]> git.ipfire.org Git - thirdparty/linux.git/commit
block: protect hctx attributes/params using q->elevator_lock
authorNilay Shroff <nilay@linux.ibm.com>
Thu, 6 Mar 2025 09:39:53 +0000 (15:09 +0530)
committerJens Axboe <axboe@kernel.dk>
Mon, 10 Mar 2025 13:31:06 +0000 (07:31 -0600)
commit5abba4cebec0a591ca7e7f55701e42cd5dc059af
tree4a316ee6217a9ac1ab559b040508d94ef077c75d
parent5e40f4452dc9a3fb44d13bb6bc7032f3911a2675
block: protect hctx attributes/params using q->elevator_lock

Currently, hctx attributes (nr_tags, nr_reserved_tags, and cpu_list)
are protected using `q->sysfs_lock`. However, these attributes can be
updated in multiple scenarios:
  - During the driver's probe method.
  - When updating nr_hw_queues.
  - When writing to the sysfs attribute nr_requests,
    which can modify nr_tags.
The nr_requests attribute is already protected using q->elevator_lock,
but none of the update paths actually use q->sysfs_lock to protect hctx
attributes. So to ensure proper synchronization, replace q->sysfs_lock
with q->elevator_lock when reading hctx attributes through sysfs.

Additionally, blk_mq_update_nr_hw_queues allocates and updates hctx.
The allocation of hctx is protected using q->elevator_lock, however,
updating hctx params happens without any protection, so safeguard hctx
param update path by also using q->elevator_lock.

Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
Link: https://lore.kernel.org/r/20250306093956.2818808-1-nilay@linux.ibm.com
[axboe: wrap comment at 80 chars]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
block/blk-mq-sysfs.c
block/blk-mq.c
include/linux/blkdev.h