kernel/linux.git/block/blk-throttle.c, branch v6.12.80

blk-throttle: fix access race during throttle policy activation

2026-03-25T10:08:45+00:00

[ Upstream commit bd9fd5be6bc0836820500f68fff144609fbd85a9 ] On repeated cold boots we occasionally hit a NULL pointer crash in blk_should_throtl() when throttling is consulted before the throttle policy is fully enabled for the queue. Checking only q->td != NULL is insufficient during early initialization, so blkg_to_pd() for the throttle policy can still return NULL and blkg_to_tg() becomes NULL, which later gets dereferenced. Unable to handle kernel NULL pointer dereference at virtual address 0000000000000156 ... pc : submit_bio_noacct+0x14c/0x4c8 lr : submit_bio_noacct+0x48/0x4c8 sp : ffff800087f0b690 x29: ffff800087f0b690 x28: 0000000000005f90 x27: ffff00068af393c0 x26: 0000000000080000 x25: 000000000002fbc0 x24: ffff000684ddcc70 x23: 0000000000000000 x22: 0000000000000000 x21: 0000000000000000 x20: 0000000000080000 x19: ffff000684ddcd08 x18: ffffffffffffffff x17: 0000000000000000 x16: ffff80008132a550 x15: 0000ffff98020fff x14: 0000000000000000 x13: 1fffe000d11d7021 x12: ffff000688eb810c x11: ffff00077ec4bb80 x10: ffff000688dcb720 x9 : ffff80008068ef60 x8 : 00000a6fb8a86e85 x7 : 000000000000111e x6 : 0000000000000002 x5 : 0000000000000246 x4 : 0000000000015cff x3 : 0000000000394500 x2 : ffff000682e35e40 x1 : 0000000000364940 x0 : 000000000000001a Call trace: submit_bio_noacct+0x14c/0x4c8 verity_map+0x178/0x2c8 __map_bio+0x228/0x250 dm_submit_bio+0x1c4/0x678 __submit_bio+0x170/0x230 submit_bio_noacct_nocheck+0x16c/0x388 submit_bio_noacct+0x16c/0x4c8 submit_bio+0xb4/0x210 f2fs_submit_read_bio+0x4c/0xf0 f2fs_mpage_readpages+0x3b0/0x5f0 f2fs_readahead+0x90/0xe8 Tighten blk_throtl_activated() to also require that the throttle policy bit is set on the queue: return q->td != NULL && test_bit(blkcg_policy_throtl.plid, q->blkcg_pols); This prevents blk_should_throtl() from accessing throttle group state until policy data has been attached to blkgs. Fixes: a3166c51702b ("blk-throttle: delay initialization until configuration") Co-developed-by: Liang Jie Signed-off-by: Liang Jie Signed-off-by: Han Guangjiang Reviewed-by: Yu Kuai Signed-off-by: Jens Axboe Signed-off-by: Robert Garcia Signed-off-by: Greg Kroah-Hartman

block/blk-throttle: Fix throttle slice time for SSDs

2025-12-18T12:55:00+00:00

[ Upstream commit f76581f9f1d29e32e120b0242974ba266e79de58 ] Commit d61fcfa4bb18 ("blk-throttle: choose a small throtl_slice for SSD") introduced device type specific throttle slices if BLK_DEV_THROTTLING_LOW was enabled. Commit bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") removed support for BLK_DEV_THROTTLING_LOW, but left the device type specific throttle slices in place. This effectively changed throttling behavior on systems with SSD which now use a different and non-configurable slice time compared to non-SSD devices. Practical impact is that throughput tests with low configured throttle values (65536 bps) experience less than expected throughput on SSDs, presumably due to rounding errors associated with the small throttle slice time used for those devices. The same tests pass when setting the throttle values to 65536 * 4 = 262144 bps. The original code sets the throttle slice time to DFL_THROTL_SLICE_HD if CONFIG_BLK_DEV_THROTTLING_LOW is disabled. Restore that code to fix the problem. With that, DFL_THROTL_SLICE_SSD is no longer necessary. Revert to the original code and re-introduce DFL_THROTL_SLICE to replace both DFL_THROTL_SLICE_HD and DFL_THROTL_SLICE_SSD. This effectively reverts commit d61fcfa4bb18 ("blk-throttle: choose a small throtl_slice for SSD"). While at it, also remove MAX_THROTL_SLICE since it is not used anymore. Fixes: bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") Cc: Yu Kuai Cc: Tejun Heo Signed-off-by: Guenter Roeck Signed-off-by: Khazhismel Kumykov Reviewed-by: Yu Kuai Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

blk-throttle: don't take carryover for prioritized processing of metadata

2025-05-29T09:02:29+00:00

[ Upstream commit a9fc8868b350cbf4ff730a4ea9651319cc669516 ] Commit 29390bb5661d ("blk-throttle: support prioritized processing of metadata") takes bytes/ios carryover for prioritized processing of metadata. Turns out we can support it by charging it directly without trimming slice, and the result is same with carryover. Cc: Tejun Heo Cc: Josef Bacik Cc: Yu Kuai Signed-off-by: Ming Lei Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20250305043123.3938491-3-ming.lei@redhat.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin

Merge tag 'for-6.12/block-20240913' of git://git.kernel.dk/linux

2024-09-16T11:33:06+00:00

Pull block updates from Jens Axboe: - MD changes via Song: - md-bitmap refactoring (Yu Kuai) - raid5 performance optimization (Artur Paszkiewicz) - Other small fixes (Yu Kuai, Chen Ni) - Add a sysfs entry 'new_level' (Xiao Ni) - Improve information reported in /proc/mdstat (Mateusz Kusiak) - NVMe changes via Keith: - Asynchronous namespace scanning (Stuart) - TCP TLS updates (Hannes) - RDMA queue controller validation (Niklas) - Align field names to the spec (Anuj) - Metadata support validation (Puranjay) - A syntax cleanup (Shen) - Fix a Kconfig linking error (Arnd) - New queue-depth quirk (Keith) - Add missing unplug trace event (Keith) - blk-iocost fixes (Colin, Konstantin) - t10-pi modular removal and fixes (Alexey) - Fix for potential BLKSECDISCARD overflow (Alexey) - bio splitting cleanups and fixes (Christoph) - Deal with folios rather than rather than pages, speeding up how the block layer handles bigger IOs (Kundan) - Use spinlocks rather than bit spinlocks in zram (Sebastian, Mike) - Reduce zoned device overhead in ublk (Ming) - Add and use sendpages_ok() for drbd and nvme-tcp (Ofir) - Fix regression in partition error pointer checking (Riyan) - Add support for write zeroes and rotational status in nbd (Wouter) - Add Yu Kuai as new BFQ maintainer. The scheduler has been unmaintained for quite a while. - Various sets of fixes for BFQ (Yu Kuai) - Misc fixes and cleanups (Alvaro, Christophe, Li, Md Haris, Mikhail, Yang) * tag 'for-6.12/block-20240913' of git://git.kernel.dk/linux: (120 commits) nvme-pci: qdepth 1 quirk block: fix potential invalid pointer dereference in blk_add_partition blk_iocost: make read-only static array vrate_adj_pct const block: unpin user pages belonging to a folio at once mm: release number of pages of a folio block: introduce folio awareness and add a bigger size from folio block: Added folio-ized version of bio_add_hw_page() block, bfq: factor out a helper to split bfqq in bfq_init_rq() block, bfq: remove local variable 'bfqq_already_existing' in bfq_init_rq() block, bfq: remove local variable 'split' in bfq_init_rq() block, bfq: remove bfq_log_bfqg() block, bfq: merge bfq_release_process_ref() into bfq_put_cooperator() block, bfq: fix procress reference leakage for bfqq in merge chain block, bfq: fix uaf for accessing waker_bfqq after splitting blk-throttle: support prioritized processing of metadata blk-throttle: remove last_low_overflow_time drbd: Add NULL check for net_conf to prevent dereference in state validation nvme-tcp: fix link failure for TCP auth blk-mq: add missing unplug trace event mtip32xx: Remove redundant null pointer checks in mtip_hw_debugfs_init() ...

blk-throttle: support prioritized processing of metadata

2024-09-10T22:31:41+00:00

Currently, blk-throttle handle all IO fifo, hence if data IO is throttled and then meta IO is dispatched, the meta IO will have to wait for the data IO, causing priority inversion problems. This patch support to handle metadata first and then pay debt while throttling data. Test script: use cgroup v1 to throttle root cgroup, then create new dir and file while write back is throttled test() { mkdir /mnt/test/xxx touch /mnt/test/xxx/1 sync /mnt/test/xxx sync /mnt/test/xxx } mkfs.ext4 -F /dev/nvme0n1 -E lazy_itable_init=0,lazy_journal_init=0 mount /dev/nvme0n1 /mnt/test echo "259:0 $((1024*1024))" > /sys/fs/cgroup/blkio/blkio.throttle.write_bps_device dd if=/dev/zero of=/mnt/test/foo1 bs=16M count=1 conv=fdatasync status=none & sleep 4 time test echo "259:0 0" > /sys/fs/cgroup/blkio/blkio.throttle.write_bps_device sleep 1 umount /dev/nvme0n1 Test result: time cost for creating new dir and file before this patch: 14s after this patch: 0.1s Signed-off-by: Yu Kuai Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20240903135149.271857-3-yukuai1@huaweicloud.com Signed-off-by: Jens Axboe

blk-throttle: remove last_low_overflow_time

2024-09-10T22:31:41+00:00

last_low_overflow_time is not used anymore after commit bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW"). Signed-off-by: Yu Kuai Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20240903135149.271857-2-yukuai1@huaweicloud.com Signed-off-by: Jens Axboe

blk-throttle: remove more latency dead-code

2024-07-27T16:33:25+00:00

The struct 'latency_bucket' and the #define 'request_bucket_index' are unused since commit bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") and the 'LATENCY_BUCKET_SIZE' #define was only used by the 'request_bucket_index' define. Remove them. Signed-off-by: Dr. David Alan Gilbert Link: https://lore.kernel.org/r/20240727155824.1000042-1-linux@treblig.org Signed-off-by: Jens Axboe

blk-throttle: fix lower control under super low iops limit

2024-06-28T20:55:02+00:00

User will configure allowed iops limit in 1s, and calculate_io_allowed() will calculate allowed iops in the slice by: limit * HZ / throtl_slice However, if limit is quite low, the result can be 0, then allowed IO in the slice is 0, this will cause missing dispatch and control will be lower than limit. For example, set iops_limit to 5 with HD disk, and test will found that iops will be 3. This is usually not a big deal, because user will unlikely to configure such low iops limit, however, this is still a problem in the extreme scene. Fix the problem by making sure the wait time calculated by tg_within_iops_limit() should allow at least one IO to be dispatched. Signed-off-by: Yu Kuai Acked-by: Tejun Heo Link: https://lore.kernel.org/r/20240618062108.3680835-1-yukuai1@huaweicloud.com Signed-off-by: Jens Axboe

blk-throttle: Fix incorrect display of io.max

2024-05-31T01:44:29+00:00

Commit bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") attempts to revert the code change introduced by commit cd5ab1b0fcb4 ("blk-throttle: add .low interface"). However, it leaves behind the bps_conf[] and iops_conf[] fields in the throtl_grp structure which aren't set anywhere in the new blk-throttle.c code but are still being used by tg_prfill_limit() to display the limits in io.max. Now io.max always displays the following values if a block queue is used: : rbps=0 wbps=0 riops=0 wiops=0 Fix this problem by removing bps_conf[] and iops_conf[] and use bps[] and iops[] instead to complete the revert. Fixes: bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") Reported-by: Justin Forbes Closes: https://github.com/containers/podman/issues/22701#issuecomment-2120627789 Signed-off-by: Waiman Long Acked-by: Tejun Heo Reviewed-by: Yu Kuai Link: https://lore.kernel.org/r/20240530134547.970075-1-longman@redhat.com Signed-off-by: Jens Axboe

blk-throttle: remove unused struct 'avg_latency_bucket'

2024-05-22T17:30:47+00:00

'avg_latency_bucket' is unused since commit bf20ab538c81 ("blk-throttle: remove CONFIG_BLK_DEV_THROTTLING_LOW") Remove it. Signed-off-by: Dr. David Alan Gilbert Link: https://lore.kernel.org/r/20240522172458.334173-1-linux@treblig.org Signed-off-by: Jens Axboe