c/pve-qemu-qoup
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update submodule and patches to QEMU 8.2.2

Browse Source 
This version includes both the AioContext lock and the block graph
lock, so there might be some deadlocks lurking. It's not possible to
disable the block graph lock like was done in QEMU 8.1, because there
are no changes like the function bdrv_schedule_unref() that require
it. QEMU 9.0 will finally get rid of the AioContext locking.

During live-restore with a VirtIO SCSI drive with iothread there is a
known racy deadlock related to the AioContext lock. Not new [1], but
not sure if more likely now. Should be fixed in QEMU 9.0.

The block graph lock comes with annotations that can be checked by
clang's TSA. This required changes to the block drivers, i.e.
alloc-track, pbs, zeroinit as well as taking the appropriate locks
in pve-backup, savevm-async, vma-reader.

Local variable shadowing is prohibited via a compiler flag now,
required slight adaptation in vma.c.

Major changes only affect alloc-track:

* It is not possible to call a generated co-wrapper like
  bdrv_get_info() while holding the block graph lock exclusively [0],
  which does happen during initialization of alloc-track when the
  backing hd is set and the refresh_limits driver callback is invoked.

  The bdrv_get_info() call to get the cluster size is moved to
  directly after opening the file child in track_open().

  The important thing is that at least the request alignment for the
  write target is used, because then the RMW cycle in bdrv_pwritev
  will gather enough data from the backing file. Partial cluster
  allocations in the target are not a fundamental issue, because the
  driver returns its allocation status based on the bitmap, so any
  other data that maps to the same cluster will still be copied later
  by a stream job (or during writes to that cluster).

* Replacing the node cannot be done in the
  track_co_change_backing_file() callback, because it is a coroutine
  and cannot hold the block graph lock exclusively. So it is moved to
  the stream job itself with the auto-remove option not having an
  effect anymore (qemu-server would always set it anyways).

  In the future, there could either be a special option for the stream
  job, or maybe the upcoming blockdev-replace QMP command can be used.

  Replacing the backing child is actually already done in the stream
  job, so no need to do it in the track_co_change_backing_file()
  callback. It also cannot be called from a coroutine. Looking at the
  implementation in the qcow2 driver, it doesn't seem to be intended
  to change the backing child itself, just update driver-internal
  state.

Other changes:

* alloc-track: Error out early when used without auto-remove. Since
  replacing the node now happens in the stream job, where the option
  cannot be read from (it's internal to the driver), it will always be
  treated as 'on'. Makes sure to have users beside qemu-server notice
  the change (should they even exist). The option can be fully dropped
  in the future while adding a version guard in qemu-server.

* alloc-track: Avoid seemingly superfluous child permission update.
  Doesn't seem necessary nowadays (maybe after commit "alloc-track:
  fix deadlock during drop" where the dropping is not rescheduled and
  delayed anymore or some upstream change). Replacing the block node
  will already update the permissions of the new node (which was the
  file child before). Should there really be some issue, instead of
  having a drop state, this could also be just based off the fact
  whether there is still a backing child.

  Dumping the cumulative (shared) permissions for the BDS with a debug
  print yields the same values after this patch and with QEMU 8.1,
  namely 3 and 5.

* PBS block driver: compile unconditionally. Proxmox VE always needs
  it and something in the build process changed to make it not enabled
  by default. Probably would need to move the build option to meson
  otherwise.

* backup: job unreferencing during cleanup needs to happen outside of
  coroutine, so it was moved to before invoking the clean

* mirror: Cherry-pick stable fix to avoid potential deadlock.

* savevm-async: migrate_init now can fail, so propagate potential
  error.

* savevm-async: compression counters are not accessible outside
  migration/ram-compress now, so drop code that prophylactically set
  it to zero.

[0]: https://lore.kernel.org/qemu-devel/220be383-3b0d-4938-b584-69ad214e5d5d@proxmox.com/
[1]: https://lore.kernel.org/qemu-devel/e13b488e-bf13-44f2-acca-e724d14f43fd@proxmox.com/

Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
This commit is contained in:
Fiona Ebner 2024-04-25 17:21:28 +02:00 committed by Thomas Lamprecht
parent 2e71c17f5b
commit f1eed34ac7
73 changed files with 901 additions and 1576 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

88
debian/patches/bitmap-mirror/0001-drive-mirror-add-support-for-sync-bitmap-mode-never.patch vendored
View File

@ -27,18 +27,18 @@ Signed-off-by: Ma Haocong <mahaocong@didichuxing.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
[FE: rebased for 8.1.1]
[FE: rebased for 8.2.2]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/mirror.c | 98 +++++++++++++++++++++-----
block/mirror.c | 99 ++++++++++++++++++++------
blockdev.c | 38 +++++++++-
include/block/block_int-global-state.h | 4 +-
qapi/block-core.json | 25 ++++++-
tests/unit/test-block-iothread.c | 4 +-
5 files changed, 142 insertions(+), 27 deletions(-)
5 files changed, 142 insertions(+), 28 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index d3cacd1708..1ff42c8af1 100644
index abbddb39e4..ed14c8b498 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -51,7 +51,7 @@ typedef struct MirrorBlockJob {
@ -50,7 +50,7 @@ index d3cacd1708..1ff42c8af1 100644
BlockMirrorBackingMode backing_mode;
/* Whether the target image requires explicit zero-initialization */
bool zero_target;
@@ -65,6 +65,8 @@ typedef struct MirrorBlockJob {
@@ -73,6 +73,8 @@ typedef struct MirrorBlockJob {
size_t buf_size;
int64_t bdev_length;
unsigned long *cow_bitmap;
@ -59,9 +59,9 @@ index d3cacd1708..1ff42c8af1 100644
BdrvDirtyBitmap *dirty_bitmap;
BdrvDirtyBitmapIter *dbi;
uint8_t *buf;
@@ -705,7 +707,8 @@ static int mirror_exit_common(Job *job)
bdrv_child_refresh_perms(mirror_top_bs, mirror_top_bs->backing,
@@ -724,7 +726,8 @@ 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;
@ -69,7 +69,7 @@ index d3cacd1708..1ff42c8af1 100644
BlockDriverState *unfiltered_target = bdrv_skip_filters(target_bs);
if (bdrv_cow_bs(unfiltered_target) != backing) {
@@ -809,6 +812,16 @@ static void mirror_abort(Job *job)
@@ -831,6 +834,16 @@ static void mirror_abort(Job *job)
assert(ret == 0);
}
@ -86,7 +86,7 @@ index d3cacd1708..1ff42c8af1 100644
static void coroutine_fn mirror_throttle(MirrorBlockJob *s)
{
int64_t now = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
@@ -997,7 +1010,8 @@ static int coroutine_fn mirror_run(Job *job, Error **errp)
@@ -1027,7 +1040,8 @@ 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);
@ -96,7 +96,7 @@ index d3cacd1708..1ff42c8af1 100644
ret = mirror_dirty_init(s);
if (ret < 0 || job_is_cancelled(&s->common.job)) {
goto immediate_exit;
@@ -1251,6 +1265,7 @@ static const BlockJobDriver mirror_job_driver = {
@@ -1323,6 +1337,7 @@ static const BlockJobDriver mirror_job_driver = {
.run = mirror_run,
.prepare = mirror_prepare,
.abort = mirror_abort,
@ -104,7 +104,7 @@ index d3cacd1708..1ff42c8af1 100644
.pause = mirror_pause,
.complete = mirror_complete,
.cancel = mirror_cancel,
@@ -1267,6 +1282,7 @@ static const BlockJobDriver commit_active_job_driver = {
@@ -1341,6 +1356,7 @@ static const BlockJobDriver commit_active_job_driver = {
.run = mirror_run,
.prepare = mirror_prepare,
.abort = mirror_abort,
@ -112,7 +112,7 @@ index d3cacd1708..1ff42c8af1 100644
.pause = mirror_pause,
.complete = mirror_complete,
.cancel = commit_active_cancel,
@@ -1658,7 +1674,10 @@ static BlockJob *mirror_start_job(
@@ -1733,7 +1749,10 @@ static BlockJob *mirror_start_job(
BlockCompletionFunc *cb,
void *opaque,
const BlockJobDriver *driver,
@ -124,9 +124,9 @@ index d3cacd1708..1ff42c8af1 100644
bool auto_complete, const char *filter_node_name,
bool is_mirror, MirrorCopyMode copy_mode,
Error **errp)
@@ -1670,10 +1689,39 @@ static BlockJob *mirror_start_job(
uint64_t target_perms, target_shared_perms;
int ret;
@@ -1747,10 +1766,39 @@ static BlockJob *mirror_start_job(
GLOBAL_STATE_CODE();
- if (granularity == 0) {
- granularity = bdrv_get_default_bitmap_granularity(target);
@ -166,7 +166,7 @@ index d3cacd1708..1ff42c8af1 100644
assert(is_power_of_2(granularity));
if (buf_size < 0) {
@@ -1804,7 +1852,9 @@ static BlockJob *mirror_start_job(
@@ -1890,7 +1938,9 @@ static BlockJob *mirror_start_job(
s->replaces = g_strdup(replaces);
s->on_source_error = on_source_error;
s->on_target_error = on_target_error;
@ -176,10 +176,10 @@ index d3cacd1708..1ff42c8af1 100644
+ s->bitmap_mode = bitmap_mode;
s->backing_mode = backing_mode;
s->zero_target = zero_target;
s->copy_mode = copy_mode;
@@ -1825,6 +1875,18 @@ static BlockJob *mirror_start_job(
bdrv_disable_dirty_bitmap(s->dirty_bitmap);
}
qatomic_set(&s->copy_mode, copy_mode);
@@ -1916,6 +1966,18 @@ static BlockJob *mirror_start_job(
*/
bdrv_disable_dirty_bitmap(s->dirty_bitmap);
+ if (s->sync_bitmap) {
+ bdrv_dirty_bitmap_set_busy(s->sync_bitmap, true);
@ -193,10 +193,10 @@ index d3cacd1708..1ff42c8af1 100644
+ }
+ }
+
bdrv_graph_wrlock(bs);
ret = block_job_add_bdrv(&s->common, "source", bs, 0,
BLK_PERM_WRITE_UNCHANGED | BLK_PERM_WRITE |
BLK_PERM_CONSISTENT_READ,
@@ -1902,6 +1964,9 @@ fail:
@@ -1998,6 +2060,9 @@ fail:
if (s->dirty_bitmap) {
bdrv_release_dirty_bitmap(s->dirty_bitmap);
}
@ -206,7 +206,7 @@ index d3cacd1708..1ff42c8af1 100644
job_early_fail(&s->common.job);
}
@@ -1919,31 +1984,25 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
@@ -2020,35 +2085,28 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
BlockDriverState *target, const char *replaces,
int creation_flags, int64_t speed,
uint32_t granularity, int64_t buf_size,
@ -231,8 +231,12 @@ index d3cacd1708..1ff42c8af1 100644
- MirrorSyncMode_str(mode));
- return;
- }
-
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,
@ -243,7 +247,7 @@ index d3cacd1708..1ff42c8af1 100644
}
BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
@@ -1970,7 +2029,8 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
@@ -2075,7 +2133,8 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
job_id, bs, creation_flags, base, NULL, speed, 0, 0,
MIRROR_LEAVE_BACKING_CHAIN, false,
on_error, on_error, true, cb, opaque,
@ -254,10 +258,10 @@ index d3cacd1708..1ff42c8af1 100644
errp);
if (!job) {
diff --git a/blockdev.c b/blockdev.c
index c28462a633..a402fa4bf7 100644
index c91f49e7b6..8c8e8b604a 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2849,6 +2849,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -2903,6 +2903,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
BlockDriverState *target,
const char *replaces,
enum MirrorSyncMode sync,
@ -267,7 +271,7 @@ index c28462a633..a402fa4bf7 100644
BlockMirrorBackingMode backing_mode,
bool zero_target,
bool has_speed, int64_t speed,
@@ -2867,6 +2870,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -2921,6 +2924,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
{
BlockDriverState *unfiltered_bs;
int job_flags = JOB_DEFAULT;
@ -275,7 +279,7 @@ index c28462a633..a402fa4bf7 100644
GLOBAL_STATE_CODE();
GRAPH_RDLOCK_GUARD_MAINLOOP();
@@ -2921,6 +2925,29 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -2975,6 +2979,29 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
sync = MIRROR_SYNC_MODE_FULL;
}
@ -305,7 +309,7 @@ index c28462a633..a402fa4bf7 100644
if (!replaces) {
/* We want to mirror from @bs, but keep implicit filters on top */
unfiltered_bs = bdrv_skip_implicit_filters(bs);
@@ -2966,8 +2993,8 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -3030,8 +3057,8 @@ 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,
@ -316,7 +320,7 @@ index c28462a633..a402fa4bf7 100644
on_source_error, on_target_error, unmap, filter_node_name,
copy_mode, errp);
}
@@ -3115,6 +3142,8 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
@@ -3189,6 +3216,8 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
blockdev_mirror_common(arg->job_id, bs, target_bs,
arg->replaces, arg->sync,
@ -325,7 +329,7 @@ index c28462a633..a402fa4bf7 100644
backing_mode, zero_target,
arg->has_speed, arg->speed,
arg->has_granularity, arg->granularity,
@@ -3136,6 +3165,8 @@ void qmp_blockdev_mirror(const char *job_id,
@@ -3210,6 +3239,8 @@ void qmp_blockdev_mirror(const char *job_id,
const char *device, const char *target,
const char *replaces,
MirrorSyncMode sync,
@ -334,7 +338,7 @@ index c28462a633..a402fa4bf7 100644
bool has_speed, int64_t speed,
bool has_granularity, uint32_t granularity,
bool has_buf_size, int64_t buf_size,
@@ -3184,7 +3215,8 @@ void qmp_blockdev_mirror(const char *job_id,
@@ -3258,7 +3289,8 @@ void qmp_blockdev_mirror(const char *job_id,
}
blockdev_mirror_common(job_id, bs, target_bs,
@ -345,7 +349,7 @@ index c28462a633..a402fa4bf7 100644
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 da5fb31089..32f0f9858a 100644
index ef31c58bb3..57265a617a 100644
--- a/include/block/block_int-global-state.h
+++ b/include/block/block_int-global-state.h
@@ -152,7 +152,9 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
@ -360,10 +364,10 @@ index da5fb31089..32f0f9858a 100644
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
diff --git a/qapi/block-core.json b/qapi/block-core.json
index bca1a0c372..a5cea82139 100644
index ca390c5700..8db0986e9e 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -2145,6 +2145,15 @@
@@ -2163,6 +2163,15 @@
# destination (all the disk, only the sectors allocated in the
# topmost image, or only new I/O).
#
@ -379,7 +383,7 @@ index bca1a0c372..a5cea82139 100644
# @granularity: granularity of the dirty bitmap, default is 64K if the
# image format doesn't have clusters, 4K if the clusters are
# smaller than that, else the cluster size. Must be a power of 2
@@ -2187,7 +2196,9 @@
@@ -2205,7 +2214,9 @@
{ 'struct': 'DriveMirror',
'data': { '*job-id': 'str', 'device': 'str', 'target': 'str',
'*format': 'str', '*node-name': 'str', '*replaces': 'str',
@ -390,7 +394,7 @@ index bca1a0c372..a5cea82139 100644
'*speed': 'int', '*granularity': 'uint32',
'*buf-size': 'int', '*on-source-error': 'BlockdevOnError',
'*on-target-error': 'BlockdevOnError',
@@ -2471,6 +2482,15 @@
@@ -2489,6 +2500,15 @@
# destination (all the disk, only the sectors allocated in the
# topmost image, or only new I/O).
#
@ -406,7 +410,7 @@ index bca1a0c372..a5cea82139 100644
# @granularity: granularity of the dirty bitmap, default is 64K if the
# image format doesn't have clusters, 4K if the clusters are
# smaller than that, else the cluster size. Must be a power of 2
@@ -2521,7 +2541,8 @@
@@ -2539,7 +2559,8 @@
{ 'command': 'blockdev-mirror',
'data': { '*job-id': 'str', 'device': 'str', 'target': 'str',
'*replaces': 'str',
@ -417,12 +421,12 @@ index bca1a0c372..a5cea82139 100644
'*buf-size': 'int', '*on-source-error': 'BlockdevOnError',
'*on-target-error': 'BlockdevOnError',
diff --git a/tests/unit/test-block-iothread.c b/tests/unit/test-block-iothread.c
index d727a5fee8..8a34aa2328 100644
index 9b15d2768c..54acd47188 100644
--- a/tests/unit/test-block-iothread.c
+++ b/tests/unit/test-block-iothread.c
@@ -757,8 +757,8 @@ static void test_propagate_mirror(void)
@@ -766,8 +766,8 @@ static void test_propagate_mirror(void)
/* Start a mirror job */
aio_context_acquire(main_ctx);
mirror_start("job0", src, target, NULL, JOB_DEFAULT, 0, 0, 0,
- MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false,
- BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
@ -430,4 +434,4 @@ index d727a5fee8..8a34aa2328 100644
+ false, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
false, "filter_node", MIRROR_COPY_MODE_BACKGROUND,
&error_abort);
WITH_JOB_LOCK_GUARD() {
aio_context_release(main_ctx);

14
debian/patches/bitmap-mirror/0002-drive-mirror-add-support-for-conditional-and-always-.patch vendored
View File

@ -24,10 +24,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 18 insertions(+), 6 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 1ff42c8af1..11b8a8e959 100644
index ed14c8b498..beb98f2e07 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -682,8 +682,6 @@ static int mirror_exit_common(Job *job)
@@ -695,8 +695,6 @@ static int mirror_exit_common(Job *job)
bdrv_unfreeze_backing_chain(mirror_top_bs, target_bs);
}
@ -36,9 +36,9 @@ index 1ff42c8af1..11b8a8e959 100644
/* Make sure that the source BDS doesn't go away during bdrv_replace_node,
* before we can call bdrv_drained_end */
bdrv_ref(src);
@@ -788,6 +786,18 @@ static int mirror_exit_common(Job *job)
block_job_remove_all_bdrv(bjob);
bdrv_replace_node(mirror_top_bs, mirror_top_bs->backing->bs, &error_abort);
@@ -810,6 +808,18 @@ static int mirror_exit_common(Job *job)
bdrv_drained_end(target_bs);
bdrv_unref(target_bs);
+ if (s->sync_bitmap) {
+ if (s->bitmap_mode == BITMAP_SYNC_MODE_ALWAYS ||
@ -55,7 +55,7 @@ index 1ff42c8af1..11b8a8e959 100644
bs_opaque->job = NULL;
bdrv_drained_end(src);
@@ -1699,10 +1709,6 @@ static BlockJob *mirror_start_job(
@@ -1776,10 +1786,6 @@ static BlockJob *mirror_start_job(
" sync mode",
MirrorSyncMode_str(sync_mode));
return NULL;
@ -66,7 +66,7 @@ index 1ff42c8af1..11b8a8e959 100644
}
} else if (bitmap) {
error_setg(errp,
@@ -1719,6 +1725,12 @@ static BlockJob *mirror_start_job(
@@ -1796,6 +1802,12 @@ static BlockJob *mirror_start_job(
return NULL;
}
granularity = bdrv_dirty_bitmap_granularity(bitmap);

4
debian/patches/bitmap-mirror/0003-mirror-add-check-for-bitmap-mode-without-bitmap.patch vendored
View File

@ -16,10 +16,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 3 insertions(+)
diff --git a/blockdev.c b/blockdev.c
index a402fa4bf7..01b0ab0549 100644
index 8c8e8b604a..cdfe7b33b1 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2946,6 +2946,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -3000,6 +3000,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
if (bdrv_dirty_bitmap_check(bitmap, BDRV_BITMAP_ALLOW_RO, errp)) {
return;
}

8
debian/patches/bitmap-mirror/0004-mirror-switch-to-bdrv_dirty_bitmap_merge_internal.patch vendored
View File

@ -16,10 +16,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 4 insertions(+), 7 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 11b8a8e959..00f2665ca4 100644
index beb98f2e07..1aa75e7384 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -792,8 +792,8 @@ static int mirror_exit_common(Job *job)
@@ -814,8 +814,8 @@ static int mirror_exit_common(Job *job)
job->ret == 0 && ret == 0)) {
/* Success; synchronize copy back to sync. */
bdrv_clear_dirty_bitmap(s->sync_bitmap, NULL);
@ -30,7 +30,7 @@ index 11b8a8e959..00f2665ca4 100644
}
}
bdrv_release_dirty_bitmap(s->dirty_bitmap);
@@ -1892,11 +1892,8 @@ static BlockJob *mirror_start_job(
@@ -1983,11 +1983,8 @@ static BlockJob *mirror_start_job(
}
if (s->sync_mode == MIRROR_SYNC_MODE_BITMAP) {
@ -43,4 +43,4 @@ index 11b8a8e959..00f2665ca4 100644
+ NULL, true);
}
ret = block_job_add_bdrv(&s->common, "source", bs, 0,
bdrv_graph_wrlock(bs);

14
debian/patches/bitmap-mirror/0006-mirror-move-some-checks-to-qmp.patch vendored
View File

@ -12,7 +12,7 @@ uniform w.r.t. backup block jobs.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
[FE: rebase for 8.0]
[FE: rebase for 8.2.2]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/mirror.c | 28 +++------------
@ -21,12 +21,12 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
3 files changed, 70 insertions(+), 59 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index 00f2665ca4..60cf574de5 100644
index 1aa75e7384..a02efbfbbd 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -1699,31 +1699,13 @@ static BlockJob *mirror_start_job(
uint64_t target_perms, target_shared_perms;
int ret;
@@ -1776,31 +1776,13 @@ static BlockJob *mirror_start_job(
GLOBAL_STATE_CODE();
- if (sync_mode == MIRROR_SYNC_MODE_INCREMENTAL) {
- error_setg(errp, "Sync mode '%s' not supported",
@ -62,10 +62,10 @@ index 00f2665ca4..60cf574de5 100644
if (bitmap_mode != BITMAP_SYNC_MODE_NEVER) {
diff --git a/blockdev.c b/blockdev.c
index 01b0ab0549..cd5f205ad1 100644
index cdfe7b33b1..38a40e3e32 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2925,7 +2925,36 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
@@ -2979,7 +2979,36 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
sync = MIRROR_SYNC_MODE_FULL;
}

2
debian/patches/extra/0001-monitor-qmp-fix-race-with-clients-disconnecting-earl.patch vendored
View File

@ -78,7 +78,7 @@ index 252de85681..8db28f9272 100644
/**
diff --git a/monitor/monitor.c b/monitor/monitor.c
index dc352f9e9d..56e1307014 100644
index 01ede1babd..5681bca346 100644
--- a/monitor/monitor.c
+++ b/monitor/monitor.c
@@ -117,6 +117,21 @@ bool monitor_cur_is_qmp(void)

10
debian/patches/extra/0003-ide-avoid-potential-deadlock-when-draining-during-tr.patch vendored
View File

@ -55,10 +55,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/hw/ide/core.c b/hw/ide/core.c
index c3508acbb1..289347af58 100644
index 8a0579bff4..254255f8dc 100644
--- a/hw/ide/core.c
+++ b/hw/ide/core.c
@@ -444,7 +444,7 @@ static void ide_trim_bh_cb(void *opaque)
@@ -456,7 +456,7 @@ static void ide_trim_bh_cb(void *opaque)
iocb->bh = NULL;
qemu_aio_unref(iocb);
@ -67,7 +67,7 @@ index c3508acbb1..289347af58 100644
blk_dec_in_flight(blk);
}
@@ -504,6 +504,8 @@ static void ide_issue_trim_cb(void *opaque, int ret)
@@ -516,6 +516,8 @@ static void ide_issue_trim_cb(void *opaque, int ret)
done:
iocb->aiocb = NULL;
if (iocb->bh) {
@ -76,7 +76,7 @@ index c3508acbb1..289347af58 100644
replay_bh_schedule_event(iocb->bh);
}
}
@@ -516,9 +518,6 @@ BlockAIOCB *ide_issue_trim(
@@ -528,9 +530,6 @@ BlockAIOCB *ide_issue_trim(
IDEDevice *dev = s->unit ? s->bus->slave : s->bus->master;
TrimAIOCB *iocb;
@ -86,7 +86,7 @@ index c3508acbb1..289347af58 100644
iocb = blk_aio_get(&trim_aiocb_info, s->blk, cb, cb_opaque);
iocb->s = s;
iocb->bh = qemu_bh_new_guarded(ide_trim_bh_cb, iocb,
@@ -742,8 +741,9 @@ void ide_cancel_dma_sync(IDEState *s)
@@ -754,8 +753,9 @@ void ide_cancel_dma_sync(IDEState *s)
*/
if (s->bus->dma->aiocb) {
trace_ide_cancel_dma_sync_remaining();

6
debian/patches/extra/0004-migration-block-dirty-bitmap-fix-loading-bitmap-when.patch vendored
View File

@ -20,10 +20,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
1 file changed, 6 insertions(+)
diff --git a/migration/block-dirty-bitmap.c b/migration/block-dirty-bitmap.c
index 032fc5f405..e1ae3b7316 100644
index 24347ab0f7..0070b13b6f 100644
--- a/migration/block-dirty-bitmap.c
+++ b/migration/block-dirty-bitmap.c
@@ -805,8 +805,11 @@ static int dirty_bitmap_load_start(QEMUFile *f, DBMLoadState *s)
@@ -809,8 +809,11 @@ static int dirty_bitmap_load_start(QEMUFile *f, DBMLoadState *s)
"destination", bdrv_dirty_bitmap_name(s->bitmap));
return -EINVAL;
} else {
@ -35,7 +35,7 @@ index 032fc5f405..e1ae3b7316 100644
if (!s->bitmap) {
error_report_err(local_err);
return -EINVAL;
@@ -833,7 +836,10 @@ static int dirty_bitmap_load_start(QEMUFile *f, DBMLoadState *s)
@@ -837,7 +840,10 @@ static int dirty_bitmap_load_start(QEMUFile *f, DBMLoadState *s)
bdrv_disable_dirty_bitmap(s->bitmap);
if (flags & DIRTY_BITMAP_MIG_START_FLAG_ENABLED) {

140
debian/patches/extra/0005-Revert-Revert-graph-lock-Disable-locking-for-now.patch vendored
View File

@ -1,140 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Thu, 28 Sep 2023 10:07:03 +0200
Subject: [PATCH] Revert "Revert "graph-lock: Disable locking for now""
This reverts commit 3cce22defb4b0e47cf135444e30cc673cff5ebad.
There are still some issues with graph locking, e.g. deadlocks during
backup canceling [0]. Because the AioContext locks still exist, it
should be safe to disable locking again.
From the original 80fc5d2600 ("graph-lock: Disable locking for now"):
> We don't currently rely on graph locking yet. It is supposed to replace
> the AioContext lock eventually to enable multiqueue support, but as long
> as we still have the AioContext lock, it is sufficient without the graph
> lock. Once the AioContext lock goes away, the deadlock doesn't exist any
> more either and this commit can be reverted. (Of course, it can also be
> reverted while the AioContext lock still exists if the callers have been
> fixed.)
[0]: https://lists.nongnu.org/archive/html/qemu-devel/2023-09/msg00729.html
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/graph-lock.c | 24 ++++++++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/block/graph-lock.c b/block/graph-lock.c
index 5e66f01ae8..5c2873262a 100644
--- a/block/graph-lock.c
+++ b/block/graph-lock.c
@@ -30,8 +30,10 @@ BdrvGraphLock graph_lock;
/* Protects the list of aiocontext and orphaned_reader_count */
static QemuMutex aio_context_list_lock;
+#if 0
/* Written and read with atomic operations. */
static int has_writer;
+#endif
/*
* A reader coroutine could move from an AioContext to another.
@@ -88,6 +90,7 @@ void unregister_aiocontext(AioContext *ctx)
g_free(ctx->bdrv_graph);
}
+#if 0
static uint32_t reader_count(void)
{
BdrvGraphRWlock *brdv_graph;
@@ -105,12 +108,19 @@ static uint32_t reader_count(void)
assert((int32_t)rd >= 0);
return rd;
}
+#endif
void bdrv_graph_wrlock(BlockDriverState *bs)
{
+#if 0
AioContext *ctx = NULL;
GLOBAL_STATE_CODE();
+ /*
+ * TODO Some callers hold an AioContext lock when this is called, which
+ * causes deadlocks. Reenable once the AioContext locking is cleaned up (or
+ * AioContext locks are gone).
+ */
assert(!qatomic_read(&has_writer));
/*
@@ -158,11 +168,13 @@ void bdrv_graph_wrlock(BlockDriverState *bs)
if (ctx) {
aio_context_acquire(bdrv_get_aio_context(bs));
}
+#endif
}
void bdrv_graph_wrunlock(void)
{
GLOBAL_STATE_CODE();
+#if 0
QEMU_LOCK_GUARD(&aio_context_list_lock);
assert(qatomic_read(&has_writer));
@@ -174,10 +186,13 @@ void bdrv_graph_wrunlock(void)
/* Wake up all coroutine that are waiting to read the graph */
qemu_co_enter_all(&reader_queue, &aio_context_list_lock);
+#endif
}
void coroutine_fn bdrv_graph_co_rdlock(void)
{
+ /* TODO Reenable when wrlock is reenabled */
+#if 0
BdrvGraphRWlock *bdrv_graph;
bdrv_graph = qemu_get_current_aio_context()->bdrv_graph;
@@ -237,10 +252,12 @@ void coroutine_fn bdrv_graph_co_rdlock(void)
qemu_co_queue_wait(&reader_queue, &aio_context_list_lock);
}
}
+#endif
}
void coroutine_fn bdrv_graph_co_rdunlock(void)
{
+#if 0
BdrvGraphRWlock *bdrv_graph;
bdrv_graph = qemu_get_current_aio_context()->bdrv_graph;
@@ -258,6 +275,7 @@ void coroutine_fn bdrv_graph_co_rdunlock(void)
if (qatomic_read(&has_writer)) {
aio_wait_kick();
}
+#endif
}
void bdrv_graph_rdlock_main_loop(void)
@@ -275,13 +293,19 @@ void bdrv_graph_rdunlock_main_loop(void)
void assert_bdrv_graph_readable(void)
{
/* reader_count() is slow due to aio_context_list_lock lock contention */
+ /* TODO Reenable when wrlock is reenabled */
+#if 0
#ifdef CONFIG_DEBUG_GRAPH_LOCK
assert(qemu_in_main_thread() || reader_count());
#endif
+#endif
}
void assert_bdrv_graph_writable(void)
{
assert(qemu_in_main_thread());
+ /* TODO Reenable when wrlock is reenabled */
+#if 0
assert(qatomic_read(&has_writer));
+#endif
}

4
debian/patches/extra/0007-Revert-x86-acpi-workaround-Windows-not-handling-name.patch → debian/patches/extra/0005-Revert-x86-acpi-workaround-Windows-not-handling-name.patch vendored
View File

@ -24,10 +24,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
index bb12b0ad43..de14d3c3da 100644
index 1e178341de..d8694b338e 100644
--- a/hw/i386/acpi-build.c
+++ b/hw/i386/acpi-build.c
@@ -362,13 +362,9 @@ Aml *aml_pci_device_dsm(void)
@@ -357,13 +357,9 @@ Aml *aml_pci_device_dsm(void)
{
Aml *params = aml_local(0);
Aml *pkg = aml_package(2);

57
debian/patches/extra/0006-migration-states-workaround-snapshot-performance-reg.patch vendored
View File

@ -1,57 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Thu, 28 Sep 2023 11:19:14 +0200
Subject: [PATCH] migration states: workaround snapshot performance regression
Commit 813cd616 ("migration: Use migration_transferred_bytes() to
calculate rate_limit") introduced a prohibitive performance regression
when taking a snapshot [0]. The reason turns out to be the flushing
done by migration_transferred_bytes()
Just use a _noflush version of the relevant function as a workaround
until upstream fixes the issue. This is inspired by a not-applied
upstream series [1], but doing the very minimum to avoid the
regression.
[0]: https://gitlab.com/qemu-project/qemu/-/issues/1821
[1]: https://lists.nongnu.org/archive/html/qemu-devel/2023-05/msg07708.html
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
migration/migration-stats.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/migration/migration-stats.c b/migration/migration-stats.c
index 095d6d75bb..8073c8ebaa 100644
--- a/migration/migration-stats.c
+++ b/migration/migration-stats.c
@@ -18,6 +18,20 @@
MigrationAtomicStats mig_stats;
+/*
+ * Same as migration_transferred_bytes below, but using the _noflush
+ * variant of qemu_file_transferred() to avoid a performance
+ * regression in migration_rate_exceeded().
+ */
+static uint64_t migration_transferred_bytes_noflush(QEMUFile *f)
+{
+ uint64_t multifd = stat64_get(&mig_stats.multifd_bytes);
+ uint64_t qemu_file = qemu_file_transferred_noflush(f);
+
+ trace_migration_transferred_bytes(qemu_file, multifd);
+ return qemu_file + multifd;
+}
+
bool migration_rate_exceeded(QEMUFile *f)
{
if (qemu_file_get_error(f)) {
@@ -25,7 +39,7 @@ bool migration_rate_exceeded(QEMUFile *f)
}
uint64_t rate_limit_start = stat64_get(&mig_stats.rate_limit_start);
- uint64_t rate_limit_current = migration_transferred_bytes(f);
+ uint64_t rate_limit_current = migration_transferred_bytes_noflush(f);
uint64_t rate_limit_used = rate_limit_current - rate_limit_start;
uint64_t rate_limit_max = stat64_get(&mig_stats.rate_limit_max);

28
debian/patches/extra/0012-qemu_init-increase-NOFILE-soft-limit-on-POSIX.patch → debian/patches/extra/0006-qemu_init-increase-NOFILE-soft-limit-on-POSIX.patch vendored
View File

@ -33,26 +33,26 @@ Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
include/sysemu/os-posix.h | 1 +
include/sysemu/os-win32.h | 5 +++++
os-posix.c | 22 ++++++++++++++++++++++
softmmu/vl.c | 2 ++
system/vl.c | 2 ++
4 files changed, 30 insertions(+)
diff --git a/include/sysemu/os-posix.h b/include/sysemu/os-posix.h
index 1030d39904..edc415aff5 100644
index dff32ae185..b881ac6c6f 100644
--- a/include/sysemu/os-posix.h
+++ b/include/sysemu/os-posix.h
@@ -48,6 +48,7 @@ void os_setup_early_signal_handling(void);
void os_set_proc_name(const char *s);
void os_setup_signal_handling(void);
@@ -51,6 +51,7 @@ bool is_daemonized(void);
void os_daemonize(void);
bool os_set_runas(const char *user_id);
void os_set_chroot(const char *path);
+void os_setup_limits(void);
void os_setup_post(void);
int os_mlock(void);
diff --git a/include/sysemu/os-win32.h b/include/sysemu/os-win32.h
index 91aa0d7ec0..f6e23fe01e 100644
index 1047d260cb..106f155037 100644
--- a/include/sysemu/os-win32.h
+++ b/include/sysemu/os-win32.h
@@ -129,6 +129,11 @@ static inline int os_mlock(void)
@@ -128,6 +128,11 @@ static inline int os_mlock(void)
return -ENOSYS;
}
@ -65,7 +65,7 @@ index 91aa0d7ec0..f6e23fe01e 100644
#if !defined(lseek)
diff --git a/os-posix.c b/os-posix.c
index cfcb96533c..0cc1d991b1 100644
index 52ef6990ff..a4284e2c07 100644
--- a/os-posix.c
+++ b/os-posix.c
@@ -24,6 +24,7 @@
@ -76,7 +76,7 @@ index cfcb96533c..0cc1d991b1 100644
#include <sys/wait.h>
#include <pwd.h>
#include <grp.h>
@@ -286,6 +287,27 @@ void os_daemonize(void)
@@ -256,6 +257,27 @@ void os_daemonize(void)
}
}
@ -104,11 +104,11 @@ index cfcb96533c..0cc1d991b1 100644
void os_setup_post(void)
{
int fd = 0;
diff --git a/softmmu/vl.c b/softmmu/vl.c
index c9e9ede237..ba6ad8a8df 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -2713,6 +2713,8 @@ void qemu_init(int argc, char **argv)
diff --git a/system/vl.c b/system/vl.c
index e18fa3ce46..d2a3b3f457 100644
--- a/system/vl.c
+++ b/system/vl.c
@@ -2782,6 +2782,8 @@ void qemu_init(int argc, char **argv)
error_init(argv[0]);
qemu_init_exec_dir(argv[0]);

79
debian/patches/extra/0007-mirror-Don-t-call-job_pause_point-under-graph-lock.patch vendored Normal file
View File

@ -0,0 +1,79 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Kevin Wolf <kwolf@redhat.com>
Date: Wed, 13 Mar 2024 16:30:00 +0100
Subject: [PATCH] mirror: Don't call job_pause_point() under graph lock
Calling job_pause_point() while holding the graph reader lock
potentially results in a deadlock: bdrv_graph_wrlock() first drains
everything, including the mirror job, which pauses it. The job is only
unpaused at the end of the drain section, which is when the graph writer
lock has been successfully taken. However, if the job happens to be
paused at a pause point where it still holds the reader lock, the writer
lock can't be taken as long as the job is still paused.
Mark job_pause_point() as GRAPH_UNLOCKED and fix mirror accordingly.
Cc: qemu-stable@nongnu.org
Buglink: https://issues.redhat.com/browse/RHEL-28125
Fixes: 004915a96a7a ("block: Protect bs->backing with graph_lock")
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20240313153000.33121-1-kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
(cherry picked from commit ae5a40e8581185654a667fbbf7e4adbc2a2a3e45)
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/mirror.c | 10 ++++++----
include/qemu/job.h | 2 +-
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/block/mirror.c b/block/mirror.c
index cd9d3ad4a8..abbddb39e4 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -479,9 +479,9 @@ static unsigned mirror_perform(MirrorBlockJob *s, int64_t offset,
return bytes_handled;
}
-static void coroutine_fn GRAPH_RDLOCK mirror_iteration(MirrorBlockJob *s)
+static void coroutine_fn GRAPH_UNLOCKED mirror_iteration(MirrorBlockJob *s)
{
- BlockDriverState *source = s->mirror_top_bs->backing->bs;
+ BlockDriverState *source;
MirrorOp *pseudo_op;
int64_t offset;
/* At least the first dirty chunk is mirrored in one iteration. */
@@ -489,6 +489,10 @@ static void coroutine_fn GRAPH_RDLOCK mirror_iteration(MirrorBlockJob *s)
bool write_zeroes_ok = bdrv_can_write_zeroes_with_unmap(blk_bs(s->target));
int max_io_bytes = MAX(s->buf_size / MAX_IN_FLIGHT, MAX_IO_BYTES);
+ bdrv_graph_co_rdlock();
+ source = s->mirror_top_bs->backing->bs;
+ bdrv_graph_co_rdunlock();
+
bdrv_dirty_bitmap_lock(s->dirty_bitmap);
offset = bdrv_dirty_iter_next(s->dbi);
if (offset < 0) {
@@ -1078,9 +1082,7 @@ static int coroutine_fn mirror_run(Job *job, Error **errp)
mirror_wait_for_free_in_flight_slot(s);
continue;
} else if (cnt != 0) {
- bdrv_graph_co_rdlock();
mirror_iteration(s);
- bdrv_graph_co_rdunlock();
}
}
diff --git a/include/qemu/job.h b/include/qemu/job.h
index e502787dd8..b4bc2e174b 100644
--- a/include/qemu/job.h
+++ b/include/qemu/job.h
@@ -503,7 +503,7 @@ void job_enter(Job *job);
*
* Called with job_mutex *not* held.
*/
-void coroutine_fn job_pause_point(Job *job);
+void coroutine_fn GRAPH_UNLOCKED job_pause_point(Job *job);
/**
* @job: The job that calls the function.

34
debian/patches/extra/0008-target-i386-the-sgx_epc_get_section-stub-is-reachabl.patch vendored
View File

@ -1,34 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Paolo Bonzini <pbonzini@redhat.com>
Date: Tue, 1 Feb 2022 20:09:41 +0100
Subject: [PATCH] target/i386: the sgx_epc_get_section stub is reachable
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
The sgx_epc_get_section stub is reachable from cpu_x86_cpuid. It
should not assert, instead it should just return true just like
the "real" sgx_epc_get_section does when SGX is disabled.
Reported-by: Vladimír Beneš <vbenes@redhat.com>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20220201190941.106001-1-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry-picked from commit 219615740425d9683588207b40a365e6741691a6)
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
hw/i386/sgx-stub.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/i386/sgx-stub.c b/hw/i386/sgx-stub.c
index 26833eb233..16b1dfd90b 100644
--- a/hw/i386/sgx-stub.c
+++ b/hw/i386/sgx-stub.c
@@ -34,5 +34,5 @@ void pc_machine_init_sgx_epc(PCMachineState *pcms)
bool sgx_epc_get_section(int section_nr, uint64_t *addr, uint64_t *size)
{
- g_assert_not_reached();
+ return true;
}

86
debian/patches/extra/0009-ui-clipboard-mark-type-as-not-available-when-there-i.patch vendored
View File

@ -1,86 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Wed, 24 Jan 2024 11:57:48 +0100
Subject: [PATCH] ui/clipboard: mark type as not available when there is no
data
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
With VNC, a client can send a non-extended VNC_MSG_CLIENT_CUT_TEXT
message with len=0. In qemu_clipboard_set_data(), the clipboard info
will be updated setting data to NULL (because g_memdup(data, size)
returns NULL when size is 0). If the client does not set the
VNC_ENCODING_CLIPBOARD_EXT feature when setting up the encodings, then
the 'request' callback for the clipboard peer is not initialized.
Later, because data is NULL, qemu_clipboard_request() can be reached
via vdagent_chr_write() and vdagent_clipboard_recv_request() and
there, the clipboard owner's 'request' callback will be attempted to
be called, but that is a NULL pointer.
In particular, this can happen when using the KRDC (22.12.3) VNC
client.
Another scenario leading to the same issue is with two clients (say
noVNC and KRDC):
The noVNC client sets the extension VNC_FEATURE_CLIPBOARD_EXT and
initializes its cbpeer.
The KRDC client does not, but triggers a vnc_client_cut_text() (note
it's not the _ext variant)). There, a new clipboard info with it as
the 'owner' is created and via qemu_clipboard_set_data() is called,
which in turn calls qemu_clipboard_update() with that info.
In qemu_clipboard_update(), the notifier for the noVNC client will be
called, i.e. vnc_clipboard_notify() and also set vs->cbinfo for the
noVNC client. The 'owner' in that clipboard info is the clipboard peer
for the KRDC client, which did not initialize the 'request' function.
That sounds correct to me, it is the owner of that clipboard info.
Then when noVNC sends a VNC_MSG_CLIENT_CUT_TEXT message (it did set
the VNC_FEATURE_CLIPBOARD_EXT feature correctly, so a check for it
passes), that clipboard info is passed to qemu_clipboard_request() and
the original segfault still happens.
Fix the issue by handling updates with size 0 differently. In
particular, mark in the clipboard info that the type is not available.
While at it, switch to g_memdup2(), because g_memdup() is deprecated.
Cc: qemu-stable@nongnu.org
Fixes: CVE-2023-6683
Reported-by: Markus Frank <m.frank@proxmox.com>
Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Tested-by: Markus Frank <m.frank@proxmox.com>
(picked from https://lists.nongnu.org/archive/html/qemu-stable/2024-01/msg00228.html)
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
ui/clipboard.c | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/ui/clipboard.c b/ui/clipboard.c
index 3d14bffaf8..b3f6fa3c9e 100644
--- a/ui/clipboard.c
+++ b/ui/clipboard.c
@@ -163,9 +163,15 @@ void qemu_clipboard_set_data(QemuClipboardPeer *peer,
}
g_free(info->types[type].data);
- info->types[type].data = g_memdup(data, size);
- info->types[type].size = size;
- info->types[type].available = true;
+ if (size) {
+ info->types[type].data = g_memdup2(data, size);
+ info->types[type].size = size;
+ info->types[type].available = true;
+ } else {
+ info->types[type].data = NULL;
+ info->types[type].size = 0;
+ info->types[type].available = false;
+ }
if (update) {
qemu_clipboard_update(info);

65
debian/patches/extra/0010-virtio-scsi-Attach-event-vq-notifier-with-no_poll.patch vendored
View File

@ -1,65 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Hanna Czenczek <hreitz@redhat.com>
Date: Fri, 2 Feb 2024 16:31:56 +0100
Subject: [PATCH] virtio-scsi: Attach event vq notifier with no_poll
As of commit 38738f7dbbda90fbc161757b7f4be35b52205552 ("virtio-scsi:
don't waste CPU polling the event virtqueue"), we only attach an io_read
notifier for the virtio-scsi event virtqueue instead, and no polling
notifiers. During operation, the event virtqueue is typically
non-empty, but none of the buffers are intended to be used immediately.
Instead, they only get used when certain events occur. Therefore, it
makes no sense to continuously poll it when non-empty, because it is
supposed to be and stay non-empty.
We do this by using virtio_queue_aio_attach_host_notifier_no_poll()
instead of virtio_queue_aio_attach_host_notifier() for the event
virtqueue.
Commit 766aa2de0f29b657148e04599320d771c36fd126 ("virtio-scsi: implement
BlockDevOps->drained_begin()") however has virtio_scsi_drained_end() use
virtio_queue_aio_attach_host_notifier() for all virtqueues, including
the event virtqueue. This can lead to it being polled again, undoing
the benefit of commit 38738f7dbbda90fbc161757b7f4be35b52205552.
Fix it by using virtio_queue_aio_attach_host_notifier_no_poll() for the
event virtqueue.
("virtio-scsi: implement BlockDevOps->drained_begin()")
Reported-by: Fiona Ebner <f.ebner@proxmox.com>
Fixes: 766aa2de0f29b657148e04599320d771c36fd126
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Tested-by: Fiona Ebner <f.ebner@proxmox.com>
Reviewed-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
hw/scsi/virtio-scsi.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index 45b95ea070..ad24a882fd 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -1148,6 +1148,7 @@ static void virtio_scsi_drained_begin(SCSIBus *bus)
static void virtio_scsi_drained_end(SCSIBus *bus)
{
VirtIOSCSI *s = container_of(bus, VirtIOSCSI, bus);
+ VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(s);
VirtIODevice *vdev = VIRTIO_DEVICE(s);
uint32_t total_queues = VIRTIO_SCSI_VQ_NUM_FIXED +
s->parent_obj.conf.num_queues;
@@ -1165,7 +1166,11 @@ static void virtio_scsi_drained_end(SCSIBus *bus)
for (uint32_t i = 0; i < total_queues; i++) {
VirtQueue *vq = virtio_get_queue(vdev, i);
- virtio_queue_aio_attach_host_notifier(vq, s->ctx);
+ if (vq == vs->event_vq) {
+ virtio_queue_aio_attach_host_notifier_no_poll(vq, s->ctx);
+ } else {
+ virtio_queue_aio_attach_host_notifier(vq, s->ctx);
+ }
}
}

125
debian/patches/extra/0011-virtio-Re-enable-notifications-after-drain.patch vendored
View File

@ -1,125 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Hanna Czenczek <hreitz@redhat.com>
Date: Fri, 2 Feb 2024 16:31:57 +0100
Subject: [PATCH] virtio: Re-enable notifications after drain
During drain, we do not care about virtqueue notifications, which is why
we remove the handlers on it. When removing those handlers, whether vq
notifications are enabled or not depends on whether we were in polling
mode or not; if not, they are enabled (by default); if so, they have
been disabled by the io_poll_start callback.
Because we do not care about those notifications after removing the
handlers, this is fine. However, we have to explicitly ensure they are
enabled when re-attaching the handlers, so we will resume receiving
notifications. We do this in virtio_queue_aio_attach_host_notifier*().
If such a function is called while we are in a polling section,
attaching the notifiers will then invoke the io_poll_start callback,
re-disabling notifications.
Because we will always miss virtqueue updates in the drained section, we
also need to poll the virtqueue once after attaching the notifiers.
Buglink: https://issues.redhat.com/browse/RHEL-3934
Signed-off-by: Hanna Czenczek <hreitz@redhat.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
hw/virtio/virtio.c | 42 ++++++++++++++++++++++++++++++++++++++++++
include/block/aio.h | 7 ++++++-
2 files changed, 48 insertions(+), 1 deletion(-)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 969c25f4cf..02cce83111 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -3526,6 +3526,17 @@ static void virtio_queue_host_notifier_aio_poll_end(EventNotifier *n)
void virtio_queue_aio_attach_host_notifier(VirtQueue *vq, AioContext *ctx)
{
+ /*
+ * virtio_queue_aio_detach_host_notifier() can leave notifications disabled.
+ * Re-enable them. (And if detach has not been used before, notifications
+ * being enabled is still the default state while a notifier is attached;
+ * see virtio_queue_host_notifier_aio_poll_end(), which will always leave
+ * notifications enabled once the polling section is left.)
+ */
+ if (!virtio_queue_get_notification(vq)) {
+ virtio_queue_set_notification(vq, 1);
+ }
+
aio_set_event_notifier(ctx, &vq->host_notifier,
virtio_queue_host_notifier_read,
virtio_queue_host_notifier_aio_poll,
@@ -3533,6 +3544,13 @@ void virtio_queue_aio_attach_host_notifier(VirtQueue *vq, AioContext *ctx)
aio_set_event_notifier_poll(ctx, &vq->host_notifier,
virtio_queue_host_notifier_aio_poll_begin,
virtio_queue_host_notifier_aio_poll_end);
+
+ /*
+ * We will have ignored notifications about new requests from the guest
+ * while no notifiers were attached, so "kick" the virt queue to process
+ * those requests now.
+ */
+ event_notifier_set(&vq->host_notifier);
}
/*
@@ -3543,14 +3561,38 @@ void virtio_queue_aio_attach_host_notifier(VirtQueue *vq, AioContext *ctx)
*/
void virtio_queue_aio_attach_host_notifier_no_poll(VirtQueue *vq, AioContext *ctx)
{
+ /* See virtio_queue_aio_attach_host_notifier() */
+ if (!virtio_queue_get_notification(vq)) {
+ virtio_queue_set_notification(vq, 1);
+ }
+
aio_set_event_notifier(ctx, &vq->host_notifier,
virtio_queue_host_notifier_read,
NULL, NULL);
+
+ /*
+ * See virtio_queue_aio_attach_host_notifier().
+ * Note that this may be unnecessary for the type of virtqueues this
+ * function is used for. Still, it will not hurt to have a quick look into
+ * whether we can/should process any of the virtqueue elements.
+ */
+ event_notifier_set(&vq->host_notifier);
}
void virtio_queue_aio_detach_host_notifier(VirtQueue *vq, AioContext *ctx)
{
aio_set_event_notifier(ctx, &vq->host_notifier, NULL, NULL, NULL);
+
+ /*
+ * aio_set_event_notifier_poll() does not guarantee whether io_poll_end()
+ * will run after io_poll_begin(), so by removing the notifier, we do not
+ * know whether virtio_queue_host_notifier_aio_poll_end() has run after a
+ * previous virtio_queue_host_notifier_aio_poll_begin(), i.e. whether
+ * notifications are enabled or disabled. It does not really matter anyway;
+ * we just removed the notifier, so we do not care about notifications until
+ * we potentially re-attach it. The attach_host_notifier functions will
+ * ensure that notifications are enabled again when they are needed.
+ */
}
void virtio_queue_host_notifier_read(EventNotifier *n)
diff --git a/include/block/aio.h b/include/block/aio.h
index 32042e8905..79efadfa48 100644
--- a/include/block/aio.h
+++ b/include/block/aio.h
@@ -498,9 +498,14 @@ void aio_set_event_notifier(AioContext *ctx,
AioPollFn *io_poll,
EventNotifierHandler *io_poll_ready);
-/* Set polling begin/end callbacks for an event notifier that has already been
+/*
+ * Set polling begin/end callbacks for an event notifier that has already been
* registered with aio_set_event_notifier. Do nothing if the event notifier is
* not registered.
+ *
+ * Note that if the io_poll_end() callback (or the entire notifier) is removed
+ * during polling, it will not be called, so an io_poll_begin() is not
+ * necessarily always followed by an io_poll_end().
*/
void aio_set_event_notifier_poll(AioContext *ctx,
EventNotifier *notifier,

61
debian/patches/extra/0013-virtio-blk-avoid-using-ioeventfd-state-in-irqfd-cond.patch vendored
View File

@ -1,61 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Stefan Hajnoczi <stefanha@redhat.com>
Date: Mon, 22 Jan 2024 12:26:25 -0500
Subject: [PATCH] virtio-blk: avoid using ioeventfd state in irqfd conditional
Requests that complete in an IOThread use irqfd to notify the guest
while requests that complete in the main loop thread use the traditional
qdev irq code path. The reason for this conditional is that the irq code
path requires the BQL:
if (s->ioeventfd_started && !s->ioeventfd_disabled) {
virtio_notify_irqfd(vdev, req->vq);
} else {
virtio_notify(vdev, req->vq);
}
There is a corner case where the conditional invokes the irq code path
instead of the irqfd code path:
static void virtio_blk_stop_ioeventfd(VirtIODevice *vdev)
{
...
/*
* Set ->ioeventfd_started to false before draining so that host notifiers
* are not detached/attached anymore.
*/
s->ioeventfd_started = false;
/* Wait for virtio_blk_dma_restart_bh() and in flight I/O to complete */
blk_drain(s->conf.conf.blk);
During blk_drain() the conditional produces the wrong result because
ioeventfd_started is false.
Use qemu_in_iothread() instead of checking the ioeventfd state.
Cc: qemu-stable@nongnu.org
Buglink: https://issues.redhat.com/browse/RHEL-15394
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20240122172625.415386-1-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
[FE: backport: dataplane -> ioeventfd rework didn't happen yet]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
hw/block/virtio-blk.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index 39e7f23fab..61bd1f6859 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -64,7 +64,7 @@ static void virtio_blk_req_complete(VirtIOBlockReq *req, unsigned char status)
iov_discard_undo(&req->inhdr_undo);
iov_discard_undo(&req->outhdr_undo);
virtqueue_push(req->vq, &req->elem, req->in_len);
- if (s->dataplane_started && !s->dataplane_disabled) {
+ if (qemu_in_iothread()) {
virtio_blk_data_plane_notify(s->dataplane, req->vq);
} else {
virtio_notify(vdev, req->vq);

2
debian/patches/pve/0001-PVE-Config-block-file-change-locking-default-to-off.patch vendored
View File

@ -14,7 +14,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 7f540b03ed..ca551baa42 100644
index b862406c71..bc09aefe3b 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -563,7 +563,7 @@ static QemuOptsList raw_runtime_opts = {

4
debian/patches/pve/0002-PVE-Config-Adjust-network-script-path-to-etc-kvm.patch vendored
View File

@ -9,10 +9,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/include/net/net.h b/include/net/net.h
index 685ec58318..22edf4ee96 100644
index ffbd2c8d56..e857c75f4c 100644
--- a/include/net/net.h
+++ b/include/net/net.h
@@ -260,8 +260,8 @@ void netdev_add(QemuOpts *opts, Error **errp);
@@ -263,8 +263,8 @@ void netdev_add(QemuOpts *opts, Error **errp);
int net_hub_id_for_client(NetClientState *nc, int *id);
NetClientState *net_hub_port_find(int hub_id);

4
debian/patches/pve/0003-PVE-Config-set-the-CPU-model-to-kvm64-32-instead-of-.patch vendored
View File

@ -10,10 +10,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 0893b794e9..6d650a58b9 100644
index 705d925e6c..ba269e5f1e 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -2243,9 +2243,9 @@ uint64_t cpu_get_tsc(CPUX86State *env);
@@ -2281,9 +2281,9 @@ uint64_t cpu_get_tsc(CPUX86State *env);
#define CPU_RESOLVING_TYPE TYPE_X86_CPU
#ifdef TARGET_X86_64

2
debian/patches/pve/0004-PVE-Config-ui-spice-default-to-pve-certificates.patch vendored
View File

@ -9,7 +9,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git a/ui/spice-core.c b/ui/spice-core.c
index 52a59386d7..b20c25aee0 100644
index db21db2c94..0eb138cf43 100644
--- a/ui/spice-core.c
+++ b/ui/spice-core.c
@@ -691,32 +691,35 @@ static void qemu_spice_init(void)

2
debian/patches/pve/0005-PVE-Config-glusterfs-no-default-logfile-if-daemonize.patch vendored
View File

@ -9,7 +9,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 11 insertions(+), 4 deletions(-)
diff --git a/block/gluster.c b/block/gluster.c
index ad5fadbe79..d0011085c4 100644
index cc74af06dc..3ba9bbfa5e 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -43,7 +43,7 @@

2
debian/patches/pve/0006-PVE-Config-rbd-block-rbd-disable-rbd_cache_writethro.patch vendored
View File

@ -18,7 +18,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+)
diff --git a/block/rbd.c b/block/rbd.c
index 978671411e..a4749f3b1b 100644
index 84bb2fa5d7..63f60d41be 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -963,6 +963,8 @@ static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx,

10
debian/patches/pve/0007-PVE-Up-glusterfs-allow-partial-reads.patch vendored
View File

@ -16,7 +16,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/block/gluster.c b/block/gluster.c
index d0011085c4..2df3d6e35d 100644
index 3ba9bbfa5e..34936eb855 100644
--- a/block/gluster.c
+++ b/block/gluster.c
@@ -58,6 +58,7 @@ typedef struct GlusterAIOCB {
@ -39,7 +39,7 @@ index d0011085c4..2df3d6e35d 100644
}
aio_co_schedule(acb->aio_context, acb->coroutine);
@@ -1021,6 +1024,7 @@ static coroutine_fn int qemu_gluster_co_pwrite_zeroes(BlockDriverState *bs,
@@ -1023,6 +1026,7 @@ static coroutine_fn int qemu_gluster_co_pwrite_zeroes(BlockDriverState *bs,
acb.ret = 0;
acb.coroutine = qemu_coroutine_self();
acb.aio_context = bdrv_get_aio_context(bs);
@ -47,7 +47,7 @@ index d0011085c4..2df3d6e35d 100644
ret = glfs_zerofill_async(s->fd, offset, bytes, gluster_finish_aiocb, &acb);
if (ret < 0) {
@@ -1201,9 +1205,11 @@ static coroutine_fn int qemu_gluster_co_rw(BlockDriverState *bs,
@@ -1203,9 +1207,11 @@ static coroutine_fn int qemu_gluster_co_rw(BlockDriverState *bs,
acb.aio_context = bdrv_get_aio_context(bs);
if (write) {
@ -59,7 +59,7 @@ index d0011085c4..2df3d6e35d 100644
ret = glfs_preadv_async(s->fd, qiov->iov, qiov->niov, offset, 0,
gluster_finish_aiocb, &acb);
}
@@ -1266,6 +1272,7 @@ static coroutine_fn int qemu_gluster_co_flush_to_disk(BlockDriverState *bs)
@@ -1268,6 +1274,7 @@ static coroutine_fn int qemu_gluster_co_flush_to_disk(BlockDriverState *bs)
acb.ret = 0;
acb.coroutine = qemu_coroutine_self();
acb.aio_context = bdrv_get_aio_context(bs);
@ -67,7 +67,7 @@ index d0011085c4..2df3d6e35d 100644
ret = glfs_fsync_async(s->fd, gluster_finish_aiocb, &acb);
if (ret < 0) {
@@ -1314,6 +1321,7 @@ static coroutine_fn int qemu_gluster_co_pdiscard(BlockDriverState *bs,
@@ -1316,6 +1323,7 @@ static coroutine_fn int qemu_gluster_co_pdiscard(BlockDriverState *bs,
acb.ret = 0;
acb.coroutine = qemu_coroutine_self();
acb.aio_context = bdrv_get_aio_context(bs);

4
debian/patches/pve/0008-PVE-Up-qemu-img-return-success-on-info-without-snaps.patch vendored
View File

@ -9,10 +9,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/qemu-img.c b/qemu-img.c
index 78433f3746..25d427edd1 100644
index 5a77f67719..de51233825 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -3062,7 +3062,8 @@ static int img_info(int argc, char **argv)
@@ -3079,7 +3079,8 @@ static int img_info(int argc, char **argv)
list = collect_image_info_list(image_opts, filename, fmt, chain,
force_share);
if (!list) {

14
debian/patches/pve/0009-PVE-Up-qemu-img-dd-add-osize-and-read-from-to-stdin-.patch vendored
View File

@ -38,7 +38,7 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2 files changed, 133 insertions(+), 73 deletions(-)
diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx
index 1b1dab5b17..d1616c045a 100644
index 068692d13e..73e0bb1d2c 100644
--- a/qemu-img-cmds.hx
+++ b/qemu-img-cmds.hx
@@ -58,9 +58,9 @@ SRST
@ -54,10 +54,10 @@ index 1b1dab5b17..d1616c045a 100644
DEF("info", img_info,
diff --git a/qemu-img.c b/qemu-img.c
index 25d427edd1..220e6ec577 100644
index de51233825..ad770f6570 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -4899,10 +4899,12 @@ static int img_bitmap(int argc, char **argv)
@@ -4997,10 +4997,12 @@ static int img_bitmap(int argc, char **argv)
#define C_IF 04
#define C_OF 010
#define C_SKIP 020
@ -70,7 +70,7 @@ index 25d427edd1..220e6ec577 100644
};
struct DdIo {
@@ -4978,6 +4980,19 @@ static int img_dd_skip(const char *arg,
@@ -5076,6 +5078,19 @@ static int img_dd_skip(const char *arg,
return 0;
}
@ -90,7 +90,7 @@ index 25d427edd1..220e6ec577 100644
static int img_dd(int argc, char **argv)
{
int ret = 0;
@@ -5018,6 +5033,7 @@ static int img_dd(int argc, char **argv)
@@ -5116,6 +5131,7 @@ static int img_dd(int argc, char **argv)
{ "if", img_dd_if, C_IF },
{ "of", img_dd_of, C_OF },
{ "skip", img_dd_skip, C_SKIP },
@ -98,7 +98,7 @@ index 25d427edd1..220e6ec577 100644
{ NULL, NULL, 0 }
};
const struct option long_options[] = {
@@ -5093,91 +5109,112 @@ static int img_dd(int argc, char **argv)
@@ -5191,91 +5207,112 @@ static int img_dd(int argc, char **argv)
arg = NULL;
}
@ -275,7 +275,7 @@ index 25d427edd1..220e6ec577 100644
}
if (dd.flags & C_SKIP && (in.offset > INT64_MAX / in.bsz ||
@@ -5194,20 +5231,43 @@ static int img_dd(int argc, char **argv)
@@ -5292,20 +5329,43 @@ static int img_dd(int argc, char **argv)
in.buf = g_new(uint8_t, in.bsz);
for (out_pos = 0; in_pos < size; ) {

14
debian/patches/pve/0010-PVE-Up-qemu-img-dd-add-isize-parameter.patch vendored
View File

@ -16,10 +16,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
1 file changed, 25 insertions(+), 3 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index 220e6ec577..58bf9b43d1 100644
index ad770f6570..b0839a5d18 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -4900,11 +4900,13 @@ static int img_bitmap(int argc, char **argv)
@@ -4998,11 +4998,13 @@ static int img_bitmap(int argc, char **argv)
#define C_OF 010
#define C_SKIP 020
#define C_OSIZE 040
@ -33,7 +33,7 @@ index 220e6ec577..58bf9b43d1 100644
};
struct DdIo {
@@ -4993,6 +4995,19 @@ static int img_dd_osize(const char *arg,
@@ -5091,6 +5093,19 @@ static int img_dd_osize(const char *arg,
return 0;
}
@ -53,7 +53,7 @@ index 220e6ec577..58bf9b43d1 100644
static int img_dd(int argc, char **argv)
{
int ret = 0;
@@ -5007,12 +5022,14 @@ static int img_dd(int argc, char **argv)
@@ -5105,12 +5120,14 @@ static int img_dd(int argc, char **argv)
int c, i;
const char *out_fmt = "raw";
const char *fmt = NULL;
@ -69,7 +69,7 @@ index 220e6ec577..58bf9b43d1 100644
};
struct DdIo in = {
.bsz = 512, /* Block size is by default 512 bytes */
@@ -5034,6 +5051,7 @@ static int img_dd(int argc, char **argv)
@@ -5132,6 +5149,7 @@ static int img_dd(int argc, char **argv)
{ "of", img_dd_of, C_OF },
{ "skip", img_dd_skip, C_SKIP },
{ "osize", img_dd_osize, C_OSIZE },
@ -77,7 +77,7 @@ index 220e6ec577..58bf9b43d1 100644
{ NULL, NULL, 0 }
};
const struct option long_options[] = {
@@ -5230,9 +5248,10 @@ static int img_dd(int argc, char **argv)
@@ -5328,9 +5346,10 @@ static int img_dd(int argc, char **argv)
in.buf = g_new(uint8_t, in.bsz);
@ -90,7 +90,7 @@ index 220e6ec577..58bf9b43d1 100644
if (blk1) {
in_ret = blk_pread(blk1, in_pos, bytes, in.buf, 0);
if (in_ret == 0) {
@@ -5241,6 +5260,9 @@ static int img_dd(int argc, char **argv)
@@ -5339,6 +5358,9 @@ static int img_dd(int argc, char **argv)
} else {
in_ret = read(STDIN_FILENO, in.buf, bytes);
if (in_ret == 0) {

20
debian/patches/pve/0011-PVE-Up-qemu-img-dd-add-n-skip_create.patch vendored
View File

@ -13,10 +13,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
3 files changed, 26 insertions(+), 12 deletions(-)
diff --git a/docs/tools/qemu-img.rst b/docs/tools/qemu-img.rst
index 15aeddc6d8..5e713e231d 100644
index 4459c065f1..c6bc902182 100644
--- a/docs/tools/qemu-img.rst
+++ b/docs/tools/qemu-img.rst
@@ -208,6 +208,10 @@ Parameters to convert subcommand:
@@ -212,6 +212,10 @@ Parameters to convert subcommand:
Parameters to dd subcommand:
@ -27,7 +27,7 @@ index 15aeddc6d8..5e713e231d 100644
.. program:: qemu-img-dd
.. option:: bs=BLOCK_SIZE
@@ -488,7 +492,7 @@ Command description:
@@ -492,7 +496,7 @@ Command description:
it doesn't need to be specified separately in this case.
@ -36,7 +36,7 @@ index 15aeddc6d8..5e713e231d 100644
dd copies from *INPUT* file to *OUTPUT* file converting it from
*FMT* format to *OUTPUT_FMT* format.
@@ -499,6 +503,11 @@ Command description:
@@ -503,6 +507,11 @@ Command description:
The size syntax is similar to :manpage:`dd(1)`'s size syntax.
@ -49,7 +49,7 @@ index 15aeddc6d8..5e713e231d 100644
Give information about the disk image *FILENAME*. Use it in
diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx
index d1616c045a..b5b0bb4467 100644
index 73e0bb1d2c..32749b416e 100644
--- a/qemu-img-cmds.hx
+++ b/qemu-img-cmds.hx
@@ -58,9 +58,9 @@ SRST
@ -65,10 +65,10 @@ index d1616c045a..b5b0bb4467 100644
DEF("info", img_info,
diff --git a/qemu-img.c b/qemu-img.c
index 58bf9b43d1..9d414d639b 100644
index b0839a5d18..4acdf75879 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -5024,7 +5024,7 @@ static int img_dd(int argc, char **argv)
@@ -5122,7 +5122,7 @@ static int img_dd(int argc, char **argv)
const char *fmt = NULL;
int64_t size = 0, readsize = 0;
int64_t out_pos, in_pos;
@ -77,7 +77,7 @@ index 58bf9b43d1..9d414d639b 100644
struct DdInfo dd = {
.flags = 0,
.count = 0,
@@ -5062,7 +5062,7 @@ static int img_dd(int argc, char **argv)
@@ -5160,7 +5160,7 @@ static int img_dd(int argc, char **argv)
{ 0, 0, 0, 0 }
};
@ -86,7 +86,7 @@ index 58bf9b43d1..9d414d639b 100644
if (c == EOF) {
break;
}
@@ -5082,6 +5082,9 @@ static int img_dd(int argc, char **argv)
@@ -5180,6 +5180,9 @@ static int img_dd(int argc, char **argv)
case 'h':
help();
break;
@ -96,7 +96,7 @@ index 58bf9b43d1..9d414d639b 100644
case 'U':
force_share = true;
break;
@@ -5212,13 +5215,15 @@ static int img_dd(int argc, char **argv)
@@ -5310,13 +5313,15 @@ static int img_dd(int argc, char **argv)
size - in.bsz * in.offset, &error_abort);
}

20
debian/patches/pve/0012-qemu-img-dd-add-l-option-for-loading-a-snapshot.patch vendored
View File

@ -12,10 +12,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
3 files changed, 36 insertions(+), 7 deletions(-)
diff --git a/docs/tools/qemu-img.rst b/docs/tools/qemu-img.rst
index 5e713e231d..9390d5e5cf 100644
index c6bc902182..fe2241856e 100644
--- a/docs/tools/qemu-img.rst
+++ b/docs/tools/qemu-img.rst
@@ -492,10 +492,10 @@ Command description:
@@ -496,10 +496,10 @@ Command description:
it doesn't need to be specified separately in this case.
@ -30,7 +30,7 @@ index 5e713e231d..9390d5e5cf 100644
The data is by default read and written using blocks of 512 bytes but can be
modified by specifying *BLOCK_SIZE*. If count=\ *BLOCKS* is specified
diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx
index b5b0bb4467..36f97e1f19 100644
index 32749b416e..40ee9c0ed8 100644
--- a/qemu-img-cmds.hx
+++ b/qemu-img-cmds.hx
@@ -58,9 +58,9 @@ SRST
@ -46,10 +46,10 @@ index b5b0bb4467..36f97e1f19 100644
DEF("info", img_info,
diff --git a/qemu-img.c b/qemu-img.c
index 9d414d639b..e13a12137b 100644
index 4acdf75879..a9632b5951 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -5016,6 +5016,7 @@ static int img_dd(int argc, char **argv)
@@ -5114,6 +5114,7 @@ static int img_dd(int argc, char **argv)
BlockDriver *drv = NULL, *proto_drv = NULL;
BlockBackend *blk1 = NULL, *blk2 = NULL;
QemuOpts *opts = NULL;
@ -57,7 +57,7 @@ index 9d414d639b..e13a12137b 100644
QemuOptsList *create_opts = NULL;
Error *local_err = NULL;
bool image_opts = false;
@@ -5025,6 +5026,7 @@ static int img_dd(int argc, char **argv)
@@ -5123,6 +5124,7 @@ static int img_dd(int argc, char **argv)
int64_t size = 0, readsize = 0;
int64_t out_pos, in_pos;
bool force_share = false, skip_create = false;
@ -65,7 +65,7 @@ index 9d414d639b..e13a12137b 100644
struct DdInfo dd = {
.flags = 0,
.count = 0,
@@ -5062,7 +5064,7 @@ static int img_dd(int argc, char **argv)
@@ -5160,7 +5162,7 @@ static int img_dd(int argc, char **argv)
{ 0, 0, 0, 0 }
};
@ -74,7 +74,7 @@ index 9d414d639b..e13a12137b 100644
if (c == EOF) {
break;
}
@@ -5085,6 +5087,19 @@ static int img_dd(int argc, char **argv)
@@ -5183,6 +5185,19 @@ static int img_dd(int argc, char **argv)
case 'n':
skip_create = true;
break;
@ -94,7 +94,7 @@ index 9d414d639b..e13a12137b 100644
case 'U':
force_share = true;
break;
@@ -5144,11 +5159,24 @@ static int img_dd(int argc, char **argv)
@@ -5242,11 +5257,24 @@ static int img_dd(int argc, char **argv)
if (dd.flags & C_IF) {
blk1 = img_open(image_opts, in.filename, fmt, 0, false, false,
force_share);
@ -120,7 +120,7 @@ index 9d414d639b..e13a12137b 100644
}
if (dd.flags & C_OSIZE) {
@@ -5303,6 +5331,7 @@ static int img_dd(int argc, char **argv)
@@ -5401,6 +5429,7 @@ static int img_dd(int argc, char **argv)
out:
g_free(arg);
qemu_opts_del(opts);

10
debian/patches/pve/0013-PVE-virtio-balloon-improve-query-balloon.patch vendored
View File

@ -18,10 +18,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
4 files changed, 82 insertions(+), 4 deletions(-)
diff --git a/hw/core/machine-hmp-cmds.c b/hw/core/machine-hmp-cmds.c
index c3e55ef9e9..0e32e6201f 100644
index a6ff6a4875..e7f74d1c63 100644
--- a/hw/core/machine-hmp-cmds.c
+++ b/hw/core/machine-hmp-cmds.c
@@ -169,7 +169,35 @@ void hmp_info_balloon(Monitor *mon, const QDict *qdict)
@@ -175,7 +175,35 @@ void hmp_info_balloon(Monitor *mon, const QDict *qdict)
return;
}
@ -103,10 +103,10 @@ index d004cf29d2..2660ed520b 100644
static void virtio_balloon_to_target(void *opaque, ram_addr_t target)
diff --git a/qapi/machine.json b/qapi/machine.json
index a08b6576ca..5c9a4d55f4 100644
index b6d634b30d..03a72efc11 100644
--- a/qapi/machine.json
+++ b/qapi/machine.json
@@ -1063,9 +1063,29 @@
@@ -1087,9 +1087,29 @@
# @actual: the logical size of the VM in bytes Formula used:
# logical_vm_size = vm_ram_size - balloon_size
#
@ -138,7 +138,7 @@ index a08b6576ca..5c9a4d55f4 100644
##
# @query-balloon:
diff --git a/qapi/pragma.json b/qapi/pragma.json
index 7f810b0e97..325e684411 100644
index 0aa4eeddd3..eae9f54700 100644
--- a/qapi/pragma.json
+++ b/qapi/pragma.json
@@ -35,6 +35,7 @@

6
debian/patches/pve/0014-PVE-qapi-modify-query-machines.patch vendored
View File

@ -30,10 +30,10 @@ index 3860a50c3b..40821e2317 100644
info->default_cpu_type = g_strdup(mc->default_cpu_type);
}
diff --git a/qapi/machine.json b/qapi/machine.json
index 5c9a4d55f4..fbb61f18e4 100644
index 03a72efc11..297ad0e0e5 100644
--- a/qapi/machine.json
+++ b/qapi/machine.json
@@ -139,6 +139,8 @@
@@ -146,6 +146,8 @@
#
# @is-default: whether the machine is default
#
@ -42,7 +42,7 @@ index 5c9a4d55f4..fbb61f18e4 100644
# @cpu-max: maximum number of CPUs supported by the machine type
# (since 1.5)
#
@@ -163,7 +165,7 @@
@@ -170,7 +172,7 @@
##
{ 'struct': 'MachineInfo',
'data': { 'name': 'str', '*alias': 'str',

6
debian/patches/pve/0015-PVE-qapi-modify-spice-query.patch vendored
View File

@ -14,10 +14,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
2 files changed, 7 insertions(+)
diff --git a/qapi/ui.json b/qapi/ui.json
index 006616aa77..dfd1d3e36b 100644
index a0158baf23..20d8994f49 100644
--- a/qapi/ui.json
+++ b/qapi/ui.json
@@ -317,11 +317,14 @@
@@ -318,11 +318,14 @@
#
# @channels: a list of @SpiceChannel for each active spice channel
#
@ -33,7 +33,7 @@ index 006616aa77..dfd1d3e36b 100644
'if': 'CONFIG_SPICE' }
diff --git a/ui/spice-core.c b/ui/spice-core.c
index b20c25aee0..26baeb7846 100644
index 0eb138cf43..56d677efe1 100644
--- a/ui/spice-core.c
+++ b/ui/spice-core.c
@@ -548,6 +548,10 @@ static SpiceInfo *qmp_query_spice_real(Error **errp)

13
debian/patches/pve/0016-PVE-add-IOChannel-implementation-for-savevm-async.patch vendored
View File

@ -16,19 +16,19 @@ Additionally, allows tracking the current position from the outside
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
migration/channel-savevm-async.c | 183 +++++++++++++++++++++++++++++++
migration/channel-savevm-async.c | 184 +++++++++++++++++++++++++++++++
migration/channel-savevm-async.h | 51 +++++++++
migration/meson.build | 1 +
3 files changed, 235 insertions(+)
3 files changed, 236 insertions(+)
create mode 100644 migration/channel-savevm-async.c
create mode 100644 migration/channel-savevm-async.h
diff --git a/migration/channel-savevm-async.c b/migration/channel-savevm-async.c
new file mode 100644
index 0000000000..aab081ce07
index 0000000000..081a192f49
--- /dev/null
+++ b/migration/channel-savevm-async.c
@@ -0,0 +1,183 @@
@@ -0,0 +1,184 @@
+/*
+ * QIO Channel implementation to be used by savevm-async QMP calls
+ */
@ -175,8 +175,9 @@ index 0000000000..aab081ce07
+
+static void
+qio_channel_savevm_async_set_aio_fd_handler(QIOChannel *ioc,
+ AioContext *ctx,
+ AioContext *read_ctx,
+ IOHandler *io_read,
+ AioContext *write_ctx,
+ IOHandler *io_write,
+ void *opaque)
+{
@ -270,7 +271,7 @@ index 0000000000..17ae2cb261
+
+#endif /* QIO_CHANNEL_SAVEVM_ASYNC_H */
diff --git a/migration/meson.build b/migration/meson.build
index 1ae28523a1..37ddcb5d60 100644
index 92b1cc4297..0e689eac09 100644
--- a/migration/meson.build
+++ b/migration/meson.build
@@ -13,6 +13,7 @@ system_ss.add(files(

83
debian/patches/pve/0017-PVE-add-savevm-async-for-background-state-snapshots.patch vendored
View File

@ -27,7 +27,7 @@ Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
[FE: further improve aborting
adapt to removal of QEMUFileOps
improve condition for entering final stage
adapt to QAPI and other changes for 8.0]
adapt to QAPI and other changes for 8.2]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
hmp-commands-info.hx | 13 +
@ -35,13 +35,13 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
include/migration/snapshot.h | 2 +
include/monitor/hmp.h | 3 +
migration/meson.build | 1 +
migration/savevm-async.c | 531 +++++++++++++++++++++++++++++++++++
migration/savevm-async.c | 534 +++++++++++++++++++++++++++++++++++
monitor/hmp-cmds.c | 38 +++
qapi/migration.json | 34 +++
qapi/misc.json | 16 ++
qemu-options.hx | 12 +
softmmu/vl.c | 10 +
11 files changed, 677 insertions(+)
system/vl.c | 10 +
11 files changed, 680 insertions(+)
create mode 100644 migration/savevm-async.c
diff --git a/hmp-commands-info.hx b/hmp-commands-info.hx
@ -69,10 +69,10 @@ index f5b37eb74a..10fdd822e0 100644
.name = "balloon",
.args_type = "",
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 2cbd0f77a0..e352f86872 100644
index 765349ed14..893c3bd240 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1865,3 +1865,20 @@ SRST
@@ -1875,3 +1875,20 @@ SRST
List event channels in the guest
ERST
#endif
@ -126,10 +126,10 @@ index 13f9a2dedb..7a7def7530 100644
void coroutine_fn hmp_screendump(Monitor *mon, const QDict *qdict);
void hmp_chardev_add(Monitor *mon, const QDict *qdict);
diff --git a/migration/meson.build b/migration/meson.build
index 37ddcb5d60..07f6057acc 100644
index 0e689eac09..8f9d122187 100644
--- a/migration/meson.build
+++ b/migration/meson.build
@@ -26,6 +26,7 @@ system_ss.add(files(
@@ -27,6 +27,7 @@ system_ss.add(files(
'options.c',
'postcopy-ram.c',
'savevm.c',
@ -139,10 +139,10 @@ index 37ddcb5d60..07f6057acc 100644
'threadinfo.c',
diff --git a/migration/savevm-async.c b/migration/savevm-async.c
new file mode 100644
index 0000000000..e9fc18fb10
index 0000000000..8f63c4c637
--- /dev/null
+++ b/migration/savevm-async.c
@@ -0,0 +1,531 @@
@@ -0,0 +1,534 @@
+#include "qemu/osdep.h"
+#include "migration/channel-savevm-async.h"
+#include "migration/migration.h"
@ -441,21 +441,25 @@ index 0000000000..e9fc18fb10
+ * so move there now and after every flush.
+ */
+ aio_co_reschedule_self(qemu_get_aio_context());
+ for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
+ bdrv_graph_co_rdlock();
+ bs = bdrv_first(&it);
+ bdrv_graph_co_rdunlock();
+ while (bs) {
+ /* target has BDRV_O_NO_FLUSH, no sense calling bdrv_flush on it */
+ if (bs == blk_bs(snap_state.target)) {
+ continue;
+ }
+
+ AioContext *bs_ctx = bdrv_get_aio_context(bs);
+ if (bs_ctx != qemu_get_aio_context()) {
+ DPRINTF("savevm: async flushing drive %s\n", bs->filename);
+ aio_co_reschedule_self(bs_ctx);
+ bdrv_graph_co_rdlock();
+ bdrv_flush(bs);
+ bdrv_graph_co_rdunlock();
+ aio_co_reschedule_self(qemu_get_aio_context());
+ if (bs != blk_bs(snap_state.target)) {
+ AioContext *bs_ctx = bdrv_get_aio_context(bs);
+ if (bs_ctx != qemu_get_aio_context()) {
+ DPRINTF("savevm: async flushing drive %s\n", bs->filename);
+ aio_co_reschedule_self(bs_ctx);
+ bdrv_graph_co_rdlock();
+ bdrv_flush(bs);
+ bdrv_graph_co_rdunlock();
+ aio_co_reschedule_self(qemu_get_aio_context());
+ }
+ }
+ bdrv_graph_co_rdlock();
+ bs = bdrv_next(&it);
+ bdrv_graph_co_rdunlock();
+ }
+
+ DPRINTF("timing: async flushing took %ld ms\n",
@ -535,9 +539,10 @@ index 0000000000..e9fc18fb10
+ * State is cleared in process_savevm_co, but has to be initialized
+ * here (blocking main thread, from QMP) to avoid race conditions.
+ */
+ migrate_init(ms);
+ if (migrate_init(ms, errp)) {
+ return;
+ }
+ memset(&mig_stats, 0, sizeof(mig_stats));
+ memset(&compression_counters, 0, sizeof(compression_counters));
+ ms->to_dst_file = snap_state.file;
+
+ error_setg(&snap_state.blocker, "block device is in use by savevm");
@ -546,10 +551,8 @@ index 0000000000..e9fc18fb10
+ snap_state.state = SAVE_STATE_ACTIVE;
+ snap_state.finalize_bh = qemu_bh_new(process_savevm_finalize, &snap_state);
+ snap_state.co = qemu_coroutine_create(&process_savevm_co, NULL);
+ qemu_mutex_unlock_iothread();
+ qemu_savevm_state_header(snap_state.file);
+ qemu_savevm_state_setup(snap_state.file);
+ qemu_mutex_lock_iothread();
+
+ /* Async processing from here on out happens in iohandler context, so let
+ * the target bdrv have its home there.
@ -675,7 +678,7 @@ index 0000000000..e9fc18fb10
+ return ret;
+}
diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
index 6c559b48c8..91be698308 100644
index 871898ac46..ef4634e5c1 100644
--- a/monitor/hmp-cmds.c
+++ b/monitor/hmp-cmds.c
@@ -22,6 +22,7 @@
@ -685,7 +688,7 @@ index 6c559b48c8..91be698308 100644
+#include "qapi/qapi-commands-migration.h"
#include "qapi/qapi-commands-misc.h"
#include "qapi/qmp/qdict.h"
#include "qapi/qmp/qerror.h"
#include "qemu/cutils.h"
@@ -443,3 +444,40 @@ void hmp_info_mtree(Monitor *mon, const QDict *qdict)
mtree_info(flatview, dispatch_tree, owner, disabled);
@ -728,10 +731,10 @@ index 6c559b48c8..91be698308 100644
+ }
+}
diff --git a/qapi/migration.json b/qapi/migration.json
index 8843e74b59..aca0ca1ac1 100644
index 197d3faa43..b41465fbe9 100644
--- a/qapi/migration.json
+++ b/qapi/migration.json
@@ -291,6 +291,40 @@
@@ -298,6 +298,40 @@
'*dirty-limit-throttle-time-per-round': 'uint64',
'*dirty-limit-ring-full-time': 'uint64'} }
@ -800,10 +803,10 @@ index cda2effa81..94a58bb0bf 100644
# @CommandLineParameterType:
#
diff --git a/qemu-options.hx b/qemu-options.hx
index 8073f5edf5..dc1ececc9c 100644
index b6b4ad9e67..881b0b3c43 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -4483,6 +4483,18 @@ SRST
@@ -4590,6 +4590,18 @@ SRST
Start right away with a saved state (``loadvm`` in monitor)
ERST
@ -822,11 +825,11 @@ index 8073f5edf5..dc1ececc9c 100644
#ifndef _WIN32
DEF("daemonize", 0, QEMU_OPTION_daemonize, \
"-daemonize daemonize QEMU after initializing\n", QEMU_ARCH_ALL)
diff --git a/softmmu/vl.c b/softmmu/vl.c
index ba6ad8a8df..ddeace306e 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -164,6 +164,7 @@ static const char *accelerators;
diff --git a/system/vl.c b/system/vl.c
index d2a3b3f457..57f7ba0525 100644
--- a/system/vl.c
+++ b/system/vl.c
@@ -163,6 +163,7 @@ static const char *accelerators;
static bool have_custom_ram_size;
static const char *ram_memdev_id;
static QDict *machine_opts_dict;
@ -834,7 +837,7 @@ index ba6ad8a8df..ddeace306e 100644
static QTAILQ_HEAD(, ObjectOption) object_opts = QTAILQ_HEAD_INITIALIZER(object_opts);
static QTAILQ_HEAD(, DeviceOption) device_opts = QTAILQ_HEAD_INITIALIZER(device_opts);
static int display_remote;
@@ -2647,6 +2648,12 @@ void qmp_x_exit_preconfig(Error **errp)
@@ -2715,6 +2716,12 @@ void qmp_x_exit_preconfig(Error **errp)
if (loadvm) {
load_snapshot(loadvm, NULL, false, NULL, &error_fatal);
@ -847,7 +850,7 @@ index ba6ad8a8df..ddeace306e 100644
}
if (replay_mode != REPLAY_MODE_NONE) {
replay_vmstate_init();
@@ -3196,6 +3203,9 @@ void qemu_init(int argc, char **argv)
@@ -3265,6 +3272,9 @@ void qemu_init(int argc, char **argv)
case QEMU_OPTION_loadvm:
loadvm = optarg;
break;

63
debian/patches/pve/0018-PVE-add-optional-buffer-size-to-QEMUFile.patch vendored
View File

@ -13,18 +13,18 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
[FE: adapt to removal of QEMUFileOps]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
migration/qemu-file.c | 49 +++++++++++++++++++++++++++-------------
migration/qemu-file.c | 50 +++++++++++++++++++++++++++-------------
migration/qemu-file.h | 2 ++
migration/savevm-async.c | 5 ++--
3 files changed, 38 insertions(+), 18 deletions(-)
3 files changed, 39 insertions(+), 18 deletions(-)
diff --git a/migration/qemu-file.c b/migration/qemu-file.c
index 19c33c9985..e9ffff0f0a 100644
index 94231ff295..afda98292f 100644
--- a/migration/qemu-file.c
+++ b/migration/qemu-file.c
@@ -33,8 +33,8 @@
#include "options.h"
@@ -34,8 +34,8 @@
#include "qapi/error.h"
#include "rdma.h"
-#define IO_BUF_SIZE 32768
-#define MAX_IOV_SIZE MIN_CONST(IOV_MAX, 64)
@ -32,8 +32,8 @@ index 19c33c9985..e9ffff0f0a 100644
+#define MAX_IOV_SIZE MIN_CONST(IOV_MAX, 256)
struct QEMUFile {
const QEMUFileHooks *hooks;
@@ -46,7 +46,8 @@ struct QEMUFile {
QIOChannel *ioc;
@@ -43,7 +43,8 @@ struct QEMUFile {
int buf_index;
int buf_size; /* 0 when writing */
@ -43,7 +43,7 @@ index 19c33c9985..e9ffff0f0a 100644
DECLARE_BITMAP(may_free, MAX_IOV_SIZE);
struct iovec iov[MAX_IOV_SIZE];
@@ -100,7 +101,9 @@ int qemu_file_shutdown(QEMUFile *f)
@@ -97,7 +98,9 @@ int qemu_file_shutdown(QEMUFile *f)
return 0;
}
@ -54,7 +54,7 @@ index 19c33c9985..e9ffff0f0a 100644
{
QEMUFile *f;
@@ -109,6 +112,8 @@ static QEMUFile *qemu_file_new_impl(QIOChannel *ioc, bool is_writable)
@@ -106,6 +109,8 @@ static QEMUFile *qemu_file_new_impl(QIOChannel *ioc, bool is_writable)
object_ref(ioc);
f->ioc = ioc;
f->is_writable = is_writable;
@ -63,7 +63,7 @@ index 19c33c9985..e9ffff0f0a 100644
return f;
}
@@ -119,17 +124,27 @@ static QEMUFile *qemu_file_new_impl(QIOChannel *ioc, bool is_writable)
@@ -116,17 +121,27 @@ static QEMUFile *qemu_file_new_impl(QIOChannel *ioc, bool is_writable)
*/
QEMUFile *qemu_file_get_return_path(QEMUFile *f)
{
@ -93,8 +93,8 @@ index 19c33c9985..e9ffff0f0a 100644
+ return qemu_file_new_impl(ioc, false, buffer_size);
}
void qemu_file_set_hooks(QEMUFile *f, const QEMUFileHooks *hooks)
@@ -375,7 +390,7 @@ static ssize_t coroutine_mixed_fn qemu_fill_buffer(QEMUFile *f)
/*
@@ -320,7 +335,7 @@ static ssize_t coroutine_mixed_fn qemu_fill_buffer(QEMUFile *f)
do {
len = qio_channel_read(f->ioc,
(char *)f->buf + pending,
@ -103,16 +103,17 @@ index 19c33c9985..e9ffff0f0a 100644
&local_error);
if (len == QIO_CHANNEL_ERR_BLOCK) {
if (qemu_in_coroutine()) {
@@ -425,6 +440,8 @@ int qemu_fclose(QEMUFile *f)
@@ -360,6 +375,9 @@ int qemu_fclose(QEMUFile *f)
ret = ret2;
}
g_clear_pointer(&f->ioc, object_unref);
+
+ free(f->buf);
+
/* If any error was spotted before closing, we should report it
* instead of the close() return value.
*/
@@ -479,7 +496,7 @@ static void add_buf_to_iovec(QEMUFile *f, size_t len)
error_free(f->last_error_obj);
g_free(f);
trace_qemu_file_fclose();
@@ -408,7 +426,7 @@ static void add_buf_to_iovec(QEMUFile *f, size_t len)
{
if (!add_to_iovec(f, f->buf + f->buf_index, len, false)) {
f->buf_index += len;
@ -121,7 +122,7 @@ index 19c33c9985..e9ffff0f0a 100644
qemu_fflush(f);
}
}
@@ -504,7 +521,7 @@ void qemu_put_buffer(QEMUFile *f, const uint8_t *buf, size_t size)
@@ -433,7 +451,7 @@ void qemu_put_buffer(QEMUFile *f, const uint8_t *buf, size_t size)
}
while (size > 0) {
@ -130,7 +131,7 @@ index 19c33c9985..e9ffff0f0a 100644
if (l > size) {
l = size;
}
@@ -549,8 +566,8 @@ size_t coroutine_mixed_fn qemu_peek_buffer(QEMUFile *f, uint8_t **buf, size_t si
@@ -478,8 +496,8 @@ size_t coroutine_mixed_fn qemu_peek_buffer(QEMUFile *f, uint8_t **buf, size_t si
size_t index;
assert(!qemu_file_is_writable(f));
@ -141,7 +142,7 @@ index 19c33c9985..e9ffff0f0a 100644
/* The 1st byte to read from */
index = f->buf_index + offset;
@@ -600,7 +617,7 @@ size_t coroutine_mixed_fn qemu_get_buffer(QEMUFile *f, uint8_t *buf, size_t size
@@ -529,7 +547,7 @@ size_t coroutine_mixed_fn qemu_get_buffer(QEMUFile *f, uint8_t *buf, size_t size
size_t res;
uint8_t *src;
@ -150,7 +151,7 @@ index 19c33c9985..e9ffff0f0a 100644
if (res == 0) {
return done;
}
@@ -634,7 +651,7 @@ size_t coroutine_mixed_fn qemu_get_buffer(QEMUFile *f, uint8_t *buf, size_t size
@@ -563,7 +581,7 @@ size_t coroutine_mixed_fn qemu_get_buffer(QEMUFile *f, uint8_t *buf, size_t size
*/
size_t coroutine_mixed_fn qemu_get_buffer_in_place(QEMUFile *f, uint8_t **buf, size_t size)
{
@ -159,7 +160,7 @@ index 19c33c9985..e9ffff0f0a 100644
size_t res;
uint8_t *src = NULL;
@@ -659,7 +676,7 @@ int coroutine_mixed_fn qemu_peek_byte(QEMUFile *f, int offset)
@@ -588,7 +606,7 @@ int coroutine_mixed_fn qemu_peek_byte(QEMUFile *f, int offset)
int index = f->buf_index + offset;
assert(!qemu_file_is_writable(f));
@ -168,7 +169,7 @@ index 19c33c9985..e9ffff0f0a 100644
if (index >= f->buf_size) {
qemu_fill_buffer(f);
@@ -777,7 +794,7 @@ static int qemu_compress_data(z_stream *stream, uint8_t *dest, size_t dest_len,
@@ -702,7 +720,7 @@ static int qemu_compress_data(z_stream *stream, uint8_t *dest, size_t dest_len,
ssize_t qemu_put_compression_data(QEMUFile *f, z_stream *stream,
const uint8_t *p, size_t size)
{
@ -178,24 +179,24 @@ index 19c33c9985..e9ffff0f0a 100644
if (blen < compressBound(size)) {
return -1;
diff --git a/migration/qemu-file.h b/migration/qemu-file.h
index 47015f5201..1312b7c903 100644
index 8aec9fabf7..5d0b18c51c 100644
--- a/migration/qemu-file.h
+++ b/migration/qemu-file.h
@@ -63,7 +63,9 @@ typedef struct QEMUFileHooks {
} QEMUFileHooks;
@@ -30,7 +30,9 @@
#include "io/channel.h"
QEMUFile *qemu_file_new_input(QIOChannel *ioc);
+QEMUFile *qemu_file_new_input_sized(QIOChannel *ioc, size_t buffer_size);
QEMUFile *qemu_file_new_output(QIOChannel *ioc);
+QEMUFile *qemu_file_new_output_sized(QIOChannel *ioc, size_t buffer_size);
void qemu_file_set_hooks(QEMUFile *f, const QEMUFileHooks *hooks);
int qemu_fclose(QEMUFile *f);
/*
diff --git a/migration/savevm-async.c b/migration/savevm-async.c
index e9fc18fb10..80624fada8 100644
index 8f63c4c637..f8d1c2f2b1 100644
--- a/migration/savevm-async.c
+++ b/migration/savevm-async.c
@@ -378,7 +378,7 @@ void qmp_savevm_start(const char *statefile, Error **errp)
@@ -382,7 +382,7 @@ void qmp_savevm_start(const char *statefile, Error **errp)
QIOChannel *ioc = QIO_CHANNEL(qio_channel_savevm_async_new(snap_state.target,
&snap_state.bs_pos));
@ -204,7 +205,7 @@ index e9fc18fb10..80624fada8 100644
if (!snap_state.file) {
error_set(errp, ERROR_CLASS_GENERIC_ERROR, "failed to open '%s'", statefile);
@@ -496,7 +496,8 @@ int load_snapshot_from_blockdev(const char *filename, Error **errp)
@@ -499,7 +499,8 @@ int load_snapshot_from_blockdev(const char *filename, Error **errp)
blk_op_block_all(be, blocker);
/* restore the VM state */

71
debian/patches/pve/0019-PVE-block-add-the-zeroinit-block-driver-filter.patch vendored
View File

@ -4,21 +4,22 @@ Date: Mon, 6 Apr 2020 12:16:47 +0200
Subject: [PATCH] PVE: block: add the zeroinit block driver filter
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
[FE: adapt to changed function signatures]
[FE: adapt to changed function signatures
adhere to block graph lock requirements]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/meson.build | 1 +
block/zeroinit.c | 200 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 201 insertions(+)
block/zeroinit.c | 214 ++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 215 insertions(+)
create mode 100644 block/zeroinit.c
diff --git a/block/meson.build b/block/meson.build
index 529fc172c6..1833c71ce9 100644
index 59ff6d380c..8ded0dc18b 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -40,6 +40,7 @@ block_ss.add(files(
'throttle-groups.c',
@@ -39,6 +39,7 @@ block_ss.add(files(
'throttle.c',
'throttle-groups.c',
'write-threshold.c',
+ 'zeroinit.c',
), zstd, zlib, gnutls)
@ -26,10 +27,10 @@ index 529fc172c6..1833c71ce9 100644
system_ss.add(when: 'CONFIG_TCG', if_true: files('blkreplay.c'))
diff --git a/block/zeroinit.c b/block/zeroinit.c
new file mode 100644
index 0000000000..1257342724
index 0000000000..1f2032bf99
--- /dev/null
+++ b/block/zeroinit.c
@@ -0,0 +1,200 @@
@@ -0,0 +1,214 @@
+/*
+ * Filter to fake a zero-initialized block device.
+ *
@ -44,6 +45,7 @@ index 0000000000..1257342724
+#include "qapi/error.h"
+#include "block/block_int.h"
+#include "block/block-io.h"
+#include "block/graph-lock.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qstring.h"
+#include "qemu/cutils.h"
@ -94,6 +96,7 @@ index 0000000000..1257342724
+ Error **errp)
+{
+ BDRVZeroinitState *s = bs->opaque;
+ BdrvChild *file = NULL;
+ QemuOpts *opts;
+ Error *local_err = NULL;
+ int ret;
@ -109,10 +112,13 @@ index 0000000000..1257342724
+ }
+
+ /* Open the raw file */
+ bs->file = bdrv_open_child(qemu_opt_get(opts, "x-next"), options, "next",
+ bs, &child_of_bds,
+ BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY,
+ false, &local_err);
+ file = bdrv_open_child(qemu_opt_get(opts, "x-next"), options, "next", bs,
+ &child_of_bds,
+ BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY, false,
+ &local_err);
+ bdrv_graph_wrlock(bs);
+ bs->file = file;
+ bdrv_graph_wrunlock(bs);
+ if (local_err) {
+ ret = -EINVAL;
+ error_propagate(errp, local_err);
@ -125,7 +131,9 @@ index 0000000000..1257342724
+ ret = 0;
+fail:
+ if (ret < 0) {
+ bdrv_graph_wrlock(bs);
+ bdrv_unref_child(bs, bs->file);
+ bdrv_graph_wrunlock(bs);
+ }
+ qemu_opts_del(opts);
+ return ret;
@ -137,19 +145,22 @@ index 0000000000..1257342724
+ (void)s;
+}
+
+static coroutine_fn int64_t zeroinit_co_getlength(BlockDriverState *bs)
+static coroutine_fn int64_t GRAPH_RDLOCK
+zeroinit_co_getlength(BlockDriverState *bs)
+{
+ return bdrv_co_getlength(bs->file->bs);
+}
+
+static int coroutine_fn zeroinit_co_preadv(BlockDriverState *bs,
+ int64_t offset, int64_t bytes, QEMUIOVector *qiov, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+zeroinit_co_preadv(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ return bdrv_co_preadv(bs->file, offset, bytes, qiov, flags);
+}
+
+static int coroutine_fn zeroinit_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset,
+ int64_t bytes, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+zeroinit_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ BdrvRequestFlags flags)
+{
+ BDRVZeroinitState *s = bs->opaque;
+ if (offset >= s->extents)
@ -157,8 +168,9 @@ index 0000000000..1257342724
+ return bdrv_pwrite_zeroes(bs->file, offset, bytes, flags);
+}
+
+static int coroutine_fn zeroinit_co_pwritev(BlockDriverState *bs,
+ int64_t offset, int64_t bytes, QEMUIOVector *qiov, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+zeroinit_co_pwritev(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ BDRVZeroinitState *s = bs->opaque;
+ int64_t extents = offset + bytes;
@ -167,32 +179,35 @@ index 0000000000..1257342724
+ return bdrv_co_pwritev(bs->file, offset, bytes, qiov, flags);
+}
+
+static coroutine_fn int zeroinit_co_flush(BlockDriverState *bs)
+static coroutine_fn int GRAPH_RDLOCK
+zeroinit_co_flush(BlockDriverState *bs)
+{
+ return bdrv_co_flush(bs->file->bs);
+}
+
+static int zeroinit_has_zero_init(BlockDriverState *bs)
+static int GRAPH_RDLOCK
+zeroinit_has_zero_init(BlockDriverState *bs)
+{
+ BDRVZeroinitState *s = bs->opaque;
+ return s->has_zero_init;
+}
+
+static int coroutine_fn zeroinit_co_pdiscard(BlockDriverState *bs,
+ int64_t offset, int64_t bytes)
+static int coroutine_fn GRAPH_RDLOCK
+zeroinit_co_pdiscard(BlockDriverState *bs, int64_t offset, int64_t bytes)
+{
+ return bdrv_co_pdiscard(bs->file, offset, bytes);
+}
+
+static int zeroinit_co_truncate(BlockDriverState *bs, int64_t offset,
+ _Bool exact, PreallocMode prealloc,
+ BdrvRequestFlags req_flags, Error **errp)
+static int GRAPH_RDLOCK
+zeroinit_co_truncate(BlockDriverState *bs, int64_t offset, _Bool exact,
+ PreallocMode prealloc, BdrvRequestFlags req_flags,
+ Error **errp)
+{
+ return bdrv_co_truncate(bs->file, offset, exact, prealloc, req_flags, errp);
+}
+
+static coroutine_fn int zeroinit_co_get_info(BlockDriverState *bs,
+ BlockDriverInfo *bdi)
+static coroutine_fn int GRAPH_RDLOCK
+zeroinit_co_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
+{
+ return bdrv_co_get_info(bs->file->bs, bdi);
+}

20
debian/patches/pve/0020-PVE-Add-dummy-id-command-line-parameter.patch vendored
View File

@ -10,14 +10,14 @@ Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
qemu-options.hx | 3 +++
softmmu/vl.c | 8 ++++++++
system/vl.c | 8 ++++++++
2 files changed, 11 insertions(+)
diff --git a/qemu-options.hx b/qemu-options.hx
index dc1ececc9c..848d2dfdd1 100644
index 881b0b3c43..c17374916c 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -1197,6 +1197,9 @@ legacy PC, they are not recommended for modern configurations.
@@ -1218,6 +1218,9 @@ legacy PC, they are not recommended for modern configurations.
ERST
@ -27,11 +27,11 @@ index dc1ececc9c..848d2dfdd1 100644
DEF("fda", HAS_ARG, QEMU_OPTION_fda,
"-fda/-fdb file use 'file' as floppy disk 0/1 image\n", QEMU_ARCH_ALL)
DEF("fdb", HAS_ARG, QEMU_OPTION_fdb, "", QEMU_ARCH_ALL)
diff --git a/softmmu/vl.c b/softmmu/vl.c
index ddeace306e..3ee90b3b94 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -2683,6 +2683,7 @@ void qemu_init(int argc, char **argv)
diff --git a/system/vl.c b/system/vl.c
index 57f7ba0525..56d715c818 100644
--- a/system/vl.c
+++ b/system/vl.c
@@ -2751,6 +2751,7 @@ void qemu_init(int argc, char **argv)
MachineClass *machine_class;
bool userconfig = true;
FILE *vmstate_dump_file = NULL;
@ -39,7 +39,7 @@ index ddeace306e..3ee90b3b94 100644
qemu_add_opts(&qemu_drive_opts);
qemu_add_drive_opts(&qemu_legacy_drive_opts);
@@ -3308,6 +3309,13 @@ void qemu_init(int argc, char **argv)
@@ -3377,6 +3378,13 @@ void qemu_init(int argc, char **argv)
machine_parse_property_opt(qemu_find_opts("smp-opts"),
"smp", optarg);
break;
@ -50,6 +50,6 @@ index ddeace306e..3ee90b3b94 100644
+ exit(1);
+ }
+ break;
#ifdef CONFIG_VNC
case QEMU_OPTION_vnc:
vnc_parse(optarg);
break;

4
debian/patches/pve/0021-PVE-Config-Revert-target-i386-disable-LINT0-after-re.patch vendored
View File

@ -11,10 +11,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 9 insertions(+)
diff --git a/hw/intc/apic_common.c b/hw/intc/apic_common.c
index 4a34f03047..59b917e50c 100644
index bccb4241c2..bd0db7567a 100644
--- a/hw/intc/apic_common.c
+++ b/hw/intc/apic_common.c
@@ -252,6 +252,15 @@ static void apic_reset_common(DeviceState *dev)
@@ -251,6 +251,15 @@ static void apic_reset_common(DeviceState *dev)
info->vapic_base_update(s);
apic_init_reset(dev);

6
debian/patches/pve/0022-PVE-Up-Config-file-posix-make-locking-optiono-on-cre.patch vendored
View File

@ -13,7 +13,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2 files changed, 42 insertions(+), 20 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index ca551baa42..8b3b83e9d4 100644
index bc09aefe3b..36a53d8682 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -2873,6 +2873,7 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
@ -119,10 +119,10 @@ index ca551baa42..8b3b83e9d4 100644
};
return raw_co_create(&options, errp);
diff --git a/qapi/block-core.json b/qapi/block-core.json
index a5cea82139..bb471c078d 100644
index 8db0986e9e..299e3fc350 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -4880,7 +4880,8 @@
@@ -4935,7 +4935,8 @@
'size': 'size',
'*preallocation': 'PreallocMode',
'*nocow': 'bool',

4
debian/patches/pve/0024-PVE-Compat-4.0-used-balloon-qemu-4-0-config-size-fal.patch vendored
View File

@ -26,10 +26,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/hw/core/machine.c b/hw/core/machine.c
index f0d35c6401..1427983543 100644
index 0c17398141..36621d58a7 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -148,7 +148,8 @@ GlobalProperty hw_compat_4_0[] = {
@@ -152,7 +152,8 @@ GlobalProperty hw_compat_4_0[] = {
{ "virtio-vga", "edid", "false" },
{ "virtio-gpu-device", "edid", "false" },
{ "virtio-device", "use-started", "false" },

28
debian/patches/pve/0025-PVE-Allow-version-code-in-machine-type.patch vendored
View File

@ -17,7 +17,7 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
hw/core/machine-qmp-cmds.c | 5 +++++
include/hw/boards.h | 2 ++
qapi/machine.json | 4 +++-
softmmu/vl.c | 25 +++++++++++++++++++++++++
system/vl.c | 25 +++++++++++++++++++++++++
4 files changed, 35 insertions(+), 1 deletion(-)
diff --git a/hw/core/machine-qmp-cmds.c b/hw/core/machine-qmp-cmds.c
@ -37,10 +37,10 @@ index 40821e2317..ee93ddd69a 100644
if (mc->default_cpu_type) {
diff --git a/include/hw/boards.h b/include/hw/boards.h
index ed83360198..f8b88cd86a 100644
index da85f86efb..1aa0987020 100644
--- a/include/hw/boards.h
+++ b/include/hw/boards.h
@@ -235,6 +235,8 @@ struct MachineClass {
@@ -240,6 +240,8 @@ struct MachineClass {
const char *desc;
const char *deprecation_reason;
@ -50,10 +50,10 @@ index ed83360198..f8b88cd86a 100644
void (*reset)(MachineState *state, ShutdownCause reason);
void (*wakeup)(MachineState *state);
diff --git a/qapi/machine.json b/qapi/machine.json
index fbb61f18e4..7da3c519ba 100644
index 297ad0e0e5..a9fd40d844 100644
--- a/qapi/machine.json
+++ b/qapi/machine.json
@@ -161,6 +161,8 @@
@@ -168,6 +168,8 @@
#
# @acpi: machine type supports ACPI (since 8.0)
#
@ -62,7 +62,7 @@ index fbb61f18e4..7da3c519ba 100644
# Since: 1.2
##
{ 'struct': 'MachineInfo',
@@ -168,7 +170,7 @@
@@ -175,7 +177,7 @@
'*is-default': 'bool', '*is-current': 'bool', 'cpu-max': 'int',
'hotpluggable-cpus': 'bool', 'numa-mem-supported': 'bool',
'deprecated': 'bool', '*default-cpu-type': 'str',
@ -71,19 +71,19 @@ index fbb61f18e4..7da3c519ba 100644
##
# @query-machines:
diff --git a/softmmu/vl.c b/softmmu/vl.c
index 3ee90b3b94..4b6d0b82fd 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -1597,6 +1597,7 @@ static const QEMUOption *lookup_opt(int argc, char **argv,
diff --git a/system/vl.c b/system/vl.c
index 56d715c818..87f03e61a1 100644
--- a/system/vl.c
+++ b/system/vl.c
@@ -1660,6 +1660,7 @@ static const QEMUOption *lookup_opt(int argc, char **argv,
static MachineClass *select_machine(QDict *qdict, Error **errp)
{
const char *optarg = qdict_get_try_str(qdict, "type");
const char *machine_type = qdict_get_try_str(qdict, "type");
+ const char *pvever = qdict_get_try_str(qdict, "pvever");
GSList *machines = object_class_get_list(TYPE_MACHINE, false);
MachineClass *machine_class;
Error *local_err = NULL;
@@ -1614,6 +1615,11 @@ static MachineClass *select_machine(QDict *qdict, Error **errp)
@@ -1677,6 +1678,11 @@ static MachineClass *select_machine(QDict *qdict, Error **errp)
}
}
@ -95,7 +95,7 @@ index 3ee90b3b94..4b6d0b82fd 100644
g_slist_free(machines);
if (local_err) {
error_append_hint(&local_err, "Use -machine help to list supported machines\n");
@@ -3250,12 +3256,31 @@ void qemu_init(int argc, char **argv)
@@ -3319,12 +3325,31 @@ void qemu_init(int argc, char **argv)
case QEMU_OPTION_machine:
{
bool help;

6
debian/patches/pve/0026-block-backup-move-bcs-bitmap-initialization-to-job-c.patch vendored
View File

@ -25,7 +25,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/block/backup.c b/block/backup.c
index db3791f4d1..39410dcf8d 100644
index 8aae5836d7..2516eac5a7 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -237,8 +237,8 @@ static void backup_init_bcs_bitmap(BackupBlockJob *job)
@ -48,9 +48,9 @@ index db3791f4d1..39410dcf8d 100644
if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
int64_t offset = 0;
int64_t count;
@@ -495,6 +493,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
block_job_add_bdrv(&job->common, "target", target, 0, BLK_PERM_ALL,
@@ -501,6 +499,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
&error_abort);
bdrv_graph_wrunlock(target);
+ backup_init_bcs_bitmap(job);
+

50
debian/patches/pve/0027-PVE-Backup-add-vma-backup-format-code.patch vendored
View File

@ -15,21 +15,21 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/meson.build | 2 +
meson.build | 5 +
vma-reader.c | 867 ++++++++++++++++++++++++++++++++++++++++++++
vma-reader.c | 870 ++++++++++++++++++++++++++++++++++++++++++++
vma-writer.c | 818 +++++++++++++++++++++++++++++++++++++++++
vma.c | 900 ++++++++++++++++++++++++++++++++++++++++++++++
vma.c | 901 ++++++++++++++++++++++++++++++++++++++++++++++
vma.h | 150 ++++++++
6 files changed, 2742 insertions(+)
6 files changed, 2746 insertions(+)
create mode 100644 vma-reader.c
create mode 100644 vma-writer.c
create mode 100644 vma.c
create mode 100644 vma.h
diff --git a/block/meson.build b/block/meson.build
index 1833c71ce9..59b71ba9f3 100644
index 8ded0dc18b..e709b67d37 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -43,6 +43,8 @@ block_ss.add(files(
@@ -42,6 +42,8 @@ block_ss.add(files(
'zeroinit.c',
), zstd, zlib, gnutls)
@ -39,10 +39,10 @@ index 1833c71ce9..59b71ba9f3 100644
system_ss.add(files('block-ram-registrar.c'))
diff --git a/meson.build b/meson.build
index a9c4f28247..cd95530d3b 100644
index 6c77d9687d..8cb1ccd5e1 100644
--- a/meson.build
+++ b/meson.build
@@ -1778,6 +1778,8 @@ endif
@@ -1802,6 +1802,8 @@ endif
has_gettid = cc.has_function('gettid')
@ -51,7 +51,7 @@ index a9c4f28247..cd95530d3b 100644
# libselinux
selinux = dependency('libselinux',
required: get_option('selinux'),
@@ -3908,6 +3910,9 @@ if have_tools
@@ -3975,6 +3977,9 @@ if have_tools
dependencies: [blockdev, qemuutil, gnutls, selinux],
install: true)
@ -63,10 +63,10 @@ index a9c4f28247..cd95530d3b 100644
subdir('contrib/elf2dmp')
diff --git a/vma-reader.c b/vma-reader.c
new file mode 100644
index 0000000000..81a891c6b1
index 0000000000..d0b6721812
--- /dev/null
+++ b/vma-reader.c
@@ -0,0 +1,867 @@
@@ -0,0 +1,870 @@
+/*
+ * VMA: Virtual Machine Archive
+ *
@ -88,6 +88,7 @@ index 0000000000..81a891c6b1
+#include "qemu/ratelimit.h"
+#include "vma.h"
+#include "block/block.h"
+#include "block/graph-lock.h"
+#include "sysemu/block-backend.h"
+
+static unsigned char zero_vma_block[VMA_BLOCK_SIZE];
@ -600,8 +601,10 @@ index 0000000000..81a891c6b1
+ } else {
+ int res = blk_pwrite(target, sector_num * BDRV_SECTOR_SIZE, nb_sectors * BDRV_SECTOR_SIZE, buf, 0);
+ if (res < 0) {
+ bdrv_graph_rdlock_main_loop();
+ error_setg(errp, "blk_pwrite to %s failed (%d)",
+ bdrv_get_device_name(blk_bs(target)), res);
+ bdrv_graph_rdunlock_main_loop();
+ return -1;
+ }
+ }
@ -1760,10 +1763,10 @@ index 0000000000..126b296647
+}
diff --git a/vma.c b/vma.c
new file mode 100644
index 0000000000..347f6283ca
index 0000000000..bb715e9061
--- /dev/null
+++ b/vma.c
@@ -0,0 +1,900 @@
@@ -0,0 +1,901 @@
+/*
+ * VMA: Virtual Machine Archive
+ *
@ -2076,17 +2079,17 @@ index 0000000000..347f6283ca
+ inbuf);
+ }
+
+ RestoreMap *map = g_new0(RestoreMap, 1);
+ map->devname = g_strdup(devname);
+ map->path = g_strdup(path);
+ map->format = format;
+ map->throttling_bps = bps_value;
+ map->throttling_group = group;
+ map->cache = cache;
+ map->write_zero = write_zero;
+ map->skip = skip;
+ RestoreMap *restore_map = g_new0(RestoreMap, 1);
+ restore_map->devname = g_strdup(devname);
+ restore_map->path = g_strdup(path);
+ restore_map->format = format;
+ restore_map->throttling_bps = bps_value;
+ restore_map->throttling_group = group;
+ restore_map->cache = cache;
+ restore_map->write_zero = write_zero;
+ restore_map->skip = skip;
+
+ g_hash_table_insert(devmap, map->devname, map);
+ g_hash_table_insert(devmap, restore_map->devname, restore_map);
+
+ };
+ }
@ -2385,7 +2388,7 @@ index 0000000000..347f6283ca
+
+static int create_archive(int argc, char **argv)
+{
+ int i, c;
+ int c;
+ int verbose = 0;
+ const char *archivename;
+ GList *backup_coroutines = NULL;
@ -2543,6 +2546,7 @@ index 0000000000..347f6283ca
+ vma_writer_get_status(vmaw, &vmastat);
+
+ if (verbose) {
+ int i;
+ for (i = 0; i < 256; i++) {
+ VmaStreamInfo *si = &vmastat.stream_info[i];
+ if (si->size) {

14
debian/patches/pve/0028-PVE-Backup-add-backup-dump-block-driver.patch vendored
View File

@ -199,7 +199,7 @@ index 0000000000..e46abf1070
+ return bs;
+}
diff --git a/block/backup.c b/block/backup.c
index 39410dcf8d..af87fa6aa9 100644
index 2516eac5a7..aec140e0c8 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -29,28 +29,6 @@
@ -231,7 +231,7 @@ index 39410dcf8d..af87fa6aa9 100644
static const BlockJobDriver backup_job_driver;
static void backup_cleanup_sync_bitmap(BackupBlockJob *job, int ret)
@@ -457,6 +435,14 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
@@ -461,6 +439,14 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
}
cluster_size = block_copy_cluster_size(bcs);
@ -247,7 +247,7 @@ index 39410dcf8d..af87fa6aa9 100644
if (perf->max_chunk && perf->max_chunk < cluster_size) {
error_setg(errp, "Required max-chunk (%" PRIi64 ") is less than backup "
diff --git a/block/meson.build b/block/meson.build
index 59b71ba9f3..6fde9f7dcd 100644
index e709b67d37..f7d1b7ac42 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -4,6 +4,7 @@ block_ss.add(files(
@ -255,11 +255,11 @@ index 59b71ba9f3..6fde9f7dcd 100644
'amend.c',
'backup.c',
+ 'backup-dump.c',
'copy-before-write.c',
'blkdebug.c',
'blklogwrites.c',
'blkverify.c',
diff --git a/include/block/block_int-common.h b/include/block/block_int-common.h
index 74195c3004..0a0339eee4 100644
index 4e31d161c5..2fc6c37bc9 100644
--- a/include/block/block_int-common.h
+++ b/include/block/block_int-common.h
@@ -26,6 +26,7 @@
@ -312,10 +312,10 @@ index 74195c3004..0a0339eee4 100644
BDRV_TRACKED_READ,
BDRV_TRACKED_WRITE,
diff --git a/job.c b/job.c
index 72d57f0934..93e22d180b 100644
index 99a2e54b54..d7a18b8e6a 100644
--- a/job.c
+++ b/job.c
@@ -330,7 +330,8 @@ static bool job_started_locked(Job *job)
@@ -331,7 +331,8 @@ static bool job_started_locked(Job *job)
}
/* Called with job_mutex held. */

12
debian/patches/pve/0029-PVE-Add-sequential-job-transaction-support.patch vendored
View File

@ -11,7 +11,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2 files changed, 46 insertions(+)
diff --git a/include/qemu/job.h b/include/qemu/job.h
index e502787dd8..963cf2bef5 100644
index b4bc2e174b..4586dc2d3c 100644
--- a/include/qemu/job.h
+++ b/include/qemu/job.h
@@ -381,6 +381,18 @@ void job_unlock(void);
@ -34,10 +34,10 @@ index e502787dd8..963cf2bef5 100644
* Release a reference that was previously acquired with job_txn_add_job or
* job_txn_new. If it's the last reference to the object, it will be freed.
diff --git a/job.c b/job.c
index 93e22d180b..2b31f1e14f 100644
index d7a18b8e6a..868df1b4ce 100644
--- a/job.c
+++ b/job.c
@@ -93,6 +93,8 @@ struct JobTxn {
@@ -94,6 +94,8 @@ struct JobTxn {
/* Reference count */
int refcnt;
@ -46,7 +46,7 @@ index 93e22d180b..2b31f1e14f 100644
};
void job_lock(void)
@@ -118,6 +120,25 @@ JobTxn *job_txn_new(void)
@@ -119,6 +121,25 @@ JobTxn *job_txn_new(void)
return txn;
}
@ -72,7 +72,7 @@ index 93e22d180b..2b31f1e14f 100644
/* Called with job_mutex held. */
static void job_txn_ref_locked(JobTxn *txn)
{
@@ -1057,6 +1078,12 @@ static void job_completed_txn_success_locked(Job *job)
@@ -1058,6 +1079,12 @@ static void job_completed_txn_success_locked(Job *job)
*/
QLIST_FOREACH(other_job, &txn->jobs, txn_list) {
if (!job_is_completed_locked(other_job)) {
@ -85,7 +85,7 @@ index 93e22d180b..2b31f1e14f 100644
return;
}
assert(other_job->ret == 0);
@@ -1268,6 +1295,13 @@ int job_finish_sync_locked(Job *job,
@@ -1269,6 +1296,13 @@ int job_finish_sync_locked(Job *job,
return -EBUSY;
}

66
debian/patches/pve/0030-PVE-Backup-Proxmox-backup-patches-for-QEMU.patch vendored
View File

@ -94,20 +94,20 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
monitor/hmp-cmds.c | 72 +++
proxmox-backup-client.c | 146 +++++
proxmox-backup-client.h | 60 ++
pve-backup.c | 1089 ++++++++++++++++++++++++++++++++
pve-backup.c | 1103 ++++++++++++++++++++++++++++++++
qapi/block-core.json | 229 +++++++
qapi/common.json | 14 +
qapi/machine.json | 16 +-
14 files changed, 1704 insertions(+), 14 deletions(-)
14 files changed, 1718 insertions(+), 14 deletions(-)
create mode 100644 proxmox-backup-client.c
create mode 100644 proxmox-backup-client.h
create mode 100644 pve-backup.c
diff --git a/block/meson.build b/block/meson.build
index 6fde9f7dcd..6d468f89e5 100644
index f7d1b7ac42..9df99aceb5 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -45,6 +45,11 @@ block_ss.add(files(
@@ -44,6 +44,11 @@ block_ss.add(files(
), zstd, zlib, gnutls)
block_ss.add(files('../vma-writer.c'), libuuid)
@ -120,10 +120,10 @@ index 6fde9f7dcd..6d468f89e5 100644
system_ss.add(when: 'CONFIG_TCG', if_true: files('blkreplay.c'))
system_ss.add(files('block-ram-registrar.c'))
diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
index ca2599de44..6efe28cef5 100644
index c729cbf1eb..1656859e03 100644
--- a/block/monitor/block-hmp-cmds.c
+++ b/block/monitor/block-hmp-cmds.c
@@ -1029,3 +1029,42 @@ void hmp_change_medium(Monitor *mon, const char *device, const char *target,
@@ -1037,3 +1037,42 @@ void hmp_change_medium(Monitor *mon, const char *device, const char *target,
qmp_blockdev_change_medium(device, NULL, target, arg, true, force,
!!read_only, read_only_mode, errp);
}
@ -167,7 +167,7 @@ index ca2599de44..6efe28cef5 100644
+ hmp_handle_error(mon, error);
+}
diff --git a/blockdev.c b/blockdev.c
index cd5f205ad1..7793143d76 100644
index 38a40e3e32..3049811be8 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -37,6 +37,7 @@
@ -204,7 +204,7 @@ index 10fdd822e0..15937793c1 100644
{
.name = "usernet",
diff --git a/hmp-commands.hx b/hmp-commands.hx
index e352f86872..0c8b6725fb 100644
index 893c3bd240..5c1ffbc602 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -101,6 +101,35 @@ ERST
@ -265,10 +265,10 @@ index 7a7def7530..cba7afe70c 100644
void hmp_device_add(Monitor *mon, const QDict *qdict);
void hmp_device_del(Monitor *mon, const QDict *qdict);
diff --git a/meson.build b/meson.build
index cd95530d3b..d53976d621 100644
index 8cb1ccd5e1..955f579308 100644
--- a/meson.build
+++ b/meson.build
@@ -1779,6 +1779,7 @@ endif
@@ -1803,6 +1803,7 @@ endif
has_gettid = cc.has_function('gettid')
libuuid = cc.find_library('uuid', required: true)
@ -277,7 +277,7 @@ index cd95530d3b..d53976d621 100644
# libselinux
selinux = dependency('libselinux',
diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
index 91be698308..5b9c231a4c 100644
index ef4634e5c1..6e25279f42 100644
--- a/monitor/hmp-cmds.c
+++ b/monitor/hmp-cmds.c
@@ -21,6 +21,7 @@
@ -586,10 +586,10 @@ index 0000000000..8cbf645b2c
+#endif /* PROXMOX_BACKUP_CLIENT_H */
diff --git a/pve-backup.c b/pve-backup.c
new file mode 100644
index 0000000000..ae3d137e12
index 0000000000..903afcd7e9
--- /dev/null
+++ b/pve-backup.c
@@ -0,0 +1,1089 @@
@@ -0,0 +1,1103 @@
+#include "proxmox-backup-client.h"
+#include "vma.h"
+
@ -600,6 +600,7 @@ index 0000000000..ae3d137e12
+#include "block/block_int-global-state.h"
+#include "block/blockjob.h"
+#include "block/dirty-bitmap.h"
+#include "block/graph-lock.h"
+#include "qapi/qapi-commands-block.h"
+#include "qapi/qmp/qerror.h"
+#include "qemu/cutils.h"
@ -928,13 +929,6 @@ index 0000000000..ae3d137e12
+ }
+ }
+
+ if (di->job) {
+ WITH_JOB_LOCK_GUARD() {
+ job_unref_locked(&di->job->job);
+ di->job = NULL;
+ }
+ }
+
+ // remove self from job list
+ backup_state.di_list = g_list_remove(backup_state.di_list, di);
+
@ -954,6 +948,16 @@ index 0000000000..ae3d137e12
+ di->completed_ret = ret;
+
+ /*
+ * Needs to happen outside of coroutine, because it takes the graph write lock.
+ */
+ if (di->job) {
+ WITH_JOB_LOCK_GUARD() {
+ job_unref_locked(&di->job->job);
+ di->job = NULL;
+ }
+ }
+
+ /*
+ * Schedule stream cleanup in async coroutine. close_image and finish might
+ * take a while, so we can't block on them here. This way it also doesn't
+ * matter if we're already running in a coroutine or not.
@ -1178,7 +1182,7 @@ index 0000000000..ae3d137e12
+ * case of an error, errp will be set, but the returned value might still be a
+ * list.
+ */
+static GList coroutine_fn *get_device_info(
+static GList coroutine_fn GRAPH_RDLOCK *get_device_info(
+ const char *devlist,
+ Error **errp)
+{
@ -1279,7 +1283,9 @@ index 0000000000..ae3d137e12
+ /* Todo: try to auto-detect format based on file name */
+ format = has_format ? format : BACKUP_FORMAT_VMA;
+
+ bdrv_graph_co_rdlock();
+ di_list = get_device_info(devlist, &local_err);
+ bdrv_graph_co_rdunlock();
+ if (local_err) {
+ error_propagate(errp, local_err);
+ goto err;
@ -1292,7 +1298,11 @@ index 0000000000..ae3d137e12
+ while (l) {
+ PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+ l = g_list_next(l);
+ if (bdrv_op_is_blocked(di->bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp)) {
+
+ bdrv_graph_co_rdlock();
+ bool blocked = bdrv_op_is_blocked(di->bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp);
+ bdrv_graph_co_rdunlock();
+ if (blocked) {
+ goto err;
+ }
+
@ -1376,7 +1386,9 @@ index 0000000000..ae3d137e12
+
+ di->block_size = dump_cb_block_size;
+
+ bdrv_graph_co_rdlock();
+ const char *devname = bdrv_get_device_name(di->bs);
+ bdrv_graph_co_rdunlock();
+ PBSBitmapAction action = PBS_BITMAP_ACTION_NOT_USED;
+ size_t dirty = di->size;
+
@ -1452,7 +1464,9 @@ index 0000000000..ae3d137e12
+ goto err_mutex;
+ }
+
+ bdrv_graph_co_rdlock();
+ const char *devname = bdrv_get_device_name(di->bs);
+ bdrv_graph_co_rdunlock();
+ di->dev_id = vma_writer_register_stream(vmaw, devname, di->size);
+ if (di->dev_id <= 0) {
+ error_set(errp, ERROR_CLASS_GENERIC_ERROR,
@ -1680,10 +1694,10 @@ index 0000000000..ae3d137e12
+ return ret;
+}
diff --git a/qapi/block-core.json b/qapi/block-core.json
index bb471c078d..1b8462a51b 100644
index 299e3fc350..c155d74230 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -839,6 +839,235 @@
@@ -841,6 +841,235 @@
{ 'command': 'query-block', 'returns': ['BlockInfo'],
'allow-preconfig': true }
@ -1942,7 +1956,7 @@ index 6fed9cde1a..630a2a8f9a 100644
+##
+{ 'struct': 'UuidInfo', 'data': {'UUID': 'str'} }
diff --git a/qapi/machine.json b/qapi/machine.json
index 7da3c519ba..888457f810 100644
index a9fd40d844..d97f024173 100644
--- a/qapi/machine.json
+++ b/qapi/machine.json
@@ -4,6 +4,8 @@
@ -1954,7 +1968,7 @@ index 7da3c519ba..888457f810 100644
##
# = Machines
##
@@ -230,20 +232,6 @@
@@ -237,20 +239,6 @@
##
{ 'command': 'query-target', 'returns': 'TargetInfo' }

4
debian/patches/pve/0031-PVE-Backup-pbs-restore-new-command-to-restore-from-p.patch vendored
View File

@ -14,10 +14,10 @@ Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
create mode 100644 pbs-restore.c
diff --git a/meson.build b/meson.build
index d53976d621..c3330310d9 100644
index 955f579308..c70f2a4937 100644
--- a/meson.build
+++ b/meson.build
@@ -3914,6 +3914,10 @@ if have_tools
@@ -3981,6 +3981,10 @@ if have_tools
vma = executable('vma', files('vma.c', 'vma-reader.c') + genh,
dependencies: [authz, block, crypto, io, qom], install: true)

91
debian/patches/pve/0032-PVE-Add-PBS-block-driver-to-map-backup-archives-into.patch vendored
View File

@ -14,35 +14,33 @@ Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
getlength is now a coroutine function]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/meson.build | 3 +
block/pbs.c | 305 +++++++++++++++++++++++++++++++++++++++++++
configure | 9 ++
block/meson.build | 2 +
block/pbs.c | 307 +++++++++++++++++++++++++++++++++++++++++++
meson.build | 2 +-
qapi/block-core.json | 13 ++
qapi/pragma.json | 1 +
6 files changed, 332 insertions(+), 1 deletion(-)
5 files changed, 324 insertions(+), 1 deletion(-)
create mode 100644 block/pbs.c
diff --git a/block/meson.build b/block/meson.build
index 6d468f89e5..becc99ac4e 100644
index 9df99aceb5..549c0c7103 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -50,6 +50,9 @@ block_ss.add(files(
@@ -49,6 +49,8 @@ block_ss.add(files(
'../pve-backup.c',
), libproxmox_backup_qemu)
+block_ss.add(when: 'CONFIG_PBS_BDRV', if_true: files('pbs.c'))
+block_ss.add(when: 'CONFIG_PBS_BDRV', if_true: libproxmox_backup_qemu)
+block_ss.add(files('pbs.c'), libproxmox_backup_qemu)
+
system_ss.add(when: 'CONFIG_TCG', if_true: files('blkreplay.c'))
system_ss.add(files('block-ram-registrar.c'))
diff --git a/block/pbs.c b/block/pbs.c
new file mode 100644
index 0000000000..a2211e0f3b
index 0000000000..dd72356bd3
--- /dev/null
+++ b/block/pbs.c
@@ -0,0 +1,305 @@
@@ -0,0 +1,307 @@
+/*
+ * Proxmox Backup Server read-only block driver
+ */
@ -234,7 +232,8 @@ index 0000000000..a2211e0f3b
+ proxmox_restore_disconnect(s->conn);
+}
+
+static coroutine_fn int64_t pbs_co_getlength(BlockDriverState *bs)
+static coroutine_fn int64_t GRAPH_RDLOCK
+pbs_co_getlength(BlockDriverState *bs)
+{
+ BDRVPBSState *s = bs->opaque;
+ return s->length;
@ -251,9 +250,9 @@ index 0000000000..a2211e0f3b
+ aio_co_schedule(rcb->ctx, rcb->co);
+}
+
+static coroutine_fn int pbs_co_preadv(BlockDriverState *bs,
+ int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+static coroutine_fn int GRAPH_RDLOCK
+pbs_co_preadv(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ BDRVPBSState *s = bs->opaque;
+ int ret;
@ -298,16 +297,17 @@ index 0000000000..a2211e0f3b
+ return 0;
+}
+
+static coroutine_fn int pbs_co_pwritev(BlockDriverState *bs,
+ int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+static coroutine_fn int GRAPH_RDLOCK
+pbs_co_pwritev(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ fprintf(stderr, "pbs-bdrv: cannot write to backup file, make sure "
+ "any attached disk devices are set to read-only!\n");
+ return -EPERM;
+}
+
+static void pbs_refresh_filename(BlockDriverState *bs)
+static void GRAPH_RDLOCK
+pbs_refresh_filename(BlockDriverState *bs)
+{
+ BDRVPBSState *s = bs->opaque;
+ if (s->namespace) {
@ -348,52 +348,11 @@ index 0000000000..a2211e0f3b
+}
+
+block_init(bdrv_pbs_init);
diff --git a/configure b/configure
index 133f4e3235..f5a830c1f3 100755
--- a/configure
+++ b/configure
@@ -256,6 +256,7 @@ qemu_suffix="qemu"
softmmu="yes"
linux_user=""
bsd_user=""
+pbs_bdrv="yes"
plugins="$default_feature"
ninja=""
python=
@@ -809,6 +810,10 @@ for opt do
;;
--enable-download) download="enabled"; git_submodules_action=update;
;;
+ --disable-pbs-bdrv) pbs_bdrv="no"
+ ;;
+ --enable-pbs-bdrv) pbs_bdrv="yes"
+ ;;
--enable-plugins) if test "$mingw32" = "yes"; then
error_exit "TCG plugins not currently supported on Windows platforms"
else
@@ -959,6 +964,7 @@ cat << EOF
bsd-user all BSD usermode emulation targets
pie Position Independent Executables
debug-tcg TCG debugging (default is disabled)
+ pbs-bdrv Proxmox backup server read-only block driver support
NOTE: The object files are built at the place where configure is launched
EOF
@@ -1744,6 +1750,9 @@ if test "$solaris" = "yes" ; then
fi
echo "SRC_PATH=$source_path" >> $config_host_mak
echo "TARGET_DIRS=$target_list" >> $config_host_mak
+if test "$pbs_bdrv" = "yes" ; then
+ echo "CONFIG_PBS_BDRV=y" >> $config_host_mak
+fi
# XXX: suppress that
if [ "$bsd" = "yes" ] ; then
diff --git a/meson.build b/meson.build
index c3330310d9..cbfc9a43fb 100644
index c70f2a4937..18c3a312eb 100644
--- a/meson.build
+++ b/meson.build
@@ -4319,7 +4319,7 @@ summary_info += {'bzip2 support': libbzip2}
@@ -4390,7 +4390,7 @@ summary_info += {'bzip2 support': libbzip2}
summary_info += {'lzfse support': liblzfse}
summary_info += {'zstd support': zstd}
summary_info += {'NUMA host support': numa}
@ -403,10 +362,10 @@ index c3330310d9..cbfc9a43fb 100644
summary_info += {'libdaxctl support': libdaxctl}
summary_info += {'libudev': libudev}
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 1b8462a51b..d67a6d448a 100644
index c155d74230..a4050268ca 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -3396,6 +3396,7 @@
@@ -3451,6 +3451,7 @@
'parallels', 'preallocate', 'qcow', 'qcow2', 'qed', 'quorum',
'raw', 'rbd',
{ 'name': 'replication', 'if': 'CONFIG_REPLICATION' },
@ -414,7 +373,7 @@ index 1b8462a51b..d67a6d448a 100644
'ssh', 'throttle', 'vdi', 'vhdx',
{ 'name': 'virtio-blk-vfio-pci', 'if': 'CONFIG_BLKIO' },
{ 'name': 'virtio-blk-vhost-user', 'if': 'CONFIG_BLKIO' },
@@ -3482,6 +3483,17 @@
@@ -3537,6 +3538,17 @@
{ 'struct': 'BlockdevOptionsNull',
'data': { '*size': 'int', '*latency-ns': 'uint64', '*read-zeroes': 'bool' } }
@ -432,7 +391,7 @@ index 1b8462a51b..d67a6d448a 100644
##
# @BlockdevOptionsNVMe:
#
@@ -4890,6 +4902,7 @@
@@ -4945,6 +4957,7 @@
'nfs': 'BlockdevOptionsNfs',
'null-aio': 'BlockdevOptionsNull',
'null-co': 'BlockdevOptionsNull',
@ -441,7 +400,7 @@ index 1b8462a51b..d67a6d448a 100644
'nvme-io_uring': { 'type': 'BlockdevOptionsNvmeIoUring',
'if': 'CONFIG_BLKIO' },
diff --git a/qapi/pragma.json b/qapi/pragma.json
index 325e684411..b6079f6a0e 100644
index eae9f54700..03467983b3 100644
--- a/qapi/pragma.json
+++ b/qapi/pragma.json
@@ -45,6 +45,7 @@

14
debian/patches/pve/0033-PVE-redirect-stderr-to-journal-when-daemonized.patch vendored
View File

@ -14,10 +14,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/meson.build b/meson.build
index cbfc9a43fb..8206270272 100644
index 18c3a312eb..b598a0450b 100644
--- a/meson.build
+++ b/meson.build
@@ -1779,6 +1779,7 @@ endif
@@ -1803,6 +1803,7 @@ endif
has_gettid = cc.has_function('gettid')
libuuid = cc.find_library('uuid', required: true)
@ -25,7 +25,7 @@ index cbfc9a43fb..8206270272 100644
libproxmox_backup_qemu = cc.find_library('proxmox_backup_qemu', required: true)
# libselinux
@@ -3406,6 +3407,7 @@ if have_block
@@ -3468,6 +3469,7 @@ if have_block
# os-posix.c contains POSIX-specific functions used by qemu-storage-daemon,
# os-win32.c does not
blockdev_ss.add(when: 'CONFIG_POSIX', if_true: files('os-posix.c'))
@ -34,7 +34,7 @@ index cbfc9a43fb..8206270272 100644
endif
diff --git a/os-posix.c b/os-posix.c
index 0cc1d991b1..f33d9901cf 100644
index a4284e2c07..197a2120fd 100644
--- a/os-posix.c
+++ b/os-posix.c
@@ -29,6 +29,8 @@
@ -44,9 +44,9 @@ index 0cc1d991b1..f33d9901cf 100644
+#include <systemd/sd-journal.h>
+#include <syslog.h>
/* Needed early for CONFIG_BSD etc. */
#include "net/slirp.h"
@@ -332,9 +334,10 @@ void os_setup_post(void)
#include "qemu/error-report.h"
#include "qemu/log.h"
@@ -302,9 +304,10 @@ void os_setup_post(void)
dup2(fd, 0);
dup2(fd, 1);

20
debian/patches/pve/0034-PVE-Migrate-dirty-bitmap-state-via-savevm.patch vendored
View File

@ -26,10 +26,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
create mode 100644 migration/pbs-state.c
diff --git a/include/migration/misc.h b/include/migration/misc.h
index 7dcc0b5c2c..4c940b2475 100644
index 1bc8902e6d..43eb0e46d6 100644
--- a/include/migration/misc.h
+++ b/include/migration/misc.h
@@ -77,4 +77,7 @@ bool migration_in_bg_snapshot(void);
@@ -80,4 +80,7 @@ bool migration_in_bg_snapshot(void);
/* migration/block-dirty-bitmap.c */
void dirty_bitmap_mig_init(void);
@ -38,7 +38,7 @@ index 7dcc0b5c2c..4c940b2475 100644
+
#endif
diff --git a/migration/meson.build b/migration/meson.build
index 07f6057acc..343994d891 100644
index 8f9d122187..2b9d2cc881 100644
--- a/migration/meson.build
+++ b/migration/meson.build
@@ -7,7 +7,9 @@ migration_files = files(
@ -52,10 +52,10 @@ index 07f6057acc..343994d891 100644
system_ss.add(files(
'block-dirty-bitmap.c',
diff --git a/migration/migration.c b/migration/migration.c
index 7a4c8beb5d..0a955a2a18 100644
index 982ab85f04..b84924442d 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -162,6 +162,7 @@ void migration_object_init(void)
@@ -202,6 +202,7 @@ void migration_object_init(void)
blk_mig_init();
ram_mig_init();
dirty_bitmap_mig_init();
@ -174,10 +174,10 @@ index 0000000000..887e998b9e
+ NULL);
+}
diff --git a/pve-backup.c b/pve-backup.c
index ae3d137e12..e6b17b797e 100644
index 903afcd7e9..777db7938e 100644
--- a/pve-backup.c
+++ b/pve-backup.c
@@ -1082,6 +1082,7 @@ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
@@ -1096,6 +1096,7 @@ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
ret->pbs_library_version = g_strdup(proxmox_backup_qemu_version());
ret->pbs_dirty_bitmap = true;
ret->pbs_dirty_bitmap_savevm = true;
@ -186,10 +186,10 @@ index ae3d137e12..e6b17b797e 100644
ret->pbs_masterkey = true;
ret->backup_max_workers = true;
diff --git a/qapi/block-core.json b/qapi/block-core.json
index d67a6d448a..09de550c95 100644
index a4050268ca..7b977459fa 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -991,6 +991,11 @@
@@ -993,6 +993,11 @@
# @pbs-dirty-bitmap-savevm: True if 'dirty-bitmaps' migration capability can
# safely be set for savevm-async.
#
@ -201,7 +201,7 @@ index d67a6d448a..09de550c95 100644
# @pbs-masterkey: True if the QMP backup call supports the 'master_keyfile'
# parameter.
#
@@ -1001,6 +1006,7 @@
@@ -1003,6 +1008,7 @@
'data': { 'pbs-dirty-bitmap': 'bool',
'query-bitmap-info': 'bool',
'pbs-dirty-bitmap-savevm': 'bool',

2
debian/patches/pve/0035-migration-block-dirty-bitmap-migrate-other-bitmaps-e.patch vendored
View File

@ -19,7 +19,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/migration/block-dirty-bitmap.c b/migration/block-dirty-bitmap.c
index e1ae3b7316..285dd1d148 100644
index 0070b13b6f..954be621ba 100644
--- a/migration/block-dirty-bitmap.c
+++ b/migration/block-dirty-bitmap.c
@@ -540,7 +540,7 @@ static int add_bitmaps_to_list(DBMSaveState *s, BlockDriverState *bs,

4
debian/patches/pve/0036-PVE-fall-back-to-open-iscsi-initiatorname.patch vendored
View File

@ -21,10 +21,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 30 insertions(+)
diff --git a/block/iscsi.c b/block/iscsi.c
index 34f97ab646..398782963d 100644
index 2ff14b7472..46f275fbf7 100644
--- a/block/iscsi.c
+++ b/block/iscsi.c
@@ -1391,12 +1391,42 @@ static char *get_initiator_name(QemuOpts *opts)
@@ -1392,12 +1392,42 @@ static char *get_initiator_name(QemuOpts *opts)
const char *name;
char *iscsi_name;
UuidInfo *uuid_info;

2
debian/patches/pve/0037-PVE-block-stream-increase-chunk-size.patch vendored
View File

@ -11,7 +11,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/stream.c b/block/stream.c
index e522bbdec5..afed72db55 100644
index 01fe7c0f16..87a97cb8a2 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -27,7 +27,7 @@ enum {

4
debian/patches/pve/0038-block-io-accept-NULL-qiov-in-bdrv_pad_request.patch vendored
View File

@ -20,10 +20,10 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
1 file changed, 16 insertions(+), 13 deletions(-)
diff --git a/block/io.c b/block/io.c
index 83d1b1dfdc..e927881e40 100644
index d202987770..d42c811bd7 100644
--- a/block/io.c
+++ b/block/io.c
@@ -1723,22 +1723,25 @@ static int bdrv_pad_request(BlockDriverState *bs,
@@ -1756,22 +1756,25 @@ static int bdrv_pad_request(BlockDriverState *bs,
return 0;
}

243
debian/patches/pve/0039-block-add-alloc-track-driver.patch vendored
View File

@ -19,27 +19,33 @@ well.
This only worked if the target supports backing images, so up until now
only for qcow2, with alloc-track any driver for the target can be used.
If 'auto-remove' is set, alloc-track will automatically detach itself
once the backing image is removed. It will be replaced by 'file'.
Replacing the node cannot be done in the
track_co_change_backing_file() callback, because replacing a node
cannot happen in a coroutine and requires the block graph lock
exclusively. Could either become a special option for the stream job,
or maybe the upcoming blockdev-replace QMP command can be used in the
future.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
[FE: adapt to changed function signatures
make error return value consistent with QEMU
avoid premature break during read]
avoid premature break during read
adhere to block graph lock requirements]
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/alloc-track.c | 352 ++++++++++++++++++++++++++++++++++++++++++++
block/alloc-track.c | 366 ++++++++++++++++++++++++++++++++++++++++++++
block/meson.build | 1 +
2 files changed, 353 insertions(+)
block/stream.c | 34 ++++
3 files changed, 401 insertions(+)
create mode 100644 block/alloc-track.c
diff --git a/block/alloc-track.c b/block/alloc-track.c
new file mode 100644
index 0000000000..b75d7c6460
index 0000000000..14698c362e
--- /dev/null
+++ b/block/alloc-track.c
@@ -0,0 +1,352 @@
@@ -0,0 +1,366 @@
+/*
+ * Node to allow backing images to be applied to any node. Assumes a blank
+ * image to begin with, only new writes are tracked as allocated, thus this
@ -56,9 +62,11 @@ index 0000000000..b75d7c6460
+#include "qapi/error.h"
+#include "block/block_int.h"
+#include "block/dirty-bitmap.h"
+#include "block/graph-lock.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qstring.h"
+#include "qemu/cutils.h"
+#include "qemu/error-report.h"
+#include "qemu/option.h"
+#include "qemu/module.h"
+#include "sysemu/block-backend.h"
@ -67,12 +75,12 @@ index 0000000000..b75d7c6460
+
+typedef enum DropState {
+ DropNone,
+ DropRequested,
+ DropInProgress,
+} DropState;
+
+typedef struct {
+ BdrvDirtyBitmap *bitmap;
+ uint64_t granularity;
+ DropState drop_state;
+ bool auto_remove;
+} BDRVAllocTrackState;
@ -91,26 +99,29 @@ index 0000000000..b75d7c6460
+ },
+};
+
+static void track_refresh_limits(BlockDriverState *bs, Error **errp)
+static void GRAPH_RDLOCK
+track_refresh_limits(BlockDriverState *bs, Error **errp)
+{
+ BlockDriverInfo bdi;
+ BDRVAllocTrackState *s = bs->opaque;
+
+ if (!bs->file) {
+ return;
+ }
+
+ /* always use alignment from underlying write device so RMW cycle for
+ * bdrv_pwritev reads data from our backing via track_co_preadv (no partial
+ * cluster allocation in 'file') */
+ bdrv_get_info(bs->file->bs, &bdi);
+ /*
+ * Always use alignment from underlying write device so RMW cycle for
+ * bdrv_pwritev reads data from our backing via track_co_preadv. Also use at
+ * least the bitmap granularity.
+ */
+ bs->bl.request_alignment = MAX(bs->file->bs->bl.request_alignment,
+ MAX(bdi.cluster_size, BDRV_SECTOR_SIZE));
+ s->granularity);
+}
+
+static int track_open(BlockDriverState *bs, QDict *options, int flags,
+ Error **errp)
+{
+ BDRVAllocTrackState *s = bs->opaque;
+ BdrvChild *file = NULL;
+ QemuOpts *opts;
+ Error *local_err = NULL;
+ int ret = 0;
@ -126,18 +137,45 @@ index 0000000000..b75d7c6460
+ s->auto_remove = qemu_opt_get_bool(opts, TRACK_OPT_AUTO_REMOVE, false);
+
+ /* open the target (write) node, backing will be attached by block layer */
+ bs->file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds,
+ BDRV_CHILD_DATA | BDRV_CHILD_METADATA, false,
+ &local_err);
+ file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds,
+ BDRV_CHILD_DATA | BDRV_CHILD_METADATA, false,
+ &local_err);
+ bdrv_graph_wrlock(bs);
+ bs->file = file;
+ bdrv_graph_wrunlock(bs);
+ if (local_err) {
+ ret = -EINVAL;
+ error_propagate(errp, local_err);
+ goto fail;
+ }
+
+ bdrv_graph_rdlock_main_loop();
+ BlockDriverInfo bdi = {0};
+ ret = bdrv_get_info(bs->file->bs, &bdi);
+ if (ret < 0) {
+ /*
+ * Not a hard failure. Worst that can happen is partial cluster
+ * allocation in the write target. However, the driver here returns its
+ * allocation status based on the dirty bitmap, so any other data that
+ * maps to such a cluster will still be copied later by a stream job (or
+ * during writes to that cluster).
+ */
+ warn_report("alloc-track: unable to query cluster size for write target: %s",
+ strerror(ret));
+ }
+ ret = 0;
+ /*
+ * Always consider alignment from underlying write device so RMW cycle for
+ * bdrv_pwritev reads data from our backing via track_co_preadv. Also try to
+ * avoid partial cluster allocation in the write target by considering the
+ * cluster size.
+ */
+ s->granularity = MAX(bs->file->bs->bl.request_alignment,
+ MAX(bdi.cluster_size, BDRV_SECTOR_SIZE));
+ track_refresh_limits(bs, errp);
+ uint64_t gran = bs->bl.request_alignment;
+ s->bitmap = bdrv_create_dirty_bitmap(bs->file->bs, gran, NULL, &local_err);
+ s->bitmap = bdrv_create_dirty_bitmap(bs->file->bs, s->granularity, NULL,
+ &local_err);
+ bdrv_graph_rdunlock_main_loop();
+ if (local_err) {
+ ret = -EIO;
+ error_propagate(errp, local_err);
@ -148,7 +186,9 @@ index 0000000000..b75d7c6460
+
+fail:
+ if (ret < 0) {
+ bdrv_graph_wrlock(bs);
+ bdrv_unref_child(bs, bs->file);
+ bdrv_graph_wrunlock(bs);
+ if (s->bitmap) {
+ bdrv_release_dirty_bitmap(s->bitmap);
+ }
@ -165,13 +205,15 @@ index 0000000000..b75d7c6460
+ }
+}
+
+static coroutine_fn int64_t track_co_getlength(BlockDriverState *bs)
+static coroutine_fn int64_t GRAPH_RDLOCK
+track_co_getlength(BlockDriverState *bs)
+{
+ return bdrv_co_getlength(bs->file->bs);
+}
+
+static int coroutine_fn track_co_preadv(BlockDriverState *bs,
+ int64_t offset, int64_t bytes, QEMUIOVector *qiov, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_preadv(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ BDRVAllocTrackState *s = bs->opaque;
+ QEMUIOVector local_qiov;
@ -229,36 +271,39 @@ index 0000000000..b75d7c6460
+ return ret;
+}
+
+static int coroutine_fn track_co_pwritev(BlockDriverState *bs,
+ int64_t offset, int64_t bytes, QEMUIOVector *qiov, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_pwritev(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ QEMUIOVector *qiov, BdrvRequestFlags flags)
+{
+ return bdrv_co_pwritev(bs->file, offset, bytes, qiov, flags);
+}
+
+static int coroutine_fn track_co_pwrite_zeroes(BlockDriverState *bs,
+ int64_t offset, int64_t bytes, BdrvRequestFlags flags)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_pwrite_zeroes(BlockDriverState *bs, int64_t offset, int64_t bytes,
+ BdrvRequestFlags flags)
+{
+ return bdrv_co_pwrite_zeroes(bs->file, offset, bytes, flags);
+}
+
+static int coroutine_fn track_co_pdiscard(BlockDriverState *bs,
+ int64_t offset, int64_t bytes)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_pdiscard(BlockDriverState *bs, int64_t offset, int64_t bytes)
+{
+ return bdrv_co_pdiscard(bs->file, offset, bytes);
+}
+
+static coroutine_fn int track_co_flush(BlockDriverState *bs)
+static coroutine_fn int GRAPH_RDLOCK
+track_co_flush(BlockDriverState *bs)
+{
+ return bdrv_co_flush(bs->file->bs);
+}
+
+static int coroutine_fn track_co_block_status(BlockDriverState *bs,
+ bool want_zero,
+ int64_t offset,
+ int64_t bytes,
+ int64_t *pnum,
+ int64_t *map,
+ BlockDriverState **file)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_block_status(BlockDriverState *bs, bool want_zero,
+ int64_t offset,
+ int64_t bytes,
+ int64_t *pnum,
+ int64_t *map,
+ BlockDriverState **file)
+{
+ BDRVAllocTrackState *s = bs->opaque;
+
@ -284,10 +329,10 @@ index 0000000000..b75d7c6460
+ return 0;
+}
+
+static void track_child_perm(BlockDriverState *bs, BdrvChild *c,
+ BdrvChildRole role, BlockReopenQueue *reopen_queue,
+ uint64_t perm, uint64_t shared,
+ uint64_t *nperm, uint64_t *nshared)
+static void GRAPH_RDLOCK
+track_child_perm(BlockDriverState *bs, BdrvChild *c, BdrvChildRole role,
+ BlockReopenQueue *reopen_queue, uint64_t perm, uint64_t shared,
+ uint64_t *nperm, uint64_t *nshared)
+{
+ BDRVAllocTrackState *s = bs->opaque;
+
@ -310,53 +355,28 @@ index 0000000000..b75d7c6460
+ }
+}
+
+static void track_drop(void *opaque)
+static int coroutine_fn GRAPH_RDLOCK
+track_co_change_backing_file(BlockDriverState *bs, const char *backing_file,
+ const char *backing_fmt)
+{
+ BlockDriverState *bs = (BlockDriverState*)opaque;
+ BlockDriverState *file = bs->file->bs;
+ BDRVAllocTrackState *s = bs->opaque;
+
+ assert(file);
+
+ /* we rely on the fact that we're not used anywhere else, so let's wait
+ * until we're only used once - in the drive connected to the guest (and one
+ * ref is held by bdrv_ref in track_change_backing_file) */
+ if (bs->refcnt > 2) {
+ aio_bh_schedule_oneshot(qemu_get_aio_context(), track_drop, opaque);
+ return;
+ }
+ AioContext *aio_context = bdrv_get_aio_context(bs);
+ aio_context_acquire(aio_context);
+
+ bdrv_drained_begin(bs);
+
+ /* now that we're drained, we can safely set 'DropInProgress' */
+ s->drop_state = DropInProgress;
+ bdrv_child_refresh_perms(bs, bs->file, &error_abort);
+
+ bdrv_replace_node(bs, file, &error_abort);
+ bdrv_set_backing_hd(bs, NULL, &error_abort);
+ bdrv_drained_end(bs);
+ bdrv_unref(bs);
+ aio_context_release(aio_context);
+}
+
+static int track_change_backing_file(BlockDriverState *bs,
+ const char *backing_file,
+ const char *backing_fmt)
+{
+ BDRVAllocTrackState *s = bs->opaque;
+ if (s->auto_remove && s->drop_state == DropNone &&
+ backing_file == NULL && backing_fmt == NULL)
+ {
+ /* backing file has been disconnected, there's no longer any use for
+ * this node, so let's remove ourselves from the block graph - we need
+ * to schedule this for later however, since when this function is
+ * called, the blockjob modifying us is probably not done yet and has a
+ * blocker on 'bs' */
+ s->drop_state = DropRequested;
+ bdrv_ref(bs);
+ aio_bh_schedule_oneshot(qemu_get_aio_context(), track_drop, (void*)bs);
+ /*
+ * Note that the actual backing file graph change is already done in the
+ * stream job itself with bdrv_set_backing_hd_drained(), so no need to
+ * actually do anything here. But still needs to be implemented, to make
+ * our caller (i.e. bdrv_co_change_backing_file() do the right thing).
+ *
+ * FIXME
+ * We'd like to auto-remove ourselves from the block graph, but it cannot
+ * be done from a coroutine. Currently done in the stream job, where it
+ * kinda fits better, but in the long-term, a special parameter would be
+ * nice (or done via qemu-server via upcoming blockdev-replace QMP command).
+ */
+ if (backing_file == NULL) {
+ BDRVAllocTrackState *s = bs->opaque;
+ bdrv_drained_begin(bs);
+ s->drop_state = DropInProgress;
+ bdrv_child_refresh_perms(bs, bs->file, &error_abort);
+ bdrv_drained_end(bs);
+ }
+
+ return 0;
@ -383,7 +403,7 @@ index 0000000000..b75d7c6460
+ .supports_backing = true,
+
+ .bdrv_co_block_status = track_co_block_status,
+ .bdrv_change_backing_file = track_change_backing_file,
+ .bdrv_co_change_backing_file = track_co_change_backing_file,
+};
+
+static void bdrv_alloc_track_init(void)
@ -393,7 +413,7 @@ index 0000000000..b75d7c6460
+
+block_init(bdrv_alloc_track_init);
diff --git a/block/meson.build b/block/meson.build
index becc99ac4e..0a69836593 100644
index 549c0c7103..73777a1620 100644
--- a/block/meson.build
+++ b/block/meson.build
@@ -2,6 +2,7 @@ block_ss.add(genh)
@ -404,3 +424,48 @@ index becc99ac4e..0a69836593 100644
'amend.c',
'backup.c',
'backup-dump.c',
diff --git a/block/stream.c b/block/stream.c
index 87a97cb8a2..3a04e95ee2 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -114,6 +114,40 @@ static int stream_prepare(Job *job)
ret = -EPERM;
goto out;
}
+
+ /*
+ * This cannot be done in the co_change_backing_file callback, because
+ * bdrv_replace_node() cannot be done in a coroutine. The latter also
+ * requires the graph lock exclusively. Only required for the
+ * alloc-track driver.
+ *
+ * The long-term plan is to either have an explicit parameter for the
+ * stream job or use the upcoming blockdev-replace QMP command.
+ */
+ if (base_id == NULL && strcmp(unfiltered_bs->drv->format_name, "alloc-track") == 0) {
+ BlockDriverState *file_bs;
+
+ bdrv_graph_rdlock_main_loop();
+ file_bs = unfiltered_bs->file->bs;
+ bdrv_graph_rdunlock_main_loop();
+
+ bdrv_ref(unfiltered_bs); // unrefed by bdrv_replace_node()
+ bdrv_drained_begin(file_bs);
+ bdrv_graph_wrlock(s->target_bs);
+
+ bdrv_replace_node(unfiltered_bs, file_bs, &local_err);
+
+ bdrv_graph_wrunlock(s->target_bs);
+ bdrv_drained_end(file_bs);
+ bdrv_unref(unfiltered_bs);
+
+ if (local_err) {
+ error_prepend(&local_err, "failed to replace alloc-track node: ");
+ error_report_err(local_err);
+ ret = -EPERM;
+ goto out;
+ }
+ }
}
out:

8
debian/patches/pve/0040-Revert-block-rbd-workaround-for-ceph-issue-53784.patch vendored
View File

@ -13,10 +13,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 2 insertions(+), 40 deletions(-)
diff --git a/block/rbd.c b/block/rbd.c
index a4749f3b1b..53e0396b51 100644
index 63f60d41be..367db42dce 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -1511,7 +1511,6 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
@@ -1515,7 +1515,6 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
int status, r;
RBDDiffIterateReq req = { .offs = offset };
uint64_t features, flags;
@ -24,7 +24,7 @@ index a4749f3b1b..53e0396b51 100644
assert(offset + bytes <= s->image_size);
@@ -1539,43 +1538,7 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
@@ -1543,43 +1542,7 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
return status;
}
@ -69,7 +69,7 @@ index a4749f3b1b..53e0396b51 100644
qemu_rbd_diff_iterate_cb, &req);
if (r < 0 && r != QEMU_RBD_EXIT_DIFF_ITERATE2) {
return status;
@@ -1594,8 +1557,7 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
@@ -1598,8 +1561,7 @@ static int coroutine_fn qemu_rbd_co_block_status(BlockDriverState *bs,
status = BDRV_BLOCK_ZERO | BDRV_BLOCK_OFFSET_VALID;
}

4
debian/patches/pve/0041-Revert-block-rbd-fix-handling-of-holes-in-.bdrv_co_b.patch vendored
View File

@ -14,10 +14,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/block/rbd.c b/block/rbd.c
index 53e0396b51..0913a0af39 100644
index 367db42dce..347b121626 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -1470,11 +1470,11 @@ static int qemu_rbd_diff_iterate_cb(uint64_t offs, size_t len,
@@ -1474,11 +1474,11 @@ static int qemu_rbd_diff_iterate_cb(uint64_t offs, size_t len,
RBDDiffIterateReq *req = opaque;
assert(req->offs + req->bytes <= offs);

6
debian/patches/pve/0042-Revert-block-rbd-implement-bdrv_co_block_status.patch vendored
View File

@ -24,7 +24,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 112 deletions(-)
diff --git a/block/rbd.c b/block/rbd.c
index 0913a0af39..1dab254517 100644
index 347b121626..e61b359b97 100644
--- a/block/rbd.c
+++ b/block/rbd.c
@@ -108,12 +108,6 @@ typedef struct RBDTask {
@ -40,7 +40,7 @@ index 0913a0af39..1dab254517 100644
static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx,
BlockdevOptionsRbd *opts, bool cache,
const char *keypairs, const char *secretid,
@@ -1456,111 +1450,6 @@ static ImageInfoSpecific *qemu_rbd_get_specific_info(BlockDriverState *bs,
@@ -1460,111 +1454,6 @@ static ImageInfoSpecific *qemu_rbd_get_specific_info(BlockDriverState *bs,
return spec_info;
}
@ -152,7 +152,7 @@ index 0913a0af39..1dab254517 100644
static int64_t coroutine_fn qemu_rbd_co_getlength(BlockDriverState *bs)
{
BDRVRBDState *s = bs->opaque;
@@ -1796,7 +1685,6 @@ static BlockDriver bdrv_rbd = {
@@ -1800,7 +1689,6 @@ static BlockDriver bdrv_rbd = {
#ifdef LIBRBD_SUPPORTS_WRITE_ZEROES
.bdrv_co_pwrite_zeroes = qemu_rbd_co_pwrite_zeroes,
#endif

43
debian/patches/pve/0043-alloc-track-error-out-when-auto-remove-is-not-set.patch vendored Normal file
View File

@ -0,0 +1,43 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Tue, 26 Mar 2024 14:57:51 +0100
Subject: [PATCH] alloc-track: error out when auto-remove is not set
Since replacing the node now happens in the stream job, where the
option cannot be read from (it's internal to the driver), it will
always be treated as on.
qemu-server will always set it, make sure to have other users notice
the change (should they even exist). The option can be fully dropped
in the future while adding a version guard in qemu-server.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/alloc-track.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/block/alloc-track.c b/block/alloc-track.c
index 14698c362e..dad8fe6375 100644
--- a/block/alloc-track.c
+++ b/block/alloc-track.c
@@ -34,7 +34,6 @@ typedef struct {
BdrvDirtyBitmap *bitmap;
uint64_t granularity;
DropState drop_state;
- bool auto_remove;
} BDRVAllocTrackState;
static QemuOptsList runtime_opts = {
@@ -86,7 +85,11 @@ static int track_open(BlockDriverState *bs, QDict *options, int flags,
goto fail;
}
- s->auto_remove = qemu_opt_get_bool(opts, TRACK_OPT_AUTO_REMOVE, false);
+ if (!qemu_opt_get_bool(opts, TRACK_OPT_AUTO_REMOVE, false)) {
+ error_setg(errp, "alloc-track: requires auto-remove option to be set to on");
+ ret = -EINVAL;
+ goto fail;
+ }
/* open the target (write) node, backing will be attached by block layer */
file = bdrv_open_child(NULL, options, "file", bs, &child_of_bds,

154
debian/patches/pve/0043-alloc-track-fix-deadlock-during-drop.patch vendored
View File

@ -1,154 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Thu, 6 Apr 2023 14:59:31 +0200
Subject: [PATCH] alloc-track: fix deadlock during drop
by replacing the block node directly after changing the backing file
instead of rescheduling it.
With changes in QEMU 8.0, calling bdrv_get_info (and bdrv_unref)
during drop can lead to a deadlock when using iothread (only triggered
with multiple disks, except during debugging where it also triggered
with one disk sometimes):
1. job_unref_locked acquires the AioContext and calls job->driver->free
2. track_drop gets scheduled
3. bdrv_graph_wrlock is called and polls which leads to track_drop being
called
4. track_drop acquires the AioContext recursively
5. bdrv_get_info is a wrapped coroutine (since 8.0) and thus polls for
bdrv_co_get_info. This releases the AioContext, but only once! The
documentation for the AIO_WAIT_WHILE macro states that the
AioContext lock needs to be acquired exactly once, but there does
not seem to be a way for track_drop to know if it acquired the lock
recursively or not (without adding further hacks).
6. Because the AioContext is still held by the main thread once, it can't
be acquired before entering bdrv_co_get_info in co_schedule_bh_cb
which happens in the iothread
When doing the operation in change_backing_file, the AioContext has
already been acquired by the caller, so the issue with the recursive
lock goes away.
The comment explaining why delaying the replace is necessary is
> we need to schedule this for later however, since when this function
> is called, the blockjob modifying us is probably not done yet and
> has a blocker on 'bs'
However, there is no check for blockers in bdrv_replace_node. It would
need to be done by us, the caller, with check_to_replace_node.
Furthermore, the mirror job also does its call to bdrv_replace_node
while there is an active blocker (inserted by mirror itself) and they
use a specialized version to check for blockers instead of
check_to_replace_node there. Alloc-track could also do something
similar to check for other blockers, but it should be fine to rely on
Proxmox VE that no other operation with the blockdev is going on.
Mirror also drains the target before replacing the node, but the
target can have other users. In case of alloc-track the file child
should not be accessible by anybody else and so there can't be an
in-flight operation for the file child when alloc-track is drained.
The rescheduling based on refcounting is a hack and it doesn't seem to
be necessary anymore. It's not clear what the original issue from the
comment was. Testing with older builds with track_drop done directly
without rescheduling also didn't lead to any noticable issue for me.
One issue it might have been is the one fixed by b1e1af394d
("block/stream: Drain subtree around graph change"), where
block-stream had a use-after-free if the base node changed at an
inconvenient time (which alloc-track's auto-drop does).
It's also not possible to just not auto-replace the alloc-track. Not
replacing it at all leads to other operations like block resize
hanging, and there is no good way to replace it manually via QMP
(there is x-blockdev-change, but it is experimental and doesn't
implement the required operation yet). Also, it's just cleaner in
general to not leave unnecessary block nodes lying around.
Suggested-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
block/alloc-track.c | 54 ++++++++++++++-------------------------------
1 file changed, 16 insertions(+), 38 deletions(-)
diff --git a/block/alloc-track.c b/block/alloc-track.c
index b75d7c6460..76da140a68 100644
--- a/block/alloc-track.c
+++ b/block/alloc-track.c
@@ -25,7 +25,6 @@
typedef enum DropState {
DropNone,
- DropRequested,
DropInProgress,
} DropState;
@@ -268,37 +267,6 @@ static void track_child_perm(BlockDriverState *bs, BdrvChild *c,
}
}
-static void track_drop(void *opaque)
-{
- BlockDriverState *bs = (BlockDriverState*)opaque;
- BlockDriverState *file = bs->file->bs;
- BDRVAllocTrackState *s = bs->opaque;
-
- assert(file);
-
- /* we rely on the fact that we're not used anywhere else, so let's wait
- * until we're only used once - in the drive connected to the guest (and one
- * ref is held by bdrv_ref in track_change_backing_file) */
- if (bs->refcnt > 2) {
- aio_bh_schedule_oneshot(qemu_get_aio_context(), track_drop, opaque);
- return;
- }
- AioContext *aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
- bdrv_drained_begin(bs);
-
- /* now that we're drained, we can safely set 'DropInProgress' */
- s->drop_state = DropInProgress;
- bdrv_child_refresh_perms(bs, bs->file, &error_abort);
-
- bdrv_replace_node(bs, file, &error_abort);
- bdrv_set_backing_hd(bs, NULL, &error_abort);
- bdrv_drained_end(bs);
- bdrv_unref(bs);
- aio_context_release(aio_context);
-}
-
static int track_change_backing_file(BlockDriverState *bs,
const char *backing_file,
const char *backing_fmt)
@@ -308,13 +276,23 @@ static int track_change_backing_file(BlockDriverState *bs,
backing_file == NULL && backing_fmt == NULL)
{
/* backing file has been disconnected, there's no longer any use for
- * this node, so let's remove ourselves from the block graph - we need
- * to schedule this for later however, since when this function is
- * called, the blockjob modifying us is probably not done yet and has a
- * blocker on 'bs' */
- s->drop_state = DropRequested;
+ * this node, so let's remove ourselves from the block graph */
+ BlockDriverState *file = bs->file->bs;
+
+ /* Just to be sure, because bdrv_replace_node unrefs it */
bdrv_ref(bs);
- aio_bh_schedule_oneshot(qemu_get_aio_context(), track_drop, (void*)bs);
+ bdrv_drained_begin(bs);
+
+ /* now that we're drained, we can safely set 'DropInProgress' */
+ s->drop_state = DropInProgress;
+
+ bdrv_child_refresh_perms(bs, bs->file, &error_abort);
+
+ bdrv_replace_node(bs, file, &error_abort);
+ bdrv_set_backing_hd(bs, NULL, &error_abort);
+
+ bdrv_drained_end(bs);
+ bdrv_unref(bs);
}
return 0;

84
debian/patches/pve/0044-alloc-track-avoid-seemingly-superfluous-child-permis.patch vendored Normal file
View File

@ -0,0 +1,84 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Wed, 27 Mar 2024 11:15:39 +0100
Subject: [PATCH] alloc-track: avoid seemingly superfluous child permission
update
Doesn't seem necessary nowadays (maybe after commit "alloc-track: fix
deadlock during drop" where the dropping is not rescheduled and delayed
anymore or some upstream change). Should there really be some issue,
instead of having a drop state, this could also be just based off the
fact whether there is still a backing child.
Dumping the cumulative (shared) permissions for the BDS with a debug
print yields the same values after this patch and with QEMU 8.1,
namely 3 and 5.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
---
block/alloc-track.c | 26 --------------------------
1 file changed, 26 deletions(-)
diff --git a/block/alloc-track.c b/block/alloc-track.c
index dad8fe6375..7aff9763ad 100644
--- a/block/alloc-track.c
+++ b/block/alloc-track.c
@@ -25,15 +25,9 @@
#define TRACK_OPT_AUTO_REMOVE "auto-remove"
-typedef enum DropState {
- DropNone,
- DropInProgress,
-} DropState;
-
typedef struct {
BdrvDirtyBitmap *bitmap;
uint64_t granularity;
- DropState drop_state;
} BDRVAllocTrackState;
static QemuOptsList runtime_opts = {
@@ -137,8 +131,6 @@ static int track_open(BlockDriverState *bs, QDict *options, int flags,
goto fail;
}
- s->drop_state = DropNone;
-
fail:
if (ret < 0) {
bdrv_graph_wrlock(bs);
@@ -289,18 +281,8 @@ track_child_perm(BlockDriverState *bs, BdrvChild *c, BdrvChildRole role,
BlockReopenQueue *reopen_queue, uint64_t perm, uint64_t shared,
uint64_t *nperm, uint64_t *nshared)
{
- BDRVAllocTrackState *s = bs->opaque;
-
*nshared = BLK_PERM_ALL;
- /* in case we're currently dropping ourselves, claim to not use any
- * permissions at all - which is fine, since from this point on we will
- * never issue a read or write anymore */
- if (s->drop_state == DropInProgress) {
- *nperm = 0;
- return;
- }
-
if (role & BDRV_CHILD_DATA) {
*nperm = perm & DEFAULT_PERM_PASSTHROUGH;
} else {
@@ -326,14 +308,6 @@ track_co_change_backing_file(BlockDriverState *bs, const char *backing_file,
* kinda fits better, but in the long-term, a special parameter would be
* nice (or done via qemu-server via upcoming blockdev-replace QMP command).
*/
- if (backing_file == NULL) {
- BDRVAllocTrackState *s = bs->opaque;
- bdrv_drained_begin(bs);
- s->drop_state = DropInProgress;
- bdrv_child_refresh_perms(bs, bs->file, &error_abort);
- bdrv_drained_end(bs);
- }
-
return 0;
}

191
debian/patches/pve/0044-migration-for-snapshots-hold-the-BQL-during-setup-ca.patch vendored
View File

@ -1,191 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Fri, 5 May 2023 13:39:53 +0200
Subject: [PATCH] migration: for snapshots, hold the BQL during setup callbacks
In spirit, this is a partial revert of commit 9b09503752 ("migration:
run setup callbacks out of big lock"), but only for the snapshot case.
For snapshots, the bdrv_writev_vmstate() function is used during setup
(in QIOChannelBlock backing the QEMUFile), but not holding the BQL
while calling it could lead to an assertion failure. To understand
how, first note the following:
1. Generated coroutine wrappers for block layer functions spawn the
coroutine and use AIO_WAIT_WHILE()/aio_poll() to wait for it.
2. If the host OS switches threads at an inconvenient time, it can
happen that a bottom half scheduled for the main thread's AioContext
is executed as part of a vCPU thread's aio_poll().
An example leading to the assertion failure is as follows:
main thread:
1. A snapshot-save QMP command gets issued.
2. snapshot_save_job_bh() is scheduled.
vCPU thread:
3. aio_poll() for the main thread's AioContext is called (e.g. when
the guest writes to a pflash device, as part of blk_pwrite which is a
generated coroutine wrapper).
4. snapshot_save_job_bh() is executed as part of aio_poll().
3. qemu_savevm_state() is called.
4. qemu_mutex_unlock_iothread() is called. Now
qemu_get_current_aio_context() returns 0x0.
5. bdrv_writev_vmstate() is executed during the usual savevm setup.
But this function is a generated coroutine wrapper, so it uses
AIO_WAIT_WHILE. There, the assertion
assert(qemu_get_current_aio_context() == qemu_get_aio_context());
will fail.
To fix it, ensure that the BQL is held during setup. To avoid changing
the behavior for migration too, introduce conditionals for the setup
callbacks that need the BQL and only take the lock if it's not already
held.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
include/migration/register.h | 2 +-
migration/block-dirty-bitmap.c | 15 ++++++++++++---
migration/block.c | 15 ++++++++++++---
migration/ram.c | 16 +++++++++++++---
migration/savevm.c | 2 --
5 files changed, 38 insertions(+), 12 deletions(-)
diff --git a/include/migration/register.h b/include/migration/register.h
index 90914f32f5..c728fd9120 100644
--- a/include/migration/register.h
+++ b/include/migration/register.h
@@ -43,9 +43,9 @@ typedef struct SaveVMHandlers {
* by other locks.
*/
int (*save_live_iterate)(QEMUFile *f, void *opaque);
+ int (*save_setup)(QEMUFile *f, void *opaque);
/* This runs outside the iothread lock! */
- int (*save_setup)(QEMUFile *f, void *opaque);
/* Note for save_live_pending:
* must_precopy:
* - must be migrated in precopy or in stopped state
diff --git a/migration/block-dirty-bitmap.c b/migration/block-dirty-bitmap.c
index 285dd1d148..f7ee5a74d9 100644
--- a/migration/block-dirty-bitmap.c
+++ b/migration/block-dirty-bitmap.c
@@ -1219,10 +1219,17 @@ static int dirty_bitmap_save_setup(QEMUFile *f, void *opaque)
{
DBMSaveState *s = &((DBMState *)opaque)->save;
SaveBitmapState *dbms = NULL;
+ bool release_lock = false;
- qemu_mutex_lock_iothread();
+ /* For snapshots, the BQL is held during setup. */
+ if (!qemu_mutex_iothread_locked()) {
+ qemu_mutex_lock_iothread();
+ release_lock = true;
+ }
if (init_dirty_bitmap_migration(s) < 0) {
- qemu_mutex_unlock_iothread();
+ if (release_lock) {
+ qemu_mutex_unlock_iothread();
+ }
return -1;
}
@@ -1230,7 +1237,9 @@ static int dirty_bitmap_save_setup(QEMUFile *f, void *opaque)
send_bitmap_start(f, s, dbms);
}
qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS);
- qemu_mutex_unlock_iothread();
+ if (release_lock) {
+ qemu_mutex_unlock_iothread();
+ }
return 0;
}
diff --git a/migration/block.c b/migration/block.c
index 86c2256a2b..8423e0c9f9 100644
--- a/migration/block.c
+++ b/migration/block.c
@@ -725,21 +725,30 @@ static void block_migration_cleanup(void *opaque)
static int block_save_setup(QEMUFile *f, void *opaque)
{
int ret;
+ bool release_lock = false;
trace_migration_block_save("setup", block_mig_state.submitted,
block_mig_state.transferred);
- qemu_mutex_lock_iothread();
+ /* For snapshots, the BQL is held during setup. */
+ if (!qemu_mutex_iothread_locked()) {
+ qemu_mutex_lock_iothread();
+ release_lock = true;
+ }
ret = init_blk_migration(f);
if (ret < 0) {
- qemu_mutex_unlock_iothread();
+ if (release_lock) {
+ qemu_mutex_unlock_iothread();
+ }
return ret;
}
/* start track dirty blocks */
ret = set_dirty_tracking();
- qemu_mutex_unlock_iothread();
+ if (release_lock) {
+ qemu_mutex_unlock_iothread();
+ }
if (ret) {
return ret;
diff --git a/migration/ram.c b/migration/ram.c
index 6e1514f69f..6a1aec7031 100644
--- a/migration/ram.c
+++ b/migration/ram.c
@@ -2896,8 +2896,16 @@ static void migration_bitmap_clear_discarded_pages(RAMState *rs)
static void ram_init_bitmaps(RAMState *rs)
{
- /* For memory_global_dirty_log_start below. */
- qemu_mutex_lock_iothread();
+ bool release_lock = false;
+
+ /*
+ * For memory_global_dirty_log_start below.
+ * For snapshots, the BQL is held during setup.
+ */
+ if (!qemu_mutex_iothread_locked()) {
+ qemu_mutex_lock_iothread();
+ release_lock = true;
+ }
qemu_mutex_lock_ramlist();
WITH_RCU_READ_LOCK_GUARD() {
@@ -2909,7 +2917,9 @@ static void ram_init_bitmaps(RAMState *rs)
}
}
qemu_mutex_unlock_ramlist();
- qemu_mutex_unlock_iothread();
+ if (release_lock) {
+ qemu_mutex_unlock_iothread();
+ }
/*
* After an eventual first bitmap sync, fixup the initial bitmap
diff --git a/migration/savevm.c b/migration/savevm.c
index d60c4f487a..3c015722f7 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -1625,10 +1625,8 @@ static int qemu_savevm_state(QEMUFile *f, Error **errp)
reset_vfio_bytes_transferred();
ms->to_dst_file = f;
- qemu_mutex_unlock_iothread();
qemu_savevm_state_header(f);
qemu_savevm_state_setup(f);
- qemu_mutex_lock_iothread();
while (qemu_file_get_error(f) == 0) {
if (qemu_savevm_state_iterate(f, false) > 0) {

4
debian/patches/pve/0046-block-copy-before-write-fix-permission.patch → debian/patches/pve/0045-block-copy-before-write-fix-permission.patch vendored
View File

@ -33,10 +33,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index b866e42271..a2dddf6f57 100644
index 13972879b1..dbdbbca44e 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -364,9 +364,13 @@ static void cbw_child_perm(BlockDriverState *bs, BdrvChild *c,
@@ -364,9 +364,13 @@ cbw_child_perm(BlockDriverState *bs, BdrvChild *c, BdrvChildRole role,
perm, shared, nperm, nshared);
if (!QLIST_EMPTY(&bs->parents)) {

30
debian/patches/pve/0045-savevm-async-don-t-hold-BQL-during-setup.patch vendored
View File

@ -1,30 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Fiona Ebner <f.ebner@proxmox.com>
Date: Fri, 5 May 2023 15:30:16 +0200
Subject: [PATCH] savevm-async: don't hold BQL during setup
See commit "migration: for snapshots, hold the BQL during setup
callbacks" for why. This is separate, because a version of that one
will hopefully land upstream.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
migration/savevm-async.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/migration/savevm-async.c b/migration/savevm-async.c
index 80624fada8..b1d85a4b41 100644
--- a/migration/savevm-async.c
+++ b/migration/savevm-async.c
@@ -401,10 +401,8 @@ void qmp_savevm_start(const char *statefile, Error **errp)
snap_state.state = SAVE_STATE_ACTIVE;
snap_state.finalize_bh = qemu_bh_new(process_savevm_finalize, &snap_state);
snap_state.co = qemu_coroutine_create(&process_savevm_co, NULL);
- qemu_mutex_unlock_iothread();
qemu_savevm_state_header(snap_state.file);
qemu_savevm_state_setup(snap_state.file);
- qemu_mutex_lock_iothread();
/* Async processing from here on out happens in iohandler context, so let
* the target bdrv have its home there.

4
debian/patches/pve/0047-block-copy-before-write-support-unligned-snapshot-di.patch → debian/patches/pve/0046-block-copy-before-write-support-unligned-snapshot-di.patch vendored
View File

@ -15,7 +15,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index a2dddf6f57..0a219c2b75 100644
index dbdbbca44e..2cbf6f9346 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -325,14 +325,24 @@ static int coroutine_fn GRAPH_RDLOCK
@ -45,4 +45,4 @@ index a2dddf6f57..0a219c2b75 100644
+ return bdrv_co_pdiscard(s->target, aligned_offset, aligned_bytes);
}
static void cbw_refresh_filename(BlockDriverState *bs)
static void GRAPH_RDLOCK cbw_refresh_filename(BlockDriverState *bs)

10
debian/patches/pve/0048-block-copy-before-write-create-block_copy-bitmap-in-.patch → debian/patches/pve/0047-block-copy-before-write-create-block_copy-bitmap-in-.patch vendored
View File

@ -28,10 +28,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
4 files changed, 60 insertions(+), 58 deletions(-)
diff --git a/block/block-copy.c b/block/block-copy.c
index e13d7bc6b6..b61685f1a2 100644
index 9ee3dd7ef5..8fca2c3698 100644
--- a/block/block-copy.c
+++ b/block/block-copy.c
@@ -346,6 +346,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
@@ -351,6 +351,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
}
BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
@ -39,7 +39,7 @@ index e13d7bc6b6..b61685f1a2 100644
const BdrvDirtyBitmap *bitmap,
Error **errp)
{
@@ -360,7 +361,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
@@ -367,7 +368,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
return NULL;
}
@ -49,10 +49,10 @@ index e13d7bc6b6..b61685f1a2 100644
if (!copy_bitmap) {
return NULL;
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index 0a219c2b75..d3b95bd600 100644
index 2cbf6f9346..afa5f473d2 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -470,7 +470,7 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
@@ -472,7 +472,7 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
((BDRV_REQ_FUA | BDRV_REQ_MAY_UNMAP | BDRV_REQ_NO_FALLBACK) &
bs->file->bs->supported_zero_flags);

48
debian/patches/pve/0049-qapi-blockdev-backup-add-discard-source-parameter.patch → debian/patches/pve/0048-qapi-blockdev-backup-add-discard-source-parameter.patch vendored
View File

@ -45,7 +45,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
10 files changed, 37 insertions(+), 8 deletions(-)
diff --git a/block/backup.c b/block/backup.c
index af87fa6aa9..3dc955f625 100644
index aec140e0c8..f19b751fe6 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -332,7 +332,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
@ -57,7 +57,7 @@ index af87fa6aa9..3dc955f625 100644
const char *filter_node_name,
BackupPerf *perf,
BlockdevOnError on_source_error,
@@ -429,7 +429,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
@@ -433,7 +433,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
goto error;
}
@ -68,7 +68,7 @@ index af87fa6aa9..3dc955f625 100644
goto error;
}
diff --git a/block/block-copy.c b/block/block-copy.c
index b61685f1a2..3c61e52bae 100644
index 8fca2c3698..7e3b378528 100644
--- a/block/block-copy.c
+++ b/block/block-copy.c
@@ -137,6 +137,7 @@ typedef struct BlockCopyState {
@ -79,7 +79,7 @@ index b61685f1a2..3c61e52bae 100644
BlockReqList reqs;
QLIST_HEAD(, BlockCopyCallState) calls;
/*
@@ -348,6 +349,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
@@ -353,6 +354,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
BlockDriverState *copy_bitmap_bs,
const BdrvDirtyBitmap *bitmap,
@ -87,7 +87,7 @@ index b61685f1a2..3c61e52bae 100644
Error **errp)
{
ERRP_GUARD();
@@ -409,6 +411,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
@@ -418,6 +420,7 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
cluster_size),
};
@ -95,7 +95,7 @@ index b61685f1a2..3c61e52bae 100644
block_copy_set_copy_opts(s, false, false);
ratelimit_init(&s->rate_limit);
@@ -580,6 +583,12 @@ static coroutine_fn int block_copy_task_entry(AioTask *task)
@@ -589,6 +592,12 @@ static coroutine_fn int block_copy_task_entry(AioTask *task)
co_put_to_shres(s->mem, t->req.bytes);
block_copy_task_end(t, ret);
@ -109,7 +109,7 @@ index b61685f1a2..3c61e52bae 100644
}
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index d3b95bd600..3503702d71 100644
index afa5f473d2..5506f66857 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -44,6 +44,7 @@ typedef struct BDRVCopyBeforeWriteState {
@ -120,16 +120,16 @@ index d3b95bd600..3503702d71 100644
/*
* @lock: protects access to @access_bitmap, @done_bitmap and
@@ -357,6 +358,8 @@ static void cbw_child_perm(BlockDriverState *bs, BdrvChild *c,
uint64_t perm, uint64_t shared,
uint64_t *nperm, uint64_t *nshared)
@@ -357,6 +358,8 @@ cbw_child_perm(BlockDriverState *bs, BdrvChild *c, BdrvChildRole role,
uint64_t perm, uint64_t shared,
uint64_t *nperm, uint64_t *nshared)
{
+ BDRVCopyBeforeWriteState *s = bs->opaque;
+
if (!(role & BDRV_CHILD_FILTERED)) {
/*
* Target child
@@ -381,6 +384,10 @@ static void cbw_child_perm(BlockDriverState *bs, BdrvChild *c,
@@ -381,6 +384,10 @@ cbw_child_perm(BlockDriverState *bs, BdrvChild *c, BdrvChildRole role,
* start
*/
*nperm = *nperm | BLK_PERM_CONSISTENT_READ;
@ -140,7 +140,7 @@ index d3b95bd600..3503702d71 100644
*nshared &= ~(BLK_PERM_WRITE | BLK_PERM_RESIZE);
}
}
@@ -470,7 +477,9 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
@@ -472,7 +479,9 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
((BDRV_REQ_FUA | BDRV_REQ_MAY_UNMAP | BDRV_REQ_NO_FALLBACK) &
bs->file->bs->supported_zero_flags);
@ -151,7 +151,7 @@ index d3b95bd600..3503702d71 100644
if (!s->bcs) {
error_prepend(errp, "Cannot create block-copy-state: ");
ret = -EINVAL;
@@ -544,12 +553,14 @@ BlockDriver bdrv_cbw_filter = {
@@ -546,12 +555,14 @@ static BlockDriver bdrv_cbw_filter = {
BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
BlockDriverState *target,
const char *filter_node_name,
@ -166,7 +166,7 @@ index d3b95bd600..3503702d71 100644
assert(source->total_sectors == target->total_sectors);
GLOBAL_STATE_CODE();
@@ -562,7 +573,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
@@ -564,7 +575,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
qdict_put_str(opts, "file", bdrv_get_node_name(source));
qdict_put_str(opts, "target", bdrv_get_node_name(target));
@ -188,10 +188,10 @@ index 6e72bb25e9..01af0cd3c4 100644
Error **errp);
void bdrv_cbw_drop(BlockDriverState *bs);
diff --git a/block/replication.c b/block/replication.c
index ea4bf1aa80..39ad78cf98 100644
index 5ded5f1ca9..bd75a6aee3 100644
--- a/block/replication.c
+++ b/block/replication.c
@@ -579,8 +579,8 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
@@ -604,8 +604,8 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
s->backup_job = backup_job_create(
NULL, s->secondary_disk->bs, s->hidden_disk->bs,
@ -203,10 +203,10 @@ index ea4bf1aa80..39ad78cf98 100644
BLOCKDEV_ON_ERROR_REPORT, JOB_INTERNAL,
backup_job_completed, bs, NULL, &local_err);
diff --git a/blockdev.c b/blockdev.c
index 7793143d76..ce3fef924c 100644
index 3049811be8..e167ad1e54 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2802,7 +2802,7 @@ static BlockJob *do_backup_common(BackupCommon *backup,
@@ -2854,7 +2854,7 @@ static BlockJob *do_backup_common(BackupCommon *backup,
job = backup_job_create(backup->job_id, bs, target_bs, backup->speed,
backup->sync, bmap, backup->bitmap_mode,
@ -216,10 +216,10 @@ index 7793143d76..ce3fef924c 100644
&perf,
backup->on_source_error,
diff --git a/include/block/block-common.h b/include/block/block-common.h
index e15395f2cb..913a8b259c 100644
index d7599564db..7f56364e73 100644
--- a/include/block/block-common.h
+++ b/include/block/block-common.h
@@ -234,6 +234,8 @@ typedef enum {
@@ -246,6 +246,8 @@ typedef enum {
read-write fails */
#define BDRV_O_IO_URING 0x40000 /* use io_uring instead of the thread pool */
@ -241,7 +241,7 @@ index 8b41643bfa..bdc703bacd 100644
/* Function should be called prior any actual copy request */
diff --git a/include/block/block_int-global-state.h b/include/block/block_int-global-state.h
index 32f0f9858a..546f2b5532 100644
index 57265a617a..df731688b4 100644
--- a/include/block/block_int-global-state.h
+++ b/include/block/block_int-global-state.h
@@ -189,7 +189,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
@ -254,10 +254,10 @@ index 32f0f9858a..546f2b5532 100644
BackupPerf *perf,
BlockdevOnError on_source_error,
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 09de550c95..4297e5beda 100644
index 7b977459fa..82960797dc 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -1816,6 +1816,9 @@
@@ -1834,6 +1834,9 @@
# node specified by @drive. If this option is not given, a node
# name is autogenerated. (Since: 4.2)
#
@ -267,7 +267,7 @@ index 09de550c95..4297e5beda 100644
# @x-perf: Performance options. (Since 6.0)
#
# Features:
@@ -1837,6 +1840,7 @@
@@ -1855,6 +1858,7 @@
'*on-target-error': 'BlockdevOnError',
'*auto-finalize': 'bool', '*auto-dismiss': 'bool',
'*filter-node-name': 'str',

20
debian/patches/pve/0050-copy-before-write-allow-specifying-minimum-cluster-s.patch → debian/patches/pve/0049-copy-before-write-allow-specifying-minimum-cluster-s.patch vendored
View File

@ -25,7 +25,7 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
4 files changed, 23 insertions(+), 6 deletions(-)
diff --git a/block/block-copy.c b/block/block-copy.c
index 3c61e52bae..c9a722a5a6 100644
index 7e3b378528..adb1cbb440 100644
--- a/block/block-copy.c
+++ b/block/block-copy.c
@@ -310,6 +310,7 @@ void block_copy_set_copy_opts(BlockCopyState *s, bool use_copy_range,
@ -36,7 +36,7 @@ index 3c61e52bae..c9a722a5a6 100644
Error **errp)
{
int ret;
@@ -330,7 +331,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
@@ -335,7 +336,7 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
"used. If the actual block size of the target exceeds "
"this default, the backup may be unusable",
BLOCK_COPY_CLUSTER_SIZE_DEFAULT);
@ -45,7 +45,7 @@ index 3c61e52bae..c9a722a5a6 100644
} else if (ret < 0 && !target_does_cow) {
error_setg_errno(errp, -ret,
"Couldn't determine the cluster size of the target image, "
@@ -340,16 +341,18 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
@@ -345,16 +346,18 @@ static int64_t block_copy_calculate_cluster_size(BlockDriverState *target,
return ret;
} else if (ret < 0 && target_does_cow) {
/* Not fatal; just trudge on ahead. */
@ -66,9 +66,9 @@ index 3c61e52bae..c9a722a5a6 100644
Error **errp)
{
ERRP_GUARD();
@@ -358,7 +361,13 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
BdrvDirtyBitmap *copy_bitmap;
bool is_fleecing;
@@ -365,7 +368,13 @@ BlockCopyState *block_copy_state_new(BdrvChild *source, BdrvChild *target,
GLOBAL_STATE_CODE();
- cluster_size = block_copy_calculate_cluster_size(target->bs, errp);
+ if (min_cluster_size && !is_power_of_2(min_cluster_size)) {
@ -82,10 +82,10 @@ index 3c61e52bae..c9a722a5a6 100644
return NULL;
}
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index 3503702d71..4a8c5bdb62 100644
index 5506f66857..fbc5aeb9eb 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -479,7 +479,8 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
@@ -481,7 +481,8 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
s->discard_source = flags & BDRV_O_CBW_DISCARD_SOURCE;
s->bcs = block_copy_state_new(bs->file, s->target, bs, bitmap,
@ -108,10 +108,10 @@ index bdc703bacd..77857c6c68 100644
/* Function should be called prior any actual copy request */
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 4297e5beda..33e7e3c090 100644
index 82960797dc..f2fec625cc 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -4825,12 +4825,18 @@
@@ -4880,12 +4880,18 @@
# @on-cbw-error parameter will decide how this failure is handled.
# Default 0. (Since 7.1)
#

18
debian/patches/pve/0051-backup-add-minimum-cluster-size-to-performance-optio.patch → debian/patches/pve/0050-backup-add-minimum-cluster-size-to-performance-optio.patch vendored
View File

@ -23,10 +23,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
5 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/block/backup.c b/block/backup.c
index 3dc955f625..ac5bd81338 100644
index f19b751fe6..4367278d68 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -430,7 +430,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
@@ -434,7 +434,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
}
cbw = bdrv_cbw_append(bs, target, filter_node_name, discard_source,
@ -36,10 +36,10 @@ index 3dc955f625..ac5bd81338 100644
goto error;
}
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index 4a8c5bdb62..9ca5ec5e5c 100644
index fbc5aeb9eb..3d5523992c 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -555,6 +555,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
@@ -557,6 +557,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
BlockDriverState *target,
const char *filter_node_name,
bool discard_source,
@ -47,7 +47,7 @@ index 4a8c5bdb62..9ca5ec5e5c 100644
BlockCopyState **bcs,
Error **errp)
{
@@ -573,6 +574,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
@@ -575,6 +576,7 @@ BlockDriverState *bdrv_cbw_append(BlockDriverState *source,
}
qdict_put_str(opts, "file", bdrv_get_node_name(source));
qdict_put_str(opts, "target", bdrv_get_node_name(target));
@ -68,10 +68,10 @@ index 01af0cd3c4..dc6cafe7fa 100644
Error **errp);
void bdrv_cbw_drop(BlockDriverState *bs);
diff --git a/blockdev.c b/blockdev.c
index ce3fef924c..5ae1dde73c 100644
index e167ad1e54..35ec5d8f0b 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -2729,6 +2729,9 @@ static BlockJob *do_backup_common(BackupCommon *backup,
@@ -2781,6 +2781,9 @@ static BlockJob *do_backup_common(BackupCommon *backup,
if (backup->x_perf->has_max_chunk) {
perf.max_chunk = backup->x_perf->max_chunk;
}
@ -82,10 +82,10 @@ index ce3fef924c..5ae1dde73c 100644
if ((backup->sync == MIRROR_SYNC_MODE_BITMAP) ||
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 33e7e3c090..58fd637e86 100644
index f2fec625cc..48eec4ef29 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -1757,11 +1757,16 @@
@@ -1775,11 +1775,16 @@
# it should not be less than job cluster size which is calculated
# as maximum of target image cluster size and 64k. Default 0.
#

51
debian/patches/pve/0052-PVE-backup-add-fleecing-option.patch → debian/patches/pve/0051-PVE-backup-add-fleecing-option.patch vendored
View File

@ -63,15 +63,15 @@ Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
---
block/monitor/block-hmp-cmds.c | 1 +
pve-backup.c | 143 ++++++++++++++++++++++++++++++++-
pve-backup.c | 145 ++++++++++++++++++++++++++++++++-
qapi/block-core.json | 8 +-
3 files changed, 148 insertions(+), 4 deletions(-)
3 files changed, 150 insertions(+), 4 deletions(-)
diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
index 6efe28cef5..ca29cc4281 100644
index 1656859e03..f6cc9e5cf7 100644
--- a/block/monitor/block-hmp-cmds.c
+++ b/block/monitor/block-hmp-cmds.c
@@ -1064,6 +1064,7 @@ void coroutine_fn hmp_backup(Monitor *mon, const QDict *qdict)
@@ -1072,6 +1072,7 @@ void coroutine_fn hmp_backup(Monitor *mon, const QDict *qdict)
NULL, NULL,
devlist, qdict_haskey(qdict, "speed"), speed,
false, 0, // BackupPerf max-workers
@ -80,21 +80,22 @@ index 6efe28cef5..ca29cc4281 100644
hmp_handle_error(mon, error);
diff --git a/pve-backup.c b/pve-backup.c
index e6b17b797e..00aaff6509 100644
index 777db7938e..4c728951ac 100644
--- a/pve-backup.c
+++ b/pve-backup.c
@@ -7,8 +7,10 @@
@@ -7,9 +7,11 @@
#include "sysemu/blockdev.h"
#include "block/block_int-global-state.h"
#include "block/blockjob.h"
+#include "block/copy-before-write.h"
#include "block/dirty-bitmap.h"
#include "block/graph-lock.h"
#include "qapi/qapi-commands-block.h"
+#include "qapi/qmp/qdict.h"
#include "qapi/qmp/qerror.h"
#include "qemu/cutils.h"
@@ -80,8 +82,15 @@ static void pvebackup_init(void)
@@ -81,8 +83,15 @@ static void pvebackup_init(void)
// initialize PVEBackupState at startup
opts_init(pvebackup_init);
@ -110,7 +111,7 @@ index e6b17b797e..00aaff6509 100644
size_t size;
uint64_t block_size;
uint8_t dev_id;
@@ -361,6 +370,25 @@ static void pvebackup_complete_cb(void *opaque, int ret)
@@ -355,6 +364,25 @@ static void pvebackup_complete_cb(void *opaque, int ret)
PVEBackupDevInfo *di = opaque;
di->completed_ret = ret;
@ -134,9 +135,9 @@ index e6b17b797e..00aaff6509 100644
+ }
+
/*
* Schedule stream cleanup in async coroutine. close_image and finish might
* take a while, so we can't block on them here. This way it also doesn't
@@ -521,9 +549,82 @@ static void create_backup_jobs_bh(void *opaque) {
* Needs to happen outside of coroutine, because it takes the graph write lock.
*/
@@ -525,9 +553,84 @@ static void create_backup_jobs_bh(void *opaque) {
bdrv_drained_begin(di->bs);
@ -144,7 +145,9 @@ index e6b17b797e..00aaff6509 100644
+
+ BlockDriverState *source_bs = di->bs;
+ bool discard_source = false;
+ bdrv_graph_co_rdlock();
+ const char *job_id = bdrv_get_device_name(di->bs);
+ bdrv_graph_co_rdunlock();
+ if (di->fleecing.bs) {
+ QDict *cbw_opts = qdict_new();
+ qdict_put_str(cbw_opts, "driver", "copy-before-write");
@ -221,7 +224,7 @@ index e6b17b797e..00aaff6509 100644
BLOCKDEV_ON_ERROR_REPORT, JOB_DEFAULT, pvebackup_complete_cb, di, backup_state.txn,
&local_err);
@@ -581,6 +682,14 @@ static void create_backup_jobs_bh(void *opaque) {
@@ -585,6 +688,14 @@ static void create_backup_jobs_bh(void *opaque) {
aio_co_enter(data->ctx, data->co);
}
@ -236,15 +239,15 @@ index e6b17b797e..00aaff6509 100644
/*
* Returns a list of device infos, which needs to be freed by the caller. In
* case of an error, errp will be set, but the returned value might still be a
@@ -588,6 +697,7 @@ static void create_backup_jobs_bh(void *opaque) {
@@ -592,6 +703,7 @@ static void create_backup_jobs_bh(void *opaque) {
*/
static GList coroutine_fn *get_device_info(
static GList coroutine_fn GRAPH_RDLOCK *get_device_info(
const char *devlist,
+ bool fleecing,
Error **errp)
{
gchar **devs = NULL;
@@ -611,6 +721,31 @@ static GList coroutine_fn *get_device_info(
@@ -615,6 +727,31 @@ static GList coroutine_fn GRAPH_RDLOCK *get_device_info(
}
PVEBackupDevInfo *di = g_new0(PVEBackupDevInfo, 1);
di->bs = bs;
@ -276,7 +279,7 @@ index e6b17b797e..00aaff6509 100644
di_list = g_list_append(di_list, di);
d++;
}
@@ -660,6 +795,7 @@ UuidInfo coroutine_fn *qmp_backup(
@@ -664,6 +801,7 @@ UuidInfo coroutine_fn *qmp_backup(
const char *devlist,
bool has_speed, int64_t speed,
bool has_max_workers, int64_t max_workers,
@ -284,16 +287,16 @@ index e6b17b797e..00aaff6509 100644
Error **errp)
{
assert(qemu_in_coroutine());
@@ -687,7 +823,7 @@ UuidInfo coroutine_fn *qmp_backup(
/* Todo: try to auto-detect format based on file name */
@@ -692,7 +830,7 @@ UuidInfo coroutine_fn *qmp_backup(
format = has_format ? format : BACKUP_FORMAT_VMA;
bdrv_graph_co_rdlock();
- di_list = get_device_info(devlist, &local_err);
+ di_list = get_device_info(devlist, has_fleecing && fleecing, &local_err);
bdrv_graph_co_rdunlock();
if (local_err) {
error_propagate(errp, local_err);
goto err;
@@ -1086,5 +1222,6 @@ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
@@ -1100,5 +1238,6 @@ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
ret->query_bitmap_info = true;
ret->pbs_masterkey = true;
ret->backup_max_workers = true;
@ -301,10 +304,10 @@ index e6b17b797e..00aaff6509 100644
return ret;
}
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 58fd637e86..0bc5f42677 100644
index 48eec4ef29..1c036e488e 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -933,6 +933,10 @@
@@ -935,6 +935,10 @@
#
# @max-workers: see @BackupPerf for details. Default 16.
#
@ -315,7 +318,7 @@ index 58fd637e86..0bc5f42677 100644
# Returns: the uuid of the backup job
#
##
@@ -953,7 +957,8 @@
@@ -955,7 +959,8 @@
'*firewall-file': 'str',
'*devlist': 'str',
'*speed': 'int',
@ -325,7 +328,7 @@ index 58fd637e86..0bc5f42677 100644
'returns': 'UuidInfo', 'coroutine': true }
##
@@ -1009,6 +1014,7 @@
@@ -1011,6 +1016,7 @@
'pbs-dirty-bitmap-migration': 'bool',
'pbs-masterkey': 'bool',
'pbs-library-version': 'str',

31
debian/patches/series vendored
View File

@ -2,15 +2,9 @@ extra/0001-monitor-qmp-fix-race-with-clients-disconnecting-earl.patch
extra/0002-scsi-megasas-Internal-cdbs-have-16-byte-length.patch
extra/0003-ide-avoid-potential-deadlock-when-draining-during-tr.patch
extra/0004-migration-block-dirty-bitmap-fix-loading-bitmap-when.patch
extra/0005-Revert-Revert-graph-lock-Disable-locking-for-now.patch
extra/0006-migration-states-workaround-snapshot-performance-reg.patch
extra/0007-Revert-x86-acpi-workaround-Windows-not-handling-name.patch
extra/0008-target-i386-the-sgx_epc_get_section-stub-is-reachabl.patch
extra/0009-ui-clipboard-mark-type-as-not-available-when-there-i.patch
extra/0010-virtio-scsi-Attach-event-vq-notifier-with-no_poll.patch
extra/0011-virtio-Re-enable-notifications-after-drain.patch
extra/0012-qemu_init-increase-NOFILE-soft-limit-on-POSIX.patch
extra/0013-virtio-blk-avoid-using-ioeventfd-state-in-irqfd-cond.patch
extra/0005-Revert-x86-acpi-workaround-Windows-not-handling-name.patch
extra/0006-qemu_init-increase-NOFILE-soft-limit-on-POSIX.patch
extra/0007-mirror-Don-t-call-job_pause_point-under-graph-lock.patch
bitmap-mirror/0001-drive-mirror-add-support-for-sync-bitmap-mode-never.patch
bitmap-mirror/0002-drive-mirror-add-support-for-conditional-and-always-.patch
bitmap-mirror/0003-mirror-add-check-for-bitmap-mode-without-bitmap.patch
@ -59,13 +53,12 @@ pve/0039-block-add-alloc-track-driver.patch
pve/0040-Revert-block-rbd-workaround-for-ceph-issue-53784.patch
pve/0041-Revert-block-rbd-fix-handling-of-holes-in-.bdrv_co_b.patch
pve/0042-Revert-block-rbd-implement-bdrv_co_block_status.patch
pve/0043-alloc-track-fix-deadlock-during-drop.patch
pve/0044-migration-for-snapshots-hold-the-BQL-during-setup-ca.patch
pve/0045-savevm-async-don-t-hold-BQL-during-setup.patch
pve/0046-block-copy-before-write-fix-permission.patch
pve/0047-block-copy-before-write-support-unligned-snapshot-di.patch
pve/0048-block-copy-before-write-create-block_copy-bitmap-in-.patch
pve/0049-qapi-blockdev-backup-add-discard-source-parameter.patch
pve/0050-copy-before-write-allow-specifying-minimum-cluster-s.patch
pve/0051-backup-add-minimum-cluster-size-to-performance-optio.patch
pve/0052-PVE-backup-add-fleecing-option.patch
pve/0043-alloc-track-error-out-when-auto-remove-is-not-set.patch
pve/0044-alloc-track-avoid-seemingly-superfluous-child-permis.patch
pve/0045-block-copy-before-write-fix-permission.patch
pve/0046-block-copy-before-write-support-unligned-snapshot-di.patch
pve/0047-block-copy-before-write-create-block_copy-bitmap-in-.patch
pve/0048-qapi-blockdev-backup-add-discard-source-parameter.patch
pve/0049-copy-before-write-allow-specifying-minimum-cluster-s.patch
pve/0050-backup-add-minimum-cluster-size-to-performance-optio.patch
pve/0051-PVE-backup-add-fleecing-option.patch

2
qemu

@ -1 +1 @@
Subproject commit 20a1b341a0af1fef84cec9e521d33da0e8d9ecf3
Subproject commit 11aa0b1ff115b86160c4d37e7c37e6a6b13b77ea