From: Chao Leng <lengchao@huawei.com>
To: Christoph Hellwig <hch@lst.de>, Jens Axboe <axboe@kernel.dk>,
Keith Busch <kbusch@kernel.org>, Sagi Grimberg <sagi@grimberg.me>
Cc: Ming Lei <ming.lei@redhat.com>, <linux-nvme@lists.infradead.org>,
<linux-block@vger.kernel.org>, Hannes Reinecke <hare@suse.de>
Subject: Re: [PATCH 14/17] blk-mq: move the srcu_struct used for quiescing to the tagset
Date: Wed, 26 Oct 2022 16:48:55 +0800 [thread overview]
Message-ID: <12eb7ad8-6b70-092a-978c-a2c1ba595ad4@huawei.com> (raw)
In-Reply-To: <20221025144020.260458-15-hch@lst.de>
On 2022/10/25 22:40, Christoph Hellwig wrote:
> All I/O submissions have fairly similar latencies, and a tagset-wide
> quiesce is a fairly common operation.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> Reviewed-by: Keith Busch <kbusch@kernel.org>
> Reviewed-by: Ming Lei <ming.lei@redhat.com>
> Reviewed-by: Chao Leng <lengchao@huawei.com>
> Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
> Reviewed-by: Hannes Reinecke <hare@suse.de>
> ---
> block/blk-core.c | 27 +++++----------------------
> block/blk-mq.c | 33 +++++++++++++++++++++++++--------
> block/blk-mq.h | 14 +++++++-------
> block/blk-sysfs.c | 9 ++-------
> block/blk.h | 9 +--------
> block/genhd.c | 2 +-
> include/linux/blk-mq.h | 4 ++++
> include/linux/blkdev.h | 9 ---------
> 8 files changed, 45 insertions(+), 62 deletions(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 17667159482e0..3a2ed8dadf738 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -65,7 +65,6 @@ DEFINE_IDA(blk_queue_ida);
> * For queue allocation
> */
> struct kmem_cache *blk_requestq_cachep;
> -struct kmem_cache *blk_requestq_srcu_cachep;
>
> /*
> * Controlling structure to kblockd
> @@ -373,26 +372,20 @@ static void blk_timeout_work(struct work_struct *work)
> {
> }
>
> -struct request_queue *blk_alloc_queue(int node_id, bool alloc_srcu)
> +struct request_queue *blk_alloc_queue(int node_id)
> {
> struct request_queue *q;
>
> - q = kmem_cache_alloc_node(blk_get_queue_kmem_cache(alloc_srcu),
> - GFP_KERNEL | __GFP_ZERO, node_id);
> + q = kmem_cache_alloc_node(blk_requestq_cachep, GFP_KERNEL | __GFP_ZERO,
> + node_id);
> if (!q)
> return NULL;
>
> - if (alloc_srcu) {
> - blk_queue_flag_set(QUEUE_FLAG_HAS_SRCU, q);
> - if (init_srcu_struct(q->srcu) != 0)
> - goto fail_q;
> - }
> -
> q->last_merge = NULL;
>
> q->id = ida_alloc(&blk_queue_ida, GFP_KERNEL);
> if (q->id < 0)
> - goto fail_srcu;
> + goto fail_q;
>
> q->stats = blk_alloc_queue_stats();
> if (!q->stats)
> @@ -435,11 +428,8 @@ struct request_queue *blk_alloc_queue(int node_id, bool alloc_srcu)
> blk_free_queue_stats(q->stats);
> fail_id:
> ida_free(&blk_queue_ida, q->id);
> -fail_srcu:
> - if (alloc_srcu)
> - cleanup_srcu_struct(q->srcu);
> fail_q:
> - kmem_cache_free(blk_get_queue_kmem_cache(alloc_srcu), q);
> + kmem_cache_free(blk_requestq_cachep, q);
> return NULL;
> }
>
> @@ -1184,9 +1174,6 @@ int __init blk_dev_init(void)
> sizeof_field(struct request, cmd_flags));
> BUILD_BUG_ON(REQ_OP_BITS + REQ_FLAG_BITS > 8 *
> sizeof_field(struct bio, bi_opf));
> - BUILD_BUG_ON(ALIGN(offsetof(struct request_queue, srcu),
> - __alignof__(struct request_queue)) !=
> - sizeof(struct request_queue));
>
> /* used for unplugging and affects IO latency/throughput - HIGHPRI */
> kblockd_workqueue = alloc_workqueue("kblockd",
> @@ -1197,10 +1184,6 @@ int __init blk_dev_init(void)
> blk_requestq_cachep = kmem_cache_create("request_queue",
> sizeof(struct request_queue), 0, SLAB_PANIC, NULL);
>
> - blk_requestq_srcu_cachep = kmem_cache_create("request_queue_srcu",
> - sizeof(struct request_queue) +
> - sizeof(struct srcu_struct), 0, SLAB_PANIC, NULL);
> -
> blk_debugfs_root = debugfs_create_dir("block", NULL);
>
> return 0;
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 802fdd3d737e3..6cbf34921e33f 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -261,8 +261,8 @@ EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_nowait);
> */
> void blk_mq_wait_quiesce_done(struct request_queue *q)
> {
> - if (blk_queue_has_srcu(q))
> - synchronize_srcu(q->srcu);
> + if (q->tag_set->flags & BLK_MQ_F_BLOCKING)
> + synchronize_srcu(q->tag_set->srcu);
> else
> synchronize_rcu();
> }
> @@ -3971,7 +3971,7 @@ static struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set,
> struct request_queue *q;
> int ret;
>
> - q = blk_alloc_queue(set->numa_node, set->flags & BLK_MQ_F_BLOCKING);
> + q = blk_alloc_queue(set->numa_node);
> if (!q)
> return ERR_PTR(-ENOMEM);
> q->queuedata = queuedata;
> @@ -4138,9 +4138,6 @@ static void blk_mq_update_poll_flag(struct request_queue *q)
> int blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
> struct request_queue *q)
> {
> - WARN_ON_ONCE(blk_queue_has_srcu(q) !=
> - !!(set->flags & BLK_MQ_F_BLOCKING));
> -
> /* mark the queue as mq asap */
> q->mq_ops = set->ops;
>
> @@ -4398,9 +4395,19 @@ int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set)
> */
> if (set->nr_maps == 1 && set->nr_hw_queues > nr_cpu_ids)
> set->nr_hw_queues = nr_cpu_ids;
> +
> + if (set->flags & BLK_MQ_F_BLOCKING) {
> + set->srcu = kmalloc(sizeof(*set->srcu), GFP_KERNEL);
Memory with contiguous physical addresses is not necessary, maybe it is a better choice to use kvmalloc,
because sizeof(*set->srcu) is a little large.
kvmalloc() is more friendly to scenarios where memory is insufficient and running for a long time.
next prev parent reply other threads:[~2022-10-26 8:49 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-25 14:40 per-tagset SRCU struct and quiesce v2 Christoph Hellwig
2022-10-25 14:40 ` [PATCH 01/17] block: set the disk capacity to 0 in blk_mark_disk_dead Christoph Hellwig
2022-10-26 12:39 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 02/17] nvme-pci: refactor the tagset handling in nvme_reset_work Christoph Hellwig
2022-10-26 12:46 ` Sagi Grimberg
2022-10-30 9:17 ` Christoph Hellwig
2022-10-25 14:40 ` [PATCH 03/17] nvme-pci: don't warn about the lack of I/O queues for admin controllers Christoph Hellwig
2022-10-26 12:49 ` Sagi Grimberg
2022-10-30 9:18 ` Christoph Hellwig
2022-10-25 14:40 ` [PATCH 04/17] nvme: don't call nvme_kill_queues from nvme_remove_namespaces Christoph Hellwig
2022-10-25 17:43 ` Keith Busch
2022-10-25 20:17 ` Sagi Grimberg
2022-10-30 9:22 ` Christoph Hellwig
2022-10-25 14:40 ` [PATCH 05/17] nvme: don't remove namespaces in nvme_passthru_end Christoph Hellwig
2022-10-26 12:50 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 06/17] nvme: remove the NVME_NS_DEAD check in nvme_remove_invalid_namespaces Christoph Hellwig
2022-10-26 12:50 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 07/17] nvme: remove the NVME_NS_DEAD check in nvme_validate_ns Christoph Hellwig
2022-10-26 12:52 ` Sagi Grimberg
2022-10-30 9:28 ` Christoph Hellwig
2022-10-25 14:40 ` [PATCH 08/17] nvme: don't unquiesce the admin queue in nvme_kill_queues Christoph Hellwig
2022-10-26 12:53 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 09/17] nvme: don't unquiesce the I/O queues " Christoph Hellwig
2022-10-26 12:54 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 10/17] nvme-pci: mark the namespaces dead earlier in nvme_remove Christoph Hellwig
2022-10-25 18:53 ` Keith Busch
2022-10-26 12:55 ` Sagi Grimberg
2022-10-26 12:57 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 11/17] nvme-pci: don't unquiesce the I/O queues in nvme_remove_dead_ctrl Christoph Hellwig
2022-10-26 8:34 ` Chao Leng
2022-10-26 12:58 ` Sagi Grimberg
2022-10-27 2:46 ` Chao Leng
2022-10-25 14:40 ` [PATCH 12/17] nvme-pci: don't unquiesce the I/O queues in apple_nvme_reset_work Christoph Hellwig
2022-10-26 8:37 ` Chao Leng
2022-10-26 12:58 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 13/17] blk-mq: skip non-mq queues in blk_mq_quiesce_queue Christoph Hellwig
2022-10-25 14:40 ` [PATCH 14/17] blk-mq: move the srcu_struct used for quiescing to the tagset Christoph Hellwig
2022-10-26 8:48 ` Chao Leng [this message]
2022-10-26 13:01 ` Sagi Grimberg
2022-10-27 2:49 ` Chao Leng
2022-10-27 10:02 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 15/17] blk-mq: pass a tagset to blk_mq_wait_quiesce_done Christoph Hellwig
2022-10-25 14:40 ` [PATCH 16/17] blk-mq: add tagset quiesce interface Christoph Hellwig
2022-10-26 8:51 ` Chao Leng
2022-10-26 13:02 ` Sagi Grimberg
2022-10-25 14:40 ` [PATCH 17/17] nvme: use blk_mq_[un]quiesce_tagset Christoph Hellwig
2022-10-26 13:03 ` Sagi Grimberg
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=12eb7ad8-6b70-092a-978c-a2c1ba595ad4@huawei.com \
--to=lengchao@huawei.com \
--cc=axboe@kernel.dk \
--cc=hare@suse.de \
--cc=hch@lst.de \
--cc=kbusch@kernel.org \
--cc=linux-block@vger.kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=ming.lei@redhat.com \
--cc=sagi@grimberg.me \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).