All of lore.kernel.org
 help / color / mirror / Atom feed
From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
To: qemu-block@nongnu.org
Cc: qemu-devel@nongnu.org, eblake@redhat.com, armbru@redhat.com,
	crosa@redhat.com, ehabkost@redhat.com, berrange@redhat.com,
	pbonzini@redhat.com, vsementsov@virtuozzo.com, jsnow@redhat.com,
	mreitz@redhat.com, kwolf@redhat.com, den@openvz.org
Subject: [PATCH v2 09/33] block-copy: always set BDRV_REQ_SERIALISING flag
Date: Thu, 20 May 2021 17:21:41 +0300	[thread overview]
Message-ID: <20210520142205.607501-10-vsementsov@virtuozzo.com> (raw)
In-Reply-To: <20210520142205.607501-1-vsementsov@virtuozzo.com>

It won't hurt in common case, so let's not bother with detecting image
fleecing.

Also, we want to simplify initialization interface of copy-before-write
filter as we are going to make it public.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
---
 block/copy-before-write.h  |  2 +-
 include/block/block-copy.h |  3 +--
 block/backup.c             | 20 +-------------------
 block/block-copy.c         | 29 ++++++++++++++++++++++++++---
 block/copy-before-write.c  |  4 ++--
 5 files changed, 31 insertions(+), 27 deletions(-)

diff --git a/block/copy-before-write.h b/block/copy-before-write.h
index e284dfb6a7..538aab8bdb 100644
--- a/block/copy-before-write.h
+++ b/block/copy-before-write.h
@@ -33,7 +33,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
                                   BlockDriverState *target,
                                   const char *filter_node_name,
                                   uint64_t cluster_size,
-                                  BdrvRequestFlags write_flags,
+                                  bool compress,
                                   BlockCopyState **bcs,
                                   Error **errp);
 void bdrv_cbw_drop(BlockDriverState *bs);
diff --git a/include/block/block-copy.h b/include/block/block-copy.h
index 338f2ea7fd..c013a20e1e 100644
--- a/include/block/block-copy.h
+++ b/include/block/block-copy.h
@@ -24,8 +24,7 @@ typedef struct BlockCopyCallState BlockCopyCallState;
 
 BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
                                      int64_t cluster_size, bool use_copy_range,
-                                     BdrvRequestFlags write_flags,
-                                     Error **errp);
+                                     bool compress, Error **errp);
 
 void block_copy_set_progress_meter(BlockCopyState *s, ProgressMeter *pm);
 
diff --git a/block/backup.c b/block/backup.c
index d41dd30e25..29db3a0ef6 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -407,7 +407,6 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
     int64_t len, target_len;
     BackupBlockJob *job = NULL;
     int64_t cluster_size;
-    BdrvRequestFlags write_flags;
     BlockDriverState *cbw = NULL;
     BlockCopyState *bcs = NULL;
 
@@ -504,25 +503,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
         goto error;
     }
 
-    /*
-     * If source is in backing chain of target assume that target is going to be
-     * used for "image fleecing", i.e. it should represent a kind of snapshot of
-     * source at backup-start point in time. And target is going to be read by
-     * somebody (for example, used as NBD export) during backup job.
-     *
-     * In this case, we need to add BDRV_REQ_SERIALISING write flag to avoid
-     * intersection of backup writes and third party reads from target,
-     * otherwise reading from target we may occasionally read already updated by
-     * guest data.
-     *
-     * For more information see commit f8d59dfb40bb and test
-     * tests/qemu-iotests/222
-     */
-    write_flags = (bdrv_chain_contains(target, bs) ? BDRV_REQ_SERIALISING : 0) |
-                  (compress ? BDRV_REQ_WRITE_COMPRESSED : 0),
-
     cbw = bdrv_cbw_append(bs, target, filter_node_name,
-                          cluster_size, write_flags, &bcs, errp);
+                          cluster_size, compress, &bcs, errp);
     if (!cbw) {
         goto error;
     }
diff --git a/block/block-copy.c b/block/block-copy.c
index 9b4af00614..daa1a2bf9f 100644
--- a/block/block-copy.c
+++ b/block/block-copy.c
@@ -245,7 +245,7 @@ static uint32_t block_copy_max_transfer(BdrvChild *source, BdrvChild *target)
 
 BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
                                      int64_t cluster_size, bool use_copy_range,
-                                     BdrvRequestFlags write_flags, Error **errp)
+                                     bool compress, Error **errp)
 {
     BlockCopyState *s;
     BdrvDirtyBitmap *copy_bitmap;
@@ -257,6 +257,28 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
     }
     bdrv_disable_dirty_bitmap(copy_bitmap);
 
+    /*
+     * Why we always set BDRV_REQ_SERIALISING write flag:
+     *
+     * Assume source is in backing chain of target assume that target is going
+     * to be used for "image fleecing", i.e. it should represent a kind of
+     * snapshot of source at backup-start point in time. And target is going to
+     * be read by somebody (for example, used as NBD export) during backup job.
+     *
+     * In this case, we need to add BDRV_REQ_SERIALISING write flag to avoid
+     * intersection of backup writes and third party reads from target,
+     * otherwise reading from target we may occasionally read already updated by
+     * guest data.
+     *
+     * For more information see commit f8d59dfb40bb and test
+     * tests/qemu-iotests/222
+     *
+     * Other cases? The only known reasonable case is "just copy to target, and
+     * target is not used for something else". In this case BDRV_REQ_SERIALISING
+     * change nothing, so let's not bother with detecting the "image fleecing"
+     * case and enabling BDRV_REQ_SERIALISING only for it.
+     */
+
     s = g_new(BlockCopyState, 1);
     *s = (BlockCopyState) {
         .source = source,
@@ -264,7 +286,8 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
         .copy_bitmap = copy_bitmap,
         .cluster_size = cluster_size,
         .len = bdrv_dirty_bitmap_size(copy_bitmap),
-        .write_flags = write_flags,
+        .write_flags = BDRV_REQ_SERIALISING |
+            (compress ? BDRV_REQ_WRITE_COMPRESSED : 0),
         .mem = shres_create(BLOCK_COPY_MAX_MEM),
     };
 
@@ -277,7 +300,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
          */
         s->use_copy_range = false;
         s->copy_size = cluster_size;
-    } else if (write_flags & BDRV_REQ_WRITE_COMPRESSED) {
+    } else if (compress) {
         /* Compression supports only cluster-size writes and no copy-range. */
         s->use_copy_range = false;
         s->copy_size = cluster_size;
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index bc795adb87..235251a640 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -170,7 +170,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
                                   BlockDriverState *target,
                                   const char *filter_node_name,
                                   uint64_t cluster_size,
-                                  BdrvRequestFlags write_flags,
+                                  bool compress,
                                   BlockCopyState **bcs,
                                   Error **errp)
 {
@@ -216,7 +216,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
 
     state->cluster_size = cluster_size;
     state->bcs = block_copy_state_new(top->backing, state->target,
-                                      cluster_size, false, write_flags, errp);
+                                      cluster_size, false, compress, errp);
     if (!state->bcs) {
         error_prepend(errp, "Cannot create block-copy-state: ");
         goto fail;
-- 
2.29.2



  parent reply	other threads:[~2021-05-20 14:28 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-20 14:21 [PATCH v2 00/33] block: publish backup-top filter Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 01/33] block: rename bdrv_replace_child to bdrv_replace_child_tran Vladimir Sementsov-Ogievskiy
2021-05-31 12:52   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 02/33] block: comment graph-modifying function not updating permissions Vladimir Sementsov-Ogievskiy
2021-05-31 12:56   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 03/33] block: introduce bdrv_replace_child_bs() Vladimir Sementsov-Ogievskiy
2021-05-31 14:18   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 04/33] block: introduce blk_replace_bs Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 05/33] qdev-properties: PropertyInfo: add realized_set_allowed field Vladimir Sementsov-Ogievskiy
2021-05-31 15:51   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 06/33] qdev: allow setting drive property for realized device Vladimir Sementsov-Ogievskiy
2021-05-31 15:59   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 07/33] block: rename backup-top to copy-before-write Vladimir Sementsov-Ogievskiy
2021-05-31 16:08   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 08/33] block/backup: drop support for copy_range Vladimir Sementsov-Ogievskiy
2021-05-28 15:29   ` Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` Vladimir Sementsov-Ogievskiy [this message]
2021-05-20 14:21 ` [PATCH v2 10/33] block/backup: move cluster size calculation to block-copy Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 11/33] block/copy-before-write: relax permission requirements when no parents Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 12/33] block/copy-before-write: drop extra bdrv_unref on failure path Vladimir Sementsov-Ogievskiy
2021-05-31 16:19   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 13/33] block/copy-before-write: use file child instead of backing Vladimir Sementsov-Ogievskiy
2021-05-31 16:22   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 14/33] block/copy-before-write: bdrv_cbw_append(): replace child at last Vladimir Sementsov-Ogievskiy
2021-05-31 16:55   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 15/33] block/copy-before-write: introduce cbw_init() Vladimir Sementsov-Ogievskiy
2021-05-31 16:57   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 16/33] block/copy-before-write: cbw_init(): rename variables Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 17/33] block/copy-before-write: cbw_init(): use file child after attaching Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 18/33] block/copy-before-write: cbw_init(): use options Vladimir Sementsov-Ogievskiy
2021-05-31 17:03   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 19/33] block/copy-before-write: initialize block-copy bitmap Vladimir Sementsov-Ogievskiy
2021-05-31 17:10   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 20/33] block/block-copy: make setting progress optional Vladimir Sementsov-Ogievskiy
2021-08-10 15:22   ` Hanna Reitz
2021-05-20 14:21 ` [PATCH v2 21/33] block/copy-before-write: make public block driver Vladimir Sementsov-Ogievskiy
2021-05-20 14:21 ` [PATCH v2 22/33] qapi: publish copy-before-write filter Vladimir Sementsov-Ogievskiy
2021-06-01  9:08   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 23/33] python/qemu/machine.py: refactor _qemu_args() Vladimir Sementsov-Ogievskiy
2021-06-01 10:05   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 24/33] python/qemu/machine: QEMUMachine: improve qmp() method Vladimir Sementsov-Ogievskiy
2021-06-01 10:19   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 25/33] iotests.py: VM: add own __enter__ method Vladimir Sementsov-Ogievskiy
2021-06-01 10:58   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 26/33] iotests/222: fix pylint and mypy complains Vladimir Sementsov-Ogievskiy
2021-06-01 11:16   ` Max Reitz
2021-05-20 14:21 ` [PATCH v2 27/33] iotests/222: constantly use single quotes for strings Vladimir Sementsov-Ogievskiy
2021-06-01 11:17   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 28/33] iotests: move 222 to tests/image-fleecing Vladimir Sementsov-Ogievskiy
2021-06-01 11:17   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 29/33] iotests.py: hmp_qemu_io: support qdev Vladimir Sementsov-Ogievskiy
2021-06-01 11:19   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 30/33] iotests/image-fleecing: proper source device Vladimir Sementsov-Ogievskiy
2021-06-01 11:29   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 31/33] iotests/image-fleecing: rename tgt_node Vladimir Sementsov-Ogievskiy
2021-06-01 11:46   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 32/33] iotests/image-fleecing: prepare for adding new test-case Vladimir Sementsov-Ogievskiy
2021-06-01 11:47   ` Max Reitz
2021-05-20 14:22 ` [PATCH v2 33/33] iotests/image-fleecing: add test-case for copy-before-write filter Vladimir Sementsov-Ogievskiy
2021-06-01 12:02   ` Max Reitz
2021-05-31 17:11 ` [PATCH v2 00/33] block: publish backup-top filter Max Reitz
2021-05-31 17:35   ` Vladimir Sementsov-Ogievskiy

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=20210520142205.607501-10-vsementsov@virtuozzo.com \
    --to=vsementsov@virtuozzo.com \
    --cc=armbru@redhat.com \
    --cc=berrange@redhat.com \
    --cc=crosa@redhat.com \
    --cc=den@openvz.org \
    --cc=eblake@redhat.com \
    --cc=ehabkost@redhat.com \
    --cc=jsnow@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=pbonzini@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 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.