redonkable
/
alistair23-linux
1
0
Fork 0
Code Releases Activity
alistair23-linux/fs
History
Douglas Anderson a48b609c8d bdev: Reduce time holding bd_mutex in sync in blkdev_close() 
[ Upstream commit b849dd84b6 ]

While trying to "dd" to the block device for a USB stick, I
encountered a hung task warning (blocked for > 120 seconds).  I
managed to come up with an easy way to reproduce this on my system
(where /dev/sdb is the block device for my USB stick) with:

  while true; do dd if=/dev/zero of=/dev/sdb bs=4M; done

With my reproduction here are the relevant bits from the hung task
detector:

 INFO: task udevd:294 blocked for more than 122 seconds.
 ...
 udevd           D    0   294      1 0x00400008
 Call trace:
  ...
  mutex_lock_nested+0x40/0x50
  __blkdev_get+0x7c/0x3d4
  blkdev_get+0x118/0x138
  blkdev_open+0x94/0xa8
  do_dentry_open+0x268/0x3a0
  vfs_open+0x34/0x40
  path_openat+0x39c/0xdf4
  do_filp_open+0x90/0x10c
  do_sys_open+0x150/0x3c8
  ...

 ...
 Showing all locks held in the system:
 ...
 1 lock held by dd/2798:
  #0: ffffff814ac1a3b8 (&bdev->bd_mutex){+.+.}, at: __blkdev_put+0x50/0x204
 ...
 dd              D    0  2798   2764 0x00400208
 Call trace:
  ...
  schedule+0x8c/0xbc
  io_schedule+0x1c/0x40
  wait_on_page_bit_common+0x238/0x338
  __lock_page+0x5c/0x68
  write_cache_pages+0x194/0x500
  generic_writepages+0x64/0xa4
  blkdev_writepages+0x24/0x30
  do_writepages+0x48/0xa8
  __filemap_fdatawrite_range+0xac/0xd8
  filemap_write_and_wait+0x30/0x84
  __blkdev_put+0x88/0x204
  blkdev_put+0xc4/0xe4
  blkdev_close+0x28/0x38
  __fput+0xe0/0x238
  ____fput+0x1c/0x28
  task_work_run+0xb0/0xe4
  do_notify_resume+0xfc0/0x14bc
  work_pending+0x8/0x14

The problem appears related to the fact that my USB disk is terribly
slow and that I have a lot of RAM in my system to cache things.
Specifically my writes seem to be happening at ~15 MB/s and I've got
~4 GB of RAM in my system that can be used for buffering.  To write 4
GB of buffer to disk thus takes ~4000 MB / ~15 MB/s = ~267 seconds.

The 267 second number is a problem because in __blkdev_put() we call
sync_blockdev() while holding the bd_mutex.  Any other callers who
want the bd_mutex will be blocked for the whole time.

The problem is made worse because I believe blkdev_put() specifically
tells other tasks (namely udev) to go try to access the device at right
around the same time we're going to hold the mutex for a long time.

Putting some traces around this (after disabling the hung task detector),
I could confirm:
 dd:    437.608600: __blkdev_put() right before sync_blockdev() for sdb
 udevd: 437.623901: blkdev_open() right before blkdev_get() for sdb
 dd:    661.468451: __blkdev_put() right after sync_blockdev() for sdb
 udevd: 663.820426: blkdev_open() right after blkdev_get() for sdb

A simple fix for this is to realize that sync_blockdev() works fine if
you're not holding the mutex.  Also, it's not the end of the world if
you sync a little early (though it can have performance impacts).
Thus we can make a guess that we're going to need to do the sync and
then do it without holding the mutex.  We still do one last sync with
the mutex but it should be much, much faster.

With this, my hung task warnings for my test case are gone.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
 		2020-10-01 13:17:55 +02:00
..
9p 9p: Fix memory leak in v9fs_mount 2020-08-19 08:16:24 +02:00
adfs Merge branch 'work.adfs' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 11:33:22 -07:00
affs affs: fix basic permission bits to actually work 2020-09-09 19:12:34 +02:00
afs rxrpc: Make rxrpc_kernel_get_srtt() indicate validity 2020-09-09 19:12:23 +02:00
autofs autofs: fix a leak in autofs_expire_indirect() 2019-10-25 00:03:11 -04:00
befs fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
bfs fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
btrfs btrfs: fix setting last_trans for reloc roots 2020-10-01 13:17:55 +02:00
cachefiles cachefiles: Fix race between read_waiter and read_copier involving op->to_do 2020-06-03 08:21:11 +02:00
ceph ceph: ensure we have a new cap before continuing in fill_inode 2020-10-01 13:17:29 +02:00
cifs CIFS: Properly process SMB3 lease breaks 2020-10-01 13:17:21 +02:00
coda y2038: add inode timestamp clamping 2019-09-19 09:42:37 -07:00
configfs configfs: fix config_item refcnt leak in configfs_rmdir() 2020-05-27 17:46:30 +02:00
cramfs cramfs: fix usage on non-MTD device 2019-11-23 21:44:49 -05:00
crypto fscrypt: don't evict dirty inodes after removing key 2020-03-18 07:17:53 +01:00
debugfs debugfs: Fix module state check condition 2020-09-17 13:47:55 +02:00
devpts devpts_pty_kill(): don't bother with d_delete() 2019-09-03 09:30:56 -04:00
dlm dlm: Fix kobject memleak 2020-08-19 08:16:21 +02:00
ecryptfs ecryptfs: replace BUG_ON with error handling code 2020-02-28 17:22:26 +01:00
efivarfs Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
efs fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
erofs erofs: fix extended inode could cross boundary 2020-08-19 08:16:26 +02:00
exportfs exportfs_decode_fh(): negative pinned may become positive without the parent locked 2019-11-10 11:56:05 -05:00
ext2 ext2: don't update mtime on COW faults 2020-09-09 19:12:30 +02:00
ext4 ext4: mark block bitmap corrupted when found instead of BUGON 2020-10-01 13:17:42 +02:00
f2fs f2fs: stop GC when the victim becomes fully valid 2020-10-01 13:17:21 +02:00
fat fat: don't allow to mount if the FAT length == 0 2020-06-17 16:40:36 +02:00
freevxfs fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
fscache Revert "Merge tag 'keys-acl-20190703' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs" 2019-07-10 18:43:43 -07:00
fuse fuse: fix weird page warning 2020-07-29 10:18:28 +02:00
gfs2 gfs2: clean up iopen glock mess in gfs2_create_inode 2020-10-01 13:17:20 +02:00
hfs treewide: Add SPDX license identifier - Makefile/Kconfig 2019-05-21 10:50:46 +02:00
hfsplus hfsplus: fix crash and filesystem corruption when deleting files 2020-04-17 10:50:22 +02:00
hostfs This pull request contains the following changes for UML: 2019-05-12 17:52:13 -04:00
hpfs fs: hpfs: Initialize filesystem timestamp ranges 2019-08-30 08:11:25 -07:00
hugetlbfs hugetlbfs: prevent filesystem stacking of hugetlbfs 2020-09-03 11:26:48 +02:00
iomap iomap: Fix overflow in iomap_page_mkwrite 2020-10-01 13:17:15 +02:00
isofs y2038: add inode timestamp clamping 2019-09-19 09:42:37 -07:00
jbd2 jbd2: abort journal if free a async write error metadata buffer 2020-09-03 11:26:56 +02:00
jffs2 jffs2: fix UAF problem 2020-08-26 10:40:56 +02:00
jfs y2038: add inode timestamp clamping 2019-09-19 09:42:37 -07:00
kernfs kernfs: do not call fsnotify() with name without a parent 2020-08-19 08:16:12 +02:00
lockd lockd: Make two symbols static 2019-07-03 17:52:09 -04:00
minix fs/minix: remove expected error message in block_to_path() 2020-08-21 13:05:37 +02:00
nfs NFS: Fix races nfs_page_group_destroy() vs nfs_destroy_unlinked_subrequests() 2020-10-01 13:17:52 +02:00
nfs_common treewide: Add SPDX license identifier - Makefile/Kconfig 2019-05-21 10:50:46 +02:00
nfsd nfsd: Don't add locks to closed or closing open stateids 2020-10-01 13:17:43 +02:00
nilfs2 nilfs2: fix null pointer dereference at nilfs_segctor_do_construct() 2020-06-17 16:40:29 +02:00
nls treewide: Add SPDX license identifier - Makefile/Kconfig 2019-05-21 10:50:46 +02:00
notify fanotify: fix ignore mask logic for events on child and on dir 2020-06-17 16:40:24 +02:00
ntfs utimes: Clamp the timestamps in notify_change() 2020-02-11 04:35:12 -08:00
ocfs2 ocfs2: change slot number type s16 to u16 2020-08-21 13:05:26 +02:00
omfs fs: omfs: Initialize filesystem timestamp ranges 2019-08-30 08:11:25 -07:00
openpromfs Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
orangefs orangefs: get rid of knob code... 2020-08-21 13:05:29 +02:00
overlayfs ovl: fix unneeded call to ovl_change_flags() 2020-07-22 09:33:12 +02:00
proc proc: io_accounting: Use new infrastructure to fix deadlocks in execve 2020-10-01 13:17:48 +02:00
pstore pstore: Fix linking when crypto API disabled 2020-08-19 08:16:27 +02:00
qnx4 fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
qnx6 fs: Fill in max and min timestamps in superblock 2019-08-30 07:27:17 -07:00
quota fs: avoid softlockups in s_inodes iterators 2020-01-12 12:21:37 +01:00
ramfs vfs: Convert ramfs, shmem, tmpfs, devtmpfs, rootfs to use the new mount API 2019-09-12 21:05:34 -04:00
reiserfs reiserfs: prevent NULL pointer dereference in reiserfs_insert_item() 2020-02-24 08:37:00 +01:00
romfs romfs: fix uninitialized memory leak in romfs_dev_read() 2020-08-26 10:40:51 +02:00
squashfs Merge branch 'work.mount2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-09-19 10:06:57 -07:00
sysfs Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
sysv fs: sysv: Initialize filesystem timestamp ranges 2019-08-30 07:27:18 -07:00
tracefs tracing: Do not create tracefs files if tracefs lockdown is in effect 2019-10-12 20:49:07 -04:00
ubifs ubifs: Fix out-of-bounds memory access caused by abnormal value of node_len 2020-10-01 13:17:51 +02:00
udf udf: Fix free space reporting for metadata and virtual partitions 2020-02-24 08:36:44 +01:00
ufs fs/ufs: avoid potential u32 multiplication overflow 2020-08-21 13:05:37 +02:00
unicode unicode: make array 'token' static const, makes object smaller 2019-09-17 11:48:24 -04:00
verity fs-verity: support builtin file signatures 2019-08-12 19:33:50 -07:00
xfs xfs: prohibit fs freezing when using empty transactions 2020-10-01 13:17:49 +02:00
Kconfig fs-verity for 5.4 2019-09-18 16:59:14 -07:00
Kconfig.binfmt binfmt_flat: make support for old format binaries optional 2019-06-24 09:16:47 +10:00
Makefile fs-verity for 5.4 2019-09-18 16:59:14 -07:00
aio.c aio: fix async fsync creds 2020-06-17 16:40:24 +02:00
anon_inodes.c Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
attr.c utimes: Clamp the timestamps in notify_change() 2020-02-11 04:35:12 -08:00
bad_inode.c
…
binfmt_aout.c treewide: Add SPDX license identifier for more missed files 2019-05-21 10:50:45 +02:00
binfmt_elf.c fs/binfmt_elf.c: allocate initialized memory in fill_thread_core_info() 2020-06-03 08:21:27 +02:00
binfmt_elf_fdpic.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 152 2019-05-30 11:26:32 -07:00
binfmt_em86.c treewide: Add SPDX license identifier for more missed files 2019-05-21 10:50:45 +02:00
binfmt_flat.c binfmt_flat: revert "binfmt_flat: don't offset the data start" 2020-09-03 11:26:39 +02:00
binfmt_misc.c Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
binfmt_script.c treewide: Add SPDX license identifier for more missed files 2019-05-21 10:50:45 +02:00
block_dev.c bdev: Reduce time holding bd_mutex in sync in blkdev_close() 2020-10-01 13:17:55 +02:00
buffer.c fs: prevent BUG_ON in submit_bh_wbc() 2020-09-03 11:26:57 +02:00
char_dev.c chardev: Avoid potential use-after-free in 'chrdev_open()' 2020-01-14 20:08:18 +01:00
compat.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 500 2019-06-19 17:09:55 +02:00
compat_binfmt_elf.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 193 2019-05-30 11:29:21 -07:00
compat_ioctl.c fix compat handling of FICLONERANGE, FIDEDUPERANGE and FS_IOC_FIEMAP 2020-01-09 10:20:05 +01:00
coredump.c coredump: fix crash when umh is disabled 2020-05-14 07:58:27 +02:00
d_path.c [PATCH] fix d_absolute_path() interplay with fsmount() 2019-08-30 19:31:09 -04:00
dax.c dax: pass NOWAIT flag to iomap_apply 2020-03-05 16:43:36 +01:00
dcache.c fix dget_parent() fastpath race 2020-10-01 13:17:19 +02:00
dcookies.c treewide: Add SPDX license identifier for missed files 2019-05-21 10:50:45 +02:00
direct-io.c fs/direct-io.c: fix kernel-doc warning 2019-10-14 15:04:01 -07:00
drop_caches.c fs: avoid softlockups in s_inodes iterators 2020-01-12 12:21:37 +01:00
eventfd.c eventfd: track eventfd_signal() recursion depth 2020-02-11 04:35:37 -08:00
eventpoll.c fix regression in "epoll: Keep a reference on files added to the check list" 2020-09-09 19:12:28 +02:00
exec.c exec: Add exec_update_mutex to replace cred_guard_mutex 2020-10-01 13:17:47 +02:00
fcntl.c fs: mark expected switch fall-throughs 2019-04-08 18:21:02 -05:00
fhandle.c fs/handle.c - fix up kerneldoc 2019-08-07 21:51:47 -04:00
file.c fix multiplication overflow in copy_fdtable() 2020-05-27 17:46:12 +02:00
file_table.c vfs: Export flush_delayed_fput for use by knfsd. 2019-08-19 11:00:39 -04:00
filesystems.c fs/filesystems.c: downgrade user-reachable WARN_ONCE() to pr_warn_once() 2020-04-17 10:50:21 +02:00
fs-writeback.c writeback: Fix sync livelock due to b_dirty_time processing 2020-09-03 11:27:04 +02:00
fs_context.c vfs: subtype handling moved to fuse 2019-09-06 21:28:49 +02:00
fs_parser.c vfs: Make fs_parse() handle fs_param_is_fd-type params better 2019-09-12 21:06:14 -04:00
fs_pin.c switch the remnants of releasing the mountpoint away from fs_pin 2019-07-16 22:52:37 -04:00
fs_struct.c treewide: Add SPDX license identifier for missed files 2019-05-21 10:50:45 +02:00
fs_types.c
…
fsopen.c Merge branch 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2019-07-19 10:42:02 -07:00
inode.c futex: Fix inode life-time issue 2020-03-25 08:25:58 +01:00
internal.h fs: move guard_bio_eod() after bio_set_op_attrs 2020-01-17 19:48:21 +01:00
io_uring.c io_uring: Fix NULL pointer dereference in io_sq_wq_submit_work() 2020-09-03 11:27:11 +02:00
ioctl.c compat_ioctl: add compat_ptr_ioctl() 2019-12-17 19:55:30 +01:00
libfs.c libfs: fix infoleak in simple_attr_read() 2020-04-01 11:02:17 +02:00
locks.c locks: reinstate locks_delete_block optimization 2020-03-25 08:25:41 +01:00
mbcache.c treewide: Add SPDX license identifier for more missed files 2019-05-21 10:50:45 +02:00
mount.h switch the remnants of releasing the mountpoint away from fs_pin 2019-07-16 22:52:37 -04:00
mpage.c fs: move guard_bio_eod() after bio_set_op_attrs 2020-01-17 19:48:21 +01:00
namei.c namei: only return -ECHILD from follow_dotdot_rcu() 2020-03-05 16:43:48 +01:00
namespace.c fs/namespace.c: fix use-after-free of mount in mnt_warn_timestamp_expiry() 2019-10-16 23:15:09 -04:00
no-block.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 152 2019-05-30 11:26:32 -07:00
nsfs.c vfs: Convert nsfs to use the new mount API 2019-05-25 18:00:06 -04:00
open.c cifs_atomic_open(): fix double-put on late allocation failure 2020-03-18 07:17:51 +01:00
pipe.c vfs: Convert pipe to use the new mount API 2019-05-25 18:00:07 -04:00
pnode.c propagate_one(): mnt_set_mountpoint() needs mount_lock 2020-05-02 08:48:44 +02:00
pnode.h treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 209 2019-05-30 11:29:53 -07:00
posix_acl.c treewide: Add SPDX license identifier for missed files 2019-05-21 10:50:45 +02:00
proc_namespace.c vfs: subtype handling moved to fuse 2019-09-06 21:28:49 +02:00
read_write.c fs: allow deduplication of eof block into the end of the destination file 2020-02-11 04:35:23 -08:00
readdir.c readdir: be more conservative with directory entry names 2020-01-29 16:45:31 +01:00
select.c fs/select.c: use struct_size() in kmalloc() 2019-07-16 19:23:25 -07:00
seq_file.c seq_file: fix problem when seeking mid-record 2019-08-13 16:06:52 -07:00
signalfd.c fs/signalfd.c: fix inconsistent return codes for signalfd4 2020-08-26 10:40:58 +02:00
splice.c splice: only read in as much information as there is pipe buffer space 2019-12-17 19:56:52 +01:00
stack.c treewide: Add SPDX license identifier for missed files 2019-05-21 10:50:45 +02:00
stat.c
…
statfs.c vfs: Fix EOVERFLOW testing in put_compat_statfs64 2019-10-03 14:21:35 -07:00
super.c Fix use after free in get_tree_bdev() 2020-05-06 08:15:15 +02:00
sync.c fs/sync.c: sync_file_range(2) may use WB_SYNC_ALL writeback 2019-05-14 09:47:50 -07:00
timerfd.c timerfd: Prepare for PREEMPT_RT 2019-08-01 20:51:23 +02:00
userfaultfd.c userfaultfd: require CAP_SYS_PTRACE for UFFD_FEATURE_EVENT_FORK 2020-01-04 19:18:32 +01:00
utimes.c utimes: Clamp the timestamps in notify_change() 2020-02-11 04:35:12 -08:00
xattr.c xattr: break delegations in {set,remove}xattr 2020-08-11 15:33:39 +02:00