a2>] ? SyS_tee+0x226/0x226
[471626.716718] [<ffffffff811962c2>] sync_fs_one_sb+0x20/0x22
[471626.717672] [<ffffffff8116fc01>] iterate_supers+0x75/0xc2
[471626.718800] [<ffffffff8119669a>] sys_sync+0x52/0x80
[471626.719990] [<ffffffff8147cd97>] entry_SYSCALL_64_fastpath+0x12/0x6f
[471626.721835] ---[ end trace baf57f43d76693f4 ]---
[471626.722954] BTRFS: error (device sdc) in btrfs_write_dirty_block_groups:3740: errno=-2 No such entry
This is a very rare situation and it happened due to a race between a free
space endio worker and writing the space caches for dirty block groups at
a transaction's commit critical section. The steps leading to this are:
1) A task calls btrfs_commit_transaction() and starts the writeout of the
space caches for all currently dirty block groups (i.e. it calls
btrfs_start_dirty_block_groups());
2) The previous step starts writeback for space caches;
3) When the writeback finishes it queues jobs for free space endio work
queue (fs_info->endio_freespace_worker) that execute
btrfs_finish_ordered_io();
4) The task committing the transaction sets the transaction's state
to TRANS_STATE_COMMIT_DOING and shortly after calls
btrfs_write_dirty_block_groups();
5) A free space endio job joins the transaction, through
btrfs_join_transaction_nolock(), and updates a free space inode item
in the root tree through btrfs_update_inode_fallback();
6) Updating the free space inode item resulted in COWing one or more
nodes/leaves of the root tree, and that resulted in creating a new
metadata block group, which gets added to the transaction's list
of dirty block groups (this is a very rare case);
7) The free space endio job has not released yet its transaction handle
at this point, so the new metadata block group was not yet fully
created (didn't go through btrfs_create_pending_block_groups() yet);
8) The transaction commit task sees the new metadata block group in
the transaction's list of dirty block groups and processes it.
When it attempts to update the block group's block group item in
the extent tree, through write_one_cache_group(), it isn't able
to find it and aborts the transaction with error -ENOENT - this
is because the free space endio job hasn't yet released its
transaction handle (which calls btrfs_create_pending_block_groups())
and therefore the block group item was not yet added to the extent
tree.
Fix this waiting for free space endio jobs if we fail to find a block
group item in the extent tree and then retry once updating the block
group item.
Signed-off-by: Filipe Manana <fdmanana@suse.com>
This is a short term solution to make sure btrfs_run_delayed_refs()
doesn't change the extent tree while we are scanning it to create the
free space tree.
Longer term we need to synchronize scanning the block groups one by one,
similar to what happens during a balance.
Signed-off-by: Chris Mason <clm@fb.com>
We call btrfs_write_dirty_block_groups() in the critical section of a
transaction's commit, when no other tasks can join the transaction and
add more block groups to the transaction's list of dirty block groups,
so we not taking the dirty block groups spinlock when checking for the
list's emptyness, grabbing its first element or deleting elements from
it.
However there's a special and rare case where we can have a concurrent
task adding elements to this list. We trigger writeback for space
caches before at btrfs_start_dirty_block_groups() and in past iterations
of the loop at btrfs_write_dirty_block_groups(), this means that when
the writeback finishes (which happens asynchronously) it creates a
task for the endio free space work queue that executes
btrfs_finish_ordered_io() - this function is able to join the transaction,
through btrfs_join_transaction_nolock(), and update the free space cache's
inode item in the root tree, which can result in COWing nodes of this tree
and therefore allocation of a new block group can happen, which gets added
to the transaction's list of dirty block groups while the transaction
commit task is operating on it concurrently.
So fix this by taking the dirty block groups spinlock before doing
operations on the dirty block groups list at
btrfs_write_dirty_block_groups().
Signed-off-by: Filipe Manana <fdmanana@suse.com>
The free space tree is updated in tandem with the extent tree. There are
only a handful of places where we need to hook in:
1. Block group creation
2. Block group deletion
3. Delayed refs (extent creation and deletion)
4. Block group caching
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
The free space cache has turned out to be a scalability bottleneck on
large, busy filesystems. When the cache for a lot of block groups needs
to be written out, we can get extremely long commit times; if this
happens in the critical section, things are especially bad because we
block new transactions from happening.
The main problem with the free space cache is that it has to be written
out in its entirety and is managed in an ad hoc fashion. Using a B-tree
to store free space fixes this: updates can be done as needed and we get
all of the benefits of using a B-tree: checksumming, RAID handling,
well-understood behavior.
With the free space tree, we get commit times that are about the same as
the no cache case with load times slower than the free space cache case
but still much faster than the no cache case. Free space is represented
with extents until it becomes more space-efficient to use bitmaps,
giving us similar space overhead to the free space cache.
The operations on the free space tree are: adding and removing free
space, handling the creation and deletion of block groups, and loading
the free space for a block group. We can also create the free space tree
by walking the extent tree and clear the free space tree.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
We're also going to load the free space tree from caching_thread(), so
we should refactor some of the common code.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Chris Mason <clm@fb.com>
As of my previous change titled "Btrfs: fix scrub preventing unused block
groups from being deleted", the following warning at
extent-tree.c:btrfs_delete_unused_bgs() can be hit when we mount the a
filesysten with "-o discard":
10263 void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
10264 {
(...)
10405 if (trimming) {
10406 WARN_ON(!list_empty(&block_group->bg_list));
10407 spin_lock(&trans->transaction->deleted_bgs_lock);
10408 list_move(&block_group->bg_list,
10409 &trans->transaction->deleted_bgs);
10410 spin_unlock(&trans->transaction->deleted_bgs_lock);
10411 btrfs_get_block_group(block_group);
10412 }
(...)
This happens because scrub can now add back the block group to the list of
unused block groups (fs_info->unused_bgs). This is dangerous because we
are moving the block group from the unused block groups list to the list
of deleted block groups without holding the lock that protects the source
list (fs_info->unused_bgs_lock).
The following diagram illustrates how this happens:
CPU 1 CPU 2
cleaner_kthread()
btrfs_delete_unused_bgs()
sees bg X in list
fs_info->unused_bgs
deletes bg X from list
fs_info->unused_bgs
scrub_enumerate_chunks()
searches device tree using
its commit root
finds device extent for
block group X
gets block group X from the tree
fs_info->block_group_cache_tree
(via btrfs_lookup_block_group())
sets bg X to RO (again)
scrub_chunk(bg X)
sets bg X back to RW mode
adds bg X to the list
fs_info->unused_bgs again,
since it's still unused and
currently not in that list
sets bg X to RO mode
btrfs_remove_chunk(bg X)
--> discard is enabled and bg X
is in the fs_info->unused_bgs
list again so the warning is
triggered
--> we move it from that list into
the transaction's delete_bgs
list, but we can have another
task currently manipulating
the first list (fs_info->unused_bgs)
Fix this by using the same lock (fs_info->unused_bgs_lock) to protect both
the list of unused block groups and the list of deleted block groups. This
makes it safe and there's not much worry for more lock contention, as this
lock is seldom used and only the cleaner kthread adds elements to the list
of deleted block groups. The warning goes away too, as this was previously
an impossible case (and would have been better a BUG_ON/ASSERT) but it's
not impossible anymore.
Reproduced with fstest btrfs/073 (using MOUNT_OPTIONS="-o discard").
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Commit 0ed4792 ('btrfs: qgroup: Switch to new extent-oriented qgroup
mechanism.') removed our qgroup accounting during
btrfs_drop_snapshot(). Predictably, this results in qgroup numbers
going bad shortly after a snapshot is removed.
Fix this by adding a dirty extent record when we encounter extents during
our shared subtree walk. This effectively restores the functionality we had
with the original shared subtree walking code in 1152651 (btrfs: qgroup:
account shared subtrees during snapshot delete).
The idea with the original patch (and this one) is that shared subtrees can
get skipped during drop_snapshot. The shared subtree walk then allows us a
chance to visit those extents and add them to the qgroup work for later
processing. This ultimately makes the accounting for drop snapshot work.
The new qgroup code nicely handles all the other extents during the tree
walk via the ref dec/inc functions so we don't have to add actions beyond
what we had originally.
Signed-off-by: Mark Fasheh <mfasheh@suse.de>
Signed-off-by: Chris Mason <clm@fb.com>
When a block group becomes unused and the cleaner kthread is currently
running, we can end up getting the current transaction aborted with error
-ENOENT when we try to commit the transaction, leading to the following
trace:
[59779.258768] WARNING: CPU: 3 PID: 5990 at fs/btrfs/extent-tree.c:3740 btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]()
[59779.272594] BTRFS: Transaction aborted (error -2)
(...)
[59779.291137] Call Trace:
[59779.291621] [<ffffffff812566f4>] dump_stack+0x4e/0x79
[59779.292543] [<ffffffff8104d0a6>] warn_slowpath_common+0x9f/0xb8
[59779.293435] [<ffffffffa04cb81f>] ? btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]
[59779.295000] [<ffffffff8104d107>] warn_slowpath_fmt+0x48/0x50
[59779.296138] [<ffffffffa04c2721>] ? write_one_cache_group.isra.32+0x77/0x82 [btrfs]
[59779.297663] [<ffffffffa04cb81f>] btrfs_write_dirty_block_groups+0x17c/0x214 [btrfs]
[59779.299141] [<ffffffffa0549b0d>] commit_cowonly_roots+0x1de/0x261 [btrfs]
[59779.300359] [<ffffffffa04dd5b6>] btrfs_commit_transaction+0x4c4/0x99c [btrfs]
[59779.301805] [<ffffffffa04b5df4>] btrfs_sync_fs+0x145/0x1ad [btrfs]
[59779.302893] [<ffffffff81196634>] sync_filesystem+0x7f/0x93
(...)
[59779.318186] ---[ end trace 577e2daff90da33a ]---
The following diagram illustrates a sequence of steps leading to this
problem:
CPU 1 CPU 2
<at transaction N>
adds bg A to list
fs_info->unused_bgs
adds bg B to list
fs_info->unused_bgs
<transaction kthread
commits transaction N
and wakes up the
cleaner kthread>
cleaner kthread
delete_unused_bgs()
sees bg A in list
fs_info->unused_bgs
btrfs_start_transaction()
<transaction N + 1 starts>
deletes bg A
update_block_group(bg C)
--> adds bg C to list
fs_info->unused_bgs
deletes bg B
sees bg C in the list
fs_info->unused_bgs
btrfs_remove_chunk(bg C)
btrfs_remove_block_group(bg C)
--> checks if the block group
is in a dirty list, and
because it isn't now, it
does nothing
--> the block group item
is deleted from the
extent tree
--> adds bg C to list
transaction->dirty_bgs
some task calls
btrfs_commit_transaction(t N + 1)
commit_cowonly_roots()
btrfs_write_dirty_block_groups()
--> sees bg C in cur_trans->dirty_bgs
--> calls write_one_cache_group()
which returns -ENOENT because
it did not find the block group
item in the extent tree
--> transaction aborte with -ENOENT
because write_one_cache_group()
returned that error
So fix this by adding a block group to the list of dirty block groups
before adding it to the list of unused block groups.
This happened on a stress test using fsstress plus concurrent calls to
fallocate 20G and truncate (releasing part of the space allocated with
fallocate).
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
