* [PATCH v5 RESEND] block: plug attempts to batch allocate tags multiple times [not found] <CGME20251113080643epcas5p154d9edfbf0f99e0d4c08408ad6cf68f4@epcas5p1.samsung.com> @ 2025-11-13 8:02 ` Xue He 2025-11-13 10:14 ` Ming Lei 0 siblings, 1 reply; 2+ messages in thread From: Xue He @ 2025-11-13 8:02 UTC (permalink / raw) To: axboe, yukuai; +Cc: akpm, linux-block, linux-kernel, hexue This patch aims to enable batch allocation of sufficient tags after batch IO submission with plug mechanism, thereby avoiding the need for frequent individual requests when the initial allocation is insufficient. ------------------------------------------------------------ Perf: base code: __blk_mq_alloc_requests() 1.31% patch: __blk_mq_alloc_requests() 0.7% ------------------------------------------------------------ --- changes since v1: - Modify multiple batch registrations into a single loop to achieve the batch quantity changes since v2: - Modify the call location of remainder handling - Refactoring sbitmap cleanup time changes since v3: - Add handle operation in loop - Add helper sbitmap_find_bits_in_word changes since v4: - Split blk-mq.c changes from sbitmap Signed-off-by: hexue <xue01.he@samsung.com> --- block/blk-mq.c | 39 ++++++++++++++++++++++----------------- 1 file changed, 22 insertions(+), 17 deletions(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 09f579414161..64cd0a3c7cbf 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -467,26 +467,31 @@ __blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data) unsigned long tag_mask; int i, nr = 0; - tag_mask = blk_mq_get_tags(data, data->nr_tags, &tag_offset); - if (unlikely(!tag_mask)) - return NULL; + do { + tag_mask = blk_mq_get_tags(data, data->nr_tags, &tag_offset); + if (unlikely(!tag_mask)) { + if (nr == 0) + return NULL; + break; + } + tags = blk_mq_tags_from_data(data); + for (i = 0; tag_mask; i++) { + if (!(tag_mask & (1UL << i))) + continue; + tag = tag_offset + i; + prefetch(tags->static_rqs[tag]); + tag_mask &= ~(1UL << i); + rq = blk_mq_rq_ctx_init(data, tags, tag); + rq_list_add_head(data->cached_rqs, rq); + data->nr_tags--; + nr++; + } + if (!(data->rq_flags & RQF_SCHED_TAGS)) + blk_mq_add_active_requests(data->hctx, nr); + } while (data->nr_tags); - tags = blk_mq_tags_from_data(data); - for (i = 0; tag_mask; i++) { - if (!(tag_mask & (1UL << i))) - continue; - tag = tag_offset + i; - prefetch(tags->static_rqs[tag]); - tag_mask &= ~(1UL << i); - rq = blk_mq_rq_ctx_init(data, tags, tag); - rq_list_add_head(data->cached_rqs, rq); - nr++; - } - if (!(data->rq_flags & RQF_SCHED_TAGS)) - blk_mq_add_active_requests(data->hctx, nr); /* caller already holds a reference, add for remainder */ percpu_ref_get_many(&data->q->q_usage_counter, nr - 1); - data->nr_tags -= nr; return rq_list_pop(data->cached_rqs); } -- 2.34.1 ^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v5 RESEND] block: plug attempts to batch allocate tags multiple times 2025-11-13 8:02 ` [PATCH v5 RESEND] block: plug attempts to batch allocate tags multiple times Xue He @ 2025-11-13 10:14 ` Ming Lei 0 siblings, 0 replies; 2+ messages in thread From: Ming Lei @ 2025-11-13 10:14 UTC (permalink / raw) To: Xue He; +Cc: axboe, yukuai, akpm, linux-block, linux-kernel On Thu, Nov 13, 2025 at 08:02:02AM +0000, Xue He wrote: > This patch aims to enable batch allocation of sufficient tags after > batch IO submission with plug mechanism, thereby avoiding the need for > frequent individual requests when the initial allocation is > insufficient. > > ------------------------------------------------------------ > Perf: > base code: __blk_mq_alloc_requests() 1.31% > patch: __blk_mq_alloc_requests() 0.7% > ------------------------------------------------------------ Can you include the workload with perf together? > > --- > changes since v1: > - Modify multiple batch registrations into a single loop to achieve > the batch quantity > > changes since v2: > - Modify the call location of remainder handling > - Refactoring sbitmap cleanup time > > changes since v3: > - Add handle operation in loop > - Add helper sbitmap_find_bits_in_word > > changes since v4: > - Split blk-mq.c changes from sbitmap > > Signed-off-by: hexue <xue01.he@samsung.com> > --- > block/blk-mq.c | 39 ++++++++++++++++++++++----------------- > 1 file changed, 22 insertions(+), 17 deletions(-) > > diff --git a/block/blk-mq.c b/block/blk-mq.c > index 09f579414161..64cd0a3c7cbf 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -467,26 +467,31 @@ __blk_mq_alloc_requests_batch(struct blk_mq_alloc_data *data) > unsigned long tag_mask; > int i, nr = 0; > > - tag_mask = blk_mq_get_tags(data, data->nr_tags, &tag_offset); > - if (unlikely(!tag_mask)) > - return NULL; > + do { > + tag_mask = blk_mq_get_tags(data, data->nr_tags, &tag_offset); > + if (unlikely(!tag_mask)) { > + if (nr == 0) > + return NULL; > + break; > + } > + tags = blk_mq_tags_from_data(data); > + for (i = 0; tag_mask; i++) { > + if (!(tag_mask & (1UL << i))) > + continue; > + tag = tag_offset + i; > + prefetch(tags->static_rqs[tag]); > + tag_mask &= ~(1UL << i); > + rq = blk_mq_rq_ctx_init(data, tags, tag); > + rq_list_add_head(data->cached_rqs, rq); > + data->nr_tags--; > + nr++; > + } > + if (!(data->rq_flags & RQF_SCHED_TAGS)) > + blk_mq_add_active_requests(data->hctx, nr); Here not only less-efficient, but also a over-counting bug, please move the above two lines after `percpu_ref_get_many`. Thanks, Ming ^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-11-13 10:15 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <CGME20251113080643epcas5p154d9edfbf0f99e0d4c08408ad6cf68f4@epcas5p1.samsung.com>
2025-11-13 8:02 ` [PATCH v5 RESEND] block: plug attempts to batch allocate tags multiple times Xue He
2025-11-13 10:14 ` Ming Lei
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox