All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ming Lei <ming.lei@redhat.com>
To: Jens Axboe <axboe@kernel.dk>
Cc: linux-block@vger.kernel.org, Yu Kuai <yukuai3@huawei.com>,
	Ming Lei <ming.lei@redhat.com>
Subject: [PATCH 1/6] blk-mq: figure out correct numa node for hw queue
Date: Mon, 28 Feb 2022 17:04:25 +0800	[thread overview]
Message-ID: <20220228090430.1064267-2-ming.lei@redhat.com> (raw)
In-Reply-To: <20220228090430.1064267-1-ming.lei@redhat.com>

The current code always uses default queue map and hw queue index
for figuring out the numa node for hw queue, this way isn't correct
because blk-mq supports three queue maps, and the correct queue map
should be used for the specified hw queue.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
---
 block/blk-mq.c | 36 ++++++++++++++++++++++++++++++------
 1 file changed, 30 insertions(+), 6 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index a05ce7725031..931add81813b 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -3107,15 +3107,41 @@ void blk_mq_free_rq_map(struct blk_mq_tags *tags)
 	blk_mq_free_tags(tags);
 }
 
+static int hctx_idx_to_type(struct blk_mq_tag_set *set,
+		unsigned int hctx_idx)
+{
+	int j;
+
+	for (j = 0; j < set->nr_maps; j++) {
+		unsigned int start =  set->map[j].queue_offset;
+		unsigned int end = start + set->map[j].nr_queues;
+
+		if (hctx_idx >= start && hctx_idx < end)
+			break;
+	}
+
+	if (j >= set->nr_maps)
+		j = HCTX_TYPE_DEFAULT;
+
+	return j;
+}
+
+static int blk_mq_get_hctx_node(struct blk_mq_tag_set *set,
+		unsigned int hctx_idx)
+{
+	int type = hctx_idx_to_type(set, hctx_idx);
+
+	return blk_mq_hw_queue_to_node(&set->map[type], hctx_idx);
+}
+
 static struct blk_mq_tags *blk_mq_alloc_rq_map(struct blk_mq_tag_set *set,
 					       unsigned int hctx_idx,
 					       unsigned int nr_tags,
 					       unsigned int reserved_tags)
 {
 	struct blk_mq_tags *tags;
-	int node;
+	int node = blk_mq_get_hctx_node(set, hctx_idx);
 
-	node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], hctx_idx);
 	if (node == NUMA_NO_NODE)
 		node = set->numa_node;
 
@@ -3165,9 +3191,8 @@ static int blk_mq_alloc_rqs(struct blk_mq_tag_set *set,
 {
 	unsigned int i, j, entries_per_page, max_order = 4;
 	size_t rq_size, left;
-	int node;
+	int node = blk_mq_get_hctx_node(set, hctx_idx);
 
-	node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], hctx_idx);
 	if (node == NUMA_NO_NODE)
 		node = set->numa_node;
 
@@ -3941,10 +3966,9 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set,
 	/* protect against switching io scheduler  */
 	mutex_lock(&q->sysfs_lock);
 	for (i = 0; i < set->nr_hw_queues; i++) {
-		int node;
+		int node = blk_mq_get_hctx_node(set, i);
 		struct blk_mq_hw_ctx *hctx;
 
-		node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], i);
 		/*
 		 * If the hw queue has been mapped to another numa node,
 		 * we need to realloc the hctx. If allocation fails, fallback
-- 
2.31.1


  reply	other threads:[~2022-02-28  9:05 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-28  9:04 [PATCH 0/6] blk-mq: update_nr_hw_queues related improvement & bugfix Ming Lei
2022-02-28  9:04 ` Ming Lei [this message]
2022-03-01 13:30   ` [PATCH 1/6] blk-mq: figure out correct numa node for hw queue Christoph Hellwig
2022-03-01 19:19   ` John Garry
2022-03-02  1:47     ` Ming Lei
2022-03-02  9:02       ` John Garry
2022-03-02  9:22         ` Ming Lei
2022-03-02 10:11           ` John Garry
2022-02-28  9:04 ` [PATCH 2/6] blk-mq: simplify reallocation of hw ctxs a bit Ming Lei
2022-03-01 13:30   ` Christoph Hellwig
2022-02-28  9:04 ` [PATCH 3/6] blk-mq: re-config poll after queue map is changed Ming Lei
2022-03-01 13:31   ` Christoph Hellwig
2022-02-28  9:04 ` [PATCH 4/6] block: mtip32xx: don't touch q->queue_hw_ctx Ming Lei
2022-03-01 13:32   ` Christoph Hellwig
2022-02-28  9:04 ` [PATCH 5/6] blk-mq: add helper of blk_mq_get_hctx to retrieve hctx via its index Ming Lei
2022-03-01 13:33   ` Christoph Hellwig
2022-02-28  9:04 ` [PATCH 6/6] blk-mq: manage hctx map via xarray Ming Lei
2022-02-28 17:36   ` kernel test robot
2022-02-28 17:57   ` kernel test robot
2022-03-01  9:08     ` Ming Lei
2022-03-01  9:08       ` Ming Lei
2022-03-01 13:37   ` Christoph Hellwig
2022-03-02  2:06     ` Ming Lei
2022-03-02  8:24       ` Christoph Hellwig
2022-03-02  9:53         ` Ming Lei
2022-03-02  9:59           ` Christoph Hellwig

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=20220228090430.1064267-2-ming.lei@redhat.com \
    --to=ming.lei@redhat.com \
    --cc=axboe@kernel.dk \
    --cc=linux-block@vger.kernel.org \
    --cc=yukuai3@huawei.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.