From: Hannes Reinecke <hare@suse.de>
To: Christoph Hellwig <hch@lst.de>, Sagi Grimberg <sagi@grimberg.me>,
Keith Busch <kbusch@kernel.org>, Jens Axboe <axboe@kernel.dk>
Cc: Chao Leng <lengchao@huawei.com>,
linux-block@vger.kernel.org, linux-nvme@lists.infradead.org
Subject: Re: [PATCH 1/2] blk-mq: add a blk_mq_submit_bio_direct API
Date: Mon, 22 Mar 2021 12:23:19 +0100 [thread overview]
Message-ID: <df54545f-0f2c-9bc1-e448-d8595088b7f5@suse.de> (raw)
In-Reply-To: <20210322073726.788347-2-hch@lst.de>
On 3/22/21 8:37 AM, Christoph Hellwig wrote:
> This adds (back) and API for simple stacking drivers to submit a bio to
> blk-mq queue. The prime aim is to avoid blocking on the queue freeze
> percpu ref, as a multipath driver really does not want to get blocked
> on that when an underlying device is undergoing error recovery. It also
> happens to optimize away the small overhead of the curren->bio_list based
> recursion avoidance.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> block/blk-core.c | 2 +-
> block/blk-mq.c | 37 +++++++++++++++++++++++++++++++++++++
> block/blk.h | 1 +
> include/linux/blk-mq.h | 1 +
> 4 files changed, 40 insertions(+), 1 deletion(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index fc60ff20849738..4344f3c9058282 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -792,7 +792,7 @@ static inline blk_status_t blk_check_zone_append(struct request_queue *q,
> return BLK_STS_OK;
> }
>
> -static noinline_for_stack bool submit_bio_checks(struct bio *bio)
> +noinline_for_stack bool submit_bio_checks(struct bio *bio)
> {
> struct block_device *bdev = bio->bi_bdev;
> struct request_queue *q = bdev->bd_disk->queue;
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index d4d7c1caa43966..4ff85692843b49 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -2286,6 +2286,43 @@ blk_qc_t blk_mq_submit_bio(struct bio *bio)
> return BLK_QC_T_NONE;
> }
>
> +/**
> + * blk_mq_submit_bio_direct - hand a bio directly to the driver for I/O
> + * @bio: The bio describing the location in memory and on the device.
> + *
> + * This function behaves similar to submit_bio_noacct(), but does never waits
> + * for the queue to be unfreozen, instead it return false and lets the caller
> + * deal with the fallout. It also does not protect against recursion and thus
> + * must only be used if the called driver is known to be blk-mq based.
> + */
> +bool blk_mq_submit_bio_direct(struct bio *bio, blk_qc_t *qc)
> +{
> + struct gendisk *disk = bio->bi_bdev->bd_disk;
> + struct request_queue *q = disk->queue;
> +
> + if (WARN_ON_ONCE(!current->bio_list) ||
> + WARN_ON_ONCE(disk->fops->submit_bio)) {
> + bio_io_error(bio);
> + goto fail;
> + }
> + if (!submit_bio_checks(bio))
> + goto fail;
> +
> + if (unlikely(blk_queue_enter(q, BLK_MQ_REQ_NOWAIT)))
> + return false;
> + if (!blk_crypto_bio_prep(&bio))
> + goto fail_queue_exit;
> + *qc = blk_mq_submit_bio(bio);
> + return true;
> +
> +fail_queue_exit:
> + blk_queue_exit(disk->queue);
> +fail:
> + *qc = BLK_QC_T_NONE;
> + return true;
> +}
> +EXPORT_SYMBOL_GPL(blk_mq_submit_bio_direct);
> +
> void blk_mq_free_rqs(struct blk_mq_tag_set *set, struct blk_mq_tags *tags,
> unsigned int hctx_idx)
> {
> diff --git a/block/blk.h b/block/blk.h
> index 3b53e44b967e4e..c4c66b2a9ffb19 100644
> --- a/block/blk.h
> +++ b/block/blk.h
> @@ -221,6 +221,7 @@ ssize_t part_timeout_show(struct device *, struct device_attribute *, char *);
> ssize_t part_timeout_store(struct device *, struct device_attribute *,
> const char *, size_t);
>
> +bool submit_bio_checks(struct bio *bio);
> void __blk_queue_split(struct bio **bio, unsigned int *nr_segs);
> int ll_back_merge_fn(struct request *req, struct bio *bio,
> unsigned int nr_segs);
> diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
> index 2c473c9b899089..6804f397106ada 100644
> --- a/include/linux/blk-mq.h
> +++ b/include/linux/blk-mq.h
> @@ -615,6 +615,7 @@ static inline void blk_rq_bio_prep(struct request *rq, struct bio *bio,
> }
>
> blk_qc_t blk_mq_submit_bio(struct bio *bio);
> +bool blk_mq_submit_bio_direct(struct bio *bio, blk_qc_t *qc);
> void blk_mq_hctx_set_fq_lock_class(struct blk_mq_hw_ctx *hctx,
> struct lock_class_key *key);
>
>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
next prev parent reply other threads:[~2021-03-22 11:23 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-22 7:37 fix nvme-tcp and nvme-rdma controller reset hangs when using multipath Christoph Hellwig
2021-03-22 7:37 ` [PATCH 1/2] blk-mq: add a blk_mq_submit_bio_direct API Christoph Hellwig
2021-03-22 11:23 ` Hannes Reinecke [this message]
2021-03-22 15:30 ` Keith Busch
2021-03-22 7:37 ` [PATCH 2/2] nvme-multipath: don't block on blk_queue_enter of the underlying device Christoph Hellwig
2021-03-22 11:22 ` Hannes Reinecke
2021-03-22 15:31 ` Keith Busch
2021-03-23 2:57 ` Sagi Grimberg
2021-03-23 3:23 ` Sagi Grimberg
2021-03-23 7:04 ` Chao Leng
2021-03-23 7:36 ` Sagi Grimberg
2021-03-23 8:13 ` Chao Leng
2021-03-23 16:17 ` Christoph Hellwig
2021-03-23 16:15 ` Christoph Hellwig
2021-03-23 18:13 ` Sagi Grimberg
2021-03-23 18:22 ` Christoph Hellwig
2021-03-23 19:00 ` Sagi Grimberg
2021-03-23 19:01 ` Christoph Hellwig
2021-03-23 19:10 ` Sagi Grimberg
2021-03-23 7:28 ` Hannes Reinecke
2021-03-23 7:31 ` Sagi Grimberg
2021-03-23 8:36 ` Hannes Reinecke
2021-03-23 14:53 ` Keith Busch
2021-03-23 16:19 ` 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=df54545f-0f2c-9bc1-e448-d8595088b7f5@suse.de \
--to=hare@suse.de \
--cc=axboe@kernel.dk \
--cc=hch@lst.de \
--cc=kbusch@kernel.org \
--cc=lengchao@huawei.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-nvme@lists.infradead.org \
--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).