diff options
| author | Nilay Shroff <nilay@linux.ibm.com> | 2025-05-15 19:14:39 +0530 |
|---|---|---|
| committer | Jens Axboe <axboe@kernel.dk> | 2025-05-15 12:14:18 -0600 |
| commit | 532b9e11b8540eb543ebec9cba851c5691e10b5b (patch) | |
| tree | 9e798fc6aea5ea032c907f27246478c764b9ef0c | |
| parent | dbc5ba08ec5ff799a32f27ac79702a2cfc2200cd (diff) | |
| download | ipsec-532b9e11b8540eb543ebec9cba851c5691e10b5b.tar.gz | |
block: fix elv_update_nr_hw_queues() to reattach elevator
When nr_hw_queues is updated, the elevator needs to be switched to
ensure that we exit elevator and reattach it to ensure that hctx->
sched_tags is correctly allocated for the new hardware queues.
However, elv_update_nr_hw_queues() currently only switches the
elevator if the queue is not registered. This is incorrect, as it
prevents reattaching the elevator after updating nr_hw_queues, which
in turn inhibits allocation of sched_tags.
Fix this by allowing the elevator switch if the queue is registered,
ensuring proper reattachment and resource allocation.
Fixes: 596dce110b7d ("block: simplify elevator reattachment for updating nr_hw_queues")
Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
Link: https://lore.kernel.org/r/20250515134511.548270-1-nilay@linux.ibm.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
| -rw-r--r-- | block/elevator.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/block/elevator.c b/block/elevator.c index f8d72bd2061030..ab22542e6cf0d1 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -697,7 +697,7 @@ void elv_update_nr_hw_queues(struct request_queue *q) WARN_ON_ONCE(q->mq_freeze_depth == 0); mutex_lock(&q->elevator_lock); - if (q->elevator && !blk_queue_dying(q) && !blk_queue_registered(q)) { + if (q->elevator && !blk_queue_dying(q) && blk_queue_registered(q)) { ctx.name = q->elevator->type->elevator_name; /* force to reattach elevator after nr_hw_queue is updated */ |