block: queue data commands from the flush state machine at the head
PATCH 1/1 block: queue data commands from the flush state machine at the head
We used to insert the data commands following a pre-flush to the head of the queue until commit 1e82fadfc6b ("blk-mq: do not do head insertions post-pre-flush commands"). Not doing this seems to cause hangs of such commands on NFS workloads when exported from file systems with SATA SSDs. I have no idea why this would starve these workloads, but doing a semantic revert of this patch (which looks quite different due to various other changes) fixes the hangs.
--- block/blk-flush.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
block/blk-flush.c +1 -1
@@ -189,7 +189,7 @@ static void blk_flush_complete_seq(struct request *rq,case REQ_FSEQ_DATA :list_move_tail (&rq->flush.list, &fq->flush_data_in_flight);spin_lock (&q->requeue_lock); -list_add_tail (&rq->queuelist, &q->flush_list); +list_add (&rq->queuelist, &q->requeue_list);spin_unlock (&q->requeue_lock);blk_mq_kick_requeue_list (q);break ;
Signature · 1 line
2.39.2
Quoted 9 lines · Christoph Hellwig
We used to insert the data commands following a pre-flush to the head of the queue until commit
1e82fadfc6b("blk-mq: do not do head insertions post-pre-flush commands"). Not doing this seems to cause hangs of such commands on NFS workloads when exported from file systems with SATA SSDs. I have no idea why this would starve these workloads, but doing a semantic revert of this patch (which looks quite different due to various other changes) fixes the hangs.[...]
Applied, thanks!
[1/1] block: queue data commands from the flush state machine at the head
commit: 9f87fc4d72f52b26ac3e19df5e4584227fe6740c
Best regards,
Signature · 1 line
Jens Axboe