summaryrefslogtreecommitdiff
path: root/tools/perf/scripts/python/bin/flamegraph-record
diff options
context:
space:
mode:
authorChao Shi <coshi036@gmail.com>2026-07-27 16:12:57 -0400
committerJens Axboe <axboe@kernel.dk>2026-07-29 05:23:20 -0600
commit26cb8ebbfaf713c82e142d08828d4d765057633b (patch)
tree8ec5bfe2d71b06293e64d08614749c179ef3101e /tools/perf/scripts/python/bin/flamegraph-record
parente65848e4ce352bac9e3465099354c8b8f845391f (diff)
block: stop the timeout timer when releasing a never added disk
disk_release() undoes blk_mq_init_allocated_queue() for a disk whose probe failed before add_disk(), but it only calls blk_mq_exit_queue(). Nothing there stops q->timeout, and that timer rolls forward: it stays pending until it next expires, not until the last request completes. So if the driver issued any I/O before adding the disk, the request_queue is freed while still linked into a timer wheel bucket. Commit 6f8191fdf41d ("block: simplify disk shutdown") dropped the blk_cleanup_queue() call that used to stop it. __del_gendisk() and blk_mq_destroy_queue() still do; only the probe failure path lost it. nvme gets there because nvme_update_ns_info() submits Report Zones or FDP io-mgmt-recv on ns->queue before the disk is added, so a later failure - a concurrent reset setting NVME_CTRL_FROZEN, or device_add_disk() failing - lands in put_disk() with the timer armed: BUG: KASAN: slab-use-after-free in detach_if_pending+0x30c/0x340 Write of size 8 at addr ffff888004d71310 by task kworker/u8:2/37 __timer_delete_sync+0x156/0x240 kernel/time/timer.c:1621 blk_sync_queue+0x22/0x40 block/blk-core.c:222 nvme_sync_queues+0x100/0x150 drivers/nvme/host/core.c:5362 nvme_reset_work+0x138/0x930 drivers/nvme/host/pci.c:3264 Allocated by task 34: __blk_mq_alloc_disk+0x33/0x100 block/blk-mq.c:4462 nvme_alloc_ns+0x290/0x3870 drivers/nvme/host/core.c:4146 Freed by task 0: blk_free_queue_rcu+0x3a/0x50 block/blk-core.c:254 rcu_core+0xc10/0x1730 kernel/rcu/tree.c:2857 The queue being synced there is ctrl->admin_q, only a victim sharing a timer wheel bucket with the freed queue's dangling entry; other runs tripped in enqueue_timer(), __run_timers() or blk_mq_timeout_work(). Failing nvme_alloc_ns() with a debug patch makes it deterministic: one leaked timer trips KASAN within seconds, while 1987 patched releases produced no splat. Stop the timer and the queue work items before blk_mq_exit_queue(), like blk_mq_destroy_queue() does. Found by FuzzNvme. Fixes: 6f8191fdf41d ("block: simplify disk shutdown") Acked-by: Weidong Zhu <weizhu@fiu.edu> Signed-off-by: Chao Shi <coshi036@gmail.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Link: https://patch.msgid.link/20260727201257.211635-1-coshi036@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
Diffstat (limited to 'tools/perf/scripts/python/bin/flamegraph-record')
0 files changed, 0 insertions, 0 deletions