qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Peter Lieven <pl@dlhnet.de>
To: "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Cc: kwolf@redhat.com, Paolo Bonzini <pbonzini@redhat.com>,
	ronnie sahlberg <ronniesahlberg@gmail.com>
Subject: [Qemu-devel] [PATCH] iscsi: partly avoid iovec linearization in iscsi_aio_writev
Date: Mon, 19 Nov 2012 15:58:31 +0100	[thread overview]
Message-ID: <50AA4917.5000402@dlhnet.de> (raw)

libiscsi expects all write16 data in a linear buffer. If the
iovec only contains one buffer we can skip the linearization
step as well as the additional malloc/free and pass the
buffer directly.

Reported-by: Ronnie Sahlberg <ronniesahlberg@gmail.com>
Signed-off-by: Peter Lieven <pl@kamp.de>
---
  block/iscsi.c |   24 +++++++++++++++---------
  1 file changed, 15 insertions(+), 9 deletions(-)

diff --git a/block/iscsi.c b/block/iscsi.c
index d0b1a10..f12148e 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -199,8 +199,10 @@ iscsi_aio_write16_cb(struct iscsi_context *iscsi, int status,
  
      trace_iscsi_aio_write16_cb(iscsi, status, acb, acb->canceled);
  
-    g_free(acb->buf);
-
+    if (acb->buf != NULL) {
+        g_free(acb->buf);
+    }
+
      if (acb->canceled != 0) {
          return;
      }
@@ -244,11 +246,18 @@ iscsi_aio_writev(BlockDriverState *bs, int64_t sector_num,
      acb->bh         = NULL;
      acb->status     = -EINPROGRESS;
  
-    /* XXX we should pass the iovec to write16 to avoid the extra copy */
-    /* this will allow us to get rid of 'buf' completely */
      size = nb_sectors * BDRV_SECTOR_SIZE;
-    acb->buf = g_malloc(size);
-    qemu_iovec_to_buf(acb->qiov, 0, acb->buf, size);
+    data.size = size;
+
+    /* if the iovec only contains one buffer we can pass it directly */
+    if (acb->qiov->niov == 1) {
+        acb->buf = NULL;
+        data.data = acb->qiov->iov[0].iov_base;
+    } else {
+        acb->buf = g_malloc(size);
+        qemu_iovec_to_buf(acb->qiov, 0, acb->buf, size);
+        data.data = acb->buf;
+    }
  
      acb->task = malloc(sizeof(struct scsi_task));
      if (acb->task == NULL) {
@@ -269,9 +278,6 @@ iscsi_aio_writev(BlockDriverState *bs, int64_t sector_num,
      *(uint32_t *)&acb->task->cdb[10] = htonl(num_sectors);
      acb->task->expxferlen = size;
  
-    data.data = acb->buf;
-    data.size = size;
-
      if (iscsi_scsi_command_async(iscsi, iscsilun->lun, acb->task,
                                   iscsi_aio_write16_cb,
                                   &data,
-- 
1.7.9.5

             reply	other threads:[~2012-11-19 14:58 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-11-19 14:58 Peter Lieven [this message]
2012-11-19 15:18 ` [Qemu-devel] [PATCH] iscsi: partly avoid iovec linearization in iscsi_aio_writev Paolo Bonzini
2012-11-19 15:23   ` Peter Lieven
2012-11-19 15:56   ` ronnie sahlberg
2012-11-20 15:51 ` Stefan Hajnoczi
2012-11-20 16:16   ` Paolo Bonzini

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=50AA4917.5000402@dlhnet.de \
    --to=pl@dlhnet.de \
    --cc=kwolf@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=ronniesahlberg@gmail.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).