linux.git/include/linux/backing-dev-defs.h, branch v7.2-rc1 Linux kernel source tree mm: kick writeback flusher for IOCB_DONTCACHE with targeted dirty tracking 2026-06-04T08:16:51+00:00 Jeff Layton jlayton@kernel.org 2026-05-11T11:58:29+00:00 e1bf79628453e6afac81ffa57f4f40f28e5512ff The IOCB_DONTCACHE writeback path in generic_write_sync() calls filemap_flush_range() on every write, submitting writeback inline in the writer's context. Perf lock contention profiling shows the performance problem is not lock contention but the writeback submission work itself — walking the page tree and submitting I/O blocks the writer for milliseconds, inflating p99.9 latency from 23ms (buffered) to 93ms (dontcache). Replace the inline filemap_flush_range() call with a flusher kick that drains dirty pages in the background. This moves writeback submission completely off the writer's hot path. To avoid flushing unrelated buffered dirty data, add a dedicated WB_start_dontcache bit and wb_check_start_dontcache() handler that uses the per-wb WB_DONTCACHE_DIRTY counter to determine how many pages to write back. The flusher writes back that many pages from the oldest dirty inodes (not restricted to dontcache-specific inodes). This helps preserve I/O batching while limiting the scope of expedited writeback. Like WB_start_all, the WB_start_dontcache bit coalesces multiple DONTCACHE writes into a single flusher wakeup without per-write allocations. Use test_and_clear_bit to atomically consume the kick request before reading the dirty counter and starting writeback, so that concurrent DONTCACHE writes during writeback can re-set the bit and schedule a follow-up flusher run. Read the dirty counter with wb_stat_sum() (aggregating per-CPU batches) rather than wb_stat() (which reads only the global counter) to ensure small writes below the percpu batch threshold are visible to the flusher. In filemap_dontcache_kick_writeback(), set the WB_start_dontcache bit inside the unlocked_inode_to_wb_begin/end section for correct cgroup writeback domain targeting, but defer the wb_wakeup() call until after the section ends, since wb_wakeup() uses spin_unlock_irq() which would unconditionally re-enable interrupts while the i_pages xa_lock may still be held under irqsave during a cgroup writeback switch. Pin the wb with wb_get() inside the RCU critical section before calling wb_wakeup() outside it, since cgroup bdi_writeback structures are RCU-freed and the wb pointer could become invalid after unlocked_inode_to_wb_end() drops the RCU read lock. Also add WB_REASON_DONTCACHE as a new writeback reason for tracing visibility. dontcache-bench results (same host, T6F_SKL_1920GBF, 251 GiB RAM, xfs on NVMe, fio io_uring): Buffered and direct I/O paths are unaffected by this patchset. All improvements are confined to the dontcache path: Single-stream throughput (MB/s): Before After Change seq-write/dontcache 298 897 +201% rand-write/dontcache 131 236 +80% Tail latency improvements (seq-write/dontcache): p99: 135,266 us -> 23,986 us (-82%) p99.9: 8,925,479 us -> 28,443 us (-99.7%) Multi-writer (4 jobs, sequential write): Before After Change dontcache aggregate (MB/s) 2,529 4,532 +79% dontcache p99 (us) 8,553 1,002 -88% dontcache p99.9 (us) 109,314 1,057 -99% Dontcache multi-writer throughput now matches buffered (4,532 vs 4,616 MB/s). 32-file write (Axboe test): Before After Change dontcache aggregate (MB/s) 1,548 3,499 +126% dontcache p99 (us) 10,170 602 -94% Peak dirty pages (MB) 1,837 213 -88% Dontcache now reaches 81% of buffered throughput (was 35%). Competing writers (dontcache vs buffered, separate files): Before After buffered writer 868 433 MB/s dontcache writer 415 433 MB/s Aggregate 1,284 866 MB/s Previously the buffered writer starved the dontcache writer 2:1. With per-bdi_writeback tracking, both writers now receive equal bandwidth. The aggregate matches the buffered-vs-buffered baseline (863 MB/s), indicating fair sharing regardless of I/O mode. The dontcache writer's p99.9 latency collapsed from 119 ms to 33 ms (-73%), eliminating the severe periodic stalls seen in the baseline. Both writers now share identical latency profiles, matching the buffered-vs-buffered pattern. The per-bdi_writeback dirty tracking dramatically reduces peak dirty pages in dontcache workloads, with the 32-file test dropping from 1.8 GB to 213 MB. Dontcache sequential write throughput triples and multi-writer throughput reaches parity with buffered I/O, with tail latencies collapsing by 1-2 orders of magnitude. Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://patch.msgid.link/20260511-dontcache-v7-3-2848ddce8090@kernel.org Reviewed-by: Jan Kara <jack@suse.cz> Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com> Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org> 
The IOCB_DONTCACHE writeback path in generic_write_sync() calls
filemap_flush_range() on every write, submitting writeback inline in
the writer's context.  Perf lock contention profiling shows the
performance problem is not lock contention but the writeback submission
work itself — walking the page tree and submitting I/O blocks the writer
for milliseconds, inflating p99.9 latency from 23ms (buffered) to 93ms
(dontcache).

Replace the inline filemap_flush_range() call with a flusher kick that
drains dirty pages in the background.  This moves writeback submission
completely off the writer's hot path.

To avoid flushing unrelated buffered dirty data, add a dedicated
WB_start_dontcache bit and wb_check_start_dontcache() handler that uses
the per-wb WB_DONTCACHE_DIRTY counter to determine how many pages to
write back.  The flusher writes back that many pages from the oldest dirty
inodes (not restricted to dontcache-specific inodes). This helps
preserve I/O batching while limiting the scope of expedited writeback.

Like WB_start_all, the WB_start_dontcache bit coalesces multiple
DONTCACHE writes into a single flusher wakeup without per-write
allocations.  Use test_and_clear_bit to atomically consume the kick
request before reading the dirty counter and starting writeback, so that
concurrent DONTCACHE writes during writeback can re-set the bit and
schedule a follow-up flusher run.

Read the dirty counter with wb_stat_sum() (aggregating per-CPU batches)
rather than wb_stat() (which reads only the global counter) to ensure
small writes below the percpu batch threshold are visible to the flusher.

In filemap_dontcache_kick_writeback(), set the WB_start_dontcache bit
inside the unlocked_inode_to_wb_begin/end section for correct cgroup
writeback domain targeting, but defer the wb_wakeup() call until after
the section ends, since wb_wakeup() uses spin_unlock_irq() which would
unconditionally re-enable interrupts while the i_pages xa_lock may still
be held under irqsave during a cgroup writeback switch. Pin the wb with
wb_get() inside the RCU critical section before calling wb_wakeup()
outside it, since cgroup bdi_writeback structures are RCU-freed and the
wb pointer could become invalid after unlocked_inode_to_wb_end() drops
the RCU read lock.

Also add WB_REASON_DONTCACHE as a new writeback reason for tracing
visibility.

dontcache-bench results (same host, T6F_SKL_1920GBF, 251 GiB RAM,
xfs on NVMe, fio io_uring):

Buffered and direct I/O paths are unaffected by this patchset. All
improvements are confined to the dontcache path:

Single-stream throughput (MB/s):
                        Before    After    Change
  seq-write/dontcache      298      897    +201%
  rand-write/dontcache     131      236     +80%

Tail latency improvements (seq-write/dontcache):
  p99:    135,266 us  ->  23,986 us   (-82%)
  p99.9: 8,925,479 us ->  28,443 us   (-99.7%)

Multi-writer (4 jobs, sequential write):
                                Before    After    Change
  dontcache aggregate (MB/s)     2,529    4,532     +79%
  dontcache p99 (us)             8,553    1,002     -88%
  dontcache p99.9 (us)         109,314    1,057     -99%

  Dontcache multi-writer throughput now matches buffered (4,532 vs
  4,616 MB/s).

32-file write (Axboe test):
                                Before    After    Change
  dontcache aggregate (MB/s)     1,548    3,499    +126%
  dontcache p99 (us)            10,170      602     -94%
  Peak dirty pages (MB)          1,837      213     -88%

  Dontcache now reaches 81% of buffered throughput (was 35%).

Competing writers (dontcache vs buffered, separate files):
                                Before    After
  buffered writer                  868      433 MB/s
  dontcache writer                 415      433 MB/s
  Aggregate                      1,284      866 MB/s

  Previously the buffered writer starved the dontcache writer 2:1.
  With per-bdi_writeback tracking, both writers now receive equal
  bandwidth. The aggregate matches the buffered-vs-buffered baseline
  (863 MB/s), indicating fair sharing regardless of I/O mode.

  The dontcache writer's p99.9 latency collapsed from 119 ms to
  33 ms (-73%), eliminating the severe periodic stalls seen in the
  baseline. Both writers now share identical latency profiles,
  matching the buffered-vs-buffered pattern.

The per-bdi_writeback dirty tracking dramatically reduces peak dirty
pages in dontcache workloads, with the 32-file test dropping from
1.8 GB to 213 MB. Dontcache sequential write throughput triples and
multi-writer throughput reaches parity with buffered I/O, with tail
latencies collapsing by 1-2 orders of magnitude.

Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Link: https://patch.msgid.link/20260511-dontcache-v7-3-2848ddce8090@kernel.org
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
mm: track DONTCACHE dirty pages per bdi_writeback 2026-06-04T08:16:50+00:00 Jeff Layton jlayton@kernel.org 2026-05-11T11:58:28+00:00 88d6f128d06d492b6d178c8e8c53db8c82305ae1 Add a per-wb WB_DONTCACHE_DIRTY counter that tracks the number of dirty pages with the dropbehind flag set (i.e., pages dirtied via RWF_DONTCACHE writes). Increment the counter alongside WB_RECLAIMABLE in folio_account_dirtied() when the folio has the dropbehind flag set, and decrement it in folio_clear_dirty_for_io() and folio_account_cleaned(). Also decrement it when a non-DONTCACHE lookup atomically clears the dropbehind flag on a dirty folio in __filemap_get_folio_mpol(), using folio_test_clear_dropbehind() to prevent concurrent lookups from double-decrementing the counter, and guarding the decrement with mapping_can_writeback() to match the increment path. Transfer the counter alongside WB_RECLAIMABLE in inode_do_switch_wbs() so that the stat is properly migrated when an inode switches cgroup writeback domains. The counter will be used by the writeback flusher to determine how many pages to write back when expediting writeback for IOCB_DONTCACHE writes, without flushing the entire BDI's dirty pages. Suggested-by: Jan Kara <jack@suse.cz> Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Jeff Layton <jlayton@kernel.org> Link: https://patch.msgid.link/20260511-dontcache-v7-2-2848ddce8090@kernel.org Reviewed-by: Jan Kara <jack@suse.cz> Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com> Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org> 
Add a per-wb WB_DONTCACHE_DIRTY counter that tracks the number of dirty
pages with the dropbehind flag set (i.e., pages dirtied via RWF_DONTCACHE
writes).

Increment the counter alongside WB_RECLAIMABLE in folio_account_dirtied()
when the folio has the dropbehind flag set, and decrement it in
folio_clear_dirty_for_io() and folio_account_cleaned(). Also decrement it
when a non-DONTCACHE lookup atomically clears the dropbehind flag on a
dirty folio in __filemap_get_folio_mpol(), using folio_test_clear_dropbehind()
to prevent concurrent lookups from double-decrementing the counter, and
guarding the decrement with mapping_can_writeback() to match the increment
path.

Transfer the counter alongside WB_RECLAIMABLE in inode_do_switch_wbs() so
that the stat is properly migrated when an inode switches cgroup writeback
domains.

The counter will be used by the writeback flusher to determine how many
pages to write back when expediting writeback for IOCB_DONTCACHE writes,
without flushing the entire BDI's dirty pages.

Suggested-by: Jan Kara <jack@suse.cz>
Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Link: https://patch.msgid.link/20260511-dontcache-v7-2-2848ddce8090@kernel.org
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
writeback: fix kernel-doc function name mismatch for wb_put_many() 2026-03-17T14:18:54+00:00 Kit Dallege xaum.io@gmail.com 2026-03-15T17:09:31+00:00 2727d44f5d5bc3f8e55a6a0ccf24d8105a5a400e The kernel-doc comment says wb_put but the actual function is wb_put_many. Fix the name to match. Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Kit Dallege <xaum.io@gmail.com> Link: https://patch.msgid.link/20260315170931.65852-1-xaum.io@gmail.com Reviewed-by: Jan Kara <jack@suse.cz> Signed-off-by: Christian Brauner <brauner@kernel.org> 
The kernel-doc comment says wb_put but the actual function is
wb_put_many. Fix the name to match.

Assisted-by: Claude:claude-opus-4-6
Signed-off-by: Kit Dallege <xaum.io@gmail.com>
Link: https://patch.msgid.link/20260315170931.65852-1-xaum.io@gmail.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
mm/block/fs: remove laptop_mode 2026-01-21T03:24:47+00:00 Johannes Weiner hannes@cmpxchg.org 2025-12-15T17:57:53+00:00 64dd89ae01f2708a508e028c28b7906e4702a9a7 Laptop mode was introduced to save battery, by delaying and consolidating writes and thereby maximize the time rotating hard drives wouldn't have to spin. Luckily, rotating hard drives, with their high spin-up times and power draw, are a thing of the past for battery-powered devices. Reclaim has also since changed to not write single filesystem pages anymore, and regular filesystem writeback is lumpy by design. The juice doesn't appear worth the squeeze anymore. The footprint of the feature is small, but nevertheless it's a complicating factor in mm, block, filesystems. Developers don't think about it, and it likely hasn't been tested with new reclaim and writeback changes in years. Let's sunset it. Keep the sysctl with a deprecation warning around for a few more cycles, but remove all functionality behind it. [akpm@linux-foundation.org: fix Documentation/admin-guide/laptops/index.rst] Link: https://lkml.kernel.org/r/20251216185201.GH905277@cmpxchg.org Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Suggested-by: Christoph Hellwig <hch@infradead.org> Reviewed-by: Christoph Hellwig <hch@lst.de> Acked-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: Shakeel Butt <shakeel.butt@linux.dev> Acked-by: Michal Hocko <mhocko@suse.com> Cc: Deepanshu Kartikey <kartikey406@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> 
Laptop mode was introduced to save battery, by delaying and consolidating
writes and thereby maximize the time rotating hard drives wouldn't have to
spin.

Luckily, rotating hard drives, with their high spin-up times and power
draw, are a thing of the past for battery-powered devices.  Reclaim has
also since changed to not write single filesystem pages anymore, and
regular filesystem writeback is lumpy by design.

The juice doesn't appear worth the squeeze anymore.  The footprint of the
feature is small, but nevertheless it's a complicating factor in mm,
block, filesystems.  Developers don't think about it, and it likely hasn't
been tested with new reclaim and writeback changes in years.

Let's sunset it.  Keep the sysctl with a deprecation warning around for a
few more cycles, but remove all functionality behind it.

[akpm@linux-foundation.org: fix Documentation/admin-guide/laptops/index.rst]
Link: https://lkml.kernel.org/r/20251216185201.GH905277@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Suggested-by: Christoph Hellwig <hch@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Shakeel Butt <shakeel.butt@linux.dev>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Deepanshu Kartikey <kartikey406@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Merge tag 'for-6.19/block-20251201' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux 2025-12-04T03:26:18+00:00 Linus Torvalds torvalds@linux-foundation.org 2025-12-04T03:26:18+00:00 cc25df3e2e22a956d3a0d427369367b4a901d203 Pull block updates from Jens Axboe: - Fix head insertion for mq-deadline, a regression from when priority support was added - Series simplifying and improving the ublk user copy code - Various ublk related cleanups - Fixup REQ_NOWAIT handling in loop/zloop, clearing NOWAIT when the request is punted to a thread for handling - Merge and then later revert loop dio nowait support, as it ended up causing excessive stack usage for when the inline issue code needs to dip back into the full file system code - Improve auto integrity code, making it less deadlock prone - Speedup polled IO handling, but manually managing the hctx lookups - Fixes for blk-throttle for SSD devices - Small series with fixes for the S390 dasd driver - Add support for caching zones, avoiding unnecessary report zone queries - MD pull requests via Yu: - fix null-ptr-dereference regression for dm-raid0 - fix IO hang for raid5 when array is broken with IO inflight - remove legacy 1s delay to speed up system shutdown - change maintainer's email address - data can be lost if array is created with different lbs devices, fix this problem and record lbs of the array in metadata - fix rcu protection for md_thread - fix mddev kobject lifetime regression - enable atomic writes for md-linear - some cleanups - bcache updates via Coly - remove useless discard and cache device code - improve usage of per-cpu workqueues - Reorganize the IO scheduler switching code, fixing some lockdep reports as well - Improve the block layer P2P DMA support - Add support to the block tracing code for zoned devices - Segment calculation improves, and memory alignment flexibility improvements - Set of prep and cleanups patches for ublk batching support. The actual batching hasn't been added yet, but helps shrink down the workload of getting that patchset ready for 6.20 - Fix for how the ps3 block driver handles segments offsets - Improve how block plugging handles batch tag allocations - nbd fixes for use-after-free of the configuration on device clear/put - Set of improvements and fixes for zloop - Add Damien as maintainer of the block zoned device code handling - Various other fixes and cleanups * tag 'for-6.19/block-20251201' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux: (162 commits) block/rnbd: correct all kernel-doc complaints blk-mq: use queue_hctx in blk_mq_map_queue_type md: remove legacy 1s delay in md_notify_reboot md/raid5: fix IO hang when array is broken with IO inflight md: warn about updating super block failure md/raid0: fix NULL pointer dereference in create_strip_zones() for dm-raid sbitmap: fix all kernel-doc warnings ublk: add helper of __ublk_fetch() ublk: pass const pointer to ublk_queue_is_zoned() ublk: refactor auto buffer register in ublk_dispatch_req() ublk: add `union ublk_io_buf` with improved naming ublk: add parameter `struct io_uring_cmd *` to ublk_prep_auto_buf_reg() kfifo: add kfifo_alloc_node() helper for NUMA awareness blk-mq: fix potential uaf for 'queue_hw_ctx' blk-mq: use array manage hctx map instead of xarray ublk: prevent invalid access with DEBUG s390/dasd: Use scnprintf() instead of sprintf() s390/dasd: Move device name formatting into separate function s390/dasd: Remove unnecessary debugfs_create() return checks s390/dasd: Fix gendisk parent after copy pair swap ... 
Pull block updates from Jens Axboe:

 - Fix head insertion for mq-deadline, a regression from when priority
   support was added

 - Series simplifying and improving the ublk user copy code

 - Various ublk related cleanups

 - Fixup REQ_NOWAIT handling in loop/zloop, clearing NOWAIT when the
   request is punted to a thread for handling

 - Merge and then later revert loop dio nowait support, as it ended up
   causing excessive stack usage for when the inline issue code needs to
   dip back into the full file system code

 - Improve auto integrity code, making it less deadlock prone

 - Speedup polled IO handling, but manually managing the hctx lookups

 - Fixes for blk-throttle for SSD devices

 - Small series with fixes for the S390 dasd driver

 - Add support for caching zones, avoiding unnecessary report zone
   queries

 - MD pull requests via Yu:
      - fix null-ptr-dereference regression for dm-raid0
      - fix IO hang for raid5 when array is broken with IO inflight
      - remove legacy 1s delay to speed up system shutdown
      - change maintainer's email address
      - data can be lost if array is created with different lbs devices,
        fix this problem and record lbs of the array in metadata
      - fix rcu protection for md_thread
      - fix mddev kobject lifetime regression
      - enable atomic writes for md-linear
      - some cleanups

 - bcache updates via Coly
      - remove useless discard and cache device code
      - improve usage of per-cpu workqueues

 - Reorganize the IO scheduler switching code, fixing some lockdep
   reports as well

 - Improve the block layer P2P DMA support

 - Add support to the block tracing code for zoned devices

 - Segment calculation improves, and memory alignment flexibility
   improvements

 - Set of prep and cleanups patches for ublk batching support. The
   actual batching hasn't been added yet, but helps shrink down the
   workload of getting that patchset ready for 6.20

 - Fix for how the ps3 block driver handles segments offsets

 - Improve how block plugging handles batch tag allocations

 - nbd fixes for use-after-free of the configuration on device clear/put

 - Set of improvements and fixes for zloop

 - Add Damien as maintainer of the block zoned device code handling

 - Various other fixes and cleanups

* tag 'for-6.19/block-20251201' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux: (162 commits)
  block/rnbd: correct all kernel-doc complaints
  blk-mq: use queue_hctx in blk_mq_map_queue_type
  md: remove legacy 1s delay in md_notify_reboot
  md/raid5: fix IO hang when array is broken with IO inflight
  md: warn about updating super block failure
  md/raid0: fix NULL pointer dereference in create_strip_zones() for dm-raid
  sbitmap: fix all kernel-doc warnings
  ublk: add helper of __ublk_fetch()
  ublk: pass const pointer to ublk_queue_is_zoned()
  ublk: refactor auto buffer register in ublk_dispatch_req()
  ublk: add `union ublk_io_buf` with improved naming
  ublk: add parameter `struct io_uring_cmd *` to ublk_prep_auto_buf_reg()
  kfifo: add kfifo_alloc_node() helper for NUMA awareness
  blk-mq: fix potential uaf for 'queue_hw_ctx'
  blk-mq: use array manage hctx map instead of xarray
  ublk: prevent invalid access with DEBUG
  s390/dasd: Use scnprintf() instead of sprintf()
  s390/dasd: Move device name formatting into separate function
  s390/dasd: Remove unnecessary debugfs_create() return checks
  s390/dasd: Fix gendisk parent after copy pair swap
  ...
fs: Add the __data_racy annotation to backing_dev_info.ra_pages 2025-11-18T22:00:11+00:00 Bart Van Assche bvanassche@acm.org 2025-11-14T21:04:06+00:00 42adb2d4ef24d2834cbd3bb96a6660826ae763da Some but not all .ra_pages changes happen while block layer I/O is paused with blk_mq_freeze_queue(). Filesystems may read .ra_pages even while block layer I/O is paused, e.g. from inside their .fadvise callback. Annotating all .ra_pages reads with READ_ONCE() would be cumbersome. Hence, add the __data_racy annotatation to the .ra_pages member variable. Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Christian Brauner <brauner@kernel.org> Cc: Nilay Shroff <nilay@linux.ibm.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> Reviewed-by: Nilay Shroff <nilay@linux.ibm.com> Signed-off-by: Jens Axboe <axboe@kernel.dk> 
Some but not all .ra_pages changes happen while block layer I/O is paused
with blk_mq_freeze_queue(). Filesystems may read .ra_pages even while
block layer I/O is paused, e.g. from inside their .fadvise callback.
Annotating all .ra_pages reads with READ_ONCE() would be cumbersome.
Hence, add the __data_racy annotatation to the .ra_pages member
variable.

Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Nilay Shroff <nilay@linux.ibm.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Nilay Shroff <nilay@linux.ibm.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
writeback: Add logging for slow writeback (exceeds sysctl_hung_task_timeout_secs) 2025-10-20T18:22:39+00:00 Julian Sun sunjunchao@bytedance.com 2025-09-30T07:18:29+00:00 d6e6215907640801b1f407dc9e871b19ca5a3805 When a writeback work lasts for sysctl_hung_task_timeout_secs, we want to identify that there are tasks waiting for a long time-this helps us pinpoint potential issues. Additionally, recording the starting jiffies is useful when debugging a crashed vmcore. Signed-off-by: Julian Sun <sunjunchao@bytedance.com> Signed-off-by: Christian Brauner <brauner@kernel.org> 
When a writeback work lasts for sysctl_hung_task_timeout_secs, we want
to identify that there are tasks waiting for a long time-this helps us
pinpoint potential issues.

Additionally, recording the starting jiffies is useful when debugging a
crashed vmcore.

Signed-off-by: Julian Sun <sunjunchao@bytedance.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
writeback: Wake up waiting tasks when finishing the writeback of a chunk. 2025-10-20T18:22:39+00:00 Julian Sun sunjunchao@bytedance.com 2025-09-30T08:53:15+00:00 1888635532fbbd6be4a4368621085c3a197279f8 Writing back a large number of pages can take a lots of time. This issue is exacerbated when the underlying device is slow or subject to block layer rate limiting, which in turn triggers unexpected hung task warnings. We can trigger a wake-up once a chunk has been written back and the waiting time for writeback exceeds half of sysctl_hung_task_timeout_secs. This action allows the hung task detector to be aware of the writeback progress, thereby eliminating these unexpected hung task warnings. This patch has passed the xfstests 'check -g quick' test based on ext4, with no additional failures introduced. Signed-off-by: Julian Sun <sunjunchao@bytedance.com> Reviewed-by: Jan Kara <jack@suse.cz> Suggested-by: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Christian Brauner <brauner@kernel.org> 
Writing back a large number of pages can take a lots of time.
This issue is exacerbated when the underlying device is slow or
subject to block layer rate limiting, which in turn triggers
unexpected hung task warnings.

We can trigger a wake-up once a chunk has been written back and the
waiting time for writeback exceeds half of
sysctl_hung_task_timeout_secs.
This action allows the hung task detector to be aware of the writeback
progress, thereby eliminating these unexpected hung task warnings.

This patch has passed the xfstests 'check -g quick' test based on ext4,
with no additional failures introduced.

Signed-off-by: Julian Sun <sunjunchao@bytedance.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
writeback: Avoid contention on wb->list_lock when switching inodes 2025-09-19T11:11:00+00:00 Jan Kara jack@suse.cz 2025-04-09T15:12:59+00:00 e1b849cfa6b61f1c866a908c9e8dd9b5aaab820b There can be multiple inode switch works that are trying to switch inodes to / from the same wb. This can happen in particular if some cgroup exits which owns many (thousands) inodes and we need to switch them all. In this case several inode_switch_wbs_work_fn() instances will be just spinning on the same wb->list_lock while only one of them makes forward progress. This wastes CPU cycles and quickly leads to softlockup reports and unusable system. Instead of running several inode_switch_wbs_work_fn() instances in parallel switching to the same wb and contending on wb->list_lock, run just one work item per wb and manage a queue of isw items switching to this wb. Acked-by: Tejun Heo <tj@kernel.org> Signed-off-by: Jan Kara <jack@suse.cz> 
There can be multiple inode switch works that are trying to switch
inodes to / from the same wb. This can happen in particular if some
cgroup exits which owns many (thousands) inodes and we need to switch
them all. In this case several inode_switch_wbs_work_fn() instances will
be just spinning on the same wb->list_lock while only one of them makes
forward progress. This wastes CPU cycles and quickly leads to softlockup
reports and unusable system.

Instead of running several inode_switch_wbs_work_fn() instances in
parallel switching to the same wb and contending on wb->list_lock, run
just one work item per wb and manage a queue of isw items switching to
this wb.

Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jan Kara <jack@suse.cz>
blk-wbt: Fix detection of dirty-throttled tasks 2024-02-06T16:44:03+00:00 Jan Kara jack@suse.cz 2024-01-23T17:58:26+00:00 f814bdda774c183b0cc15ec8f3b6e7c6f4527ba5 The detection of dirty-throttled tasks in blk-wbt has been subtly broken since its beginning in 2016. Namely if we are doing cgroup writeback and the throttled task is not in the root cgroup, balance_dirty_pages() will set dirty_sleep for the non-root bdi_writeback structure. However blk-wbt checks dirty_sleep only in the root cgroup bdi_writeback structure. Thus detection of recently throttled tasks is not working in this case (we noticed this when we switched to cgroup v2 and suddently writeback was slow). Since blk-wbt has no easy way to get to proper bdi_writeback and furthermore its intention has always been to work on the whole device rather than on individual cgroups, just move the dirty_sleep timestamp from bdi_writeback to backing_dev_info. That fixes the checking for recently throttled task and saves memory for everybody as a bonus. CC: stable@vger.kernel.org Fixes: b57d74aff9ab ("writeback: track if we're sleeping on progress in balance_dirty_pages()") Signed-off-by: Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20240123175826.21452-1-jack@suse.cz [axboe: fixup indentation errors] Signed-off-by: Jens Axboe <axboe@kernel.dk> 
The detection of dirty-throttled tasks in blk-wbt has been subtly broken
since its beginning in 2016. Namely if we are doing cgroup writeback and
the throttled task is not in the root cgroup, balance_dirty_pages() will
set dirty_sleep for the non-root bdi_writeback structure. However
blk-wbt checks dirty_sleep only in the root cgroup bdi_writeback
structure. Thus detection of recently throttled tasks is not working in
this case (we noticed this when we switched to cgroup v2 and suddently
writeback was slow).

Since blk-wbt has no easy way to get to proper bdi_writeback and
furthermore its intention has always been to work on the whole device
rather than on individual cgroups, just move the dirty_sleep timestamp
from bdi_writeback to backing_dev_info. That fixes the checking for
recently throttled task and saves memory for everybody as a bonus.

CC: stable@vger.kernel.org
Fixes: b57d74aff9ab ("writeback: track if we're sleeping on progress in balance_dirty_pages()")
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20240123175826.21452-1-jack@suse.cz
[axboe: fixup indentation errors]
Signed-off-by: Jens Axboe <axboe@kernel.dk>