From: jazeltq@gmail.com
To: jdurgin@redhat.com, jcody@redhat.com, dillaman@redhat.com,
kwolf@redhat.com, mreitz@redhat.com, qemu-block@nongnu.org
Cc: qemu-devel@nongnu.org, ceph-devel@vger.kernel.org,
jazeltq@gmail.com, tianqing <tianqing@unitedstack.com>
Subject: [Qemu-devel] [RFC v6] RBD: Add support readv,writev for rbd
Date: Tue, 21 Feb 2017 11:43:36 +0800 [thread overview]
Message-ID: <20170221034336.10097-1-jazeltq@gmail.com> (raw)
From: tianqing <tianqing@unitedstack.com>
Rbd can do readv and writev directly, so wo do not need to transform
iov to buf or vice versa any more.
Signed-off-by: tianqing <tianqing@unitedstack.com>
---
block/rbd.c | 79 ++++++++++++++++++++++++++++++++++++++++++-------------------
1 file changed, 54 insertions(+), 25 deletions(-)
diff --git a/block/rbd.c b/block/rbd.c
index a57b3e3..5373680 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -62,6 +62,13 @@
#define RBD_MAX_SNAP_NAME_SIZE 128
#define RBD_MAX_SNAPS 100
+/* The LIBRBD_SUPPORTS_IOVEC is defined in librbd.h */
+#ifdef LIBRBD_SUPPORTS_IOVEC
+#define LIBRBD_USE_IOVEC 1
+#else
+#define LIBRBD_USE_IOVEC 0
+#endif
+
typedef enum {
RBD_AIO_READ,
RBD_AIO_WRITE,
@@ -310,6 +317,17 @@ static int qemu_rbd_set_conf(rados_t cluster, const char *conf,
return ret;
}
+static void qemu_rbd_memset(RADOSCB *rcb, int64_t offs)
+{
+ if(LIBRBD_USE_IOVEC){
+ RBDAIOCB *acb = rcb->acb;
+ iov_memset(acb->qiov->iov, acb->qiov->niov, offs, 0,
+ acb->qiov->size - offs);
+ } else {
+ memset(rcb->buf + offs, 0, rcb->size - offs);
+ }
+}
+
static int qemu_rbd_create(const char *filename, QemuOpts *opts, Error **errp)
{
Error *local_err = NULL;
@@ -426,11 +444,11 @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
}
} else {
if (r < 0) {
- memset(rcb->buf, 0, rcb->size);
+ qemu_rbd_memset(rcb, 0);
acb->ret = r;
acb->error = 1;
} else if (r < rcb->size) {
- memset(rcb->buf + r, 0, rcb->size - r);
+ qemu_rbd_memset(rcb, r);
if (!acb->error) {
acb->ret = rcb->size;
}
@@ -440,11 +458,14 @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
}
g_free(rcb);
-
- if (acb->cmd == RBD_AIO_READ) {
- qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
+
+ if(!LIBRBD_USE_IOVEC){
+ if (acb->cmd == RBD_AIO_READ) {
+ qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
+ }
+ qemu_vfree(acb->bounce);
}
- qemu_vfree(acb->bounce);
+
acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
qemu_aio_unref(acb);
@@ -655,7 +676,6 @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
RBDAIOCB *acb;
RADOSCB *rcb = NULL;
rbd_completion_t c;
- char *buf;
int r;
BDRVRBDState *s = bs->opaque;
@@ -664,27 +684,29 @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
acb->cmd = cmd;
acb->qiov = qiov;
assert(!qiov || qiov->size == size);
- if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
- acb->bounce = NULL;
- } else {
- acb->bounce = qemu_try_blockalign(bs, qiov->size);
- if (acb->bounce == NULL) {
- goto failed;
+
+ rcb = g_new(RADOSCB, 1);
+
+ if(!LIBRBD_USE_IOVEC){
+ if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
+ acb->bounce = NULL;
+ } else {
+ acb->bounce = qemu_try_blockalign(bs, qiov->size);
+ if (acb->bounce == NULL) {
+ goto failed;
+ }
}
+ if (cmd == RBD_AIO_WRITE) {
+ qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
+ }
+ rcb->buf = acb->bounce;
}
+
acb->ret = 0;
acb->error = 0;
acb->s = s;
- if (cmd == RBD_AIO_WRITE) {
- qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
- }
-
- buf = acb->bounce;
-
- rcb = g_new(RADOSCB, 1);
rcb->acb = acb;
- rcb->buf = buf;
rcb->s = acb->s;
rcb->size = size;
r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
@@ -694,10 +716,16 @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
switch (cmd) {
case RBD_AIO_WRITE:
- r = rbd_aio_write(s->image, off, size, buf, c);
+ if(!LIBRBD_USE_IOVEC)
+ r = rbd_aio_write(s->image, off, size, rcb->buf, c);
+ else
+ r = rbd_aio_writev(s->image, qiov->iov, qiov->niov, off, c);
break;
case RBD_AIO_READ:
- r = rbd_aio_read(s->image, off, size, buf, c);
+ if(!LIBRBD_USE_IOVEC)
+ r = rbd_aio_read(s->image, off, size, rcb->buf, c);
+ else
+ r = rbd_aio_readv(s->image, qiov->iov, qiov->niov, off, c);
break;
case RBD_AIO_DISCARD:
r = rbd_aio_discard_wrapper(s->image, off, size, c);
@@ -712,14 +740,15 @@ static BlockAIOCB *rbd_start_aio(BlockDriverState *bs,
if (r < 0) {
goto failed_completion;
}
-
return &acb->common;
failed_completion:
rbd_aio_release(c);
failed:
g_free(rcb);
- qemu_vfree(acb->bounce);
+ if(!LIBRBD_USE_IOVEC)
+ qemu_vfree(acb->bounce);
+
qemu_aio_unref(acb);
return NULL;
}
--
2.10.2
next reply other threads:[~2017-02-21 3:46 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-21 3:43 jazeltq [this message]
2017-02-21 3:50 ` [Qemu-devel] [RFC v6] RBD: Add support readv,writev for rbd no-reply
2017-02-21 4:01 ` no-reply
2017-02-21 4:50 ` Jeff Cody
-- strict thread matches above, loose matches on Subject: below --
2017-02-21 6:50 jazeltq
2017-02-24 3:52 ` Jeff Cody
2017-02-24 6:09 ` Jaze Lee
2017-02-24 11:42 ` Jeff Cody
2017-02-24 12:51 ` Jaze Lee
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=20170221034336.10097-1-jazeltq@gmail.com \
--to=jazeltq@gmail.com \
--cc=ceph-devel@vger.kernel.org \
--cc=dillaman@redhat.com \
--cc=jcody@redhat.com \
--cc=jdurgin@redhat.com \
--cc=kwolf@redhat.com \
--cc=mreitz@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=tianqing@unitedstack.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).