qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: kwolf@redhat.com, famz@redhat.com, qemu-block@nongnu.org,
	mreitz@redhat.com
Subject: [Qemu-devel] [PATCH 07/11] qed: move tail of qed_aio_write_main to qed_aio_write_{cow, alloc}
Date: Wed, 31 May 2017 11:43:26 +0200	[thread overview]
Message-ID: <20170531094330.1808-8-pbonzini@redhat.com> (raw)
In-Reply-To: <20170531094330.1808-1-pbonzini@redhat.com>

This part is never called for in-place writes, move it away to avoid
the "backwards" coding style typical of callback-based code.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 block/qed.c | 68 ++++++++++++++++++++++++++++---------------------------------
 1 file changed, 31 insertions(+), 37 deletions(-)

diff --git a/block/qed.c b/block/qed.c
index ab505faa84..83a0973cfb 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -981,39 +981,11 @@ static int qed_aio_write_main(QEDAIOCB *acb)
     BDRVQEDState *s = acb_to_s(acb);
     uint64_t offset = acb->cur_cluster +
                       qed_offset_into_cluster(s, acb->cur_pos);
-    int ret;
 
     trace_qed_aio_write_main(s, acb, 0, offset, acb->cur_qiov.size);
 
     BLKDBG_EVENT(s->bs->file, BLKDBG_WRITE_AIO);
-    ret = bdrv_pwritev(s->bs->file, offset, &acb->cur_qiov);
-    if (ret < 0) {
-        return ret;
-    }
-
-    if (acb->find_cluster_ret != QED_CLUSTER_FOUND) {
-        if (s->bs->backing) {
-            /*
-             * Flush new data clusters before updating the L2 table
-             *
-             * This flush is necessary when a backing file is in use.  A crash
-             * during an allocating write could result in empty clusters in the
-             * image.  If the write only touched a subregion of the cluster,
-             * then backing image sectors have been lost in the untouched
-             * region.  The solution is to flush after writing a new data
-             * cluster and before updating the L2 table.
-             */
-            ret = bdrv_flush(s->bs->file->bs);
-            if (ret < 0) {
-                return ret;
-            }
-        }
-        ret = qed_aio_write_l2_update(acb, acb->cur_cluster);
-        if (ret < 0) {
-            return ret;
-        }
-    }
-    return 0;
+    return bdrv_pwritev(s->bs->file, offset, &acb->cur_qiov);
 }
 
 /**
@@ -1048,7 +1020,29 @@ static int qed_aio_write_cow(QEDAIOCB *acb)
         return ret;
     }
 
-    return qed_aio_write_main(acb);
+    ret = qed_aio_write_main(acb);
+    if (ret < 0) {
+        return ret;
+    }
+
+    if (s->bs->backing) {
+        /*
+         * Flush new data clusters before updating the L2 table
+         *
+         * This flush is necessary when a backing file is in use.  A crash
+         * during an allocating write could result in empty clusters in the
+         * image.  If the write only touched a subregion of the cluster,
+         * then backing image sectors have been lost in the untouched
+         * region.  The solution is to flush after writing a new data
+         * cluster and before updating the L2 table.
+         */
+        ret = bdrv_flush(s->bs->file->bs);
+        if (ret < 0) {
+            return ret;
+        }
+    }
+
+    return 0;
 }
 
 /**
@@ -1101,6 +1095,7 @@ static int qed_aio_write_alloc(QEDAIOCB *acb, size_t len)
         if (acb->find_cluster_ret == QED_CLUSTER_ZERO) {
             return 0;
         }
+        acb->cur_cluster = 1;
     } else {
         acb->cur_cluster = qed_alloc_clusters(s, acb->cur_nclusters);
     }
@@ -1113,15 +1108,14 @@ static int qed_aio_write_alloc(QEDAIOCB *acb, size_t len)
         }
     }
 
-    if (acb->flags & QED_AIOCB_ZERO) {
-        ret = qed_aio_write_l2_update(acb, 1);
-    } else {
+    if (!(acb->flags & QED_AIOCB_ZERO)) {
         ret = qed_aio_write_cow(acb);
+        if (ret < 0) {
+            return ret;
+        }
     }
-    if (ret < 0) {
-        return ret;
-    }
-    return 0;
+
+    return qed_aio_write_l2_update(acb, acb->cur_cluster);
 }
 
 /**
-- 
2.13.0

  parent reply	other threads:[~2017-05-31  9:43 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-05-31  9:43 [Qemu-devel] [RFC PATCH 00/11] Block layer thread-safety, part 2 Paolo Bonzini
2017-05-31  9:43 ` [Qemu-devel] [PATCH 01/11] qcow2: call CoQueue APIs under CoMutex Paolo Bonzini
2017-05-31 15:18   ` Eric Blake
2017-05-31  9:43 ` [Qemu-devel] [PATCH 02/11] coroutine-lock: add qemu_co_rwlock_downgrade and qemu_co_rwlock_upgrade Paolo Bonzini
2017-05-31 20:50   ` Eric Blake
2017-05-31  9:43 ` [Qemu-devel] [PATCH 03/11] vdi: make it thread-safe Paolo Bonzini
2017-05-31 20:55   ` Eric Blake
2017-05-31  9:43 ` [Qemu-devel] [PATCH 04/11] vpc: " Paolo Bonzini
2017-05-31 20:58   ` Eric Blake
2017-05-31  9:43 ` [Qemu-devel] [PATCH 05/11] vvfat: " Paolo Bonzini
2017-05-31 21:09   ` Eric Blake
2017-05-31  9:43 ` [Qemu-devel] [PATCH 06/11] nbd: " Paolo Bonzini
2017-05-31 21:20   ` Eric Blake
2017-05-31  9:43 ` Paolo Bonzini [this message]
2017-05-31  9:43 ` [Qemu-devel] [PATCH 08/11] block: invoke .bdrv_drain callback in coroutine context and from AioContext Paolo Bonzini
2017-06-08 12:46   ` Stefan Hajnoczi
2017-06-08 12:50     ` Paolo Bonzini
2017-05-31  9:43 ` [Qemu-devel] [PATCH 09/11] qed: protect table cache with CoMutex Paolo Bonzini
2017-05-31  9:43 ` [Qemu-devel] [PATCH 10/11] sheepdog: add queue_lock Paolo Bonzini
2017-05-31  9:43 ` [Qemu-devel] [PATCH 11/11] ssh: support I/O from any AioContext Paolo Bonzini
2017-06-08 12:52 ` [Qemu-devel] [RFC PATCH 00/11] Block layer thread-safety, part 2 Stefan Hajnoczi
2017-06-08 12:54   ` 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=20170531094330.1808-8-pbonzini@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=famz@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.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;
as well as URLs for NNTP newsgroup(s).