rpms/qemu-kvm
7
0
Fork 0
Code Issues Pull Requests Releases Activity

* Mon May 26 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-4

Browse Source 
- kvm-block-Expand-block-status-mode-from-bool-to-flags.patch [RHEL-88435 RHEL-88437]
- kvm-file-posix-gluster-Handle-zero-block-status-hint-bet.patch [RHEL-88435 RHEL-88437]
- kvm-block-Let-bdrv_co_is_zero_fast-consolidate-adjacent-.patch [RHEL-88435 RHEL-88437]
- kvm-block-Add-new-bdrv_co_is_all_zeroes-function.patch [RHEL-88435 RHEL-88437]
- kvm-iotests-Improve-iotest-194-to-mirror-data.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Minor-refactoring.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Pass-full-sync-mode-rather-than-bool-to-inter.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Allow-QMP-override-to-declare-target-already-.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Drop-redundant-zero_target-parameter.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Skip-pre-zeroing-destination-if-it-is-already.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Skip-writing-zeroes-when-target-is-already-ze.patch [RHEL-88435 RHEL-88437]
- kvm-iotests-common.rc-add-disk_usage-function.patch [RHEL-88435 RHEL-88437]
- kvm-tests-Add-iotest-mirror-sparse-for-recent-patches.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Reduce-I-O-when-destination-is-detect-zeroes-.patch [RHEL-88435 RHEL-88437]
- Resolves: RHEL-88435
  (--migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1])
- Resolves: RHEL-88437
  (Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1])
This commit is contained in:
Miroslav Rezanina 2025-05-26 04:53:58 -04:00
parent 9c8774ec77
commit f696e6c6e2
15 changed files with 3062 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

143
kvm-block-Add-new-bdrv_co_is_all_zeroes-function.patch Normal file
View File

@ -0,0 +1,143 @@
From 659dd2d1f7b1facbf9c548468c1b50237f7aa8e4 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:21 -0500
Subject: [PATCH 04/14] block: Add new bdrv_co_is_all_zeroes() function
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [4/14] 404590dbec0b1113872a7eb1bfe5af0450fe6a28 (ebblake/centos-qemu-kvm)
There are some optimizations that require knowing if an image starts
out as reading all zeroes, such as making blockdev-mirror faster by
skipping the copying of source zeroes to the destination. The
existing bdrv_co_is_zero_fast() is a good building block for answering
this question, but it tends to give an answer of 0 for a file we just
created via QMP 'blockdev-create' or similar (such as 'qemu-img create
-f raw'). Why? Because file-posix.c insists on allocating a tiny
header to any file rather than leaving it 100% sparse, due to some
filesystems that are unable to answer alignment probes on a hole. But
teaching file-posix.c to read the tiny header doesn't scale - the
problem of a small header is also visible when libvirt sets up an NBD
client to a just-created file on a migration destination host.
So, we need a wrapper function that handles a bit more complexity in a
common manner for all block devices - when the BDS is mostly a hole,
but has a small non-hole header, it is still worth the time to read
that header and check if it reads as all zeroes before giving up and
returning a pessimistic answer.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-19-eblake@redhat.com>
(cherry picked from commit 52726096707c5c8b90597c445de897fa64d56e73)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/io.c | 62 ++++++++++++++++++++++++++++++++++++++++
include/block/block-io.h | 2 ++
2 files changed, 64 insertions(+)
diff --git a/block/io.c b/block/io.c
index 64f4b1d22a..b6fc07e1dc 100644
--- a/block/io.c
+++ b/block/io.c
@@ -38,10 +38,14 @@
#include "qemu/error-report.h"
#include "qemu/main-loop.h"
#include "system/replay.h"
+#include "qemu/units.h"
/* Maximum bounce buffer for copy-on-read and write zeroes, in bytes */
#define MAX_BOUNCE_BUFFER (32768 << BDRV_SECTOR_BITS)
+/* Maximum read size for checking if data reads as zero, in bytes */
+#define MAX_ZERO_CHECK_BUFFER (128 * KiB)
+
static void coroutine_fn GRAPH_RDLOCK
bdrv_parent_cb_resize(BlockDriverState *bs);
@@ -2778,6 +2782,64 @@ int coroutine_fn bdrv_co_is_zero_fast(BlockDriverState *bs, int64_t offset,
return 1;
}
+/*
+ * Check @bs (and its backing chain) to see if the entire image is known
+ * to read as zeroes.
+ * Return 1 if that is the case, 0 otherwise and -errno on error.
+ * This test is meant to be fast rather than accurate so returning 0
+ * does not guarantee non-zero data; however, a return of 1 is reliable,
+ * and this function can report 1 in more cases than bdrv_co_is_zero_fast.
+ */
+int coroutine_fn bdrv_co_is_all_zeroes(BlockDriverState *bs)
+{
+ int ret;
+ int64_t pnum, bytes;
+ char *buf;
+ QEMUIOVector local_qiov;
+ IO_CODE();
+
+ bytes = bdrv_co_getlength(bs);
+ if (bytes < 0) {
+ return bytes;
+ }
+
+ /* First probe - see if the entire image reads as zero */
+ ret = bdrv_co_common_block_status_above(bs, NULL, false, BDRV_WANT_ZERO,
+ 0, bytes, &pnum, NULL, NULL,
+ NULL);
+ if (ret < 0) {
+ return ret;
+ }
+ if (ret & BDRV_BLOCK_ZERO) {
+ return bdrv_co_is_zero_fast(bs, pnum, bytes - pnum);
+ }
+
+ /*
+ * Because of the way 'blockdev-create' works, raw files tend to
+ * be created with a non-sparse region at the front to make
+ * alignment probing easier. If the block starts with only a
+ * small allocated region, it is still worth the effort to see if
+ * the rest of the image is still sparse, coupled with manually
+ * reading the first region to see if it reads zero after all.
+ */
+ if (pnum > MAX_ZERO_CHECK_BUFFER) {
+ return 0;
+ }
+ ret = bdrv_co_is_zero_fast(bs, pnum, bytes - pnum);
+ if (ret <= 0) {
+ return ret;
+ }
+ /* Only the head of the image is unknown, and it's small. Read it. */
+ buf = qemu_blockalign(bs, pnum);
+ qemu_iovec_init_buf(&local_qiov, buf, pnum);
+ ret = bdrv_driver_preadv(bs, 0, pnum, &local_qiov, 0, 0);
+ if (ret >= 0) {
+ ret = buffer_is_zero(buf, pnum);
+ }
+ qemu_vfree(buf);
+ return ret;
+}
+
int coroutine_fn bdrv_co_is_allocated(BlockDriverState *bs, int64_t offset,
int64_t bytes, int64_t *pnum)
{
diff --git a/include/block/block-io.h b/include/block/block-io.h
index b49e0537dd..b99cc98d26 100644
--- a/include/block/block-io.h
+++ b/include/block/block-io.h
@@ -161,6 +161,8 @@ bdrv_is_allocated_above(BlockDriverState *bs, BlockDriverState *base,
int coroutine_fn GRAPH_RDLOCK
bdrv_co_is_zero_fast(BlockDriverState *bs, int64_t offset, int64_t bytes);
+int coroutine_fn GRAPH_RDLOCK
+bdrv_co_is_all_zeroes(BlockDriverState *bs);
int GRAPH_RDLOCK
bdrv_apply_auto_read_only(BlockDriverState *bs, const char *errmsg,
--
2.39.3

689
kvm-block-Expand-block-status-mode-from-bool-to-flags.patch Normal file
View File

@ -0,0 +1,689 @@
From cb945ccd11d37c959f590ae5661ffe5b73f372a7 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:18 -0500
Subject: [PATCH 01/14] block: Expand block status mode from bool to flags
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/14] 12507ca1dbe44640e4a8baded4d5afd5fb4ed615 (ebblake/centos-qemu-kvm)
This patch is purely mechanical, changing bool want_zero into an
unsigned int for bitwise-or of flags. As of this patch, all
implementations are unchanged (the old want_zero==true is now
mode==BDRV_WANT_PRECISE which is a superset of BDRV_WANT_ZERO); but
the callers in io.c that used to pass want_zero==false are now
prepared for future driver changes that can now distinguish bewteen
BDRV_WANT_ZERO vs. BDRV_WANT_ALLOCATED. The next patch will actually
change the file-posix driver along those lines, now that we have
more-specific hints.
As for the background why this patch is useful: right now, the
file-posix driver recognizes that if allocation is being queried, the
entire image can be reported as allocated (there is no backing file to
refer to) - but this throws away information on whether the entire
image reads as zero (trivially true if lseek(SEEK_HOLE) at offset 0
returns -ENXIO, a bit more complicated to prove if the raw file was
created with 'qemu-img create' since we intentionally allocate a small
chunk of all-zero data to help with alignment probing). Later patches
will add a generic algorithm for seeing if an entire file reads as
zeroes.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-16-eblake@redhat.com>
(cherry picked from commit c33159dec79069514f78faecfe268439226b0f5b)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/blkdebug.c | 6 ++--
block/copy-before-write.c | 4 +--
block/coroutines.h | 4 +--
block/file-posix.c | 4 +--
block/gluster.c | 4 +--
block/io.c | 51 ++++++++++++++++----------------
block/iscsi.c | 6 ++--
block/nbd.c | 4 +--
block/null.c | 6 ++--
block/parallels.c | 6 ++--
block/qcow.c | 2 +-
block/qcow2.c | 6 ++--
block/qed.c | 6 ++--
block/quorum.c | 4 +--
block/raw-format.c | 4 +--
block/rbd.c | 6 ++--
block/snapshot-access.c | 4 +--
block/vdi.c | 4 +--
block/vmdk.c | 2 +-
block/vpc.c | 2 +-
block/vvfat.c | 6 ++--
include/block/block-common.h | 11 +++++++
include/block/block_int-common.h | 27 +++++++++--------
include/block/block_int-io.h | 4 +--
tests/unit/test-block-iothread.c | 2 +-
25 files changed, 99 insertions(+), 86 deletions(-)
diff --git a/block/blkdebug.c b/block/blkdebug.c
index 1c1967f8e0..c54aee0c84 100644
--- a/block/blkdebug.c
+++ b/block/blkdebug.c
@@ -751,9 +751,9 @@ blkdebug_co_pdiscard(BlockDriverState *bs, int64_t offset, int64_t bytes)
}
static int coroutine_fn GRAPH_RDLOCK
-blkdebug_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum, int64_t *map,
- BlockDriverState **file)
+blkdebug_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes, int64_t *pnum,
+ int64_t *map, BlockDriverState **file)
{
int err;
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index fd470f5f92..2badb3a885 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -291,8 +291,8 @@ cbw_co_preadv_snapshot(BlockDriverState *bs, int64_t offset, int64_t bytes,
}
static int coroutine_fn GRAPH_RDLOCK
-cbw_co_snapshot_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset, int64_t bytes,
+cbw_co_snapshot_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes,
int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
diff --git a/block/coroutines.h b/block/coroutines.h
index 79e5efbf75..892646bb7a 100644
--- a/block/coroutines.h
+++ b/block/coroutines.h
@@ -47,7 +47,7 @@ int coroutine_fn GRAPH_RDLOCK
bdrv_co_common_block_status_above(BlockDriverState *bs,
BlockDriverState *base,
bool include_base,
- bool want_zero,
+ unsigned int mode,
int64_t offset,
int64_t bytes,
int64_t *pnum,
@@ -78,7 +78,7 @@ int co_wrapper_mixed_bdrv_rdlock
bdrv_common_block_status_above(BlockDriverState *bs,
BlockDriverState *base,
bool include_base,
- bool want_zero,
+ unsigned int mode,
int64_t offset,
int64_t bytes,
int64_t *pnum,
diff --git a/block/file-posix.c b/block/file-posix.c
index 0d85123d0f..0c6569742f 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -3266,7 +3266,7 @@ static int find_allocation(BlockDriverState *bs, off_t start,
* well exceed it.
*/
static int coroutine_fn raw_co_block_status(BlockDriverState *bs,
- bool want_zero,
+ unsigned int mode,
int64_t offset,
int64_t bytes, int64_t *pnum,
int64_t *map,
@@ -3282,7 +3282,7 @@ static int coroutine_fn raw_co_block_status(BlockDriverState *bs,
return ret;
}
- if (!want_zero) {
+ if (mode != BDRV_WANT_PRECISE) {
*pnum = bytes;
*map = offset;
*file = bs;
diff --git a/block/gluster.c b/block/gluster.c
index c6d25ae733..8197b0ecef 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -1465,7 +1465,7 @@ exit:
* (Based on raw_co_block_status() from file-posix.c.)
*/
static int coroutine_fn qemu_gluster_co_block_status(BlockDriverState *bs,
- bool want_zero,
+ unsigned int mode,
int64_t offset,
int64_t bytes,
int64_t *pnum,
@@ -1482,7 +1482,7 @@ static int coroutine_fn qemu_gluster_co_block_status(BlockDriverState *bs,
return ret;
}
- if (!want_zero) {
+ if (mode != BDRV_WANT_PRECISE) {
*pnum = bytes;
*map = offset;
*file = bs;
diff --git a/block/io.c b/block/io.c
index ccec11386b..e328402adc 100644
--- a/block/io.c
+++ b/block/io.c
@@ -2364,10 +2364,8 @@ int bdrv_flush_all(void)
* Drivers not implementing the functionality are assumed to not support
* backing files, hence all their sectors are reported as allocated.
*
- * If 'want_zero' is true, the caller is querying for mapping
- * purposes, with a focus on valid BDRV_BLOCK_OFFSET_VALID, _DATA, and
- * _ZERO where possible; otherwise, the result favors larger 'pnum',
- * with a focus on accurate BDRV_BLOCK_ALLOCATED.
+ * 'mode' serves as a hint as to which results are favored; see the
+ * BDRV_WANT_* macros for details.
*
* If 'offset' is beyond the end of the disk image the return value is
* BDRV_BLOCK_EOF and 'pnum' is set to 0.
@@ -2387,7 +2385,7 @@ int bdrv_flush_all(void)
* set to the host mapping and BDS corresponding to the guest offset.
*/
static int coroutine_fn GRAPH_RDLOCK
-bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
+bdrv_co_do_block_status(BlockDriverState *bs, unsigned int mode,
int64_t offset, int64_t bytes,
int64_t *pnum, int64_t *map, BlockDriverState **file)
{
@@ -2476,7 +2474,7 @@ bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
local_file = bs;
local_map = aligned_offset;
} else {
- ret = bs->drv->bdrv_co_block_status(bs, want_zero, aligned_offset,
+ ret = bs->drv->bdrv_co_block_status(bs, mode, aligned_offset,
aligned_bytes, pnum, &local_map,
&local_file);
@@ -2488,10 +2486,10 @@ bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
* the cache requires an RCU update, so double check here to avoid
* such an update if possible.
*
- * Check want_zero, because we only want to update the cache when we
+ * Check mode, because we only want to update the cache when we
* have accurate information about what is zero and what is data.
*/
- if (want_zero &&
+ if (mode == BDRV_WANT_PRECISE &&
ret == (BDRV_BLOCK_DATA | BDRV_BLOCK_OFFSET_VALID) &&
QLIST_EMPTY(&bs->children))
{
@@ -2548,7 +2546,7 @@ bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
if (ret & BDRV_BLOCK_RAW) {
assert(ret & BDRV_BLOCK_OFFSET_VALID && local_file);
- ret = bdrv_co_do_block_status(local_file, want_zero, local_map,
+ ret = bdrv_co_do_block_status(local_file, mode, local_map,
*pnum, pnum, &local_map, &local_file);
goto out;
}
@@ -2560,7 +2558,7 @@ bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
if (!cow_bs) {
ret |= BDRV_BLOCK_ZERO;
- } else if (want_zero) {
+ } else if (mode == BDRV_WANT_PRECISE) {
int64_t size2 = bdrv_co_getlength(cow_bs);
if (size2 >= 0 && offset >= size2) {
@@ -2569,14 +2567,14 @@ bdrv_co_do_block_status(BlockDriverState *bs, bool want_zero,
}
}
- if (want_zero && ret & BDRV_BLOCK_RECURSE &&
+ if (mode == BDRV_WANT_PRECISE && ret & BDRV_BLOCK_RECURSE &&
local_file && local_file != bs &&
(ret & BDRV_BLOCK_DATA) && !(ret & BDRV_BLOCK_ZERO) &&
(ret & BDRV_BLOCK_OFFSET_VALID)) {
int64_t file_pnum;
int ret2;
- ret2 = bdrv_co_do_block_status(local_file, want_zero, local_map,
+ ret2 = bdrv_co_do_block_status(local_file, mode, local_map,
*pnum, &file_pnum, NULL, NULL);
if (ret2 >= 0) {
/* Ignore errors. This is just providing extra information, it
@@ -2627,7 +2625,7 @@ int coroutine_fn
bdrv_co_common_block_status_above(BlockDriverState *bs,
BlockDriverState *base,
bool include_base,
- bool want_zero,
+ unsigned int mode,
int64_t offset,
int64_t bytes,
int64_t *pnum,
@@ -2654,7 +2652,7 @@ bdrv_co_common_block_status_above(BlockDriverState *bs,
return 0;
}
- ret = bdrv_co_do_block_status(bs, want_zero, offset, bytes, pnum,
+ ret = bdrv_co_do_block_status(bs, mode, offset, bytes, pnum,
map, file);
++*depth;
if (ret < 0 || *pnum == 0 || ret & BDRV_BLOCK_ALLOCATED || bs == base) {
@@ -2671,7 +2669,7 @@ bdrv_co_common_block_status_above(BlockDriverState *bs,
for (p = bdrv_filter_or_cow_bs(bs); include_base || p != base;
p = bdrv_filter_or_cow_bs(p))
{
- ret = bdrv_co_do_block_status(p, want_zero, offset, bytes, pnum,
+ ret = bdrv_co_do_block_status(p, mode, offset, bytes, pnum,
map, file);
++*depth;
if (ret < 0) {
@@ -2734,7 +2732,8 @@ int coroutine_fn bdrv_co_block_status_above(BlockDriverState *bs,
BlockDriverState **file)
{
IO_CODE();
- return bdrv_co_common_block_status_above(bs, base, false, true, offset,
+ return bdrv_co_common_block_status_above(bs, base, false,
+ BDRV_WANT_PRECISE, offset,
bytes, pnum, map, file, NULL);
}
@@ -2765,8 +2764,9 @@ int coroutine_fn bdrv_co_is_zero_fast(BlockDriverState *bs, int64_t offset,
return 1;
}
- ret = bdrv_co_common_block_status_above(bs, NULL, false, false, offset,
- bytes, &pnum, NULL, NULL, NULL);
+ ret = bdrv_co_common_block_status_above(bs, NULL, false, BDRV_WANT_ZERO,
+ offset, bytes, &pnum, NULL, NULL,
+ NULL);
if (ret < 0) {
return ret;
@@ -2782,9 +2782,9 @@ int coroutine_fn bdrv_co_is_allocated(BlockDriverState *bs, int64_t offset,
int64_t dummy;
IO_CODE();
- ret = bdrv_co_common_block_status_above(bs, bs, true, false, offset,
- bytes, pnum ? pnum : &dummy, NULL,
- NULL, NULL);
+ ret = bdrv_co_common_block_status_above(bs, bs, true, BDRV_WANT_ALLOCATED,
+ offset, bytes, pnum ? pnum : &dummy,
+ NULL, NULL, NULL);
if (ret < 0) {
return ret;
}
@@ -2817,7 +2817,8 @@ int coroutine_fn bdrv_co_is_allocated_above(BlockDriverState *bs,
int ret;
IO_CODE();
- ret = bdrv_co_common_block_status_above(bs, base, include_base, false,
+ ret = bdrv_co_common_block_status_above(bs, base, include_base,
+ BDRV_WANT_ALLOCATED,
offset, bytes, pnum, NULL, NULL,
&depth);
if (ret < 0) {
@@ -3714,8 +3715,8 @@ bdrv_co_preadv_snapshot(BdrvChild *child, int64_t offset, int64_t bytes,
}
int coroutine_fn
-bdrv_co_snapshot_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset, int64_t bytes,
+bdrv_co_snapshot_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes,
int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
@@ -3733,7 +3734,7 @@ bdrv_co_snapshot_block_status(BlockDriverState *bs,
}
bdrv_inc_in_flight(bs);
- ret = drv->bdrv_co_snapshot_block_status(bs, want_zero, offset, bytes,
+ ret = drv->bdrv_co_snapshot_block_status(bs, mode, offset, bytes,
pnum, map, file);
bdrv_dec_in_flight(bs);
diff --git a/block/iscsi.c b/block/iscsi.c
index 2f0f4dac09..15b96ee880 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -694,9 +694,9 @@ out_unlock:
static int coroutine_fn iscsi_co_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum,
- int64_t *map,
+ unsigned int mode,
+ int64_t offset, int64_t bytes,
+ int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
IscsiLun *iscsilun = bs->opaque;
diff --git a/block/nbd.c b/block/nbd.c
index 887841bc81..d5a2b21c6d 100644
--- a/block/nbd.c
+++ b/block/nbd.c
@@ -1397,8 +1397,8 @@ nbd_client_co_pdiscard(BlockDriverState *bs, int64_t offset, int64_t bytes)
}
static int coroutine_fn GRAPH_RDLOCK nbd_client_co_block_status(
- BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
- int64_t *pnum, int64_t *map, BlockDriverState **file)
+ BlockDriverState *bs, unsigned int mode, int64_t offset,
+ int64_t bytes, int64_t *pnum, int64_t *map, BlockDriverState **file)
{
int ret, request_ret;
NBDExtent64 extent = { 0 };
diff --git a/block/null.c b/block/null.c
index dc0b1fdbd9..4e448d593d 100644
--- a/block/null.c
+++ b/block/null.c
@@ -227,9 +227,9 @@ static int null_reopen_prepare(BDRVReopenState *reopen_state,
}
static int coroutine_fn null_co_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum,
- int64_t *map,
+ unsigned int mode,
+ int64_t offset, int64_t bytes,
+ int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
BDRVNullState *s = bs->opaque;
diff --git a/block/parallels.c b/block/parallels.c
index 347ca127f3..3a375e2a8a 100644
--- a/block/parallels.c
+++ b/block/parallels.c
@@ -416,9 +416,9 @@ parallels_co_flush_to_os(BlockDriverState *bs)
}
static int coroutine_fn GRAPH_RDLOCK
-parallels_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum, int64_t *map,
- BlockDriverState **file)
+parallels_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes, int64_t *pnum,
+ int64_t *map, BlockDriverState **file)
{
BDRVParallelsState *s = bs->opaque;
int count;
diff --git a/block/qcow.c b/block/qcow.c
index da8ad4d243..8a3e7591a9 100644
--- a/block/qcow.c
+++ b/block/qcow.c
@@ -530,7 +530,7 @@ get_cluster_offset(BlockDriverState *bs, uint64_t offset, int allocate,
}
static int coroutine_fn GRAPH_RDLOCK
-qcow_co_block_status(BlockDriverState *bs, bool want_zero,
+qcow_co_block_status(BlockDriverState *bs, unsigned int mode,
int64_t offset, int64_t bytes, int64_t *pnum,
int64_t *map, BlockDriverState **file)
{
diff --git a/block/qcow2.c b/block/qcow2.c
index b6ade4755d..9fc96ba99a 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -2147,9 +2147,9 @@ static void qcow2_join_options(QDict *options, QDict *old_options)
}
static int coroutine_fn GRAPH_RDLOCK
-qcow2_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset,
- int64_t count, int64_t *pnum, int64_t *map,
- BlockDriverState **file)
+qcow2_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t count, int64_t *pnum,
+ int64_t *map, BlockDriverState **file)
{
BDRVQcow2State *s = bs->opaque;
uint64_t host_offset;
diff --git a/block/qed.c b/block/qed.c
index ac24449ffb..4a36fb3929 100644
--- a/block/qed.c
+++ b/block/qed.c
@@ -833,9 +833,9 @@ fail:
}
static int coroutine_fn GRAPH_RDLOCK
-bdrv_qed_co_block_status(BlockDriverState *bs, bool want_zero, int64_t pos,
- int64_t bytes, int64_t *pnum, int64_t *map,
- BlockDriverState **file)
+bdrv_qed_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t pos, int64_t bytes, int64_t *pnum,
+ int64_t *map, BlockDriverState **file)
{
BDRVQEDState *s = bs->opaque;
size_t len = MIN(bytes, SIZE_MAX);
diff --git a/block/quorum.c b/block/quorum.c
index 30747a6df9..ed8ce801ee 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -1226,7 +1226,7 @@ static void quorum_child_perm(BlockDriverState *bs, BdrvChild *c,
* region contains zeroes, and BDRV_BLOCK_DATA otherwise.
*/
static int coroutine_fn GRAPH_RDLOCK
-quorum_co_block_status(BlockDriverState *bs, bool want_zero,
+quorum_co_block_status(BlockDriverState *bs, unsigned int mode,
int64_t offset, int64_t count,
int64_t *pnum, int64_t *map, BlockDriverState **file)
{
@@ -1238,7 +1238,7 @@ quorum_co_block_status(BlockDriverState *bs, bool want_zero,
for (i = 0; i < s->num_children; i++) {
int64_t bytes;
ret = bdrv_co_common_block_status_above(s->children[i]->bs, NULL, false,
- want_zero, offset, count,
+ mode, offset, count,
&bytes, NULL, NULL, NULL);
if (ret < 0) {
quorum_report_bad(QUORUM_OP_TYPE_READ, offset, count,
diff --git a/block/raw-format.c b/block/raw-format.c
index e08526e2ec..df16ac1ea2 100644
--- a/block/raw-format.c
+++ b/block/raw-format.c
@@ -283,8 +283,8 @@ fail:
}
static int coroutine_fn GRAPH_RDLOCK
-raw_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum, int64_t *map,
+raw_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes, int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
BDRVRawState *s = bs->opaque;
diff --git a/block/rbd.c b/block/rbd.c
index af984fb7db..4f3d42a8e7 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -1504,9 +1504,9 @@ static int qemu_rbd_diff_iterate_cb(uint64_t offs, size_t len,
}
static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum,
- int64_t *map,
+ unsigned int mode,
+ int64_t offset, int64_t bytes,
+ int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
BDRVRBDState *s = bs->opaque;
diff --git a/block/snapshot-access.c b/block/snapshot-access.c
index 71ac83c01f..17ed2402db 100644
--- a/block/snapshot-access.c
+++ b/block/snapshot-access.c
@@ -41,11 +41,11 @@ snapshot_access_co_preadv_part(BlockDriverState *bs,
static int coroutine_fn GRAPH_RDLOCK
snapshot_access_co_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset,
+ unsigned int mode, int64_t offset,
int64_t bytes, int64_t *pnum,
int64_t *map, BlockDriverState **file)
{
- return bdrv_co_snapshot_block_status(bs->file->bs, want_zero, offset,
+ return bdrv_co_snapshot_block_status(bs->file->bs, mode, offset,
bytes, pnum, map, file);
}
diff --git a/block/vdi.c b/block/vdi.c
index a2da6ecab0..3ddc62a569 100644
--- a/block/vdi.c
+++ b/block/vdi.c
@@ -523,8 +523,8 @@ static int vdi_reopen_prepare(BDRVReopenState *state,
}
static int coroutine_fn GRAPH_RDLOCK
-vdi_co_block_status(BlockDriverState *bs, bool want_zero, int64_t offset,
- int64_t bytes, int64_t *pnum, int64_t *map,
+vdi_co_block_status(BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes, int64_t *pnum, int64_t *map,
BlockDriverState **file)
{
BDRVVdiState *s = (BDRVVdiState *)bs->opaque;
diff --git a/block/vmdk.c b/block/vmdk.c
index 2adec49912..9c7ab037e1 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -1777,7 +1777,7 @@ static inline uint64_t vmdk_find_offset_in_cluster(VmdkExtent *extent,
}
static int coroutine_fn GRAPH_RDLOCK
-vmdk_co_block_status(BlockDriverState *bs, bool want_zero,
+vmdk_co_block_status(BlockDriverState *bs, unsigned int mode,
int64_t offset, int64_t bytes, int64_t *pnum,
int64_t *map, BlockDriverState **file)
{
diff --git a/block/vpc.c b/block/vpc.c
index 0309e319f6..801ff5793f 100644
--- a/block/vpc.c
+++ b/block/vpc.c
@@ -726,7 +726,7 @@ fail:
}
static int coroutine_fn GRAPH_RDLOCK
-vpc_co_block_status(BlockDriverState *bs, bool want_zero,
+vpc_co_block_status(BlockDriverState *bs, unsigned int mode,
int64_t offset, int64_t bytes,
int64_t *pnum, int64_t *map,
BlockDriverState **file)
diff --git a/block/vvfat.c b/block/vvfat.c
index 91d69b3cc8..814796d918 100644
--- a/block/vvfat.c
+++ b/block/vvfat.c
@@ -3134,9 +3134,9 @@ vvfat_co_pwritev(BlockDriverState *bs, int64_t offset, int64_t bytes,
}
static int coroutine_fn vvfat_co_block_status(BlockDriverState *bs,
- bool want_zero, int64_t offset,
- int64_t bytes, int64_t *n,
- int64_t *map,
+ unsigned int mode,
+ int64_t offset, int64_t bytes,
+ int64_t *n, int64_t *map,
BlockDriverState **file)
{
*n = bytes;
diff --git a/include/block/block-common.h b/include/block/block-common.h
index 0b831ef87b..c8c626daea 100644
--- a/include/block/block-common.h
+++ b/include/block/block-common.h
@@ -333,6 +333,17 @@ typedef enum {
#define BDRV_BLOCK_RECURSE 0x40
#define BDRV_BLOCK_COMPRESSED 0x80
+/*
+ * Block status hints: the bitwise-or of these flags emphasize what
+ * the caller hopes to learn, and some drivers may be able to give
+ * faster answers by doing less work when the hint permits.
+ */
+#define BDRV_WANT_ZERO BDRV_BLOCK_ZERO
+#define BDRV_WANT_OFFSET_VALID BDRV_BLOCK_OFFSET_VALID
+#define BDRV_WANT_ALLOCATED BDRV_BLOCK_ALLOCATED
+#define BDRV_WANT_PRECISE (BDRV_WANT_ZERO | BDRV_WANT_OFFSET_VALID | \
+ BDRV_WANT_OFFSET_VALID)
+
typedef QTAILQ_HEAD(BlockReopenQueue, BlockReopenQueueEntry) BlockReopenQueue;
typedef struct BDRVReopenState {
diff --git a/include/block/block_int-common.h b/include/block/block_int-common.h
index ebb4e56a50..a9c0daa2a4 100644
--- a/include/block/block_int-common.h
+++ b/include/block/block_int-common.h
@@ -608,15 +608,16 @@ struct BlockDriver {
* according to the current layer, and should only need to set
* BDRV_BLOCK_DATA, BDRV_BLOCK_ZERO, BDRV_BLOCK_OFFSET_VALID,
* and/or BDRV_BLOCK_RAW; if the current layer defers to a backing
- * layer, the result should be 0 (and not BDRV_BLOCK_ZERO). See
- * block.h for the overall meaning of the bits. As a hint, the
- * flag want_zero is true if the caller cares more about precise
- * mappings (favor accurate _OFFSET_VALID/_ZERO) or false for
- * overall allocation (favor larger *pnum, perhaps by reporting
- * _DATA instead of _ZERO). The block layer guarantees input
- * clamped to bdrv_getlength() and aligned to request_alignment,
- * as well as non-NULL pnum, map, and file; in turn, the driver
- * must return an error or set pnum to an aligned non-zero value.
+ * layer, the result should be 0 (and not BDRV_BLOCK_ZERO). The
+ * caller will synthesize BDRV_BLOCK_ALLOCATED based on the
+ * non-zero results. See block.h for the overall meaning of the
+ * bits. As a hint, the flags in @mode may include a bitwise-or
+ * of BDRV_WANT_ALLOCATED, BDRV_WANT_OFFSET_VALID, or
+ * BDRV_WANT_ZERO based on what the caller is looking for in the
+ * results. The block layer guarantees input clamped to
+ * bdrv_getlength() and aligned to request_alignment, as well as
+ * non-NULL pnum, map, and file; in turn, the driver must return
+ * an error or set pnum to an aligned non-zero value.
*
* Note that @bytes is just a hint on how big of a region the
* caller wants to inspect. It is not a limit on *pnum.
@@ -628,8 +629,8 @@ struct BlockDriver {
* to clamping *pnum for return to its caller.
*/
int coroutine_fn GRAPH_RDLOCK_PTR (*bdrv_co_block_status)(
- BlockDriverState *bs,
- bool want_zero, int64_t offset, int64_t bytes, int64_t *pnum,
+ BlockDriverState *bs, unsigned int mode,
+ int64_t offset, int64_t bytes, int64_t *pnum,
int64_t *map, BlockDriverState **file);
/*
@@ -653,8 +654,8 @@ struct BlockDriver {
QEMUIOVector *qiov, size_t qiov_offset);
int coroutine_fn GRAPH_RDLOCK_PTR (*bdrv_co_snapshot_block_status)(
- BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
- int64_t *pnum, int64_t *map, BlockDriverState **file);
+ BlockDriverState *bs, unsigned int mode, int64_t offset,
+ int64_t bytes, int64_t *pnum, int64_t *map, BlockDriverState **file);
int coroutine_fn GRAPH_RDLOCK_PTR (*bdrv_co_pdiscard_snapshot)(
BlockDriverState *bs, int64_t offset, int64_t bytes);
diff --git a/include/block/block_int-io.h b/include/block/block_int-io.h
index 4a7cf2b4fd..4f94eb3c5a 100644
--- a/include/block/block_int-io.h
+++ b/include/block/block_int-io.h
@@ -38,8 +38,8 @@
int coroutine_fn GRAPH_RDLOCK bdrv_co_preadv_snapshot(BdrvChild *child,
int64_t offset, int64_t bytes, QEMUIOVector *qiov, size_t qiov_offset);
int coroutine_fn GRAPH_RDLOCK bdrv_co_snapshot_block_status(
- BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
- int64_t *pnum, int64_t *map, BlockDriverState **file);
+ BlockDriverState *bs, unsigned int mode, int64_t offset,
+ int64_t bytes, int64_t *pnum, int64_t *map, BlockDriverState **file);
int coroutine_fn GRAPH_RDLOCK bdrv_co_pdiscard_snapshot(BlockDriverState *bs,
int64_t offset, int64_t bytes);
diff --git a/tests/unit/test-block-iothread.c b/tests/unit/test-block-iothread.c
index 2b358eaaa8..e26b3be593 100644
--- a/tests/unit/test-block-iothread.c
+++ b/tests/unit/test-block-iothread.c
@@ -63,7 +63,7 @@ bdrv_test_co_truncate(BlockDriverState *bs, int64_t offset, bool exact,
}
static int coroutine_fn bdrv_test_co_block_status(BlockDriverState *bs,
- bool want_zero,
+ unsigned int mode,
int64_t offset, int64_t count,
int64_t *pnum, int64_t *map,
BlockDriverState **file)
--
2.39.3

90
kvm-block-Let-bdrv_co_is_zero_fast-consolidate-adjacent-.patch Normal file
View File

@ -0,0 +1,90 @@
From e101b9872f9b3f6c5e128f29d7c3bb91faca362b Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:20 -0500
Subject: [PATCH 03/14] block: Let bdrv_co_is_zero_fast consolidate adjacent
extents
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/14] 4520f7ef5bcc5803413541e6f48bf750af3d31e0 (ebblake/centos-qemu-kvm)
Some BDS drivers have a cap on how much block status they can supply
in one query (for example, NBD talking to an older server cannot
inspect more than 4G per query; and qcow2 tends to cap its answers
rather than cross a cluster boundary of an L1 table). Although the
existing callers of bdrv_co_is_zero_fast are not passing in that large
of a 'bytes' parameter, an upcoming caller wants to query the entire
image at once, and will thus benefit from being able to treat adjacent
zero regions in a coalesced manner, rather than claiming the region is
non-zero merely because pnum was truncated and didn't match the
incoming bytes.
While refactoring this into a loop, note that there is no need to
assign pnum prior to calling bdrv_co_common_block_status_above() (it
is guaranteed to be assigned deeper in the callstack).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-18-eblake@redhat.com>
(cherry picked from commit 31bf15d97dd1d205a3b264675f9a1b3bd1939068)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/io.c | 27 +++++++++++++++------------
1 file changed, 15 insertions(+), 12 deletions(-)
diff --git a/block/io.c b/block/io.c
index e328402adc..64f4b1d22a 100644
--- a/block/io.c
+++ b/block/io.c
@@ -2751,28 +2751,31 @@ int coroutine_fn bdrv_co_block_status(BlockDriverState *bs, int64_t offset,
* by @offset and @bytes is known to read as zeroes.
* Return 1 if that is the case, 0 otherwise and -errno on error.
* This test is meant to be fast rather than accurate so returning 0
- * does not guarantee non-zero data.
+ * does not guarantee non-zero data; but a return of 1 is reliable.
*/
int coroutine_fn bdrv_co_is_zero_fast(BlockDriverState *bs, int64_t offset,
int64_t bytes)
{
int ret;
- int64_t pnum = bytes;
+ int64_t pnum;
IO_CODE();
- if (!bytes) {
- return 1;
- }
-
- ret = bdrv_co_common_block_status_above(bs, NULL, false, BDRV_WANT_ZERO,
- offset, bytes, &pnum, NULL, NULL,
- NULL);
+ while (bytes) {
+ ret = bdrv_co_common_block_status_above(bs, NULL, false,
+ BDRV_WANT_ZERO, offset, bytes,
+ &pnum, NULL, NULL, NULL);
- if (ret < 0) {
- return ret;
+ if (ret < 0) {
+ return ret;
+ }
+ if (!(ret & BDRV_BLOCK_ZERO)) {
+ return 0;
+ }
+ offset += pnum;
+ bytes -= pnum;
}
- return (pnum == bytes) && (ret & BDRV_BLOCK_ZERO);
+ return 1;
}
int coroutine_fn bdrv_co_is_allocated(BlockDriverState *bs, int64_t offset,
--
2.39.3

64
kvm-file-posix-gluster-Handle-zero-block-status-hint-bet.patch Normal file
View File

@ -0,0 +1,64 @@
From f8d89f67817fa362a3b8ed0721775e353dac8f18 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:19 -0500
Subject: [PATCH 02/14] file-posix, gluster: Handle zero block status hint
better
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/14] c40cd3f8cda2ea1646d90fd174b5f0dbd3e1a50b (ebblake/centos-qemu-kvm)
Although the previous patch to change 'bool want_zero' into a bitmask
made no semantic change, it is now time to differentiate. When the
caller specifically wants to know what parts of the file read as zero,
we need to use lseek and actually reporting holes, rather than
short-circuiting and advertising full allocation.
This change will be utilized in later patches to let mirroring
optimize for the case when the destination already reads as zeroes.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-17-eblake@redhat.com>
(cherry picked from commit a6a0a7fb0e327d17594c971b4a39de14e025b415)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/file-posix.c | 3 ++-
block/gluster.c | 2 +-
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 0c6569742f..dea7b09b6c 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -3282,7 +3282,8 @@ static int coroutine_fn raw_co_block_status(BlockDriverState *bs,
return ret;
}
- if (mode != BDRV_WANT_PRECISE) {
+ if (!(mode & BDRV_WANT_ZERO)) {
+ /* There is no backing file - all bytes are allocated in this file. */
*pnum = bytes;
*map = offset;
*file = bs;
diff --git a/block/gluster.c b/block/gluster.c
index 8197b0ecef..e702666cbc 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -1482,7 +1482,7 @@ static int coroutine_fn qemu_gluster_co_block_status(BlockDriverState *bs,
return ret;
}
- if (mode != BDRV_WANT_PRECISE) {
+ if (!(mode & BDRV_WANT_ZERO)) {
*pnum = bytes;
*map = offset;
*file = bs;
--
2.39.3

42
kvm-iotests-Improve-iotest-194-to-mirror-data.patch Normal file
View File

@ -0,0 +1,42 @@
From 11b46a271d73631177f59ff581a408f967c30fb9 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:22 -0500
Subject: [PATCH 05/14] iotests: Improve iotest 194 to mirror data
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [5/14] 9f1fd3c7d4332ac310af4eb37e8f2122f6324294 (ebblake/centos-qemu-kvm)
Mirroring a completely sparse image to a sparse destination should be
practically instantaneous. It isn't yet, but the test will be more
realistic if it has some non-zero to mirror as well as the holes.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-20-eblake@redhat.com>
(cherry picked from commit eb89627899bb84148d272394e885725eff456ae9)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
tests/qemu-iotests/194 | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/qemu-iotests/194 b/tests/qemu-iotests/194
index c0ce82dd25..d0b9c084f5 100755
--- a/tests/qemu-iotests/194
+++ b/tests/qemu-iotests/194
@@ -34,6 +34,7 @@ with iotests.FilePath('source.img') as source_img_path, \
img_size = '1G'
iotests.qemu_img_create('-f', iotests.imgfmt, source_img_path, img_size)
+ iotests.qemu_io('-f', iotests.imgfmt, '-c', 'write 512M 1M', source_img_path)
iotests.qemu_img_create('-f', iotests.imgfmt, dest_img_path, img_size)
iotests.log('Launching VMs...')
--
2.39.3

68
kvm-iotests-common.rc-add-disk_usage-function.patch Normal file
View File

@ -0,0 +1,68 @@
From d8ed5039981b1eb81d229d8ee672d5ee28862e92 Mon Sep 17 00:00:00 2001
From: Andrey Drobyshev <andrey.drobyshev@virtuozzo.com>
Date: Fri, 9 May 2025 15:40:29 -0500
Subject: [PATCH 12/14] iotests/common.rc: add disk_usage function
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [12/14] 0a007f9d09f01b50cf4edeb8ac8217356b2cb5d2 (ebblake/centos-qemu-kvm)
Move the definition from iotests/250 to common.rc. This is used to
detect real disk usage of sparse files. In particular, we want to use
it for checking subclusters-based discards.
Signed-off-by: Andrey Drobyshev <andrey.drobyshev@virtuozzo.com>
Reviewed-by: Alexander Ivanov <alexander.ivanov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-ID: <20240913163942.423050-6-andrey.drobyshev@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-27-eblake@redhat.com>
(cherry picked from commit be9bac072ede6e6aa27079f59efcf17b56bd7b26)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
tests/qemu-iotests/250 | 5 -----
tests/qemu-iotests/common.rc | 6 ++++++
2 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/tests/qemu-iotests/250 b/tests/qemu-iotests/250
index af48f83aba..c0a0dbc0ff 100755
--- a/tests/qemu-iotests/250
+++ b/tests/qemu-iotests/250
@@ -52,11 +52,6 @@ _unsupported_imgopts data_file
# bdrv_co_truncate(bs->file) call in qcow2_co_truncate(), which might succeed
# anyway.
-disk_usage()
-{
- du --block-size=1 $1 | awk '{print $1}'
-}
-
size=2100M
_make_test_img -o "cluster_size=1M,preallocation=metadata" $size
diff --git a/tests/qemu-iotests/common.rc b/tests/qemu-iotests/common.rc
index 95c12577dd..237f746af8 100644
--- a/tests/qemu-iotests/common.rc
+++ b/tests/qemu-iotests/common.rc
@@ -140,6 +140,12 @@ _optstr_add()
fi
}
+# report real disk usage for sparse files
+disk_usage()
+{
+ du --block-size=1 "$1" | awk '{print $1}'
+}
+
# Set the variables to the empty string to turn Valgrind off
# for specific processes, e.g.
# $ VALGRIND_QEMU_IO= ./check -qcow2 -valgrind 015
--
2.39.3

295
kvm-mirror-Allow-QMP-override-to-declare-target-already-.patch Normal file
View File

@ -0,0 +1,295 @@
From bc4571743fc3bbb829101fbf294615e3b8fb3577 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:25 -0500
Subject: [PATCH 08/14] mirror: Allow QMP override to declare target already
zero
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [8/14] 4ad06d67db8c43df4e6e0b8f929b1d8c19e4b338 (ebblake/centos-qemu-kvm)
QEMU has an optimization for a just-created drive-mirror destination
that is not possible for blockdev-mirror (which can't create the
destination) - any time we know the destination starts life as all
zeroes, we can skip a pre-zeroing pass on the destination. Recent
patches have added an improved heuristic for detecting if a file
contains all zeroes, and we plan to use that heuristic in upcoming
patches. But since a heuristic cannot quickly detect all scenarios,
and there may be cases where the caller is aware of information that
QEMU cannot learn quickly, it makes sense to have a way to tell QEMU
to assume facts about the destination that can make the mirror
operation faster. Given our existing example of "qemu-img convert
--target-is-zero", it is time to expose this override in QMP for
blockdev-mirror as well.
This patch results in some slight redundancy between the older
s->zero_target (set any time mode==FULL and the destination image was
not just created - ie. clear if drive-mirror is asking to skip the
pre-zero pass) and the newly-introduced s->target_is_zero (in addition
to the QMP override, it is set when drive-mirror creates the
destination image); this will be cleaned up in the next patch.
There is also a subtlety that we must consider. When drive-mirror is
passing target_is_zero on behalf of a just-created image, we know the
image is sparse (skipping the pre-zeroing keeps it that way), so it
doesn't matter whether the destination also has "discard":"unmap" and
"detect-zeroes":"unmap". But now that we are letting the user set the
knob for target-is-zero, if the user passes a pre-existing file that
is fully allocated, it is fine to leave the file fully allocated under
"detect-zeroes":"on", but if the file is open with
"detect-zeroes":"unmap", we should really be trying harder to punch
holes in the destination for every region of zeroes copied from the
source. The easiest way to do this is to still run the pre-zeroing
pass (turning the entire destination file sparse before populating
just the allocated portions of the source), even though that currently
results in double I/O to the portions of the file that are allocated.
A later patch will add further optimizations to reduce redundant
zeroing I/O during the mirror operation.
Since "target-is-zero":true is designed for optimizations, it is okay
to silently ignore the parameter rather than erroring if the user ever
sets the parameter in a scenario where the mirror job can't exploit it
(for example, when doing "sync":"top" instead of "sync":"full", we
can't pre-zero, so setting the parameter won't make a speed
difference).
Signed-off-by: Eric Blake <eblake@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20250509204341.3553601-23-eblake@redhat.com>
Reviewed-by: Sunny Zhu <sunnyzhyy@qq.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit d17a34bfb94bda3a89d7320ae67255ded1d8c939)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 27 ++++++++++++++++++++++----
blockdev.c | 18 ++++++++++-------
include/block/block_int-global-state.h | 3 ++-
qapi/block-core.json | 8 +++++++-
tests/unit/test-block-iothread.c | 2 +-
5 files changed, 44 insertions(+), 14 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 2599b75d09..4dcb50c81a 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -55,6 +55,8 @@ typedef struct MirrorBlockJob {
BlockMirrorBackingMode backing_mode;
/* Whether the target image requires explicit zero-initialization */
bool zero_target;
+ /* Whether the target should be assumed to be already zero initialized */
+ bool target_is_zero;
/*
* To be accesssed with atomics. Written only under the BQL (required by the
* current implementation of mirror_change()).
@@ -844,12 +846,26 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
BlockDriverState *target_bs = blk_bs(s->target);
int ret = -EIO;
int64_t count;
+ bool punch_holes =
+ target_bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP &&
+ bdrv_can_write_zeroes_with_unmap(target_bs);
bdrv_graph_co_rdlock();
bs = s->mirror_top_bs->backing->bs;
bdrv_graph_co_rdunlock();
- if (s->zero_target) {
+ if (s->zero_target && (!s->target_is_zero || punch_holes)) {
+ /*
+ * Here, we are in FULL mode; our goal is to avoid writing
+ * zeroes if the destination already reads as zero, except
+ * when we are trying to punch holes. This is possible if
+ * zeroing happened externally (s->target_is_zero) or if we
+ * have a fast way to pre-zero the image (the dirty bitmap
+ * will be populated later by the non-zero portions, the same
+ * as for TOP mode). If pre-zeroing is not fast, or we need
+ * to punch holes, then our only recourse is to write the
+ * entire image.
+ */
if (!bdrv_can_write_zeroes_with_unmap(target_bs)) {
bdrv_set_dirty_bitmap(s->dirty_bitmap, 0, s->bdev_length);
return 0;
@@ -1714,7 +1730,7 @@ static BlockJob *mirror_start_job(
uint32_t granularity, int64_t buf_size,
MirrorSyncMode sync_mode,
BlockMirrorBackingMode backing_mode,
- bool zero_target,
+ bool zero_target, bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap,
@@ -1883,6 +1899,7 @@ static BlockJob *mirror_start_job(
s->sync_mode = sync_mode;
s->backing_mode = backing_mode;
s->zero_target = zero_target;
+ s->target_is_zero = target_is_zero;
qatomic_set(&s->copy_mode, copy_mode);
s->base = base;
s->base_overlay = bdrv_find_overlay(bs, base);
@@ -2011,7 +2028,7 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
int creation_flags, int64_t speed,
uint32_t granularity, int64_t buf_size,
MirrorSyncMode mode, BlockMirrorBackingMode backing_mode,
- bool zero_target,
+ bool zero_target, bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap, const char *filter_node_name,
@@ -2034,7 +2051,8 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
mirror_start_job(job_id, bs, creation_flags, target, replaces,
speed, granularity, buf_size, mode, backing_mode,
- zero_target, on_source_error, on_target_error, unmap,
+ zero_target,
+ target_is_zero, on_source_error, on_target_error, unmap,
NULL, NULL, &mirror_job_driver, base, false,
filter_node_name, true, copy_mode, false, errp);
}
@@ -2062,6 +2080,7 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
job = mirror_start_job(
job_id, bs, creation_flags, base, NULL, speed, 0, 0,
MIRROR_SYNC_MODE_TOP, MIRROR_LEAVE_BACKING_CHAIN, false,
+ false,
on_error, on_error, true, cb, opaque,
&commit_active_job_driver, base, auto_complete,
filter_node_name, false, MIRROR_COPY_MODE_BACKGROUND,
diff --git a/blockdev.c b/blockdev.c
index 1d1f27cfff..2e2fed539e 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2798,7 +2798,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
const char *replaces,
enum MirrorSyncMode sync,
BlockMirrorBackingMode backing_mode,
- bool zero_target,
+ bool zero_target, bool target_is_zero,
bool has_speed, int64_t speed,
bool has_granularity, uint32_t granularity,
bool has_buf_size, int64_t buf_size,
@@ -2909,11 +2909,10 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
/* pass the node name to replace to mirror start since it's loose coupling
* and will allow to check whether the node still exist at mirror completion
*/
- mirror_start(job_id, bs, target,
- replaces, job_flags,
+ mirror_start(job_id, bs, target, replaces, job_flags,
speed, granularity, buf_size, sync, backing_mode, zero_target,
- on_source_error, on_target_error, unmap, filter_node_name,
- copy_mode, errp);
+ target_is_zero, on_source_error, on_target_error, unmap,
+ filter_node_name, copy_mode, errp);
}
void qmp_drive_mirror(DriveMirror *arg, Error **errp)
@@ -2928,6 +2927,7 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
int64_t size;
const char *format = arg->format;
bool zero_target;
+ bool target_is_zero;
int ret;
bs = qmp_get_root_bs(arg->device, errp);
@@ -3044,6 +3044,8 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
zero_target = (arg->sync == MIRROR_SYNC_MODE_FULL &&
(arg->mode == NEW_IMAGE_MODE_EXISTING ||
!bdrv_has_zero_init(target_bs)));
+ target_is_zero = (arg->mode != NEW_IMAGE_MODE_EXISTING &&
+ bdrv_has_zero_init(target_bs));
bdrv_graph_rdunlock_main_loop();
@@ -3055,7 +3057,7 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
blockdev_mirror_common(arg->job_id, bs, target_bs,
arg->replaces, arg->sync,
- backing_mode, zero_target,
+ backing_mode, zero_target, target_is_zero,
arg->has_speed, arg->speed,
arg->has_granularity, arg->granularity,
arg->has_buf_size, arg->buf_size,
@@ -3085,6 +3087,7 @@ void qmp_blockdev_mirror(const char *job_id,
bool has_copy_mode, MirrorCopyMode copy_mode,
bool has_auto_finalize, bool auto_finalize,
bool has_auto_dismiss, bool auto_dismiss,
+ bool has_target_is_zero, bool target_is_zero,
Error **errp)
{
BlockDriverState *bs;
@@ -3115,7 +3118,8 @@ void qmp_blockdev_mirror(const char *job_id,
blockdev_mirror_common(job_id, bs, target_bs,
replaces, sync, backing_mode,
- zero_target, has_speed, speed,
+ zero_target, has_target_is_zero && target_is_zero,
+ has_speed, speed,
has_granularity, granularity,
has_buf_size, buf_size,
has_on_source_error, on_source_error,
diff --git a/include/block/block_int-global-state.h b/include/block/block_int-global-state.h
index eb2d92a226..8cf0003ce7 100644
--- a/include/block/block_int-global-state.h
+++ b/include/block/block_int-global-state.h
@@ -140,6 +140,7 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
* @mode: Whether to collapse all images in the chain to the target.
* @backing_mode: How to establish the target's backing chain after completion.
* @zero_target: Whether the target should be explicitly zero-initialized
+ * @target_is_zero: Whether the target already is zero-initialized.
* @on_source_error: The action to take upon error reading from the source.
* @on_target_error: The action to take upon error writing to the target.
* @unmap: Whether to unmap target where source sectors only contain zeroes.
@@ -159,7 +160,7 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
int creation_flags, int64_t speed,
uint32_t granularity, int64_t buf_size,
MirrorSyncMode mode, BlockMirrorBackingMode backing_mode,
- bool zero_target,
+ bool zero_target, bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap, const char *filter_node_name,
diff --git a/qapi/block-core.json b/qapi/block-core.json
index b1937780e1..7f70ec6d3c 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -2538,6 +2538,11 @@
# disappear from the query list without user intervention.
# Defaults to true. (Since 3.1)
#
+# @target-is-zero: Assume the destination reads as all zeroes before
+# the mirror started. Setting this to true can speed up the
+# mirror. Setting this to true when the destination is not
+# actually all zero can corrupt the destination. (Since 10.1)
+#
# Since: 2.6
#
# .. qmp-example::
@@ -2557,7 +2562,8 @@
'*on-target-error': 'BlockdevOnError',
'*filter-node-name': 'str',
'*copy-mode': 'MirrorCopyMode',
- '*auto-finalize': 'bool', '*auto-dismiss': 'bool' },
+ '*auto-finalize': 'bool', '*auto-dismiss': 'bool',
+ '*target-is-zero': 'bool'},
'allow-preconfig': true }
##
diff --git a/tests/unit/test-block-iothread.c b/tests/unit/test-block-iothread.c
index e26b3be593..54aed8252c 100644
--- a/tests/unit/test-block-iothread.c
+++ b/tests/unit/test-block-iothread.c
@@ -755,7 +755,7 @@ static void test_propagate_mirror(void)
/* Start a mirror job */
mirror_start("job0", src, target, NULL, JOB_DEFAULT, 0, 0, 0,
- MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false,
+ MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false, false,
BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
false, "filter_node", MIRROR_COPY_MODE_BACKGROUND,
&error_abort);
--
2.39.3

241
kvm-mirror-Drop-redundant-zero_target-parameter.patch Normal file
View File

@ -0,0 +1,241 @@
From db1a158312c2b94af1c1a50e0f13ace6ae58f0b6 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:26 -0500
Subject: [PATCH 09/14] mirror: Drop redundant zero_target parameter
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [9/14] b4cbd267c81b4758f59e0d51b947fd450caf6ef5 (ebblake/centos-qemu-kvm)
The two callers to a mirror job (drive-mirror and blockdev-mirror) set
zero_target precisely when sync mode == FULL, with the one exception
that drive-mirror skips zeroing the target if it was newly created and
reads as zero. But given the previous patch, that exception is
equally captured by target_is_zero.
Meanwhile, there is another slight wrinkle, fortunately caught by
iotest 185: if the caller uses "sync":"top" but the source has no
backing file, the code in blockdev.c was changing sync to be FULL, but
only after it had set zero_target=false. In mirror.c, prior to recent
patches, this didn't matter: the only places that inspected sync were
setting is_none_mode (both TOP and FULL had set that to false), and
mirror_start() setting base = mode == MIRROR_SYNC_MODE_TOP ?
bdrv_backing_chain_next(bs) : NULL. But now that we are passing sync
around, the slammed sync mode would result in a new pre-zeroing pass
even when the user had passed "sync":"top" in an effort to skip
pre-zeroing. Fortunately, the assignment of base when bs has no
backing chain still works out to NULL if we don't slam things. So
with the forced change of sync ripped out of blockdev.c, the sync mode
is passed through the full callstack unmolested, and we can now
reliably reconstruct the same settings as what used to be passed in by
zero_target=false, without the redundant parameter.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250509204341.3553601-24-eblake@redhat.com>
Reviewed-by: Sunny Zhu <sunnyzhyy@qq.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
[eblake: Fix regression in iotest 185]
Signed-off-by: Eric Blake <eblake@redhat.com>
(cherry picked from commit 253b43a29077de9266351e120c600a73b82e9c49)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 13 +++++--------
blockdev.c | 19 ++++---------------
include/block/block_int-global-state.h | 3 +--
tests/unit/test-block-iothread.c | 2 +-
4 files changed, 11 insertions(+), 26 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 4dcb50c81a..d04db85883 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -53,8 +53,6 @@ typedef struct MirrorBlockJob {
Error *replace_blocker;
MirrorSyncMode sync_mode;
BlockMirrorBackingMode backing_mode;
- /* Whether the target image requires explicit zero-initialization */
- bool zero_target;
/* Whether the target should be assumed to be already zero initialized */
bool target_is_zero;
/*
@@ -854,7 +852,9 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
bs = s->mirror_top_bs->backing->bs;
bdrv_graph_co_rdunlock();
- if (s->zero_target && (!s->target_is_zero || punch_holes)) {
+ if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
+ /* In TOP mode, there is no benefit to a pre-zeroing pass. */
+ } else if (!s->target_is_zero || punch_holes) {
/*
* Here, we are in FULL mode; our goal is to avoid writing
* zeroes if the destination already reads as zero, except
@@ -1730,7 +1730,7 @@ static BlockJob *mirror_start_job(
uint32_t granularity, int64_t buf_size,
MirrorSyncMode sync_mode,
BlockMirrorBackingMode backing_mode,
- bool zero_target, bool target_is_zero,
+ bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap,
@@ -1898,7 +1898,6 @@ static BlockJob *mirror_start_job(
s->on_target_error = on_target_error;
s->sync_mode = sync_mode;
s->backing_mode = backing_mode;
- s->zero_target = zero_target;
s->target_is_zero = target_is_zero;
qatomic_set(&s->copy_mode, copy_mode);
s->base = base;
@@ -2028,7 +2027,7 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
int creation_flags, int64_t speed,
uint32_t granularity, int64_t buf_size,
MirrorSyncMode mode, BlockMirrorBackingMode backing_mode,
- bool zero_target, bool target_is_zero,
+ bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap, const char *filter_node_name,
@@ -2051,7 +2050,6 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
mirror_start_job(job_id, bs, creation_flags, target, replaces,
speed, granularity, buf_size, mode, backing_mode,
- zero_target,
target_is_zero, on_source_error, on_target_error, unmap,
NULL, NULL, &mirror_job_driver, base, false,
filter_node_name, true, copy_mode, false, errp);
@@ -2080,7 +2078,6 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
job = mirror_start_job(
job_id, bs, creation_flags, base, NULL, speed, 0, 0,
MIRROR_SYNC_MODE_TOP, MIRROR_LEAVE_BACKING_CHAIN, false,
- false,
on_error, on_error, true, cb, opaque,
&commit_active_job_driver, base, auto_complete,
filter_node_name, false, MIRROR_COPY_MODE_BACKGROUND,
diff --git a/blockdev.c b/blockdev.c
index 2e2fed539e..0fa8813efe 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2798,7 +2798,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
const char *replaces,
enum MirrorSyncMode sync,
BlockMirrorBackingMode backing_mode,
- bool zero_target, bool target_is_zero,
+ bool target_is_zero,
bool has_speed, int64_t speed,
bool has_granularity, uint32_t granularity,
bool has_buf_size, int64_t buf_size,
@@ -2865,10 +2865,6 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
return;
}
- if (!bdrv_backing_chain_next(bs) && sync == MIRROR_SYNC_MODE_TOP) {
- sync = MIRROR_SYNC_MODE_FULL;
- }
-
if (!replaces) {
/* We want to mirror from @bs, but keep implicit filters on top */
unfiltered_bs = bdrv_skip_implicit_filters(bs);
@@ -2910,7 +2906,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
* and will allow to check whether the node still exist at mirror completion
*/
mirror_start(job_id, bs, target, replaces, job_flags,
- speed, granularity, buf_size, sync, backing_mode, zero_target,
+ speed, granularity, buf_size, sync, backing_mode,
target_is_zero, on_source_error, on_target_error, unmap,
filter_node_name, copy_mode, errp);
}
@@ -2926,7 +2922,6 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
int flags;
int64_t size;
const char *format = arg->format;
- bool zero_target;
bool target_is_zero;
int ret;
@@ -3041,9 +3036,6 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
}
bdrv_graph_rdlock_main_loop();
- zero_target = (arg->sync == MIRROR_SYNC_MODE_FULL &&
- (arg->mode == NEW_IMAGE_MODE_EXISTING ||
- !bdrv_has_zero_init(target_bs)));
target_is_zero = (arg->mode != NEW_IMAGE_MODE_EXISTING &&
bdrv_has_zero_init(target_bs));
bdrv_graph_rdunlock_main_loop();
@@ -3057,7 +3049,7 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
blockdev_mirror_common(arg->job_id, bs, target_bs,
arg->replaces, arg->sync,
- backing_mode, zero_target, target_is_zero,
+ backing_mode, target_is_zero,
arg->has_speed, arg->speed,
arg->has_granularity, arg->granularity,
arg->has_buf_size, arg->buf_size,
@@ -3094,7 +3086,6 @@ void qmp_blockdev_mirror(const char *job_id,
BlockDriverState *target_bs;
AioContext *aio_context;
BlockMirrorBackingMode backing_mode = MIRROR_LEAVE_BACKING_CHAIN;
- bool zero_target;
int ret;
bs = qmp_get_root_bs(device, errp);
@@ -3107,8 +3098,6 @@ void qmp_blockdev_mirror(const char *job_id,
return;
}
- zero_target = (sync == MIRROR_SYNC_MODE_FULL);
-
aio_context = bdrv_get_aio_context(bs);
ret = bdrv_try_change_aio_context(target_bs, aio_context, NULL, errp);
@@ -3118,7 +3107,7 @@ void qmp_blockdev_mirror(const char *job_id,
blockdev_mirror_common(job_id, bs, target_bs,
replaces, sync, backing_mode,
- zero_target, has_target_is_zero && target_is_zero,
+ has_target_is_zero && target_is_zero,
has_speed, speed,
has_granularity, granularity,
has_buf_size, buf_size,
diff --git a/include/block/block_int-global-state.h b/include/block/block_int-global-state.h
index 8cf0003ce7..d21bd7fd2f 100644
--- a/include/block/block_int-global-state.h
+++ b/include/block/block_int-global-state.h
@@ -139,7 +139,6 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
* @buf_size: The amount of data that can be in flight at one time.
* @mode: Whether to collapse all images in the chain to the target.
* @backing_mode: How to establish the target's backing chain after completion.
- * @zero_target: Whether the target should be explicitly zero-initialized
* @target_is_zero: Whether the target already is zero-initialized.
* @on_source_error: The action to take upon error reading from the source.
* @on_target_error: The action to take upon error writing to the target.
@@ -160,7 +159,7 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
int creation_flags, int64_t speed,
uint32_t granularity, int64_t buf_size,
MirrorSyncMode mode, BlockMirrorBackingMode backing_mode,
- bool zero_target, bool target_is_zero,
+ bool target_is_zero,
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
bool unmap, const char *filter_node_name,
diff --git a/tests/unit/test-block-iothread.c b/tests/unit/test-block-iothread.c
index 54aed8252c..e26b3be593 100644
--- a/tests/unit/test-block-iothread.c
+++ b/tests/unit/test-block-iothread.c
@@ -755,7 +755,7 @@ static void test_propagate_mirror(void)
/* Start a mirror job */
mirror_start("job0", src, target, NULL, JOB_DEFAULT, 0, 0, 0,
- MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false, false,
+ MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false,
BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
false, "filter_node", MIRROR_COPY_MODE_BACKGROUND,
&error_abort);
--
2.39.3

90
kvm-mirror-Minor-refactoring.patch Normal file
View File

@ -0,0 +1,90 @@
From e95294aecc606deacf716861d716f9178b132ed8 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:23 -0500
Subject: [PATCH 06/14] mirror: Minor refactoring
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [6/14] 22a87aca8033f3f5d10fd224dc0786633a4d040f (ebblake/centos-qemu-kvm)
Commit 5791ba52 (v9.2) pre-initialized ret in mirror_dirty_init to
silence a false positive compiler warning, even though in all code
paths where ret is used, it was guaranteed to be reassigned
beforehand. But since the function returns -errno, and -1 is not
always the right errno, it's better to initialize to -EIO.
An upcoming patch wants to track two bitmaps in
do_sync_target_write(); this will be easier if the current variables
related to the dirty bitmap are renamed.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250509204341.3553601-21-eblake@redhat.com>
(cherry picked from commit 870f8963cf1a84f8ec929b05a6d68906974a76c5)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 22 +++++++++++-----------
1 file changed, 11 insertions(+), 11 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index a53582f17b..34c6c5252e 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -841,7 +841,7 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
int64_t offset;
BlockDriverState *bs;
BlockDriverState *target_bs = blk_bs(s->target);
- int ret = -1;
+ int ret = -EIO;
int64_t count;
bdrv_graph_co_rdlock();
@@ -1341,7 +1341,7 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
{
int ret;
size_t qiov_offset = 0;
- int64_t bitmap_offset, bitmap_end;
+ int64_t dirty_bitmap_offset, dirty_bitmap_end;
if (!QEMU_IS_ALIGNED(offset, job->granularity) &&
bdrv_dirty_bitmap_get(job->dirty_bitmap, offset))
@@ -1388,11 +1388,11 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
* Tails are either clean or shrunk, so for bitmap resetting
* we safely align the range down.
*/
- bitmap_offset = QEMU_ALIGN_UP(offset, job->granularity);
- bitmap_end = QEMU_ALIGN_DOWN(offset + bytes, job->granularity);
- if (bitmap_offset < bitmap_end) {
- bdrv_reset_dirty_bitmap(job->dirty_bitmap, bitmap_offset,
- bitmap_end - bitmap_offset);
+ dirty_bitmap_offset = QEMU_ALIGN_UP(offset, job->granularity);
+ dirty_bitmap_end = QEMU_ALIGN_DOWN(offset + bytes, job->granularity);
+ if (dirty_bitmap_offset < dirty_bitmap_end) {
+ bdrv_reset_dirty_bitmap(job->dirty_bitmap, dirty_bitmap_offset,
+ dirty_bitmap_end - dirty_bitmap_offset);
}
job_progress_increase_remaining(&job->common.job, bytes);
@@ -1430,10 +1430,10 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
* at function start, and they must be still dirty, as we've locked
* the region for in-flight op.
*/
- bitmap_offset = QEMU_ALIGN_DOWN(offset, job->granularity);
- bitmap_end = QEMU_ALIGN_UP(offset + bytes, job->granularity);
- bdrv_set_dirty_bitmap(job->dirty_bitmap, bitmap_offset,
- bitmap_end - bitmap_offset);
+ dirty_bitmap_offset = QEMU_ALIGN_DOWN(offset, job->granularity);
+ dirty_bitmap_end = QEMU_ALIGN_UP(offset + bytes, job->granularity);
+ bdrv_set_dirty_bitmap(job->dirty_bitmap, dirty_bitmap_offset,
+ dirty_bitmap_end - dirty_bitmap_offset);
qatomic_set(&job->actively_synced, false);
action = mirror_error_action(job, false, -ret);
--
2.39.3

139
kvm-mirror-Pass-full-sync-mode-rather-than-bool-to-inter.patch Normal file
View File

@ -0,0 +1,139 @@
From db0b92495a4e774caafaaa148e778b575112bad2 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:24 -0500
Subject: [PATCH 07/14] mirror: Pass full sync mode rather than bool to
internals
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [7/14] e8872e3edad069ee6c76f6104c1bc277c025b5ac (ebblake/centos-qemu-kvm)
Out of the five possible values for MirrorSyncMode, INCREMENTAL and
BITMAP are already rejected up front in mirror_start, leaving NONE,
TOP, and FULL as the remaining values that the code was collapsing
into a single bool is_none_mode. Furthermore, mirror_dirty_init() is
only reachable for modes TOP and FULL, as further guided by
s->zero_target. However, upcoming patches want to further optimize
the pre-zeroing pass of a sync=full mirror in mirror_dirty_init(),
while avoiding that pass on a sync=top action. Instead of throwing
away context by collapsing these two values into
s->is_none_mode=false, it is better to pass s->sync_mode throughout
the entire operation. For active commit, the desired semantics match
sync mode TOP.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250509204341.3553601-22-eblake@redhat.com>
Reviewed-by: Sunny Zhu <sunnyzhyy@qq.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 9474d97bd7421b4fe7c806ab0949697514d11e88)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 24 ++++++++++++------------
1 file changed, 12 insertions(+), 12 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 34c6c5252e..2599b75d09 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -51,7 +51,7 @@ typedef struct MirrorBlockJob {
BlockDriverState *to_replace;
/* Used to block operations on the drive-mirror-replace target */
Error *replace_blocker;
- bool is_none_mode;
+ MirrorSyncMode sync_mode;
BlockMirrorBackingMode backing_mode;
/* Whether the target image requires explicit zero-initialization */
bool zero_target;
@@ -723,9 +723,10 @@ static int mirror_exit_common(Job *job)
&error_abort);
if (!abort && s->backing_mode == MIRROR_SOURCE_BACKING_CHAIN) {
- BlockDriverState *backing = s->is_none_mode ? src : s->base;
+ BlockDriverState *backing;
BlockDriverState *unfiltered_target = bdrv_skip_filters(target_bs);
+ backing = s->sync_mode == MIRROR_SYNC_MODE_NONE ? src : s->base;
if (bdrv_cow_bs(unfiltered_target) != backing) {
bdrv_set_backing_hd(unfiltered_target, backing, &local_err);
if (local_err) {
@@ -1020,7 +1021,7 @@ static int coroutine_fn mirror_run(Job *job, Error **errp)
mirror_free_init(s);
s->last_pause_ns = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
- if (!s->is_none_mode) {
+ if (s->sync_mode != MIRROR_SYNC_MODE_NONE) {
ret = mirror_dirty_init(s);
if (ret < 0 || job_is_cancelled(&s->common.job)) {
goto immediate_exit;
@@ -1711,6 +1712,7 @@ static BlockJob *mirror_start_job(
int creation_flags, BlockDriverState *target,
const char *replaces, int64_t speed,
uint32_t granularity, int64_t buf_size,
+ MirrorSyncMode sync_mode,
BlockMirrorBackingMode backing_mode,
bool zero_target,
BlockdevOnError on_source_error,
@@ -1719,7 +1721,7 @@ static BlockJob *mirror_start_job(
BlockCompletionFunc *cb,
void *opaque,
const BlockJobDriver *driver,
- bool is_none_mode, BlockDriverState *base,
+ BlockDriverState *base,
bool auto_complete, const char *filter_node_name,
bool is_mirror, MirrorCopyMode copy_mode,
bool base_ro,
@@ -1878,7 +1880,7 @@ static BlockJob *mirror_start_job(
s->replaces = g_strdup(replaces);
s->on_source_error = on_source_error;
s->on_target_error = on_target_error;
- s->is_none_mode = is_none_mode;
+ s->sync_mode = sync_mode;
s->backing_mode = backing_mode;
s->zero_target = zero_target;
qatomic_set(&s->copy_mode, copy_mode);
@@ -2015,7 +2017,6 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
bool unmap, const char *filter_node_name,
MirrorCopyMode copy_mode, Error **errp)
{
- bool is_none_mode;
BlockDriverState *base;
GLOBAL_STATE_CODE();
@@ -2028,14 +2029,13 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
}
bdrv_graph_rdlock_main_loop();
- is_none_mode = mode == MIRROR_SYNC_MODE_NONE;
base = mode == MIRROR_SYNC_MODE_TOP ? bdrv_backing_chain_next(bs) : NULL;
bdrv_graph_rdunlock_main_loop();
mirror_start_job(job_id, bs, creation_flags, target, replaces,
- speed, granularity, buf_size, backing_mode, zero_target,
- on_source_error, on_target_error, unmap, NULL, NULL,
- &mirror_job_driver, is_none_mode, base, false,
+ speed, granularity, buf_size, mode, backing_mode,
+ zero_target, on_source_error, on_target_error, unmap,
+ NULL, NULL, &mirror_job_driver, base, false,
filter_node_name, true, copy_mode, false, errp);
}
@@ -2061,9 +2061,9 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
job = mirror_start_job(
job_id, bs, creation_flags, base, NULL, speed, 0, 0,
- MIRROR_LEAVE_BACKING_CHAIN, false,
+ MIRROR_SYNC_MODE_TOP, MIRROR_LEAVE_BACKING_CHAIN, false,
on_error, on_error, true, cb, opaque,
- &commit_active_job_driver, false, base, auto_complete,
+ &commit_active_job_driver, base, auto_complete,
filter_node_name, false, MIRROR_COPY_MODE_BACKGROUND,
base_read_only, errp);
if (!job) {
--
2.39.3

58
kvm-mirror-Reduce-I-O-when-destination-is-detect-zeroes-.patch Normal file
View File

@ -0,0 +1,58 @@
From 9fedd14da6f1dc7aa3f0711d86f722397d080993 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Tue, 13 May 2025 17:00:45 -0500
Subject: [PATCH 14/14] mirror: Reduce I/O when destination is
detect-zeroes:unmap
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [14/14] d4ba9d88a8da00f82c2ba7ebf050152fbe1e2465 (ebblake/centos-qemu-kvm)
If we are going to punch holes in the mirror destination even for the
portions where the source image is unallocated, it is nicer to treat
the entire image as dirty and punch as we go, rather than pre-zeroing
the entire image just to re-do I/O to the allocated portions of the
image.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250513220142.535200-2-eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 9abfc81246c9cc1845080eec5920779961187c07)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 724318f037..c2c5099c95 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -920,11 +920,16 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
* zeroing happened externally (ret > 0) or if we have a fast
* way to pre-zero the image (the dirty bitmap will be
* populated later by the non-zero portions, the same as for
- * TOP mode). If pre-zeroing is not fast, then our only
- * recourse is to mark the entire image dirty. The act of
- * pre-zeroing will populate the zero bitmap.
+ * TOP mode). If pre-zeroing is not fast, or we need to visit
+ * the entire image in order to punch holes even in the
+ * non-allocated regions of the source, then just mark the
+ * entire image dirty and leave the zero bitmap clear at this
+ * point in time. Otherwise, it can be faster to pre-zero the
+ * image now, even if we re-write the allocated portions of
+ * the disk later, and the pre-zero pass will populate the
+ * zero bitmap.
*/
- if (!bdrv_can_write_zeroes_with_unmap(target_bs)) {
+ if (!bdrv_can_write_zeroes_with_unmap(target_bs) || punch_holes) {
bdrv_set_dirty_bitmap(s->dirty_bitmap, 0, s->bdev_length);
return 0;
}
--
2.39.3

180
kvm-mirror-Skip-pre-zeroing-destination-if-it-is-already.patch Normal file
View File

@ -0,0 +1,180 @@
From 92a033b6c8394c8efb5b881cbbe463eeff5711cd Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:27 -0500
Subject: [PATCH 10/14] mirror: Skip pre-zeroing destination if it is already
zero
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [10/14] 5d86d9c763a1bb49fab591a45e211d3be819ccfe (ebblake/centos-qemu-kvm)
When doing a sync=full mirroring, we can skip pre-zeroing the
destination if it already reads as zeroes and we are not also trying
to punch holes due to detect-zeroes. With this patch, there are fewer
scenarios that have to pass in an explicit target-is-zero, while still
resulting in a sparse destination remaining sparse.
A later patch will then further improve things to skip writing to the
destination for parts of the image where the source is zero; but even
with just this patch, it is possible to see a difference for any
source that does not report itself as fully allocated, coupled with a
destination BDS that can quickly report that it already reads as zero.
(For a source that reports as fully allocated, such as a file, the
rest of mirror_dirty_init() still sets the entire dirty bitmap to
true, so even though we avoided the pre-zeroing, we are not yet
avoiding all redundant I/O).
Iotest 194 detects the difference made by this patch: for a file
source (where block status reports the entire image as allocated, and
therefore we end up writing zeroes everywhere in the destination
anyways), the job length remains the same. But for a qcow2 source and
a destination that reads as all zeroes, the dirty bitmap changes to
just tracking the allocated portions of the source, which results in
faster completion and smaller job statistics. For the test to pass
with both ./check -file and -qcow2, a new python filter is needed to
mask out the now-varying job amounts (this matches the shell filters
_filter_block_job_{offset,len} in common.filter). A later test will
also be added which further validates expected sparseness, so it does
not matter that 194 is no longer explicitly looking at how many bytes
were copied.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250509204341.3553601-25-eblake@redhat.com>
Reviewed-by: Sunny Zhu <sunnyzhyy@qq.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 181a63667adf16c35b57e446def3e41c70f1fea6)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 24 ++++++++++++++++--------
tests/qemu-iotests/194 | 6 ++++--
tests/qemu-iotests/194.out | 4 ++--
tests/qemu-iotests/iotests.py | 12 +++++++++++-
4 files changed, 33 insertions(+), 13 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index d04db85883..bca99ec206 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -848,23 +848,31 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
target_bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP &&
bdrv_can_write_zeroes_with_unmap(target_bs);
+ /* Determine if the image is already zero, regardless of sync mode. */
bdrv_graph_co_rdlock();
bs = s->mirror_top_bs->backing->bs;
+ if (s->target_is_zero) {
+ ret = 1;
+ } else {
+ ret = bdrv_co_is_all_zeroes(target_bs);
+ }
bdrv_graph_co_rdunlock();
- if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
+ /* Determine if a pre-zeroing pass is necessary. */
+ if (ret < 0) {
+ return ret;
+ } else if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
/* In TOP mode, there is no benefit to a pre-zeroing pass. */
- } else if (!s->target_is_zero || punch_holes) {
+ } else if (ret == 0 || punch_holes) {
/*
* Here, we are in FULL mode; our goal is to avoid writing
* zeroes if the destination already reads as zero, except
* when we are trying to punch holes. This is possible if
- * zeroing happened externally (s->target_is_zero) or if we
- * have a fast way to pre-zero the image (the dirty bitmap
- * will be populated later by the non-zero portions, the same
- * as for TOP mode). If pre-zeroing is not fast, or we need
- * to punch holes, then our only recourse is to write the
- * entire image.
+ * zeroing happened externally (ret > 0) or if we have a fast
+ * way to pre-zero the image (the dirty bitmap will be
+ * populated later by the non-zero portions, the same as for
+ * TOP mode). If pre-zeroing is not fast, or we need to punch
+ * holes, then our only recourse is to write the entire image.
*/
if (!bdrv_can_write_zeroes_with_unmap(target_bs)) {
bdrv_set_dirty_bitmap(s->dirty_bitmap, 0, s->bdev_length);
diff --git a/tests/qemu-iotests/194 b/tests/qemu-iotests/194
index d0b9c084f5..e114c0b269 100755
--- a/tests/qemu-iotests/194
+++ b/tests/qemu-iotests/194
@@ -62,7 +62,8 @@ with iotests.FilePath('source.img') as source_img_path, \
iotests.log('Waiting for `drive-mirror` to complete...')
iotests.log(source_vm.event_wait('BLOCK_JOB_READY'),
- filters=[iotests.filter_qmp_event])
+ filters=[iotests.filter_qmp_event,
+ iotests.filter_block_job])
iotests.log('Starting migration...')
capabilities = [{'capability': 'events', 'state': True},
@@ -88,7 +89,8 @@ with iotests.FilePath('source.img') as source_img_path, \
while True:
event2 = source_vm.event_wait('BLOCK_JOB_COMPLETED')
- iotests.log(event2, filters=[iotests.filter_qmp_event])
+ iotests.log(event2, filters=[iotests.filter_qmp_event,
+ iotests.filter_block_job])
if event2['event'] == 'BLOCK_JOB_COMPLETED':
iotests.log('Stopping the NBD server on destination...')
iotests.log(dest_vm.qmp('nbd-server-stop'))
diff --git a/tests/qemu-iotests/194.out b/tests/qemu-iotests/194.out
index 6940e809cd..d02655a514 100644
--- a/tests/qemu-iotests/194.out
+++ b/tests/qemu-iotests/194.out
@@ -7,7 +7,7 @@ Launching NBD server on destination...
Starting `drive-mirror` on source...
{"return": {}}
Waiting for `drive-mirror` to complete...
-{"data": {"device": "mirror-job0", "len": 1073741824, "offset": 1073741824, "speed": 0, "type": "mirror"}, "event": "BLOCK_JOB_READY", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}}
+{"data": {"device": "mirror-job0", "len": "LEN", "offset": "OFFSET", "speed": 0, "type": "mirror"}, "event": "BLOCK_JOB_READY", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}}
Starting migration...
{"return": {}}
{"execute": "migrate-start-postcopy", "arguments": {}}
@@ -18,7 +18,7 @@ Starting migration...
{"data": {"status": "completed"}, "event": "MIGRATION", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}}
Gracefully ending the `drive-mirror` job on source...
{"return": {}}
-{"data": {"device": "mirror-job0", "len": 1073741824, "offset": 1073741824, "speed": 0, "type": "mirror"}, "event": "BLOCK_JOB_COMPLETED", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}}
+{"data": {"device": "mirror-job0", "len": "LEN", "offset": "OFFSET", "speed": 0, "type": "mirror"}, "event": "BLOCK_JOB_COMPLETED", "timestamp": {"microseconds": "USECS", "seconds": "SECS"}}
Stopping the NBD server on destination...
{"return": {}}
Wait for migration completion on target...
diff --git a/tests/qemu-iotests/iotests.py b/tests/qemu-iotests/iotests.py
index 7292c8b342..05274772ce 100644
--- a/tests/qemu-iotests/iotests.py
+++ b/tests/qemu-iotests/iotests.py
@@ -601,13 +601,23 @@ def filter_chown(msg):
return chown_re.sub("chown UID:GID", msg)
def filter_qmp_event(event):
- '''Filter a QMP event dict'''
+ '''Filter the timestamp of a QMP event dict'''
event = dict(event)
if 'timestamp' in event:
event['timestamp']['seconds'] = 'SECS'
event['timestamp']['microseconds'] = 'USECS'
return event
+def filter_block_job(event):
+ '''Filter the offset and length of a QMP block job event dict'''
+ event = dict(event)
+ if 'data' in event:
+ if 'offset' in event['data']:
+ event['data']['offset'] = 'OFFSET'
+ if 'len' in event['data']:
+ event['data']['len'] = 'LEN'
+ return event
+
def filter_qmp(qmsg, filter_fn):
'''Given a string filter, filter a QMP object's values.
filter_fn takes a (key, value) pair.'''
--
2.39.3

355
kvm-mirror-Skip-writing-zeroes-when-target-is-already-ze.patch Normal file
View File

@ -0,0 +1,355 @@
From cc72e6ec30fb113b82fcdb61f79a0fae18e31e79 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:28 -0500
Subject: [PATCH 11/14] mirror: Skip writing zeroes when target is already zero
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [11/14] 82df7fcf94606e6b6570469d3c05666da5039408 (ebblake/centos-qemu-kvm)
When mirroring, the goal is to ensure that the destination reads the
same as the source; this goal is met whether the destination is sparse
or fully-allocated (except when explicitly punching holes, then merely
reading zero is not enough to know if it is sparse, so we still want
to punch the hole). Avoiding a redundant write to zero (whether in
the background because the zero cluster was marked in the dirty
bitmap, or in the foreground because the guest is writing zeroes) when
the destination already reads as zero makes mirroring faster, and
avoids allocating the destination merely because the source reports as
allocated.
The effect is especially pronounced when the source is a raw file.
That's because when the source is a qcow2 file, the dirty bitmap only
visits the portions of the source that are allocated, which tend to be
non-zero. But when the source is a raw file,
bdrv_co_is_allocated_above() reports the entire file as allocated so
mirror_dirty_init sets the entire dirty bitmap, and it is only later
during mirror_iteration that we change to consulting the more precise
bdrv_co_block_status_above() to learn where the source reads as zero.
Remember that since a mirror operation can write a cluster more than
once (every time the guest changes the source, the destination is also
changed to keep up), and the guest can change whether a given cluster
reads as zero, is discarded, or has non-zero data over the course of
the mirror operation, we can't take the shortcut of relying on
s->target_is_zero (which is static for the life of the job) in
mirror_co_zero() to see if the destination is already zero, because
that information may be stale. Any solution we use must be dynamic in
the face of the guest writing or discarding a cluster while the mirror
has been ongoing.
We could just teach mirror_co_zero() to do a block_status() probe of
the destination, and skip the zeroes if the destination already reads
as zero, but we know from past experience that extra block_status()
calls are not always cheap (tmpfs, anyone?), especially when they are
random access rather than linear. Use of block_status() of the source
by the background task in a linear fashion is not our bottleneck (it's
a background task, after all); but since mirroring can be done while
the source is actively being changed, we don't want a slow
block_status() of the destination to occur on the hot path of the
guest trying to do random-access writes to the source.
So this patch takes a slightly different approach: any time we have to
track dirty clusters, we can also track which clusters are known to
read as zero. For sync=TOP or when we are punching holes from
"detect-zeroes":"unmap", the zero bitmap starts out empty, but
prevents a second write zero to a cluster that was already zero by an
earlier pass; for sync=FULL when we are not punching holes, the zero
bitmap starts out full if the destination reads as zero during
initialization. Either way, I/O to the destination can now avoid
redundant write zero to a cluster that already reads as zero, all
without having to do a block_status() per write on the destination.
With this patch, if I create a raw sparse destination file, connect it
with QMP 'blockdev-add' while leaving it at the default "discard":
"ignore", then run QMP 'blockdev-mirror' with "sync": "full", the
destination remains sparse rather than fully allocated. Meanwhile, a
destination image that is already fully allocated remains so unless it
was opened with "detect-zeroes": "unmap". And any time writing zeroes
is skipped, the job counters are not incremented.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250509204341.3553601-26-eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 7e277545b90874171128804e256a538fb0e8dd7e)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
block/mirror.c | 107 ++++++++++++++++++++++++++++++++++++++++++-------
1 file changed, 93 insertions(+), 14 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index bca99ec206..724318f037 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -73,6 +73,7 @@ typedef struct MirrorBlockJob {
size_t buf_size;
int64_t bdev_length;
unsigned long *cow_bitmap;
+ unsigned long *zero_bitmap;
BdrvDirtyBitmap *dirty_bitmap;
BdrvDirtyBitmapIter *dbi;
uint8_t *buf;
@@ -108,9 +109,12 @@ struct MirrorOp {
int64_t offset;
uint64_t bytes;
- /* The pointee is set by mirror_co_read(), mirror_co_zero(), and
- * mirror_co_discard() before yielding for the first time */
+ /*
+ * These pointers are set by mirror_co_read(), mirror_co_zero(), and
+ * mirror_co_discard() before yielding for the first time
+ */
int64_t *bytes_handled;
+ bool *io_skipped;
bool is_pseudo_op;
bool is_active_write;
@@ -408,15 +412,34 @@ static void coroutine_fn mirror_co_read(void *opaque)
static void coroutine_fn mirror_co_zero(void *opaque)
{
MirrorOp *op = opaque;
- int ret;
+ bool write_needed = true;
+ int ret = 0;
op->s->in_flight++;
op->s->bytes_in_flight += op->bytes;
*op->bytes_handled = op->bytes;
op->is_in_flight = true;
- ret = blk_co_pwrite_zeroes(op->s->target, op->offset, op->bytes,
- op->s->unmap ? BDRV_REQ_MAY_UNMAP : 0);
+ if (op->s->zero_bitmap) {
+ unsigned long end = DIV_ROUND_UP(op->offset + op->bytes,
+ op->s->granularity);
+ assert(QEMU_IS_ALIGNED(op->offset, op->s->granularity));
+ assert(QEMU_IS_ALIGNED(op->bytes, op->s->granularity) ||
+ op->offset + op->bytes == op->s->bdev_length);
+ if (find_next_zero_bit(op->s->zero_bitmap, end,
+ op->offset / op->s->granularity) == end) {
+ write_needed = false;
+ *op->io_skipped = true;
+ }
+ }
+ if (write_needed) {
+ ret = blk_co_pwrite_zeroes(op->s->target, op->offset, op->bytes,
+ op->s->unmap ? BDRV_REQ_MAY_UNMAP : 0);
+ }
+ if (ret >= 0 && op->s->zero_bitmap) {
+ bitmap_set(op->s->zero_bitmap, op->offset / op->s->granularity,
+ DIV_ROUND_UP(op->bytes, op->s->granularity));
+ }
mirror_write_complete(op, ret);
}
@@ -435,29 +458,43 @@ static void coroutine_fn mirror_co_discard(void *opaque)
}
static unsigned mirror_perform(MirrorBlockJob *s, int64_t offset,
- unsigned bytes, MirrorMethod mirror_method)
+ unsigned bytes, MirrorMethod mirror_method,
+ bool *io_skipped)
{
MirrorOp *op;
Coroutine *co;
int64_t bytes_handled = -1;
+ assert(QEMU_IS_ALIGNED(offset, s->granularity));
+ assert(QEMU_IS_ALIGNED(bytes, s->granularity) ||
+ offset + bytes == s->bdev_length);
op = g_new(MirrorOp, 1);
*op = (MirrorOp){
.s = s,
.offset = offset,
.bytes = bytes,
.bytes_handled = &bytes_handled,
+ .io_skipped = io_skipped,
};
qemu_co_queue_init(&op->waiting_requests);
switch (mirror_method) {
case MIRROR_METHOD_COPY:
+ if (s->zero_bitmap) {
+ bitmap_clear(s->zero_bitmap, offset / s->granularity,
+ DIV_ROUND_UP(bytes, s->granularity));
+ }
co = qemu_coroutine_create(mirror_co_read, op);
break;
case MIRROR_METHOD_ZERO:
+ /* s->zero_bitmap handled in mirror_co_zero */
co = qemu_coroutine_create(mirror_co_zero, op);
break;
case MIRROR_METHOD_DISCARD:
+ if (s->zero_bitmap) {
+ bitmap_clear(s->zero_bitmap, offset / s->granularity,
+ DIV_ROUND_UP(bytes, s->granularity));
+ }
co = qemu_coroutine_create(mirror_co_discard, op);
break;
default:
@@ -568,6 +605,7 @@ static void coroutine_fn GRAPH_UNLOCKED mirror_iteration(MirrorBlockJob *s)
int ret = -1;
int64_t io_bytes;
int64_t io_bytes_acct;
+ bool io_skipped = false;
MirrorMethod mirror_method = MIRROR_METHOD_COPY;
assert(!(offset % s->granularity));
@@ -611,8 +649,10 @@ static void coroutine_fn GRAPH_UNLOCKED mirror_iteration(MirrorBlockJob *s)
}
io_bytes = mirror_clip_bytes(s, offset, io_bytes);
- io_bytes = mirror_perform(s, offset, io_bytes, mirror_method);
- if (mirror_method != MIRROR_METHOD_COPY && write_zeroes_ok) {
+ io_bytes = mirror_perform(s, offset, io_bytes, mirror_method,
+ &io_skipped);
+ if (io_skipped ||
+ (mirror_method != MIRROR_METHOD_COPY && write_zeroes_ok)) {
io_bytes_acct = 0;
} else {
io_bytes_acct = io_bytes;
@@ -847,8 +887,10 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
bool punch_holes =
target_bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP &&
bdrv_can_write_zeroes_with_unmap(target_bs);
+ int64_t bitmap_length = DIV_ROUND_UP(s->bdev_length, s->granularity);
/* Determine if the image is already zero, regardless of sync mode. */
+ s->zero_bitmap = bitmap_new(bitmap_length);
bdrv_graph_co_rdlock();
bs = s->mirror_top_bs->backing->bs;
if (s->target_is_zero) {
@@ -862,7 +904,14 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
if (ret < 0) {
return ret;
} else if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
- /* In TOP mode, there is no benefit to a pre-zeroing pass. */
+ /*
+ * In TOP mode, there is no benefit to a pre-zeroing pass, but
+ * the zero bitmap can be set if the destination already reads
+ * as zero and we are not punching holes.
+ */
+ if (ret > 0 && !punch_holes) {
+ bitmap_set(s->zero_bitmap, 0, bitmap_length);
+ }
} else if (ret == 0 || punch_holes) {
/*
* Here, we are in FULL mode; our goal is to avoid writing
@@ -871,8 +920,9 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
* zeroing happened externally (ret > 0) or if we have a fast
* way to pre-zero the image (the dirty bitmap will be
* populated later by the non-zero portions, the same as for
- * TOP mode). If pre-zeroing is not fast, or we need to punch
- * holes, then our only recourse is to write the entire image.
+ * TOP mode). If pre-zeroing is not fast, then our only
+ * recourse is to mark the entire image dirty. The act of
+ * pre-zeroing will populate the zero bitmap.
*/
if (!bdrv_can_write_zeroes_with_unmap(target_bs)) {
bdrv_set_dirty_bitmap(s->dirty_bitmap, 0, s->bdev_length);
@@ -883,6 +933,7 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
for (offset = 0; offset < s->bdev_length; ) {
int bytes = MIN(s->bdev_length - offset,
QEMU_ALIGN_DOWN(INT_MAX, s->granularity));
+ bool ignored;
mirror_throttle(s);
@@ -898,12 +949,15 @@ static int coroutine_fn GRAPH_UNLOCKED mirror_dirty_init(MirrorBlockJob *s)
continue;
}
- mirror_perform(s, offset, bytes, MIRROR_METHOD_ZERO);
+ mirror_perform(s, offset, bytes, MIRROR_METHOD_ZERO, &ignored);
offset += bytes;
}
mirror_wait_for_all_io(s);
s->initial_zeroing_ongoing = false;
+ } else {
+ /* In FULL mode, and image already reads as zero. */
+ bitmap_set(s->zero_bitmap, 0, bitmap_length);
}
/* First part, loop on the sectors and initialize the dirty bitmap. */
@@ -1188,6 +1242,7 @@ immediate_exit:
assert(s->in_flight == 0);
qemu_vfree(s->buf);
g_free(s->cow_bitmap);
+ g_free(s->zero_bitmap);
g_free(s->in_flight_bitmap);
bdrv_dirty_iter_free(s->dbi);
@@ -1367,6 +1422,7 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
int ret;
size_t qiov_offset = 0;
int64_t dirty_bitmap_offset, dirty_bitmap_end;
+ int64_t zero_bitmap_offset, zero_bitmap_end;
if (!QEMU_IS_ALIGNED(offset, job->granularity) &&
bdrv_dirty_bitmap_get(job->dirty_bitmap, offset))
@@ -1410,8 +1466,9 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
}
/*
- * Tails are either clean or shrunk, so for bitmap resetting
- * we safely align the range down.
+ * Tails are either clean or shrunk, so for dirty bitmap resetting
+ * we safely align the range narrower. But for zero bitmap, round
+ * range wider for checking or clearing, and narrower for setting.
*/
dirty_bitmap_offset = QEMU_ALIGN_UP(offset, job->granularity);
dirty_bitmap_end = QEMU_ALIGN_DOWN(offset + bytes, job->granularity);
@@ -1419,22 +1476,44 @@ do_sync_target_write(MirrorBlockJob *job, MirrorMethod method,
bdrv_reset_dirty_bitmap(job->dirty_bitmap, dirty_bitmap_offset,
dirty_bitmap_end - dirty_bitmap_offset);
}
+ zero_bitmap_offset = offset / job->granularity;
+ zero_bitmap_end = DIV_ROUND_UP(offset + bytes, job->granularity);
job_progress_increase_remaining(&job->common.job, bytes);
job->active_write_bytes_in_flight += bytes;
switch (method) {
case MIRROR_METHOD_COPY:
+ if (job->zero_bitmap) {
+ bitmap_clear(job->zero_bitmap, zero_bitmap_offset,
+ zero_bitmap_end - zero_bitmap_offset);
+ }
ret = blk_co_pwritev_part(job->target, offset, bytes,
qiov, qiov_offset, flags);
break;
case MIRROR_METHOD_ZERO:
+ if (job->zero_bitmap) {
+ if (find_next_zero_bit(job->zero_bitmap, zero_bitmap_end,
+ zero_bitmap_offset) == zero_bitmap_end) {
+ ret = 0;
+ break;
+ }
+ }
assert(!qiov);
ret = blk_co_pwrite_zeroes(job->target, offset, bytes, flags);
+ if (job->zero_bitmap && ret >= 0) {
+ bitmap_set(job->zero_bitmap, dirty_bitmap_offset / job->granularity,
+ (dirty_bitmap_end - dirty_bitmap_offset) /
+ job->granularity);
+ }
break;
case MIRROR_METHOD_DISCARD:
+ if (job->zero_bitmap) {
+ bitmap_clear(job->zero_bitmap, zero_bitmap_offset,
+ zero_bitmap_end - zero_bitmap_offset);
+ }
assert(!qiov);
ret = blk_co_pdiscard(job->target, offset, bytes);
break;
--
2.39.3

545
kvm-tests-Add-iotest-mirror-sparse-for-recent-patches.patch Normal file
View File

@ -0,0 +1,545 @@
From 2bb881df5b93f5534e5f0b91cf1ed3e0b524c2d3 Mon Sep 17 00:00:00 2001
From: Eric Blake <eblake@redhat.com>
Date: Fri, 9 May 2025 15:40:30 -0500
Subject: [PATCH 13/14] tests: Add iotest mirror-sparse for recent patches
RH-Author: Eric Blake <eblake@redhat.com>
RH-MergeRequest: 363: blockdev-mirror: More efficient handling of sparse mirrors
RH-Jira: RHEL-88435 RHEL-88437
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [13/14] 474f12dfe9161c7e9f59cafde203e5183e2fc3f5 (ebblake/centos-qemu-kvm)
Prove that blockdev-mirror can now result in sparse raw destination
files, regardless of whether the source is raw or qcow2. By making
this a separate test, it was possible to test effects of individual
patches for the various pieces that all have to work together for a
sparse mirror to be successful.
Note that ./check -file produces different job lengths than ./check
-qcow2 (the test uses a filter to normalize); that's because when
deciding how much of the image to be mirrored, the code looks at how
much of the source image was allocated (for qcow2, this is only the
written clusters; for raw, it is the entire file). But the important
part is that the destination file ends up smaller than 3M, rather than
the 20M it used to be before this patch series.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250509204341.3553601-28-eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit c0ddcb2cbc146e64f666eaae4edc7b5db7e5814d)
Jira: https://issues.redhat.com/browse/RHEL-88435
Jira: https://issues.redhat.com/browse/RHEL-88437
Signed-off-by: Eric Blake <eblake@redhat.com>
---
tests/qemu-iotests/tests/mirror-sparse | 125 +++++++
tests/qemu-iotests/tests/mirror-sparse.out | 365 +++++++++++++++++++++
2 files changed, 490 insertions(+)
create mode 100755 tests/qemu-iotests/tests/mirror-sparse
create mode 100644 tests/qemu-iotests/tests/mirror-sparse.out
diff --git a/tests/qemu-iotests/tests/mirror-sparse b/tests/qemu-iotests/tests/mirror-sparse
new file mode 100755
index 0000000000..8c52a4e244
--- /dev/null
+++ b/tests/qemu-iotests/tests/mirror-sparse
@@ -0,0 +1,125 @@
+#!/usr/bin/env bash
+# group: rw auto quick
+#
+# Test blockdev-mirror with raw sparse destination
+#
+# Copyright (C) 2025 Red Hat, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+#
+
+seq="$(basename $0)"
+echo "QA output created by $seq"
+
+status=1 # failure is the default!
+
+_cleanup()
+{
+ _cleanup_test_img
+ _cleanup_qemu
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+cd ..
+. ./common.rc
+. ./common.filter
+. ./common.qemu
+
+_supported_fmt qcow2 raw # Format of the source. dst is always raw file
+_supported_proto file
+_supported_os Linux
+
+echo
+echo "=== Initial image setup ==="
+echo
+
+TEST_IMG="$TEST_IMG.base" _make_test_img 20M
+$QEMU_IO -c 'w 8M 2M' -f $IMGFMT "$TEST_IMG.base" | _filter_qemu_io
+
+_launch_qemu \
+ -blockdev '{"driver":"file", "cache":{"direct":true, "no-flush":false},
+ "filename":"'"$TEST_IMG.base"'", "node-name":"src-file"}' \
+ -blockdev '{"driver":"'$IMGFMT'", "node-name":"src", "file":"src-file"}'
+h1=$QEMU_HANDLE
+_send_qemu_cmd $h1 '{"execute": "qmp_capabilities"}' 'return'
+
+# Check several combinations; most should result in a sparse destination;
+# the destination should only be fully allocated if pre-allocated
+# and not punching holes due to detect-zeroes
+# do_test creation discard zeroes result
+do_test() {
+ creation=$1
+ discard=$2
+ zeroes=$3
+ expected=$4
+
+echo
+echo "=== Testing creation=$creation discard=$discard zeroes=$zeroes ==="
+echo
+
+rm -f $TEST_IMG
+if test $creation = external; then
+ truncate --size=20M $TEST_IMG
+else
+ _send_qemu_cmd $h1 '{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"'$TEST_IMG'",
+ "size":'$((20*1024*1024))', "preallocation":"'$creation'"},
+ "job-id":"job1"}}' 'concluded'
+ _send_qemu_cmd $h1 '{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}' 'return'
+fi
+_send_qemu_cmd $h1 '{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"'$TEST_IMG'", "aio":"threads",
+ "auto-read-only":true, "discard":"'$discard'",
+ "detect-zeroes":"'$zeroes'"}}' 'return'
+_send_qemu_cmd $h1 '{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}' 'return'
+_timed_wait_for $h1 '"ready"'
+_send_qemu_cmd $h1 '{"execute": "job-complete", "arguments":
+ {"id":"job2"}}' 'return' \
+ | _filter_block_job_offset | _filter_block_job_len
+_send_qemu_cmd $h1 '{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}' 'return' \
+ | _filter_block_job_offset | _filter_block_job_len
+$QEMU_IMG compare -U -f $IMGFMT -F raw $TEST_IMG.base $TEST_IMG
+result=$(disk_usage $TEST_IMG)
+if test $result -lt $((3*1024*1024)); then
+ actual=sparse
+elif test $result = $((20*1024*1024)); then
+ actual=full
+else
+ actual=unknown
+fi
+echo "Destination is $actual; expected $expected"
+}
+
+do_test external ignore off sparse
+do_test external unmap off sparse
+do_test external unmap unmap sparse
+do_test off ignore off sparse
+do_test off unmap off sparse
+do_test off unmap unmap sparse
+do_test full ignore off full
+do_test full unmap off sparse
+do_test full unmap unmap sparse
+
+_send_qemu_cmd $h1 '{"execute":"quit"}' ''
+
+# success, all done
+echo '*** done'
+rm -f $seq.full
+status=0
diff --git a/tests/qemu-iotests/tests/mirror-sparse.out b/tests/qemu-iotests/tests/mirror-sparse.out
new file mode 100644
index 0000000000..2103b891c3
--- /dev/null
+++ b/tests/qemu-iotests/tests/mirror-sparse.out
@@ -0,0 +1,365 @@
+QA output created by mirror-sparse
+
+=== Initial image setup ===
+
+Formatting 'TEST_DIR/t.IMGFMT.base', fmt=IMGFMT size=20971520
+wrote 2097152/2097152 bytes at offset 8388608
+2 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
+{"execute": "qmp_capabilities"}
+{"return": {}}
+
+=== Testing creation=external discard=ignore zeroes=off ===
+
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"ignore",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=external discard=unmap zeroes=off ===
+
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=external discard=unmap zeroes=unmap ===
+
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"unmap"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=off discard=ignore zeroes=off ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"off"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"ignore",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=off discard=unmap zeroes=off ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"off"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=off discard=unmap zeroes=unmap ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"off"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"unmap"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=full discard=ignore zeroes=off ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"full"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"ignore",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is full; expected full
+
+=== Testing creation=full discard=unmap zeroes=off ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"full"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"off"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+
+=== Testing creation=full discard=unmap zeroes=unmap ===
+
+{"execute": "blockdev-create", "arguments":
+ {"options": {"driver":"file", "filename":"TEST_DIR/t.IMGFMT",
+ "size":20971520, "preallocation":"full"},
+ "job-id":"job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job1"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job1"}}
+{"execute": "job-dismiss", "arguments":
+ {"id": "job1"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job1"}}
+{"return": {}}
+{"execute": "blockdev-add", "arguments":
+ {"node-name": "dst", "driver":"file",
+ "filename":"TEST_DIR/t.IMGFMT", "aio":"threads",
+ "auto-read-only":true, "discard":"unmap",
+ "detect-zeroes":"unmap"}}
+{"return": {}}
+{"execute":"blockdev-mirror", "arguments":
+ {"sync":"full", "device":"src", "target":"dst",
+ "job-id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "created", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "running", "id": "job2"}}
+{"return": {}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "ready", "id": "job2"}}
+{"execute": "job-complete", "arguments":
+ {"id":"job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_READY", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"return": {}}
+{"execute": "blockdev-del", "arguments":
+ {"node-name": "dst"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "waiting", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "pending", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "BLOCK_JOB_COMPLETED", "data": {"device": "job2", "len": LEN, "offset": OFFSET, "speed": 0, "type": "mirror"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "concluded", "id": "job2"}}
+{"timestamp": {"seconds": TIMESTAMP, "microseconds": TIMESTAMP}, "event": "JOB_STATUS_CHANGE", "data": {"status": "null", "id": "job2"}}
+{"return": {}}
+Images are identical.
+Destination is sparse; expected sparse
+{"execute":"quit"}
+*** done
--
2.39.3

64
qemu-kvm.spec
View File

@ -147,7 +147,7 @@ Obsoletes: %{name}-block-ssh <= %{epoch}:%{version} \
Summary: QEMU is a machine emulator and virtualizer
Name: qemu-kvm
Version: 10.0.0
Release: 3%{?rcrel}%{?dist}%{?cc_suffix}
Release: 4%{?rcrel}%{?dist}%{?cc_suffix}
# Epoch because we pushed a qemu-1.0 package. AIUI this can't ever be dropped
# Epoch 15 used for RHEL 8
# Epoch 17 used for RHEL 9 (due to release versioning offset in RHEL 8.5)
@ -202,6 +202,48 @@ Patch24: kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch
Patch25: kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch
# Fixing s390x build issues
Patch26: kvm-docs-Don-t-define-duplicate-label-in-qemu-block-driv.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch27: kvm-block-Expand-block-status-mode-from-bool-to-flags.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch28: kvm-file-posix-gluster-Handle-zero-block-status-hint-bet.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch29: kvm-block-Let-bdrv_co_is_zero_fast-consolidate-adjacent-.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch30: kvm-block-Add-new-bdrv_co_is_all_zeroes-function.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch31: kvm-iotests-Improve-iotest-194-to-mirror-data.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch32: kvm-mirror-Minor-refactoring.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch33: kvm-mirror-Pass-full-sync-mode-rather-than-bool-to-inter.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch34: kvm-mirror-Allow-QMP-override-to-declare-target-already-.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch35: kvm-mirror-Drop-redundant-zero_target-parameter.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch36: kvm-mirror-Skip-pre-zeroing-destination-if-it-is-already.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch37: kvm-mirror-Skip-writing-zeroes-when-target-is-already-ze.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch38: kvm-iotests-common.rc-add-disk_usage-function.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch39: kvm-tests-Add-iotest-mirror-sparse-for-recent-patches.patch
# For RHEL-88435 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1]
# For RHEL-88437 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1]
Patch40: kvm-mirror-Reduce-I-O-when-destination-is-detect-zeroes-.patch
%if %{have_clang}
BuildRequires: clang
@ -1284,6 +1326,26 @@ useradd -r -u 107 -g qemu -G kvm -d / -s /sbin/nologin \
%endif
%changelog
* Mon May 26 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-4
- kvm-block-Expand-block-status-mode-from-bool-to-flags.patch [RHEL-88435 RHEL-88437]
- kvm-file-posix-gluster-Handle-zero-block-status-hint-bet.patch [RHEL-88435 RHEL-88437]
- kvm-block-Let-bdrv_co_is_zero_fast-consolidate-adjacent-.patch [RHEL-88435 RHEL-88437]
- kvm-block-Add-new-bdrv_co_is_all_zeroes-function.patch [RHEL-88435 RHEL-88437]
- kvm-iotests-Improve-iotest-194-to-mirror-data.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Minor-refactoring.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Pass-full-sync-mode-rather-than-bool-to-inter.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Allow-QMP-override-to-declare-target-already-.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Drop-redundant-zero_target-parameter.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Skip-pre-zeroing-destination-if-it-is-already.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Skip-writing-zeroes-when-target-is-already-ze.patch [RHEL-88435 RHEL-88437]
- kvm-iotests-common.rc-add-disk_usage-function.patch [RHEL-88435 RHEL-88437]
- kvm-tests-Add-iotest-mirror-sparse-for-recent-patches.patch [RHEL-88435 RHEL-88437]
- kvm-mirror-Reduce-I-O-when-destination-is-detect-zeroes-.patch [RHEL-88435 RHEL-88437]
- Resolves: RHEL-88435
(--migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-10.1])
- Resolves: RHEL-88437
(Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-10.1])
* Mon May 19 2025 Miroslav Rezanina <mrezanin@redhat.com> - 10.0.0-3
- kvm-migration-postcopy-Spatial-locality-page-hint-for-pr.patch [RHEL-85635]
- kvm-meson-configure-add-valgrind-option-en-dis-able-valg.patch [RHEL-88457]