Commit Graph

379 Commits

Author SHA1 Message Date
Thomas Lamprecht
f87d0523df vma: allow partial restore
Introduce a new map line for skipping a certain drive, of the form
skip=drive-scsi0

Since in PVE, most archives are compressed and piped to vma for
restore, it's not easily possible to skip reads.

For the reader, a new skip flag for VmaRestoreState is added and the
target is allowed to be NULL if skip is specified when registering.
If
the skip flag is set, no writes will be made as well as no check for
duplicate clusters. Therefore, the flag is not set for verify.

Originally-by: Fabian Ebner <f.ebner@proxmox.com>
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Acked-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-25 10:07:37 +02:00
Thomas Lamprecht
2fd4ea2813 patches: update context
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-25 10:07:01 +02:00
Thomas Lamprecht
2653a5f029 vma: restore: call blk_unref for all opened block devices
Originally-by: Fabian Ebner <f.ebner@proxmox.com>
Link: https://lists.proxmox.com/pipermail/pve-devel/2022-April/052642.html
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-25 10:05:29 +02:00
Thomas Lamprecht
664ecf59a9 bump version to 6.2.0-4
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-22 11:52:34 +02:00
Thomas Lamprecht
4de9440f87 various stable backports
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-22 10:22:39 +02:00
Thomas Lamprecht
9b348f8c6d d/copyright: drop trailing whitespace
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-22 09:16:23 +02:00
Thomas Lamprecht
799cf8c5a3 d/control: add suggest dependency-hint for libgl1
It pulls in a lot of stuff via the libglx0 -> libglx-mesa0 dependency
chain, so only suggest it for now to avoid installing it in the
installer or via common "PVE on-top Debian" installations, VirGL
integration is experimental after all and we may drop/replace it with
the vulkan based venus one, once available (Debian 12?).

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-22 09:14:39 +02:00
Thomas Lamprecht
b02e62dba0 d/control: add libgbm to build dependencies
required for good virgl support

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-22 08:50:36 +02:00
Thomas Lamprecht
fc03e1b6bf bump version to 6.2.0-3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-15 09:09:43 +02:00
Thomas Lamprecht
c8ba14bed0 cherry-pick fix for passing some acpi slic tables
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-04-15 08:07:34 +02:00
Wolfgang Bumiller
daea534caa bump version to 6.2.0-2
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2022-03-03 12:05:37 +01:00
Fabian Ebner
27199bd753 backup: add patch to initialize bcs bitmap early enough for PBS
This is necessary for multi-disk backups where not all jobs are
immediately started after they are created. QEMU commit
06e0a9c16405c0a4c1eca33cf286cc04c42066a2 did already part of the work,
ensuring that new writes after job creation don't pass through to the
backup, but not yet for the MIRROR_SYNC_MODE_BITMAP case which is used
for PBS.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2022-03-03 11:37:17 +01:00
Thomas Lamprecht
e050683663 d/control: mark numactl a recommended package
we do not call in anywhere unconditionally

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-24 07:51:19 +01:00
Thomas Lamprecht
13184117e4 d/control: drop sdl dependency, we disable it on compile tinme
disabled via d/rules since a while...

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-24 07:51:19 +01:00
Thomas Lamprecht
aa4b14ea10 d/control: libaio1 is added by dh shlibs
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-24 07:51:19 +01:00
Thomas Lamprecht
3aa5b7598d enable zstd support
plan to use that for multifd migration, among other things

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-24 07:51:19 +01:00
Thomas Lamprecht
13d3e10aa6 compile in virgl support
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-24 07:51:19 +01:00
Thomas Lamprecht
58c3533a58 bump version to 6.2.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-18 14:23:41 +01:00
Thomas Lamprecht
fe0542bed9 d/rules: disable libssh by default
was always disabled in our clean builds, this now also avoids
auto-enabling it on "dirty" build hosts

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-18 14:21:53 +01:00
Fabian Ebner
f6d40bfdf4 add patch for loading a snapshot with qemu-img dd
Will be used when cloning from a qcow2 efidisk.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2022-02-15 14:03:07 +01:00
Fabian Ebner
107132becc fix getopt-string when introducing -n option for qemu-img dd
The colon after U is wrong, because it doesn't take an argument.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2022-02-15 14:03:07 +01:00
Fabian Ebner
4567474e95 update submodule and patches to 6.2.0
Notable changes:
* bdrv_co_p{discard,readv,writev,write_zeroes} function signatures
  changed, to using int64_t for offsets/bytes and some still had int
  rather than BrdvRequestFlags for the flags.
* job_cancel_sync now has a force parameter. Commit messages in
  73895f3838cd7fdaf185cf1dbc47be58844a966f
  4cfb3f05627ad82af473e7f7ae113c3884cd04e3
  sound like using force=true makes more sense.
* Added 3 patches coming in via qemu-stable tag, most important one is
  to work around a librbd issue.
* Added another 3 patches from qemu-devel to fix issue leading to
  crash when live migrating with iothread.
* cluster_size calculation helper changed (see patch pve/0026).
* QAPI's if conditionals now use 'CONFIG_FOO' rather than
  'defined(CONFIG_FOO)'

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2022-02-15 14:03:07 +01:00
Thomas Lamprecht
33a2d3a826 bump version to 6.1.1-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-02-14 15:53:18 +01:00
Fabian Ebner
2bf61c3eb6 vma: create: register all streams before entering coroutines
Otherwise, the header might already get written by a coroutine and
registering further streams will fail after that.

Also adds a missing g_list_free call for the other GList that's used.

Reported in the community forum:
https://forum.proxmox.com/threads/104744/

Reproducer script (increase beyond 30 if the issue isn't triggered yet):
> #!/usr/bin/perl
>
> my $dir = "./vma-create-bug";
> mkdir $dir;
>
> my $archive_path = "$dir/vzdump-qemu-104-2202_02_02-00_00_00.vma";
> unlink $archive_path;
>
> my $cmd = "vma create $archive_path -v";
> for (my $i = 0; $i < 30; $i++) {
>   system("truncate -s 1M $dir/drive-virtio$i.img");
>   $cmd .= " drive-virtio$i=$dir/drive-virtio$i.img";
> }
> system($cmd);

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2022-02-14 15:38:58 +01:00
Thomas Lamprecht
c07b2203b3 bump version to 6.1.1-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-01-13 10:57:48 +01:00
Thomas Lamprecht
ddbf7a872d update submodule and patches to 6.1.1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2022-01-13 10:56:39 +01:00
Thomas Lamprecht
95c7156d1e bump version to 6.1.0-3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-12-01 15:35:49 +01:00
Fabian Ebner
570d4ad51d fix #3738: cherry-pick "block: introduce max_hw_iov for use in scsi-generic"
which fixes the bad commit 18473467d55a20d643b6c9b3a52de42f705b4d35
that was tracked down via bisecting, and has a Cc for qemu-stable as
well.

Issue was easy enough to reproduce with a single virtio-block disk
using a few runs of dd if=/dev/urandom of=file bs=1M count=1000

Commit cc071629539dc1f303175a7e2d4ab854c0a8b20f upstream.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2021-12-01 15:34:27 +01:00
Dominik Csapak
c5e8e7c998 buildsys: fix build-dependencies on headers for 'vma' and 'pbs_restore'
both of them depend on generated header files, so we have to specify
them as sources. Otherwise, it happens (at least on some machines)
that they will be compiled before the headers are generated, aborting
the build.

Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
2021-11-18 08:11:57 +01:00
Fabian Grünbichler
7cf6b60926 fix #3728: handle machine without type
libguestfs starts their helper VMs with `-machine accel=..` without a
machine type, and our pve version suffix handling would segfault in that
case. there might be other scripted use cases that are affected as well.

this regression was introduced with the rebase of our patch set on top
of 6.1.0

Fixes: f376b2b9e2

Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2021-11-17 17:20:26 +01:00
Thomas Lamprecht
50999525c6 bump version to 6.1.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-11-16 09:38:18 +01:00
Fabian Grünbichler
edbcc10a69 cherry-pick segfault fix
this was reported multiple times in our forums[1 with backtraces, 2 & 3
with same log messages], fix is taken from upstream master.

1: https://forum.proxmox.com/threads/pve-7-0-14-1-vm-not-running-live-migration-kills-vm-post-ssd-move-pre-ram-move.99704/
2: https://forum.proxmox.com/threads/proxmox-7-0-14-1-crashes-vm-during-migrate-to-other-host.99678
3: https://forum.proxmox.com/threads/cannot-migrate-between-zfs-and-ceph.99685/#post-430152

Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2021-11-16 09:23:43 +01:00
Thomas Lamprecht
cc707c362e bump version to 6.1.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-10-13 17:58:18 +02:00
Stefan Reiter
af64ed13eb add fixup patch for qxl migration logic
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-10-13 17:58:18 +02:00
Stefan Reiter
f376b2b9e2 update and rebase to QEMU v6.1.0
Very clean rebase, only the +pve version handling needed manual fixing.
Drops two applied patches from extra/ and adds one new from upstream
(extra/0001*, fixes VNC over unix sockets) as well as 3 of my own for
allowing password changes on custom VNC displays again (as seen and
reviewed upstream, but not yet applied).

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-10-11 15:13:26 +02:00
Thomas Lamprecht
89fa943ef9 bump version to 6.0.0-4
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-09-06 07:30:05 +02:00
Stefan Reiter
26eee146bc add temporary QMP race fix
same as the initial version sent to qemu-devel, it won't be the final
fix we plan to upstream but it should be enough band-aid to
workaround how PVE uses the QMP.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
 [ Thomas: add a bit reasoning to commit message body ]
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-09-06 07:28:07 +02:00
Wolfgang Bumiller
277d33454f drop patch force-disabling smm
This drops debian/patches/pve/0005-PVE-Config-smm_available-false.patch
(and renumbers the remaining patches)

From what I could gather, this patch was originally added
due to issues with old kernels. Now we have users which
seem to run into issues *with* the patch.

All this does is toggle an option, and it's available via a
qemu CLI option anyway, so if dropping this patch causes
issues for some people we can just add an option to
qemu-server & UI control smm explicitly.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Cc: Alexandre Derumier <aderumier@odiso.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
2021-08-24 11:19:05 +02:00
Fabian Grünbichler
611b692181 bump version to 6.0.0-3
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2021-08-03 15:02:40 +02:00
Fabian Ebner
0114d3cd02 io_uring: resubmit when result is -EAGAIN
Linux SCSI can throw spurious -EAGAIN in some corner cases in its
completion path, which will end up being the result in the completed
io_uring request.

Resubmitting such requests should allow block jobs to complete, even
if such spurious errors are encountered.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2021-07-29 11:51:57 +02:00
Thomas Lamprecht
bb3e494bdc bump version to 6.0.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-06-23 11:04:47 +02:00
Stefan Reiter
403f23a0c3 enable io-uring support in QEMU builds
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-06-21 09:56:06 +02:00
Thomas Lamprecht
8893def37c bump version to 6.0.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-05-28 11:30:55 +02:00
Stefan Reiter
eb96e850ac debian: ignore submodule checks in QEMU build
...we do those manually, and the build dir is not a git repo.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-05-28 11:29:44 +02:00
Stefan Reiter
8dca018b68 udpate and rebase to QEMU v6.0.0
Mostly minor changes, bigger ones summarized:
* QEMU's internal backup code now uses a new async system, which allows
  parallel requests - the default max_workers settings is 64, I chose
  less, since 64 put enough stress on QEMU that the guest became
  practically unusable during the backup, and 16 still shows quite a
  nice measureable performance improvement. Little code changes for us
  though.
* 'malformed' QAPI parameters/functions are now a build error (i.e.
  using '_' vs '-'), I chose to just whitelist our calls in the name of
  backwards compatibility.
* monitor OOB race fix now uses the upstream variant, cherry-picked from
  origin/master since it's not in 6.0 by default
* last patch fixes a bug with snapshot rollback related to the new yank
  system

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-05-28 11:29:44 +02:00
Thomas Lamprecht
1cbf147251 bump version to 5.2.0-11
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-05-13 19:30:59 +02:00
Thomas Lamprecht
9e4cba3943 d/control: add libjson-perl to build dependencies
we use that to build the available machine/flags list

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-05-13 19:30:37 +02:00
Thomas Lamprecht
ebcd9ada10 bump version to 5.2.0-6
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-04-15 16:29:52 +02:00
Thomas Lamprecht
0a88214b72 alloc track: use coroutine version of bdrv_pwrite_zeroes
as we're in a coroutine here too

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-04-06 16:31:53 +02:00
Thomas Lamprecht
76e464784e pbs block driver: run read in the AIO context of the bs
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-04-06 16:31:53 +02:00
Thomas Lamprecht
b36e8acc31 alloc track: acquire BS AIO context during dropping
ran into this when live-restoring a backup configured for IO-threads,
got the good ol':
> qemu: qemu_mutex_unlock_impl: Operation not permitted
error.

Checking out the history of the related bdrv_backup_top_drop(*bs)
method, we can see that it used to do the AIO context acquiring too,
but in the backup path this was problematic and was changed to be
higher up in the call path in a upstream series from Stefan[0].

That said, this is a completely different code path and it is safe to
do so here. We always run from the main threads's AIO context here
and we call it only indirectly once, guarded by checking for
`s->drop_state == DropNone` and set `s->drop_state = DropRequested`
shortly before we schedule the track_drop() in a bh.

[0]: https://lists.gnu.org/archive/html/qemu-devel/2020-03/msg09139.html

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-04-06 16:27:48 +02:00
Thomas Lamprecht
aa42ea267e alloc track: keep track_drop() closer to similar block drivers
Reads just nicer with a drain begin *and* end call. Also clearing the
backing link of the alloc track BDS makes it closer to
bdrv_backup_top_drop() with which this driver has a bit in common.

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-04-06 16:27:37 +02:00
Thomas Lamprecht
6d6894394c bump version to 5.2.0-5
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-30 18:19:47 +02:00
Stefan Reiter
e79be6c6c4 add upstream fixes for qmp_block_resize
cherry-picked cleanly from 6.0 development tree, fixes an issue with
resizing RBD drives (and reportedly also on krbd or potentially other
storage backends) with iothreads.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-03-30 18:14:37 +02:00
Thomas Lamprecht
0be2cab670 bump version to 5.2.0-4
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-23 15:41:33 +01:00
Stefan Reiter
bb751cab32 Add tentative fix for QMP hang
Not exactly as sent upstream[0] since we're missing a change in our
v5.2.0 branch (irrelevant for us), but functionally works the same.

[0] https://lists.gnu.org/archive/html/qemu-devel/2021-03/msg07590.html
2021-03-22 16:52:40 +01:00
Thomas Lamprecht
154a7c0f8d d/control: drop unused python from dependencies
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-18 14:15:31 +01:00
Stefan Reiter
677d0d169f add alloc-track block driver patch
See added patches for more info, overview:
0044: slightly increase PBS performance by reducing allocations
0045: slightly increase block-stream performance for Ceph
0046: don't crash with block-stream on RBD
0047: add alloc-track driver for live restore

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-03-16 20:53:18 +01:00
Stefan Reiter
e9b36665c7 fix saving and loading dirty bitmaps in snapshots
Saving dirty bitmaps from our savevm-async code didn't work, since we
use a coroutine which holds the iothread mutex already (upstream savevm
is sync, migration uses a thread). Release the mutex before calling the
one function that (according to it's documentation) requires the lock to
*not* be held: qemu_savevm_state_pending.

Additionally, loading dirty bitmaps requires a call to
dirty_bitmap_mig_before_vm_start after "loadvm", which the upstream
savevm does explicitly afterwards - do that too.

This is exposed via the query-proxmox-support property
"pbs-dirty-bitmap-savevm".

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-03-16 20:44:06 +01:00
Thomas Lamprecht
970196fc93 bump version to 5.2.0-3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-05 16:23:16 +01:00
Thomas Lamprecht
6503e6e08d machine list: save as JSON
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-05 16:21:30 +01:00
Stefan Reiter
40e6b6e5a5 add ACPI compat patch for 5.1 and older machine types
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-03-05 15:20:14 +01:00
Stefan Reiter
a6ede89808 add static supported machines file
Same rationale as the CPU flags file, avoids calling QEMU binary just to
query machine types.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-03-05 15:20:14 +01:00
Stefan Reiter
2413972b46 move bitmap-mirror patches to seperate folder
...instead of having them in the middle of the backup related patches.
These might (hopefully) become upstream at some point as well.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-03 14:29:05 +01:00
Stefan Reiter
0c893fd820 clean up pve/ patches by squashing patches of patches
No functional change intended.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-03-03 14:29:05 +01:00
Thomas Lamprecht
25b9156448 bump version to 5.2.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-02-24 19:02:07 +01:00
Stefan Reiter
4194124719 pbs-restore: unref/close target block backend
Use blk_unref to drop the last reference, which will close the block
backend and flush all caches and outstanding writes.

This is especially important for restoring to Ceph, as the userspace
librbd caches will not be flushed if the application exits immediately,
leading to potentially incomplete restores.

Reported-by: Eneko Lacunza <elacunza@binovo.es>
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-02-24 19:02:07 +01:00
Thomas Lamprecht
42a90c4e1c d/patches: backport virtiofsd security fix
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-02-24 19:02:07 +01:00
Thomas Lamprecht
448136958e d/rules: build virtiofsd
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-02-24 19:02:07 +01:00
Thomas Lamprecht
8e231fbe8d bump version to 5.2.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-02-12 12:10:04 +01:00
Fabian Grünbichler
00fae7cdbe build: drop explicit libproxmox-backup-qemu0 dep
it ships a symbol file now, so it can be auto-generated based on the
build-dep and usage.

Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2021-02-12 12:07:12 +01:00
Stefan Reiter
0b8da68824 add PBS master key support
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-02-12 10:47:14 +01:00
Stefan Reiter
817b7667e8 Update to QEMU 5.2
Lots of patches touched and some slight changes to the build process
since QEMU switched to meson as their build system. Functionality-wise
very little rebasing required.

New patches introduced:
* pve/0058: to fix VMA backups and clean up some code in general with
  new 5.2 features now available to us (namely coroutine-enabled QMP).
* extra/0002: don't build man pages for guest agent when disabled
* extra/0003: fix live-migration with hugepages
* 0017 and 0018 are adjusted to fix snapshot abort and improve
  snap performance a bit

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2021-02-12 10:20:01 +01:00
Fabian Ebner
a16eaaffd3 fix #3084: fall back to open-iscsi initiatorname
Fixes vma restore when the target is an iSCSI storage which expects that
initiatorname. Also avoids the need to always explicitly set the initiatorname
in PVE code, thus fixing moving efidisks from and to such iSCSI storages.

Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
2021-02-06 15:09:15 +01:00
Wolfgang Bumiller
cfa22a5b2a bump version to 5.1.0-8
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2021-01-07 10:28:46 +01:00
Wolfgang Bumiller
b515d45e6b fix #3225: properly cancel jobs in 'created' state
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2021-01-07 10:26:37 +01:00
Stefan Reiter
3d785eabbe disable jemalloc
jemalloc does not play nice with our Rust library (proxmox-backup-qemu),
specifically it never releases memory allocated from Rust to the OS.
This leads to a problem with larger caches (e.g. for the PBS block driver).

It appears to be related to this GitHub issue:
https://github.com/jemalloc/jemalloc/issues/1398

The background_thread solution seems weirdly hacky, so let's disable
jemalloc entirely for now.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-12-15 14:43:34 +01:00
Thomas Lamprecht
55df4a9eb1 bump version to 5.1.0-7
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-11-25 14:09:24 +01:00
Stefan Reiter
cfe02b3b4e update patches with some pbs-state migration cleanups
...and literal cleanup, as in, call save_cleanup after success or error.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-11-25 11:49:06 +01:00
Stefan Reiter
32ee41155b update patches with squashed in 'include library version'
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>

bump build-dependency on libproxmox-backup-qemu0-dev with version query
support

Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2020-11-25 11:49:06 +01:00
Thomas Lamprecht
424f6841d5 bump version to 5.1.0-6
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-11-05 18:59:44 +01:00
Thomas Lamprecht
66eae0ae75 fix dirty-bitmap state migration freeze
The idea in general is to migrate all the state, which is small for
us, in a single step once. But, QEMU only calls save state if we
return active true.

Hardcoding is-active to return true, like done initially, makes the
migration freeze, as QEMU thinks this is never done, and only stops
calling us and finishes after a few seconds.

So, add a state with an "active" boolean, set to true when
initializing a migration, and set it to false when the state was
saved.

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-11-05 18:58:15 +01:00
Thomas Lamprecht
2130e925a8 bump version to 5.1.0-5
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-11-04 18:36:42 +01:00
Thomas Lamprecht
f36fa39113 migration/block-dirty-bitmap: migrate other bitmaps even if one fails
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-11-04 18:35:50 +01:00
Thomas Lamprecht
50eb3ea618 bump version to 5.1.0-4
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-10-29 18:09:22 +01:00
Thomas Lamprecht
3e2a0699e6 d/control: update versioned dependency for proxmox backup qemu library
to have the proxmox_export_state et al. available

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-10-29 18:07:14 +01:00
Thomas Lamprecht
d95ad93eed apply dirty-bitmap state migration + fix
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-10-29 18:05:43 +01:00
Stefan Reiter
72ae34ecce Several fixes for backup abort and error reporting
Also add my Signed-off-by to some patches where it was missing.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-10-29 17:57:47 +01:00
Thomas Lamprecht
60a607ad0f bump version to 5.1.0-3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-09-29 09:23:06 +02:00
Stefan Reiter
d333327a1b Add transaction patches and fix for blocking finish
With the transaction patches, patch 0026-PVE-Backup-modify-job-api.patch
is no longer necessary, so drop it and rebase all following patches on
top.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-09-29 09:21:15 +02:00
Thomas Lamprecht
250d87c276 bump version to 5.1.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-09-14 20:01:32 +02:00
Thomas Lamprecht
4b7a18845c cherry-pick: "usb: fix setup_len init (CVE-2020-14364)"
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-09-14 19:38:34 +02:00
Thomas Lamprecht
7895b0d523 work around #3002: revert "qemu-img convert: Don't pre-zero images"
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-09-14 19:37:45 +02:00
Stefan Reiter
437d68473c Add systemd journal logging patch
Prints QEMU errors that occur *after* the "-daemonize" fork to the
systemd journal, instead of pushing them into /dev/null like before.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-09-08 17:13:29 +02:00
Thomas Lamprecht
46c0ec19ad bump version to 5.1.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-08-20 14:48:58 +02:00
Fabian Grünbichler
a5feeebc51 allow backup of read-only block drives
this is needed for template backups with PBS until we have the backup
equivalent of 'pbs-restore'.

Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2020-08-20 14:11:39 +02:00
Stefan Reiter
60ae3775bf update to QEMU 5.1
No major semantic changes, mostly just deprecations and changed function
signatures. Drop the extra/ patches, as they have been applied upstream.

The added extra/ patch was accepted upstream[0] but has not been picked
up for 5.1. It is required for non-4M aligned backups to work with PBS.

[0] https://lists.gnu.org/archive/html/qemu-devel/2020-08/msg01671.html

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-08-20 13:40:36 +02:00
Thomas Lamprecht
878df11e78 bump version to 5.0.0-13
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-08-19 18:19:35 +02:00
Thomas Lamprecht
f00a720d7e PVE: add query-pbs-bitmap-info QMP call
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-08-19 18:11:23 +02:00
Thomas Lamprecht
c5f7dc1d72 PVE: add zero block handling to PBS dump callback
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-08-19 13:56:03 +02:00
Fabian Grünbichler
dde77467ca bump version to 5.0.0-12
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2020-08-11 11:29:21 +02:00
Fabian Grünbichler
2821f02d70 fix PBS write callback with big blocks
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
2020-08-11 11:14:36 +02:00
Oguz Bektas
95fd47ecb9 patch for possible DOS in qemu network packet processing
fixes an assertion failure in qemu network packet processing, which can
lead to DOS'ing the qemu process on the host. this affects 'e1000e' and
'vmxnet3' network devices.

patch is cherry-picked from the commit mentioned in the oss-security email.

more info on oss-security [0]

[0]: https://www.openwall.com/lists/oss-security/2020/08/10/1

Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
2020-08-11 11:08:39 +02:00
Thomas Lamprecht
41424ed9c8 bump version to 5.0.0-11
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-14 12:45:24 +02:00
Stefan Reiter
f257cc05f4 Fix dirty-bitmap PBS backup with multiple drives
"PVE backup: rename incremental to use-dirty-bitmap" merged two
variables (use_dirty_bitmap and incremental) into one, but they served
two different purposes. Rename the original use_dirty_bitmap to
"expect_only_dirty" so the new one doesn't conflict, and rework "PVE:
use proxmox_backup_check_incremental" around that semantic.

In practice, this had the effect that only one disk at a time would
have a bitmap added, as after the first "use_dirty_bitmap" would be set
to one and the rest would behave as if the QMP parameter of the same
name was unset.

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-07-14 10:46:48 +02:00
Wolfgang Bumiller
9886892f10 bump version to 5.0.0-10
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2020-07-10 13:33:44 +02:00
Wolfgang Bumiller
6d46b2ff4c fix backup qmp parameters to pass along encryption info
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2020-07-10 13:31:52 +02:00
Thomas Lamprecht
2e4f5f2a90 bump version to 5.0.0-9
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-09 13:18:48 +02:00
Thomas Lamprecht
1d606ec161 d/control: update build-dependency of libproxmox-backup-qemu0-dev
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-09 13:17:29 +02:00
Thomas Lamprecht
3499c5b45a PBS patches: block driver, adapat encrypt/compress param, add query-proxmox-support QMP cmd
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-09 13:15:49 +02:00
Thomas Lamprecht
102ddd7e59 bump version to 5.0.0-8
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-06 23:05:42 +02:00
Thomas Lamprecht
4c17eebee4 fixup: proxmox_backup_check_incremental is negated
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-06 23:05:42 +02:00
Thomas Lamprecht
dfed71b229 bump version to 5.0.0-7
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-06 22:19:12 +02:00
Thomas Lamprecht
3ab149ccdd update/add PBS integration patches
* rename "incremental" param to "use-dirty-bitmap", avoids confusion
  as the backup can be incrementally also with that param set to
  false.
* use new proxmox_backup_check_incremental
* fix setting dirty counter and adapt to new connect API semantic

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-06 22:13:12 +02:00
Thomas Lamprecht
1f8140323f bump version to 5.0.0-6
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-03 19:26:09 +02:00
Thomas Lamprecht
15b9c76e1f pbs: query-backup: set reused field also for dirty-bitmap
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-03 19:26:09 +02:00
Thomas Lamprecht
d7f4e01a34 debian/patches: squash some followup patches and regroup a bit more together
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-02 13:33:16 +02:00
Thomas Lamprecht
20be7fa0a0 backup: improve QAPI info and remove all dirty-bitmaps on failed drive-job
effectively two commits merged as one:
https://pve.proxmox.com/pipermail/pve-devel/2020-July/044185.html
https://pve.proxmox.com/pipermail/pve-devel/2020-July/044194.html

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-07-02 13:03:49 +02:00
Thomas Lamprecht
0943af81a6 bump version to 5.0.0-5
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-06-30 12:59:26 +02:00
Thomas Lamprecht
04e436ea6b d/control: build depend on newer libproxmox-backup-qemu0-dev
to ensure we have the new 'incremental' parameter for the
proxmox_backup_register_image and proxmox_backup_register_image_async
functions available

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-06-30 12:53:01 +02:00
Dietmar Maurer
c96a4a38cb add incremental backup patches
and fix typo: s/BPS/PBS/
2020-06-30 10:34:00 +02:00
Stefan Reiter
f0b53ef0b2 fix #2794: Include legacy-igd passthrough fix
See https://bugs.launchpad.net/qemu/+bug/1882784

Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
2020-06-25 13:24:34 +02:00
Thomas Lamprecht
b570f1c41e Fix backup for not 64k-aligned storages
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Tested-by: Roland Kammerer <roland.kammerer@linbit.com>
2020-06-24 16:26:30 +02:00
Thomas Lamprecht
fff7e250ee pbs-restore: flush verbose log before calling into library
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-06-02 13:20:56 +02:00
Thomas Lamprecht
5107839915 bump version to 5.0.0-4
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-30 18:48:00 +02:00
Thomas Lamprecht
bce72611f9 pbs-restore: be more verbose if asked to
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-30 18:48:00 +02:00
Thomas Lamprecht
c6979241f1 small cleanups for pbs-restore
Add trailing newline to two error messages, and drop an extra
unconditional `qdict_put_str(options, "driver", format);`
Besides that it's just formatting.

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-30 15:24:20 +02:00
Thomas Lamprecht
4e74eca7ed install pbs-restore to usr/bin
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-30 13:54:08 +02:00
Thomas Lamprecht
62b7007c2a bump version to 5.0.0-3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-27 18:54:06 +02:00
Thomas Lamprecht
f063a8aadb fix vmstate-snapshots with iothread=1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-27 18:54:06 +02:00
Thomas Lamprecht
63be960208 bump version to 5.0.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-05-04 15:23:37 +02:00
Wolfgang Bumiller
abc9e57f05 add optional buffer size to QEMUFile
and use 4M for our savevm-async buffer size

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2020-05-04 15:09:10 +02:00
Wolfgang Bumiller
24e289d025 experimentally move savevm-async back into a coroutine
Move qemu_savevm_state_{header,setup} into the main loop and
the rest of the iteration into a coroutine. The former need
to lock the iothread (and we can't unlock it in the
coroutine), and the latter can't deal with being in a
separate thread, so a coroutine it must be.

Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
2020-05-04 15:09:10 +02:00
Thomas Lamprecht
9d5bec4ded bump version to 5.0.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-29 09:54:19 +02:00
Thomas Lamprecht
fa5ebda23f bump version to 5.0.0~rc3-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-17 09:11:27 +02:00
Thomas Lamprecht
751ed3661b avoid using QemuRecMutex inside coroutines, QemuMutex to lock outside
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-17 09:09:12 +02:00
Thomas Lamprecht
9e4a452f24 bump version to 5.0.0~rc3-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-16 16:53:20 +02:00
Thomas Lamprecht
9ba27d39e0 d/control: build deps; switch libcap to libcap-ng, add libattr
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-16 16:53:20 +02:00
Thomas Lamprecht
d5619ef646 pve backup: avoid using recursive mutex
Release the lock on qemu_coroutine_yield, so coroutines don't
deadlock.

Originally-by: Stefan Reiter <s.reiter@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-16 15:20:22 +02:00
Thomas Lamprecht
b7e851a529 rebase patches onto v5.0.0-rc3
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-16 15:19:31 +02:00
Thomas Lamprecht
ad9dbe4930 bump version to 5.0.0~rc2-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-08 15:36:32 +02:00
Thomas Lamprecht
83faa3fe30 import QEMU 5.0.0-rc2 and rebase patches
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-04-08 15:36:32 +02:00
Thomas Lamprecht
e791d95bb8 bump version to 4.2.0-2
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-03-25 12:20:15 +01:00
Thomas Lamprecht
48d490fe54 patches: move upstream cherry picked to extra folder, like we did in the past
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-03-23 10:16:35 +01:00
Fabian Grünbichler
059a9447e2 add bitmap drive-mirror patches
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-03-23 10:15:22 +01:00
Dominic Jäger
66e3df5d45 vma_writer: Display more error information
Also print the reason why the function vma_writer_register_stream failed to
help debug errors like in [0].

[0] https://forum.proxmox.com/threads/backup-error-vma_writer_register_stream-drive-scsi0-failed-pve-6-1-7.65925/

Signed-off-by: Dominic Jäger <d.jaeger@proxmox.com>
Reviewed-by: Stefan Reiter <s.reiter@proxmox.com>
2020-03-19 18:58:47 +01:00
Oguz Bektas
1be32c854f slirp: re-add security patches
the first two patches were mistakenly left out during the 4.2 qemu
rebase.

also adds another patch for issue CVE-2019-14378 (heap-based BOF)

Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
2020-03-13 13:20:11 +01:00
Dietmar Maurer
20505f521e fixup backup with io-threads 2020-03-12 10:30:18 +01:00
Thomas Lamprecht
c3e8606cd9 bump version to 4.2.0-1
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-03-11 07:26:21 +01:00
Thomas Lamprecht
6402d96100 update submodule and patches for 4.2.0
The long overdue nice rebase+cleanup was done by Dietmar

Originally-by: Dietmar Maurer <dietmar@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
2020-03-11 07:26:21 +01:00