All of lore.kernel.org
 help / color / mirror / Atom feed
From: Kevin Wolf <kwolf@redhat.com>
To: qemu-block@nongnu.org
Cc: kwolf@redhat.com, vsementsov@virtuozzo.com, berto@igalia.com,
	qemu-devel@nongnu.org, mreitz@redhat.com
Subject: [PATCH v7 10/10] qcow2: Forward ZERO_WRITE flag for full preallocation
Date: Fri, 24 Apr 2020 16:27:01 +0200	[thread overview]
Message-ID: <20200424142701.67053-1-kwolf@redhat.com> (raw)
In-Reply-To: <20200424125448.63318-1-kwolf@redhat.com>

The BDRV_REQ_ZERO_WRITE is currently implemented in a way that first the
image is possibly preallocated and then the zero flag is added to all
clusters. This means that a copy-on-write operation may be needed when
writing to these clusters, despite having used preallocation, negating
one of the major benefits of preallocation.

Instead, try to forward the BDRV_REQ_ZERO_WRITE to the protocol driver,
and if the protocol driver can ensure that the new area reads as zeros,
we can skip setting the zero flag in the qcow2 layer.

Unfortunately, the same approach doesn't work for metadata
preallocation, so we'll still set the zero flag there.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
---
 block/qcow2.c              | 22 +++++++++++++++++++---
 tests/qemu-iotests/274.out |  4 ++--
 2 files changed, 21 insertions(+), 5 deletions(-)

diff --git a/block/qcow2.c b/block/qcow2.c
index 98065d7808..2ba0b17c39 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -4170,9 +4170,25 @@ static int coroutine_fn qcow2_co_truncate(BlockDriverState *bs, int64_t offset,
         /* Allocate the data area */
         new_file_size = allocation_start +
                         nb_new_data_clusters * s->cluster_size;
-        /* Image file grows, so @exact does not matter */
-        ret = bdrv_co_truncate(bs->file, new_file_size, false, prealloc, 0,
-                               errp);
+        /*
+         * Image file grows, so @exact does not matter.
+         *
+         * If we need to zero out the new area, try first whether the protocol
+         * driver can already take care of this.
+         */
+        if (flags & BDRV_REQ_ZERO_WRITE) {
+            ret = bdrv_co_truncate(bs->file, new_file_size, false, prealloc,
+                                   BDRV_REQ_ZERO_WRITE, NULL);
+            if (ret >= 0) {
+                flags &= ~BDRV_REQ_ZERO_WRITE;
+            }
+        } else {
+            ret = -1;
+        }
+        if (ret < 0) {
+            ret = bdrv_co_truncate(bs->file, new_file_size, false, prealloc, 0,
+                                   errp);
+        }
         if (ret < 0) {
             error_prepend(errp, "Failed to resize underlying file: ");
             qcow2_free_clusters(bs, allocation_start,
diff --git a/tests/qemu-iotests/274.out b/tests/qemu-iotests/274.out
index 1a796fd07c..9d6fdeb1f7 100644
--- a/tests/qemu-iotests/274.out
+++ b/tests/qemu-iotests/274.out
@@ -187,7 +187,7 @@ read 65536/65536 bytes at offset 9437184
 10 MiB (0xa00000) bytes     allocated at offset 5 MiB (0x500000)
 
 [{ "start": 0, "length": 5242880, "depth": 1, "zero": true, "data": false},
-{ "start": 5242880, "length": 10485760, "depth": 0, "zero": true, "data": false, "offset": 327680}]
+{ "start": 5242880, "length": 10485760, "depth": 0, "zero": false, "data": true, "offset": 327680}]
 
 === preallocation=full ===
 Formatting 'TEST_DIR/PID-base', fmt=qcow2 size=16777216 cluster_size=65536 lazy_refcounts=off refcount_bits=16
@@ -206,7 +206,7 @@ read 65536/65536 bytes at offset 11534336
 4 MiB (0x400000) bytes     allocated at offset 8 MiB (0x800000)
 
 [{ "start": 0, "length": 8388608, "depth": 1, "zero": true, "data": false},
-{ "start": 8388608, "length": 4194304, "depth": 0, "zero": true, "data": false, "offset": 327680}]
+{ "start": 8388608, "length": 4194304, "depth": 0, "zero": false, "data": true, "offset": 327680}]
 
 === preallocation=off ===
 Formatting 'TEST_DIR/PID-base', fmt=qcow2 size=393216 cluster_size=65536 lazy_refcounts=off refcount_bits=16
-- 
2.25.3



  parent reply	other threads:[~2020-04-24 14:28 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-24 12:54 [PATCH v7 00/10] block: Fix resize (extending) of short overlays Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 01/10] block: Add flags to BlockDriver.bdrv_co_truncate() Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 02/10] block: Add flags to bdrv(_co)_truncate() Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 03/10] block-backend: Add flags to blk_truncate() Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 04/10] qcow2: Support BDRV_REQ_ZERO_WRITE for truncate Kevin Wolf
2020-04-24 14:07   ` Max Reitz
2020-04-24 14:39   ` Eric Blake
2020-04-28 16:28   ` Eric Blake
2020-04-28 18:45     ` Kevin Wolf
2020-04-28 18:58       ` Eric Blake
2020-04-24 12:54 ` [PATCH v7 05/10] raw-format: " Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 06/10] file-posix: " Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 07/10] block: truncate: Don't make backing file data visible Kevin Wolf
2020-04-24 14:09   ` Max Reitz
2020-04-24 12:54 ` [PATCH v7 08/10] iotests: Filter testfiles out in filter_img_info() Kevin Wolf
2020-04-24 12:54 ` [PATCH v7 09/10] iotests: Test committing to short backing file Kevin Wolf
2020-04-24 14:14   ` Max Reitz
2020-04-24 14:49   ` Vladimir Sementsov-Ogievskiy
2020-04-24 14:16 ` [PATCH v7 00/10] block: Fix resize (extending) of short overlays Max Reitz
2020-04-24 14:27 ` Kevin Wolf [this message]
2020-04-24 14:28   ` [PATCH v7 10/10] qcow2: Forward ZERO_WRITE flag for full preallocation Max Reitz
2020-04-24 15:26   ` Vladimir Sementsov-Ogievskiy
2020-04-27 15:43 ` [PATCH v7 00/10] block: Fix resize (extending) of short overlays Kevin Wolf

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=20200424142701.67053-1-kwolf@redhat.com \
    --to=kwolf@redhat.com \
    --cc=berto@igalia.com \
    --cc=mreitz@redhat.com \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=vsementsov@virtuozzo.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.