From: Jens Axboe <axboe@kernel.dk>
To: "linux-block@vger.kernel.org" <linux-block@vger.kernel.org>
Subject: [PATCH] block: improve error checking in blkdev_bio_end_io_async()
Date: Tue, 19 Oct 2021 12:59:48 -0600 [thread overview]
Message-ID: <69df4731-3232-eb2a-664d-47d0db381843@kernel.dk> (raw)
Track the current error status of the dio with DIO_ERROR in the flags,
which can then avoid diving into dio->bio for the fast path of not
having any errors. This reduces the overhead of the function nicely,
which was previously dominated by this seemingly cheap check:
4.55% -1.13% [kernel.vmlinux] [k] blkdev_bio_end_io_async
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
diff --git a/block/fops.c b/block/fops.c
index d4f4fffb7d32..21d265caecff 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -127,6 +127,7 @@ enum {
DIO_MULTI_BIO = 1,
DIO_SHOULD_DIRTY = 2,
DIO_IS_SYNC = 4,
+ DIO_ERROR = 8,
};
struct blkdev_dio {
@@ -147,8 +148,10 @@ static void blkdev_bio_end_io(struct bio *bio)
struct blkdev_dio *dio = bio->bi_private;
unsigned int flags = dio->flags;
- if (bio->bi_status && !dio->bio.bi_status)
+ if (!(flags & DIO_ERROR) && !dio->bio.bi_status) {
dio->bio.bi_status = bio->bi_status;
+ flags |= DIO_ERROR;
+ }
if (!(flags & DIO_MULTI_BIO) || atomic_dec_and_test(&dio->ref)) {
if (!(flags & DIO_IS_SYNC)) {
@@ -157,7 +160,7 @@ static void blkdev_bio_end_io(struct bio *bio)
WRITE_ONCE(iocb->private, NULL);
- if (likely(!dio->bio.bi_status)) {
+ if (likely(!(flags & DIO_ERROR))) {
ret = dio->size;
iocb->ki_pos += ret;
} else {
--
Jens Axboe
next reply other threads:[~2021-10-19 18:59 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-19 18:59 Jens Axboe [this message]
2021-10-19 19:05 ` [PATCH] block: improve error checking in blkdev_bio_end_io_async() Jens Axboe
2021-10-19 19:06 ` Pavel Begunkov
2021-10-19 19:11 ` Jens Axboe
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=69df4731-3232-eb2a-664d-47d0db381843@kernel.dk \
--to=axboe@kernel.dk \
--cc=linux-block@vger.kernel.org \
/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