From: Peter Lieven <pl@kamp.de>
To: qemu-block@nongnu.org
Cc: qemu-devel@nongnu.org, famz@redhat.com, kwolf@redhat.com,
stefanha@redhat.com, mreitz@redhat.com, pbonzini@redhat.com,
Peter Lieven <pl@kamp.de>
Subject: [Qemu-devel] [PATCH V3] block/io: optimize bdrv_co_pwritev for small requests
Date: Mon, 30 May 2016 13:31:13 +0200 [thread overview]
Message-ID: <1464607873-28206-1-git-send-email-pl@kamp.de> (raw)
in a read-modify-write cycle a small request might cause
head and tail to fall into the same aligned block. Currently
QEMU reads the same block twice in this case which is
not necessary.
Signed-off-by: Peter Lieven <pl@kamp.de>
---
v1->v2: following Paolos suggestions to simplify the if condition and
adjusting the comment
v2->v3: fix iotest 077 for requests that are within the same aligned block [Fam, Kevin]
block/io.c | 8 ++++++++
tests/qemu-iotests/077 | 12 +-----------
tests/qemu-iotests/077.out | 26 --------------------------
3 files changed, 9 insertions(+), 37 deletions(-)
diff --git a/block/io.c b/block/io.c
index 2d832aa..0e4bb1e 100644
--- a/block/io.c
+++ b/block/io.c
@@ -1427,6 +1427,14 @@ int coroutine_fn bdrv_co_pwritev(BlockDriverState *bs,
bytes += offset & (align - 1);
offset = offset & ~(align - 1);
+
+ /* We have read the tail already if the request is smaller
+ * than one aligned block.
+ */
+ if (bytes < align) {
+ qemu_iovec_add(&local_qiov, head_buf + bytes, align - bytes);
+ bytes = align;
+ }
}
if ((offset + bytes) & (align - 1)) {
diff --git a/tests/qemu-iotests/077 b/tests/qemu-iotests/077
index 4dc680b..d2d2a2d 100755
--- a/tests/qemu-iotests/077
+++ b/tests/qemu-iotests/077
@@ -60,7 +60,7 @@ EOF
# Sequential RMW requests on the same physical sector
off=0x1000
-for ev in "head" "after_head" "tail" "after_tail"; do
+for ev in "head" "after_head"; do
cat <<EOF
break pwritev_rmw_$ev A
aio_write -P 10 $((off + 0x200)) 0x200
@@ -211,16 +211,6 @@ function verify_io()
echo read -P 11 0x2400 0x200
echo read -P 0 0x2600 0xa00
- echo read -P 0 0x3000 0x200
- echo read -P 10 0x3200 0x200
- echo read -P 11 0x3400 0x200
- echo read -P 0 0x3600 0xa00
-
- echo read -P 0 0x4000 0x200
- echo read -P 10 0x4200 0x200
- echo read -P 11 0x4400 0x200
- echo read -P 0 0x4600 0xa00
-
# Chained dependencies
echo read -P 10 0x5000 0x200
echo read -P 11 0x5200 0x200
diff --git a/tests/qemu-iotests/077.out b/tests/qemu-iotests/077.out
index eab14ae..16f951f 100644
--- a/tests/qemu-iotests/077.out
+++ b/tests/qemu-iotests/077.out
@@ -19,16 +19,6 @@ wrote XXX/XXX bytes at offset XXX
XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote XXX/XXX bytes at offset XXX
XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-blkdebug: Resuming request 'A'
-wrote XXX/XXX bytes at offset XXX
-XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote XXX/XXX bytes at offset XXX
-XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-blkdebug: Resuming request 'A'
-wrote XXX/XXX bytes at offset XXX
-XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-wrote XXX/XXX bytes at offset XXX
-XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote XXX/XXX bytes at offset XXX
XXX bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
wrote XXX/XXX bytes at offset XXX
@@ -114,22 +104,6 @@ read 512/512 bytes at offset 9216
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
read 2560/2560 bytes at offset 9728
2.500 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 12288
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 12800
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 13312
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 2560/2560 bytes at offset 13824
-2.500 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 16384
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 16896
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 512/512 bytes at offset 17408
-512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
-read 2560/2560 bytes at offset 17920
-2.500 KiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
read 512/512 bytes at offset 20480
512 bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
read 512/512 bytes at offset 20992
--
1.9.1
next reply other threads:[~2016-05-30 11:31 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-30 11:31 Peter Lieven [this message]
2016-05-31 23:39 ` [Qemu-devel] [PATCH V3] block/io: optimize bdrv_co_pwritev for small requests Stefan Hajnoczi
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=1464607873-28206-1-git-send-email-pl@kamp.de \
--to=pl@kamp.de \
--cc=famz@redhat.com \
--cc=kwolf@redhat.com \
--cc=mreitz@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@redhat.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 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).