This is a prep for adding support of IO polling for dm device. ->poll_fn is introduced in commit ea435e1b9392 ("block: add a poll_fn callback to struct request_queue") for supporting non-mq queues such as nvme multipath, but removed in commit 529262d56dbe ("block: remove ->poll_fn"). To add support of IO polling for dm device, support for non-mq device should be added and thus we need ->poll_fn back. commit c62b37d96b6e ("block: move ->make_request_fn to struct block_device_operations") moved all callbacks into struct block_device_operations in gendisk. But ->poll_fn can't be moved there since there's no way to fetch the corresponding gendisk from request_queue. Signed-off-by: Jeffle Xu <jefflexu@xxxxxxxxxxxxxxxxx> --- block/blk-mq.c | 30 ++++++++++++++++++++++++------ include/linux/blkdev.h | 3 +++ 2 files changed, 27 insertions(+), 6 deletions(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 696450257ac1..b521ab01eaf3 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -43,6 +43,7 @@ static DEFINE_PER_CPU(struct list_head, blk_cpu_done); +static int blk_mq_poll(struct request_queue *q, blk_qc_t cookie); static void blk_mq_poll_stats_start(struct request_queue *q); static void blk_mq_poll_stats_fn(struct blk_stat_callback *cb); @@ -3212,6 +3213,9 @@ struct request_queue *blk_mq_init_allocated_queue(struct blk_mq_tag_set *set, q->tag_set = set; + if (q->mq_ops->poll) + q->poll_fn = blk_mq_poll; + q->queue_flags |= QUEUE_FLAG_MQ_DEFAULT; if (set->nr_maps > HCTX_TYPE_POLL && set->map[HCTX_TYPE_POLL].nr_queues) @@ -3856,7 +3860,8 @@ int blk_poll(struct request_queue *q, blk_qc_t cookie, bool spin) if (current->plug) blk_flush_plug_list(current->plug, false); - hctx = q->queue_hw_ctx[blk_qc_t_to_queue_num(cookie)]; + hctx = queue_is_mq(q) ? + q->queue_hw_ctx[blk_qc_t_to_queue_num(cookie)] : NULL; /* * If we sleep, have the caller restart the poll loop to reset @@ -3864,21 +3869,26 @@ int blk_poll(struct request_queue *q, blk_qc_t cookie, bool spin) * caller is responsible for checking if the IO completed. If * the IO isn't complete, we'll get called again and will go * straight to the busy poll loop. + * + * Currently dm doesn't support hybrid polling. */ - if (blk_mq_poll_hybrid(q, hctx, cookie)) + if (hctx && blk_mq_poll_hybrid(q, hctx, cookie)) return 1; - hctx->poll_considered++; + if (hctx) + hctx->poll_considered++; state = current->state; do { int ret; - hctx->poll_invoked++; + if (hctx) + hctx->poll_invoked++; - ret = q->mq_ops->poll(hctx); + ret = q->poll_fn(q, cookie); if (ret > 0) { - hctx->poll_success++; + if (hctx) + hctx->poll_success++; __set_current_state(TASK_RUNNING); return ret; } @@ -3898,6 +3908,14 @@ int blk_poll(struct request_queue *q, blk_qc_t cookie, bool spin) } EXPORT_SYMBOL_GPL(blk_poll); +static int blk_mq_poll(struct request_queue *q, blk_qc_t cookie) +{ + struct blk_mq_hw_ctx *hctx; + + hctx = q->queue_hw_ctx[blk_qc_t_to_queue_num(cookie)]; + return q->mq_ops->poll(hctx); +} + unsigned int blk_mq_rq_cpu(struct request *rq) { return rq->mq_ctx->cpu; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 639cae2c158b..d05684449893 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -288,6 +288,8 @@ static inline unsigned short req_get_ioprio(struct request *req) struct blk_queue_ctx; +typedef int (poll_q_fn) (struct request_queue *q, blk_qc_t); + struct bio_vec; enum blk_eh_timer_return { @@ -486,6 +488,7 @@ struct request_queue { struct blk_stat_callback *poll_cb; struct blk_rq_stat poll_stat[BLK_MQ_POLL_STATS_BKTS]; + poll_q_fn *poll_fn; struct timer_list timeout; struct work_struct timeout_work; -- 2.27.0