Add bdrv_co_block_status

Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>

Makefile

@@ -19,6 +19,9 @@ submodule:
 	test -f "${SRCDIR}/configure" || git submodule update --init --recursive
 
 $(BUILDDIR): keycodemapdb | submodule
+	# check if qemu/ was used for a build
+	# if so, please run 'make distclean' in the submodule and try again
+	test ! -f $(SRCDIR)/build/config.status
 	rm -rf $(BUILDDIR)
 	cp -a $(SRCDIR) $(BUILDDIR)
 	cp -a debian $(BUILDDIR)/debian
@@ -30,7 +33,7 @@ $(BUILDDIR): keycodemapdb | submodule
 deb kvm: $(DEBS)
 $(DEB_DBG): $(DEB)
 $(DEB): $(BUILDDIR)
-	cd $(BUILDDIR); dpkg-buildpackage -b -us -uc -j
+	cd $(BUILDDIR); dpkg-buildpackage -b -us -uc -j32
 	lintian $(DEBS)
 
 .PHONY: update

debian/changelog

@@ -1,7 +1,215 @@
+pve-qemu-kvm (5.2.0-11+vitastor2) buster; urgency=medium
+
+  * Add bdrv_co_block_status implementation for QCOW2 export support
+
+ -- Vitaliy Filippov <vitalif@yourcmc.ru>  Fri, 13 Jan 2023 20:26:21 +0300
+
+pve-qemu-kvm (5.2.0-11+vitastor1) buster; urgency=medium
+
+  * Add Vitastor support
+
+ -- Vitaliy Filippov <vitalif@yourcmc.ru>  Sun, 05 Dec 2021 17:22:43 +0300
+
+pve-qemu-kvm (5.2.0-11) bullseye; urgency=medium
+
+  * re-build for Proxmox VE 7 / Debian Bullseye
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 13 May 2021 14:03:00 +0200
+
+pve-qemu-kvm (5.2.0-6) pve; urgency=medium
+
+  * improve the alloc-track and Proxmox Backup Server special block driver when
+    used with IO-threads for the new live-restore feature
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 15 Apr 2021 16:29:48 +0200
+
+pve-qemu-kvm (5.2.0-5) pve; urgency=medium
+
+  * cherry-pick fixes for a possible deadlock when resizing a disk while using
+    IO-threads
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 30 Mar 2021 18:18:18 +0200
+
+pve-qemu-kvm (5.2.0-4) pve; urgency=medium
+
+  * monitor/qmp: fix race on CHR_EVENT_CLOSED without OOB
+
+  * improve saving and loading dirty bitmaps in live-snapshots
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 23 Mar 2021 15:41:26 +0100
+
+pve-qemu-kvm (5.2.0-3) pve; urgency=medium
+
+  * backport "i386/acpi: restore device paths for pre-5.1 vms" patch
+
+  * ship list of 'i440fx' and 'q35' machine versions this QEMU version supports
+
+ -- Proxmox Support Team <support@proxmox.com>  Fri, 05 Mar 2021 16:23:06 +0100
+
+pve-qemu-kvm (5.2.0-2) pve; urgency=medium
+
+  * Proxmox Backup restore: ensure all caches are flushed for userspace
+    attached storages (for example, Ceph when librbd and not KRBD is used)
+
+  * ship VirtIOFSd daemon
+
+ -- Proxmox Support Team <support@proxmox.com>  Wed, 24 Feb 2021 18:25:14 +0100
+
+pve-qemu-kvm (5.2.0-1) pve; urgency=medium
+
+  * update to QEMU stable release 5.2.0
+
+  * fix #3084: fall back to open-iscsi initiatorname
+
+  * fix snapshot abort and improve performance in some edge cases
+
+  * add basis for Proxmox Backup Server master key support
+
+ -- Proxmox Support Team <support@proxmox.com>  Fri, 12 Feb 2021 12:09:59 +0100
+
+pve-qemu-kvm (5.1.0-8) pve; urgency=medium
+
+  * disable jemalloc, as it does not play nice with our library written in
+    rust
+
+  * fix #3225: properly cancel jobs in 'created' state when cancelling or
+    failing backup jobs
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 07 Jan 2021 10:27:33 +0100
+
+pve-qemu-kvm (5.1.0-7) pve; urgency=medium
+
+  * allow to query the loaded Proxmox Backup library version
+
+ -- Proxmox Support Team <support@proxmox.com>  Wed, 25 Nov 2020 14:09:16 +0100
+
+pve-qemu-kvm (5.1.0-6) pve; urgency=medium
+
+  * migration/block-dirty-bitmap: avoid telling QEMU that the bitmap migration
+    is active longer than required
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 05 Nov 2020 18:59:40 +0100
+
+pve-qemu-kvm (5.1.0-5) pve; urgency=medium
+
+  * migration/block-dirty-bitmap: migrate other bitmaps even if one fails
+
+ -- Proxmox Support Team <support@proxmox.com>  Wed, 04 Nov 2020 18:36:32 +0100
+
+pve-qemu-kvm (5.1.0-4) pve; urgency=medium
+
+  * several fixes for backup abort edgecases and error reporting
+
+  * allow to migrate dirty-bitmap and Proxmox Backup Server state
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 29 Oct 2020 18:09:16 +0100
+
+pve-qemu-kvm (5.1.0-3) pve; urgency=medium
+
+  * backup: make more use of coroutines and do not block on finishing
+
+  * backup: use transactions to synchronize the disk job states
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 29 Sep 2020 09:22:56 +0200
+
+pve-qemu-kvm (5.1.0-2) pve; urgency=medium
+
+  * cherry-pick fix to harden checks for USB devices (CVE-2020-14364)
+
+  * work around #3002: revert "qemu-img convert: Don't pre-zero images" as it
+    correlates with an issue when using LVM as a target storage for disk move
+    operations.
+
+ -- Proxmox Support Team <support@proxmox.com>  Mon, 14 Sep 2020 20:01:24 +0200
+
+pve-qemu-kvm (5.1.0-1) pve; urgency=medium
+
+  * update to QEMU 5.1.0
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 20 Aug 2020 13:42:10 +0200
+
+pve-qemu-kvm (5.0.0-13) pve; urgency=medium
+
+  * improve zero block handling for PBS backups
+
+  * allow querying a more detailed dirty-bitmap state per VM disk
+
+ -- Proxmox Support Team <support@proxmox.com>  Wed, 19 Aug 2020 18:19:26 +0200
+
+pve-qemu-kvm (5.0.0-12) pve; urgency=medium
+
+  * patch for possible DOS in qemu network packet processing
+
+  * fix PBS write callback with big blocks
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 11 Aug 2020 11:29:13 +0200
+
+pve-qemu-kvm (5.0.0-11) pve; urgency=medium
+
+  * improve dirty-bitmap Proxmox Backup Server backup with multiple drives
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 14 Jul 2020 12:44:25 +0200
+
+pve-qemu-kvm (5.0.0-10) pve; urgency=medium
+
+  * fix compression and encryption related backup parameters not being passed
+    on from the HMP command properly
+
+ -- Proxmox Support Team <support@proxmox.com>  Fri, 10 Jul 2020 13:32:11 +0200
+
+pve-qemu-kvm (5.0.0-9) pve; urgency=medium
+
+  * adapt to new compress and encrypt params of the proxmox-backup library
+
+  * add block driver for Proxmox Backup Server backed images
+
+  * add 'query-proxmox-support' QMP command
+
+ -- Proxmox Support Team <support@proxmox.com>  Thu, 09 Jul 2020 13:18:45 +0200
+
+pve-qemu-kvm (5.0.0-8) pve; urgency=medium
+
+  * backup: rename parameter for dirty-bitmap PBS backup to "use-dirty-bitmap"
+
+  * backup: improve checking if a previous backup is available when trying to
+    reuse it
+
+ -- Proxmox Support Team <support@proxmox.com>  Mon, 06 Jul 2020 22:58:45 +0200
+
+pve-qemu-kvm (5.0.0-6) pve; urgency=medium
+
+  * backup: improve query-backup information and remove all dirty-bitmaps on
+    failed drive-job
+
+ -- Proxmox Support Team <support@proxmox.com>  Fri, 03 Jul 2020 17:00:30 +0200
+
+pve-qemu-kvm (5.0.0-5) pve; urgency=medium
+
+  * fix backup for not 64k-aligned storages
+
+  * fix #2794: Include legacy-igd passthrough fix
+
+  * add initial support for incremental backup for running VMs and
+    Proxmox Backup Server as a target
+
+ -- Proxmox Support Team <support@proxmox.com>  Tue, 30 Jun 2020 11:12:55 +0200
+
+pve-qemu-kvm (5.0.0-4) pve; urgency=medium
+
+  * install missing restore helper binary
+
+ -- Proxmox Support Team <support@proxmox.com>  Sat, 30 May 2020 15:25:38 +0200
+
+pve-qemu-kvm (5.0.0-3) pve; urgency=medium
+
+  * ensure that a data-flush for all drives uses the correct AioContext. Fixes
+    a potential VM hang happening on some storage types if IOThreads are used.
+
+ -- Proxmox Support Team <support@proxmox.com>  Wed, 27 May 2020 14:41:31 +0200
+
 pve-qemu-kvm (5.0.0-2) pve; urgency=medium
 
-  * fix saving a VM-state (snapshot, suspend-to-disk) in combination with
-    IOThreads
+  * fix saving a VM-state (snapshot, suspend-to-disk) with QEMU 5.0
 
   * try to use bigger chunks for saving a VM-state to improve performance on
     storage backends like Ceph RBD

debian/control

@@ -16,17 +16,20 @@ Build-Depends: autotools-dev,
                libiscsi-dev (>= 1.12.0),
                libjemalloc-dev,
                libjpeg-dev,
+               libjson-perl,
                libnuma-dev,
                libpci-dev,
                libpixman-1-dev,
-               libproxmox-backup-qemu0-dev,
+               libproxmox-backup-qemu0-dev (>= 1.0.3-1),
                librbd-dev (>= 0.48),
                libsdl1.2-dev,
                libseccomp-dev,
                libspice-protocol-dev (>= 0.12.14~),
                libspice-server-dev (>= 0.14.0~),
+               libsystemd-dev,
                libusb-1.0-0-dev (>= 1.0.17-1),
                libusbredirparser-dev (>= 0.6-2),
+               meson,
                python3-minimal,
                python3-sphinx,
                quilt,
@@ -57,7 +60,6 @@ Depends: ceph-common (>= 0.48),
          libusbredirparser1 (>= 0.6-2),
          libuuid1,
          numactl,
-         python,
          ${misc:Depends},
          ${shlibs:Depends},
 Conflicts: kvm,

debian/parse-machines.pl

@@ -0,0 +1,27 @@
+#!/usr/bin/perl
+
+use warnings;
+use strict;
+
+use JSON;
+
+my $machines = [];
+
+while (<STDIN>) {
+    if (/^\s*Supported machines are:/) {
+	next;
+    }
+
+    s/^\s+//;
+    my @machine = split(/\s+/);
+    next if $machine[0] !~ m/^pc-(i440fx|q35)-(.+)$/;
+    push @$machines, {
+        'id' => $machine[0],
+        'type' => $1,
+        'version' => $2,
+    };
+}
+
+die "no QEMU machine types detected from STDIN input" if scalar (@$machines) <= 0;
+
+print to_json($machines, { utf8 => 1 }) or die "$!\n";

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

@@ -35,10 +35,10 @@ Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
  5 files changed, 145 insertions(+), 29 deletions(-)
 
 diff --git a/block/mirror.c b/block/mirror.c
-index c26fd9260d..3c9cd42c50 100644
+index 8e1ad6eceb..97843992c2 100644
 --- a/block/mirror.c
 +++ b/block/mirror.c
-@@ -49,7 +49,7 @@ typedef struct MirrorBlockJob {
+@@ -50,7 +50,7 @@ typedef struct MirrorBlockJob {
      BlockDriverState *to_replace;
      /* Used to block operations on the drive-mirror-replace target */
      Error *replace_blocker;
@@ -47,7 +47,7 @@ index c26fd9260d..3c9cd42c50 100644
      BlockMirrorBackingMode backing_mode;
      /* Whether the target image requires explicit zero-initialization */
      bool zero_target;
-@@ -64,6 +64,8 @@ typedef struct MirrorBlockJob {
+@@ -65,6 +65,8 @@ typedef struct MirrorBlockJob {
      size_t buf_size;
      int64_t bdev_length;
      unsigned long *cow_bitmap;
@@ -56,17 +56,17 @@ index c26fd9260d..3c9cd42c50 100644
      BdrvDirtyBitmap *dirty_bitmap;
      BdrvDirtyBitmapIter *dbi;
      uint8_t *buf;
-@@ -676,7 +678,8 @@ static int mirror_exit_common(Job *job)
+@@ -677,7 +679,8 @@ static int mirror_exit_common(Job *job)
      bdrv_child_refresh_perms(mirror_top_bs, mirror_top_bs->backing,
                               &error_abort);
      if (!abort && s->backing_mode == MIRROR_SOURCE_BACKING_CHAIN) {
 -        BlockDriverState *backing = s->is_none_mode ? src : s->base;
 +        BlockDriverState *backing;
 +        backing = s->sync_mode == MIRROR_SYNC_MODE_NONE ? src : s->base;
-         if (backing_bs(target_bs) != backing) {
-             bdrv_set_backing_hd(target_bs, backing, &local_err);
-             if (local_err) {
-@@ -771,6 +774,16 @@ static void mirror_abort(Job *job)
+         BlockDriverState *unfiltered_target = bdrv_skip_filters(target_bs);
+ 
+         if (bdrv_cow_bs(unfiltered_target) != backing) {
+@@ -774,6 +777,16 @@ static void mirror_abort(Job *job)
      assert(ret == 0);
  }
  
@@ -83,7 +83,7 @@ index c26fd9260d..3c9cd42c50 100644
  static void coroutine_fn mirror_throttle(MirrorBlockJob *s)
  {
      int64_t now = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
-@@ -949,7 +962,8 @@ static int coroutine_fn mirror_run(Job *job, Error **errp)
+@@ -955,7 +968,8 @@ 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);
@@ -93,7 +93,7 @@ index c26fd9260d..3c9cd42c50 100644
          ret = mirror_dirty_init(s);
          if (ret < 0 || job_is_cancelled(&s->common.job)) {
              goto immediate_exit;
-@@ -1181,6 +1195,7 @@ static const BlockJobDriver mirror_job_driver = {
+@@ -1188,6 +1202,7 @@ static const BlockJobDriver mirror_job_driver = {
          .run                    = mirror_run,
          .prepare                = mirror_prepare,
          .abort                  = mirror_abort,
@@ -101,7 +101,7 @@ index c26fd9260d..3c9cd42c50 100644
          .pause                  = mirror_pause,
          .complete               = mirror_complete,
      },
-@@ -1196,6 +1211,7 @@ static const BlockJobDriver commit_active_job_driver = {
+@@ -1203,6 +1218,7 @@ static const BlockJobDriver commit_active_job_driver = {
          .run                    = mirror_run,
          .prepare                = mirror_prepare,
          .abort                  = mirror_abort,
@@ -109,7 +109,7 @@ index c26fd9260d..3c9cd42c50 100644
          .pause                  = mirror_pause,
          .complete               = mirror_complete,
      },
-@@ -1542,7 +1558,10 @@ static BlockJob *mirror_start_job(
+@@ -1550,7 +1566,10 @@ static BlockJob *mirror_start_job(
                               BlockCompletionFunc *cb,
                               void *opaque,
                               const BlockJobDriver *driver,
@@ -121,7 +121,7 @@ index c26fd9260d..3c9cd42c50 100644
                               bool auto_complete, const char *filter_node_name,
                               bool is_mirror, MirrorCopyMode copy_mode,
                               Error **errp)
-@@ -1555,10 +1574,39 @@ static BlockJob *mirror_start_job(
+@@ -1563,10 +1582,39 @@ static BlockJob *mirror_start_job(
      Error *local_err = NULL;
      int ret;
  
@@ -163,7 +163,7 @@ index c26fd9260d..3c9cd42c50 100644
      assert(is_power_of_2(granularity));
  
      if (buf_size < 0) {
-@@ -1662,7 +1710,9 @@ static BlockJob *mirror_start_job(
+@@ -1705,7 +1753,9 @@ static BlockJob *mirror_start_job(
      s->replaces = g_strdup(replaces);
      s->on_source_error = on_source_error;
      s->on_target_error = on_target_error;
@@ -174,7 +174,7 @@ index c26fd9260d..3c9cd42c50 100644
      s->backing_mode = backing_mode;
      s->zero_target = zero_target;
      s->copy_mode = copy_mode;
-@@ -1682,6 +1732,18 @@ static BlockJob *mirror_start_job(
+@@ -1726,6 +1776,18 @@ static BlockJob *mirror_start_job(
          bdrv_disable_dirty_bitmap(s->dirty_bitmap);
      }
  
@@ -193,7 +193,7 @@ index c26fd9260d..3c9cd42c50 100644
      ret = block_job_add_bdrv(&s->common, "source", bs, 0,
                               BLK_PERM_WRITE_UNCHANGED | BLK_PERM_WRITE |
                               BLK_PERM_CONSISTENT_READ,
-@@ -1735,6 +1797,9 @@ fail:
+@@ -1803,6 +1865,9 @@ fail:
          if (s->dirty_bitmap) {
              bdrv_release_dirty_bitmap(s->dirty_bitmap);
          }
@@ -203,7 +203,7 @@ index c26fd9260d..3c9cd42c50 100644
          job_early_fail(&s->common.job);
      }
  
-@@ -1752,29 +1817,23 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
+@@ -1820,29 +1885,23 @@ 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,
@@ -227,7 +227,7 @@ index c26fd9260d..3c9cd42c50 100644
 -        return;
 -    }
 -    is_none_mode = mode == MIRROR_SYNC_MODE_NONE;
-     base = mode == MIRROR_SYNC_MODE_TOP ? backing_bs(bs) : NULL;
+     base = mode == MIRROR_SYNC_MODE_TOP ? bdrv_backing_chain_next(bs) : NULL;
      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,
@@ -238,7 +238,7 @@ index c26fd9260d..3c9cd42c50 100644
  }
  
  BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
-@@ -1800,7 +1859,8 @@ BlockJob *commit_active_start(const char *job_id, BlockDriverState *bs,
+@@ -1868,7 +1927,8 @@ 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,
@@ -249,10 +249,10 @@ index c26fd9260d..3c9cd42c50 100644
                       &local_err);
      if (local_err) {
 diff --git a/blockdev.c b/blockdev.c
-index f391c3b3c7..bbeff9c439 100644
+index fe6fb5dc1d..394920613d 100644
 --- a/blockdev.c
 +++ b/blockdev.c
-@@ -3159,6 +3159,10 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
+@@ -2930,6 +2930,10 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
                                     BlockDriverState *target,
                                     bool has_replaces, const char *replaces,
                                     enum MirrorSyncMode sync,
@@ -263,15 +263,15 @@ index f391c3b3c7..bbeff9c439 100644
                                     BlockMirrorBackingMode backing_mode,
                                     bool zero_target,
                                     bool has_speed, int64_t speed,
-@@ -3177,6 +3181,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
-                                    Error **errp)
+@@ -2949,6 +2953,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
  {
+     BlockDriverState *unfiltered_bs;
      int job_flags = JOB_DEFAULT;
 +    BdrvDirtyBitmap *bitmap = NULL;
  
      if (!has_speed) {
          speed = 0;
-@@ -3231,6 +3236,29 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
+@@ -3003,6 +3008,29 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
          sync = MIRROR_SYNC_MODE_FULL;
      }
  
@@ -298,10 +298,10 @@ index f391c3b3c7..bbeff9c439 100644
 +        }
 +    }
 +
-     if (has_replaces) {
-         BlockDriverState *to_replace_bs;
-         AioContext *replace_aio_context;
-@@ -3268,8 +3296,8 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
+     if (!has_replaces) {
+         /* We want to mirror from @bs, but keep implicit filters on top */
+         unfiltered_bs = bdrv_skip_implicit_filters(bs);
+@@ -3049,8 +3077,8 @@ 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,
@@ -312,7 +312,7 @@ index f391c3b3c7..bbeff9c439 100644
                   on_source_error, on_target_error, unmap, filter_node_name,
                   copy_mode, errp);
  }
-@@ -3410,6 +3438,8 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
+@@ -3195,6 +3223,8 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
  
      blockdev_mirror_common(arg->has_job_id ? arg->job_id : NULL, bs, target_bs,
                             arg->has_replaces, arg->replaces, arg->sync,
@@ -321,7 +321,7 @@ index f391c3b3c7..bbeff9c439 100644
                             backing_mode, zero_target,
                             arg->has_speed, arg->speed,
                             arg->has_granularity, arg->granularity,
-@@ -3432,6 +3462,8 @@ void qmp_blockdev_mirror(bool has_job_id, const char *job_id,
+@@ -3216,6 +3246,8 @@ void qmp_blockdev_mirror(bool has_job_id, const char *job_id,
                           const char *device, const char *target,
                           bool has_replaces, const char *replaces,
                           MirrorSyncMode sync,
@@ -330,7 +330,7 @@ index f391c3b3c7..bbeff9c439 100644
                           bool has_speed, int64_t speed,
                           bool has_granularity, uint32_t granularity,
                           bool has_buf_size, int64_t buf_size,
-@@ -3482,7 +3514,8 @@ void qmp_blockdev_mirror(bool has_job_id, const char *job_id,
+@@ -3265,7 +3297,8 @@ void qmp_blockdev_mirror(bool has_job_id, const char *job_id,
      }
  
      blockdev_mirror_common(has_job_id ? job_id : NULL, bs, target_bs,
@@ -341,10 +341,10 @@ index f391c3b3c7..bbeff9c439 100644
                             has_granularity, granularity,
                             has_buf_size, buf_size,
 diff --git a/include/block/block_int.h b/include/block/block_int.h
-index 6d234f1de9..180a5e00fd 100644
+index 95d9333be1..6f8eda629a 100644
 --- a/include/block/block_int.h
 +++ b/include/block/block_int.h
-@@ -1210,7 +1210,9 @@ void mirror_start(const char *job_id, BlockDriverState *bs,
+@@ -1230,7 +1230,9 @@ 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,
@@ -356,10 +356,10 @@ index 6d234f1de9..180a5e00fd 100644
                    BlockdevOnError on_source_error,
                    BlockdevOnError on_target_error,
 diff --git a/qapi/block-core.json b/qapi/block-core.json
-index 97d1f64636..8bdbccb397 100644
+index 04ad80bc1e..9db3120716 100644
 --- a/qapi/block-core.json
 +++ b/qapi/block-core.json
-@@ -2054,10 +2054,19 @@
+@@ -1971,10 +1971,19 @@
  #        (all the disk, only the sectors allocated in the topmost image, or
  #        only new I/O).
  #
@@ -380,7 +380,7 @@ index 97d1f64636..8bdbccb397 100644
  #
  # @buf-size: maximum amount of data in flight from source to
  #            target (since 1.4).
-@@ -2095,7 +2104,9 @@
+@@ -2012,7 +2021,9 @@
  { 'struct': 'DriveMirror',
    'data': { '*job-id': 'str', 'device': 'str', 'target': 'str',
              '*format': 'str', '*node-name': 'str', '*replaces': 'str',
@@ -391,7 +391,7 @@ index 97d1f64636..8bdbccb397 100644
              '*speed': 'int', '*granularity': 'uint32',
              '*buf-size': 'int', '*on-source-error': 'BlockdevOnError',
              '*on-target-error': 'BlockdevOnError',
-@@ -2362,10 +2373,19 @@
+@@ -2280,10 +2291,19 @@
  #        (all the disk, only the sectors allocated in the topmost image, or
  #        only new I/O).
  #
@@ -412,7 +412,7 @@ index 97d1f64636..8bdbccb397 100644
  #
  # @buf-size: maximum amount of data in flight from source to
  #            target
-@@ -2414,7 +2434,8 @@
+@@ -2332,7 +2352,8 @@
  { 'command': 'blockdev-mirror',
    'data': { '*job-id': 'str', 'device': 'str', 'target': 'str',
              '*replaces': 'str',
@@ -423,10 +423,10 @@ index 97d1f64636..8bdbccb397 100644
              '*buf-size': 'int', '*on-source-error': 'BlockdevOnError',
              '*on-target-error': 'BlockdevOnError',
 diff --git a/tests/test-block-iothread.c b/tests/test-block-iothread.c
-index 0c861809f0..da87a67a57 100644
+index 3f866a35c6..500ede71c8 100644
 --- a/tests/test-block-iothread.c
 +++ b/tests/test-block-iothread.c
-@@ -611,8 +611,8 @@ static void test_propagate_mirror(void)
+@@ -623,8 +623,8 @@ static void test_propagate_mirror(void)
  
      /* Start a mirror job */
      mirror_start("job0", src, target, NULL, JOB_DEFAULT, 0, 0, 0,

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

@@ -23,10 +23,10 @@ Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
  1 file changed, 18 insertions(+), 6 deletions(-)
 
 diff --git a/block/mirror.c b/block/mirror.c
-index 3c9cd42c50..08ac9827f2 100644
+index 97843992c2..d1cce079da 100644
 --- a/block/mirror.c
 +++ b/block/mirror.c
-@@ -653,8 +653,6 @@ static int mirror_exit_common(Job *job)
+@@ -654,8 +654,6 @@ static int mirror_exit_common(Job *job)
          bdrv_unfreeze_backing_chain(mirror_top_bs, target_bs);
      }
  
@@ -35,7 +35,7 @@ index 3c9cd42c50..08ac9827f2 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);
-@@ -752,6 +750,18 @@ static int mirror_exit_common(Job *job)
+@@ -755,6 +753,18 @@ static int mirror_exit_common(Job *job)
      blk_set_perm(bjob->blk, 0, BLK_PERM_ALL, &error_abort);
      blk_insert_bs(bjob->blk, mirror_top_bs, &error_abort);
  
@@ -54,7 +54,7 @@ index 3c9cd42c50..08ac9827f2 100644
      bs_opaque->job = NULL;
  
      bdrv_drained_end(src);
-@@ -1584,10 +1594,6 @@ static BlockJob *mirror_start_job(
+@@ -1592,10 +1602,6 @@ static BlockJob *mirror_start_job(
                         " sync mode",
                         MirrorSyncMode_str(sync_mode));
              return NULL;
@@ -65,7 +65,7 @@ index 3c9cd42c50..08ac9827f2 100644
          }
      } else if (bitmap) {
          error_setg(errp,
-@@ -1604,6 +1610,12 @@ static BlockJob *mirror_start_job(
+@@ -1612,6 +1618,12 @@ static BlockJob *mirror_start_job(
              return NULL;
          }
          granularity = bdrv_dirty_bitmap_granularity(bitmap);

debian/patches/bitmap-mirror/0003-mirror-add-check-for-bitmap-mode-without-bitmap.patch

@@ -15,10 +15,10 @@ Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
  1 file changed, 3 insertions(+)
 
 diff --git a/blockdev.c b/blockdev.c
-index bbeff9c439..fa3c2f5548 100644
+index 394920613d..4f8bd38b58 100644
 --- a/blockdev.c
 +++ b/blockdev.c
-@@ -3257,6 +3257,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
+@@ -3029,6 +3029,9 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
          if (bdrv_dirty_bitmap_check(bitmap, BDRV_BITMAP_ALLOW_RO, errp)) {
              return;
          }
@@ -27,4 +27,4 @@ index bbeff9c439..fa3c2f5548 100644
 +        return;
      }
  
-     if (has_replaces) {
+     if (!has_replaces) {

debian/patches/bitmap-mirror/0004-mirror-switch-to-bdrv_dirty_bitmap_merge_internal.patch

@@ -15,10 +15,10 @@ Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
  1 file changed, 4 insertions(+), 4 deletions(-)
 
 diff --git a/block/mirror.c b/block/mirror.c
-index 08ac9827f2..c56b0f87e3 100644
+index d1cce079da..e6140cf018 100644
 --- a/block/mirror.c
 +++ b/block/mirror.c
-@@ -756,8 +756,8 @@ static int mirror_exit_common(Job *job)
+@@ -759,8 +759,8 @@ 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);
@@ -29,7 +29,7 @@ index 08ac9827f2..c56b0f87e3 100644
          }
      }
      bdrv_release_dirty_bitmap(s->dirty_bitmap);
-@@ -1749,8 +1749,8 @@ static BlockJob *mirror_start_job(
+@@ -1793,8 +1793,8 @@ static BlockJob *mirror_start_job(
      }
  
      if (s->sync_mode == MIRROR_SYNC_MODE_BITMAP) {

debian/patches/bitmap-mirror/0005-iotests-add-test-for-bitmap-mirror.patch

@@ -3434,7 +3434,7 @@ index 0000000000..9b7408b6d6
 +{"error": {"class": "GenericError", "desc": "bitmap-mode must be specified if a bitmap is provided"}}
 +
 diff --git a/tests/qemu-iotests/group b/tests/qemu-iotests/group
-index 435dccd5af..939efd9c70 100644
+index 2960dff728..952dceba1f 100644
 --- a/tests/qemu-iotests/group
 +++ b/tests/qemu-iotests/group
 @@ -270,6 +270,7 @@

debian/patches/bitmap-mirror/0006-mirror-move-some-checks-to-qmp.patch

@@ -18,10 +18,10 @@ Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
  3 files changed, 70 insertions(+), 59 deletions(-)
 
 diff --git a/block/mirror.c b/block/mirror.c
-index c56b0f87e3..dbba6fc80e 100644
+index e6140cf018..3a08239a78 100644
 --- a/block/mirror.c
 +++ b/block/mirror.c
-@@ -1584,31 +1584,13 @@ static BlockJob *mirror_start_job(
+@@ -1592,31 +1592,13 @@ static BlockJob *mirror_start_job(
      Error *local_err = NULL;
      int ret;
  
@@ -59,10 +59,10 @@ index c56b0f87e3..dbba6fc80e 100644
  
          if (bitmap_mode != BITMAP_SYNC_MODE_NEVER) {
 diff --git a/blockdev.c b/blockdev.c
-index fa3c2f5548..206de2b6c2 100644
+index 4f8bd38b58..a40c6fd0f6 100644
 --- a/blockdev.c
 +++ b/blockdev.c
-@@ -3236,7 +3236,36 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
+@@ -3008,7 +3008,36 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
          sync = MIRROR_SYNC_MODE_FULL;
      }
  

debian/patches/extra/0001-Revert-qemu-img-convert-Don-t-pre-zero-images.patch

@@ -0,0 +1,33 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Thomas Lamprecht <t.lamprecht@proxmox.com>
+Date: Mon, 14 Sep 2020 19:32:21 +0200
+Subject: [PATCH] Revert "qemu-img convert: Don't pre-zero images"
+
+This reverts commit edafc70c0c8510862f2f213a3acf7067113bcd08.
+
+As it correlates with causing issues on LVM allocation
+https://bugzilla.proxmox.com/show_bug.cgi?id=3002
+---
+ qemu-img.c | 9 +++++++++
+ 1 file changed, 9 insertions(+)
+
+diff --git a/qemu-img.c b/qemu-img.c
+index 8bdea40b58..f9050bfaad 100644
+--- a/qemu-img.c
++++ b/qemu-img.c
+@@ -2104,6 +2104,15 @@ static int convert_do_copy(ImgConvertState *s)
+         s->has_zero_init = bdrv_has_zero_init(blk_bs(s->target));
+     }
+ 
++    if (!s->has_zero_init && !s->target_has_backing &&
++        bdrv_can_write_zeroes_with_unmap(blk_bs(s->target)))
++    {
++        ret = blk_make_zero(s->target, BDRV_REQ_MAY_UNMAP | BDRV_REQ_NO_FALLBACK);
++        if (ret == 0) {
++            s->has_zero_init = true;
++        }
++    }
++
+     /* Allocate buffer for copied data. For compressed images, only one cluster
+      * can be copied at a time. */
+     if (s->compressed) {

debian/patches/extra/0002-docs-don-t-install-man-page-if-guest-agent-is-disabl.patch

@@ -0,0 +1,38 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Thu, 28 Jan 2021 15:19:51 +0100
+Subject: [PATCH] docs: don't install man page if guest agent is disabled
+
+No sense outputting the qemu-ga and qemu-ga-ref man pages when the guest
+agent binary itself is disabled. This mirrors behaviour from before the
+meson switch.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ docs/meson.build | 6 ++++--
+ 1 file changed, 4 insertions(+), 2 deletions(-)
+
+diff --git a/docs/meson.build b/docs/meson.build
+index ebd85d59f9..cc6f5007f8 100644
+--- a/docs/meson.build
++++ b/docs/meson.build
+@@ -46,6 +46,8 @@ if build_docs
+                           meson.source_root() / 'docs/sphinx/qmp_lexer.py',
+                           qapi_gen_depends ]
+ 
++  have_ga = have_tools and config_host.has_key('CONFIG_GUEST_AGENT')
++
+   configure_file(output: 'index.html',
+                  input: files('index.html.in'),
+                  configuration: {'VERSION': meson.project_version()},
+@@ -53,8 +55,8 @@ if build_docs
+   manuals = [ 'devel', 'interop', 'tools', 'specs', 'system', 'user' ]
+   man_pages = {
+     'interop' : {
+-        'qemu-ga.8': (have_tools ? 'man8' : ''),
+-        'qemu-ga-ref.7': 'man7',
++        'qemu-ga.8': (have_ga ? 'man8' : ''),
++        'qemu-ga-ref.7': (have_ga ? 'man7' : ''),
+         'qemu-qmp-ref.7': 'man7',
+     },
+     'tools': {

debian/patches/extra/0003-migration-only-check-page-size-match-if-RAM-postcopy.patch

@@ -0,0 +1,31 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Thu, 4 Feb 2021 17:06:19 +0100
+Subject: [PATCH] migration: only check page size match if RAM postcopy is
+ enabled
+
+Postcopy may also be advised for dirty-bitmap migration only, in which
+case the remote page size will not be available and we'll instead read
+bogus data, blocking migration with a mismatch error if the VM uses
+hugepages.
+
+Fixes: 58110f0acb ("migration: split common postcopy out of ram postcopy")
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+---
+ migration/ram.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/migration/ram.c b/migration/ram.c
+index 7811cde643..6ace15261c 100644
+--- a/migration/ram.c
++++ b/migration/ram.c
+@@ -3521,7 +3521,7 @@ static int ram_load_precopy(QEMUFile *f)
+                         }
+                     }
+                     /* For postcopy we need to check hugepage sizes match */
+-                    if (postcopy_advised &&
++                    if (postcopy_advised && migrate_postcopy_ram() &&
+                         block->page_size != qemu_host_page_size) {
+                         uint64_t remote_page_size = qemu_get_be64(f);
+                         if (remote_page_size != block->page_size) {

debian/patches/extra/0004-virtiofsd-extract-lo_do_open-from-lo_open.patch

@@ -0,0 +1,143 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Hajnoczi <stefanha@redhat.com>
+Date: Thu, 4 Feb 2021 18:34:35 +0000
+Subject: [PATCH] virtiofsd: extract lo_do_open() from lo_open()
+
+Both lo_open() and lo_create() have similar code to open a file. Extract
+a common lo_do_open() function from lo_open() that will be used by
+lo_create() in a later commit.
+
+Since lo_do_open() does not otherwise need fuse_req_t req, convert
+lo_add_fd_mapping() to use struct lo_data *lo instead.
+
+Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20210204150208.367837-2-stefanha@redhat.com>
+Reviewed-by: Greg Kurz <groug@kaod.org>
+Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ tools/virtiofsd/passthrough_ll.c | 73 ++++++++++++++++++++------------
+ 1 file changed, 46 insertions(+), 27 deletions(-)
+
+diff --git a/tools/virtiofsd/passthrough_ll.c b/tools/virtiofsd/passthrough_ll.c
+index 97485b22b4..218e20e9d7 100644
+--- a/tools/virtiofsd/passthrough_ll.c
++++ b/tools/virtiofsd/passthrough_ll.c
+@@ -471,17 +471,17 @@ static void lo_map_remove(struct lo_map *map, size_t key)
+ }
+ 
+ /* Assumes lo->mutex is held */
+-static ssize_t lo_add_fd_mapping(fuse_req_t req, int fd)
++static ssize_t lo_add_fd_mapping(struct lo_data *lo, int fd)
+ {
+     struct lo_map_elem *elem;
+ 
+-    elem = lo_map_alloc_elem(&lo_data(req)->fd_map);
++    elem = lo_map_alloc_elem(&lo->fd_map);
+     if (!elem) {
+         return -1;
+     }
+ 
+     elem->fd = fd;
+-    return elem - lo_data(req)->fd_map.elems;
++    return elem - lo->fd_map.elems;
+ }
+ 
+ /* Assumes lo->mutex is held */
+@@ -1661,6 +1661,38 @@ static void update_open_flags(int writeback, int allow_direct_io,
+     }
+ }
+ 
++static int lo_do_open(struct lo_data *lo, struct lo_inode *inode,
++                      struct fuse_file_info *fi)
++{
++    char buf[64];
++    ssize_t fh;
++    int fd;
++
++    update_open_flags(lo->writeback, lo->allow_direct_io, fi);
++
++    sprintf(buf, "%i", inode->fd);
++    fd = openat(lo->proc_self_fd, buf, fi->flags & ~O_NOFOLLOW);
++    if (fd == -1) {
++        return errno;
++    }
++
++    pthread_mutex_lock(&lo->mutex);
++    fh = lo_add_fd_mapping(lo, fd);
++    pthread_mutex_unlock(&lo->mutex);
++    if (fh == -1) {
++        close(fd);
++        return ENOMEM;
++    }
++
++    fi->fh = fh;
++    if (lo->cache == CACHE_NONE) {
++        fi->direct_io = 1;
++    } else if (lo->cache == CACHE_ALWAYS) {
++        fi->keep_cache = 1;
++    }
++    return 0;
++}
++
+ static void lo_create(fuse_req_t req, fuse_ino_t parent, const char *name,
+                       mode_t mode, struct fuse_file_info *fi)
+ {
+@@ -1701,7 +1733,7 @@ static void lo_create(fuse_req_t req, fuse_ino_t parent, const char *name,
+         ssize_t fh;
+ 
+         pthread_mutex_lock(&lo->mutex);
+-        fh = lo_add_fd_mapping(req, fd);
++        fh = lo_add_fd_mapping(lo, fd);
+         pthread_mutex_unlock(&lo->mutex);
+         if (fh == -1) {
+             close(fd);
+@@ -1892,38 +1924,25 @@ static void lo_fsyncdir(fuse_req_t req, fuse_ino_t ino, int datasync,
+ 
+ static void lo_open(fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
+ {
+-    int fd;
+-    ssize_t fh;
+-    char buf[64];
+     struct lo_data *lo = lo_data(req);
++    struct lo_inode *inode = lo_inode(req, ino);
++    int err;
+ 
+     fuse_log(FUSE_LOG_DEBUG, "lo_open(ino=%" PRIu64 ", flags=%d)\n", ino,
+              fi->flags);
+ 
+-    update_open_flags(lo->writeback, lo->allow_direct_io, fi);
+-
+-    sprintf(buf, "%i", lo_fd(req, ino));
+-    fd = openat(lo->proc_self_fd, buf, fi->flags & ~O_NOFOLLOW);
+-    if (fd == -1) {
+-        return (void)fuse_reply_err(req, errno);
+-    }
+-
+-    pthread_mutex_lock(&lo->mutex);
+-    fh = lo_add_fd_mapping(req, fd);
+-    pthread_mutex_unlock(&lo->mutex);
+-    if (fh == -1) {
+-        close(fd);
+-        fuse_reply_err(req, ENOMEM);
++    if (!inode) {
++        fuse_reply_err(req, EBADF);
+         return;
+     }
+ 
+-    fi->fh = fh;
+-    if (lo->cache == CACHE_NONE) {
+-        fi->direct_io = 1;
+-    } else if (lo->cache == CACHE_ALWAYS) {
+-        fi->keep_cache = 1;
++    err = lo_do_open(lo, inode, fi);
++    lo_inode_put(lo, &inode);
++    if (err) {
++        fuse_reply_err(req, err);
++    } else {
++        fuse_reply_open(req, fi);
+     }
+-    fuse_reply_open(req, fi);
+ }
+ 
+ static void lo_release(fuse_req_t req, fuse_ino_t ino,

debian/patches/extra/0005-virtiofsd-optionally-return-inode-pointer-from-lo_do.patch

@@ -0,0 +1,107 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Hajnoczi <stefanha@redhat.com>
+Date: Thu, 4 Feb 2021 18:34:36 +0000
+Subject: [PATCH] virtiofsd: optionally return inode pointer from
+ lo_do_lookup()
+
+lo_do_lookup() finds an existing inode or allocates a new one. It
+increments nlookup so that the inode stays alive until the client
+releases it.
+
+Existing callers don't need the struct lo_inode so the function doesn't
+return it. Extend the function to optionally return the inode. The next
+commit will need it.
+
+Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Reviewed-by: Greg Kurz <groug@kaod.org>
+Message-Id: <20210204150208.367837-3-stefanha@redhat.com>
+Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ tools/virtiofsd/passthrough_ll.c | 29 +++++++++++++++++++++--------
+ 1 file changed, 21 insertions(+), 8 deletions(-)
+
+diff --git a/tools/virtiofsd/passthrough_ll.c b/tools/virtiofsd/passthrough_ll.c
+index 218e20e9d7..2bd050b620 100644
+--- a/tools/virtiofsd/passthrough_ll.c
++++ b/tools/virtiofsd/passthrough_ll.c
+@@ -843,11 +843,13 @@ static int do_statx(struct lo_data *lo, int dirfd, const char *pathname,
+ }
+ 
+ /*
+- * Increments nlookup and caller must release refcount using
+- * lo_inode_put(&parent).
++ * Increments nlookup on the inode on success. unref_inode_lolocked() must be
++ * called eventually to decrement nlookup again. If inodep is non-NULL, the
++ * inode pointer is stored and the caller must call lo_inode_put().
+  */
+ static int lo_do_lookup(fuse_req_t req, fuse_ino_t parent, const char *name,
+-                        struct fuse_entry_param *e)
++                        struct fuse_entry_param *e,
++                        struct lo_inode **inodep)
+ {
+     int newfd;
+     int res;
+@@ -857,6 +859,10 @@ static int lo_do_lookup(fuse_req_t req, fuse_ino_t parent, const char *name,
+     struct lo_inode *inode = NULL;
+     struct lo_inode *dir = lo_inode(req, parent);
+ 
++    if (inodep) {
++        *inodep = NULL;
++    }
++
+     /*
+      * name_to_handle_at() and open_by_handle_at() can reach here with fuse
+      * mount point in guest, but we don't have its inode info in the
+@@ -924,7 +930,14 @@ static int lo_do_lookup(fuse_req_t req, fuse_ino_t parent, const char *name,
+         pthread_mutex_unlock(&lo->mutex);
+     }
+     e->ino = inode->fuse_ino;
+-    lo_inode_put(lo, &inode);
++
++    /* Transfer ownership of inode pointer to caller or drop it */
++    if (inodep) {
++        *inodep = inode;
++    } else {
++        lo_inode_put(lo, &inode);
++    }
++
+     lo_inode_put(lo, &dir);
+ 
+     fuse_log(FUSE_LOG_DEBUG, "  %lli/%s -> %lli\n", (unsigned long long)parent,
+@@ -959,7 +972,7 @@ static void lo_lookup(fuse_req_t req, fuse_ino_t parent, const char *name)
+         return;
+     }
+ 
+-    err = lo_do_lookup(req, parent, name, &e);
++    err = lo_do_lookup(req, parent, name, &e, NULL);
+     if (err) {
+         fuse_reply_err(req, err);
+     } else {
+@@ -1067,7 +1080,7 @@ static void lo_mknod_symlink(fuse_req_t req, fuse_ino_t parent,
+         goto out;
+     }
+ 
+-    saverr = lo_do_lookup(req, parent, name, &e);
++    saverr = lo_do_lookup(req, parent, name, &e, NULL);
+     if (saverr) {
+         goto out;
+     }
+@@ -1544,7 +1557,7 @@ static void lo_do_readdir(fuse_req_t req, fuse_ino_t ino, size_t size,
+ 
+         if (plus) {
+             if (!is_dot_or_dotdot(name)) {
+-                err = lo_do_lookup(req, ino, name, &e);
++                err = lo_do_lookup(req, ino, name, &e, NULL);
+                 if (err) {
+                     goto error;
+                 }
+@@ -1742,7 +1755,7 @@ static void lo_create(fuse_req_t req, fuse_ino_t parent, const char *name,
+         }
+ 
+         fi->fh = fh;
+-        err = lo_do_lookup(req, parent, name, &e);
++        err = lo_do_lookup(req, parent, name, &e, NULL);
+     }
+     if (lo->cache == CACHE_NONE) {
+         fi->direct_io = 1;

debian/patches/extra/0006-virtiofsd-prevent-opening-of-special-files-CVE-2020-.patch

@@ -0,0 +1,296 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Hajnoczi <stefanha@redhat.com>
+Date: Thu, 4 Feb 2021 18:34:37 +0000
+Subject: [PATCH] virtiofsd: prevent opening of special files (CVE-2020-35517)
+
+A well-behaved FUSE client does not attempt to open special files with
+FUSE_OPEN because they are handled on the client side (e.g. device nodes
+are handled by client-side device drivers).
+
+The check to prevent virtiofsd from opening special files is missing in
+a few cases, most notably FUSE_OPEN. A malicious client can cause
+virtiofsd to open a device node, potentially allowing the guest to
+escape. This can be exploited by a modified guest device driver. It is
+not exploitable from guest userspace since the guest kernel will handle
+special files inside the guest instead of sending FUSE requests.
+
+This patch fixes this issue by introducing the lo_inode_open() function
+to check the file type before opening it. This is a short-term solution
+because it does not prevent a compromised virtiofsd process from opening
+device nodes on the host.
+
+Restructure lo_create() to try O_CREAT | O_EXCL first. Note that O_CREAT
+| O_EXCL does not follow symlinks, so O_NOFOLLOW masking is not
+necessary here. If the file exists and the user did not specify O_EXCL,
+open it via lo_do_open().
+
+Reported-by: Alex Xu <alex@alxu.ca>
+Fixes: CVE-2020-35517
+Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Reviewed-by: Vivek Goyal <vgoyal@redhat.com>
+Reviewed-by: Greg Kurz <groug@kaod.org>
+Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
+Message-Id: <20210204150208.367837-4-stefanha@redhat.com>
+Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ tools/virtiofsd/passthrough_ll.c | 144 ++++++++++++++++++++-----------
+ 1 file changed, 92 insertions(+), 52 deletions(-)
+
+diff --git a/tools/virtiofsd/passthrough_ll.c b/tools/virtiofsd/passthrough_ll.c
+index 2bd050b620..03c5e0d13c 100644
+--- a/tools/virtiofsd/passthrough_ll.c
++++ b/tools/virtiofsd/passthrough_ll.c
+@@ -567,6 +567,38 @@ static int lo_fd(fuse_req_t req, fuse_ino_t ino)
+     return fd;
+ }
+ 
++/*
++ * Open a file descriptor for an inode. Returns -EBADF if the inode is not a
++ * regular file or a directory.
++ *
++ * Use this helper function instead of raw openat(2) to prevent security issues
++ * when a malicious client opens special files such as block device nodes.
++ * Symlink inodes are also rejected since symlinks must already have been
++ * traversed on the client side.
++ */
++static int lo_inode_open(struct lo_data *lo, struct lo_inode *inode,
++                         int open_flags)
++{
++    g_autofree char *fd_str = g_strdup_printf("%d", inode->fd);
++    int fd;
++
++    if (!S_ISREG(inode->filetype) && !S_ISDIR(inode->filetype)) {
++        return -EBADF;
++    }
++
++    /*
++     * The file is a symlink so O_NOFOLLOW must be ignored. We checked earlier
++     * that the inode is not a special file but if an external process races
++     * with us then symlinks are traversed here. It is not possible to escape
++     * the shared directory since it is mounted as "/" though.
++     */
++    fd = openat(lo->proc_self_fd, fd_str, open_flags & ~O_NOFOLLOW);
++    if (fd < 0) {
++        return -errno;
++    }
++    return fd;
++}
++
+ static void lo_init(void *userdata, struct fuse_conn_info *conn)
+ {
+     struct lo_data *lo = (struct lo_data *)userdata;
+@@ -696,9 +728,9 @@ static void lo_setattr(fuse_req_t req, fuse_ino_t ino, struct stat *attr,
+         if (fi) {
+             truncfd = fd;
+         } else {
+-            sprintf(procname, "%i", ifd);
+-            truncfd = openat(lo->proc_self_fd, procname, O_RDWR);
++            truncfd = lo_inode_open(lo, inode, O_RDWR);
+             if (truncfd < 0) {
++                errno = -truncfd;
+                 goto out_err;
+             }
+         }
+@@ -860,7 +892,7 @@ static int lo_do_lookup(fuse_req_t req, fuse_ino_t parent, const char *name,
+     struct lo_inode *dir = lo_inode(req, parent);
+ 
+     if (inodep) {
+-        *inodep = NULL;
++        *inodep = NULL; /* in case there is an error */
+     }
+ 
+     /*
+@@ -1674,19 +1706,26 @@ static void update_open_flags(int writeback, int allow_direct_io,
+     }
+ }
+ 
++/*
++ * Open a regular file, set up an fd mapping, and fill out the struct
++ * fuse_file_info for it. If existing_fd is not negative, use that fd instead
++ * opening a new one. Takes ownership of existing_fd.
++ *
++ * Returns 0 on success or a positive errno.
++ */
+ static int lo_do_open(struct lo_data *lo, struct lo_inode *inode,
+-                      struct fuse_file_info *fi)
++                      int existing_fd, struct fuse_file_info *fi)
+ {
+-    char buf[64];
+     ssize_t fh;
+-    int fd;
++    int fd = existing_fd;
+ 
+     update_open_flags(lo->writeback, lo->allow_direct_io, fi);
+ 
+-    sprintf(buf, "%i", inode->fd);
+-    fd = openat(lo->proc_self_fd, buf, fi->flags & ~O_NOFOLLOW);
+-    if (fd == -1) {
+-        return errno;
++    if (fd < 0) {
++        fd = lo_inode_open(lo, inode, fi->flags);
++        if (fd < 0) {
++            return -fd;
++        }
+     }
+ 
+     pthread_mutex_lock(&lo->mutex);
+@@ -1709,9 +1748,10 @@ static int lo_do_open(struct lo_data *lo, struct lo_inode *inode,
+ static void lo_create(fuse_req_t req, fuse_ino_t parent, const char *name,
+                       mode_t mode, struct fuse_file_info *fi)
+ {
+-    int fd;
++    int fd = -1;
+     struct lo_data *lo = lo_data(req);
+     struct lo_inode *parent_inode;
++    struct lo_inode *inode = NULL;
+     struct fuse_entry_param e;
+     int err;
+     struct lo_cred old = {};
+@@ -1737,36 +1777,38 @@ static void lo_create(fuse_req_t req, fuse_ino_t parent, const char *name,
+ 
+     update_open_flags(lo->writeback, lo->allow_direct_io, fi);
+ 
+-    fd = openat(parent_inode->fd, name, (fi->flags | O_CREAT) & ~O_NOFOLLOW,
+-                mode);
++    /* Try to create a new file but don't open existing files */
++    fd = openat(parent_inode->fd, name, fi->flags | O_CREAT | O_EXCL, mode);
+     err = fd == -1 ? errno : 0;
+-    lo_restore_cred(&old);
+ 
+-    if (!err) {
+-        ssize_t fh;
++    lo_restore_cred(&old);
+ 
+-        pthread_mutex_lock(&lo->mutex);
+-        fh = lo_add_fd_mapping(lo, fd);
+-        pthread_mutex_unlock(&lo->mutex);
+-        if (fh == -1) {
+-            close(fd);
+-            err = ENOMEM;
+-            goto out;
+-        }
++    /* Ignore the error if file exists and O_EXCL was not given */
++    if (err && (err != EEXIST || (fi->flags & O_EXCL))) {
++        goto out;
++    }
+ 
+-        fi->fh = fh;
+-        err = lo_do_lookup(req, parent, name, &e, NULL);
++    err = lo_do_lookup(req, parent, name, &e, &inode);
++    if (err) {
++        goto out;
+     }
+-    if (lo->cache == CACHE_NONE) {
+-        fi->direct_io = 1;
+-    } else if (lo->cache == CACHE_ALWAYS) {
+-        fi->keep_cache = 1;
++
++    err = lo_do_open(lo, inode, fd, fi);
++    fd = -1; /* lo_do_open() takes ownership of fd */
++    if (err) {
++        /* Undo lo_do_lookup() nlookup ref */
++        unref_inode_lolocked(lo, inode, 1);
+     }
+ 
+ out:
++    lo_inode_put(lo, &inode);
+     lo_inode_put(lo, &parent_inode);
+ 
+     if (err) {
++        if (fd >= 0) {
++            close(fd);
++        }
++
+         fuse_reply_err(req, err);
+     } else {
+         fuse_reply_create(req, &e, fi);
+@@ -1780,7 +1822,6 @@ static struct lo_inode_plock *lookup_create_plock_ctx(struct lo_data *lo,
+                                                       pid_t pid, int *err)
+ {
+     struct lo_inode_plock *plock;
+-    char procname[64];
+     int fd;
+ 
+     plock =
+@@ -1797,12 +1838,10 @@ static struct lo_inode_plock *lookup_create_plock_ctx(struct lo_data *lo,
+     }
+ 
+     /* Open another instance of file which can be used for ofd locks. */
+-    sprintf(procname, "%i", inode->fd);
+-
+     /* TODO: What if file is not writable? */
+-    fd = openat(lo->proc_self_fd, procname, O_RDWR);
+-    if (fd == -1) {
+-        *err = errno;
++    fd = lo_inode_open(lo, inode, O_RDWR);
++    if (fd < 0) {
++        *err = -fd;
+         free(plock);
+         return NULL;
+     }
+@@ -1949,7 +1988,7 @@ static void lo_open(fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
+         return;
+     }
+ 
+-    err = lo_do_open(lo, inode, fi);
++    err = lo_do_open(lo, inode, -1, fi);
+     lo_inode_put(lo, &inode);
+     if (err) {
+         fuse_reply_err(req, err);
+@@ -2005,39 +2044,40 @@ static void lo_flush(fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
+ static void lo_fsync(fuse_req_t req, fuse_ino_t ino, int datasync,
+                      struct fuse_file_info *fi)
+ {
++    struct lo_inode *inode = lo_inode(req, ino);
++    struct lo_data *lo = lo_data(req);
+     int res;
+     int fd;
+-    char *buf;
+ 
+     fuse_log(FUSE_LOG_DEBUG, "lo_fsync(ino=%" PRIu64 ", fi=0x%p)\n", ino,
+              (void *)fi);
+ 
+-    if (!fi) {
+-        struct lo_data *lo = lo_data(req);
+-
+-        res = asprintf(&buf, "%i", lo_fd(req, ino));
+-        if (res == -1) {
+-            return (void)fuse_reply_err(req, errno);
+-        }
++    if (!inode) {
++        fuse_reply_err(req, EBADF);
++        return;
++    }
+ 
+-        fd = openat(lo->proc_self_fd, buf, O_RDWR);
+-        free(buf);
+-        if (fd == -1) {
+-            return (void)fuse_reply_err(req, errno);
++    if (!fi) {
++        fd = lo_inode_open(lo, inode, O_RDWR);
++        if (fd < 0) {
++            res = -fd;
++            goto out;
+         }
+     } else {
+         fd = lo_fi_fd(req, fi);
+     }
+ 
+     if (datasync) {
+-        res = fdatasync(fd);
++        res = fdatasync(fd) == -1 ? errno : 0;
+     } else {
+-        res = fsync(fd);
++        res = fsync(fd) == -1 ? errno : 0;
+     }
+     if (!fi) {
+         close(fd);
+     }
+-    fuse_reply_err(req, res == -1 ? errno : 0);
++out:
++    lo_inode_put(lo, &inode);
++    fuse_reply_err(req, res);
+ }
+ 
+ static void lo_read(fuse_req_t req, fuse_ino_t ino, size_t size, off_t offset,

debian/patches/extra/0007-virtiofsd-Add-_llseek-to-the-seccomp-whitelist.patch

@@ -0,0 +1,29 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Greg Kurz <groug@kaod.org>
+Date: Thu, 4 Feb 2021 18:34:38 +0000
+Subject: [PATCH] virtiofsd: Add _llseek to the seccomp whitelist
+
+This is how glibc implements lseek(2) on POWER.
+
+BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1917692
+Signed-off-by: Greg Kurz <groug@kaod.org>
+Message-Id: <20210121171540.1449777-1-groug@kaod.org>
+Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ tools/virtiofsd/passthrough_seccomp.c | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/tools/virtiofsd/passthrough_seccomp.c b/tools/virtiofsd/passthrough_seccomp.c
+index 11623f56f2..bb8ef5b17f 100644
+--- a/tools/virtiofsd/passthrough_seccomp.c
++++ b/tools/virtiofsd/passthrough_seccomp.c
+@@ -68,6 +68,7 @@ static const int syscall_whitelist[] = {
+     SCMP_SYS(linkat),
+     SCMP_SYS(listxattr),
+     SCMP_SYS(lseek),
++    SCMP_SYS(_llseek), /* For POWER */
+     SCMP_SYS(madvise),
+     SCMP_SYS(mkdirat),
+     SCMP_SYS(mknodat),

debian/patches/extra/0008-virtiofsd-Add-restart_syscall-to-the-seccomp-whiteli.patch

@@ -0,0 +1,31 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Greg Kurz <groug@kaod.org>
+Date: Thu, 4 Feb 2021 18:34:39 +0000
+Subject: [PATCH] virtiofsd: Add restart_syscall to the seccomp whitelist
+
+This is how linux restarts some system calls after SIGSTOP/SIGCONT.
+This is needed to avoid virtiofsd termination when resuming execution
+under GDB for example.
+
+Signed-off-by: Greg Kurz <groug@kaod.org>
+Message-Id: <20210201193305.136390-1-groug@kaod.org>
+Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
+Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ tools/virtiofsd/passthrough_seccomp.c | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/tools/virtiofsd/passthrough_seccomp.c b/tools/virtiofsd/passthrough_seccomp.c
+index bb8ef5b17f..44d75e0e36 100644
+--- a/tools/virtiofsd/passthrough_seccomp.c
++++ b/tools/virtiofsd/passthrough_seccomp.c
+@@ -92,6 +92,7 @@ static const int syscall_whitelist[] = {
+     SCMP_SYS(renameat),
+     SCMP_SYS(renameat2),
+     SCMP_SYS(removexattr),
++    SCMP_SYS(restart_syscall),
+     SCMP_SYS(rt_sigaction),
+     SCMP_SYS(rt_sigprocmask),
+     SCMP_SYS(rt_sigreturn),

debian/patches/extra/0009-i386-acpi-restore-device-paths-for-pre-5.1-vms.patch

@@ -0,0 +1,108 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Vitaly Cheptsov <cheptsov@ispras.ru>
+Date: Tue, 2 Mar 2021 09:21:10 -0500
+Subject: [PATCH] i386/acpi: restore device paths for pre-5.1 vms
+
+After fixing the _UID value for the primary PCI root bridge in
+af1b80ae it was discovered that this change updates Windows
+configuration in an incompatible way causing network configuration
+failure unless DHCP is used. More details provided on the list:
+
+https://lists.gnu.org/archive/html/qemu-devel/2021-02/msg08484.html
+
+This change reverts the _UID update from 1 to 0 for q35 and i440fx
+VMs before version 5.2 to maintain the original behaviour when
+upgrading.
+
+Cc: qemu-stable@nongnu.org
+Cc: qemu-devel@nongnu.org
+Reported-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+Suggested-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Vitaly Cheptsov <cheptsov@ispras.ru>
+Message-Id: <20210301195919.9333-1-cheptsov@ispras.ru>
+Tested-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+Reviewed-by: Igor Mammedov <imammedo@redhat.com>
+Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
+Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
+Fixes: af1b80ae56c9 ("i386/acpi: fix inconsistent QEMU/OVMF device paths")
+---
+ hw/i386/acpi-build.c | 4 ++--
+ hw/i386/pc_piix.c    | 2 ++
+ hw/i386/pc_q35.c     | 2 ++
+ include/hw/i386/pc.h | 1 +
+ 4 files changed, 7 insertions(+), 2 deletions(-)
+
+diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
+index 1f5c211245..b5616582a5 100644
+--- a/hw/i386/acpi-build.c
++++ b/hw/i386/acpi-build.c
+@@ -1513,7 +1513,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,
+         dev = aml_device("PCI0");
+         aml_append(dev, aml_name_decl("_HID", aml_eisaid("PNP0A03")));
+         aml_append(dev, aml_name_decl("_ADR", aml_int(0)));
+-        aml_append(dev, aml_name_decl("_UID", aml_int(0)));
++        aml_append(dev, aml_name_decl("_UID", aml_int(pcmc->pci_root_uid)));
+         aml_append(sb_scope, dev);
+         aml_append(dsdt, sb_scope);
+ 
+@@ -1530,7 +1530,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,
+         aml_append(dev, aml_name_decl("_HID", aml_eisaid("PNP0A08")));
+         aml_append(dev, aml_name_decl("_CID", aml_eisaid("PNP0A03")));
+         aml_append(dev, aml_name_decl("_ADR", aml_int(0)));
+-        aml_append(dev, aml_name_decl("_UID", aml_int(0)));
++        aml_append(dev, aml_name_decl("_UID", aml_int(pcmc->pci_root_uid)));
+         aml_append(dev, build_q35_osc_method());
+         aml_append(sb_scope, dev);
+ 
+diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
+index 13d1628f13..2524c96216 100644
+--- a/hw/i386/pc_piix.c
++++ b/hw/i386/pc_piix.c
+@@ -417,6 +417,7 @@ static void pc_i440fx_machine_options(MachineClass *m)
+ {
+     PCMachineClass *pcmc = PC_MACHINE_CLASS(m);
+     pcmc->default_nic_model = "e1000";
++    pcmc->pci_root_uid = 0;
+ 
+     m->family = "pc_piix";
+     m->desc = "Standard PC (i440FX + PIIX, 1996)";
+@@ -448,6 +449,7 @@ static void pc_i440fx_5_1_machine_options(MachineClass *m)
+     compat_props_add(m->compat_props, hw_compat_5_1, hw_compat_5_1_len);
+     compat_props_add(m->compat_props, pc_compat_5_1, pc_compat_5_1_len);
+     pcmc->kvmclock_create_always = false;
++    pcmc->pci_root_uid = 1;
+ }
+ 
+ DEFINE_I440FX_MACHINE(v5_1, "pc-i440fx-5.1", NULL,
+diff --git a/hw/i386/pc_q35.c b/hw/i386/pc_q35.c
+index a3f4959c43..c58dad5ae3 100644
+--- a/hw/i386/pc_q35.c
++++ b/hw/i386/pc_q35.c
+@@ -329,6 +329,7 @@ static void pc_q35_machine_options(MachineClass *m)
+ {
+     PCMachineClass *pcmc = PC_MACHINE_CLASS(m);
+     pcmc->default_nic_model = "e1000e";
++    pcmc->pci_root_uid = 0;
+ 
+     m->family = "pc_q35";
+     m->desc = "Standard PC (Q35 + ICH9, 2009)";
+@@ -364,6 +365,7 @@ static void pc_q35_5_1_machine_options(MachineClass *m)
+     compat_props_add(m->compat_props, hw_compat_5_1, hw_compat_5_1_len);
+     compat_props_add(m->compat_props, pc_compat_5_1, pc_compat_5_1_len);
+     pcmc->kvmclock_create_always = false;
++    pcmc->pci_root_uid = 1;
+ }
+ 
+ DEFINE_Q35_MACHINE(v5_1, "pc-q35-5.1", NULL,
+diff --git a/include/hw/i386/pc.h b/include/hw/i386/pc.h
+index 911e460097..7f8e1a791f 100644
+--- a/include/hw/i386/pc.h
++++ b/include/hw/i386/pc.h
+@@ -99,6 +99,7 @@ struct PCMachineClass {
+     int legacy_acpi_table_size;
+     unsigned acpi_data_size;
+     bool do_not_add_smb_acpi;
++    int pci_root_uid;
+ 
+     /* SMBIOS compat: */
+     bool smbios_defaults;

debian/patches/extra/0010-monitor-qmp-fix-race-on-CHR_EVENT_CLOSED-without-OOB.patch

@@ -0,0 +1,48 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Mon, 22 Mar 2021 15:20:04 +0100
+Subject: [PATCH] monitor/qmp: fix race on CHR_EVENT_CLOSED without OOB
+
+The QMP dispatcher coroutine holds the qmp_queue_lock over a yield
+point, where it expects to be rescheduled from the main context. If a
+CHR_EVENT_CLOSED event is received just then, it can race and block the
+main thread on the mutex in monitor_qmp_cleanup_queue_and_resume.
+
+Calculate need_resume immediately after we pop a request from the queue,
+so that we can release the mutex before yielding.
+
+Suggested-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ monitor/qmp.c | 11 ++++++-----
+ 1 file changed, 6 insertions(+), 5 deletions(-)
+
+diff --git a/monitor/qmp.c b/monitor/qmp.c
+index 2e37d11bd3..2aff833f7a 100644
+--- a/monitor/qmp.c
++++ b/monitor/qmp.c
+@@ -252,6 +252,12 @@ void coroutine_fn monitor_qmp_dispatcher_co(void *data)
+             }
+         }
+ 
++        mon = req_obj->mon;
++        /* qmp_oob_enabled() might change after "qmp_capabilities" */
++        need_resume = !qmp_oob_enabled(mon) ||
++            mon->qmp_requests->length == QMP_REQ_QUEUE_LEN_MAX - 1;
++        qemu_mutex_unlock(&mon->qmp_queue_lock);
++
+         if (qatomic_xchg(&qmp_dispatcher_co_busy, true) == true) {
+             /*
+              * Someone rescheduled us (probably because a new requests
+@@ -270,11 +276,6 @@ void coroutine_fn monitor_qmp_dispatcher_co(void *data)
+         aio_co_schedule(qemu_get_aio_context(), qmp_dispatcher_co);
+         qemu_coroutine_yield();
+ 
+-        mon = req_obj->mon;
+-        /* qmp_oob_enabled() might change after "qmp_capabilities" */
+-        need_resume = !qmp_oob_enabled(mon) ||
+-            mon->qmp_requests->length == QMP_REQ_QUEUE_LEN_MAX - 1;
+-        qemu_mutex_unlock(&mon->qmp_queue_lock);
+         if (req_obj->req) {
+             QDict *qdict = qobject_to(QDict, req_obj->req);
+             QObject *id = qdict ? qdict_get(qdict, "id") : NULL;

debian/patches/extra/0011-block-Fix-locking-in-qmp_block_resize.patch

@@ -0,0 +1,42 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Kevin Wolf <kwolf@redhat.com>
+Date: Thu, 3 Dec 2020 18:23:10 +0100
+Subject: [PATCH] block: Fix locking in qmp_block_resize()
+
+The drain functions assume that we hold the AioContext lock of the
+drained block node. Make sure to actually take the lock.
+
+Cc: qemu-stable@nongnu.org
+Fixes: eb94b81a94bce112e6b206df846c1551aaf6cab6
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Message-Id: <20201203172311.68232-3-kwolf@redhat.com>
+Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ blockdev.c | 5 ++++-
+ 1 file changed, 4 insertions(+), 1 deletion(-)
+
+diff --git a/blockdev.c b/blockdev.c
+index fe6fb5dc1d..9a86e9fb4b 100644
+--- a/blockdev.c
++++ b/blockdev.c
+@@ -2481,14 +2481,17 @@ void coroutine_fn qmp_block_resize(bool has_device, const char *device,
+         goto out;
+     }
+ 
++    bdrv_co_lock(bs);
+     bdrv_drained_begin(bs);
++    bdrv_co_unlock(bs);
++
+     old_ctx = bdrv_co_enter(bs);
+     blk_truncate(blk, size, false, PREALLOC_MODE_OFF, 0, errp);
+     bdrv_co_leave(bs, old_ctx);
+-    bdrv_drained_end(bs);
+ 
+ out:
+     bdrv_co_lock(bs);
++    bdrv_drained_end(bs);
+     blk_unref(blk);
+     bdrv_co_unlock(bs);
+ }

debian/patches/extra/0012-block-Fix-deadlock-in-bdrv_co_yield_to_drain.patch

@@ -0,0 +1,118 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Kevin Wolf <kwolf@redhat.com>
+Date: Thu, 3 Dec 2020 18:23:11 +0100
+Subject: [PATCH] block: Fix deadlock in bdrv_co_yield_to_drain()
+
+If bdrv_co_yield_to_drain() is called for draining a block node that
+runs in a different AioContext, it keeps that AioContext locked while it
+yields and schedules a BH in the AioContext to do the actual drain.
+
+As long as executing the BH is the very next thing that the event loop
+of the node's AioContext does, this actually happens to work, but when
+it tries to execute something else that wants to take the AioContext
+lock, it will deadlock. (In the bug report, this other thing is a
+virtio-scsi device running virtio_scsi_data_plane_handle_cmd().)
+
+Instead, always drop the AioContext lock across the yield and reacquire
+it only when the coroutine is reentered. The BH needs to unconditionally
+take the lock for itself now.
+
+This fixes the 'block_resize' QMP command on a block node that runs in
+an iothread.
+
+Cc: qemu-stable@nongnu.org
+Fixes: eb94b81a94bce112e6b206df846c1551aaf6cab6
+Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1903511
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Message-Id: <20201203172311.68232-4-kwolf@redhat.com>
+Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
+Signed-off-by: Kevin Wolf <kwolf@redhat.com>
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/io.c | 41 ++++++++++++++++++++++++-----------------
+ 1 file changed, 24 insertions(+), 17 deletions(-)
+
+diff --git a/block/io.c b/block/io.c
+index ec5e152bb7..a9f56a9ab1 100644
+--- a/block/io.c
++++ b/block/io.c
+@@ -306,17 +306,7 @@ static void bdrv_co_drain_bh_cb(void *opaque)
+ 
+     if (bs) {
+         AioContext *ctx = bdrv_get_aio_context(bs);
+-        AioContext *co_ctx = qemu_coroutine_get_aio_context(co);
+-
+-        /*
+-         * When the coroutine yielded, the lock for its home context was
+-         * released, so we need to re-acquire it here. If it explicitly
+-         * acquired a different context, the lock is still held and we don't
+-         * want to lock it a second time (or AIO_WAIT_WHILE() would hang).
+-         */
+-        if (ctx == co_ctx) {
+-            aio_context_acquire(ctx);
+-        }
++        aio_context_acquire(ctx);
+         bdrv_dec_in_flight(bs);
+         if (data->begin) {
+             assert(!data->drained_end_counter);
+@@ -328,9 +318,7 @@ static void bdrv_co_drain_bh_cb(void *opaque)
+                                 data->ignore_bds_parents,
+                                 data->drained_end_counter);
+         }
+-        if (ctx == co_ctx) {
+-            aio_context_release(ctx);
+-        }
++        aio_context_release(ctx);
+     } else {
+         assert(data->begin);
+         bdrv_drain_all_begin();
+@@ -348,13 +336,16 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
+                                                 int *drained_end_counter)
+ {
+     BdrvCoDrainData data;
++    Coroutine *self = qemu_coroutine_self();
++    AioContext *ctx = bdrv_get_aio_context(bs);
++    AioContext *co_ctx = qemu_coroutine_get_aio_context(self);
+ 
+     /* Calling bdrv_drain() from a BH ensures the current coroutine yields and
+      * other coroutines run if they were queued by aio_co_enter(). */
+ 
+     assert(qemu_in_coroutine());
+     data = (BdrvCoDrainData) {
+-        .co = qemu_coroutine_self(),
++        .co = self,
+         .bs = bs,
+         .done = false,
+         .begin = begin,
+@@ -368,13 +359,29 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
+     if (bs) {
+         bdrv_inc_in_flight(bs);
+     }
+-    replay_bh_schedule_oneshot_event(bdrv_get_aio_context(bs),
+-                                     bdrv_co_drain_bh_cb, &data);
++
++    /*
++     * Temporarily drop the lock across yield or we would get deadlocks.
++     * bdrv_co_drain_bh_cb() reaquires the lock as needed.
++     *
++     * When we yield below, the lock for the current context will be
++     * released, so if this is actually the lock that protects bs, don't drop
++     * it a second time.
++     */
++    if (ctx != co_ctx) {
++        aio_context_release(ctx);
++    }
++    replay_bh_schedule_oneshot_event(ctx, bdrv_co_drain_bh_cb, &data);
+ 
+     qemu_coroutine_yield();
+     /* If we are resumed from some other event (such as an aio completion or a
+      * timer callback), it is a bug in the caller that should be fixed. */
+     assert(data.done);
++
++    /* Reaquire the AioContext of bs if we dropped it */
++    if (ctx != co_ctx) {
++        aio_context_acquire(ctx);
++    }
+ }
+ 
+ void bdrv_do_drained_begin_quiesce(BlockDriverState *bs,

debian/patches/pve-qemu-5.2-vitastor.patch

@@ -0,0 +1,983 @@
+Index: pve-qemu-kvm-5.2.0/qapi/block-core.json
+===================================================================
+--- pve-qemu-kvm-5.2.0.orig/qapi/block-core.json
++++ pve-qemu-kvm-5.2.0/qapi/block-core.json
+@@ -3076,7 +3076,7 @@
+             'luks', 'nbd', 'nfs', 'null-aio', 'null-co', 'nvme', 'parallels',
+             'qcow', 'qcow2', 'qed', 'quorum', 'raw', 'rbd',
+             { 'name': 'replication', 'if': 'defined(CONFIG_REPLICATION)' },
+-            'sheepdog', 'pbs',
++            'sheepdog', 'pbs', 'vitastor',
+             'ssh', 'throttle', 'vdi', 'vhdx', 'vmdk', 'vpc', 'vvfat' ] }
+ 
+ ##
+@@ -3924,6 +3924,28 @@
+             '*tag': 'str' } }
+ 
+ ##
++# @BlockdevOptionsVitastor:
++#
++# Driver specific block device options for vitastor
++#
++# @image:       Image name
++# @inode:       Inode number
++# @pool:        Pool ID
++# @size:        Desired image size in bytes
++# @config-path: Path to Vitastor configuration
++# @etcd-host:   etcd connection address(es)
++# @etcd-prefix: etcd key/value prefix
++##
++{ 'struct': 'BlockdevOptionsVitastor',
++  'data': { '*inode': 'uint64',
++            '*pool': 'uint64',
++            '*size': 'uint64',
++            '*image': 'str',
++            '*config-path': 'str',
++            '*etcd-host': 'str',
++            '*etcd-prefix': 'str' } }
++
++##
+ # @ReplicationMode:
+ #
+ # An enumeration of replication modes.
+@@ -4272,6 +4294,7 @@
+       'replication': { 'type': 'BlockdevOptionsReplication',
+                        'if': 'defined(CONFIG_REPLICATION)' },
+       'sheepdog':   'BlockdevOptionsSheepdog',
++      'vitastor':   'BlockdevOptionsVitastor',
+       'ssh':        'BlockdevOptionsSsh',
+       'throttle':   'BlockdevOptionsThrottle',
+       'vdi':        'BlockdevOptionsGenericFormat',
+@@ -4662,6 +4685,17 @@
+             '*cluster-size' :   'size' } }
+ 
+ ##
++# @BlockdevCreateOptionsVitastor:
++#
++# Driver specific image creation options for Vitastor.
++#
++# @size: Size of the virtual disk in bytes
++##
++{ 'struct': 'BlockdevCreateOptionsVitastor',
++  'data': { 'location':         'BlockdevOptionsVitastor',
++            'size':             'size' } }
++
++##
+ # @BlockdevVmdkSubformat:
+ #
+ # Subformat options for VMDK images
+@@ -4923,6 +4957,7 @@
+       'qed':            'BlockdevCreateOptionsQed',
+       'rbd':            'BlockdevCreateOptionsRbd',
+       'sheepdog':       'BlockdevCreateOptionsSheepdog',
++      'vitastor':       'BlockdevCreateOptionsVitastor',
+       'ssh':            'BlockdevCreateOptionsSsh',
+       'vdi':            'BlockdevCreateOptionsVdi',
+       'vhdx':           'BlockdevCreateOptionsVhdx',
+Index: pve-qemu-kvm-5.2.0/block/meson.build
+===================================================================
+--- pve-qemu-kvm-5.2.0.orig/block/meson.build
++++ pve-qemu-kvm-5.2.0/block/meson.build
+@@ -89,6 +89,7 @@ foreach m : [
+   ['CONFIG_LIBNFS', 'nfs', libnfs, 'nfs.c'],
+   ['CONFIG_LIBSSH', 'ssh', libssh, 'ssh.c'],
+   ['CONFIG_RBD', 'rbd', rbd, 'rbd.c'],
++  ['CONFIG_VITASTOR', 'vitastor', vitastor, 'vitastor.c'],
+ ]
+   if config_host.has_key(m[0])
+     if enable_modules
+Index: pve-qemu-kvm-5.2.0/configure
+===================================================================
+--- pve-qemu-kvm-5.2.0.orig/configure
++++ pve-qemu-kvm-5.2.0/configure
+@@ -372,6 +372,7 @@ trace_backends="log"
+ trace_file="trace"
+ spice=""
+ rbd=""
++vitastor=""
+ smartcard=""
+ u2f="auto"
+ libusb=""
+@@ -1264,6 +1265,10 @@ for opt do
+   ;;
+   --enable-rbd) rbd="yes"
+   ;;
++  --disable-vitastor) vitastor="no"
++  ;;
++  --enable-vitastor) vitastor="yes"
++  ;;
+   --disable-xfsctl) xfs="no"
+   ;;
+   --enable-xfsctl) xfs="yes"
+@@ -1807,6 +1812,7 @@ disabled with --disable-FEATURE, default
+   vhost-vdpa      vhost-vdpa kernel backend support
+   spice           spice
+   rbd             rados block device (rbd)
++  vitastor        vitastor block device
+   libiscsi        iscsi support
+   libnfs          nfs support
+   smartcard       smartcard support (libcacard)
+@@ -3700,6 +3706,27 @@ EOF
+ fi
+ 
+ ##########################################
++# vitastor probe
++if test "$vitastor" != "no" ; then
++  cat > $TMPC <<EOF
++#include <vitastor_c.h>
++int main(void) {
++  vitastor_c_create_qemu(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0);
++  return 0;
++}
++EOF
++  vitastor_libs="-lvitastor_client"
++  if compile_prog "" "$vitastor_libs" ; then
++    vitastor=yes
++  else
++    if test "$vitastor" = "yes" ; then
++      feature_not_found "vitastor block device" "Install vitastor-client-dev"
++    fi
++    vitastor=no
++  fi
++fi
++
++##########################################
+ # libssh probe
+ if test "$libssh" != "no" ; then
+   if $pkg_config --exists libssh; then
+@@ -6437,6 +6464,10 @@ if test "$rbd" = "yes" ; then
+   echo "CONFIG_RBD=y" >> $config_host_mak
+   echo "RBD_LIBS=$rbd_libs" >> $config_host_mak
+ fi
++if test "$vitastor" = "yes" ; then
++  echo "CONFIG_VITASTOR=y" >> $config_host_mak
++  echo "VITASTOR_LIBS=$vitastor_libs" >> $config_host_mak
++fi
+ 
+ echo "CONFIG_COROUTINE_BACKEND=$coroutine" >> $config_host_mak
+ if test "$coroutine_pool" = "yes" ; then
+Index: pve-qemu-kvm-5.2.0/meson.build
+===================================================================
+--- pve-qemu-kvm-5.2.0.orig/meson.build
++++ pve-qemu-kvm-5.2.0/meson.build
+@@ -596,6 +596,10 @@ rbd = not_found
+ if 'CONFIG_RBD' in config_host
+   rbd = declare_dependency(link_args: config_host['RBD_LIBS'].split())
+ endif
++vitastor = not_found
++if 'CONFIG_VITASTOR' in config_host
++  vitastor = declare_dependency(link_args: config_host['VITASTOR_LIBS'].split())
++endif
+ glusterfs = not_found
+ if 'CONFIG_GLUSTERFS' in config_host
+   glusterfs = declare_dependency(compile_args: config_host['GLUSTERFS_CFLAGS'].split(),
+@@ -2151,6 +2155,7 @@ endif
+ # TODO: add back protocol and server version
+ summary_info += {'spice support':     config_host.has_key('CONFIG_SPICE')}
+ summary_info += {'rbd support':       config_host.has_key('CONFIG_RBD')}
++summary_info += {'vitastor support':  config_host.has_key('CONFIG_VITASTOR')}
+ summary_info += {'xfsctl support':    config_host.has_key('CONFIG_XFS')}
+ summary_info += {'smartcard support': config_host.has_key('CONFIG_SMARTCARD')}
+ summary_info += {'U2F support':       u2f.found()}
+Index: a/block/vitastor.c
+===================================================================
+--- /dev/null
++++ a/block/vitastor.c
+@@ -0,0 +1,797 @@
++// Copyright (c) Vitaliy Filippov, 2019+
++// License: VNPL-1.1 or GNU GPL-2.0+ (see README.md for details)
++
++// QEMU block driver
++
++#ifdef VITASTOR_SOURCE_TREE
++#define BUILD_DSO
++#define _GNU_SOURCE
++#endif
++#include "qemu/osdep.h"
++#include "qemu/main-loop.h"
++#include "block/block_int.h"
++#include "qapi/error.h"
++#include "qapi/qmp/qdict.h"
++#include "qapi/qmp/qerror.h"
++#include "qemu/uri.h"
++#include "qemu/error-report.h"
++#include "qemu/module.h"
++#include "qemu/option.h"
++
++#if QEMU_VERSION_MAJOR >= 3
++#include "qemu/units.h"
++#include "block/qdict.h"
++#include "qemu/cutils.h"
++#elif QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 10
++#include "qemu/cutils.h"
++#include "qapi/qmp/qstring.h"
++#include "qapi/qmp/qjson.h"
++#else
++#include "qapi/qmp/qint.h"
++#define qdict_put_int(options, name, num_val) qdict_put_obj(options, name, QOBJECT(qint_from_int(num_val)))
++#define qdict_put_str(options, name, value) qdict_put_obj(options, name, QOBJECT(qstring_from_str(value)))
++#define qobject_unref QDECREF
++#endif
++
++#include "vitastor_c.h"
++
++#ifdef VITASTOR_SOURCE_TREE
++void qemu_module_dummy(void)
++{
++}
++
++void DSO_STAMP_FUN(void)
++{
++}
++#endif
++
++typedef struct VitastorClient
++{
++    void *proxy;
++    void *watch;
++    char *config_path;
++    char *etcd_host;
++    char *etcd_prefix;
++    char *image;
++    int skip_parents;
++    uint64_t inode;
++    uint64_t pool;
++    uint64_t size;
++    long readonly;
++    int use_rdma;
++    char *rdma_device;
++    int rdma_port_num;
++    int rdma_gid_index;
++    int rdma_mtu;
++    QemuMutex mutex;
++
++    uint64_t last_bitmap_inode, last_bitmap_offset, last_bitmap_len;
++    uint32_t last_bitmap_granularity;
++    uint8_t *last_bitmap;
++} VitastorClient;
++
++typedef struct VitastorRPC
++{
++    BlockDriverState *bs;
++    Coroutine *co;
++    QEMUIOVector *iov;
++    long ret;
++    int complete;
++    uint64_t inode, offset, len;
++    uint32_t bitmap_granularity;
++    uint8_t *bitmap;
++} VitastorRPC;
++
++static void vitastor_co_init_task(BlockDriverState *bs, VitastorRPC *task);
++static void vitastor_co_generic_bh_cb(void *opaque, long retval);
++static void vitastor_co_read_cb(void *opaque, long retval, uint64_t version);
++static void vitastor_close(BlockDriverState *bs);
++
++static char *qemu_vitastor_next_tok(char *src, char delim, char **p)
++{
++    char *end;
++    *p = NULL;
++    for (end = src; *end; ++end)
++    {
++        if (*end == delim)
++            break;
++        if (*end == '\\' && end[1] != '\0')
++            end++;
++    }
++    if (*end == delim)
++    {
++        *p = end + 1;
++        *end = '\0';
++    }
++    return src;
++}
++
++static void qemu_vitastor_unescape(char *src)
++{
++    char *p;
++    for (p = src; *src; ++src, ++p)
++    {
++        if (*src == '\\' && src[1] != '\0')
++            src++;
++        *p = *src;
++    }
++    *p = '\0';
++}
++
++// vitastor[:key=value]*
++// vitastor[:etcd_host=127.0.0.1]:inode=1:pool=1[:rdma_gid_index=3]
++// vitastor:config_path=/etc/vitastor/vitastor.conf:image=testimg
++static void vitastor_parse_filename(const char *filename, QDict *options, Error **errp)
++{
++    const char *start;
++    char *p, *buf;
++
++    if (!strstart(filename, "vitastor:", &start))
++    {
++        error_setg(errp, "File name must start with 'vitastor:'");
++        return;
++    }
++
++    buf = g_strdup(start);
++    p = buf;
++
++    // The following are all key/value pairs
++    while (p)
++    {
++        int i;
++        char *name, *value;
++        name = qemu_vitastor_next_tok(p, '=', &p);
++        if (!p)
++        {
++            error_setg(errp, "conf option %s has no value", name);
++            break;
++        }
++        for (i = 0; i < strlen(name); i++)
++            if (name[i] == '_')
++                name[i] = '-';
++        qemu_vitastor_unescape(name);
++        value = qemu_vitastor_next_tok(p, ':', &p);
++        qemu_vitastor_unescape(value);
++        if (!strcmp(name, "inode") ||
++            !strcmp(name, "pool") ||
++            !strcmp(name, "size") ||
++            !strcmp(name, "skip-parents") ||
++            !strcmp(name, "use-rdma") ||
++            !strcmp(name, "rdma-port_num") ||
++            !strcmp(name, "rdma-gid-index") ||
++            !strcmp(name, "rdma-mtu"))
++        {
++            unsigned long long num_val;
++            if (parse_uint_full(value, &num_val, 0))
++            {
++                error_setg(errp, "Illegal %s: %s", name, value);
++                goto out;
++            }
++            qdict_put_int(options, name, num_val);
++        }
++        else
++        {
++            qdict_put_str(options, name, value);
++        }
++    }
++    if (!qdict_get_try_str(options, "image"))
++    {
++        if (!qdict_get_try_int(options, "inode", 0))
++        {
++            error_setg(errp, "one of image (name) and inode (number) must be specified");
++            goto out;
++        }
++        if (!(qdict_get_try_int(options, "inode", 0) >> (64-POOL_ID_BITS)) &&
++            !qdict_get_try_int(options, "pool", 0))
++        {
++            error_setg(errp, "pool number must be specified or included in the inode number");
++            goto out;
++        }
++        if (!qdict_get_try_int(options, "size", 0))
++        {
++            error_setg(errp, "size must be specified when inode number is used instead of image name");
++            goto out;
++        }
++    }
++
++out:
++    g_free(buf);
++    return;
++}
++
++static void coroutine_fn vitastor_co_get_metadata(VitastorRPC *task)
++{
++    BlockDriverState *bs = task->bs;
++    VitastorClient *client = bs->opaque;
++    task->co = qemu_coroutine_self();
++
++    qemu_mutex_lock(&client->mutex);
++    vitastor_c_watch_inode(client->proxy, client->image, vitastor_co_generic_bh_cb, task);
++    qemu_mutex_unlock(&client->mutex);
++
++    while (!task->complete)
++    {
++        qemu_coroutine_yield();
++    }
++}
++
++static void vitastor_aio_set_fd_handler(void *ctx, int fd, int unused1, IOHandler *fd_read, IOHandler *fd_write, void *unused2, void *opaque)
++{
++    aio_set_fd_handler(ctx, fd,
++#if QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 5 || QEMU_VERSION_MAJOR >= 3
++        0 /*is_external*/,
++#endif
++        fd_read, fd_write,
++#if QEMU_VERSION_MAJOR == 1 && QEMU_VERSION_MINOR <= 6 || QEMU_VERSION_MAJOR < 1
++        NULL /*io_flush*/,
++#endif
++#if QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 9 || QEMU_VERSION_MAJOR >= 3
++        NULL /*io_poll*/,
++#endif
++#if QEMU_VERSION_MAJOR >= 7
++        NULL /*io_poll_ready*/,
++#endif
++        opaque);
++}
++
++static int vitastor_file_open(BlockDriverState *bs, QDict *options, int flags, Error **errp)
++{
++    VitastorRPC task;
++    VitastorClient *client = bs->opaque;
++    void *image = NULL;
++    int64_t ret = 0;
++    qemu_mutex_init(&client->mutex);
++    client->config_path = g_strdup(qdict_get_try_str(options, "config-path"));
++    // FIXME: Rename to etcd_address
++    client->etcd_host = g_strdup(qdict_get_try_str(options, "etcd-host"));
++    client->etcd_prefix = g_strdup(qdict_get_try_str(options, "etcd-prefix"));
++    client->skip_parents = qdict_get_try_int(options, "skip-parents", 0);
++    client->use_rdma = qdict_get_try_int(options, "use-rdma", -1);
++    client->rdma_device = g_strdup(qdict_get_try_str(options, "rdma-device"));
++    client->rdma_port_num = qdict_get_try_int(options, "rdma-port-num", 0);
++    client->rdma_gid_index = qdict_get_try_int(options, "rdma-gid-index", 0);
++    client->rdma_mtu = qdict_get_try_int(options, "rdma-mtu", 0);
++    client->proxy = vitastor_c_create_qemu(
++        vitastor_aio_set_fd_handler, bdrv_get_aio_context(bs), client->config_path, client->etcd_host, client->etcd_prefix,
++        client->use_rdma, client->rdma_device, client->rdma_port_num, client->rdma_gid_index, client->rdma_mtu, 0
++    );
++    image = client->image = g_strdup(qdict_get_try_str(options, "image"));
++    client->readonly = (flags & BDRV_O_RDWR) ? 1 : 0;
++    // Get image metadata (size and readonly flag) or just wait until the client is ready
++    if (!image)
++        client->image = (char*)"x";
++    task.complete = 0;
++    task.bs = bs;
++    if (qemu_in_coroutine())
++    {
++        vitastor_co_get_metadata(&task);
++    }
++    else
++    {
++        bdrv_coroutine_enter(bs, qemu_coroutine_create((void(*)(void*))vitastor_co_get_metadata, &task));
++        BDRV_POLL_WHILE(bs, !task.complete);
++    }
++    client->image = image;
++    if (client->image)
++    {
++        client->watch = (void*)task.ret;
++        client->readonly = client->readonly || vitastor_c_inode_get_readonly(client->watch);
++        client->size = vitastor_c_inode_get_size(client->watch);
++        if (!vitastor_c_inode_get_num(client->watch))
++        {
++            error_setg(errp, "image does not exist");
++            vitastor_close(bs);
++            return -1;
++        }
++        if (!client->size)
++        {
++            client->size = qdict_get_try_int(options, "size", 0);
++        }
++    }
++    else
++    {
++        client->watch = NULL;
++        client->inode = qdict_get_try_int(options, "inode", 0);
++        client->pool = qdict_get_try_int(options, "pool", 0);
++        if (client->pool)
++        {
++            client->inode = (client->inode & (((uint64_t)1 << (64-POOL_ID_BITS)) - 1)) | (client->pool << (64-POOL_ID_BITS));
++        }
++        client->size = qdict_get_try_int(options, "size", 0);
++        vitastor_c_close_watch(client->proxy, (void*)task.ret);
++    }
++    if (!client->size)
++    {
++        error_setg(errp, "image size not specified");
++        vitastor_close(bs);
++        return -1;
++    }
++    bs->total_sectors = client->size / BDRV_SECTOR_SIZE;
++    //client->aio_context = bdrv_get_aio_context(bs);
++    qdict_del(options, "use-rdma");
++    qdict_del(options, "rdma-mtu");
++    qdict_del(options, "rdma-gid-index");
++    qdict_del(options, "rdma-port-num");
++    qdict_del(options, "rdma-device");
++    qdict_del(options, "config-path");
++    qdict_del(options, "etcd-host");
++    qdict_del(options, "etcd-prefix");
++    qdict_del(options, "image");
++    qdict_del(options, "inode");
++    qdict_del(options, "pool");
++    qdict_del(options, "size");
++    qdict_del(options, "skip-parents");
++    return ret;
++}
++
++static void vitastor_close(BlockDriverState *bs)
++{
++    VitastorClient *client = bs->opaque;
++    vitastor_c_destroy(client->proxy);
++    qemu_mutex_destroy(&client->mutex);
++    if (client->config_path)
++        g_free(client->config_path);
++    if (client->etcd_host)
++        g_free(client->etcd_host);
++    if (client->etcd_prefix)
++        g_free(client->etcd_prefix);
++    if (client->image)
++        g_free(client->image);
++    free(client->last_bitmap);
++    client->last_bitmap = NULL;
++}
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 2
++static int vitastor_probe_blocksizes(BlockDriverState *bs, BlockSizes *bsz)
++{
++    bsz->phys = 4096;
++    bsz->log = 512;
++    return 0;
++}
++#endif
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 12
++static int coroutine_fn vitastor_co_create_opts(
++#if QEMU_VERSION_MAJOR >= 4
++    BlockDriver *drv,
++#endif
++    const char *url, QemuOpts *opts, Error **errp)
++{
++    QDict *options;
++    int ret;
++
++    options = qdict_new();
++    vitastor_parse_filename(url, options, errp);
++    if (*errp)
++    {
++        ret = -1;
++        goto out;
++    }
++
++    // inodes don't require creation in Vitastor. FIXME: They will when there will be some metadata
++
++    ret = 0;
++out:
++    qobject_unref(options);
++    return ret;
++}
++#endif
++
++#if QEMU_VERSION_MAJOR >= 3
++static int coroutine_fn vitastor_co_truncate(BlockDriverState *bs, int64_t offset,
++#if QEMU_VERSION_MAJOR >= 4
++    bool exact,
++#endif
++    PreallocMode prealloc,
++#if QEMU_VERSION_MAJOR >= 5 && QEMU_VERSION_MINOR >= 1 || QEMU_VERSION_MAJOR > 5 || defined RHEL_BDRV_CO_TRUNCATE_FLAGS
++    BdrvRequestFlags flags,
++#endif
++    Error **errp)
++{
++    VitastorClient *client = bs->opaque;
++
++    if (prealloc != PREALLOC_MODE_OFF)
++    {
++        error_setg(errp, "Unsupported preallocation mode '%s'", PreallocMode_str(prealloc));
++        return -ENOTSUP;
++    }
++
++    // TODO: Resize inode to <offset> bytes
++    client->size = offset / BDRV_SECTOR_SIZE;
++
++    return 0;
++}
++#endif
++
++static int vitastor_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
++{
++    bdi->cluster_size = 4096;
++    return 0;
++}
++
++static int64_t vitastor_getlength(BlockDriverState *bs)
++{
++    VitastorClient *client = bs->opaque;
++    return client->size;
++}
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 0
++static void vitastor_refresh_limits(BlockDriverState *bs, Error **errp)
++#else
++static int vitastor_refresh_limits(BlockDriverState *bs)
++#endif
++{
++    bs->bl.request_alignment = 4096;
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 3
++    bs->bl.min_mem_alignment = 4096;
++#endif
++    bs->bl.opt_mem_alignment = 4096;
++#if QEMU_VERSION_MAJOR < 2 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR == 0
++    return 0;
++#endif
++}
++
++//static int64_t vitastor_get_allocated_file_size(BlockDriverState *bs)
++//{
++//    return 0;
++//}
++
++static void vitastor_co_init_task(BlockDriverState *bs, VitastorRPC *task)
++{
++    *task = (VitastorRPC) {
++        .co     = qemu_coroutine_self(),
++        .bs     = bs,
++    };
++}
++
++static void vitastor_co_generic_bh_cb(void *opaque, long retval)
++{
++    VitastorRPC *task = opaque;
++    task->ret = retval;
++    task->complete = 1;
++    if (qemu_coroutine_self() != task->co)
++    {
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 8
++        aio_co_wake(task->co);
++#else
++        qemu_coroutine_enter(task->co, NULL);
++        qemu_aio_release(task);
++#endif
++    }
++}
++
++static void vitastor_co_read_cb(void *opaque, long retval, uint64_t version)
++{
++    vitastor_co_generic_bh_cb(opaque, retval);
++}
++
++static int coroutine_fn vitastor_co_preadv(BlockDriverState *bs,
++#if QEMU_VERSION_MAJOR >= 7 || QEMU_VERSION_MAJOR == 6 && QEMU_VERSION_MINOR >= 2
++    int64_t offset, int64_t bytes, QEMUIOVector *iov, BdrvRequestFlags flags
++#else
++    uint64_t offset, uint64_t bytes, QEMUIOVector *iov, int flags
++#endif
++)
++{
++    VitastorClient *client = bs->opaque;
++    VitastorRPC task;
++    vitastor_co_init_task(bs, &task);
++    task.iov = iov;
++
++    uint64_t inode = client->watch ? vitastor_c_inode_get_num(client->watch) : client->inode;
++    qemu_mutex_lock(&client->mutex);
++    vitastor_c_read(client->proxy, inode, offset, bytes, iov->iov, iov->niov, vitastor_co_read_cb, &task);
++    qemu_mutex_unlock(&client->mutex);
++
++    while (!task.complete)
++    {
++        qemu_coroutine_yield();
++    }
++
++    return task.ret;
++}
++
++static int coroutine_fn vitastor_co_pwritev(BlockDriverState *bs,
++#if QEMU_VERSION_MAJOR >= 7 || QEMU_VERSION_MAJOR == 6 && QEMU_VERSION_MINOR >= 2
++    int64_t offset, int64_t bytes, QEMUIOVector *iov, BdrvRequestFlags flags
++#else
++    uint64_t offset, uint64_t bytes, QEMUIOVector *iov, int flags
++#endif
++)
++{
++    VitastorClient *client = bs->opaque;
++    VitastorRPC task;
++    vitastor_co_init_task(bs, &task);
++    task.iov = iov;
++
++    if (client->last_bitmap)
++    {
++        // Invalidate last bitmap on write
++        free(client->last_bitmap);
++        client->last_bitmap = NULL;
++    }
++
++    uint64_t inode = client->watch ? vitastor_c_inode_get_num(client->watch) : client->inode;
++    qemu_mutex_lock(&client->mutex);
++    vitastor_c_write(client->proxy, inode, offset, bytes, 0, iov->iov, iov->niov, vitastor_co_generic_bh_cb, &task);
++    qemu_mutex_unlock(&client->mutex);
++
++    while (!task.complete)
++    {
++        qemu_coroutine_yield();
++    }
++
++    return task.ret;
++}
++
++#if defined VITASTOR_C_API_VERSION && VITASTOR_C_API_VERSION >= 1
++#if QEMU_VERSION_MAJOR >= 2 || QEMU_VERSION_MAJOR == 1 && QEMU_VERSION_MINOR >= 7
++static void vitastor_co_read_bitmap_cb(void *opaque, long retval, uint8_t *bitmap)
++{
++    VitastorRPC *task = opaque;
++    VitastorClient *client = task->bs->opaque;
++    task->ret = retval;
++    task->complete = 1;
++    if (retval >= 0)
++    {
++        task->bitmap = bitmap;
++        if (client->last_bitmap_inode == task->inode &&
++            client->last_bitmap_offset == task->offset &&
++            client->last_bitmap_len == task->len)
++        {
++            free(client->last_bitmap);
++            client->last_bitmap = bitmap;
++        }
++    }
++    if (qemu_coroutine_self() != task->co)
++    {
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 8
++        aio_co_wake(task->co);
++#else
++        qemu_coroutine_enter(task->co, NULL);
++        qemu_aio_release(task);
++#endif
++    }
++}
++
++static int coroutine_fn vitastor_co_block_status(
++    BlockDriverState *bs, bool want_zero, int64_t offset, int64_t bytes,
++    int64_t *pnum, int64_t *map, BlockDriverState **file)
++{
++    // Allocated => return BDRV_BLOCK_DATA|BDRV_BLOCK_OFFSET_VALID
++    // Not allocated => return 0
++    // Error => return -errno
++    // Set pnum to length of the extent, `*map` = `offset`, `*file` = `bs`
++    VitastorRPC task;
++    VitastorClient *client = bs->opaque;
++    uint64_t inode = client->watch ? vitastor_c_inode_get_num(client->watch) : client->inode;
++    uint8_t bit = 0;
++    if (client->last_bitmap && client->last_bitmap_inode == inode &&
++        client->last_bitmap_offset <= offset &&
++        client->last_bitmap_offset+client->last_bitmap_len >= (want_zero ? offset+1 : offset+bytes))
++    {
++        // Use the previously read bitmap
++        task.bitmap_granularity = client->last_bitmap_granularity;
++        task.offset = client->last_bitmap_offset;
++        task.len = client->last_bitmap_len;
++        task.bitmap = client->last_bitmap;
++    }
++    else
++    {
++        // Read bitmap from this position, rounding to full inode PG blocks
++        uint32_t block_size = vitastor_c_inode_get_block_size(client->proxy, inode);
++        if (!block_size)
++            return -EAGAIN;
++        // Init coroutine
++        vitastor_co_init_task(bs, &task);
++        free(client->last_bitmap);
++        task.inode = client->last_bitmap_inode = inode;
++        task.bitmap_granularity = client->last_bitmap_granularity = vitastor_c_inode_get_bitmap_granularity(client->proxy, inode);
++        task.offset = client->last_bitmap_offset = offset / block_size * block_size;
++        task.len = client->last_bitmap_len = (offset+bytes+block_size-1) / block_size * block_size - task.offset;
++        task.bitmap = client->last_bitmap = NULL;
++        qemu_mutex_lock(&client->mutex);
++        vitastor_c_read_bitmap(client->proxy, task.inode, task.offset, task.len, !client->skip_parents, vitastor_co_read_bitmap_cb, &task);
++        qemu_mutex_unlock(&client->mutex);
++        while (!task.complete)
++        {
++            qemu_coroutine_yield();
++        }
++        if (task.ret < 0)
++        {
++            // Error
++            return task.ret;
++        }
++    }
++    if (want_zero)
++    {
++        // Get precise mapping with all holes
++        uint64_t bmp_pos = (offset-task.offset) / task.bitmap_granularity;
++        uint64_t bmp_len = task.len / task.bitmap_granularity;
++        uint64_t bmp_end = bmp_pos+1;
++        bit = (task.bitmap[bmp_pos >> 3] >> (bmp_pos & 0x7)) & 1;
++        while (bmp_end < bmp_len && ((task.bitmap[bmp_end >> 3] >> (bmp_end & 0x7)) & 1) == bit)
++        {
++            bmp_end++;
++        }
++        *pnum = (bmp_end-bmp_pos) * task.bitmap_granularity;
++    }
++    else
++    {
++        // Get larger allocated extents, possibly with false positives
++        uint64_t bmp_pos = (offset-task.offset) / task.bitmap_granularity;
++        uint64_t bmp_end = (offset+bytes-task.offset) / task.bitmap_granularity - bmp_pos;
++        while (bmp_pos < bmp_end)
++        {
++            if (!(bmp_pos & 7) && bmp_end >= bmp_pos+8)
++            {
++                bit = bit || task.bitmap[bmp_pos >> 3];
++                bmp_pos += 8;
++            }
++            else
++            {
++                bit = bit || ((task.bitmap[bmp_pos >> 3] >> (bmp_pos & 0x7)) & 1);
++                bmp_pos++;
++            }
++        }
++        *pnum = bytes;
++    }
++    if (bit)
++    {
++        *map = offset;
++        *file = bs;
++    }
++    return (bit ? (BDRV_BLOCK_DATA|BDRV_BLOCK_OFFSET_VALID) : 0);
++}
++#endif
++#if QEMU_VERSION_MAJOR == 1 && QEMU_VERSION_MINOR >= 7 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR < 12
++// QEMU 1.7-2.11
++static int64_t coroutine_fn vitastor_co_get_block_status(BlockDriverState *bs,
++    int64_t sector_num, int nb_sectors, int *pnum, BlockDriverState **file)
++{
++    int64_t map = 0;
++    int64_t pnumbytes = 0;
++    int r = vitastor_co_block_status(bs, 1, sector_num*BDRV_SECTOR_SIZE, nb_sectors*BDRV_SECTOR_SIZE, &pnumbytes, &map, &file);
++    *pnum = pnumbytes/BDRV_SECTOR_SIZE;
++    return r;
++}
++#endif
++#endif
++
++#if !( QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 7 )
++static int coroutine_fn vitastor_co_readv(BlockDriverState *bs, int64_t sector_num, int nb_sectors, QEMUIOVector *iov)
++{
++    return vitastor_co_preadv(bs, sector_num*BDRV_SECTOR_SIZE, nb_sectors*BDRV_SECTOR_SIZE, iov, 0);
++}
++
++static int coroutine_fn vitastor_co_writev(BlockDriverState *bs, int64_t sector_num, int nb_sectors, QEMUIOVector *iov)
++{
++    return vitastor_co_pwritev(bs, sector_num*BDRV_SECTOR_SIZE, nb_sectors*BDRV_SECTOR_SIZE, iov, 0);
++}
++#endif
++
++static int coroutine_fn vitastor_co_flush(BlockDriverState *bs)
++{
++    VitastorClient *client = bs->opaque;
++    VitastorRPC task;
++    vitastor_co_init_task(bs, &task);
++
++    qemu_mutex_lock(&client->mutex);
++    vitastor_c_sync(client->proxy, vitastor_co_generic_bh_cb, &task);
++    qemu_mutex_unlock(&client->mutex);
++
++    while (!task.complete)
++    {
++        qemu_coroutine_yield();
++    }
++
++    return task.ret;
++}
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 0
++static QemuOptsList vitastor_create_opts = {
++    .name = "vitastor-create-opts",
++    .head = QTAILQ_HEAD_INITIALIZER(vitastor_create_opts.head),
++    .desc = {
++        {
++            .name = BLOCK_OPT_SIZE,
++            .type = QEMU_OPT_SIZE,
++            .help = "Virtual disk size"
++        },
++        { /* end of list */ }
++    }
++};
++#else
++static QEMUOptionParameter vitastor_create_opts[] = {
++    {
++        .name = BLOCK_OPT_SIZE,
++        .type = OPT_SIZE,
++        .help = "Virtual disk size"
++    },
++    { NULL }
++};
++#endif
++
++#if QEMU_VERSION_MAJOR >= 4
++static const char *vitastor_strong_runtime_opts[] = {
++    "inode",
++    "pool",
++    "config-path",
++    "etcd-host",
++    "etcd-prefix",
++
++    NULL
++};
++#endif
++
++static BlockDriver bdrv_vitastor = {
++    .format_name                    = "vitastor",
++    .protocol_name                  = "vitastor",
++
++    .instance_size                  = sizeof(VitastorClient),
++    .bdrv_parse_filename            = vitastor_parse_filename,
++
++    .bdrv_has_zero_init             = bdrv_has_zero_init_1,
++    .bdrv_get_info                  = vitastor_get_info,
++    .bdrv_getlength                 = vitastor_getlength,
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 2
++    .bdrv_probe_blocksizes          = vitastor_probe_blocksizes,
++#endif
++    .bdrv_refresh_limits            = vitastor_refresh_limits,
++
++    // FIXME: Implement it along with per-inode statistics
++    //.bdrv_get_allocated_file_size   = vitastor_get_allocated_file_size,
++
++    .bdrv_file_open                 = vitastor_file_open,
++    .bdrv_close                     = vitastor_close,
++
++    // Option list for the create operation
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR > 0
++    .create_opts                    = &vitastor_create_opts,
++#else
++    .create_options                 = vitastor_create_opts,
++#endif
++
++    // For qmp_blockdev_create(), used by the qemu monitor / QAPI
++    // Requires patching QAPI IDL, thus unimplemented
++    //.bdrv_co_create                 = vitastor_co_create,
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 12
++    // For bdrv_create(), used by qemu-img
++    .bdrv_co_create_opts            = vitastor_co_create_opts,
++#endif
++
++#if QEMU_VERSION_MAJOR >= 3
++    .bdrv_co_truncate               = vitastor_co_truncate,
++#endif
++
++#if defined VITASTOR_C_API_VERSION && VITASTOR_C_API_VERSION >= 1
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 12
++    // For snapshot export
++    .bdrv_co_block_status           = vitastor_co_block_status,
++#elif QEMU_VERSION_MAJOR == 1 && QEMU_VERSION_MINOR >= 7 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR < 12
++    .bdrv_co_get_block_status       = vitastor_co_get_block_status,
++#endif
++#endif
++
++#if QEMU_VERSION_MAJOR >= 3 || QEMU_VERSION_MAJOR == 2 && QEMU_VERSION_MINOR >= 7
++    .bdrv_co_preadv                 = vitastor_co_preadv,
++    .bdrv_co_pwritev                = vitastor_co_pwritev,
++#else
++    .bdrv_co_readv                  = vitastor_co_readv,
++    .bdrv_co_writev                 = vitastor_co_writev,
++#endif
++
++    .bdrv_co_flush_to_disk          = vitastor_co_flush,
++
++#if QEMU_VERSION_MAJOR >= 4
++    .strong_runtime_opts            = vitastor_strong_runtime_opts,
++#endif
++};
++
++static void vitastor_block_init(void)
++{
++    bdrv_register(&bdrv_vitastor);
++}
++
++block_init(vitastor_block_init);

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

@@ -14,10 +14,10 @@ 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 7e19bbff5f..b527e82a82 100644
+index d5fd1dbcd2..bda3e606dc 100644
 --- a/block/file-posix.c
 +++ b/block/file-posix.c
-@@ -450,7 +450,7 @@ static QemuOptsList raw_runtime_opts = {
+@@ -508,7 +508,7 @@ static QemuOptsList raw_runtime_opts = {
          {
              .name = "locking",
              .type = QEMU_OPT_STRING,
@@ -26,7 +26,7 @@ index 7e19bbff5f..b527e82a82 100644
          },
          {
              .name = "pr-manager",
-@@ -550,7 +550,7 @@ static int raw_open_common(BlockDriverState *bs, QDict *options,
+@@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict *options,
          s->use_lock = false;
          break;
      case ON_OFF_AUTO_AUTO:

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

@@ -5,22 +5,21 @@ Subject: [PATCH] PVE: [Config] Adjust network script path to /etc/kvm/
 
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
- include/net/net.h | 5 +++--
- 1 file changed, 3 insertions(+), 2 deletions(-)
+ include/net/net.h | 4 ++--
+ 1 file changed, 2 insertions(+), 2 deletions(-)
 
 diff --git a/include/net/net.h b/include/net/net.h
-index 39085d9444..487e3ea1b4 100644
+index 778fc787ca..fb2db6bb75 100644
 --- a/include/net/net.h
 +++ b/include/net/net.h
-@@ -208,8 +208,9 @@ void netdev_add(QemuOpts *opts, Error **errp);
+@@ -210,8 +210,8 @@ 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);
  
--#define DEFAULT_NETWORK_SCRIPT "/etc/qemu-ifup"
--#define DEFAULT_NETWORK_DOWN_SCRIPT "/etc/qemu-ifdown"
-+#define DEFAULT_NETWORK_SCRIPT "/etc/kvm/kvm-ifup"
-+#define DEFAULT_NETWORK_DOWN_SCRIPT "/etc/kvm/kvm-ifdown"
-+
+-#define DEFAULT_NETWORK_SCRIPT CONFIG_SYSCONFDIR "/qemu-ifup"
+-#define DEFAULT_NETWORK_DOWN_SCRIPT CONFIG_SYSCONFDIR "/qemu-ifdown"
++#define DEFAULT_NETWORK_SCRIPT CONFIG_SYSCONFDIR "/kvm/kvm-ifup"
++#define DEFAULT_NETWORK_DOWN_SCRIPT CONFIG_SYSCONFDIR "/kvm/kvm-ifdown"
  #define DEFAULT_BRIDGE_HELPER CONFIG_QEMU_HELPERDIR "/qemu-bridge-helper"
  #define DEFAULT_BRIDGE_INTERFACE "br0"
  

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

@@ -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 e818fc712a..dd9bf7b3da 100644
+index 88e8586f8f..93563ee0c2 100644
 --- a/target/i386/cpu.h
 +++ b/target/i386/cpu.h
-@@ -1954,9 +1954,9 @@ uint64_t cpu_get_tsc(CPUX86State *env);
+@@ -1973,9 +1973,9 @@ uint64_t cpu_get_tsc(CPUX86State *env);
  #define CPU_RESOLVING_TYPE TYPE_X86_CPU
  
  #ifdef TARGET_X86_64

debian/patches/pve/0004-PVE-Config-ui-spice-default-to-pve-certificates.patch

@@ -9,10 +9,10 @@ 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 ecc2ec2c55..ca04965ead 100644
+index eea52f5389..d09ee7f09e 100644
 --- a/ui/spice-core.c
 +++ b/ui/spice-core.c
-@@ -668,32 +668,35 @@ void qemu_spice_init(void)
+@@ -667,32 +667,35 @@ static void qemu_spice_init(void)
  
      if (tls_port) {
          x509_dir = qemu_opt_get(opts, "x509-dir");

debian/patches/pve/0005-PVE-Config-smm_available-false.patch

@@ -10,10 +10,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  1 file changed, 1 insertion(+), 1 deletion(-)
 
 diff --git a/hw/i386/x86.c b/hw/i386/x86.c
-index b82770024c..bd05b3c79a 100644
+index 5944fc44ed..31b481b4e9 100644
 --- a/hw/i386/x86.c
 +++ b/hw/i386/x86.c
-@@ -896,7 +896,7 @@ bool x86_machine_is_smm_enabled(X86MachineState *x86ms)
+@@ -1115,7 +1115,7 @@ bool x86_machine_is_smm_enabled(const X86MachineState *x86ms)
      if (tcg_enabled() || qtest_enabled()) {
          smm_available = true;
      } else if (kvm_enabled()) {

debian/patches/pve/0006-PVE-Config-glusterfs-no-default-logfile-if-daemonize.patch

@@ -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 0aa1f2cda4..dcd1ef7ebc 100644
+index 4f1448e2bc..715f06394c 100644
 --- a/block/gluster.c
 +++ b/block/gluster.c
 @@ -42,7 +42,7 @@

debian/patches/pve/0007-PVE-Config-rbd-block-rbd-disable-rbd_cache_writethro.patch

@@ -18,10 +18,10 @@ 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 e637639a07..5717e7258c 100644
+index 9bd2bce716..c7195a2342 100644
 --- a/block/rbd.c
 +++ b/block/rbd.c
-@@ -651,6 +651,8 @@ static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx,
+@@ -609,6 +609,8 @@ static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx,
          rados_conf_set(*cluster, "rbd_cache", "false");
      }
  

debian/patches/pve/0008-PVE-Up-qmp-add-get_link_status.patch

@@ -11,10 +11,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  3 files changed, 43 insertions(+)
 
 diff --git a/net/net.c b/net/net.c
-index 38778e831d..dabfb482f0 100644
+index 6a2c3d9567..a1e9514fb8 100644
 --- a/net/net.c
 +++ b/net/net.c
-@@ -1331,6 +1331,33 @@ void hmp_info_network(Monitor *mon, const QDict *qdict)
+@@ -1277,6 +1277,33 @@ void hmp_info_network(Monitor *mon, const QDict *qdict)
      }
  }
  
@@ -49,10 +49,10 @@ index 38778e831d..dabfb482f0 100644
  {
      NetClientState *nc;
 diff --git a/qapi/net.json b/qapi/net.json
-index cebb1b52e3..f6854483b1 100644
+index a3a1336001..b8092c4e20 100644
 --- a/qapi/net.json
 +++ b/qapi/net.json
-@@ -34,6 +34,21 @@
+@@ -35,6 +35,21 @@
  ##
  { 'command': 'set_link', 'data': {'name': 'str', 'up': 'bool'} }
  

debian/patches/pve/0009-PVE-Up-glusterfs-allow-partial-reads.patch

@@ -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 dcd1ef7ebc..ac79b4bdb4 100644
+index 715f06394c..6d02170d9b 100644
 --- a/block/gluster.c
 +++ b/block/gluster.c
 @@ -57,6 +57,7 @@ typedef struct GlusterAIOCB {
@@ -27,7 +27,7 @@ index dcd1ef7ebc..ac79b4bdb4 100644
  } GlusterAIOCB;
  
  typedef struct BDRVGlusterState {
-@@ -763,8 +764,10 @@ static void gluster_finish_aiocb(struct glfs_fd *fd, ssize_t ret,
+@@ -759,8 +760,10 @@ static void gluster_finish_aiocb(struct glfs_fd *fd, ssize_t ret,
          acb->ret = 0; /* Success */
      } else if (ret < 0) {
          acb->ret = -errno; /* Read/Write failed */
@@ -39,7 +39,7 @@ index dcd1ef7ebc..ac79b4bdb4 100644
      }
  
      aio_co_schedule(acb->aio_context, acb->coroutine);
-@@ -1035,6 +1038,7 @@ static coroutine_fn int qemu_gluster_co_pwrite_zeroes(BlockDriverState *bs,
+@@ -1028,6 +1031,7 @@ 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 dcd1ef7ebc..ac79b4bdb4 100644
  
      ret = glfs_zerofill_async(s->fd, offset, size, gluster_finish_aiocb, &acb);
      if (ret < 0) {
-@@ -1216,9 +1220,11 @@ static coroutine_fn int qemu_gluster_co_rw(BlockDriverState *bs,
+@@ -1209,9 +1213,11 @@ 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 dcd1ef7ebc..ac79b4bdb4 100644
          ret = glfs_preadv_async(s->fd, qiov->iov, qiov->niov, offset, 0,
                                  gluster_finish_aiocb, &acb);
      }
-@@ -1281,6 +1287,7 @@ static coroutine_fn int qemu_gluster_co_flush_to_disk(BlockDriverState *bs)
+@@ -1275,6 +1281,7 @@ 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 dcd1ef7ebc..ac79b4bdb4 100644
  
      ret = glfs_fsync_async(s->fd, gluster_finish_aiocb, &acb);
      if (ret < 0) {
-@@ -1327,6 +1334,7 @@ static coroutine_fn int qemu_gluster_co_pdiscard(BlockDriverState *bs,
+@@ -1321,6 +1328,7 @@ 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);

debian/patches/pve/0010-PVE-Up-qemu-img-return-success-on-info-without-snaps.patch

@@ -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 821cbf610e..667c540a89 100644
+index f9050bfaad..7e6666b5f7 100644
 --- a/qemu-img.c
 +++ b/qemu-img.c
-@@ -2821,7 +2821,8 @@ static int img_info(int argc, char **argv)
+@@ -3022,7 +3022,8 @@ static int img_info(int argc, char **argv)
      list = collect_image_info_list(image_opts, filename, fmt, chain,
                                     force_share);
      if (!list) {

debian/patches/pve/0011-PVE-Up-qemu-img-dd-add-osize-and-read-from-to-stdin-.patch

@@ -33,14 +33,14 @@ Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
  qemu-img-cmds.hx |   4 +-
- qemu-img.c       | 192 +++++++++++++++++++++++++++++------------------
- 2 files changed, 122 insertions(+), 74 deletions(-)
+ qemu-img.c       | 191 +++++++++++++++++++++++++++++------------------
+ 2 files changed, 121 insertions(+), 74 deletions(-)
 
 diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx
-index c9c54de1df..0f98033658 100644
+index b3620f29e5..e70ef3dc91 100644
 --- a/qemu-img-cmds.hx
 +++ b/qemu-img-cmds.hx
-@@ -51,9 +51,9 @@ SRST
+@@ -58,9 +58,9 @@ SRST
  ERST
  
  DEF("dd", img_dd,
@@ -53,10 +53,10 @@ index c9c54de1df..0f98033658 100644
  
  DEF("info", img_info,
 diff --git a/qemu-img.c b/qemu-img.c
-index 667c540a89..6b7d1fcb51 100644
+index 7e6666b5f7..44cf942bd2 100644
 --- a/qemu-img.c
 +++ b/qemu-img.c
-@@ -4444,10 +4444,12 @@ out:
+@@ -4897,10 +4897,12 @@ static int img_bitmap(int argc, char **argv)
  #define C_IF      04
  #define C_OF      010
  #define C_SKIP    020
@@ -69,7 +69,7 @@ index 667c540a89..6b7d1fcb51 100644
  };
  
  struct DdIo {
-@@ -4526,6 +4528,20 @@ static int img_dd_skip(const char *arg,
+@@ -4976,6 +4978,19 @@ static int img_dd_skip(const char *arg,
      return 0;
  }
  
@@ -77,10 +77,9 @@ index 667c540a89..6b7d1fcb51 100644
 +                        struct DdIo *in, struct DdIo *out,
 +                        struct DdInfo *dd)
 +{
-+    dd->osize = cvtnum(arg);
++    dd->osize = cvtnum("size", arg);
 +
 +    if (dd->osize < 0) {
-+        error_report("invalid number: '%s'", arg);
 +        return 1;
 +    }
 +
@@ -90,7 +89,7 @@ index 667c540a89..6b7d1fcb51 100644
  static int img_dd(int argc, char **argv)
  {
      int ret = 0;
-@@ -4566,6 +4582,7 @@ static int img_dd(int argc, char **argv)
+@@ -5016,6 +5031,7 @@ 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 +97,7 @@ index 667c540a89..6b7d1fcb51 100644
          { NULL, NULL, 0 }
      };
      const struct option long_options[] = {
-@@ -4644,8 +4661,13 @@ static int img_dd(int argc, char **argv)
+@@ -5094,8 +5110,13 @@ static int img_dd(int argc, char **argv)
          arg = NULL;
      }
  
@@ -114,7 +113,7 @@ index 667c540a89..6b7d1fcb51 100644
          ret = -1;
          goto out;
      }
-@@ -4657,85 +4679,101 @@ static int img_dd(int argc, char **argv)
+@@ -5107,85 +5128,101 @@ static int img_dd(int argc, char **argv)
          goto out;
      }
  
@@ -280,7 +279,7 @@ index 667c540a89..6b7d1fcb51 100644
      }
  
      if (dd.flags & C_SKIP && (in.offset > INT64_MAX / in.bsz ||
-@@ -4753,11 +4791,17 @@ static int img_dd(int argc, char **argv)
+@@ -5203,11 +5240,17 @@ static int img_dd(int argc, char **argv)
  
      for (out_pos = 0; in_pos < size; block_count++) {
          int in_ret, out_ret;
@@ -302,7 +301,7 @@ index 667c540a89..6b7d1fcb51 100644
          }
          if (in_ret < 0) {
              error_report("error while reading from input image file: %s",
-@@ -4767,9 +4811,13 @@ static int img_dd(int argc, char **argv)
+@@ -5217,9 +5260,13 @@ static int img_dd(int argc, char **argv)
          }
          in_pos += in_ret;
  

debian/patches/pve/0012-PVE-Up-qemu-img-dd-add-isize-parameter.patch

@@ -11,14 +11,14 @@ an expected end of input.
 Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
- qemu-img.c | 29 ++++++++++++++++++++++++++---
- 1 file changed, 26 insertions(+), 3 deletions(-)
+ qemu-img.c | 28 +++++++++++++++++++++++++---
+ 1 file changed, 25 insertions(+), 3 deletions(-)
 
 diff --git a/qemu-img.c b/qemu-img.c
-index 6b7d1fcb51..17393b2f53 100644
+index 44cf942bd2..5ce60e8a45 100644
 --- a/qemu-img.c
 +++ b/qemu-img.c
-@@ -4445,11 +4445,13 @@ out:
+@@ -4898,11 +4898,13 @@ static int img_bitmap(int argc, char **argv)
  #define C_OF      010
  #define C_SKIP    020
  #define C_OSIZE   040
@@ -32,7 +32,7 @@ index 6b7d1fcb51..17393b2f53 100644
  };
  
  struct DdIo {
-@@ -4542,6 +4544,20 @@ static int img_dd_osize(const char *arg,
+@@ -4991,6 +4993,19 @@ static int img_dd_osize(const char *arg,
      return 0;
  }
  
@@ -40,10 +40,9 @@ index 6b7d1fcb51..17393b2f53 100644
 +                        struct DdIo *in, struct DdIo *out,
 +                        struct DdInfo *dd)
 +{
-+    dd->isize = cvtnum(arg);
++    dd->isize = cvtnum("size", arg);
 +
 +    if (dd->isize < 0) {
-+        error_report("invalid number: '%s'", arg);
 +        return 1;
 +    }
 +
@@ -53,7 +52,7 @@ index 6b7d1fcb51..17393b2f53 100644
  static int img_dd(int argc, char **argv)
  {
      int ret = 0;
-@@ -4556,12 +4572,14 @@ static int img_dd(int argc, char **argv)
+@@ -5005,12 +5020,14 @@ static int img_dd(int argc, char **argv)
      int c, i;
      const char *out_fmt = "raw";
      const char *fmt = NULL;
@@ -69,7 +68,7 @@ index 6b7d1fcb51..17393b2f53 100644
      };
      struct DdIo in = {
          .bsz = 512, /* Block size is by default 512 bytes */
-@@ -4583,6 +4601,7 @@ static int img_dd(int argc, char **argv)
+@@ -5032,6 +5049,7 @@ 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 +76,7 @@ index 6b7d1fcb51..17393b2f53 100644
          { NULL, NULL, 0 }
      };
      const struct option long_options[] = {
-@@ -4789,14 +4808,18 @@ static int img_dd(int argc, char **argv)
+@@ -5238,14 +5256,18 @@ static int img_dd(int argc, char **argv)
  
      in.buf = g_new(uint8_t, in.bsz);
  

debian/patches/pve/0013-PVE-Up-qemu-img-dd-add-n-skip_create.patch

@@ -1,7 +1,7 @@
 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
 From: Alexandre Derumier <aderumier@odiso.com>
 Date: Mon, 6 Apr 2020 12:16:42 +0200
-Subject: [PATCH] PVE: [Up] qemu-img dd : add -n skip_create
+Subject: [PATCH] PVE: [Up] qemu-img dd: add -n skip_create
 
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
@@ -9,10 +9,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  1 file changed, 14 insertions(+), 9 deletions(-)
 
 diff --git a/qemu-img.c b/qemu-img.c
-index 17393b2f53..574bb3c73d 100644
+index 5ce60e8a45..86bfd0288b 100644
 --- a/qemu-img.c
 +++ b/qemu-img.c
-@@ -4574,7 +4574,7 @@ static int img_dd(int argc, char **argv)
+@@ -5022,7 +5022,7 @@ static int img_dd(int argc, char **argv)
      const char *fmt = NULL;
      int64_t size = 0, readsize = 0;
      int64_t block_count = 0, out_pos, in_pos;
@@ -21,7 +21,7 @@ index 17393b2f53..574bb3c73d 100644
      struct DdInfo dd = {
          .flags = 0,
          .count = 0,
-@@ -4612,7 +4612,7 @@ static int img_dd(int argc, char **argv)
+@@ -5060,7 +5060,7 @@ static int img_dd(int argc, char **argv)
          { 0, 0, 0, 0 }
      };
  
@@ -30,7 +30,7 @@ index 17393b2f53..574bb3c73d 100644
          if (c == EOF) {
              break;
          }
-@@ -4632,6 +4632,9 @@ static int img_dd(int argc, char **argv)
+@@ -5080,6 +5080,9 @@ static int img_dd(int argc, char **argv)
          case 'h':
              help();
              break;
@@ -40,7 +40,7 @@ index 17393b2f53..574bb3c73d 100644
          case 'U':
              force_share = true;
              break;
-@@ -4772,13 +4775,15 @@ static int img_dd(int argc, char **argv)
+@@ -5220,13 +5223,15 @@ static int img_dd(int argc, char **argv)
                                  size - in.bsz * in.offset, &error_abort);
          }
  

debian/patches/pve/0014-PVE-virtio-balloon-improve-query-balloon.patch

@@ -10,14 +10,14 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
  hw/virtio/virtio-balloon.c | 33 +++++++++++++++++++++++++++++++--
  monitor/hmp-cmds.c         | 30 +++++++++++++++++++++++++++++-
- qapi/misc.json             | 22 +++++++++++++++++++++-
+ qapi/machine.json          | 22 +++++++++++++++++++++-
  3 files changed, 81 insertions(+), 4 deletions(-)
 
 diff --git a/hw/virtio/virtio-balloon.c b/hw/virtio/virtio-balloon.c
-index a4729f7fc9..97c1c16ccf 100644
+index b22b5beda3..6e581439bf 100644
 --- a/hw/virtio/virtio-balloon.c
 +++ b/hw/virtio/virtio-balloon.c
-@@ -713,8 +713,37 @@ static uint64_t virtio_balloon_get_features(VirtIODevice *vdev, uint64_t f,
+@@ -805,8 +805,37 @@ static uint64_t virtio_balloon_get_features(VirtIODevice *vdev, uint64_t f,
  static void virtio_balloon_stat(void *opaque, BalloonInfo *info)
  {
      VirtIOBalloon *dev = opaque;
@@ -58,10 +58,10 @@ index a4729f7fc9..97c1c16ccf 100644
  
  static void virtio_balloon_to_target(void *opaque, ram_addr_t target)
 diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
-index 9b94e67879..0c6f6ff331 100644
+index 65d8ff4849..705f08a8f1 100644
 --- a/monitor/hmp-cmds.c
 +++ b/monitor/hmp-cmds.c
-@@ -653,7 +653,35 @@ void hmp_info_balloon(Monitor *mon, const QDict *qdict)
+@@ -695,7 +695,35 @@ void hmp_info_balloon(Monitor *mon, const QDict *qdict)
          return;
      }
  
@@ -98,13 +98,13 @@ index 9b94e67879..0c6f6ff331 100644
  
      qapi_free_BalloonInfo(info);
  }
-diff --git a/qapi/misc.json b/qapi/misc.json
-index 99b90ac80b..e2a6678eae 100644
---- a/qapi/misc.json
-+++ b/qapi/misc.json
-@@ -225,10 +225,30 @@
- #
- # @actual: the number of bytes the balloon currently contains
+diff --git a/qapi/machine.json b/qapi/machine.json
+index 7c9a263778..3e59199280 100644
+--- a/qapi/machine.json
++++ b/qapi/machine.json
+@@ -1205,10 +1205,30 @@
+ # @actual: the logical size of the VM in bytes
+ #          Formula used: logical_vm_size = vm_ram_size - balloon_size
  #
 +# @last_update: time when stats got updated from guest
 +#

debian/patches/pve/0015-PVE-qapi-modify-query-machines.patch

@@ -13,10 +13,10 @@ Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
  2 files changed, 9 insertions(+), 1 deletion(-)
 
 diff --git a/hw/core/machine-qmp-cmds.c b/hw/core/machine-qmp-cmds.c
-index eed5aeb2f7..1953633e82 100644
+index 5362c80a18..3fcb82ce2f 100644
 --- a/hw/core/machine-qmp-cmds.c
 +++ b/hw/core/machine-qmp-cmds.c
-@@ -230,6 +230,12 @@ MachineInfoList *qmp_query_machines(Error **errp)
+@@ -234,6 +234,12 @@ MachineInfoList *qmp_query_machines(Error **errp)
          info->hotpluggable_cpus = mc->has_hotpluggable_cpus;
          info->numa_mem_supported = mc->numa_mem_supported;
          info->deprecated = !!mc->deprecation_reason;
@@ -30,10 +30,10 @@ index eed5aeb2f7..1953633e82 100644
              info->default_cpu_type = g_strdup(mc->default_cpu_type);
              info->has_default_cpu_type = true;
 diff --git a/qapi/machine.json b/qapi/machine.json
-index ff7b5032e3..f6cf28f9fd 100644
+index 3e59199280..dfc1a49d3c 100644
 --- a/qapi/machine.json
 +++ b/qapi/machine.json
-@@ -340,6 +340,8 @@
+@@ -318,6 +318,8 @@
  #
  # @is-default: whether the machine is default
  #
@@ -42,12 +42,12 @@ index ff7b5032e3..f6cf28f9fd 100644
  # @cpu-max: maximum number of CPUs supported by the machine type
  #           (since 1.5.0)
  #
-@@ -359,7 +361,7 @@
+@@ -339,7 +341,7 @@
  ##
  { 'struct': 'MachineInfo',
    'data': { 'name': 'str', '*alias': 'str',
 -            '*is-default': 'bool', 'cpu-max': 'int',
 +            '*is-default': 'bool', '*is-current': 'bool', 'cpu-max': 'int',
              'hotpluggable-cpus': 'bool',  'numa-mem-supported': 'bool',
-             'deprecated': 'bool', '*default-cpu-type': 'str' } }
- 
+             'deprecated': 'bool', '*default-cpu-type': 'str',
+             '*default-ram-id': 'str' } }

debian/patches/pve/0016-PVE-qapi-modify-spice-query.patch

@@ -12,10 +12,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  2 files changed, 8 insertions(+)
 
 diff --git a/qapi/ui.json b/qapi/ui.json
-index e16e98a060..feda6ef090 100644
+index 6c7b33cb72..39ff301d1e 100644
 --- a/qapi/ui.json
 +++ b/qapi/ui.json
-@@ -213,11 +213,14 @@
+@@ -215,11 +215,14 @@
  #
  # @channels: a list of @SpiceChannel for each active spice channel
  #
@@ -31,10 +31,10 @@ index e16e98a060..feda6ef090 100644
    'if': 'defined(CONFIG_SPICE)' }
  
 diff --git a/ui/spice-core.c b/ui/spice-core.c
-index ca04965ead..243466c13d 100644
+index d09ee7f09e..da3d2644d1 100644
 --- a/ui/spice-core.c
 +++ b/ui/spice-core.c
-@@ -539,6 +539,11 @@ SpiceInfo *qmp_query_spice(Error **errp)
+@@ -538,6 +538,11 @@ static SpiceInfo *qmp_query_spice_real(Error **errp)
      micro = SPICE_SERVER_VERSION & 0xff;
      info->compiled_version = g_strdup_printf("%d.%d.%d", major, minor, micro);
  

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

@@ -1,44 +1,48 @@
 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
 From: Dietmar Maurer <dietmar@proxmox.com>
 Date: Mon, 6 Apr 2020 12:16:46 +0200
-Subject: [PATCH] PVE: internal snapshot async
+Subject: [PATCH] PVE: add savevm-async for background state snapshots
 
-Truncate at 1024 boundary (Fabian Ebner will send a patch for stable)
+Put 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.
+
+Truncate output file at 1024 boundary.
+
+Do not block the VM and save the state on aborting a snapshot, as the
+snapshot will be invalid anyway.
+
+Also, when aborting, wait for the target file to be closed, otherwise a
+client might run into race-conditions when trying to remove the file
+still opened by QEMU.
 
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
+Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
+[improve aborting]
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
 ---
- Makefile.objs                |   1 +
  hmp-commands-info.hx         |  13 +
- hmp-commands.hx              |  32 +++
+ hmp-commands.hx              |  33 ++
  include/migration/snapshot.h |   1 +
  include/monitor/hmp.h        |   5 +
- monitor/hmp-cmds.c           |  57 +++++
- qapi/migration.json          |  34 +++
- qapi/misc.json               |  32 +++
+ migration/meson.build        |   1 +
+ migration/savevm-async.c     | 598 +++++++++++++++++++++++++++++++++++
+ monitor/hmp-cmds.c           |  57 ++++
+ qapi/migration.json          |  34 ++
+ qapi/misc.json               |  32 ++
  qemu-options.hx              |  12 +
- savevm-async.c               | 464 +++++++++++++++++++++++++++++++++++
  softmmu/vl.c                 |  10 +
- 11 files changed, 661 insertions(+)
- create mode 100644 savevm-async.c
+ 11 files changed, 796 insertions(+)
+ create mode 100644 migration/savevm-async.c
 
-diff --git a/Makefile.objs b/Makefile.objs
-index a7c967633a..d0b4dde836 100644
---- a/Makefile.objs
-+++ b/Makefile.objs
-@@ -47,6 +47,7 @@ common-obj-y += bootdevice.o iothread.o
- common-obj-y += dump/
- common-obj-y += job-qmp.o
- common-obj-y += monitor/
-+common-obj-y += savevm-async.o
- common-obj-y += net/
- common-obj-y += qdev-monitor.o
- common-obj-$(CONFIG_WIN32) += os-win32.o
 diff --git a/hmp-commands-info.hx b/hmp-commands-info.hx
-index ca5198438d..89fea71972 100644
+index 117ba25f91..b3b797ca28 100644
 --- a/hmp-commands-info.hx
 +++ b/hmp-commands-info.hx
-@@ -579,6 +579,19 @@ SRST
+@@ -580,6 +580,19 @@ SRST
      Show current migration xbzrle cache size.
  ERST
  
@@ -59,10 +63,10 @@ index ca5198438d..89fea71972 100644
          .name       = "balloon",
          .args_type  = "",
 diff --git a/hmp-commands.hx b/hmp-commands.hx
-index 7f0f3974ad..81fe305d07 100644
+index ff2d7aa8f3..d294c234a5 100644
 --- a/hmp-commands.hx
 +++ b/hmp-commands.hx
-@@ -1814,3 +1814,35 @@ ERST
+@@ -1866,3 +1866,36 @@ ERST
          .flags      = "p",
      },
  
@@ -96,7 +100,8 @@ index 7f0f3974ad..81fe305d07 100644
 +        .args_type  = "",
 +        .params     = "",
 +        .help       = "Resume VM after snaphot.",
-+        .cmd = hmp_savevm_end,
++        .cmd        = hmp_savevm_end,
++        .coroutine  = true,
 +    },
 diff --git a/include/migration/snapshot.h b/include/migration/snapshot.h
 index c85b6ec75b..4411b7121d 100644
@@ -110,7 +115,7 @@ index c85b6ec75b..4411b7121d 100644
  
  #endif
 diff --git a/include/monitor/hmp.h b/include/monitor/hmp.h
-index e33ca5a911..601827d43f 100644
+index ed2913fd18..4e06f89e8e 100644
 --- a/include/monitor/hmp.h
 +++ b/include/monitor/hmp.h
 @@ -25,6 +25,7 @@ void hmp_info_status(Monitor *mon, const QDict *qdict);
@@ -132,191 +137,24 @@ index e33ca5a911..601827d43f 100644
  void hmp_sendkey(Monitor *mon, const QDict *qdict);
  void hmp_screendump(Monitor *mon, const QDict *qdict);
  void hmp_chardev_add(Monitor *mon, const QDict *qdict);
-diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
-index 0c6f6ff331..39c7474cea 100644
---- a/monitor/hmp-cmds.c
-+++ b/monitor/hmp-cmds.c
-@@ -1876,6 +1876,63 @@ void hmp_info_memory_devices(Monitor *mon, const QDict *qdict)
-     hmp_handle_error(mon, err);
- }
- 
-+void hmp_savevm_start(Monitor *mon, const QDict *qdict)
-+{
-+    Error *errp = NULL;
-+    const char *statefile = qdict_get_try_str(qdict, "statefile");
-+
-+    qmp_savevm_start(statefile != NULL, statefile, &errp);
-+    hmp_handle_error(mon, errp);
-+}
-+
-+void hmp_snapshot_drive(Monitor *mon, const QDict *qdict)
-+{
-+    Error *errp = NULL;
-+    const char *name = qdict_get_str(qdict, "name");
-+    const char *device = qdict_get_str(qdict, "device");
-+
-+    qmp_snapshot_drive(device, name, &errp);
-+    hmp_handle_error(mon, errp);
-+}
-+
-+void hmp_delete_drive_snapshot(Monitor *mon, const QDict *qdict)
-+{
-+    Error *errp = NULL;
-+    const char *name = qdict_get_str(qdict, "name");
-+    const char *device = qdict_get_str(qdict, "device");
-+
-+    qmp_delete_drive_snapshot(device, name, &errp);
-+    hmp_handle_error(mon, errp);
-+}
-+
-+void hmp_savevm_end(Monitor *mon, const QDict *qdict)
-+{
-+    Error *errp = NULL;
-+
-+    qmp_savevm_end(&errp);
-+    hmp_handle_error(mon, errp);
-+}
-+
-+void hmp_info_savevm(Monitor *mon, const QDict *qdict)
-+{
-+    SaveVMInfo *info;
-+    info = qmp_query_savevm(NULL);
-+
-+    if (info->has_status) {
-+        monitor_printf(mon, "savevm status: %s\n", info->status);
-+        monitor_printf(mon, "total time: %" PRIu64 " milliseconds\n",
-+                       info->total_time);
-+    } else {
-+        monitor_printf(mon, "savevm status: not running\n");
-+    }
-+    if (info->has_bytes) {
-+        monitor_printf(mon, "Bytes saved: %"PRIu64"\n", info->bytes);
-+    }
-+    if (info->has_error) {
-+        monitor_printf(mon, "Error: %s\n", info->error);
-+    }
-+}
-+
- void hmp_info_iothreads(Monitor *mon, const QDict *qdict)
- {
-     IOThreadInfoList *info_list = qmp_query_iothreads(NULL);
-diff --git a/qapi/migration.json b/qapi/migration.json
-index eca2981d0a..081663d67a 100644
---- a/qapi/migration.json
-+++ b/qapi/migration.json
-@@ -222,6 +222,40 @@
-            '*compression': 'CompressionStats',
-            '*socket-address': ['SocketAddress'] } }
- 
-+##
-+# @SaveVMInfo:
-+#
-+# Information about current migration process.
-+#
-+# @status: string describing the current savevm status.
-+#          This can be 'active', 'completed', 'failed'.
-+#          If this field is not returned, no savevm process
-+#          has been initiated
-+#
-+# @error: string containing error message is status is failed.
-+#
-+# @total-time: total amount of milliseconds since savevm started.
-+#        If savevm has ended, it returns the total save time
-+#
-+# @bytes: total amount of data transfered
-+#
-+# Since: 1.3
-+##
-+{ 'struct': 'SaveVMInfo',
-+  'data': {'*status': 'str', '*error': 'str',
-+           '*total-time': 'int', '*bytes': 'int'} }
-+
-+##
-+# @query-savevm:
-+#
-+# Returns information about current savevm process.
-+#
-+# Returns: @SaveVMInfo
-+#
-+# Since: 1.3
-+##
-+{ 'command': 'query-savevm', 'returns': 'SaveVMInfo' }
-+
- ##
- # @query-migrate:
- #
-diff --git a/qapi/misc.json b/qapi/misc.json
-index e2a6678eae..0868de22b7 100644
---- a/qapi/misc.json
-+++ b/qapi/misc.json
-@@ -1165,6 +1165,38 @@
- ##
- { 'command': 'query-fdsets', 'returns': ['FdsetInfo'] }
- 
-+##
-+# @savevm-start:
-+#
-+# Prepare for snapshot and halt VM. Save VM state to statefile.
-+#
-+##
-+{ 'command': 'savevm-start', 'data': { '*statefile': 'str' } }
-+
-+##
-+# @snapshot-drive:
-+#
-+# Create an internal drive snapshot.
-+#
-+##
-+{ 'command': 'snapshot-drive', 'data': { 'device': 'str', 'name': 'str' } }
-+
-+##
-+# @delete-drive-snapshot:
-+#
-+# Delete a drive snapshot.
-+#
-+##
-+{ 'command': 'delete-drive-snapshot', 'data': { 'device': 'str', 'name': 'str' } }
-+
-+##
-+# @savevm-end:
-+#
-+# Resume VM after a snapshot.
-+#
-+##
-+{ 'command': 'savevm-end' }
-+
- ##
- # @AcpiTableOptions:
- #
-diff --git a/qemu-options.hx b/qemu-options.hx
-index 292d4e7c0c..55eef64ddf 100644
---- a/qemu-options.hx
-+++ b/qemu-options.hx
-@@ -3832,6 +3832,18 @@ SRST
-     Start right away with a saved state (``loadvm`` in monitor)
- ERST
- 
-+DEF("loadstate", HAS_ARG, QEMU_OPTION_loadstate, \
-+    "-loadstate file\n" \
-+    "                start right away with a saved state\n",
-+    QEMU_ARCH_ALL)
-+SRST
-+``-loadstate file``
-+  Start right away with a saved state. This option does not rollback
-+  disk state like @code{loadvm}, so user must make sure that disk
-+  have correct state. @var{file} can be any valid device URL. See the section
-+  for "Device URL Syntax" for more information.
-+ERST
-+
- #ifndef _WIN32
- DEF("daemonize", 0, QEMU_OPTION_daemonize, \
-     "-daemonize      daemonize QEMU after initializing\n", QEMU_ARCH_ALL)
-diff --git a/savevm-async.c b/savevm-async.c
+diff --git a/migration/meson.build b/migration/meson.build
+index 980e37865c..e62b79b60f 100644
+--- a/migration/meson.build
++++ b/migration/meson.build
+@@ -23,6 +23,7 @@ softmmu_ss.add(files(
+   'multifd-zlib.c',
+   'postcopy-ram.c',
+   'savevm.c',
++  'savevm-async.c',
+   'socket.c',
+   'tls.c',
+ ))
+diff --git a/migration/savevm-async.c b/migration/savevm-async.c
 new file mode 100644
-index 0000000000..54ceeae26c
+index 0000000000..593a619088
 --- /dev/null
-+++ b/savevm-async.c
-@@ -0,0 +1,464 @@
++++ b/migration/savevm-async.c
+@@ -0,0 +1,598 @@
 +#include "qemu/osdep.h"
 +#include "migration/migration.h"
 +#include "migration/savevm.h"
@@ -335,6 +173,7 @@ index 0000000000..54ceeae26c
 +#include "qapi/qapi-commands-misc.h"
 +#include "qapi/qapi-commands-block.h"
 +#include "qemu/cutils.h"
++#include "qemu/timer.h"
 +#include "qemu/main-loop.h"
 +#include "qemu/rcu.h"
 +
@@ -369,10 +208,17 @@ index 0000000000..54ceeae26c
 +    int saved_vm_running;
 +    QEMUFile *file;
 +    int64_t total_time;
-+    QEMUBH *cleanup_bh;
-+    QemuThread thread;
++    QEMUBH *finalize_bh;
++    Coroutine *co;
++    QemuCoSleepState *target_close_wait;
 +} snap_state;
 +
++static bool savevm_aborted(void)
++{
++    return snap_state.state == SAVE_STATE_CANCELLED ||
++        snap_state.state == SAVE_STATE_ERROR;
++}
++
 +SaveVMInfo *qmp_query_savevm(Error **errp)
 +{
 +    SaveVMInfo *info = g_malloc0(sizeof(*info));
@@ -425,17 +271,23 @@ index 0000000000..54ceeae26c
 +    }
 +
 +    if (snap_state.target) {
-+        /* try to truncate, but ignore errors (will fail on block devices).
-+         * note1: bdrv_read() need whole blocks, so we need to round up
-+         * note2: PVE requires 1024 (BDRV_SECTOR_SIZE*2) alignment
-+         */
-+        size_t size = QEMU_ALIGN_UP(snap_state.bs_pos, BDRV_SECTOR_SIZE*2);
-+        blk_truncate(snap_state.target, size, false, PREALLOC_MODE_OFF, NULL);
++        if (!savevm_aborted()) {
++            /* try to truncate, but ignore errors (will fail on block devices).
++            * note1: bdrv_read() need whole blocks, so we need to round up
++            * note2: PVE requires 1024 (BDRV_SECTOR_SIZE*2) alignment
++            */
++            size_t size = QEMU_ALIGN_UP(snap_state.bs_pos, BDRV_SECTOR_SIZE*2);
++            blk_truncate(snap_state.target, size, false, PREALLOC_MODE_OFF, 0, NULL);
++        }
 +        blk_op_unblock_all(snap_state.target, snap_state.blocker);
 +        error_free(snap_state.blocker);
 +        snap_state.blocker = NULL;
 +        blk_unref(snap_state.target);
 +        snap_state.target = NULL;
++
++        if (snap_state.target_close_wait) {
++            qemu_co_sleep_wake(snap_state.target_close_wait);
++        }
 +    }
 +
 +    return ret;
@@ -477,6 +329,7 @@ index 0000000000..54ceeae26c
 +    BlkRwCo *rwco = opaque;
 +    rwco->ret = blk_co_pwritev(snap_state.target, rwco->offset, rwco->qiov->size,
 +                               rwco->qiov, 0);
++    aio_wait_kick();
 +}
 +
 +static ssize_t block_state_writev_buffer(void *opaque, struct iovec *iov,
@@ -514,19 +367,63 @@ index 0000000000..54ceeae26c
 +    .close =          block_state_close,
 +};
 +
-+static void process_savevm_cleanup(void *opaque)
++static void process_savevm_finalize(void *opaque)
 +{
 +    int ret;
-+    qemu_bh_delete(snap_state.cleanup_bh);
-+    snap_state.cleanup_bh = NULL;
-+    qemu_mutex_unlock_iothread();
-+    qemu_thread_join(&snap_state.thread);
-+    qemu_mutex_lock_iothread();
++    AioContext *iohandler_ctx = iohandler_get_aio_context();
++    MigrationState *ms = migrate_get_current();
++
++    bool aborted = savevm_aborted();
++
++#ifdef DEBUG_SAVEVM_STATE
++    int64_t start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
++#endif
++
++    qemu_bh_delete(snap_state.finalize_bh);
++    snap_state.finalize_bh = NULL;
++    snap_state.co = NULL;
++
++    /* We need to own the target bdrv's context for the following functions,
++     * so move it back. It can stay in the main context and live out its live
++     * there, since we're done with it after this method ends anyway.
++     */
++    aio_context_acquire(iohandler_ctx);
++    blk_set_aio_context(snap_state.target, qemu_get_aio_context(), NULL);
++    aio_context_release(iohandler_ctx);
++
++    ret = vm_stop_force_state(RUN_STATE_FINISH_MIGRATE);
++    if (ret < 0) {
++        save_snapshot_error("vm_stop_force_state error %d", ret);
++    }
++
++    if (!aborted) {
++        /* skip state saving if we aborted, snapshot will be invalid anyway */
++        (void)qemu_savevm_state_complete_precopy(snap_state.file, false, false);
++        ret = qemu_file_get_error(snap_state.file);
++        if (ret < 0) {
++                save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
++        }
++    }
++
++    DPRINTF("state saving complete\n");
++    DPRINTF("timing: process_savevm_finalize (state saving) took %ld ms\n",
++        qemu_clock_get_ms(QEMU_CLOCK_REALTIME) - start_time);
++
++    /* clear migration state */
++    migrate_set_state(&ms->state, MIGRATION_STATUS_SETUP,
++        ret || aborted ? MIGRATION_STATUS_FAILED : MIGRATION_STATUS_COMPLETED);
++    ms->to_dst_file = NULL;
++
++    qemu_savevm_state_cleanup();
++
 +    ret = save_snapshot_cleanup();
 +    if (ret < 0) {
 +        save_snapshot_error("save_snapshot_cleanup error %d", ret);
 +    } else if (snap_state.state == SAVE_STATE_ACTIVE) {
 +        snap_state.state = SAVE_STATE_COMPLETED;
++    } else if (aborted) {
++        save_snapshot_error("process_savevm_cleanup: found aborted state: %d",
++                            snap_state.state);
 +    } else {
 +        save_snapshot_error("process_savevm_cleanup: invalid state: %d",
 +                            snap_state.state);
@@ -535,82 +432,98 @@ index 0000000000..54ceeae26c
 +        vm_start();
 +        snap_state.saved_vm_running = false;
 +    }
++
++    DPRINTF("timing: process_savevm_finalize (full) took %ld ms\n",
++        qemu_clock_get_ms(QEMU_CLOCK_REALTIME) - start_time);
 +}
 +
-+static void *process_savevm_thread(void *opaque)
++static void coroutine_fn process_savevm_co(void *opaque)
 +{
 +    int ret;
 +    int64_t maxlen;
++    BdrvNextIterator it;
++    BlockDriverState *bs = NULL;
 +
-+    rcu_register_thread();
++#ifdef DEBUG_SAVEVM_STATE
++    int64_t start_time = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
++#endif
 +
-+    qemu_savevm_state_header(snap_state.file);
-+    qemu_savevm_state_setup(snap_state.file);
 +    ret = qemu_file_get_error(snap_state.file);
-+
 +    if (ret < 0) {
 +        save_snapshot_error("qemu_savevm_state_setup failed");
-+        rcu_unregister_thread();
-+        return NULL;
++        return;
 +    }
 +
 +    while (snap_state.state == SAVE_STATE_ACTIVE) {
 +        uint64_t pending_size, pend_precopy, pend_compatible, pend_postcopy;
 +
++        /* pending is expected to be called without iothread lock */
++        qemu_mutex_unlock_iothread();
 +        qemu_savevm_state_pending(snap_state.file, 0, &pend_precopy, &pend_compatible, &pend_postcopy);
++        qemu_mutex_lock_iothread();
++
 +        pending_size = pend_precopy + pend_compatible + pend_postcopy;
 +
 +        maxlen = blk_getlength(snap_state.target) - 30*1024*1024;
 +
 +        if (pending_size > 400000 && snap_state.bs_pos + pending_size < maxlen) {
-+            qemu_mutex_lock_iothread();
 +            ret = qemu_savevm_state_iterate(snap_state.file, false);
 +            if (ret < 0) {
 +                save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
 +                break;
 +            }
-+            qemu_mutex_unlock_iothread();
-+            DPRINTF("savevm inerate pending size %lu ret %d\n", pending_size, ret);
++            DPRINTF("savevm iterate pending size %lu ret %d\n", pending_size, ret);
 +        } else {
-+            qemu_mutex_lock_iothread();
 +            qemu_system_wakeup_request(QEMU_WAKEUP_REASON_OTHER, NULL);
 +            ret = global_state_store();
 +            if (ret) {
 +                save_snapshot_error("global_state_store error %d", ret);
 +                break;
 +            }
-+            ret = vm_stop_force_state(RUN_STATE_FINISH_MIGRATE);
-+            if (ret < 0) {
-+                save_snapshot_error("vm_stop_force_state error %d", ret);
-+                break;
-+            }
-+            DPRINTF("savevm inerate finished\n");
-+            /* upstream made the return value here inconsistent
-+             * (-1 instead of 'ret' in one case and 0 after flush which can
-+             * still set a file error...)
-+             */
-+            (void)qemu_savevm_state_complete_precopy(snap_state.file, false, false);
-+            ret = qemu_file_get_error(snap_state.file);
-+            if (ret < 0) {
-+                    save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
-+                    break;
-+            }
-+            qemu_savevm_state_cleanup();
-+            DPRINTF("save complete\n");
++
++            DPRINTF("savevm iterate complete\n");
 +            break;
 +        }
 +    }
 +
-+    qemu_bh_schedule(snap_state.cleanup_bh);
-+    qemu_mutex_unlock_iothread();
++    DPRINTF("timing: process_savevm_co took %ld ms\n",
++        qemu_clock_get_ms(QEMU_CLOCK_REALTIME) - start_time);
 +
-+    rcu_unregister_thread();
-+    return NULL;
++#ifdef DEBUG_SAVEVM_STATE
++    int64_t start_time_flush = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
++#endif
++    /* If a drive runs in an IOThread we can flush it async, and only
++     * need to sync-flush whatever IO happens between now and
++     * vm_stop_force_state. bdrv_next can only be called from main AioContext,
++     * 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)) {
++        /* 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_flush(bs);
++            aio_co_reschedule_self(qemu_get_aio_context());
++        }
++    }
++
++    DPRINTF("timing: async flushing took %ld ms\n",
++        qemu_clock_get_ms(QEMU_CLOCK_REALTIME) - start_time_flush);
++
++    qemu_bh_schedule(snap_state.finalize_bh);
 +}
 +
 +void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
 +{
 +    Error *local_err = NULL;
++    MigrationState *ms = migrate_get_current();
++    AioContext *iohandler_ctx = iohandler_get_aio_context();
 +
 +    int bdrv_oflags = BDRV_O_RDWR | BDRV_O_RESIZE | BDRV_O_NO_FLUSH;
 +
@@ -620,6 +533,17 @@ index 0000000000..54ceeae26c
 +        return;
 +    }
 +
++    if (migration_is_running(ms->state)) {
++        error_set(errp, ERROR_CLASS_GENERIC_ERROR, QERR_MIGRATION_ACTIVE);
++        return;
++    }
++
++    if (migrate_use_block()) {
++        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
++                  "Block migration and snapshots are incompatible");
++        return;
++    }
++
 +    /* initialize snapshot info */
 +    snap_state.saved_vm_running = runstate_is_running();
 +    snap_state.bs_pos = 0;
@@ -658,14 +582,32 @@ index 0000000000..54ceeae26c
 +        goto restart;
 +    }
 +
++    /*
++     * qemu_savevm_* paths use migration code and expect a migration state.
++     * 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);
++    memset(&ram_counters, 0, sizeof(ram_counters));
++    ms->to_dst_file = snap_state.file;
 +
 +    error_setg(&snap_state.blocker, "block device is in use by savevm");
 +    blk_op_block_all(snap_state.target, snap_state.blocker);
 +
 +    snap_state.state = SAVE_STATE_ACTIVE;
-+    snap_state.cleanup_bh = qemu_bh_new(process_savevm_cleanup, &snap_state);
-+    qemu_thread_create(&snap_state.thread, "savevm-async", process_savevm_thread,
-+                       NULL, QEMU_THREAD_JOINABLE);
++    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.
++     */
++    blk_set_aio_context(snap_state.target, iohandler_ctx, &local_err);
++
++    aio_co_schedule(iohandler_ctx, snap_state.co);
 +
 +    return;
 +
@@ -675,11 +617,14 @@ index 0000000000..54ceeae26c
 +
 +    if (snap_state.saved_vm_running) {
 +        vm_start();
++        snap_state.saved_vm_running = false;
 +    }
 +}
 +
-+void qmp_savevm_end(Error **errp)
++void coroutine_fn qmp_savevm_end(Error **errp)
 +{
++    int64_t timeout;
++
 +    if (snap_state.state == SAVE_STATE_DONE) {
 +        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
 +                  "VM snapshot not started\n");
@@ -688,14 +633,38 @@ index 0000000000..54ceeae26c
 +
 +    if (snap_state.state == SAVE_STATE_ACTIVE) {
 +        snap_state.state = SAVE_STATE_CANCELLED;
-+        return;
++        goto wait_for_close;
 +    }
 +
 +    if (snap_state.saved_vm_running) {
 +        vm_start();
++        snap_state.saved_vm_running = false;
 +    }
 +
 +    snap_state.state = SAVE_STATE_DONE;
++
++wait_for_close:
++    if (!snap_state.target) {
++        DPRINTF("savevm-end: no target file open\n");
++        return;
++    }
++
++    /* wait until cleanup is done before returning, this ensures that after this
++     * call exits the statefile will be closed and can be removed immediately */
++    DPRINTF("savevm-end: waiting for cleanup\n");
++    timeout = 30L * 1000 * 1000 * 1000;
++    qemu_co_sleep_ns_wakeable(QEMU_CLOCK_REALTIME, timeout,
++                              &snap_state.target_close_wait);
++    snap_state.target_close_wait = NULL;
++    if (snap_state.target) {
++        save_snapshot_error("timeout waiting for target file close in "
++                            "qmp_savevm_end");
++        /* we cannot assume the snapshot finished in this case, so leave the
++         * state alone - caller has to figure something out */
++        return;
++    }
++
++    DPRINTF("savevm-end: cleanup done\n");
 +}
 +
 +// FIXME: Deprecated
@@ -764,6 +733,9 @@ index 0000000000..54ceeae26c
 +    qemu_system_reset(SHUTDOWN_CAUSE_NONE);
 +    ret = qemu_loadvm_state(f);
 +
++    /* dirty bitmap migration has a special case we need to trigger manually */
++    dirty_bitmap_mig_before_vm_start();
++
 +    qemu_fclose(f);
 +    migration_incoming_state_destroy();
 +    if (ret < 0) {
@@ -781,11 +753,190 @@ index 0000000000..54ceeae26c
 +    }
 +    return ret;
 +}
+diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
+index 705f08a8f1..77ab152aab 100644
+--- a/monitor/hmp-cmds.c
++++ b/monitor/hmp-cmds.c
+@@ -1949,6 +1949,63 @@ void hmp_info_memory_devices(Monitor *mon, const QDict *qdict)
+     hmp_handle_error(mon, err);
+ }
+ 
++void hmp_savevm_start(Monitor *mon, const QDict *qdict)
++{
++    Error *errp = NULL;
++    const char *statefile = qdict_get_try_str(qdict, "statefile");
++
++    qmp_savevm_start(statefile != NULL, statefile, &errp);
++    hmp_handle_error(mon, errp);
++}
++
++void hmp_snapshot_drive(Monitor *mon, const QDict *qdict)
++{
++    Error *errp = NULL;
++    const char *name = qdict_get_str(qdict, "name");
++    const char *device = qdict_get_str(qdict, "device");
++
++    qmp_snapshot_drive(device, name, &errp);
++    hmp_handle_error(mon, errp);
++}
++
++void hmp_delete_drive_snapshot(Monitor *mon, const QDict *qdict)
++{
++    Error *errp = NULL;
++    const char *name = qdict_get_str(qdict, "name");
++    const char *device = qdict_get_str(qdict, "device");
++
++    qmp_delete_drive_snapshot(device, name, &errp);
++    hmp_handle_error(mon, errp);
++}
++
++void coroutine_fn hmp_savevm_end(Monitor *mon, const QDict *qdict)
++{
++    Error *errp = NULL;
++
++    qmp_savevm_end(&errp);
++    hmp_handle_error(mon, errp);
++}
++
++void hmp_info_savevm(Monitor *mon, const QDict *qdict)
++{
++    SaveVMInfo *info;
++    info = qmp_query_savevm(NULL);
++
++    if (info->has_status) {
++        monitor_printf(mon, "savevm status: %s\n", info->status);
++        monitor_printf(mon, "total time: %" PRIu64 " milliseconds\n",
++                       info->total_time);
++    } else {
++        monitor_printf(mon, "savevm status: not running\n");
++    }
++    if (info->has_bytes) {
++        monitor_printf(mon, "Bytes saved: %"PRIu64"\n", info->bytes);
++    }
++    if (info->has_error) {
++        monitor_printf(mon, "Error: %s\n", info->error);
++    }
++}
++
+ void hmp_info_iothreads(Monitor *mon, const QDict *qdict)
+ {
+     IOThreadInfoList *info_list = qmp_query_iothreads(NULL);
+diff --git a/qapi/migration.json b/qapi/migration.json
+index 3c75820527..cb3627884c 100644
+--- a/qapi/migration.json
++++ b/qapi/migration.json
+@@ -242,6 +242,40 @@
+            '*compression': 'CompressionStats',
+            '*socket-address': ['SocketAddress'] } }
+ 
++##
++# @SaveVMInfo:
++#
++# Information about current migration process.
++#
++# @status: string describing the current savevm status.
++#          This can be 'active', 'completed', 'failed'.
++#          If this field is not returned, no savevm process
++#          has been initiated
++#
++# @error: string containing error message is status is failed.
++#
++# @total-time: total amount of milliseconds since savevm started.
++#              If savevm has ended, it returns the total save time
++#
++# @bytes: total amount of data transfered
++#
++# Since: 1.3
++##
++{ 'struct': 'SaveVMInfo',
++  'data': {'*status': 'str', '*error': 'str',
++           '*total-time': 'int', '*bytes': 'int'} }
++
++##
++# @query-savevm:
++#
++# Returns information about current savevm process.
++#
++# Returns: @SaveVMInfo
++#
++# Since: 1.3
++##
++{ 'command': 'query-savevm', 'returns': 'SaveVMInfo' }
++
+ ##
+ # @query-migrate:
+ #
+diff --git a/qapi/misc.json b/qapi/misc.json
+index 40df513856..4f5333d960 100644
+--- a/qapi/misc.json
++++ b/qapi/misc.json
+@@ -476,6 +476,38 @@
+ ##
+ { 'command': 'query-fdsets', 'returns': ['FdsetInfo'] }
+ 
++##
++# @savevm-start:
++#
++# Prepare for snapshot and halt VM. Save VM state to statefile.
++#
++##
++{ 'command': 'savevm-start', 'data': { '*statefile': 'str' } }
++
++##
++# @snapshot-drive:
++#
++# Create an internal drive snapshot.
++#
++##
++{ 'command': 'snapshot-drive', 'data': { 'device': 'str', 'name': 'str' } }
++
++##
++# @delete-drive-snapshot:
++#
++# Delete a drive snapshot.
++#
++##
++{ 'command': 'delete-drive-snapshot', 'data': { 'device': 'str', 'name': 'str' } }
++
++##
++# @savevm-end:
++#
++# Resume VM after a snapshot.
++#
++##
++{ 'command': 'savevm-end', 'coroutine': true }
++
+ ##
+ # @CommandLineParameterType:
+ #
+diff --git a/qemu-options.hx b/qemu-options.hx
+index 104632ea34..c1352312c2 100644
+--- a/qemu-options.hx
++++ b/qemu-options.hx
+@@ -3903,6 +3903,18 @@ SRST
+     Start right away with a saved state (``loadvm`` in monitor)
+ ERST
+ 
++DEF("loadstate", HAS_ARG, QEMU_OPTION_loadstate, \
++    "-loadstate file\n" \
++    "                start right away with a saved state\n",
++    QEMU_ARCH_ALL)
++SRST
++``-loadstate file``
++  Start right away with a saved state. This option does not rollback
++  disk state like @code{loadvm}, so user must make sure that disk
++  have correct state. @var{file} can be any valid device URL. See the section
++  for "Device URL Syntax" for more information.
++ERST
++
+ #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 32c0047889..4b45eb0c37 100644
+index e6e0ad5a92..03152c816c 100644
 --- a/softmmu/vl.c
 +++ b/softmmu/vl.c
-@@ -2827,6 +2827,7 @@ void qemu_init(int argc, char **argv, char **envp)
+@@ -2878,6 +2878,7 @@ void qemu_init(int argc, char **argv, char **envp)
      int optind;
      const char *optarg;
      const char *loadvm = NULL;
@@ -793,7 +944,7 @@ index 32c0047889..4b45eb0c37 100644
      MachineClass *machine_class;
      const char *cpu_option;
      const char *vga_model = NULL;
-@@ -3391,6 +3392,9 @@ void qemu_init(int argc, char **argv, char **envp)
+@@ -3439,6 +3440,9 @@ void qemu_init(int argc, char **argv, char **envp)
              case QEMU_OPTION_loadvm:
                  loadvm = optarg;
                  break;
@@ -803,7 +954,7 @@ index 32c0047889..4b45eb0c37 100644
              case QEMU_OPTION_full_screen:
                  dpy.has_full_screen = true;
                  dpy.full_screen = true;
-@@ -4447,6 +4451,12 @@ void qemu_init(int argc, char **argv, char **envp)
+@@ -4478,6 +4482,12 @@ void qemu_init(int argc, char **argv, char **envp)
              autostart = 0;
              exit(1);
          }

debian/patches/pve/0018-PVE-add-optional-buffer-size-to-QEMUFile.patch

@@ -1,31 +1,35 @@
 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
 From: Wolfgang Bumiller <w.bumiller@proxmox.com>
 Date: Mon, 4 May 2020 11:05:08 +0200
-Subject: [PATCH] add optional buffer size to QEMUFile
+Subject: [PATCH] PVE: add optional buffer size to QEMUFile
 
 So we can use a 4M buffer for savevm-async which should
 increase performance storing the state onto ceph.
 
 Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
+[increase max IOV count in QEMUFile to actually write more data]
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
 ---
- migration/qemu-file.c | 36 ++++++++++++++++++++++++------------
- migration/qemu-file.h |  1 +
- savevm-async.c        |  4 ++--
- 3 files changed, 27 insertions(+), 14 deletions(-)
+ migration/qemu-file.c    | 38 +++++++++++++++++++++++++-------------
+ migration/qemu-file.h    |  1 +
+ migration/savevm-async.c |  4 ++--
+ 3 files changed, 28 insertions(+), 15 deletions(-)
 
 diff --git a/migration/qemu-file.c b/migration/qemu-file.c
-index 1c3a358a14..7362e51c71 100644
+index be21518c57..1926b5202c 100644
 --- a/migration/qemu-file.c
 +++ b/migration/qemu-file.c
-@@ -30,7 +30,7 @@
+@@ -30,8 +30,8 @@
  #include "trace.h"
  #include "qapi/error.h"
  
 -#define IO_BUF_SIZE 32768
+-#define MAX_IOV_SIZE MIN_CONST(IOV_MAX, 64)
 +#define DEFAULT_IO_BUF_SIZE 32768
- #define MAX_IOV_SIZE MIN(IOV_MAX, 64)
++#define MAX_IOV_SIZE MIN_CONST(IOV_MAX, 256)
  
  struct QEMUFile {
+     const QEMUFileOps *ops;
 @@ -45,7 +45,8 @@ struct QEMUFile {
                      when reading */
      int buf_index;
@@ -159,11 +163,11 @@ index a9b6d6ccb7..8752d27c74 100644
  void qemu_file_set_hooks(QEMUFile *f, const QEMUFileHooks *hooks);
  int qemu_get_fd(QEMUFile *f);
  int qemu_fclose(QEMUFile *f);
-diff --git a/savevm-async.c b/savevm-async.c
-index af865b9a0a..c3fe741c38 100644
---- a/savevm-async.c
-+++ b/savevm-async.c
-@@ -338,7 +338,7 @@ void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
+diff --git a/migration/savevm-async.c b/migration/savevm-async.c
+index 593a619088..cc2552d977 100644
+--- a/migration/savevm-async.c
++++ b/migration/savevm-async.c
+@@ -418,7 +418,7 @@ void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
          goto restart;
      }
  
@@ -172,7 +176,7 @@ index af865b9a0a..c3fe741c38 100644
  
      if (!snap_state.file) {
          error_set(errp, ERROR_CLASS_GENERIC_ERROR, "failed to open '%s'", statefile);
-@@ -454,7 +454,7 @@ int load_snapshot_from_blockdev(const char *filename, Error **errp)
+@@ -567,7 +567,7 @@ int load_snapshot_from_blockdev(const char *filename, Error **errp)
      blk_op_block_all(be, blocker);
  
      /* restore the VM state */

debian/patches/pve/0019-PVE-block-add-the-zeroinit-block-driver-filter.patch

@@ -5,29 +5,29 @@ Subject: [PATCH] PVE: block: add the zeroinit block driver filter
 
 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
- block/Makefile.objs |   1 +
- block/zeroinit.c    | 197 ++++++++++++++++++++++++++++++++++++++++++++
- 2 files changed, 198 insertions(+)
+ block/meson.build |   1 +
+ block/zeroinit.c  | 196 ++++++++++++++++++++++++++++++++++++++++++++++
+ 2 files changed, 197 insertions(+)
  create mode 100644 block/zeroinit.c
 
-diff --git a/block/Makefile.objs b/block/Makefile.objs
-index 3635b6b4c1..1282445672 100644
---- a/block/Makefile.objs
-+++ b/block/Makefile.objs
-@@ -11,6 +11,7 @@ block-obj-$(CONFIG_QED) += qed.o qed-l2-cache.o qed-table.o qed-cluster.o
- block-obj-$(CONFIG_QED) += qed-check.o
- block-obj-y += vhdx.o vhdx-endian.o vhdx-log.o
- block-obj-y += quorum.o
-+block-obj-y += zeroinit.o
- block-obj-y += blkdebug.o blkverify.o blkreplay.o
- block-obj-$(CONFIG_PARALLELS) += parallels.o
- block-obj-y += blklogwrites.o
+diff --git a/block/meson.build b/block/meson.build
+index 5dcc1e5cce..c10d544864 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -39,6 +39,7 @@ block_ss.add(files(
+   'vmdk.c',
+   'vpc.c',
+   'write-threshold.c',
++  'zeroinit.c',
+ ), zstd, zlib)
+ 
+ softmmu_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..ff38388d94
+index 0000000000..5529627f7e
 --- /dev/null
 +++ b/block/zeroinit.c
-@@ -0,0 +1,197 @@
+@@ -0,0 +1,196 @@
 +/*
 + * Filter to fake a zero-initialized block device.
 + *
@@ -107,7 +107,7 @@ index 0000000000..ff38388d94
 +
 +    /* Open the raw file */
 +    bs->file = bdrv_open_child(qemu_opt_get(opts, "x-next"), options, "next",
-+                               bs, &child_file, false, &local_err);
++                               bs, &child_of_bds, BDRV_CHILD_FILTERED, false, &local_err);
 +    if (local_err) {
 +        ret = -EINVAL;
 +        error_propagate(errp, local_err);
@@ -180,9 +180,10 @@ index 0000000000..ff38388d94
 +}
 +
 +static int zeroinit_co_truncate(BlockDriverState *bs, int64_t offset,
-+                                _Bool exact, PreallocMode prealloc, Error **errp)
++                                _Bool exact, PreallocMode prealloc,
++                                BdrvRequestFlags req_flags, Error **errp)
 +{
-+    return bdrv_co_truncate(bs->file, offset, exact, prealloc, errp);
++    return bdrv_co_truncate(bs->file, offset, exact, prealloc, req_flags, errp);
 +}
 +
 +static int zeroinit_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
@@ -199,7 +200,7 @@ index 0000000000..ff38388d94
 +    .bdrv_file_open                   = zeroinit_open,
 +    .bdrv_close                       = zeroinit_close,
 +    .bdrv_getlength                   = zeroinit_getlength,
-+    .bdrv_child_perm                  = bdrv_filter_default_perms,
++    .bdrv_child_perm                  = bdrv_default_perms,
 +    .bdrv_co_flush_to_disk            = zeroinit_co_flush,
 +
 +    .bdrv_co_pwrite_zeroes            = zeroinit_co_pwrite_zeroes,
@@ -211,8 +212,6 @@ index 0000000000..ff38388d94
 +
 +    .bdrv_has_zero_init               = zeroinit_has_zero_init,
 +
-+    .bdrv_co_block_status             = bdrv_co_block_status_from_file,
-+
 +    .bdrv_co_pdiscard                 = zeroinit_co_pdiscard,
 +
 +    .bdrv_co_truncate                 = zeroinit_co_truncate,

debian/patches/pve/0020-PVE-Add-dummy-id-command-line-parameter.patch

@@ -14,10 +14,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  2 files changed, 11 insertions(+)
 
 diff --git a/qemu-options.hx b/qemu-options.hx
-index 55eef64ddf..e11b4f8ff5 100644
+index c1352312c2..9a0cb6780e 100644
 --- a/qemu-options.hx
 +++ b/qemu-options.hx
-@@ -904,6 +904,9 @@ DEFHEADING()
+@@ -906,6 +906,9 @@ DEFHEADING()
  
  DEFHEADING(Block device options:)
  
@@ -28,10 +28,10 @@ index 55eef64ddf..e11b4f8ff5 100644
      "-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 4b45eb0c37..9de81875fd 100644
+index 03152c816c..da204d24f0 100644
 --- a/softmmu/vl.c
 +++ b/softmmu/vl.c
-@@ -2815,6 +2815,7 @@ static void create_default_memdev(MachineState *ms, const char *path)
+@@ -2866,6 +2866,7 @@ static char *find_datadir(void)
  void qemu_init(int argc, char **argv, char **envp)
  {
      int i;
@@ -39,7 +39,7 @@ index 4b45eb0c37..9de81875fd 100644
      int snapshot, linux_boot;
      const char *initrd_filename;
      const char *kernel_filename, *kernel_cmdline;
-@@ -3518,6 +3519,13 @@ void qemu_init(int argc, char **argv, char **envp)
+@@ -3557,6 +3558,13 @@ void qemu_init(int argc, char **argv, char **envp)
                      exit(1);
                  }
                  break;

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

@@ -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 9ec0f2deb2..a00d45251f 100644
+index 502e94effc..590ef6ec8e 100644
 --- a/hw/intc/apic_common.c
 +++ b/hw/intc/apic_common.c
-@@ -259,6 +259,15 @@ static void apic_reset_common(DeviceState *dev)
+@@ -278,6 +278,15 @@ static void apic_reset_common(DeviceState *dev)
      info->vapic_base_update(s);
  
      apic_init_reset(dev);

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

@@ -13,10 +13,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  2 files changed, 43 insertions(+), 21 deletions(-)
 
 diff --git a/block/file-posix.c b/block/file-posix.c
-index b527e82a82..36ebd0967e 100644
+index bda3e606dc..037839622e 100644
 --- a/block/file-posix.c
 +++ b/block/file-posix.c
-@@ -2309,6 +2309,7 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
+@@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
      int fd;
      uint64_t perm, shared;
      int result = 0;
@@ -24,7 +24,7 @@ index b527e82a82..36ebd0967e 100644
  
      /* Validate options and set default values */
      assert(options->driver == BLOCKDEV_DRIVER_FILE);
-@@ -2342,19 +2343,22 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
+@@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
      perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
      shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
  
@@ -59,7 +59,7 @@ index b527e82a82..36ebd0967e 100644
      }
  
      /* Clear the file by truncating it to 0 */
-@@ -2387,13 +2391,15 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
+@@ -2494,13 +2498,15 @@ raw_co_create(BlockdevCreateOptions *options, Error **errp)
      }
  
  out_unlock:
@@ -82,7 +82,7 @@ index b527e82a82..36ebd0967e 100644
      }
  
  out_close:
-@@ -2416,6 +2422,7 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
+@@ -2525,6 +2531,7 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
      PreallocMode prealloc;
      char *buf = NULL;
      Error *local_err = NULL;
@@ -90,7 +90,7 @@ index b527e82a82..36ebd0967e 100644
  
      /* Skip file: protocol prefix */
      strstart(filename, "file:", &filename);
-@@ -2433,6 +2440,18 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
+@@ -2547,6 +2554,18 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
          return -EINVAL;
      }
  
@@ -109,16 +109,16 @@ index b527e82a82..36ebd0967e 100644
      options = (BlockdevCreateOptions) {
          .driver     = BLOCKDEV_DRIVER_FILE,
          .u.file     = {
-@@ -2442,6 +2461,8 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
-             .preallocation      = prealloc,
-             .has_nocow          = true,
+@@ -2558,6 +2577,8 @@ static int coroutine_fn raw_co_create_opts(BlockDriver *drv,
              .nocow              = nocow,
+             .has_extent_size_hint = has_extent_size_hint,
+             .extent_size_hint   = extent_size_hint,
 +            .has_locking        = true,
 +            .locking            = locking,
          },
      };
      return raw_co_create(&options, errp);
-@@ -2983,7 +3004,7 @@ static int raw_check_perm(BlockDriverState *bs, uint64_t perm, uint64_t shared,
+@@ -3104,7 +3125,7 @@ static int raw_check_perm(BlockDriverState *bs, uint64_t perm, uint64_t shared,
      }
  
      /* Copy locks to the new fd */
@@ -128,15 +128,15 @@ index b527e82a82..36ebd0967e 100644
                                     false, errp);
          if (ret < 0) {
 diff --git a/qapi/block-core.json b/qapi/block-core.json
-index 943df1926a..4c55464f86 100644
+index 9db3120716..d285622589 100644
 --- a/qapi/block-core.json
 +++ b/qapi/block-core.json
-@@ -4183,7 +4183,8 @@
-   'data': { 'filename':         'str',
-             'size':             'size',
-             '*preallocation':   'PreallocMode',
--            '*nocow':           'bool' } }
-+            '*nocow':           'bool',
+@@ -4224,7 +4224,8 @@
+             'size':                 'size',
+             '*preallocation':       'PreallocMode',
+             '*nocow':               'bool',
+-            '*extent-size-hint':    'size'} }
++            '*extent-size-hint':    'size',
 +            '*locking':         'OnOffAuto' } }
  
  ##

debian/patches/pve/0022-PVE-savevm-async-kick-AIO-wait-on-block-state-write.patch

@@ -1,22 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Thomas Lamprecht <t.lamprecht@proxmox.com>
-Date: Mon, 6 Apr 2020 12:16:51 +0200
-Subject: [PATCH] PVE: savevm-async: kick AIO wait on block state write
-
-Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
----
- savevm-async.c | 1 +
- 1 file changed, 1 insertion(+)
-
-diff --git a/savevm-async.c b/savevm-async.c
-index 54ceeae26c..393d55af2a 100644
---- a/savevm-async.c
-+++ b/savevm-async.c
-@@ -158,6 +158,7 @@ static void coroutine_fn block_state_write_entry(void *opaque) {
-     BlkRwCo *rwco = opaque;
-     rwco->ret = blk_co_pwritev(snap_state.target, rwco->offset, rwco->qiov->size,
-                                rwco->qiov, 0);
-+    aio_wait_kick();
- }
- 
- static ssize_t block_state_writev_buffer(void *opaque, struct iovec *iov,

debian/patches/pve/0023-PVE-monitor-disable-oob-capability.patch

@@ -18,10 +18,10 @@ Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  1 file changed, 1 insertion(+), 2 deletions(-)
 
 diff --git a/monitor/qmp.c b/monitor/qmp.c
-index f89e7daf27..ed5e39fcf7 100644
+index b42f8c6af3..2e37d11bd3 100644
 --- a/monitor/qmp.c
 +++ b/monitor/qmp.c
-@@ -406,8 +406,7 @@ void monitor_init_qmp(Chardev *chr, bool pretty, Error **errp)
+@@ -466,8 +466,7 @@ void monitor_init_qmp(Chardev *chr, bool pretty, Error **errp)
      qemu_chr_fe_set_echo(&mon->common.chr, true);
  
      /* Note: we run QMP monitor in I/O thread when @chr supports that */

debian/patches/pve/0023-PVE-move-snapshot-cleanup-into-bottom-half.patch

@@ -1,38 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Wolfgang Bumiller <w.bumiller@proxmox.com>
-Date: Mon, 6 Apr 2020 12:16:52 +0200
-Subject: [PATCH] PVE: move snapshot cleanup into bottom half
-
-as per:
-    (0ceccd858a8d) migration: qemu_savevm_state_cleanup() in cleanup
-
-may affect held locks and therefore change assumptions made
-by that function!
-
-Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
-Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
----
- savevm-async.c | 3 ++-
- 1 file changed, 2 insertions(+), 1 deletion(-)
-
-diff --git a/savevm-async.c b/savevm-async.c
-index 393d55af2a..790e27ae37 100644
---- a/savevm-async.c
-+++ b/savevm-async.c
-@@ -201,6 +201,8 @@ static void process_savevm_cleanup(void *opaque)
-     int ret;
-     qemu_bh_delete(snap_state.cleanup_bh);
-     snap_state.cleanup_bh = NULL;
-+    qemu_savevm_state_cleanup();
-+
-     qemu_mutex_unlock_iothread();
-     qemu_thread_join(&snap_state.thread);
-     qemu_mutex_lock_iothread();
-@@ -277,7 +279,6 @@ static void *process_savevm_thread(void *opaque)
-                     save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
-                     break;
-             }
--            qemu_savevm_state_cleanup();
-             DPRINTF("save complete\n");
-             break;
-         }

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

@@ -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 c1a444cb75..9f56ecc4e8 100644
+index d0408049b5..5b38cf9356 100644
 --- a/hw/core/machine.c
 +++ b/hw/core/machine.c
-@@ -56,7 +56,8 @@ GlobalProperty hw_compat_4_0[] = {
+@@ -78,7 +78,8 @@ GlobalProperty hw_compat_4_0[] = {
      { "virtio-vga",     "edid", "false" },
      { "virtio-gpu-device", "edid", "false" },
      { "virtio-device", "use-started", "false" },

debian/patches/pve/0025-PVE-Allow-version-code-in-machine-type.patch

@@ -13,15 +13,15 @@ Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
 ---
  hw/core/machine-qmp-cmds.c |  6 ++++++
  include/hw/boards.h        |  2 ++
- qapi/machine.json          |  3 ++-
+ qapi/machine.json          |  4 +++-
  softmmu/vl.c               | 15 ++++++++++++++-
- 4 files changed, 24 insertions(+), 2 deletions(-)
+ 4 files changed, 25 insertions(+), 2 deletions(-)
 
 diff --git a/hw/core/machine-qmp-cmds.c b/hw/core/machine-qmp-cmds.c
-index 1953633e82..ca8c0dc53d 100644
+index 3fcb82ce2f..7868241bd5 100644
 --- a/hw/core/machine-qmp-cmds.c
 +++ b/hw/core/machine-qmp-cmds.c
-@@ -234,6 +234,12 @@ MachineInfoList *qmp_query_machines(Error **errp)
+@@ -238,6 +238,12 @@ MachineInfoList *qmp_query_machines(Error **errp)
          if (strcmp(mc->name, MACHINE_GET_CLASS(current_machine)->name) == 0) {
              info->has_is_current = true;
              info->is_current = true;
@@ -35,10 +35,10 @@ index 1953633e82..ca8c0dc53d 100644
  
          if (mc->default_cpu_type) {
 diff --git a/include/hw/boards.h b/include/hw/boards.h
-index fd4d62b501..dd395e9232 100644
+index a49e3a6b44..8e0a8c5571 100644
 --- a/include/hw/boards.h
 +++ b/include/hw/boards.h
-@@ -170,6 +170,8 @@ struct MachineClass {
+@@ -165,6 +165,8 @@ struct MachineClass {
      const char *desc;
      const char *deprecation_reason;
  
@@ -48,24 +48,32 @@ index fd4d62b501..dd395e9232 100644
      void (*reset)(MachineState *state);
      void (*wakeup)(MachineState *state);
 diff --git a/qapi/machine.json b/qapi/machine.json
-index f6cf28f9fd..a7f9c79a59 100644
+index dfc1a49d3c..32fc674042 100644
 --- a/qapi/machine.json
 +++ b/qapi/machine.json
-@@ -363,7 +363,8 @@
-   'data': { 'name': 'str', '*alias': 'str',
+@@ -337,6 +337,8 @@
+ #
+ # @default-ram-id: the default ID of initial RAM memory backend (since 5.2)
+ #
++# @pve-version: custom PVE version suffix specified as 'machine+pveN'
++#
+ # Since: 1.2.0
+ ##
+ { 'struct': 'MachineInfo',
+@@ -344,7 +346,7 @@
              '*is-default': 'bool', '*is-current': 'bool', 'cpu-max': 'int',
              'hotpluggable-cpus': 'bool',  'numa-mem-supported': 'bool',
--            'deprecated': 'bool', '*default-cpu-type': 'str' } }
-+            'deprecated': 'bool', '*default-cpu-type': 'str',
-+            '*pve-version': 'str' } }
+             'deprecated': 'bool', '*default-cpu-type': 'str',
+-            '*default-ram-id': 'str' } }
++            '*default-ram-id': 'str', '*pve-version': 'str' } }
  
  ##
  # @query-machines:
 diff --git a/softmmu/vl.c b/softmmu/vl.c
-index 9de81875fd..8340c4ca53 100644
+index da204d24f0..5b5512128e 100644
 --- a/softmmu/vl.c
 +++ b/softmmu/vl.c
-@@ -2300,6 +2300,8 @@ static MachineClass *machine_parse(const char *name, GSList *machines)
+@@ -2325,6 +2325,8 @@ static MachineClass *machine_parse(const char *name, GSList *machines)
  {
      MachineClass *mc;
      GSList *el;
@@ -74,7 +82,7 @@ index 9de81875fd..8340c4ca53 100644
  
      if (is_help_option(name)) {
          printf("Supported machines are:\n");
-@@ -2316,12 +2318,23 @@ static MachineClass *machine_parse(const char *name, GSList *machines)
+@@ -2341,12 +2343,23 @@ static MachineClass *machine_parse(const char *name, GSList *machines)
          exit(0);
      }
  

debian/patches/pve/0026-PVE-Backup-add-vma-backup-format-code.patch

@@ -4,51 +4,54 @@ Date: Mon, 6 Apr 2020 12:16:57 +0200
 Subject: [PATCH] PVE-Backup: add vma backup format code
 
 ---
- Makefile      |   3 +-
- Makefile.objs |   1 +
- vma-reader.c  | 857 ++++++++++++++++++++++++++++++++++++++++++++++++++
- vma-writer.c  | 771 +++++++++++++++++++++++++++++++++++++++++++++
- vma.c         | 837 ++++++++++++++++++++++++++++++++++++++++++++++++
- vma.h         | 150 +++++++++
- 6 files changed, 2618 insertions(+), 1 deletion(-)
+ block/meson.build |   2 +
+ meson.build       |   5 +
+ vma-reader.c      | 857 ++++++++++++++++++++++++++++++++++++++++++++++
+ vma-writer.c      | 790 ++++++++++++++++++++++++++++++++++++++++++
+ vma.c             | 839 +++++++++++++++++++++++++++++++++++++++++++++
+ vma.h             | 150 ++++++++
+ 6 files changed, 2643 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/Makefile b/Makefile
-index 8a9113e666..74c2039005 100644
---- a/Makefile
-+++ b/Makefile
-@@ -479,7 +479,7 @@ dummy := $(call unnest-vars,, \
+diff --git a/block/meson.build b/block/meson.build
+index c10d544864..feffbc8623 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -42,6 +42,8 @@ block_ss.add(files(
+   'zeroinit.c',
+ ), zstd, zlib)
  
- include $(SRC_PATH)/tests/Makefile.include
++block_ss.add(files('../vma-writer.c'), libuuid)
++
+ softmmu_ss.add(when: 'CONFIG_TCG', if_true: files('blkreplay.c'))
  
--all: $(DOCS) $(if $(BUILD_DOCS),sphinxdocs) $(TOOLS) $(HELPERS-y) recurse-all modules $(vhost-user-json-y)
-+all: $(DOCS) $(if $(BUILD_DOCS),sphinxdocs) $(TOOLS) vma$(EXESUF) $(HELPERS-y) recurse-all modules $(vhost-user-json-y)
+ block_ss.add(when: 'CONFIG_QCOW1', if_true: files('qcow.c'))
+diff --git a/meson.build b/meson.build
+index e3386196ba..d5b660516b 100644
+--- a/meson.build
++++ b/meson.build
+@@ -725,6 +725,8 @@ keyutils = dependency('libkeyutils', required: false,
  
- qemu-version.h: FORCE
- 	$(call quiet-command, \
-@@ -608,6 +608,7 @@ qemu-img$(EXESUF): qemu-img.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io
- qemu-nbd$(EXESUF): qemu-nbd.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-obj-y) $(qom-obj-y) $(COMMON_LDADDS)
- qemu-io$(EXESUF): qemu-io.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-obj-y) $(qom-obj-y) $(COMMON_LDADDS)
- qemu-storage-daemon$(EXESUF): qemu-storage-daemon.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(chardev-obj-y) $(io-obj-y) $(qom-obj-y) $(storage-daemon-obj-y) $(COMMON_LDADDS)
-+vma$(EXESUF): vma.o vma-reader.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-obj-y) $(qom-obj-y) $(COMMON_LDADDS)
+ has_gettid = cc.has_function('gettid')
  
- qemu-bridge-helper$(EXESUF): qemu-bridge-helper.o $(COMMON_LDADDS)
++libuuid = cc.find_library('uuid', required: true)
++
+ # Malloc tests
  
-diff --git a/Makefile.objs b/Makefile.objs
-index d0b4dde836..05031a3da7 100644
---- a/Makefile.objs
-+++ b/Makefile.objs
-@@ -18,6 +18,7 @@ block-obj-y += block.o blockjob.o job.o
- block-obj-y += block/ scsi/
- block-obj-y += qemu-io-cmds.o
- block-obj-$(CONFIG_REPLICATION) += replication.o
-+block-obj-y += vma-writer.o
- 
- block-obj-m = block/
+ malloc = []
+@@ -1907,6 +1909,9 @@ if have_tools
+   qemu_nbd = executable('qemu-nbd', files('qemu-nbd.c'),
+                dependencies: [blockdev, qemuutil], install: true)
  
++  vma = executable('vma', files('vma.c', 'vma-reader.c'),
++                   dependencies: [authz, block, crypto, io, qom], install: true)
++
+   subdir('storage-daemon')
+   subdir('contrib/rdmacm-mux')
+   subdir('contrib/elf2dmp')
 diff --git a/vma-reader.c b/vma-reader.c
 new file mode 100644
 index 0000000000..2b1d1cdab3
@@ -914,10 +917,10 @@ index 0000000000..2b1d1cdab3
 +
 diff --git a/vma-writer.c b/vma-writer.c
 new file mode 100644
-index 0000000000..fe86b18a60
+index 0000000000..11d8321ffd
 --- /dev/null
 +++ b/vma-writer.c
-@@ -0,0 +1,771 @@
+@@ -0,0 +1,790 @@
 +/*
 + * VMA: Virtual Machine Archive
 + *
@@ -1213,20 +1216,20 @@ index 0000000000..fe86b18a60
 +
 +        if ((stat(filename, &st) == 0) && S_ISFIFO(st.st_mode)) {
 +            oflags = O_NONBLOCK|O_WRONLY;
-+            vmaw->fd = qemu_open(filename, oflags, 0644);
++            vmaw->fd = qemu_open(filename, oflags, errp);
 +        } else if (strstart(filename, "/dev/fdset/", &tmp_id_str)) {
 +            oflags = O_NONBLOCK|O_WRONLY;
-+            vmaw->fd = qemu_open(filename, oflags, 0644);
++            vmaw->fd = qemu_open(filename, oflags, errp);
 +        } else if (strstart(filename, "/dev/fdname/", &tmp_id_str)) {
-+            vmaw->fd = monitor_get_fd(cur_mon, tmp_id_str, errp);
++            vmaw->fd = monitor_get_fd(monitor_cur(), tmp_id_str, errp);
 +            if (vmaw->fd < 0) {
 +                goto err;
 +            }
 +            /* try to use O_NONBLOCK */
 +            fcntl(vmaw->fd, F_SETFL, fcntl(vmaw->fd, F_GETFL)|O_NONBLOCK);
 +        } else  {
-+            oflags = O_NONBLOCK|O_DIRECT|O_WRONLY|O_CREAT|O_EXCL;
-+            vmaw->fd = qemu_open(filename, oflags, 0644);
++            oflags = O_NONBLOCK|O_DIRECT|O_WRONLY|O_EXCL;
++            vmaw->fd = qemu_create(filename, oflags, 0644, errp);
 +        }
 +
 +        if (vmaw->fd < 0) {
@@ -1553,17 +1556,33 @@ index 0000000000..fe86b18a60
 +
 +    DPRINTF("VMA WRITE %d %zd\n", dev_id, cluster_num);
 +
++    uint64_t dev_size = vmaw->stream_info[dev_id].size;
 +    uint16_t mask = 0;
 +
 +    if (buf) {
 +        int i;
 +        int bit = 1;
++        uint64_t byte_offset = cluster_num * VMA_CLUSTER_SIZE;
 +        for (i = 0; i < 16; i++) {
 +            const unsigned char *vmablock = buf + (i*VMA_BLOCK_SIZE);
-+            if (!buffer_is_zero(vmablock, VMA_BLOCK_SIZE)) {
++
++            // Note: If the source is not 64k-aligned, we might reach 4k blocks
++            // after the end of the device. Always mark these as zero in the
++            // mask, so the restore handles them correctly.
++            if (byte_offset < dev_size &&
++                !buffer_is_zero(vmablock, VMA_BLOCK_SIZE))
++            {
 +                mask |= bit;
 +                memcpy(vmaw->outbuf + vmaw->outbuf_pos, vmablock,
 +                       VMA_BLOCK_SIZE);
++
++                // prevent memory leakage on unaligned last block
++                if (byte_offset + VMA_BLOCK_SIZE > dev_size) {
++                    uint64_t real_data_in_block = dev_size - byte_offset;
++                    memset(vmaw->outbuf + vmaw->outbuf_pos + real_data_in_block,
++                           0, VMA_BLOCK_SIZE - real_data_in_block);
++                }
++
 +                vmaw->outbuf_pos += VMA_BLOCK_SIZE;
 +            } else {
 +                DPRINTF("VMA WRITE %zd ZERO BLOCK %d\n", cluster_num, i);
@@ -1571,6 +1590,7 @@ index 0000000000..fe86b18a60
 +                *zero_bytes += VMA_BLOCK_SIZE;
 +            }
 +
++            byte_offset += VMA_BLOCK_SIZE;
 +            bit = bit << 1;
 +        }
 +    } else {
@@ -1596,8 +1616,8 @@ index 0000000000..fe86b18a60
 +
 +    if (dev_id != vmaw->vmstate_stream) {
 +        uint64_t last = (cluster_num + 1) * VMA_CLUSTER_SIZE;
-+        if (last > vmaw->stream_info[dev_id].size) {
-+            uint64_t diff = last - vmaw->stream_info[dev_id].size;
++        if (last > dev_size) {
++            uint64_t diff = last - dev_size;
 +            if (diff >= VMA_CLUSTER_SIZE) {
 +                vma_writer_set_error(vmaw, "vma_writer_write: "
 +                                     "read after last cluster");
@@ -1687,14 +1707,16 @@ index 0000000000..fe86b18a60
 +        g_checksum_free(vmaw->md5csum);
 +    }
 +
++    qemu_vfree(vmaw->headerbuf);
++    qemu_vfree(vmaw->outbuf);
 +    g_free(vmaw);
 +}
 diff --git a/vma.c b/vma.c
 new file mode 100644
-index 0000000000..a82752448a
+index 0000000000..2eea2fc281
 --- /dev/null
 +++ b/vma.c
-@@ -0,0 +1,837 @@
+@@ -0,0 +1,839 @@
 +/*
 + * VMA: Virtual Machine Archive
 + *
@@ -2262,6 +2284,7 @@ index 0000000000..a82752448a
 +            g_warning("vma_writer_close failed %s", error_get_pretty(err));
 +        }
 +    }
++    qemu_vfree(buf);
 +}
 +
 +static int create_archive(int argc, char **argv)
@@ -2429,6 +2452,7 @@ index 0000000000..a82752448a
 +        g_error("creating vma archive failed");
 +    }
 +
++    vma_writer_destroy(vmaw);
 +    return 0;
 +}
 +

debian/patches/pve/0027-PVE-Backup-add-backup-dump-block-driver.patch

@@ -8,32 +8,20 @@ Subject: [PATCH] PVE-Backup: add backup-dump block driver
 - block/backup.c - backup-job-create: also consider source cluster size
 - job.c: make job_should_pause non-static
 ---
- block/Makefile.objs       |   1 +
- block/backup-dump.c       | 169 ++++++++++++++++++++++++++++++++++++++
+ block/backup-dump.c       | 168 ++++++++++++++++++++++++++++++++++++++
  block/backup.c            |  23 ++----
+ block/meson.build         |   1 +
  include/block/block_int.h |  30 +++++++
  job.c                     |   3 +-
- 5 files changed, 207 insertions(+), 19 deletions(-)
+ 5 files changed, 206 insertions(+), 19 deletions(-)
  create mode 100644 block/backup-dump.c
 
-diff --git a/block/Makefile.objs b/block/Makefile.objs
-index 1282445672..8af7073c83 100644
---- a/block/Makefile.objs
-+++ b/block/Makefile.objs
-@@ -34,6 +34,7 @@ block-obj-$(CONFIG_RBD) += rbd.o
- block-obj-$(CONFIG_GLUSTERFS) += gluster.o
- block-obj-$(CONFIG_VXHS) += vxhs.o
- block-obj-$(CONFIG_LIBSSH) += ssh.o
-+block-obj-y += backup-dump.o
- block-obj-y += accounting.o dirty-bitmap.o
- block-obj-y += write-threshold.o
- block-obj-y += backup.o
 diff --git a/block/backup-dump.c b/block/backup-dump.c
 new file mode 100644
-index 0000000000..3066ab0698
+index 0000000000..93d7f46950
 --- /dev/null
 +++ b/block/backup-dump.c
-@@ -0,0 +1,169 @@
+@@ -0,0 +1,168 @@
 +/*
 + * BlockDriver to send backup data stream to a callback function
 + *
@@ -61,7 +49,6 @@ index 0000000000..3066ab0698
 +    BDRVBackupDumpState *s = bs->opaque;
 +
 +    bdi->cluster_size = s->dump_cb_block_size;
-+    bdi->unallocated_blocks_are_zero = true;
 +    return 0;
 +}
 +
@@ -142,7 +129,7 @@ index 0000000000..3066ab0698
 +static void qemu_backup_dump_child_perm(
 +    BlockDriverState *bs,
 +    BdrvChild *c,
-+    const BdrvChildRole *role,
++    BdrvChildRole role,
 +    BlockReopenQueue *reopen_queue,
 +    uint64_t perm, uint64_t shared,
 +    uint64_t *nperm, uint64_t *nshared)
@@ -204,7 +191,7 @@ index 0000000000..3066ab0698
 +    return bs;
 +}
 diff --git a/block/backup.c b/block/backup.c
-index ecd93e91e0..cf8f5ad25d 100644
+index 9afa0bf3b4..3df3d532d5 100644
 --- a/block/backup.c
 +++ b/block/backup.c
 @@ -32,24 +32,6 @@
@@ -232,7 +219,7 @@ index ecd93e91e0..cf8f5ad25d 100644
  static const BlockJobDriver backup_job_driver;
  
  static void backup_progress_bytes_callback(int64_t bytes, void *opaque)
-@@ -411,6 +393,11 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
+@@ -423,6 +405,11 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
          goto error;
      }
  
@@ -244,11 +231,23 @@ index ecd93e91e0..cf8f5ad25d 100644
      /*
       * If source is in backing chain of target assume that target is going to be
       * used for "image fleecing", i.e. it should represent a kind of snapshot of
+diff --git a/block/meson.build b/block/meson.build
+index feffbc8623..2507af1168 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -4,6 +4,7 @@ block_ss.add(files(
+   'aio_task.c',
+   'amend.c',
+   'backup.c',
++  'backup-dump.c',
+   'backup-top.c',
+   'blkdebug.c',
+   'blklogwrites.c',
 diff --git a/include/block/block_int.h b/include/block/block_int.h
-index 336f71e69d..62e5579723 100644
+index 6f8eda629a..5455102da8 100644
 --- a/include/block/block_int.h
 +++ b/include/block/block_int.h
-@@ -60,6 +60,36 @@
+@@ -63,6 +63,36 @@
  
  #define BLOCK_PROBE_BUF_SIZE        512
  
@@ -286,7 +285,7 @@ index 336f71e69d..62e5579723 100644
      BDRV_TRACKED_READ,
      BDRV_TRACKED_WRITE,
 diff --git a/job.c b/job.c
-index e82253e041..bcbbb0be02 100644
+index 8fecf38960..f9884e7d9d 100644
 --- a/job.c
 +++ b/job.c
 @@ -269,7 +269,8 @@ static bool job_started(Job *job)

debian/patches/pve/0027-PVE-Backup-modify-job-api.patch

@@ -1,92 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Wolfgang Bumiller <w.bumiller@proxmox.com>
-Date: Mon, 6 Apr 2020 12:16:56 +0200
-Subject: [PATCH] PVE-Backup: modify job api
-
-Introduce a pause_count parameter to start a backup in
-paused mode. This way backups of multiple drives can be
-started up sequentially via the completion callback while
-having been started at the same point in time.
-
-Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
----
- block/backup.c            | 3 +++
- block/replication.c       | 2 +-
- blockdev.c                | 3 ++-
- include/block/block_int.h | 1 +
- job.c                     | 2 +-
- 5 files changed, 8 insertions(+), 3 deletions(-)
-
-diff --git a/block/backup.c b/block/backup.c
-index a7a7dcaf4c..ecd93e91e0 100644
---- a/block/backup.c
-+++ b/block/backup.c
-@@ -338,6 +338,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
-                   BlockdevOnError on_target_error,
-                   int creation_flags,
-                   BlockCompletionFunc *cb, void *opaque,
-+                  int pause_count,
-                   JobTxn *txn, Error **errp)
- {
-     int64_t len;
-@@ -459,6 +460,8 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
-     block_job_add_bdrv(&job->common, "target", target, 0, BLK_PERM_ALL,
-                        &error_abort);
- 
-+    job->common.job.pause_count += pause_count;
-+
-     return &job->common;
- 
-  error:
-diff --git a/block/replication.c b/block/replication.c
-index da013c2041..17246a822c 100644
---- a/block/replication.c
-+++ b/block/replication.c
-@@ -554,7 +554,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
-                                 0, MIRROR_SYNC_MODE_NONE, NULL, 0, false, NULL,
-                                 BLOCKDEV_ON_ERROR_REPORT,
-                                 BLOCKDEV_ON_ERROR_REPORT, JOB_INTERNAL,
--                                backup_job_completed, bs, NULL, &local_err);
-+                                backup_job_completed, bs, 0, NULL, &local_err);
-         if (local_err) {
-             error_propagate(errp, local_err);
-             backup_job_cleanup(bs);
-diff --git a/blockdev.c b/blockdev.c
-index 5faddaa705..65c358e4ef 100644
---- a/blockdev.c
-+++ b/blockdev.c
-@@ -3114,7 +3114,8 @@ static BlockJob *do_backup_common(BackupCommon *backup,
-                             backup->filter_node_name,
-                             backup->on_source_error,
-                             backup->on_target_error,
--                            job_flags, NULL, NULL, txn, errp);
-+                            job_flags, NULL, NULL, 0, txn, errp);
-+
-     return job;
- }
- 
-diff --git a/include/block/block_int.h b/include/block/block_int.h
-index 4c3587ea19..336f71e69d 100644
---- a/include/block/block_int.h
-+++ b/include/block/block_int.h
-@@ -1219,6 +1219,7 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
-                             BlockdevOnError on_target_error,
-                             int creation_flags,
-                             BlockCompletionFunc *cb, void *opaque,
-+                            int pause_count,
-                             JobTxn *txn, Error **errp);
- 
- BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs,
-diff --git a/job.c b/job.c
-index 53be57a3a0..e82253e041 100644
---- a/job.c
-+++ b/job.c
-@@ -918,7 +918,7 @@ void job_start(Job *job)
-     job->co = qemu_coroutine_create(job_co_entry, job);
-     job->pause_count--;
-     job->busy = true;
--    job->paused = false;
-+    job->paused = job->pause_count > 0;
-     job_state_transition(job, JOB_STATUS_RUNNING);
-     aio_co_enter(job->aio_context, job->co);
- }

debian/patches/pve/0029-PVE-Backup-pbs-restore-new-command-to-restore-from-p.patch

@@ -4,40 +4,34 @@ Date: Mon, 6 Apr 2020 12:17:01 +0200
 Subject: [PATCH] PVE-Backup: pbs-restore - new command to restore from proxmox
  backup server
 
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
 ---
- Makefile      |   4 +-
- pbs-restore.c | 208 ++++++++++++++++++++++++++++++++++++++++++++++++++
- 2 files changed, 211 insertions(+), 1 deletion(-)
+ meson.build   |   4 +
+ pbs-restore.c | 224 ++++++++++++++++++++++++++++++++++++++++++++++++++
+ 2 files changed, 228 insertions(+)
  create mode 100644 pbs-restore.c
 
-diff --git a/Makefile b/Makefile
-index dbd9542ae4..7c1fb58e18 100644
---- a/Makefile
-+++ b/Makefile
-@@ -479,7 +479,7 @@ dummy := $(call unnest-vars,, \
- 
- include $(SRC_PATH)/tests/Makefile.include
- 
--all: $(DOCS) $(if $(BUILD_DOCS),sphinxdocs) $(TOOLS) vma$(EXESUF) $(HELPERS-y) recurse-all modules $(vhost-user-json-y)
-+all: $(DOCS) $(if $(BUILD_DOCS),sphinxdocs) $(TOOLS) vma$(EXESUF) pbs-restore$(EXESUF) $(HELPERS-y) recurse-all modules $(vhost-user-json-y)
- 
- qemu-version.h: FORCE
- 	$(call quiet-command, \
-@@ -610,6 +610,8 @@ qemu-io$(EXESUF): qemu-io.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-o
- qemu-storage-daemon$(EXESUF): qemu-storage-daemon.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(chardev-obj-y) $(io-obj-y) $(qom-obj-y) $(storage-daemon-obj-y) $(COMMON_LDADDS)
- qemu-storage-daemon$(EXESUF): LIBS += -lproxmox_backup_qemu
- vma$(EXESUF): vma.o vma-reader.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-obj-y) $(qom-obj-y) $(COMMON_LDADDS)
-+pbs-restore$(EXESUF): pbs-restore.o $(authz-obj-y) $(block-obj-y) $(crypto-obj-y) $(io-obj-y) $(qom-obj-y) $(COMMON_LDADDS)
-+pbs-restore$(EXESUF): LIBS += -lproxmox_backup_qemu
- 
- qemu-bridge-helper$(EXESUF): qemu-bridge-helper.o $(COMMON_LDADDS)
+diff --git a/meson.build b/meson.build
+index 3094f98c47..6f1fafee14 100644
+--- a/meson.build
++++ b/meson.build
+@@ -1913,6 +1913,10 @@ if have_tools
+   vma = executable('vma', files('vma.c', 'vma-reader.c'),
+                    dependencies: [authz, block, crypto, io, qom], install: true)
  
++  pbs_restore = executable('pbs-restore', files('pbs-restore.c'),
++                  dependencies: [authz, block, crypto, io, qom,
++                    libproxmox_backup_qemu], install: true)
++
+   subdir('storage-daemon')
+   subdir('contrib/rdmacm-mux')
+   subdir('contrib/elf2dmp')
 diff --git a/pbs-restore.c b/pbs-restore.c
 new file mode 100644
-index 0000000000..f65de8b890
+index 0000000000..4d3f925a1b
 --- /dev/null
 +++ b/pbs-restore.c
-@@ -0,0 +1,208 @@
+@@ -0,0 +1,224 @@
 +/*
 + * Qemu image restore helper for Proxmox Backup
 + *
@@ -124,7 +118,7 @@ index 0000000000..f65de8b890
 +
 +    error_init(argv[0]);
 +
-+    for(;;) {
++    for (;;) {
 +        static const struct option long_options[] = {
 +            {"help", no_argument, 0, 'h'},
 +            {"skip-zero", no_argument, 0, 'S'},
@@ -138,31 +132,31 @@ index 0000000000..f65de8b890
 +        if (c == -1) {
 +            break;
 +        }
-+        switch(c) {
-+        case ':':
-+            fprintf(stderr, "missing argument for option '%s'", argv[optind - 1]);
-+            return -1;
-+        case '?':
-+            fprintf(stderr, "unrecognized option '%s'", argv[optind - 1]);
-+            return -1;
-+        case 'f':
-+            format = g_strdup(argv[optind - 1]);
-+            break;
-+        case 'r':
-+            repository = g_strdup(argv[optind - 1]);
-+            break;
-+        case 'k':
-+            keyfile = g_strdup(argv[optind - 1]);
-+            break;
-+        case 'v':
-+            verbose = true;
-+            break;
-+        case 'S':
-+            skip_zero = true;
-+            break;
-+        case 'h':
-+            help();
-+            return 0;
++        switch (c) {
++            case ':':
++                fprintf(stderr, "missing argument for option '%s'\n", argv[optind - 1]);
++                return -1;
++            case '?':
++                fprintf(stderr, "unrecognized option '%s'\n", argv[optind - 1]);
++                return -1;
++            case 'f':
++                format = g_strdup(argv[optind - 1]);
++                break;
++            case 'r':
++                repository = g_strdup(argv[optind - 1]);
++                break;
++            case 'k':
++                keyfile = g_strdup(argv[optind - 1]);
++                break;
++            case 'v':
++                verbose = true;
++                break;
++            case 'S':
++                skip_zero = true;
++                break;
++            case 'h':
++                help();
++                return 0;
 +        }
 +    }
 +
@@ -197,31 +191,42 @@ index 0000000000..f65de8b890
 +    bdrv_init();
 +    module_call_init(MODULE_INIT_QOM);
 +
++    if (verbose) {
++        fprintf(stderr, "connecting to repository '%s'\n", repository);
++    }
 +    char *pbs_error = NULL;
-+    ProxmoxRestoreHandle *conn = proxmox_restore_connect(
++    ProxmoxRestoreHandle *conn = proxmox_restore_new(
 +        repository, snapshot, password, keyfile, key_password, fingerprint, &pbs_error);
 +    if (conn == NULL) {
 +        fprintf(stderr, "restore failed: %s\n", pbs_error);
 +        return -1;
 +    }
 +
++    int res = proxmox_restore_connect(conn, &pbs_error);
++    if (res < 0 || pbs_error) {
++        fprintf(stderr, "restore failed (connection error): %s\n", pbs_error);
++        return -1;
++    }
++
 +    QDict *options = qdict_new();
-+    qdict_put_str(options, "driver", format);
 +
 +    if (format) {
 +        qdict_put_str(options, "driver", format);
 +    }
 +
++
++    if (verbose) {
++        fprintf(stderr, "open block backend for target '%s'\n", target);
++    }
 +    Error *local_err = NULL;
 +    int flags = BDRV_O_RDWR;
-+
 +    BlockBackend *blk = blk_new_open(target, NULL, options, flags, &local_err);
 +    if (!blk) {
 +        fprintf(stderr, "%s\n", error_get_pretty(local_err));
 +        return -1;
 +    }
 +
-+    CallbackData *callback_data = calloc(sizeof( CallbackData), 1);
++    CallbackData *callback_data = calloc(sizeof(CallbackData), 1);
 +
 +    callback_data->target = blk;
 +    callback_data->skip_zero = skip_zero;
@@ -229,7 +234,11 @@ index 0000000000..f65de8b890
 +
 +    // blk_set_enable_write_cache(blk, !writethrough);
 +
-+    int res = proxmox_restore_image(
++    if (verbose) {
++        fprintf(stderr, "starting to restore snapshot '%s'\n", snapshot);
++        fflush(stderr); // ensure we do not get printed after the progress log
++    }
++    res = proxmox_restore_image(
 +        conn,
 +        archive_name,
 +        write_callback,
@@ -238,6 +247,7 @@ index 0000000000..f65de8b890
 +        verbose);
 +
 +    proxmox_restore_disconnect(conn);
++    blk_unref(blk);
 +
 +    if (res < 0) {
 +        fprintf(stderr, "restore failed: %s\n", pbs_error);

debian/patches/pve/0030-PVE-Backup-Add-dirty-bitmap-tracking-for-incremental.patch

@@ -0,0 +1,452 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Mon, 29 Jun 2020 11:06:03 +0200
+Subject: [PATCH] PVE-Backup: Add dirty-bitmap tracking for incremental backups
+
+Uses QEMU's existing MIRROR_SYNC_MODE_BITMAP and a dirty-bitmap on top
+of all backed-up drives. This will only execute the data-write callback
+for any changed chunks, the PBS rust code will reuse chunks from the
+previous index for everything it doesn't receive if reuse_index is true.
+
+On error or cancellation, remove all dirty bitmaps to ensure
+consistency.
+
+Add PBS/incremental specific information to query backup info QMP and
+HMP commands.
+
+Only supported for PBS backups.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+---
+ block/monitor/block-hmp-cmds.c |   1 +
+ monitor/hmp-cmds.c             |  45 ++++++++++----
+ proxmox-backup-client.c        |   3 +-
+ proxmox-backup-client.h        |   1 +
+ pve-backup.c                   | 103 ++++++++++++++++++++++++++++++---
+ qapi/block-core.json           |  12 +++-
+ 6 files changed, 142 insertions(+), 23 deletions(-)
+
+diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
+index 9ba7c774a2..056d14deee 100644
+--- a/block/monitor/block-hmp-cmds.c
++++ b/block/monitor/block-hmp-cmds.c
+@@ -1039,6 +1039,7 @@ void hmp_backup(Monitor *mon, const QDict *qdict)
+         false, NULL, // PBS fingerprint
+         false, NULL, // PBS backup-id
+         false, 0, // PBS backup-time
++        false, false, // PBS incremental
+         true, dir ? BACKUP_FORMAT_DIR : BACKUP_FORMAT_VMA,
+         false, NULL, false, NULL, !!devlist,
+         devlist, qdict_haskey(qdict, "speed"), speed, &error);
+diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
+index 182e79c943..604026bb37 100644
+--- a/monitor/hmp-cmds.c
++++ b/monitor/hmp-cmds.c
+@@ -221,19 +221,42 @@ void hmp_info_backup(Monitor *mon, const QDict *qdict)
+             monitor_printf(mon, "End time: %s", ctime(&info->end_time));
+         }
+ 
+-        int per = (info->has_total && info->total &&
+-            info->has_transferred && info->transferred) ?
+-            (info->transferred * 100)/info->total : 0;
+-        int zero_per = (info->has_total && info->total &&
+-                        info->has_zero_bytes && info->zero_bytes) ?
+-            (info->zero_bytes * 100)/info->total : 0;
+         monitor_printf(mon, "Backup file: %s\n", info->backup_file);
+         monitor_printf(mon, "Backup uuid: %s\n", info->uuid);
+-        monitor_printf(mon, "Total size: %zd\n", info->total);
+-        monitor_printf(mon, "Transferred bytes: %zd (%d%%)\n",
+-                       info->transferred, per);
+-        monitor_printf(mon, "Zero bytes: %zd (%d%%)\n",
+-                       info->zero_bytes, zero_per);
++
++        if (!(info->has_total && info->total))  {
++            // this should not happen normally
++            monitor_printf(mon, "Total size: %d\n", 0);
++        } else {
++            bool incremental = false;
++            size_t total_or_dirty = info->total;
++            if (info->has_transferred) {
++                if (info->has_dirty && info->dirty) {
++                     if (info->dirty < info->total) {
++                        total_or_dirty = info->dirty;
++                        incremental = true;
++                    }
++                }
++            }
++
++            int per = (info->transferred * 100)/total_or_dirty;
++
++            monitor_printf(mon, "Backup mode: %s\n", incremental ? "incremental" : "full");
++
++            int zero_per = (info->has_zero_bytes && info->zero_bytes) ?
++                (info->zero_bytes * 100)/info->total : 0;
++            monitor_printf(mon, "Total size: %zd\n", info->total);
++            monitor_printf(mon, "Transferred bytes: %zd (%d%%)\n",
++                           info->transferred, per);
++            monitor_printf(mon, "Zero bytes: %zd (%d%%)\n",
++                           info->zero_bytes, zero_per);
++
++            if (info->has_reused) {
++                int reused_per = (info->reused * 100)/total_or_dirty;
++                monitor_printf(mon, "Reused bytes: %zd (%d%%)\n",
++                               info->reused, reused_per);
++            }
++        }
+     }
+ 
+     qapi_free_BackupStatus(info);
+diff --git a/proxmox-backup-client.c b/proxmox-backup-client.c
+index a8f6653a81..4ce7bc0b5e 100644
+--- a/proxmox-backup-client.c
++++ b/proxmox-backup-client.c
+@@ -89,6 +89,7 @@ proxmox_backup_co_register_image(
+     ProxmoxBackupHandle *pbs,
+     const char *device_name,
+     uint64_t size,
++    bool incremental,
+     Error **errp)
+ {
+     Coroutine *co = qemu_coroutine_self();
+@@ -98,7 +99,7 @@ proxmox_backup_co_register_image(
+     int pbs_res = -1;
+ 
+     proxmox_backup_register_image_async(
+-        pbs, device_name, size ,proxmox_backup_schedule_wake, &waker, &pbs_res, &pbs_err);
++        pbs, device_name, size, incremental, proxmox_backup_schedule_wake, &waker, &pbs_res, &pbs_err);
+     qemu_coroutine_yield();
+     if (pbs_res < 0) {
+         if (errp) error_setg(errp, "backup register image failed: %s", pbs_err ? pbs_err : "unknown error");
+diff --git a/proxmox-backup-client.h b/proxmox-backup-client.h
+index 1dda8b7d8f..8cbf645b2c 100644
+--- a/proxmox-backup-client.h
++++ b/proxmox-backup-client.h
+@@ -32,6 +32,7 @@ proxmox_backup_co_register_image(
+     ProxmoxBackupHandle *pbs,
+     const char *device_name,
+     uint64_t size,
++    bool incremental,
+     Error **errp);
+ 
+ 
+diff --git a/pve-backup.c b/pve-backup.c
+index d40f3f2fd6..1cd9d31d7c 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -28,6 +28,8 @@
+  *
+  */
+ 
++const char *PBS_BITMAP_NAME = "pbs-incremental-dirty-bitmap";
++
+ static struct PVEBackupState {
+     struct {
+         // Everithing accessed from qmp_backup_query command is protected using lock
+@@ -39,7 +41,9 @@ static struct PVEBackupState {
+         uuid_t uuid;
+         char uuid_str[37];
+         size_t total;
++        size_t dirty;
+         size_t transferred;
++        size_t reused;
+         size_t zero_bytes;
+     } stat;
+     int64_t speed;
+@@ -66,6 +70,7 @@ typedef struct PVEBackupDevInfo {
+     uint8_t dev_id;
+     bool completed;
+     char targetfile[PATH_MAX];
++    BdrvDirtyBitmap *bitmap;
+     BlockDriverState *target;
+ } PVEBackupDevInfo;
+ 
+@@ -105,11 +110,12 @@ static bool pvebackup_error_or_canceled(void)
+     return error_or_canceled;
+ }
+ 
+-static void pvebackup_add_transfered_bytes(size_t transferred, size_t zero_bytes)
++static void pvebackup_add_transfered_bytes(size_t transferred, size_t zero_bytes, size_t reused)
+ {
+     qemu_mutex_lock(&backup_state.stat.lock);
+     backup_state.stat.zero_bytes += zero_bytes;
+     backup_state.stat.transferred += transferred;
++    backup_state.stat.reused += reused;
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ }
+ 
+@@ -148,7 +154,8 @@ pvebackup_co_dump_pbs_cb(
+         pvebackup_propagate_error(local_err);
+         return pbs_res;
+     } else {
+-        pvebackup_add_transfered_bytes(size, !buf ? size : 0);
++        size_t reused = (pbs_res == 0) ? size : 0;
++        pvebackup_add_transfered_bytes(size, !buf ? size : 0, reused);
+     }
+ 
+     return size;
+@@ -208,11 +215,11 @@ pvebackup_co_dump_vma_cb(
+         } else {
+             if (remaining >= VMA_CLUSTER_SIZE) {
+                 assert(ret == VMA_CLUSTER_SIZE);
+-                pvebackup_add_transfered_bytes(VMA_CLUSTER_SIZE, zero_bytes);
++                pvebackup_add_transfered_bytes(VMA_CLUSTER_SIZE, zero_bytes, 0);
+                 remaining -= VMA_CLUSTER_SIZE;
+             } else {
+                 assert(ret == remaining);
+-                pvebackup_add_transfered_bytes(remaining, zero_bytes);
++                pvebackup_add_transfered_bytes(remaining, zero_bytes, 0);
+                 remaining = 0;
+             }
+         }
+@@ -248,6 +255,18 @@ static void coroutine_fn pvebackup_co_cleanup(void *unused)
+             if (local_err != NULL) {
+                 pvebackup_propagate_error(local_err);
+             }
++        } else {
++            // on error or cancel we cannot ensure synchronization of dirty
++            // bitmaps with backup server, so remove all and do full backup next
++            GList *l = backup_state.di_list;
++            while (l) {
++                PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
++                l = g_list_next(l);
++
++                if (di->bitmap) {
++                    bdrv_release_dirty_bitmap(di->bitmap);
++                }
++            }
+         }
+ 
+         proxmox_backup_disconnect(backup_state.pbs);
+@@ -303,6 +322,12 @@ static void pvebackup_complete_cb(void *opaque, int ret)
+     // remove self from job queue
+     backup_state.di_list = g_list_remove(backup_state.di_list, di);
+ 
++    if (di->bitmap && ret < 0) {
++        // on error or cancel we cannot ensure synchronization of dirty
++        // bitmaps with backup server, so remove all and do full backup next
++        bdrv_release_dirty_bitmap(di->bitmap);
++    }
++
+     g_free(di);
+ 
+     qemu_mutex_unlock(&backup_state.backup_mutex);
+@@ -470,12 +495,18 @@ static bool create_backup_jobs(void) {
+ 
+         assert(di->target != NULL);
+ 
++        MirrorSyncMode sync_mode = MIRROR_SYNC_MODE_FULL;
++        BitmapSyncMode bitmap_mode = BITMAP_SYNC_MODE_NEVER;
++        if (di->bitmap) {
++            sync_mode = MIRROR_SYNC_MODE_BITMAP;
++            bitmap_mode = BITMAP_SYNC_MODE_ON_SUCCESS;
++        }
+         AioContext *aio_context = bdrv_get_aio_context(di->bs);
+         aio_context_acquire(aio_context);
+ 
+         BlockJob *job = backup_job_create(
+-            NULL, di->bs, di->target, backup_state.speed, MIRROR_SYNC_MODE_FULL, NULL,
+-            BITMAP_SYNC_MODE_NEVER, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
++            NULL, di->bs, di->target, backup_state.speed, sync_mode, di->bitmap,
++            bitmap_mode, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
+             JOB_DEFAULT, pvebackup_complete_cb, di, 1, NULL, &local_err);
+ 
+         aio_context_release(aio_context);
+@@ -526,6 +557,8 @@ typedef struct QmpBackupTask {
+     const char *fingerprint;
+     bool has_fingerprint;
+     int64_t backup_time;
++    bool has_use_dirty_bitmap;
++    bool use_dirty_bitmap;
+     bool has_format;
+     BackupFormat format;
+     bool has_config_file;
+@@ -617,6 +650,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     }
+ 
+     size_t total = 0;
++    size_t dirty = 0;
+ 
+     l = di_list;
+     while (l) {
+@@ -654,6 +688,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+         int dump_cb_block_size = PROXMOX_BACKUP_DEFAULT_CHUNK_SIZE; // Hardcoded (4M)
+         firewall_name = "fw.conf";
+ 
++        bool use_dirty_bitmap = task->has_use_dirty_bitmap && task->use_dirty_bitmap;
++
+         char *pbs_err = NULL;
+         pbs = proxmox_backup_new(
+             task->backup_file,
+@@ -673,7 +709,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             goto err;
+         }
+ 
+-        if (proxmox_backup_co_connect(pbs, task->errp) < 0)
++        int connect_result = proxmox_backup_co_connect(pbs, task->errp);
++        if (connect_result < 0)
+             goto err;
+ 
+         /* register all devices */
+@@ -684,9 +721,40 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+ 
+             const char *devname = bdrv_get_device_name(di->bs);
+ 
+-            int dev_id = proxmox_backup_co_register_image(pbs, devname, di->size, task->errp);
+-            if (dev_id < 0)
++            BdrvDirtyBitmap *bitmap = bdrv_find_dirty_bitmap(di->bs, PBS_BITMAP_NAME);
++            bool expect_only_dirty = false;
++
++            if (use_dirty_bitmap) {
++                if (bitmap == NULL) {
++                    bitmap = bdrv_create_dirty_bitmap(di->bs, dump_cb_block_size, PBS_BITMAP_NAME, task->errp);
++                    if (!bitmap) {
++                        goto err;
++                    }
++                } else {
++                    expect_only_dirty = proxmox_backup_check_incremental(pbs, devname, di->size) != 0;
++                }
++
++                if (expect_only_dirty) {
++                    dirty += bdrv_get_dirty_count(bitmap);
++                } else {
++                    /* mark entire bitmap as dirty to make full backup */
++                    bdrv_set_dirty_bitmap(bitmap, 0, di->size);
++                    dirty += di->size;
++                }
++                di->bitmap = bitmap;
++            } else {
++                dirty += di->size;
++
++                /* after a full backup the old dirty bitmap is invalid anyway */
++                if (bitmap != NULL) {
++                    bdrv_release_dirty_bitmap(bitmap);
++                }
++            }
++
++            int dev_id = proxmox_backup_co_register_image(pbs, devname, di->size, expect_only_dirty, task->errp);
++            if (dev_id < 0) {
+                 goto err;
++            }
+ 
+             if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_pbs_cb, di, task->errp))) {
+                 goto err;
+@@ -695,6 +763,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             di->dev_id = dev_id;
+         }
+     } else if (format == BACKUP_FORMAT_VMA) {
++        dirty = total;
++
+         vmaw = vma_writer_create(task->backup_file, uuid, &local_err);
+         if (!vmaw) {
+             if (local_err) {
+@@ -722,6 +792,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             }
+         }
+     } else if (format == BACKUP_FORMAT_DIR) {
++        dirty = total;
++
+         if (mkdir(task->backup_file, 0640) != 0) {
+             error_setg_errno(task->errp, errno, "can't create directory '%s'\n",
+                              task->backup_file);
+@@ -794,8 +866,10 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     char *uuid_str = g_strdup(backup_state.stat.uuid_str);
+ 
+     backup_state.stat.total = total;
++    backup_state.stat.dirty = dirty;
+     backup_state.stat.transferred = 0;
+     backup_state.stat.zero_bytes = 0;
++    backup_state.stat.reused = format == BACKUP_FORMAT_PBS && dirty >= total ? 0 : total - dirty;
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+@@ -819,6 +893,10 @@ err:
+         PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+         l = g_list_next(l);
+ 
++        if (di->bitmap) {
++            bdrv_release_dirty_bitmap(di->bitmap);
++        }
++
+         if (di->target) {
+             bdrv_unref(di->target);
+         }
+@@ -860,6 +938,7 @@ UuidInfo *qmp_backup(
+     bool has_fingerprint, const char *fingerprint,
+     bool has_backup_id, const char *backup_id,
+     bool has_backup_time, int64_t backup_time,
++    bool has_use_dirty_bitmap, bool use_dirty_bitmap,
+     bool has_format, BackupFormat format,
+     bool has_config_file, const char *config_file,
+     bool has_firewall_file, const char *firewall_file,
+@@ -878,6 +957,8 @@ UuidInfo *qmp_backup(
+         .backup_id = backup_id,
+         .has_backup_time = has_backup_time,
+         .backup_time = backup_time,
++        .has_use_dirty_bitmap = has_use_dirty_bitmap,
++        .use_dirty_bitmap = use_dirty_bitmap,
+         .has_format = has_format,
+         .format = format,
+         .has_config_file = has_config_file,
+@@ -946,10 +1027,14 @@ BackupStatus *qmp_query_backup(Error **errp)
+ 
+     info->has_total = true;
+     info->total = backup_state.stat.total;
++    info->has_dirty = true;
++    info->dirty = backup_state.stat.dirty;
+     info->has_zero_bytes = true;
+     info->zero_bytes = backup_state.stat.zero_bytes;
+     info->has_transferred = true;
+     info->transferred = backup_state.stat.transferred;
++    info->has_reused = true;
++    info->reused = backup_state.stat.reused;
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 9054db608c..d4e1c98c50 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -758,8 +758,13 @@
+ #
+ # @total: total amount of bytes involved in the backup process
+ #
++# @dirty: with incremental mode (PBS) this is the amount of bytes involved
++#         in the backup process which are marked dirty.
++#
+ # @transferred: amount of bytes already backed up.
+ #
++# @reused: amount of bytes reused due to deduplication.
++#
+ # @zero-bytes: amount of 'zero' bytes detected.
+ #
+ # @start-time: time (epoch) when backup job started.
+@@ -772,8 +777,8 @@
+ #
+ ##
+ { 'struct': 'BackupStatus',
+-  'data': {'*status': 'str', '*errmsg': 'str', '*total': 'int',
+-           '*transferred': 'int', '*zero-bytes': 'int',
++  'data': {'*status': 'str', '*errmsg': 'str', '*total': 'int', '*dirty': 'int',
++           '*transferred': 'int', '*zero-bytes': 'int', '*reused': 'int',
+            '*start-time': 'int', '*end-time': 'int',
+            '*backup-file': 'str', '*uuid': 'str' } }
+ 
+@@ -816,6 +821,8 @@
+ #
+ # @backup-time: backup timestamp (Unix epoch, required for format 'pbs')
+ #
++# @use-dirty-bitmap: use dirty bitmap to detect incremental changes since last job (optional for format 'pbs')
++#
+ # Returns: the uuid of the backup job
+ #
+ ##
+@@ -826,6 +833,7 @@
+                                     '*fingerprint': 'str',
+                                     '*backup-id': 'str',
+                                     '*backup-time': 'int',
++                                    '*use-dirty-bitmap': 'bool',
+                                     '*format': 'BackupFormat',
+                                     '*config-file': 'str',
+                                     '*firewall-file': 'str',

debian/patches/pve/0031-PVE-Backup-aquire-aio_context-before-calling-backup_.patch

@@ -1,39 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Dietmar Maurer <dietmar@proxmox.com>
-Date: Mon, 6 Apr 2020 12:17:00 +0200
-Subject: [PATCH] PVE-Backup: aquire aio_context before calling
- backup_job_create
-
-And do not set target in same aoi_context as source, because
-this is already done in bdrv_backup_top_append ...
-
-Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
----
- pve-backup.c | 8 +++-----
- 1 file changed, 3 insertions(+), 5 deletions(-)
-
-diff --git a/pve-backup.c b/pve-backup.c
-index 9ae89fb679..38dd33e28b 100644
---- a/pve-backup.c
-+++ b/pve-backup.c
-@@ -757,17 +757,15 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-         PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-         l = g_list_next(l);
- 
--        // make sure target runs in same aoi_context as source
-         AioContext *aio_context = bdrv_get_aio_context(di->bs);
-         aio_context_acquire(aio_context);
--        GSList *ignore = NULL;
--        bdrv_set_aio_context_ignore(di->target, aio_context, &ignore);
--        g_slist_free(ignore);
--        aio_context_release(aio_context);
- 
-         job = backup_job_create(NULL, di->bs, di->target, backup_state.speed, MIRROR_SYNC_MODE_FULL, NULL,
-                                 BITMAP_SYNC_MODE_NEVER, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
-                                 JOB_DEFAULT, pvebackup_complete_cb, di, 1, NULL, &local_err);
-+
-+        aio_context_release(aio_context);
-+
-         if (!job || local_err != NULL) {
-             qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
-             error_setg(&backup_state.stat.error, "backup_job_create failed");

debian/patches/pve/0031-PVE-various-PBS-fixes.patch

@@ -0,0 +1,218 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Dietmar Maurer <dietmar@proxmox.com>
+Date: Thu, 9 Jul 2020 12:53:08 +0200
+Subject: [PATCH] PVE: various PBS fixes
+
+pbs: fix crypt and compress parameters
+Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
+
+PVE: handle PBS write callback with big blocks correctly
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+
+PVE: add zero block handling to PBS dump callback
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/monitor/block-hmp-cmds.c |  4 ++-
+ pve-backup.c                   | 57 +++++++++++++++++++++++++++-------
+ qapi/block-core.json           |  6 ++++
+ 3 files changed, 54 insertions(+), 13 deletions(-)
+
+diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
+index 056d14deee..46c63b1cf9 100644
+--- a/block/monitor/block-hmp-cmds.c
++++ b/block/monitor/block-hmp-cmds.c
+@@ -1039,7 +1039,9 @@ void hmp_backup(Monitor *mon, const QDict *qdict)
+         false, NULL, // PBS fingerprint
+         false, NULL, // PBS backup-id
+         false, 0, // PBS backup-time
+-        false, false, // PBS incremental
++        false, false, // PBS use-dirty-bitmap
++        false, false, // PBS compress
++        false, false, // PBS encrypt
+         true, dir ? BACKUP_FORMAT_DIR : BACKUP_FORMAT_VMA,
+         false, NULL, false, NULL, !!devlist,
+         devlist, qdict_haskey(qdict, "speed"), speed, &error);
+diff --git a/pve-backup.c b/pve-backup.c
+index 1cd9d31d7c..b8182aaf89 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -8,6 +8,7 @@
+ #include "block/blockjob.h"
+ #include "qapi/qapi-commands-block.h"
+ #include "qapi/qmp/qerror.h"
++#include "qemu/cutils.h"
+ 
+ /* PVE backup state and related function */
+ 
+@@ -67,6 +68,7 @@ opts_init(pvebackup_init);
+ typedef struct PVEBackupDevInfo {
+     BlockDriverState *bs;
+     size_t size;
++    uint64_t block_size;
+     uint8_t dev_id;
+     bool completed;
+     char targetfile[PATH_MAX];
+@@ -135,10 +137,13 @@ pvebackup_co_dump_pbs_cb(
+     PVEBackupDevInfo *di = opaque;
+ 
+     assert(backup_state.pbs);
++    assert(buf);
+ 
+     Error *local_err = NULL;
+     int pbs_res = -1;
+ 
++    bool is_zero_block = size == di->block_size && buffer_is_zero(buf, size);
++
+     qemu_co_mutex_lock(&backup_state.dump_callback_mutex);
+ 
+     // avoid deadlock if job is cancelled
+@@ -147,17 +152,29 @@ pvebackup_co_dump_pbs_cb(
+         return -1;
+     }
+ 
+-    pbs_res = proxmox_backup_co_write_data(backup_state.pbs, di->dev_id, buf, start, size, &local_err);
+-    qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
++    uint64_t transferred = 0;
++    uint64_t reused = 0;
++    while (transferred < size) {
++        uint64_t left = size - transferred;
++        uint64_t to_transfer = left < di->block_size ? left : di->block_size;
+ 
+-    if (pbs_res < 0) {
+-        pvebackup_propagate_error(local_err);
+-        return pbs_res;
+-    } else {
+-        size_t reused = (pbs_res == 0) ? size : 0;
+-        pvebackup_add_transfered_bytes(size, !buf ? size : 0, reused);
++        pbs_res = proxmox_backup_co_write_data(backup_state.pbs, di->dev_id,
++            is_zero_block ? NULL : buf + transferred, start + transferred,
++            to_transfer, &local_err);
++        transferred += to_transfer;
++
++        if (pbs_res < 0) {
++            pvebackup_propagate_error(local_err);
++            qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
++            return pbs_res;
++        }
++
++        reused += pbs_res == 0 ? to_transfer : 0;
+     }
+ 
++    qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
++    pvebackup_add_transfered_bytes(size, is_zero_block ? size : 0, reused);
++
+     return size;
+ }
+ 
+@@ -178,6 +195,7 @@ pvebackup_co_dump_vma_cb(
+     int ret = -1;
+ 
+     assert(backup_state.vmaw);
++    assert(buf);
+ 
+     uint64_t remaining = size;
+ 
+@@ -204,9 +222,7 @@ pvebackup_co_dump_vma_cb(
+         qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
+ 
+         ++cluster_num;
+-        if (buf) {
+-            buf += VMA_CLUSTER_SIZE;
+-        }
++        buf += VMA_CLUSTER_SIZE;
+         if (ret < 0) {
+             Error *local_err = NULL;
+             vma_writer_error_propagate(backup_state.vmaw, &local_err);
+@@ -567,6 +583,10 @@ typedef struct QmpBackupTask {
+     const char *firewall_file;
+     bool has_devlist;
+     const char *devlist;
++    bool has_compress;
++    bool compress;
++    bool has_encrypt;
++    bool encrypt;
+     bool has_speed;
+     int64_t speed;
+     Error **errp;
+@@ -690,6 +710,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+ 
+         bool use_dirty_bitmap = task->has_use_dirty_bitmap && task->use_dirty_bitmap;
+ 
++
+         char *pbs_err = NULL;
+         pbs = proxmox_backup_new(
+             task->backup_file,
+@@ -699,8 +720,10 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             task->has_password ? task->password : NULL,
+             task->has_keyfile ? task->keyfile : NULL,
+             task->has_key_password ? task->key_password : NULL,
++            task->has_compress ? task->compress : true,
++            task->has_encrypt ? task->encrypt : task->has_keyfile,
+             task->has_fingerprint ? task->fingerprint : NULL,
+-            &pbs_err);
++             &pbs_err);
+ 
+         if (!pbs) {
+             error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
+@@ -719,6 +742,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+             l = g_list_next(l);
+ 
++            di->block_size = dump_cb_block_size;
++
+             const char *devname = bdrv_get_device_name(di->bs);
+ 
+             BdrvDirtyBitmap *bitmap = bdrv_find_dirty_bitmap(di->bs, PBS_BITMAP_NAME);
+@@ -939,6 +964,8 @@ UuidInfo *qmp_backup(
+     bool has_backup_id, const char *backup_id,
+     bool has_backup_time, int64_t backup_time,
+     bool has_use_dirty_bitmap, bool use_dirty_bitmap,
++    bool has_compress, bool compress,
++    bool has_encrypt, bool encrypt,
+     bool has_format, BackupFormat format,
+     bool has_config_file, const char *config_file,
+     bool has_firewall_file, const char *firewall_file,
+@@ -949,6 +976,8 @@ UuidInfo *qmp_backup(
+         .backup_file = backup_file,
+         .has_password = has_password,
+         .password = password,
++        .has_keyfile = has_keyfile,
++        .keyfile = keyfile,
+         .has_key_password = has_key_password,
+         .key_password = key_password,
+         .has_fingerprint = has_fingerprint,
+@@ -959,6 +988,10 @@ UuidInfo *qmp_backup(
+         .backup_time = backup_time,
+         .has_use_dirty_bitmap = has_use_dirty_bitmap,
+         .use_dirty_bitmap = use_dirty_bitmap,
++        .has_compress = has_compress,
++        .compress = compress,
++        .has_encrypt = has_encrypt,
++        .encrypt = encrypt,
+         .has_format = has_format,
+         .format = format,
+         .has_config_file = has_config_file,
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index d4e1c98c50..0fda1e3fd3 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -823,6 +823,10 @@
+ #
+ # @use-dirty-bitmap: use dirty bitmap to detect incremental changes since last job (optional for format 'pbs')
+ #
++# @compress: use compression (optional for format 'pbs', defaults to true)
++#
++# @encrypt: use encryption ((optional for format 'pbs', defaults to true if there is a keyfile)
++#
+ # Returns: the uuid of the backup job
+ #
+ ##
+@@ -834,6 +838,8 @@
+                                     '*backup-id': 'str',
+                                     '*backup-time': 'int',
+                                     '*use-dirty-bitmap': 'bool',
++                                    '*compress': 'bool',
++                                    '*encrypt': 'bool',
+                                     '*format': 'BackupFormat',
+                                     '*config-file': 'str',
+                                     '*firewall-file': 'str',

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

@@ -0,0 +1,400 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Wed, 8 Jul 2020 09:50:54 +0200
+Subject: [PATCH] PVE: Add PBS block driver to map backup archives into VMs
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+[error cleanups, file_open implementation]
+Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
+---
+ block/meson.build    |   3 +
+ block/pbs.c          | 271 +++++++++++++++++++++++++++++++++++++++++++
+ configure            |   9 ++
+ meson.build          |   1 +
+ qapi/block-core.json |  14 ++-
+ 5 files changed, 297 insertions(+), 1 deletion(-)
+ create mode 100644 block/pbs.c
+
+diff --git a/block/meson.build b/block/meson.build
+index dfae565db3..a070060e53 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -49,6 +49,9 @@ 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)
++
+ 
+ softmmu_ss.add(when: 'CONFIG_TCG', if_true: files('blkreplay.c'))
+ 
+diff --git a/block/pbs.c b/block/pbs.c
+new file mode 100644
+index 0000000000..1481a2bfd1
+--- /dev/null
++++ b/block/pbs.c
+@@ -0,0 +1,271 @@
++/*
++ * Proxmox Backup Server read-only block driver
++ */
++
++#include "qemu/osdep.h"
++#include "qapi/error.h"
++#include "qapi/qmp/qdict.h"
++#include "qapi/qmp/qstring.h"
++#include "qemu/module.h"
++#include "qemu/option.h"
++#include "qemu/cutils.h"
++#include "block/block_int.h"
++
++#include <proxmox-backup-qemu.h>
++
++#define PBS_OPT_REPOSITORY "repository"
++#define PBS_OPT_SNAPSHOT "snapshot"
++#define PBS_OPT_ARCHIVE "archive"
++#define PBS_OPT_KEYFILE "keyfile"
++#define PBS_OPT_PASSWORD "password"
++#define PBS_OPT_FINGERPRINT "fingerprint"
++#define PBS_OPT_ENCRYPTION_PASSWORD "key_password"
++
++typedef struct {
++    ProxmoxRestoreHandle *conn;
++    char aid;
++    int64_t length;
++
++    char *repository;
++    char *snapshot;
++    char *archive;
++} BDRVPBSState;
++
++static QemuOptsList runtime_opts = {
++    .name = "pbs",
++    .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head),
++    .desc = {
++        {
++            .name = PBS_OPT_REPOSITORY,
++            .type = QEMU_OPT_STRING,
++            .help = "The server address and repository to connect to.",
++        },
++        {
++            .name = PBS_OPT_SNAPSHOT,
++            .type = QEMU_OPT_STRING,
++            .help = "The snapshot to read.",
++        },
++        {
++            .name = PBS_OPT_ARCHIVE,
++            .type = QEMU_OPT_STRING,
++            .help = "Which archive within the snapshot should be accessed.",
++        },
++        {
++            .name = PBS_OPT_PASSWORD,
++            .type = QEMU_OPT_STRING,
++            .help = "Server password. Can be passed as env var 'PBS_PASSWORD'.",
++        },
++        {
++            .name = PBS_OPT_FINGERPRINT,
++            .type = QEMU_OPT_STRING,
++            .help = "Server fingerprint. Can be passed as env var 'PBS_FINGERPRINT'.",
++        },
++        {
++            .name = PBS_OPT_ENCRYPTION_PASSWORD,
++            .type = QEMU_OPT_STRING,
++            .help = "Optional: Key password. Can be passed as env var 'PBS_ENCRYPTION_PASSWORD'.",
++        },
++        {
++            .name = PBS_OPT_KEYFILE,
++            .type = QEMU_OPT_STRING,
++            .help = "Optional: The path to the keyfile to use.",
++        },
++        { /* end of list */ }
++    },
++};
++
++
++// filename format:
++// pbs:repository=<repo>,snapshot=<snap>,password=<pw>,key_password=<kpw>,fingerprint=<fp>,archive=<archive>
++static void pbs_parse_filename(const char *filename, QDict *options,
++                                     Error **errp)
++{
++
++    if (!strstart(filename, "pbs:", &filename)) {
++        if (errp) error_setg(errp, "pbs_parse_filename failed - missing 'pbs:' prefix");
++    }
++
++
++    QemuOpts *opts = qemu_opts_parse_noisily(&runtime_opts, filename, false);
++    if (!opts) {
++        if (errp) error_setg(errp, "pbs_parse_filename failed");
++        return;
++    }
++
++    qemu_opts_to_qdict(opts, options);
++
++    qemu_opts_del(opts);
++}
++
++static int pbs_open(BlockDriverState *bs, QDict *options, int flags,
++                    Error **errp)
++{
++    QemuOpts *opts;
++    BDRVPBSState *s = bs->opaque;
++    char *pbs_error = NULL;
++
++    opts = qemu_opts_create(&runtime_opts, NULL, 0, &error_abort);
++    qemu_opts_absorb_qdict(opts, options, &error_abort);
++
++    s->repository = g_strdup(qemu_opt_get(opts, PBS_OPT_REPOSITORY));
++    s->snapshot = g_strdup(qemu_opt_get(opts, PBS_OPT_SNAPSHOT));
++    s->archive = g_strdup(qemu_opt_get(opts, PBS_OPT_ARCHIVE));
++    const char *keyfile = qemu_opt_get(opts, PBS_OPT_KEYFILE);
++    const char *password = qemu_opt_get(opts, PBS_OPT_PASSWORD);
++    const char *fingerprint = qemu_opt_get(opts, PBS_OPT_FINGERPRINT);
++    const char *key_password = qemu_opt_get(opts, PBS_OPT_ENCRYPTION_PASSWORD);
++
++    if (!password) {
++        password = getenv("PBS_PASSWORD");
++    }
++    if (!fingerprint) {
++        fingerprint = getenv("PBS_FINGERPRINT");
++    }
++    if (!key_password) {
++        key_password = getenv("PBS_ENCRYPTION_PASSWORD");
++    }
++
++    /* connect to PBS server in read mode */
++    s->conn = proxmox_restore_new(s->repository, s->snapshot, password,
++        keyfile, key_password, fingerprint, &pbs_error);
++
++    /* invalidates qemu_opt_get char pointers from above */
++    qemu_opts_del(opts);
++
++    if (!s->conn) {
++        if (pbs_error && errp) error_setg(errp, "PBS restore_new failed: %s", pbs_error);
++        if (pbs_error) proxmox_backup_free_error(pbs_error);
++        return -ENOMEM;
++    }
++
++    int ret = proxmox_restore_connect(s->conn, &pbs_error);
++    if (ret < 0) {
++        if (pbs_error && errp) error_setg(errp, "PBS connect failed: %s", pbs_error);
++        if (pbs_error) proxmox_backup_free_error(pbs_error);
++        return -ECONNREFUSED;
++    }
++
++    /* acquire handle and length */
++    s->aid = proxmox_restore_open_image(s->conn, s->archive, &pbs_error);
++    if (s->aid < 0) {
++        if (pbs_error && errp) error_setg(errp, "PBS open_image failed: %s", pbs_error);
++        if (pbs_error) proxmox_backup_free_error(pbs_error);
++        return -ENODEV;
++    }
++    s->length = proxmox_restore_get_image_length(s->conn, s->aid, &pbs_error);
++    if (s->length < 0) {
++        if (pbs_error && errp) error_setg(errp, "PBS get_image_length failed: %s", pbs_error);
++        if (pbs_error) proxmox_backup_free_error(pbs_error);
++        return -EINVAL;
++    }
++
++    return 0;
++}
++
++static int pbs_file_open(BlockDriverState *bs, QDict *options, int flags,
++                         Error **errp)
++{
++    return pbs_open(bs, options, flags, errp);
++}
++
++static void pbs_close(BlockDriverState *bs) {
++    BDRVPBSState *s = bs->opaque;
++    g_free(s->repository);
++    g_free(s->snapshot);
++    g_free(s->archive);
++    proxmox_restore_disconnect(s->conn);
++}
++
++static int64_t pbs_getlength(BlockDriverState *bs)
++{
++    BDRVPBSState *s = bs->opaque;
++    return s->length;
++}
++
++typedef struct ReadCallbackData {
++    Coroutine *co;
++    AioContext *ctx;
++} ReadCallbackData;
++
++static void read_callback(void *callback_data)
++{
++    ReadCallbackData *rcb = callback_data;
++    aio_co_schedule(rcb->ctx, rcb->co);
++}
++
++static coroutine_fn int pbs_co_preadv(BlockDriverState *bs,
++                                      uint64_t offset, uint64_t bytes,
++                                      QEMUIOVector *qiov, int flags)
++{
++    BDRVPBSState *s = bs->opaque;
++    int ret;
++    char *pbs_error = NULL;
++    uint8_t *buf = malloc(bytes);
++
++    ReadCallbackData rcb = {
++        .co = qemu_coroutine_self(),
++        .ctx = bdrv_get_aio_context(bs),
++    };
++
++    proxmox_restore_read_image_at_async(s->conn, s->aid, buf, offset, bytes,
++                                        read_callback, (void *) &rcb, &ret, &pbs_error);
++
++    qemu_coroutine_yield();
++
++    if (ret < 0) {
++        fprintf(stderr, "error during PBS read: %s\n", pbs_error ? pbs_error : "unknown error");
++        if (pbs_error) proxmox_backup_free_error(pbs_error);
++        return -EIO;
++    }
++
++    qemu_iovec_from_buf(qiov, 0, buf, bytes);
++    free(buf);
++
++    return ret;
++}
++
++static coroutine_fn int pbs_co_pwritev(BlockDriverState *bs,
++                                       uint64_t offset, uint64_t bytes,
++                                       QEMUIOVector *qiov, int 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)
++{
++    BDRVPBSState *s = bs->opaque;
++    snprintf(bs->exact_filename, sizeof(bs->exact_filename), "%s/%s(%s)",
++             s->repository, s->snapshot, s->archive);
++}
++
++static const char *const pbs_strong_runtime_opts[] = {
++    NULL
++};
++
++static BlockDriver bdrv_pbs_co = {
++    .format_name            = "pbs",
++    .protocol_name          = "pbs",
++    .instance_size          = sizeof(BDRVPBSState),
++
++    .bdrv_parse_filename    = pbs_parse_filename,
++
++    .bdrv_file_open         = pbs_file_open,
++    .bdrv_open              = pbs_open,
++    .bdrv_close             = pbs_close,
++    .bdrv_getlength         = pbs_getlength,
++
++    .bdrv_co_preadv         = pbs_co_preadv,
++    .bdrv_co_pwritev        = pbs_co_pwritev,
++
++    .bdrv_refresh_filename  = pbs_refresh_filename,
++    .strong_runtime_opts    = pbs_strong_runtime_opts,
++};
++
++static void bdrv_pbs_init(void)
++{
++    bdrv_register(&bdrv_pbs_co);
++}
++
++block_init(bdrv_pbs_init);
+diff --git a/configure b/configure
+index 18c26e0389..33d9933871 100755
+--- a/configure
++++ b/configure
+@@ -436,6 +436,7 @@ vvfat="yes"
+ qed="yes"
+ parallels="yes"
+ sheepdog="no"
++pbs_bdrv="yes"
+ libxml2=""
+ debug_mutex="no"
+ libpmem=""
+@@ -1461,6 +1462,10 @@ for opt do
+   ;;
+   --enable-sheepdog) sheepdog="yes"
+   ;;
++  --disable-pbs-bdrv) pbs_bdrv="no"
++  ;;
++  --enable-pbs-bdrv) pbs_bdrv="yes"
++  ;;
+   --disable-vhost-user) vhost_user="no"
+   ;;
+   --enable-vhost-user) vhost_user="yes"
+@@ -1843,6 +1848,7 @@ disabled with --disable-FEATURE, default is enabled if available:
+   qed             qed image format support
+   parallels       parallels image format support
+   sheepdog        sheepdog block driver support (deprecated)
++  pbs-bdrv        Proxmox backup server read-only block driver support
+   crypto-afalg    Linux AF_ALG crypto backend driver
+   capstone        capstone disassembler support
+   debug-mutex     mutex debugging support
+@@ -6682,6 +6688,9 @@ if test "$sheepdog" = "yes" ; then
+   add_to deprecated_features "sheepdog"
+   echo "CONFIG_SHEEPDOG=y" >> $config_host_mak
+ fi
++if test "$pbs_bdrv" = "yes" ; then
++  echo "CONFIG_PBS_BDRV=y" >> $config_host_mak
++fi
+ if test "$pty_h" = "yes" ; then
+   echo "HAVE_PTY_H=y" >> $config_host_mak
+ fi
+diff --git a/meson.build b/meson.build
+index 6f1fafee14..4d156d35ce 100644
+--- a/meson.build
++++ b/meson.build
+@@ -2199,6 +2199,7 @@ summary_info += {'vvfat support':     config_host.has_key('CONFIG_VVFAT')}
+ summary_info += {'qed support':       config_host.has_key('CONFIG_QED')}
+ summary_info += {'parallels support': config_host.has_key('CONFIG_PARALLELS')}
+ summary_info += {'sheepdog support':  config_host.has_key('CONFIG_SHEEPDOG')}
++summary_info += {'PBS bdrv support':  config_host.has_key('CONFIG_PBS_BDRV')}
+ summary_info += {'capstone':          capstone_opt == 'disabled' ? false : capstone_opt}
+ summary_info += {'libpmem support':   config_host.has_key('CONFIG_LIBPMEM')}
+ summary_info += {'libdaxctl support': config_host.has_key('CONFIG_LIBDAXCTL')}
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 0fda1e3fd3..553112d998 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -2975,7 +2975,7 @@
+             'luks', 'nbd', 'nfs', 'null-aio', 'null-co', 'nvme', 'parallels',
+             'qcow', 'qcow2', 'qed', 'quorum', 'raw', 'rbd',
+             { 'name': 'replication', 'if': 'defined(CONFIG_REPLICATION)' },
+-            'sheepdog',
++            'sheepdog', 'pbs',
+             'ssh', 'throttle', 'vdi', 'vhdx', 'vmdk', 'vpc', 'vvfat' ] }
+ 
+ ##
+@@ -3039,6 +3039,17 @@
+ { 'struct': 'BlockdevOptionsNull',
+   'data': { '*size': 'int', '*latency-ns': 'uint64', '*read-zeroes': 'bool' } }
+ 
++##
++# @BlockdevOptionsPbs:
++#
++# Driver specific block device options for the PBS backend.
++#
++##
++{ 'struct': 'BlockdevOptionsPbs',
++  'data': { 'repository': 'str', 'snapshot': 'str', 'archive': 'str',
++            '*keyfile': 'str', '*password': 'str', '*fingerprint': 'str',
++            '*key_password': 'str' } }
++
+ ##
+ # @BlockdevOptionsNVMe:
+ #
+@@ -4148,6 +4159,7 @@
+       'nfs':        'BlockdevOptionsNfs',
+       'null-aio':   'BlockdevOptionsNull',
+       'null-co':    'BlockdevOptionsNull',
++      'pbs':        'BlockdevOptionsPbs',
+       'nvme':       'BlockdevOptionsNVMe',
+       'parallels':  'BlockdevOptionsGenericFormat',
+       'qcow2':      'BlockdevOptionsQcow2',

debian/patches/pve/0033-PVE-Backup-avoid-coroutines-to-fix-AIO-freeze-cleanu.patch

@@ -1,884 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Dietmar Maurer <dietmar@proxmox.com>
-Date: Mon, 6 Apr 2020 12:17:02 +0200
-Subject: [PATCH] PVE-Backup: avoid coroutines to fix AIO freeze, cleanups
-
-We observed various AIO pool loop freezes, so we decided to avoid
-coroutines and restrict ourselfes using similar code as upstream
-(see blockdev.c: do_backup_common).
-
-* avoid coroutine for job related code (causes hangs with iothreads)
-    - this imply to use normal QemuRecMutex instead of CoMutex
-
-* split pvebackup_co_dump_cb into:
-    - pvebackup_co_dump_pbs_cb and
-    - pvebackup_co_dump_pbs_cb
-
-* new helper functions
-    - pvebackup_propagate_error
-    - pvebackup_error_or_canceled
-    - pvebackup_add_transfered_bytes
-
-* avoid cancel flag (not needed)
-
-* simplify backup_cancel logic
-
-There is progress on upstream to support running qmp commands inside
-coroutines, see:
-
-https://lists.gnu.org/archive/html/qemu-devel/2020-02/msg04852.html
-
-We should consider using that when it is available in upstream qemu.
----
- pve-backup.c | 611 +++++++++++++++++++++++++--------------------------
- 1 file changed, 299 insertions(+), 312 deletions(-)
-
-diff --git a/pve-backup.c b/pve-backup.c
-index 38dd33e28b..169f0c68d0 100644
---- a/pve-backup.c
-+++ b/pve-backup.c
-@@ -11,11 +11,10 @@
- 
- /* PVE backup state and related function */
- 
--
- static struct PVEBackupState {
-     struct {
--        // Everithing accessed from qmp command, protected using rwlock
--        CoRwlock rwlock;
-+        // Everithing accessed from qmp_backup_query command is protected using lock
-+        QemuRecMutex lock;
-         Error *error;
-         time_t start_time;
-         time_t end_time;
-@@ -25,19 +24,18 @@ static struct PVEBackupState {
-         size_t total;
-         size_t transferred;
-         size_t zero_bytes;
--        bool cancel;
-     } stat;
-     int64_t speed;
-     VmaWriter *vmaw;
-     ProxmoxBackupHandle *pbs;
-     GList *di_list;
--    CoMutex backup_mutex;
-+    QemuRecMutex backup_mutex;
- } backup_state;
- 
- static void pvebackup_init(void)
- {
--    qemu_co_rwlock_init(&backup_state.stat.rwlock);
--    qemu_co_mutex_init(&backup_state.backup_mutex);
-+    qemu_rec_mutex_init(&backup_state.stat.lock);
-+    qemu_rec_mutex_init(&backup_state.backup_mutex);
- }
- 
- // initialize PVEBackupState at startup
-@@ -52,10 +50,54 @@ typedef struct PVEBackupDevInfo {
-     BlockDriverState *target;
- } PVEBackupDevInfo;
- 
--static void pvebackup_co_run_next_job(void);
-+static void pvebackup_run_next_job(void);
-+
-+static BlockJob *
-+lookup_active_block_job(PVEBackupDevInfo *di)
-+{
-+    if (!di->completed && di->bs) {
-+        for (BlockJob *job = block_job_next(NULL); job; job = block_job_next(job)) {
-+            if (job->job.driver->job_type != JOB_TYPE_BACKUP) {
-+                continue;
-+            }
-+
-+            BackupBlockJob *bjob = container_of(job, BackupBlockJob, common);
-+            if (bjob && bjob->source_bs == di->bs) {
-+                return job;
-+            }
-+        }
-+    }
-+    return NULL;
-+}
-+
-+static void pvebackup_propagate_error(Error *err)
-+{
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    error_propagate(&backup_state.stat.error, err);
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+}
-+
-+static bool pvebackup_error_or_canceled(void)
-+{
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    bool error_or_canceled = !!backup_state.stat.error;
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+
-+    return error_or_canceled;
-+}
- 
-+static void pvebackup_add_transfered_bytes(size_t transferred, size_t zero_bytes)
-+{
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    backup_state.stat.zero_bytes += zero_bytes;
-+    backup_state.stat.transferred += transferred;
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+}
-+
-+// This may get called from multiple coroutines in multiple io-threads
-+// Note1: this may get called after job_cancel()
- static int coroutine_fn
--pvebackup_co_dump_cb(
-+pvebackup_co_dump_pbs_cb(
-     void *opaque,
-     uint64_t start,
-     uint64_t bytes,
-@@ -67,137 +109,129 @@ pvebackup_co_dump_cb(
-     const unsigned char *buf = pbuf;
-     PVEBackupDevInfo *di = opaque;
- 
--    qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--    bool cancel = backup_state.stat.cancel;
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+    assert(backup_state.pbs);
-+
-+    Error *local_err = NULL;
-+    int pbs_res = -1;
-+
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
--    if (cancel) {
--        return size; // return success
-+    // avoid deadlock if job is cancelled
-+    if (pvebackup_error_or_canceled()) {
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        return -1;
-     }
- 
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
-+    pbs_res = proxmox_backup_co_write_data(backup_state.pbs, di->dev_id, buf, start, size, &local_err);
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
--    int ret = -1;
-+    if (pbs_res < 0) {
-+        pvebackup_propagate_error(local_err);
-+        return pbs_res;
-+    } else {
-+        pvebackup_add_transfered_bytes(size, !buf ? size : 0);
-+    }
- 
--    if (backup_state.vmaw) {
--        size_t zero_bytes = 0;
--        uint64_t remaining = size;
--
--        uint64_t cluster_num = start / VMA_CLUSTER_SIZE;
--        if ((cluster_num * VMA_CLUSTER_SIZE) != start) {
--            qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--            if (!backup_state.stat.error) {
--                qemu_co_rwlock_upgrade(&backup_state.stat.rwlock);
--                error_setg(&backup_state.stat.error,
--                           "got unaligned write inside backup dump "
--                           "callback (sector %ld)", start);
--            }
--            qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--            qemu_co_mutex_unlock(&backup_state.backup_mutex);
--            return -1; // not aligned to cluster size
--        }
-+    return size;
-+}
- 
--        while (remaining > 0) {
--            ret = vma_writer_write(backup_state.vmaw, di->dev_id, cluster_num,
--                                   buf, &zero_bytes);
--            ++cluster_num;
--            if (buf) {
--                buf += VMA_CLUSTER_SIZE;
--            }
--            if (ret < 0) {
--                qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--                if (!backup_state.stat.error) {
--                    qemu_co_rwlock_upgrade(&backup_state.stat.rwlock);
--                    vma_writer_error_propagate(backup_state.vmaw, &backup_state.stat.error);
--                }
--                qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+// This may get called from multiple coroutines in multiple io-threads
-+static int coroutine_fn
-+pvebackup_co_dump_vma_cb(
-+    void *opaque,
-+    uint64_t start,
-+    uint64_t bytes,
-+    const void *pbuf)
-+{
-+    assert(qemu_in_coroutine());
- 
--                qemu_co_mutex_unlock(&backup_state.backup_mutex);
--                return ret;
--            } else {
--                qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--                backup_state.stat.zero_bytes += zero_bytes;
--                if (remaining >= VMA_CLUSTER_SIZE) {
--                    backup_state.stat.transferred += VMA_CLUSTER_SIZE;
--                    remaining -= VMA_CLUSTER_SIZE;
--                } else {
--                    backup_state.stat.transferred += remaining;
--                    remaining = 0;
--                }
--                qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--            }
--        }
--    } else if (backup_state.pbs) {
--        Error *local_err = NULL;
--        int pbs_res = -1;
-+    const uint64_t size = bytes;
-+    const unsigned char *buf = pbuf;
-+    PVEBackupDevInfo *di = opaque;
- 
--        pbs_res = proxmox_backup_co_write_data(backup_state.pbs, di->dev_id, buf, start, size, &local_err);
- 
--        qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
-+    int ret = -1;
- 
--        if (pbs_res < 0) {
--            error_propagate(&backup_state.stat.error, local_err);
--            qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--            qemu_co_mutex_unlock(&backup_state.backup_mutex);
--            return pbs_res;
--        } else {
--            if (!buf) {
--                backup_state.stat.zero_bytes += size;
--            }
--            backup_state.stat.transferred += size;
-+    assert(backup_state.vmaw);
-+
-+    uint64_t remaining = size;
-+
-+    uint64_t cluster_num = start / VMA_CLUSTER_SIZE;
-+    if ((cluster_num * VMA_CLUSTER_SIZE) != start) {
-+        Error *local_err = NULL;
-+        error_setg(&local_err,
-+                   "got unaligned write inside backup dump "
-+                   "callback (sector %ld)", start);
-+        pvebackup_propagate_error(local_err);
-+        return -1; // not aligned to cluster size
-+    }
-+
-+    while (remaining > 0) {
-+        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+        // avoid deadlock if job is cancelled
-+        if (pvebackup_error_or_canceled()) {
-+            qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+            return -1;
-         }
- 
--        qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+        size_t zero_bytes = 0;
-+        ret = vma_writer_write(backup_state.vmaw, di->dev_id, cluster_num, buf, &zero_bytes);
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
--    } else {
--        qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--        if (!buf) {
--            backup_state.stat.zero_bytes += size;
-+        ++cluster_num;
-+        if (buf) {
-+            buf += VMA_CLUSTER_SIZE;
-+        }
-+        if (ret < 0) {
-+            Error *local_err = NULL;
-+            vma_writer_error_propagate(backup_state.vmaw, &local_err);
-+            pvebackup_propagate_error(local_err);
-+            return ret;
-+        } else {
-+            if (remaining >= VMA_CLUSTER_SIZE) {
-+                assert(ret == VMA_CLUSTER_SIZE);
-+                pvebackup_add_transfered_bytes(VMA_CLUSTER_SIZE, zero_bytes);
-+                remaining -= VMA_CLUSTER_SIZE;
-+            } else {
-+                assert(ret == remaining);
-+                pvebackup_add_transfered_bytes(remaining, zero_bytes);
-+                remaining = 0;
-+            }
-         }
--        backup_state.stat.transferred += size;
--        qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-     }
- 
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
--
-     return size;
- }
- 
--static void coroutine_fn pvebackup_co_cleanup(void)
-+static void coroutine_fn pvebackup_co_cleanup(void *unused)
- {
-     assert(qemu_in_coroutine());
- 
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
--    qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
-     backup_state.stat.end_time = time(NULL);
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
- 
-     if (backup_state.vmaw) {
-         Error *local_err = NULL;
-         vma_writer_close(backup_state.vmaw, &local_err);
- 
-         if (local_err != NULL) {
--            qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--            error_propagate(&backup_state.stat.error, local_err);
--            qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--        }
-+            pvebackup_propagate_error(local_err);
-+         }
- 
-         backup_state.vmaw = NULL;
-     }
- 
-     if (backup_state.pbs) {
--        qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--        bool error_or_canceled = backup_state.stat.error || backup_state.stat.cancel;
--        if (!error_or_canceled) {
-+        if (!pvebackup_error_or_canceled()) {
-             Error *local_err = NULL;
-             proxmox_backup_co_finish(backup_state.pbs, &local_err);
-             if (local_err != NULL) {
--                qemu_co_rwlock_upgrade(&backup_state.stat.rwlock);
--                error_propagate(&backup_state.stat.error, local_err);
--             }
-+                pvebackup_propagate_error(local_err);
-+            }
-         }
--        qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
- 
-         proxmox_backup_disconnect(backup_state.pbs);
-         backup_state.pbs = NULL;
-@@ -205,43 +239,14 @@ static void coroutine_fn pvebackup_co_cleanup(void)
- 
-     g_list_free(backup_state.di_list);
-     backup_state.di_list = NULL;
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- }
- 
--typedef struct PVEBackupCompeteCallbackData {
--    PVEBackupDevInfo *di;
--    int result;
--} PVEBackupCompeteCallbackData;
--
--static void coroutine_fn pvebackup_co_complete_cb(void *opaque)
-+static void coroutine_fn pvebackup_complete_stream(void *opaque)
- {
--    assert(qemu_in_coroutine());
--
--    PVEBackupCompeteCallbackData *cb_data = opaque;
--
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
--
--    PVEBackupDevInfo *di = cb_data->di;
--    int ret = cb_data->result;
--
--    di->completed = true;
--
--    qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--    bool error_or_canceled = backup_state.stat.error || backup_state.stat.cancel;
--
--    if (ret < 0 && !backup_state.stat.error) {
--        qemu_co_rwlock_upgrade(&backup_state.stat.rwlock);
--        error_setg(&backup_state.stat.error, "job failed with err %d - %s",
--                   ret, strerror(-ret));
--    }
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--
--    di->bs = NULL;
-+    PVEBackupDevInfo *di = opaque;
- 
--    if (di->target) {
--        bdrv_unref(di->target);
--        di->target = NULL;
--    }
-+    bool error_or_canceled = pvebackup_error_or_canceled();
- 
-     if (backup_state.vmaw) {
-         vma_writer_close_stream(backup_state.vmaw, di->dev_id);
-@@ -251,108 +256,96 @@ static void coroutine_fn pvebackup_co_complete_cb(void *opaque)
-         Error *local_err = NULL;
-         proxmox_backup_co_close_image(backup_state.pbs, di->dev_id, &local_err);
-         if (local_err != NULL) {
--            qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--            error_propagate(&backup_state.stat.error, local_err);
--            qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+            pvebackup_propagate_error(local_err);
-         }
-     }
-+}
- 
--    // remove self from job queue
--    backup_state.di_list = g_list_remove(backup_state.di_list, di);
--    g_free(di);
-+static void pvebackup_complete_cb(void *opaque, int ret)
-+{
-+    assert(!qemu_in_coroutine());
-+
-+    PVEBackupDevInfo *di = opaque;
- 
--    int pending_jobs = g_list_length(backup_state.di_list);
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+    di->completed = true;
- 
--    if (pending_jobs > 0) {
--        pvebackup_co_run_next_job();
--    } else {
--        pvebackup_co_cleanup();
-+    if (ret < 0) {
-+        Error *local_err = NULL;
-+        error_setg(&local_err, "job failed with err %d - %s", ret, strerror(-ret));
-+        pvebackup_propagate_error(local_err);
-     }
--}
- 
--static void pvebackup_complete_cb(void *opaque, int ret)
--{
--    // This can be called from the main loop, or from a coroutine
--    PVEBackupCompeteCallbackData cb_data = {
--        .di = opaque,
--        .result = ret,
--    };
-+    di->bs = NULL;
- 
--    if (qemu_in_coroutine()) {
--        pvebackup_co_complete_cb(&cb_data);
--    } else {
--        block_on_coroutine_fn(pvebackup_co_complete_cb, &cb_data);
--    }
--}
-+    assert(di->target == NULL);
- 
--static void coroutine_fn pvebackup_co_cancel(void *opaque)
--{
--    assert(qemu_in_coroutine());
-+    block_on_coroutine_fn(pvebackup_complete_stream, di);
- 
--    qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--    backup_state.stat.cancel = true;
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+    // remove self from job queue
-+    backup_state.di_list = g_list_remove(backup_state.di_list, di);
- 
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
-+    g_free(di);
- 
--    // Avoid race between block jobs and backup-cancel command:
--    if (!(backup_state.vmaw || backup_state.pbs)) {
--        qemu_co_mutex_unlock(&backup_state.backup_mutex);
--        return;
--    }
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
--    qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--    if (!backup_state.stat.error) {
--        qemu_co_rwlock_upgrade(&backup_state.stat.rwlock);
--        error_setg(&backup_state.stat.error, "backup cancelled");
--    }
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+    pvebackup_run_next_job();
-+}
-+
-+static void pvebackup_cancel(void)
-+{
-+    Error *cancel_err = NULL;
-+    error_setg(&cancel_err, "backup canceled");
-+    pvebackup_propagate_error(cancel_err);
-+
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
-     if (backup_state.vmaw) {
-         /* make sure vma writer does not block anymore */
--        vma_writer_set_error(backup_state.vmaw, "backup cancelled");
-+        vma_writer_set_error(backup_state.vmaw, "backup canceled");
-     }
- 
-     if (backup_state.pbs) {
--        proxmox_backup_abort(backup_state.pbs, "backup cancelled");
-+        proxmox_backup_abort(backup_state.pbs, "backup canceled");
-     }
- 
--    bool running_jobs = 0;
--    GList *l = backup_state.di_list;
--    while (l) {
--        PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
--        l = g_list_next(l);
--        if (!di->completed && di->bs) {
--            for (BlockJob *job = block_job_next(NULL); job; job = block_job_next(job)) {
--                if (job->job.driver->job_type != JOB_TYPE_BACKUP) {
--                    continue;
--                }
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
--                BackupBlockJob *bjob = container_of(job, BackupBlockJob, common);
--                if (bjob && bjob->source_bs == di->bs) {
--                    AioContext *aio_context = job->job.aio_context;
--                    aio_context_acquire(aio_context);
-+    for(;;) {
- 
--                    if (!di->completed) {
--                        running_jobs += 1;
--                        job_cancel(&job->job, false);
--                    }
--                    aio_context_release(aio_context);
--                }
-+        BlockJob *next_job = NULL;
-+
-+        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+
-+        GList *l = backup_state.di_list;
-+        while (l) {
-+            PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-+            l = g_list_next(l);
-+
-+            BlockJob *job = lookup_active_block_job(di);
-+            if (job != NULL) {
-+                next_job = job;
-+                break;
-             }
-         }
--    }
- 
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
--    if (running_jobs == 0) pvebackup_co_cleanup(); // else job will call completion handler
-+        if (next_job) {
-+            AioContext *aio_context = next_job->job.aio_context;
-+            aio_context_acquire(aio_context);
-+            job_cancel_sync(&next_job->job);
-+            aio_context_release(aio_context);
-+        } else {
-+            break;
-+        }
-+    }
- }
- 
- void qmp_backup_cancel(Error **errp)
- {
--    block_on_coroutine_fn(pvebackup_co_cancel, NULL);
-+    pvebackup_cancel();
- }
- 
- static int coroutine_fn pvebackup_co_add_config(
-@@ -406,46 +399,97 @@ static int coroutine_fn pvebackup_co_add_config(
- 
- bool job_should_pause(Job *job);
- 
--static void coroutine_fn pvebackup_co_run_next_job(void)
-+static void pvebackup_run_next_job(void)
- {
--    assert(qemu_in_coroutine());
-+    assert(!qemu_in_coroutine());
- 
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
-     GList *l = backup_state.di_list;
-     while (l) {
-         PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-         l = g_list_next(l);
--        if (!di->completed && di->bs) {
--            for (BlockJob *job = block_job_next(NULL); job; job = block_job_next(job)) {
--                if (job->job.driver->job_type != JOB_TYPE_BACKUP) {
--                    continue;
--                }
- 
--                BackupBlockJob *bjob = container_of(job, BackupBlockJob, common);
--                if (bjob && bjob->source_bs == di->bs) {
--                    AioContext *aio_context = job->job.aio_context;
--                    qemu_co_mutex_unlock(&backup_state.backup_mutex);
--                    aio_context_acquire(aio_context);
--
--                    if (job_should_pause(&job->job)) {
--                        qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--                        bool error_or_canceled = backup_state.stat.error || backup_state.stat.cancel;
--                        qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--
--                        if (error_or_canceled) {
--                            job_cancel(&job->job, false);
--                        } else {
--                            job_resume(&job->job);
--                        }
--                    }
--                    aio_context_release(aio_context);
--                    return;
-+        BlockJob *job = lookup_active_block_job(di);
-+
-+        if (job) {
-+            qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+
-+            AioContext *aio_context = job->job.aio_context;
-+            aio_context_acquire(aio_context);
-+
-+            if (job_should_pause(&job->job)) {
-+                bool error_or_canceled = pvebackup_error_or_canceled();
-+                if (error_or_canceled) {
-+                    job_cancel_sync(&job->job);
-+                } else {
-+                    job_resume(&job->job);
-                 }
-             }
-+            aio_context_release(aio_context);
-+            return;
-         }
-     }
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+
-+    block_on_coroutine_fn(pvebackup_co_cleanup, NULL); // no more jobs, run cleanup
-+}
-+
-+static bool create_backup_jobs(void) {
-+
-+    assert(!qemu_in_coroutine());
-+
-+    Error *local_err = NULL;
-+
-+    /* create and start all jobs (paused state) */
-+    GList *l =  backup_state.di_list;
-+    while (l) {
-+        PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-+        l = g_list_next(l);
-+
-+        assert(di->target != NULL);
-+
-+        AioContext *aio_context = bdrv_get_aio_context(di->bs);
-+        aio_context_acquire(aio_context);
-+
-+        BlockJob *job = backup_job_create(
-+            NULL, di->bs, di->target, backup_state.speed, MIRROR_SYNC_MODE_FULL, NULL,
-+            BITMAP_SYNC_MODE_NEVER, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
-+            JOB_DEFAULT, pvebackup_complete_cb, di, 1, NULL, &local_err);
-+
-+        aio_context_release(aio_context);
-+
-+        if (!job || local_err != NULL) {
-+            Error *create_job_err = NULL;
-+            error_setg(&create_job_err, "backup_job_create failed: %s",
-+                       local_err ? error_get_pretty(local_err) : "null");
-+
-+            pvebackup_propagate_error(create_job_err);
-+            break;
-+        }
-+        job_start(&job->job);
-+
-+        bdrv_unref(di->target);
-+        di->target = NULL;
-+    }
-+
-+    bool errors = pvebackup_error_or_canceled();
-+
-+    if (errors) {
-+        l = backup_state.di_list;
-+        while (l) {
-+            PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-+            l = g_list_next(l);
-+
-+            if (di->target) {
-+                bdrv_unref(di->target);
-+                di->target = NULL;
-+            }
-+        }
-+    }
-+
-+    return errors;
- }
- 
- typedef struct QmpBackupTask {
-@@ -476,7 +520,7 @@ typedef struct QmpBackupTask {
-     UuidInfo *result;
- } QmpBackupTask;
- 
--static void coroutine_fn pvebackup_co_start(void *opaque)
-+static void coroutine_fn pvebackup_co_prepare(void *opaque)
- {
-     assert(qemu_in_coroutine());
- 
-@@ -495,15 +539,14 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-     GList *di_list = NULL;
-     GList *l;
-     UuidInfo *uuid_info;
--    BlockJob *job;
- 
-     const char *config_name = "qemu-server.conf";
-     const char *firewall_name = "qemu-server.fw";
- 
--    qemu_co_mutex_lock(&backup_state.backup_mutex);
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
- 
-     if (backup_state.di_list) {
--        qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-         error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
-                   "previous backup not finished");
-         return;
-@@ -631,7 +674,7 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-             if (dev_id < 0)
-                 goto err;
- 
--            if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_cb, di, task->errp))) {
-+            if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_pbs_cb, di, task->errp))) {
-                 goto err;
-             }
- 
-@@ -652,7 +695,7 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-             PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
-             l = g_list_next(l);
- 
--            if (!(di->target = bdrv_backup_dump_create(VMA_CLUSTER_SIZE, di->size, pvebackup_co_dump_cb, di, task->errp))) {
-+            if (!(di->target = bdrv_backup_dump_create(VMA_CLUSTER_SIZE, di->size, pvebackup_co_dump_vma_cb, di, task->errp))) {
-                 goto err;
-             }
- 
-@@ -717,9 +760,7 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-     }
-     /* initialize global backup_state now */
- 
--    qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--
--    backup_state.stat.cancel = false;
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
- 
-     if (backup_state.stat.error) {
-         error_free(backup_state.stat.error);
-@@ -742,7 +783,7 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
-     backup_state.stat.transferred = 0;
-     backup_state.stat.zero_bytes = 0;
- 
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
- 
-     backup_state.speed = (task->has_speed && task->speed > 0) ? task->speed : 0;
- 
-@@ -751,45 +792,7 @@ static void coroutine_fn pvebackup_co_start(void *opaque)
- 
-     backup_state.di_list = di_list;
- 
--    /* start all jobs (paused state) */
--    l = di_list;
--    while (l) {
--        PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
--        l = g_list_next(l);
--
--        AioContext *aio_context = bdrv_get_aio_context(di->bs);
--        aio_context_acquire(aio_context);
--
--        job = backup_job_create(NULL, di->bs, di->target, backup_state.speed, MIRROR_SYNC_MODE_FULL, NULL,
--                                BITMAP_SYNC_MODE_NEVER, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
--                                JOB_DEFAULT, pvebackup_complete_cb, di, 1, NULL, &local_err);
--
--        aio_context_release(aio_context);
--
--        if (!job || local_err != NULL) {
--            qemu_co_rwlock_wrlock(&backup_state.stat.rwlock);
--            error_setg(&backup_state.stat.error, "backup_job_create failed");
--            qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--            break;
--        }
--        job_start(&job->job);
--        if (di->target) {
--            bdrv_unref(di->target);
--            di->target = NULL;
--        }
--    }
--
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
--
--    qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
--    bool no_errors = !backup_state.stat.error;
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--
--    if (no_errors) {
--        pvebackup_co_run_next_job(); // run one job
--    } else {
--        pvebackup_co_cancel(NULL);
--    }
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
-     uuid_info = g_malloc0(sizeof(*uuid_info));
-     uuid_info->UUID = uuid_str;
-@@ -833,7 +836,7 @@ err:
-         rmdir(backup_dir);
-     }
- 
--    qemu_co_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- 
-     task->result = NULL;
-     return;
-@@ -878,32 +881,28 @@ UuidInfo *qmp_backup(
-         .errp = errp,
-     };
- 
--    block_on_coroutine_fn(pvebackup_co_start, &task);
-+    block_on_coroutine_fn(pvebackup_co_prepare, &task);
-+
-+    if (*errp == NULL) {
-+        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+        create_backup_jobs();
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        pvebackup_run_next_job();
-+    }
- 
-     return task.result;
- }
- 
--
--typedef struct QmpQueryBackupTask {
--    Error **errp;
--    BackupStatus *result;
--} QmpQueryBackupTask;
--
--static void coroutine_fn pvebackup_co_query(void *opaque)
-+BackupStatus *qmp_query_backup(Error **errp)
- {
--    assert(qemu_in_coroutine());
--
--    QmpQueryBackupTask *task = opaque;
--
-     BackupStatus *info = g_malloc0(sizeof(*info));
- 
--    qemu_co_rwlock_rdlock(&backup_state.stat.rwlock);
-+    qemu_rec_mutex_lock(&backup_state.stat.lock);
- 
-     if (!backup_state.stat.start_time) {
-         /* not started, return {} */
--        task->result = info;
--        qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--        return;
-+        qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+        return info;
-     }
- 
-     info->has_status = true;
-@@ -939,19 +938,7 @@ static void coroutine_fn pvebackup_co_query(void *opaque)
-     info->has_transferred = true;
-     info->transferred = backup_state.stat.transferred;
- 
--    task->result = info;
-+    qemu_rec_mutex_unlock(&backup_state.stat.lock);
- 
--    qemu_co_rwlock_unlock(&backup_state.stat.rwlock);
--}
--
--BackupStatus *qmp_query_backup(Error **errp)
--{
--    QmpQueryBackupTask task = {
--        .errp = errp,
--        .result = NULL,
--    };
--
--    block_on_coroutine_fn(pvebackup_co_query, &task);
--
--    return task.result;
-+    return info;
- }

debian/patches/pve/0033-PVE-add-query_proxmox_support-QMP-command.patch

@@ -0,0 +1,74 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Wed, 8 Jul 2020 11:57:53 +0200
+Subject: [PATCH] PVE: add query_proxmox_support QMP command
+
+Generic interface for future use, currently used for PBS dirty-bitmap
+backup support.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
+[PVE: query-proxmox-support: include library version]
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ pve-backup.c         |  9 +++++++++
+ qapi/block-core.json | 29 +++++++++++++++++++++++++++++
+ 2 files changed, 38 insertions(+)
+
+diff --git a/pve-backup.c b/pve-backup.c
+index b8182aaf89..98e79552ef 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -1073,3 +1073,12 @@ BackupStatus *qmp_query_backup(Error **errp)
+ 
+     return info;
+ }
++
++ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
++{
++    ProxmoxSupportStatus *ret = g_malloc0(sizeof(*ret));
++    ret->pbs_library_version = g_strdup(proxmox_backup_qemu_version());
++    ret->pbs_dirty_bitmap = true;
++    ret->pbs_dirty_bitmap_savevm = true;
++    return ret;
++}
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 553112d998..f3608390c4 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -868,6 +868,35 @@
+ ##
+ { 'command': 'backup-cancel' }
+ 
++##
++# @ProxmoxSupportStatus:
++#
++# Contains info about supported features added by Proxmox.
++#
++# @pbs-dirty-bitmap: True if dirty-bitmap-incremental backups to PBS are
++#                    supported.
++#
++# @pbs-dirty-bitmap-savevm: True if 'dirty-bitmaps' migration capability can
++#                           safely be set for savevm-async.
++#
++# @pbs-library-version: Running version of libproxmox-backup-qemu0 library.
++#
++##
++{ 'struct': 'ProxmoxSupportStatus',
++  'data': { 'pbs-dirty-bitmap': 'bool',
++            'pbs-dirty-bitmap-savevm': 'bool',
++            'pbs-library-version': 'str' } }
++
++##
++# @query-proxmox-support:
++#
++# Returns information about supported features added by Proxmox.
++#
++# Returns: @ProxmoxSupportStatus
++#
++##
++{ 'command': 'query-proxmox-support', 'returns': 'ProxmoxSupportStatus' }
++
+ ##
+ # @BlockDeviceTimedStats:
+ #

debian/patches/pve/0034-PVE-add-query-pbs-bitmap-info-QMP-call.patch

@@ -0,0 +1,440 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Wed, 19 Aug 2020 17:02:00 +0200
+Subject: [PATCH] PVE: add query-pbs-bitmap-info QMP call
+
+Returns advanced information about dirty bitmaps used (or not used) for
+the latest PBS backup.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ monitor/hmp-cmds.c   |  28 ++++++-----
+ pve-backup.c         | 117 ++++++++++++++++++++++++++++++++-----------
+ qapi/block-core.json |  56 +++++++++++++++++++++
+ 3 files changed, 159 insertions(+), 42 deletions(-)
+
+diff --git a/monitor/hmp-cmds.c b/monitor/hmp-cmds.c
+index 604026bb37..95f4e7f5c1 100644
+--- a/monitor/hmp-cmds.c
++++ b/monitor/hmp-cmds.c
+@@ -198,6 +198,7 @@ void hmp_info_mice(Monitor *mon, const QDict *qdict)
+ void hmp_info_backup(Monitor *mon, const QDict *qdict)
+ {
+     BackupStatus *info;
++    PBSBitmapInfoList *bitmap_info;
+ 
+     info = qmp_query_backup(NULL);
+ 
+@@ -228,26 +229,29 @@ void hmp_info_backup(Monitor *mon, const QDict *qdict)
+             // this should not happen normally
+             monitor_printf(mon, "Total size: %d\n", 0);
+         } else {
+-            bool incremental = false;
+             size_t total_or_dirty = info->total;
+-            if (info->has_transferred) {
+-                if (info->has_dirty && info->dirty) {
+-                     if (info->dirty < info->total) {
+-                        total_or_dirty = info->dirty;
+-                        incremental = true;
+-                    }
+-                }
++            bitmap_info = qmp_query_pbs_bitmap_info(NULL);
++
++            while (bitmap_info) {
++                monitor_printf(mon, "Drive %s:\n",
++                        bitmap_info->value->drive);
++                monitor_printf(mon, "  bitmap action: %s\n",
++                        PBSBitmapAction_str(bitmap_info->value->action));
++                monitor_printf(mon, "  size: %zd\n",
++                        bitmap_info->value->size);
++                monitor_printf(mon, "  dirty: %zd\n",
++                        bitmap_info->value->dirty);
++                bitmap_info = bitmap_info->next;
+             }
+ 
+-            int per = (info->transferred * 100)/total_or_dirty;
+-
+-            monitor_printf(mon, "Backup mode: %s\n", incremental ? "incremental" : "full");
++            qapi_free_PBSBitmapInfoList(bitmap_info);
+ 
+             int zero_per = (info->has_zero_bytes && info->zero_bytes) ?
+                 (info->zero_bytes * 100)/info->total : 0;
+             monitor_printf(mon, "Total size: %zd\n", info->total);
++            int trans_per = (info->transferred * 100)/total_or_dirty;
+             monitor_printf(mon, "Transferred bytes: %zd (%d%%)\n",
+-                           info->transferred, per);
++                           info->transferred, trans_per);
+             monitor_printf(mon, "Zero bytes: %zd (%d%%)\n",
+                            info->zero_bytes, zero_per);
+ 
+diff --git a/pve-backup.c b/pve-backup.c
+index 98e79552ef..8305105fd5 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -46,6 +46,7 @@ static struct PVEBackupState {
+         size_t transferred;
+         size_t reused;
+         size_t zero_bytes;
++        GList *bitmap_list;
+     } stat;
+     int64_t speed;
+     VmaWriter *vmaw;
+@@ -670,7 +671,6 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     }
+ 
+     size_t total = 0;
+-    size_t dirty = 0;
+ 
+     l = di_list;
+     while (l) {
+@@ -691,18 +691,33 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+ 
+     uuid_generate(uuid);
+ 
++    qemu_mutex_lock(&backup_state.stat.lock);
++    backup_state.stat.reused = 0;
++
++    /* clear previous backup's bitmap_list */
++    if (backup_state.stat.bitmap_list) {
++        GList *bl = backup_state.stat.bitmap_list;
++        while (bl) {
++            g_free(((PBSBitmapInfo *)bl->data)->drive);
++            g_free(bl->data);
++            bl = g_list_next(bl);
++        }
++        g_list_free(backup_state.stat.bitmap_list);
++        backup_state.stat.bitmap_list = NULL;
++    }
++
+     if (format == BACKUP_FORMAT_PBS) {
+         if (!task->has_password) {
+             error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'password'");
+-            goto err;
++            goto err_mutex;
+         }
+         if (!task->has_backup_id) {
+             error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-id'");
+-            goto err;
++            goto err_mutex;
+         }
+         if (!task->has_backup_time) {
+             error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-time'");
+-            goto err;
++            goto err_mutex;
+         }
+ 
+         int dump_cb_block_size = PROXMOX_BACKUP_DEFAULT_CHUNK_SIZE; // Hardcoded (4M)
+@@ -729,12 +744,12 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
+                       "proxmox_backup_new failed: %s", pbs_err);
+             proxmox_backup_free_error(pbs_err);
+-            goto err;
++            goto err_mutex;
+         }
+ 
+         int connect_result = proxmox_backup_co_connect(pbs, task->errp);
+         if (connect_result < 0)
+-            goto err;
++            goto err_mutex;
+ 
+         /* register all devices */
+         l = di_list;
+@@ -745,6 +760,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             di->block_size = dump_cb_block_size;
+ 
+             const char *devname = bdrv_get_device_name(di->bs);
++            PBSBitmapAction action = PBS_BITMAP_ACTION_NOT_USED;
++            size_t dirty = di->size;
+ 
+             BdrvDirtyBitmap *bitmap = bdrv_find_dirty_bitmap(di->bs, PBS_BITMAP_NAME);
+             bool expect_only_dirty = false;
+@@ -753,49 +770,59 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+                 if (bitmap == NULL) {
+                     bitmap = bdrv_create_dirty_bitmap(di->bs, dump_cb_block_size, PBS_BITMAP_NAME, task->errp);
+                     if (!bitmap) {
+-                        goto err;
++                        goto err_mutex;
+                     }
++                    action = PBS_BITMAP_ACTION_NEW;
+                 } else {
+                     expect_only_dirty = proxmox_backup_check_incremental(pbs, devname, di->size) != 0;
+                 }
+ 
+                 if (expect_only_dirty) {
+-                    dirty += bdrv_get_dirty_count(bitmap);
++                    /* track clean chunks as reused */
++                    dirty = MIN(bdrv_get_dirty_count(bitmap), di->size);
++                    backup_state.stat.reused += di->size - dirty;
++                    action = PBS_BITMAP_ACTION_USED;
+                 } else {
+                     /* mark entire bitmap as dirty to make full backup */
+                     bdrv_set_dirty_bitmap(bitmap, 0, di->size);
+-                    dirty += di->size;
++                    if (action != PBS_BITMAP_ACTION_NEW) {
++                        action = PBS_BITMAP_ACTION_INVALID;
++                    }
+                 }
+                 di->bitmap = bitmap;
+             } else {
+-                dirty += di->size;
+-
+                 /* after a full backup the old dirty bitmap is invalid anyway */
+                 if (bitmap != NULL) {
+                     bdrv_release_dirty_bitmap(bitmap);
++                    action = PBS_BITMAP_ACTION_NOT_USED_REMOVED;
+                 }
+             }
+ 
+             int dev_id = proxmox_backup_co_register_image(pbs, devname, di->size, expect_only_dirty, task->errp);
+             if (dev_id < 0) {
+-                goto err;
++                goto err_mutex;
+             }
+ 
+             if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_pbs_cb, di, task->errp))) {
+-                goto err;
++                goto err_mutex;
+             }
+ 
+             di->dev_id = dev_id;
++
++            PBSBitmapInfo *info = g_malloc(sizeof(*info));
++            info->drive = g_strdup(devname);
++            info->action = action;
++            info->size = di->size;
++            info->dirty = dirty;
++            backup_state.stat.bitmap_list = g_list_append(backup_state.stat.bitmap_list, info);
+         }
+     } else if (format == BACKUP_FORMAT_VMA) {
+-        dirty = total;
+-
+         vmaw = vma_writer_create(task->backup_file, uuid, &local_err);
+         if (!vmaw) {
+             if (local_err) {
+                 error_propagate(task->errp, local_err);
+             }
+-            goto err;
++            goto err_mutex;
+         }
+ 
+         /* register all devices for vma writer */
+@@ -805,7 +832,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             l = g_list_next(l);
+ 
+             if (!(di->target = bdrv_backup_dump_create(VMA_CLUSTER_SIZE, di->size, pvebackup_co_dump_vma_cb, di, task->errp))) {
+-                goto err;
++                goto err_mutex;
+             }
+ 
+             const char *devname = bdrv_get_device_name(di->bs);
+@@ -813,16 +840,14 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             if (di->dev_id <= 0) {
+                 error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
+                           "register_stream failed");
+-                goto err;
++                goto err_mutex;
+             }
+         }
+     } else if (format == BACKUP_FORMAT_DIR) {
+-        dirty = total;
+-
+         if (mkdir(task->backup_file, 0640) != 0) {
+             error_setg_errno(task->errp, errno, "can't create directory '%s'\n",
+                              task->backup_file);
+-            goto err;
++            goto err_mutex;
+         }
+         backup_dir = task->backup_file;
+ 
+@@ -839,18 +864,18 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+                             di->size, flags, false, &local_err);
+             if (local_err) {
+                 error_propagate(task->errp, local_err);
+-                goto err;
++                goto err_mutex;
+             }
+ 
+             di->target = bdrv_open(di->targetfile, NULL, NULL, flags, &local_err);
+             if (!di->target) {
+                 error_propagate(task->errp, local_err);
+-                goto err;
++                goto err_mutex;
+             }
+         }
+     } else {
+         error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "unknown backup format");
+-        goto err;
++        goto err_mutex;
+     }
+ 
+ 
+@@ -858,7 +883,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     if (task->has_config_file) {
+         if (pvebackup_co_add_config(task->config_file, config_name, format, backup_dir,
+                                     vmaw, pbs, task->errp) != 0) {
+-            goto err;
++            goto err_mutex;
+         }
+     }
+ 
+@@ -866,12 +891,11 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     if (task->has_firewall_file) {
+         if (pvebackup_co_add_config(task->firewall_file, firewall_name, format, backup_dir,
+                                     vmaw, pbs, task->errp) != 0) {
+-            goto err;
++            goto err_mutex;
+         }
+     }
+     /* initialize global backup_state now */
+-
+-    qemu_mutex_lock(&backup_state.stat.lock);
++    /* note: 'reused' and 'bitmap_list' are initialized earlier */
+ 
+     if (backup_state.stat.error) {
+         error_free(backup_state.stat.error);
+@@ -891,10 +915,9 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     char *uuid_str = g_strdup(backup_state.stat.uuid_str);
+ 
+     backup_state.stat.total = total;
+-    backup_state.stat.dirty = dirty;
++    backup_state.stat.dirty = total - backup_state.stat.reused;
+     backup_state.stat.transferred = 0;
+     backup_state.stat.zero_bytes = 0;
+-    backup_state.stat.reused = format == BACKUP_FORMAT_PBS && dirty >= total ? 0 : total - dirty;
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+@@ -911,6 +934,9 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     task->result = uuid_info;
+     return;
+ 
++err_mutex:
++    qemu_mutex_unlock(&backup_state.stat.lock);
++
+ err:
+ 
+     l = di_list;
+@@ -1074,11 +1100,42 @@ BackupStatus *qmp_query_backup(Error **errp)
+     return info;
+ }
+ 
++PBSBitmapInfoList *qmp_query_pbs_bitmap_info(Error **errp)
++{
++    PBSBitmapInfoList *head = NULL, **p_next = &head;
++
++    qemu_mutex_lock(&backup_state.stat.lock);
++
++    GList *l = backup_state.stat.bitmap_list;
++    while (l) {
++        PBSBitmapInfo *info = (PBSBitmapInfo *)l->data;
++        l = g_list_next(l);
++
++        /* clone bitmap info to avoid auto free after QMP marshalling */
++        PBSBitmapInfo *info_ret = g_malloc0(sizeof(*info_ret));
++        info_ret->drive = g_strdup(info->drive);
++        info_ret->action = info->action;
++        info_ret->size = info->size;
++        info_ret->dirty = info->dirty;
++
++        PBSBitmapInfoList *info_list = g_malloc0(sizeof(*info_list));
++        info_list->value = info_ret;
++
++        *p_next = info_list;
++        p_next = &info_list->next;
++    }
++
++    qemu_mutex_unlock(&backup_state.stat.lock);
++
++    return head;
++}
++
+ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
+ {
+     ProxmoxSupportStatus *ret = g_malloc0(sizeof(*ret));
+     ret->pbs_library_version = g_strdup(proxmox_backup_qemu_version());
+     ret->pbs_dirty_bitmap = true;
+     ret->pbs_dirty_bitmap_savevm = true;
++    ret->query_bitmap_info = true;
+     return ret;
+ }
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index f3608390c4..f57fda122c 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -876,6 +876,8 @@
+ # @pbs-dirty-bitmap: True if dirty-bitmap-incremental backups to PBS are
+ #                    supported.
+ #
++# @query-bitmap-info: True if the 'query-pbs-bitmap-info' QMP call is supported.
++#
+ # @pbs-dirty-bitmap-savevm: True if 'dirty-bitmaps' migration capability can
+ #                           safely be set for savevm-async.
+ #
+@@ -884,6 +886,7 @@
+ ##
+ { 'struct': 'ProxmoxSupportStatus',
+   'data': { 'pbs-dirty-bitmap': 'bool',
++            'query-bitmap-info': 'bool',
+             'pbs-dirty-bitmap-savevm': 'bool',
+             'pbs-library-version': 'str' } }
+ 
+@@ -897,6 +900,59 @@
+ ##
+ { 'command': 'query-proxmox-support', 'returns': 'ProxmoxSupportStatus' }
+ 
++##
++# @PBSBitmapAction:
++#
++# An action taken on a dirty-bitmap when a backup job was started.
++#
++# @not-used: Bitmap mode was not enabled.
++#
++# @not-used-removed: Bitmap mode was not enabled, but a bitmap from a
++#                    previous backup still existed and was removed.
++#
++# @new: A new bitmap was attached to the drive for this backup.
++#
++# @used: An existing bitmap will be used to only backup changed data.
++#
++# @invalid: A bitmap existed, but had to be cleared since it's associated
++#           base snapshot did not match the base given for the current job or
++#           the crypt mode has changed.
++#
++##
++{ 'enum': 'PBSBitmapAction',
++  'data': ['not-used', 'not-used-removed', 'new', 'used', 'invalid'] }
++
++##
++# @PBSBitmapInfo:
++#
++# Contains information about dirty bitmaps used for each drive in a PBS backup.
++#
++# @drive: The underlying drive.
++#
++# @action: The action that was taken when the backup started.
++#
++# @size: The total size of the drive.
++#
++# @dirty: How much of the drive is considered dirty and will be backed up,
++#         or 'size' if everything will be.
++#
++##
++{ 'struct': 'PBSBitmapInfo',
++  'data': { 'drive': 'str', 'action': 'PBSBitmapAction', 'size': 'int',
++            'dirty': 'int' } }
++
++##
++# @query-pbs-bitmap-info:
++#
++# Returns information about dirty bitmaps used on the most recently started
++# backup. Returns nothing when the last backup was not using PBS or if no
++# backup occured in this session.
++#
++# Returns: @PBSBitmapInfo
++#
++##
++{ 'command': 'query-pbs-bitmap-info', 'returns': ['PBSBitmapInfo'] }
++
+ ##
+ # @BlockDeviceTimedStats:
+ #

debian/patches/pve/0035-PVE-redirect-stderr-to-journal-when-daemonized.patch

@@ -0,0 +1,60 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Tue, 12 Jan 2021 14:12:20 +0100
+Subject: [PATCH] PVE: redirect stderr to journal when daemonized
+
+QEMU uses the logging for error messages usually, so LOG_ERR is most
+fitting.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ meson.build | 2 ++
+ os-posix.c  | 7 +++++--
+ 2 files changed, 7 insertions(+), 2 deletions(-)
+
+diff --git a/meson.build b/meson.build
+index 4d156d35ce..737ea9e5d7 100644
+--- a/meson.build
++++ b/meson.build
+@@ -726,6 +726,7 @@ keyutils = dependency('libkeyutils', required: false,
+ has_gettid = cc.has_function('gettid')
+ 
+ libuuid = cc.find_library('uuid', required: true)
++libsystemd = cc.find_library('systemd', required: true)
+ libproxmox_backup_qemu = cc.find_library('proxmox_backup_qemu', required: true)
+ 
+ # Malloc tests
+@@ -1539,6 +1540,7 @@ blockdev_ss.add(files(
+ # 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'))
++blockdev_ss.add(when: 'CONFIG_POSIX', if_true: libsystemd)
+ softmmu_ss.add(when: 'CONFIG_WIN32', if_true: [files('os-win32.c')])
+ 
+ common_ss.add(files('cpus-common.c'))
+diff --git a/os-posix.c b/os-posix.c
+index 1de2839554..ac4f652923 100644
+--- a/os-posix.c
++++ b/os-posix.c
+@@ -28,6 +28,8 @@
+ #include <pwd.h>
+ #include <grp.h>
+ #include <libgen.h>
++#include <systemd/sd-journal.h>
++#include <syslog.h>
+ 
+ #include "qemu-common.h"
+ /* Needed early for CONFIG_BSD etc. */
+@@ -288,9 +290,10 @@ void os_setup_post(void)
+ 
+         dup2(fd, 0);
+         dup2(fd, 1);
+-        /* In case -D is given do not redirect stderr to /dev/null */
++        /* In case -D is given do not redirect stderr to journal */
+         if (!qemu_logfile) {
+-            dup2(fd, 2);
++            int journal_fd = sd_journal_stream_fd("QEMU", LOG_ERR, 0);
++            dup2(journal_fd, 2);
+         }
+ 
+         close(fd);

debian/patches/pve/0036-PVE-Add-sequential-job-transaction-support.patch

@@ -0,0 +1,97 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Thu, 20 Aug 2020 14:31:59 +0200
+Subject: [PATCH] PVE: Add sequential job transaction support
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ include/qemu/job.h | 12 ++++++++++++
+ job.c              | 31 +++++++++++++++++++++++++++++++
+ 2 files changed, 43 insertions(+)
+
+diff --git a/include/qemu/job.h b/include/qemu/job.h
+index 32aabb1c60..f7a6a0926a 100644
+--- a/include/qemu/job.h
++++ b/include/qemu/job.h
+@@ -280,6 +280,18 @@ typedef enum JobCreateFlags {
+  */
+ JobTxn *job_txn_new(void);
+ 
++/**
++ * Create a new transaction and set it to sequential mode, i.e. run all jobs
++ * one after the other instead of at the same time.
++ */
++JobTxn *job_txn_new_seq(void);
++
++/**
++ * Helper method to start the first job in a sequential transaction to kick it
++ * off. Other jobs will be run after this one completes.
++ */
++void job_txn_start_seq(JobTxn *txn);
++
+ /**
+  * 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 f9884e7d9d..05b7797e82 100644
+--- a/job.c
++++ b/job.c
+@@ -72,6 +72,8 @@ struct JobTxn {
+ 
+     /* Reference count */
+     int refcnt;
++
++    bool sequential;
+ };
+ 
+ /* Right now, this mutex is only needed to synchronize accesses to job->busy
+@@ -102,6 +104,25 @@ JobTxn *job_txn_new(void)
+     return txn;
+ }
+ 
++JobTxn *job_txn_new_seq(void)
++{
++    JobTxn *txn = job_txn_new();
++    txn->sequential = true;
++    return txn;
++}
++
++void job_txn_start_seq(JobTxn *txn)
++{
++    assert(txn->sequential);
++    assert(!txn->aborting);
++
++    Job *first = QLIST_FIRST(&txn->jobs);
++    assert(first);
++    assert(first->status == JOB_STATUS_CREATED);
++
++    job_start(first);
++}
++
+ static void job_txn_ref(JobTxn *txn)
+ {
+     txn->refcnt++;
+@@ -841,6 +862,9 @@ static void job_completed_txn_success(Job *job)
+      */
+     QLIST_FOREACH(other_job, &txn->jobs, txn_list) {
+         if (!job_is_completed(other_job)) {
++            if (txn->sequential) {
++                job_start(other_job);
++            }
+             return;
+         }
+         assert(other_job->ret == 0);
+@@ -1011,6 +1035,13 @@ int job_finish_sync(Job *job, void (*finish)(Job *, Error **errp), Error **errp)
+         return -EBUSY;
+     }
+ 
++    /* in a sequential transaction jobs with status CREATED can appear at time
++     * of cancelling, these have not begun work so job_enter won't do anything,
++     * let's ensure they are marked as ABORTING if required */
++    if (job->status == JOB_STATUS_CREATED && job->txn->sequential) {
++        job_update_rc(job);
++    }
++
+     AIO_WAIT_WHILE(job->aio_context,
+                    (job_enter(job), !job_is_completed(job)));
+ 

debian/patches/pve/0037-PVE-Backup-Use-a-transaction-to-synchronize-job-stat.patch

@@ -0,0 +1,292 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Thu, 20 Aug 2020 14:25:00 +0200
+Subject: [PATCH] PVE-Backup: Use a transaction to synchronize job states
+
+By using a JobTxn, we can sync dirty bitmaps only when *all* jobs were
+successful - meaning we don't need to remove them when the backup fails,
+since QEMU's BITMAP_SYNC_MODE_ON_SUCCESS will now handle that for us.
+
+To keep the rate-limiting and IO impact from before, we use a sequential
+transaction, so drives will still be backed up one after the other.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ pve-backup.c | 167 +++++++++++++++------------------------------------
+ 1 file changed, 49 insertions(+), 118 deletions(-)
+
+diff --git a/pve-backup.c b/pve-backup.c
+index 8305105fd5..d7f2b2206f 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -52,6 +52,7 @@ static struct PVEBackupState {
+     VmaWriter *vmaw;
+     ProxmoxBackupHandle *pbs;
+     GList *di_list;
++    JobTxn *txn;
+     QemuMutex backup_mutex;
+     CoMutex dump_callback_mutex;
+ } backup_state;
+@@ -71,32 +72,12 @@ typedef struct PVEBackupDevInfo {
+     size_t size;
+     uint64_t block_size;
+     uint8_t dev_id;
+-    bool completed;
+     char targetfile[PATH_MAX];
+     BdrvDirtyBitmap *bitmap;
+     BlockDriverState *target;
++    BlockJob *job;
+ } PVEBackupDevInfo;
+ 
+-static void pvebackup_run_next_job(void);
+-
+-static BlockJob *
+-lookup_active_block_job(PVEBackupDevInfo *di)
+-{
+-    if (!di->completed && di->bs) {
+-        for (BlockJob *job = block_job_next(NULL); job; job = block_job_next(job)) {
+-            if (job->job.driver->job_type != JOB_TYPE_BACKUP) {
+-                continue;
+-            }
+-
+-            BackupBlockJob *bjob = container_of(job, BackupBlockJob, common);
+-            if (bjob && bjob->source_bs == di->bs) {
+-                return job;
+-            }
+-        }
+-    }
+-    return NULL;
+-}
+-
+ static void pvebackup_propagate_error(Error *err)
+ {
+     qemu_mutex_lock(&backup_state.stat.lock);
+@@ -272,18 +253,6 @@ static void coroutine_fn pvebackup_co_cleanup(void *unused)
+             if (local_err != NULL) {
+                 pvebackup_propagate_error(local_err);
+             }
+-        } else {
+-            // on error or cancel we cannot ensure synchronization of dirty
+-            // bitmaps with backup server, so remove all and do full backup next
+-            GList *l = backup_state.di_list;
+-            while (l) {
+-                PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+-                l = g_list_next(l);
+-
+-                if (di->bitmap) {
+-                    bdrv_release_dirty_bitmap(di->bitmap);
+-                }
+-            }
+         }
+ 
+         proxmox_backup_disconnect(backup_state.pbs);
+@@ -322,8 +291,6 @@ static void pvebackup_complete_cb(void *opaque, int ret)
+ 
+     qemu_mutex_lock(&backup_state.backup_mutex);
+ 
+-    di->completed = true;
+-
+     if (ret < 0) {
+         Error *local_err = NULL;
+         error_setg(&local_err, "job failed with err %d - %s", ret, strerror(-ret));
+@@ -336,20 +303,17 @@ static void pvebackup_complete_cb(void *opaque, int ret)
+ 
+     block_on_coroutine_fn(pvebackup_complete_stream, di);
+ 
+-    // remove self from job queue
++    // remove self from job list
+     backup_state.di_list = g_list_remove(backup_state.di_list, di);
+ 
+-    if (di->bitmap && ret < 0) {
+-        // on error or cancel we cannot ensure synchronization of dirty
+-        // bitmaps with backup server, so remove all and do full backup next
+-        bdrv_release_dirty_bitmap(di->bitmap);
+-    }
+-
+     g_free(di);
+ 
+-    qemu_mutex_unlock(&backup_state.backup_mutex);
++    /* call cleanup if we're the last job */
++    if (!g_list_first(backup_state.di_list)) {
++        block_on_coroutine_fn(pvebackup_co_cleanup, NULL);
++    }
+ 
+-    pvebackup_run_next_job();
++    qemu_mutex_unlock(&backup_state.backup_mutex);
+ }
+ 
+ static void pvebackup_cancel(void)
+@@ -371,36 +335,28 @@ static void pvebackup_cancel(void)
+         proxmox_backup_abort(backup_state.pbs, "backup canceled");
+     }
+ 
+-    qemu_mutex_unlock(&backup_state.backup_mutex);
+-
+-    for(;;) {
+-
+-        BlockJob *next_job = NULL;
+-
+-        qemu_mutex_lock(&backup_state.backup_mutex);
+-
+-        GList *l = backup_state.di_list;
+-        while (l) {
+-            PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+-            l = g_list_next(l);
++    /* it's enough to cancel one job in the transaction, the rest will follow
++     * automatically */
++    GList *bdi = g_list_first(backup_state.di_list);
++    BlockJob *cancel_job = bdi && bdi->data ?
++        ((PVEBackupDevInfo *)bdi->data)->job :
++        NULL;
+ 
+-            BlockJob *job = lookup_active_block_job(di);
+-            if (job != NULL) {
+-                next_job = job;
+-                break;
+-            }
+-        }
++    /* ref the job before releasing the mutex, just to be safe */
++    if (cancel_job) {
++        job_ref(&cancel_job->job);
++    }
+ 
+-        qemu_mutex_unlock(&backup_state.backup_mutex);
++    /* job_cancel_sync may enter the job, so we need to release the
++     * backup_mutex to avoid deadlock */
++    qemu_mutex_unlock(&backup_state.backup_mutex);
+ 
+-        if (next_job) {
+-            AioContext *aio_context = next_job->job.aio_context;
+-            aio_context_acquire(aio_context);
+-            job_cancel_sync(&next_job->job);
+-            aio_context_release(aio_context);
+-        } else {
+-            break;
+-        }
++    if (cancel_job) {
++        AioContext *aio_context = cancel_job->job.aio_context;
++        aio_context_acquire(aio_context);
++        job_cancel_sync(&cancel_job->job);
++        job_unref(&cancel_job->job);
++        aio_context_release(aio_context);
+     }
+ }
+ 
+@@ -459,51 +415,19 @@ static int coroutine_fn pvebackup_co_add_config(
+     goto out;
+ }
+ 
+-bool job_should_pause(Job *job);
+-
+-static void pvebackup_run_next_job(void)
+-{
+-    assert(!qemu_in_coroutine());
+-
+-    qemu_mutex_lock(&backup_state.backup_mutex);
+-
+-    GList *l = backup_state.di_list;
+-    while (l) {
+-        PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+-        l = g_list_next(l);
+-
+-        BlockJob *job = lookup_active_block_job(di);
+-
+-        if (job) {
+-            qemu_mutex_unlock(&backup_state.backup_mutex);
+-
+-            AioContext *aio_context = job->job.aio_context;
+-            aio_context_acquire(aio_context);
+-
+-            if (job_should_pause(&job->job)) {
+-                bool error_or_canceled = pvebackup_error_or_canceled();
+-                if (error_or_canceled) {
+-                    job_cancel_sync(&job->job);
+-                } else {
+-                    job_resume(&job->job);
+-                }
+-            }
+-            aio_context_release(aio_context);
+-            return;
+-        }
+-    }
+-
+-    block_on_coroutine_fn(pvebackup_co_cleanup, NULL); // no more jobs, run cleanup
+-
+-    qemu_mutex_unlock(&backup_state.backup_mutex);
+-}
+-
+ static bool create_backup_jobs(void) {
+ 
+     assert(!qemu_in_coroutine());
+ 
+     Error *local_err = NULL;
+ 
++    /* create job transaction to synchronize bitmap commit and cancel all
++     * jobs in case one errors */
++    if (backup_state.txn) {
++        job_txn_unref(backup_state.txn);
++    }
++    backup_state.txn = job_txn_new_seq();
++
+     /* create and start all jobs (paused state) */
+     GList *l =  backup_state.di_list;
+     while (l) {
+@@ -524,7 +448,7 @@ static bool create_backup_jobs(void) {
+         BlockJob *job = backup_job_create(
+             NULL, di->bs, di->target, backup_state.speed, sync_mode, di->bitmap,
+             bitmap_mode, false, NULL, BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
+-            JOB_DEFAULT, pvebackup_complete_cb, di, 1, NULL, &local_err);
++            JOB_DEFAULT, pvebackup_complete_cb, di, backup_state.txn, &local_err);
+ 
+         aio_context_release(aio_context);
+ 
+@@ -536,7 +460,8 @@ static bool create_backup_jobs(void) {
+             pvebackup_propagate_error(create_job_err);
+             break;
+         }
+-        job_start(&job->job);
++
++        di->job = job;
+ 
+         bdrv_unref(di->target);
+         di->target = NULL;
+@@ -554,6 +479,10 @@ static bool create_backup_jobs(void) {
+                 bdrv_unref(di->target);
+                 di->target = NULL;
+             }
++
++            if (di->job) {
++                job_unref(&di->job->job);
++            }
+         }
+     }
+ 
+@@ -944,10 +873,6 @@ err:
+         PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+         l = g_list_next(l);
+ 
+-        if (di->bitmap) {
+-            bdrv_release_dirty_bitmap(di->bitmap);
+-        }
+-
+         if (di->target) {
+             bdrv_unref(di->target);
+         }
+@@ -1036,9 +961,15 @@ UuidInfo *qmp_backup(
+     block_on_coroutine_fn(pvebackup_co_prepare, &task);
+ 
+     if (*errp == NULL) {
+-        create_backup_jobs();
++        bool errors = create_backup_jobs();
+         qemu_mutex_unlock(&backup_state.backup_mutex);
+-        pvebackup_run_next_job();
++
++        if (!errors) {
++            /* start the first job in the transaction
++             * note: this might directly enter the job, so we need to do this
++             * after unlocking the backup_mutex */
++            job_txn_start_seq(backup_state.txn);
++        }
+     } else {
+         qemu_mutex_unlock(&backup_state.backup_mutex);
+     }

debian/patches/pve/0038-PVE-Backup-Don-t-block-on-finishing-and-cleanup-crea.patch

@@ -0,0 +1,500 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Mon, 28 Sep 2020 13:40:51 +0200
+Subject: [PATCH] PVE-Backup: Don't block on finishing and cleanup
+ create_backup_jobs
+
+proxmox_backup_co_finish is already async, but previously we would wait
+for the coroutine using block_on_coroutine_fn(). Avoid this by
+scheduling pvebackup_co_complete_stream (and thus pvebackup_co_cleanup)
+as a real coroutine when calling from pvebackup_complete_cb. This is ok,
+since complete_stream uses the backup_mutex internally to synchronize,
+and other streams can happily continue writing in the meantime anyway.
+
+To accomodate, backup_mutex is converted to a CoMutex. This means
+converting every user to a coroutine. This is not just useful here, but
+will come in handy once this series[0] is merged, and QMP calls can be
+yield-able coroutines too. Then we can also finally get rid of
+block_on_coroutine_fn.
+
+Cases of aio_context_acquire/release from within what is now a coroutine
+are changed to aio_co_reschedule_self, which works since a running
+coroutine always holds the aio lock for the context it is running in.
+
+job_cancel_sync is called from a BH since it can't be run from a
+coroutine (uses AIO_WAIT_WHILE internally).
+
+Same thing for create_backup_jobs, which is converted to a BH too.
+
+To communicate the finishing state, a new property is introduced to
+query-backup: 'finishing'. A new state is explicitly not used, since
+that would break compatibility with older qemu-server versions.
+
+Also fix create_backup_jobs:
+
+No more weird bool returns, just the standard "errp" format used
+everywhere else too. With this, if backup_job_create fails, the error
+message is actually returned over QMP and can be shown to the user.
+
+To facilitate correct cleanup on such an error, we call
+create_backup_jobs as a bottom half directly from pvebackup_co_prepare.
+This additionally allows us to actually hold the backup_mutex during
+operation.
+
+Also add a job_cancel_sync before job_unref, since a job must be in
+STATUS_NULL to be deleted by unref, which could trigger an assert
+before.
+
+[0] https://lists.gnu.org/archive/html/qemu-devel/2020-09/msg03515.html
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ pve-backup.c         | 217 ++++++++++++++++++++++++++++---------------
+ qapi/block-core.json |   5 +-
+ 2 files changed, 144 insertions(+), 78 deletions(-)
+
+diff --git a/pve-backup.c b/pve-backup.c
+index d7f2b2206f..e671ed8d48 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -33,7 +33,9 @@ const char *PBS_BITMAP_NAME = "pbs-incremental-dirty-bitmap";
+ 
+ static struct PVEBackupState {
+     struct {
+-        // Everithing accessed from qmp_backup_query command is protected using lock
++        // Everything accessed from qmp_backup_query command is protected using
++        // this lock. Do NOT hold this lock for long times, as it is sometimes
++        // acquired from coroutines, and thus any wait time may block the guest.
+         QemuMutex lock;
+         Error *error;
+         time_t start_time;
+@@ -47,20 +49,22 @@ static struct PVEBackupState {
+         size_t reused;
+         size_t zero_bytes;
+         GList *bitmap_list;
++        bool finishing;
++        bool starting;
+     } stat;
+     int64_t speed;
+     VmaWriter *vmaw;
+     ProxmoxBackupHandle *pbs;
+     GList *di_list;
+     JobTxn *txn;
+-    QemuMutex backup_mutex;
++    CoMutex backup_mutex;
+     CoMutex dump_callback_mutex;
+ } backup_state;
+ 
+ static void pvebackup_init(void)
+ {
+     qemu_mutex_init(&backup_state.stat.lock);
+-    qemu_mutex_init(&backup_state.backup_mutex);
++    qemu_co_mutex_init(&backup_state.backup_mutex);
+     qemu_co_mutex_init(&backup_state.dump_callback_mutex);
+ }
+ 
+@@ -72,6 +76,7 @@ typedef struct PVEBackupDevInfo {
+     size_t size;
+     uint64_t block_size;
+     uint8_t dev_id;
++    int completed_ret; // INT_MAX if not completed
+     char targetfile[PATH_MAX];
+     BdrvDirtyBitmap *bitmap;
+     BlockDriverState *target;
+@@ -227,12 +232,12 @@ pvebackup_co_dump_vma_cb(
+ }
+ 
+ // assumes the caller holds backup_mutex
+-static void coroutine_fn pvebackup_co_cleanup(void *unused)
++static void coroutine_fn pvebackup_co_cleanup(void)
+ {
+     assert(qemu_in_coroutine());
+ 
+     qemu_mutex_lock(&backup_state.stat.lock);
+-    backup_state.stat.end_time = time(NULL);
++    backup_state.stat.finishing = true;
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+     if (backup_state.vmaw) {
+@@ -261,35 +266,29 @@ static void coroutine_fn pvebackup_co_cleanup(void *unused)
+ 
+     g_list_free(backup_state.di_list);
+     backup_state.di_list = NULL;
++
++    qemu_mutex_lock(&backup_state.stat.lock);
++    backup_state.stat.end_time = time(NULL);
++    backup_state.stat.finishing = false;
++    qemu_mutex_unlock(&backup_state.stat.lock);
+ }
+ 
+-// assumes the caller holds backup_mutex
+-static void coroutine_fn pvebackup_complete_stream(void *opaque)
++static void coroutine_fn pvebackup_co_complete_stream(void *opaque)
+ {
+     PVEBackupDevInfo *di = opaque;
++    int ret = di->completed_ret;
+ 
+-    bool error_or_canceled = pvebackup_error_or_canceled();
+-
+-    if (backup_state.vmaw) {
+-        vma_writer_close_stream(backup_state.vmaw, di->dev_id);
++    qemu_mutex_lock(&backup_state.stat.lock);
++    bool starting = backup_state.stat.starting;
++    qemu_mutex_unlock(&backup_state.stat.lock);
++    if (starting) {
++        /* in 'starting' state, no tasks have been run yet, meaning we can (and
++         * must) skip all cleanup, as we don't know what has and hasn't been
++         * initialized yet. */
++        return;
+     }
+ 
+-    if (backup_state.pbs && !error_or_canceled) {
+-        Error *local_err = NULL;
+-        proxmox_backup_co_close_image(backup_state.pbs, di->dev_id, &local_err);
+-        if (local_err != NULL) {
+-            pvebackup_propagate_error(local_err);
+-        }
+-    }
+-}
+-
+-static void pvebackup_complete_cb(void *opaque, int ret)
+-{
+-    assert(!qemu_in_coroutine());
+-
+-    PVEBackupDevInfo *di = opaque;
+-
+-    qemu_mutex_lock(&backup_state.backup_mutex);
++    qemu_co_mutex_lock(&backup_state.backup_mutex);
+ 
+     if (ret < 0) {
+         Error *local_err = NULL;
+@@ -301,7 +300,19 @@ static void pvebackup_complete_cb(void *opaque, int ret)
+ 
+     assert(di->target == NULL);
+ 
+-    block_on_coroutine_fn(pvebackup_complete_stream, di);
++    bool error_or_canceled = pvebackup_error_or_canceled();
++
++    if (backup_state.vmaw) {
++        vma_writer_close_stream(backup_state.vmaw, di->dev_id);
++    }
++
++    if (backup_state.pbs && !error_or_canceled) {
++        Error *local_err = NULL;
++        proxmox_backup_co_close_image(backup_state.pbs, di->dev_id, &local_err);
++        if (local_err != NULL) {
++            pvebackup_propagate_error(local_err);
++        }
++    }
+ 
+     // remove self from job list
+     backup_state.di_list = g_list_remove(backup_state.di_list, di);
+@@ -310,21 +321,49 @@ static void pvebackup_complete_cb(void *opaque, int ret)
+ 
+     /* call cleanup if we're the last job */
+     if (!g_list_first(backup_state.di_list)) {
+-        block_on_coroutine_fn(pvebackup_co_cleanup, NULL);
++        pvebackup_co_cleanup();
+     }
+ 
+-    qemu_mutex_unlock(&backup_state.backup_mutex);
++    qemu_co_mutex_unlock(&backup_state.backup_mutex);
+ }
+ 
+-static void pvebackup_cancel(void)
++static void pvebackup_complete_cb(void *opaque, int ret)
+ {
+-    assert(!qemu_in_coroutine());
++    PVEBackupDevInfo *di = opaque;
++    di->completed_ret = ret;
++
++    /*
++     * 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.
++     * Note: di is a pointer to an entry in the global backup_state struct, so
++     * it stays valid.
++     */
++    Coroutine *co = qemu_coroutine_create(pvebackup_co_complete_stream, di);
++    aio_co_enter(qemu_get_aio_context(), co);
++}
++
++/*
++ * job_cancel(_sync) does not like to be called from coroutines, so defer to
++ * main loop processing via a bottom half.
++ */
++static void job_cancel_bh(void *opaque) {
++    CoCtxData *data = (CoCtxData*)opaque;
++    Job *job = (Job*)data->data;
++    AioContext *job_ctx = job->aio_context;
++    aio_context_acquire(job_ctx);
++    job_cancel_sync(job);
++    aio_context_release(job_ctx);
++    aio_co_enter(data->ctx, data->co);
++}
+ 
++static void coroutine_fn pvebackup_co_cancel(void *opaque)
++{
+     Error *cancel_err = NULL;
+     error_setg(&cancel_err, "backup canceled");
+     pvebackup_propagate_error(cancel_err);
+ 
+-    qemu_mutex_lock(&backup_state.backup_mutex);
++    qemu_co_mutex_lock(&backup_state.backup_mutex);
+ 
+     if (backup_state.vmaw) {
+         /* make sure vma writer does not block anymore */
+@@ -342,27 +381,22 @@ static void pvebackup_cancel(void)
+         ((PVEBackupDevInfo *)bdi->data)->job :
+         NULL;
+ 
+-    /* ref the job before releasing the mutex, just to be safe */
+     if (cancel_job) {
+-        job_ref(&cancel_job->job);
++        CoCtxData data = {
++            .ctx = qemu_get_current_aio_context(),
++            .co = qemu_coroutine_self(),
++            .data = &cancel_job->job,
++        };
++        aio_bh_schedule_oneshot(data.ctx, job_cancel_bh, &data);
++        qemu_coroutine_yield();
+     }
+ 
+-    /* job_cancel_sync may enter the job, so we need to release the
+-     * backup_mutex to avoid deadlock */
+-    qemu_mutex_unlock(&backup_state.backup_mutex);
+-
+-    if (cancel_job) {
+-        AioContext *aio_context = cancel_job->job.aio_context;
+-        aio_context_acquire(aio_context);
+-        job_cancel_sync(&cancel_job->job);
+-        job_unref(&cancel_job->job);
+-        aio_context_release(aio_context);
+-    }
++    qemu_co_mutex_unlock(&backup_state.backup_mutex);
+ }
+ 
+ void qmp_backup_cancel(Error **errp)
+ {
+-    pvebackup_cancel();
++    block_on_coroutine_fn(pvebackup_co_cancel, NULL);
+ }
+ 
+ // assumes the caller holds backup_mutex
+@@ -415,10 +449,18 @@ static int coroutine_fn pvebackup_co_add_config(
+     goto out;
+ }
+ 
+-static bool create_backup_jobs(void) {
++/*
++ * backup_job_create can *not* be run from a coroutine (and requires an
++ * acquired AioContext), so this can't either.
++ * The caller is responsible that backup_mutex is held nonetheless.
++ */
++static void create_backup_jobs_bh(void *opaque) {
+ 
+     assert(!qemu_in_coroutine());
+ 
++    CoCtxData *data = (CoCtxData*)opaque;
++    Error **errp = (Error**)data->data;
++
+     Error *local_err = NULL;
+ 
+     /* create job transaction to synchronize bitmap commit and cancel all
+@@ -452,24 +494,19 @@ static bool create_backup_jobs(void) {
+ 
+         aio_context_release(aio_context);
+ 
+-        if (!job || local_err != NULL) {
+-            Error *create_job_err = NULL;
+-            error_setg(&create_job_err, "backup_job_create failed: %s",
+-                       local_err ? error_get_pretty(local_err) : "null");
++        di->job = job;
+ 
+-            pvebackup_propagate_error(create_job_err);
++        if (!job || local_err) {
++            error_setg(errp, "backup_job_create failed: %s",
++                       local_err ? error_get_pretty(local_err) : "null");
+             break;
+         }
+ 
+-        di->job = job;
+-
+         bdrv_unref(di->target);
+         di->target = NULL;
+     }
+ 
+-    bool errors = pvebackup_error_or_canceled();
+-
+-    if (errors) {
++    if (*errp) {
+         l = backup_state.di_list;
+         while (l) {
+             PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+@@ -481,12 +518,17 @@ static bool create_backup_jobs(void) {
+             }
+ 
+             if (di->job) {
++                AioContext *ctx = di->job->job.aio_context;
++                aio_context_acquire(ctx);
++                job_cancel_sync(&di->job->job);
+                 job_unref(&di->job->job);
++                aio_context_release(ctx);
+             }
+         }
+     }
+ 
+-    return errors;
++    /* return */
++    aio_co_enter(data->ctx, data->co);
+ }
+ 
+ typedef struct QmpBackupTask {
+@@ -523,11 +565,12 @@ typedef struct QmpBackupTask {
+     UuidInfo *result;
+ } QmpBackupTask;
+ 
+-// assumes the caller holds backup_mutex
+ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+ {
+     assert(qemu_in_coroutine());
+ 
++    qemu_co_mutex_lock(&backup_state.backup_mutex);
++
+     QmpBackupTask *task = opaque;
+ 
+     task->result = NULL; // just to be sure
+@@ -548,8 +591,9 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     const char *firewall_name = "qemu-server.fw";
+ 
+     if (backup_state.di_list) {
+-         error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
++        error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
+                   "previous backup not finished");
++        qemu_co_mutex_unlock(&backup_state.backup_mutex);
+         return;
+     }
+ 
+@@ -616,6 +660,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+         }
+         di->size = size;
+         total += size;
++
++        di->completed_ret = INT_MAX;
+     }
+ 
+     uuid_generate(uuid);
+@@ -847,6 +893,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     backup_state.stat.dirty = total - backup_state.stat.reused;
+     backup_state.stat.transferred = 0;
+     backup_state.stat.zero_bytes = 0;
++    backup_state.stat.finishing = false;
++    backup_state.stat.starting = true;
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+@@ -861,6 +909,33 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     uuid_info->UUID = uuid_str;
+ 
+     task->result = uuid_info;
++
++    /* Run create_backup_jobs_bh outside of coroutine (in BH) but keep
++    * backup_mutex locked. This is fine, a CoMutex can be held across yield
++    * points, and we'll release it as soon as the BH reschedules us.
++    */
++    CoCtxData waker = {
++        .co = qemu_coroutine_self(),
++        .ctx = qemu_get_current_aio_context(),
++        .data = &local_err,
++    };
++    aio_bh_schedule_oneshot(waker.ctx, create_backup_jobs_bh, &waker);
++    qemu_coroutine_yield();
++
++    if (local_err) {
++        error_propagate(task->errp, local_err);
++        goto err;
++    }
++
++    qemu_co_mutex_unlock(&backup_state.backup_mutex);
++
++    qemu_mutex_lock(&backup_state.stat.lock);
++    backup_state.stat.starting = false;
++    qemu_mutex_unlock(&backup_state.stat.lock);
++
++    /* start the first job in the transaction */
++    job_txn_start_seq(backup_state.txn);
++
+     return;
+ 
+ err_mutex:
+@@ -883,6 +958,7 @@ err:
+         g_free(di);
+     }
+     g_list_free(di_list);
++    backup_state.di_list = NULL;
+ 
+     if (devs) {
+         g_strfreev(devs);
+@@ -903,6 +979,8 @@ err:
+     }
+ 
+     task->result = NULL;
++
++    qemu_co_mutex_unlock(&backup_state.backup_mutex);
+     return;
+ }
+ 
+@@ -956,24 +1034,8 @@ UuidInfo *qmp_backup(
+         .errp = errp,
+     };
+ 
+-    qemu_mutex_lock(&backup_state.backup_mutex);
+-
+     block_on_coroutine_fn(pvebackup_co_prepare, &task);
+ 
+-    if (*errp == NULL) {
+-        bool errors = create_backup_jobs();
+-        qemu_mutex_unlock(&backup_state.backup_mutex);
+-
+-        if (!errors) {
+-            /* start the first job in the transaction
+-             * note: this might directly enter the job, so we need to do this
+-             * after unlocking the backup_mutex */
+-            job_txn_start_seq(backup_state.txn);
+-        }
+-    } else {
+-        qemu_mutex_unlock(&backup_state.backup_mutex);
+-    }
+-
+     return task.result;
+ }
+ 
+@@ -1025,6 +1087,7 @@ BackupStatus *qmp_query_backup(Error **errp)
+     info->transferred = backup_state.stat.transferred;
+     info->has_reused = true;
+     info->reused = backup_state.stat.reused;
++    info->finishing = backup_state.stat.finishing;
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index f57fda122c..9b827cbe43 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -775,12 +775,15 @@
+ #
+ # @uuid: uuid for this backup job
+ #
++# @finishing: if status='active' and finishing=true, then the backup process is
++#             waiting for the target to finish.
++#
+ ##
+ { 'struct': 'BackupStatus',
+   'data': {'*status': 'str', '*errmsg': 'str', '*total': 'int', '*dirty': 'int',
+            '*transferred': 'int', '*zero-bytes': 'int', '*reused': 'int',
+            '*start-time': 'int', '*end-time': 'int',
+-           '*backup-file': 'str', '*uuid': 'str' } }
++           '*backup-file': 'str', '*uuid': 'str', 'finishing': 'bool' } }
+ 
+ ##
+ # @BackupFormat:

debian/patches/pve/0039-PVE-Migrate-dirty-bitmap-state-via-savevm.patch

@@ -0,0 +1,211 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Thu, 22 Oct 2020 17:34:18 +0200
+Subject: [PATCH] PVE: Migrate dirty bitmap state via savevm
+
+QEMU provides 'savevm' registrations as a mechanism for arbitrary state
+to be migrated along with a VM. Use this to send a serialized version of
+dirty bitmap state data from proxmox-backup-qemu, and restore it on the
+target node.
+
+Also add a flag to query-proxmox-support so qemu-server can determine if
+safe migration is possible and makes sense.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ include/migration/misc.h |   3 ++
+ migration/meson.build    |   2 +
+ migration/pbs-state.c    | 106 +++++++++++++++++++++++++++++++++++++++
+ pve-backup.c             |   1 +
+ qapi/block-core.json     |   6 +++
+ softmmu/vl.c             |   1 +
+ 6 files changed, 119 insertions(+)
+ create mode 100644 migration/pbs-state.c
+
+diff --git a/include/migration/misc.h b/include/migration/misc.h
+index 34e7d75713..f83816dd3c 100644
+--- a/include/migration/misc.h
++++ b/include/migration/misc.h
+@@ -75,4 +75,7 @@ bool migration_in_incoming_postcopy(void);
+ /* migration/block-dirty-bitmap.c */
+ void dirty_bitmap_mig_init(void);
+ 
++/* migration/pbs-state.c */
++void pbs_state_mig_init(void);
++
+ #endif
+diff --git a/migration/meson.build b/migration/meson.build
+index e62b79b60f..b90a04aa75 100644
+--- a/migration/meson.build
++++ b/migration/meson.build
+@@ -7,8 +7,10 @@ migration_files = files(
+   'qemu-file-channel.c',
+   'qemu-file.c',
+   'qjson.c',
++  'pbs-state.c',
+ )
+ softmmu_ss.add(migration_files)
++softmmu_ss.add(libproxmox_backup_qemu)
+ 
+ softmmu_ss.add(files(
+   'block-dirty-bitmap.c',
+diff --git a/migration/pbs-state.c b/migration/pbs-state.c
+new file mode 100644
+index 0000000000..29f2b3860d
+--- /dev/null
++++ b/migration/pbs-state.c
+@@ -0,0 +1,106 @@
++/*
++ * PBS (dirty-bitmap) state migration
++ */
++
++#include "qemu/osdep.h"
++#include "migration/misc.h"
++#include "qemu-file.h"
++#include "migration/vmstate.h"
++#include "migration/register.h"
++#include "proxmox-backup-qemu.h"
++
++typedef struct PBSState {
++    bool active;
++} PBSState;
++
++/* state is accessed via this static variable directly, 'opaque' is NULL */
++static PBSState pbs_state;
++
++static void pbs_state_save_pending(QEMUFile *f, void *opaque,
++                                      uint64_t max_size,
++                                      uint64_t *res_precopy_only,
++                                      uint64_t *res_compatible,
++                                      uint64_t *res_postcopy_only)
++{
++    /* we send everything in save_setup, so nothing is ever pending */
++}
++
++/* receive PBS state via f and deserialize, called on target */
++static int pbs_state_load(QEMUFile *f, void *opaque, int version_id)
++{
++    /* safe cast, we cannot migrate to target with less bits than source */
++    size_t buf_size = (size_t)qemu_get_be64(f);
++
++    uint8_t *buf = (uint8_t *)malloc(buf_size);
++    size_t read = qemu_get_buffer(f, buf, buf_size);
++
++    if (read < buf_size) {
++        fprintf(stderr, "error receiving PBS state: not enough data\n");
++        return -EIO;
++    }
++
++    proxmox_import_state(buf, buf_size);
++
++    free(buf);
++    return 0;
++}
++
++/* serialize PBS state and send to target via f, called on source */
++static int pbs_state_save_setup(QEMUFile *f, void *opaque)
++{
++    size_t buf_size;
++    uint8_t *buf = proxmox_export_state(&buf_size);
++
++    /* LV encoding */
++    qemu_put_be64(f, buf_size);
++    qemu_put_buffer(f, buf, buf_size);
++
++    proxmox_free_state_buf(buf);
++    pbs_state.active = false;
++    return 0;
++}
++
++static bool pbs_state_is_active(void *opaque)
++{
++    /* we need to return active exactly once, else .save_setup is never called,
++     * but if we'd just return true the migration doesn't make progress since
++     * it'd be waiting for us */
++    return pbs_state.active;
++}
++
++static bool pbs_state_is_active_iterate(void *opaque)
++{
++    /* we don't iterate, everything is sent in save_setup */
++    return pbs_state_is_active(opaque);
++}
++
++static bool pbs_state_has_postcopy(void *opaque)
++{
++    /* PBS state can't change during a migration (since that's blocking any
++     * potential backups), so we can copy everything before the VM is stopped */
++    return false;
++}
++
++static void pbs_state_save_cleanup(void *opaque)
++{
++    /* reset active after migration succeeds or fails */
++    pbs_state.active = false;
++}
++
++static SaveVMHandlers savevm_pbs_state_handlers = {
++    .save_setup = pbs_state_save_setup,
++    .has_postcopy = pbs_state_has_postcopy,
++    .save_live_pending = pbs_state_save_pending,
++    .is_active_iterate = pbs_state_is_active_iterate,
++    .load_state = pbs_state_load,
++    .is_active = pbs_state_is_active,
++    .save_cleanup = pbs_state_save_cleanup,
++};
++
++void pbs_state_mig_init(void)
++{
++    pbs_state.active = true;
++    register_savevm_live("pbs-state", 0, 1,
++                         &savevm_pbs_state_handlers,
++                         NULL);
++}
+diff --git a/pve-backup.c b/pve-backup.c
+index e671ed8d48..bd2647e5f3 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -1130,6 +1130,7 @@ 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;
++    ret->pbs_dirty_bitmap_migration = true;
+     ret->query_bitmap_info = true;
+     return ret;
+ }
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 9b827cbe43..30eb1262ff 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -884,6 +884,11 @@
+ # @pbs-dirty-bitmap-savevm: True if 'dirty-bitmaps' migration capability can
+ #                           safely be set for savevm-async.
+ #
++# @pbs-dirty-bitmap-migration: True if safe migration of dirty-bitmaps including
++#                              PBS state is supported. Enabling 'dirty-bitmaps'
++#                              migration cap if this is false/unset may lead
++#                              to crashes on migration!
++#
+ # @pbs-library-version: Running version of libproxmox-backup-qemu0 library.
+ #
+ ##
+@@ -891,6 +896,7 @@
+   'data': { 'pbs-dirty-bitmap': 'bool',
+             'query-bitmap-info': 'bool',
+             'pbs-dirty-bitmap-savevm': 'bool',
++            'pbs-dirty-bitmap-migration': 'bool',
+             'pbs-library-version': 'str' } }
+ 
+ ##
+diff --git a/softmmu/vl.c b/softmmu/vl.c
+index 5b5512128e..6721889fee 100644
+--- a/softmmu/vl.c
++++ b/softmmu/vl.c
+@@ -4304,6 +4304,7 @@ void qemu_init(int argc, char **argv, char **envp)
+     blk_mig_init();
+     ram_mig_init();
+     dirty_bitmap_mig_init();
++    pbs_state_mig_init();
+ 
+     qemu_opts_foreach(qemu_find_opts("mon"),
+                       mon_init_func, NULL, &error_fatal);

debian/patches/pve/0040-PVE-savevm-async-set-up-migration-state.patch

@@ -1,88 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Stefan Reiter <s.reiter@proxmox.com>
-Date: Wed, 8 Apr 2020 15:29:03 +0200
-Subject: [PATCH] PVE: savevm-async: set up migration state
-
-code mostly adapted from upstream savevm.c
-
-Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
----
- savevm-async.c | 30 ++++++++++++++++++++++++++++--
- 1 file changed, 28 insertions(+), 2 deletions(-)
-
-diff --git a/savevm-async.c b/savevm-async.c
-index 790e27ae37..a38b15d652 100644
---- a/savevm-async.c
-+++ b/savevm-async.c
-@@ -225,6 +225,7 @@ static void *process_savevm_thread(void *opaque)
- {
-     int ret;
-     int64_t maxlen;
-+    MigrationState *ms = migrate_get_current();
- 
-     rcu_register_thread();
- 
-@@ -234,8 +235,7 @@ static void *process_savevm_thread(void *opaque)
- 
-     if (ret < 0) {
-         save_snapshot_error("qemu_savevm_state_setup failed");
--        rcu_unregister_thread();
--        return NULL;
-+        goto out;
-     }
- 
-     while (snap_state.state == SAVE_STATE_ACTIVE) {
-@@ -287,6 +287,12 @@ static void *process_savevm_thread(void *opaque)
-     qemu_bh_schedule(snap_state.cleanup_bh);
-     qemu_mutex_unlock_iothread();
- 
-+out:
-+    /* set migration state accordingly and clear soon-to-be stale file */
-+    migrate_set_state(&ms->state, MIGRATION_STATUS_SETUP,
-+                      ret ? MIGRATION_STATUS_FAILED : MIGRATION_STATUS_COMPLETED);
-+    ms->to_dst_file = NULL;
-+
-     rcu_unregister_thread();
-     return NULL;
- }
-@@ -294,6 +300,7 @@ static void *process_savevm_thread(void *opaque)
- void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
- {
-     Error *local_err = NULL;
-+    MigrationState *ms = migrate_get_current();
- 
-     int bdrv_oflags = BDRV_O_RDWR | BDRV_O_RESIZE | BDRV_O_NO_FLUSH;
- 
-@@ -303,6 +310,17 @@ void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
-         return;
-     }
- 
-+    if (migration_is_running(ms->state)) {
-+        error_set(errp, ERROR_CLASS_GENERIC_ERROR, QERR_MIGRATION_ACTIVE);
-+        return;
-+    }
-+
-+    if (migrate_use_block()) {
-+        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
-+                  "Block migration and snapshots are incompatible");
-+        return;
-+    }
-+
-     /* initialize snapshot info */
-     snap_state.saved_vm_running = runstate_is_running();
-     snap_state.bs_pos = 0;
-@@ -341,6 +359,14 @@ void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
-         goto restart;
-     }
- 
-+    /*
-+     * qemu_savevm_* paths use migration code and expect a migration state.
-+     * State is cleared in process_savevm_thread, but has to be initialized
-+     * here (blocking main thread, from QMP) to avoid race conditions.
-+     */
-+    migrate_init(ms);
-+    memset(&ram_counters, 0, sizeof(ram_counters));
-+    ms->to_dst_file = snap_state.file;
- 
-     error_setg(&snap_state.blocker, "block device is in use by savevm");
-     blk_op_block_all(snap_state.target, snap_state.blocker);

debian/patches/pve/0040-migration-block-dirty-bitmap-migrate-other-bitmaps-e.patch

@@ -0,0 +1,32 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Tue, 3 Nov 2020 14:57:32 +0100
+Subject: [PATCH] migration/block-dirty-bitmap: migrate other bitmaps even if
+ one fails
+
+If the checks in bdrv_dirty_bitmap_check fail, that only means that this
+one specific bitmap cannot be migrated. That is not an error condition
+for any other bitmaps on the same block device.
+
+Fixes dirty-bitmap migration with sync=bitmap, as the bitmaps used for
+that are obviously marked as "busy", which would cause none at all to be
+transferred.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ migration/block-dirty-bitmap.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/migration/block-dirty-bitmap.c b/migration/block-dirty-bitmap.c
+index c61d382be8..26e4e5c99c 100644
+--- a/migration/block-dirty-bitmap.c
++++ b/migration/block-dirty-bitmap.c
+@@ -534,7 +534,7 @@ static int add_bitmaps_to_list(DBMSaveState *s, BlockDriverState *bs,
+ 
+         if (bdrv_dirty_bitmap_check(bitmap, BDRV_BITMAP_DEFAULT, &local_err)) {
+             error_report_err(local_err);
+-            return -1;
++            continue;
+         }
+ 
+         if (bitmap_aliases) {

debian/patches/pve/0041-PVE-Backup-avoid-use-QemuRecMutex-inside-coroutines.patch

@@ -1,211 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Dietmar Maurer <dietmar@proxmox.com>
-Date: Fri, 17 Apr 2020 08:57:47 +0200
-Subject: [PATCH] PVE Backup: avoid use QemuRecMutex inside coroutines
-
----
- pve-backup.c | 59 +++++++++++++++++++++++++++++++++-------------------
- 1 file changed, 38 insertions(+), 21 deletions(-)
-
-diff --git a/pve-backup.c b/pve-backup.c
-index 169f0c68d0..dddf430399 100644
---- a/pve-backup.c
-+++ b/pve-backup.c
-@@ -11,6 +11,23 @@
- 
- /* PVE backup state and related function */
- 
-+/*
-+ * Note: A resume from a qemu_coroutine_yield can happen in a different thread,
-+ * so you may not use normal mutexes within coroutines:
-+ *
-+ * ---bad-example---
-+ * qemu_rec_mutex_lock(lock)
-+ * ...
-+ * qemu_coroutine_yield() // wait for something
-+ * // we are now inside a different thread
-+ * qemu_rec_mutex_unlock(lock) // Crash - wrong thread!!
-+ * ---end-bad-example--
-+ *
-+ * ==> Always use CoMutext inside coroutines.
-+ * ==> Never acquire/release AioContext withing coroutines (because that use QemuRecMutex)
-+ *
-+ */
-+
- static struct PVEBackupState {
-     struct {
-         // Everithing accessed from qmp_backup_query command is protected using lock
-@@ -30,12 +47,14 @@ static struct PVEBackupState {
-     ProxmoxBackupHandle *pbs;
-     GList *di_list;
-     QemuRecMutex backup_mutex;
-+    CoMutex dump_callback_mutex;
- } backup_state;
- 
- static void pvebackup_init(void)
- {
-     qemu_rec_mutex_init(&backup_state.stat.lock);
-     qemu_rec_mutex_init(&backup_state.backup_mutex);
-+    qemu_co_mutex_init(&backup_state.dump_callback_mutex);
- }
- 
- // initialize PVEBackupState at startup
-@@ -114,16 +133,16 @@ pvebackup_co_dump_pbs_cb(
-     Error *local_err = NULL;
-     int pbs_res = -1;
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+    qemu_co_mutex_lock(&backup_state.dump_callback_mutex);
- 
-     // avoid deadlock if job is cancelled
-     if (pvebackup_error_or_canceled()) {
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
-         return -1;
-     }
- 
-     pbs_res = proxmox_backup_co_write_data(backup_state.pbs, di->dev_id, buf, start, size, &local_err);
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
- 
-     if (pbs_res < 0) {
-         pvebackup_propagate_error(local_err);
-@@ -149,7 +168,6 @@ pvebackup_co_dump_vma_cb(
-     const unsigned char *buf = pbuf;
-     PVEBackupDevInfo *di = opaque;
- 
--
-     int ret = -1;
- 
-     assert(backup_state.vmaw);
-@@ -167,16 +185,16 @@ pvebackup_co_dump_vma_cb(
-     }
- 
-     while (remaining > 0) {
--        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+        qemu_co_mutex_lock(&backup_state.dump_callback_mutex);
-         // avoid deadlock if job is cancelled
-         if (pvebackup_error_or_canceled()) {
--            qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+            qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
-             return -1;
-         }
- 
-         size_t zero_bytes = 0;
-         ret = vma_writer_write(backup_state.vmaw, di->dev_id, cluster_num, buf, &zero_bytes);
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_co_mutex_unlock(&backup_state.dump_callback_mutex);
- 
-         ++cluster_num;
-         if (buf) {
-@@ -203,12 +221,11 @@ pvebackup_co_dump_vma_cb(
-     return size;
- }
- 
-+// assumes the caller holds backup_mutex
- static void coroutine_fn pvebackup_co_cleanup(void *unused)
- {
-     assert(qemu_in_coroutine());
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
--
-     qemu_rec_mutex_lock(&backup_state.stat.lock);
-     backup_state.stat.end_time = time(NULL);
-     qemu_rec_mutex_unlock(&backup_state.stat.lock);
-@@ -239,9 +256,9 @@ static void coroutine_fn pvebackup_co_cleanup(void *unused)
- 
-     g_list_free(backup_state.di_list);
-     backup_state.di_list = NULL;
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- }
- 
-+// assumes the caller holds backup_mutex
- static void coroutine_fn pvebackup_complete_stream(void *opaque)
- {
-     PVEBackupDevInfo *di = opaque;
-@@ -295,6 +312,8 @@ static void pvebackup_complete_cb(void *opaque, int ret)
- 
- static void pvebackup_cancel(void)
- {
-+    assert(!qemu_in_coroutine());
-+
-     Error *cancel_err = NULL;
-     error_setg(&cancel_err, "backup canceled");
-     pvebackup_propagate_error(cancel_err);
-@@ -348,6 +367,7 @@ void qmp_backup_cancel(Error **errp)
-     pvebackup_cancel();
- }
- 
-+// assumes the caller holds backup_mutex
- static int coroutine_fn pvebackup_co_add_config(
-     const char *file,
-     const char *name,
-@@ -431,9 +451,9 @@ static void pvebackup_run_next_job(void)
-         }
-     }
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
--
-     block_on_coroutine_fn(pvebackup_co_cleanup, NULL); // no more jobs, run cleanup
-+
-+    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
- }
- 
- static bool create_backup_jobs(void) {
-@@ -520,6 +540,7 @@ typedef struct QmpBackupTask {
-     UuidInfo *result;
- } QmpBackupTask;
- 
-+// assumes the caller holds backup_mutex
- static void coroutine_fn pvebackup_co_prepare(void *opaque)
- {
-     assert(qemu_in_coroutine());
-@@ -543,11 +564,8 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
-     const char *config_name = "qemu-server.conf";
-     const char *firewall_name = "qemu-server.fw";
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
--
-     if (backup_state.di_list) {
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
--        error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
-+         error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
-                   "previous backup not finished");
-         return;
-     }
-@@ -792,8 +810,6 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
- 
-     backup_state.di_list = di_list;
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
--
-     uuid_info = g_malloc0(sizeof(*uuid_info));
-     uuid_info->UUID = uuid_str;
- 
-@@ -836,8 +852,6 @@ err:
-         rmdir(backup_dir);
-     }
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
--
-     task->result = NULL;
-     return;
- }
-@@ -881,13 +895,16 @@ UuidInfo *qmp_backup(
-         .errp = errp,
-     };
- 
-+    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+
-     block_on_coroutine_fn(pvebackup_co_prepare, &task);
- 
-     if (*errp == NULL) {
--        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-         create_backup_jobs();
-         qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-         pvebackup_run_next_job();
-+    } else {
-+        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-     }
- 
-     return task.result;

debian/patches/pve/0041-PVE-fall-back-to-open-iscsi-initiatorname.patch

@@ -0,0 +1,68 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Fabian Ebner <f.ebner@proxmox.com>
+Date: Tue, 17 Nov 2020 10:51:05 +0100
+Subject: [PATCH] PVE: fall back to open-iscsi initiatorname
+
+When no explicit option is given, try reading the initiator name from
+/etc/iscsi/initiatorname.iscsi and only use the generic fallback, i.e.
+iqn.2008-11.org.linux-kvmXXX, as a third alternative.
+
+This avoids the need to add an explicit option for vma and to explicitly set it
+for each call to qemu that deals with iSCSI disks, while still allowing to set
+the option if a different name is needed.
+
+According to RFC 3720, an initiator name is at most 223 bytes long, so the
+4 KiB buffer is big enough, even if many whitespaces are used.
+
+Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
+---
+ block/iscsi.c | 30 ++++++++++++++++++++++++++++++
+ 1 file changed, 30 insertions(+)
+
+diff --git a/block/iscsi.c b/block/iscsi.c
+index e30a7e3606..6c70bbe351 100644
+--- a/block/iscsi.c
++++ b/block/iscsi.c
+@@ -1374,12 +1374,42 @@ static char *get_initiator_name(QemuOpts *opts)
+     const char *name;
+     char *iscsi_name;
+     UuidInfo *uuid_info;
++    FILE *name_fh;
+ 
+     name = qemu_opt_get(opts, "initiator-name");
+     if (name) {
+         return g_strdup(name);
+     }
+ 
++    name_fh = fopen("/etc/iscsi/initiatorname.iscsi", "r");
++    if (name_fh) {
++        const char *key = "InitiatorName";
++        char buffer[4096];
++        char *line;
++
++        while ((line = fgets(buffer, sizeof(buffer), name_fh))) {
++            line = g_strstrip(line);
++            if (!strncmp(line, key, strlen(key))) {
++                line = strchr(line, '=');
++                if (!line || strlen(line) == 1) {
++                    continue;
++                }
++                line++;
++                g_strstrip(line);
++                if (!strlen(line)) {
++                    continue;
++                }
++                name = line;
++                break;
++            }
++        }
++        fclose(name_fh);
++
++        if (name) {
++            return g_strdup(name);
++        }
++    }
++
+     uuid_info = qmp_query_uuid(NULL);
+     if (strcmp(uuid_info->UUID, UUID_NONE) == 0) {
+         name = qemu_get_vm_name();

debian/patches/pve/0042-PVE-Backup-use-QemuMutex-instead-of-QemuRecMutex.patch

@@ -1,227 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Dietmar Maurer <dietmar@proxmox.com>
-Date: Fri, 17 Apr 2020 08:57:48 +0200
-Subject: [PATCH] PVE Backup: use QemuMutex instead of QemuRecMutex
-
-We acquire/release all mutexes outside coroutines now, so we can now
-correctly use a normal mutex.
----
- pve-backup.c | 58 ++++++++++++++++++++++++++--------------------------
- 1 file changed, 29 insertions(+), 29 deletions(-)
-
-diff --git a/pve-backup.c b/pve-backup.c
-index dddf430399..bb917ee972 100644
---- a/pve-backup.c
-+++ b/pve-backup.c
-@@ -31,7 +31,7 @@
- static struct PVEBackupState {
-     struct {
-         // Everithing accessed from qmp_backup_query command is protected using lock
--        QemuRecMutex lock;
-+        QemuMutex lock;
-         Error *error;
-         time_t start_time;
-         time_t end_time;
-@@ -46,14 +46,14 @@ static struct PVEBackupState {
-     VmaWriter *vmaw;
-     ProxmoxBackupHandle *pbs;
-     GList *di_list;
--    QemuRecMutex backup_mutex;
-+    QemuMutex backup_mutex;
-     CoMutex dump_callback_mutex;
- } backup_state;
- 
- static void pvebackup_init(void)
- {
--    qemu_rec_mutex_init(&backup_state.stat.lock);
--    qemu_rec_mutex_init(&backup_state.backup_mutex);
-+    qemu_mutex_init(&backup_state.stat.lock);
-+    qemu_mutex_init(&backup_state.backup_mutex);
-     qemu_co_mutex_init(&backup_state.dump_callback_mutex);
- }
- 
-@@ -91,26 +91,26 @@ lookup_active_block_job(PVEBackupDevInfo *di)
- 
- static void pvebackup_propagate_error(Error *err)
- {
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
-     error_propagate(&backup_state.stat.error, err);
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- }
- 
- static bool pvebackup_error_or_canceled(void)
- {
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
-     bool error_or_canceled = !!backup_state.stat.error;
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- 
-     return error_or_canceled;
- }
- 
- static void pvebackup_add_transfered_bytes(size_t transferred, size_t zero_bytes)
- {
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
-     backup_state.stat.zero_bytes += zero_bytes;
-     backup_state.stat.transferred += transferred;
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- }
- 
- // This may get called from multiple coroutines in multiple io-threads
-@@ -226,9 +226,9 @@ static void coroutine_fn pvebackup_co_cleanup(void *unused)
- {
-     assert(qemu_in_coroutine());
- 
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
-     backup_state.stat.end_time = time(NULL);
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- 
-     if (backup_state.vmaw) {
-         Error *local_err = NULL;
-@@ -284,7 +284,7 @@ static void pvebackup_complete_cb(void *opaque, int ret)
- 
-     PVEBackupDevInfo *di = opaque;
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+    qemu_mutex_lock(&backup_state.backup_mutex);
- 
-     di->completed = true;
- 
-@@ -305,7 +305,7 @@ static void pvebackup_complete_cb(void *opaque, int ret)
- 
-     g_free(di);
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_mutex_unlock(&backup_state.backup_mutex);
- 
-     pvebackup_run_next_job();
- }
-@@ -318,7 +318,7 @@ static void pvebackup_cancel(void)
-     error_setg(&cancel_err, "backup canceled");
-     pvebackup_propagate_error(cancel_err);
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+    qemu_mutex_lock(&backup_state.backup_mutex);
- 
-     if (backup_state.vmaw) {
-         /* make sure vma writer does not block anymore */
-@@ -329,13 +329,13 @@ static void pvebackup_cancel(void)
-         proxmox_backup_abort(backup_state.pbs, "backup canceled");
-     }
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_mutex_unlock(&backup_state.backup_mutex);
- 
-     for(;;) {
- 
-         BlockJob *next_job = NULL;
- 
--        qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+        qemu_mutex_lock(&backup_state.backup_mutex);
- 
-         GList *l = backup_state.di_list;
-         while (l) {
-@@ -349,7 +349,7 @@ static void pvebackup_cancel(void)
-             }
-         }
- 
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_mutex_unlock(&backup_state.backup_mutex);
- 
-         if (next_job) {
-             AioContext *aio_context = next_job->job.aio_context;
-@@ -423,7 +423,7 @@ static void pvebackup_run_next_job(void)
- {
-     assert(!qemu_in_coroutine());
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+    qemu_mutex_lock(&backup_state.backup_mutex);
- 
-     GList *l = backup_state.di_list;
-     while (l) {
-@@ -433,7 +433,7 @@ static void pvebackup_run_next_job(void)
-         BlockJob *job = lookup_active_block_job(di);
- 
-         if (job) {
--            qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+            qemu_mutex_unlock(&backup_state.backup_mutex);
- 
-             AioContext *aio_context = job->job.aio_context;
-             aio_context_acquire(aio_context);
-@@ -453,7 +453,7 @@ static void pvebackup_run_next_job(void)
- 
-     block_on_coroutine_fn(pvebackup_co_cleanup, NULL); // no more jobs, run cleanup
- 
--    qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+    qemu_mutex_unlock(&backup_state.backup_mutex);
- }
- 
- static bool create_backup_jobs(void) {
-@@ -778,7 +778,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
-     }
-     /* initialize global backup_state now */
- 
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
- 
-     if (backup_state.stat.error) {
-         error_free(backup_state.stat.error);
-@@ -801,7 +801,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
-     backup_state.stat.transferred = 0;
-     backup_state.stat.zero_bytes = 0;
- 
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- 
-     backup_state.speed = (task->has_speed && task->speed > 0) ? task->speed : 0;
- 
-@@ -895,16 +895,16 @@ UuidInfo *qmp_backup(
-         .errp = errp,
-     };
- 
--    qemu_rec_mutex_lock(&backup_state.backup_mutex);
-+    qemu_mutex_lock(&backup_state.backup_mutex);
- 
-     block_on_coroutine_fn(pvebackup_co_prepare, &task);
- 
-     if (*errp == NULL) {
-         create_backup_jobs();
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_mutex_unlock(&backup_state.backup_mutex);
-         pvebackup_run_next_job();
-     } else {
--        qemu_rec_mutex_unlock(&backup_state.backup_mutex);
-+        qemu_mutex_unlock(&backup_state.backup_mutex);
-     }
- 
-     return task.result;
-@@ -914,11 +914,11 @@ BackupStatus *qmp_query_backup(Error **errp)
- {
-     BackupStatus *info = g_malloc0(sizeof(*info));
- 
--    qemu_rec_mutex_lock(&backup_state.stat.lock);
-+    qemu_mutex_lock(&backup_state.stat.lock);
- 
-     if (!backup_state.stat.start_time) {
-         /* not started, return {} */
--        qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+        qemu_mutex_unlock(&backup_state.stat.lock);
-         return info;
-     }
- 
-@@ -955,7 +955,7 @@ BackupStatus *qmp_query_backup(Error **errp)
-     info->has_transferred = true;
-     info->transferred = backup_state.stat.transferred;
- 
--    qemu_rec_mutex_unlock(&backup_state.stat.lock);
-+    qemu_mutex_unlock(&backup_state.stat.lock);
- 
-     return info;
- }

debian/patches/pve/0042-PVE-Use-coroutine-QMP-for-backup-cancel_backup.patch

@@ -0,0 +1,597 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Tue, 26 Jan 2021 15:45:30 +0100
+Subject: [PATCH] PVE: Use coroutine QMP for backup/cancel_backup
+
+Finally turn backup QMP calls into coroutines, now that it's possible.
+This has the benefit that calls are asynchronous to the main loop, i.e.
+long running operations like connecting to a PBS server will no longer
+hang the VM.
+
+Additionally, it allows us to get rid of block_on_coroutine_fn, which
+was always a hacky workaround.
+
+While we're already spring cleaning, also remove the QmpBackupTask
+struct, since we can now put the 'prepare' function directly into
+qmp_backup and thus no longer need those giant walls of text.
+
+(Note that for our patches to work with 5.2.0 this change is actually
+required, otherwise monitor_get_fd() fails as we're not in a QMP
+coroutine, but one we start ourselves - we could of course set the
+monitor for that coroutine ourselves, but let's just fix it the right
+way instead)
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/monitor/block-hmp-cmds.c |   4 +-
+ hmp-commands.hx                |   2 +
+ proxmox-backup-client.c        |  31 -----
+ pve-backup.c                   | 232 ++++++++++-----------------------
+ qapi/block-core.json           |   4 +-
+ 5 files changed, 77 insertions(+), 196 deletions(-)
+
+diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
+index 46c63b1cf9..11c84d5508 100644
+--- a/block/monitor/block-hmp-cmds.c
++++ b/block/monitor/block-hmp-cmds.c
+@@ -1013,7 +1013,7 @@ void hmp_info_snapshots(Monitor *mon, const QDict *qdict)
+     g_free(global_snapshots);
+ }
+ 
+-void hmp_backup_cancel(Monitor *mon, const QDict *qdict)
++void coroutine_fn hmp_backup_cancel(Monitor *mon, const QDict *qdict)
+ {
+     Error *error = NULL;
+ 
+@@ -1022,7 +1022,7 @@ void hmp_backup_cancel(Monitor *mon, const QDict *qdict)
+     hmp_handle_error(mon, error);
+ }
+ 
+-void hmp_backup(Monitor *mon, const QDict *qdict)
++void coroutine_fn hmp_backup(Monitor *mon, const QDict *qdict)
+ {
+     Error *error = NULL;
+ 
+diff --git a/hmp-commands.hx b/hmp-commands.hx
+index 0c6b944850..54de3f80e6 100644
+--- a/hmp-commands.hx
++++ b/hmp-commands.hx
+@@ -108,6 +108,7 @@ ERST
+ 		    "\n\t\t\t Use -d to dump data into a directory instead"
+ 		    "\n\t\t\t of using VMA format.",
+         .cmd = hmp_backup,
++        .coroutine  = true,
+     },
+ 
+ SRST
+@@ -121,6 +122,7 @@ ERST
+         .params     = "",
+         .help       = "cancel the current VM backup",
+         .cmd        = hmp_backup_cancel,
++        .coroutine  = true,
+     },
+ 
+ SRST
+diff --git a/proxmox-backup-client.c b/proxmox-backup-client.c
+index 4ce7bc0b5e..0923037dec 100644
+--- a/proxmox-backup-client.c
++++ b/proxmox-backup-client.c
+@@ -5,37 +5,6 @@
+ 
+ /* Proxmox Backup Server client bindings using coroutines */
+ 
+-typedef struct BlockOnCoroutineWrapper {
+-    AioContext *ctx;
+-    CoroutineEntry *entry;
+-    void *entry_arg;
+-    bool finished;
+-} BlockOnCoroutineWrapper;
+-
+-static void coroutine_fn block_on_coroutine_wrapper(void *opaque)
+-{
+-    BlockOnCoroutineWrapper *wrapper = opaque;
+-    wrapper->entry(wrapper->entry_arg);
+-    wrapper->finished = true;
+-    aio_wait_kick();
+-}
+-
+-void block_on_coroutine_fn(CoroutineEntry *entry, void *entry_arg)
+-{
+-    assert(!qemu_in_coroutine());
+-
+-    AioContext *ctx = qemu_get_current_aio_context();
+-    BlockOnCoroutineWrapper wrapper = {
+-        .finished = false,
+-        .entry = entry,
+-        .entry_arg = entry_arg,
+-        .ctx = ctx,
+-    };
+-    Coroutine *wrapper_co = qemu_coroutine_create(block_on_coroutine_wrapper, &wrapper);
+-    aio_co_enter(ctx, wrapper_co);
+-    AIO_WAIT_WHILE(ctx, !wrapper.finished);
+-}
+-
+ // This is called from another thread, so we use aio_co_schedule()
+ static void proxmox_backup_schedule_wake(void *data) {
+     CoCtxData *waker = (CoCtxData *)data;
+diff --git a/pve-backup.c b/pve-backup.c
+index bd2647e5f3..dec9c0d188 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -357,7 +357,7 @@ static void job_cancel_bh(void *opaque) {
+     aio_co_enter(data->ctx, data->co);
+ }
+ 
+-static void coroutine_fn pvebackup_co_cancel(void *opaque)
++void coroutine_fn qmp_backup_cancel(Error **errp)
+ {
+     Error *cancel_err = NULL;
+     error_setg(&cancel_err, "backup canceled");
+@@ -394,11 +394,6 @@ static void coroutine_fn pvebackup_co_cancel(void *opaque)
+     qemu_co_mutex_unlock(&backup_state.backup_mutex);
+ }
+ 
+-void qmp_backup_cancel(Error **errp)
+-{
+-    block_on_coroutine_fn(pvebackup_co_cancel, NULL);
+-}
+-
+ // assumes the caller holds backup_mutex
+ static int coroutine_fn pvebackup_co_add_config(
+     const char *file,
+@@ -531,50 +526,27 @@ static void create_backup_jobs_bh(void *opaque) {
+     aio_co_enter(data->ctx, data->co);
+ }
+ 
+-typedef struct QmpBackupTask {
+-    const char *backup_file;
+-    bool has_password;
+-    const char *password;
+-    bool has_keyfile;
+-    const char *keyfile;
+-    bool has_key_password;
+-    const char *key_password;
+-    bool has_backup_id;
+-    const char *backup_id;
+-    bool has_backup_time;
+-    const char *fingerprint;
+-    bool has_fingerprint;
+-    int64_t backup_time;
+-    bool has_use_dirty_bitmap;
+-    bool use_dirty_bitmap;
+-    bool has_format;
+-    BackupFormat format;
+-    bool has_config_file;
+-    const char *config_file;
+-    bool has_firewall_file;
+-    const char *firewall_file;
+-    bool has_devlist;
+-    const char *devlist;
+-    bool has_compress;
+-    bool compress;
+-    bool has_encrypt;
+-    bool encrypt;
+-    bool has_speed;
+-    int64_t speed;
+-    Error **errp;
+-    UuidInfo *result;
+-} QmpBackupTask;
+-
+-static void coroutine_fn pvebackup_co_prepare(void *opaque)
++UuidInfo coroutine_fn *qmp_backup(
++    const char *backup_file,
++    bool has_password, const char *password,
++    bool has_keyfile, const char *keyfile,
++    bool has_key_password, const char *key_password,
++    bool has_fingerprint, const char *fingerprint,
++    bool has_backup_id, const char *backup_id,
++    bool has_backup_time, int64_t backup_time,
++    bool has_use_dirty_bitmap, bool use_dirty_bitmap,
++    bool has_compress, bool compress,
++    bool has_encrypt, bool encrypt,
++    bool has_format, BackupFormat format,
++    bool has_config_file, const char *config_file,
++    bool has_firewall_file, const char *firewall_file,
++    bool has_devlist, const char *devlist,
++    bool has_speed, int64_t speed, Error **errp)
+ {
+     assert(qemu_in_coroutine());
+ 
+     qemu_co_mutex_lock(&backup_state.backup_mutex);
+ 
+-    QmpBackupTask *task = opaque;
+-
+-    task->result = NULL; // just to be sure
+-
+     BlockBackend *blk;
+     BlockDriverState *bs = NULL;
+     const char *backup_dir = NULL;
+@@ -591,17 +563,17 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     const char *firewall_name = "qemu-server.fw";
+ 
+     if (backup_state.di_list) {
+-        error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
++        error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                   "previous backup not finished");
+         qemu_co_mutex_unlock(&backup_state.backup_mutex);
+-        return;
++        return NULL;
+     }
+ 
+     /* Todo: try to auto-detect format based on file name */
+-    BackupFormat format = task->has_format ? task->format : BACKUP_FORMAT_VMA;
++    format = has_format ? format : BACKUP_FORMAT_VMA;
+ 
+-    if (task->has_devlist) {
+-        devs = g_strsplit_set(task->devlist, ",;:", -1);
++    if (has_devlist) {
++        devs = g_strsplit_set(devlist, ",;:", -1);
+ 
+         gchar **d = devs;
+         while (d && *d) {
+@@ -609,14 +581,14 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             if (blk) {
+                 bs = blk_bs(blk);
+                 if (!bdrv_is_inserted(bs)) {
+-                    error_setg(task->errp, QERR_DEVICE_HAS_NO_MEDIUM, *d);
++                    error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, *d);
+                     goto err;
+                 }
+                 PVEBackupDevInfo *di = g_new0(PVEBackupDevInfo, 1);
+                 di->bs = bs;
+                 di_list = g_list_append(di_list, di);
+             } else {
+-                error_set(task->errp, ERROR_CLASS_DEVICE_NOT_FOUND,
++                error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
+                           "Device '%s' not found", *d);
+                 goto err;
+             }
+@@ -639,7 +611,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     }
+ 
+     if (!di_list) {
+-        error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "empty device list");
++        error_set(errp, ERROR_CLASS_GENERIC_ERROR, "empty device list");
+         goto err;
+     }
+ 
+@@ -649,13 +621,13 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     while (l) {
+         PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+         l = g_list_next(l);
+-        if (bdrv_op_is_blocked(di->bs, BLOCK_OP_TYPE_BACKUP_SOURCE, task->errp)) {
++        if (bdrv_op_is_blocked(di->bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp)) {
+             goto err;
+         }
+ 
+         ssize_t size = bdrv_getlength(di->bs);
+         if (size < 0) {
+-            error_setg_errno(task->errp, -di->size, "bdrv_getlength failed");
++            error_setg_errno(errp, -di->size, "bdrv_getlength failed");
+             goto err;
+         }
+         di->size = size;
+@@ -682,47 +654,44 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     }
+ 
+     if (format == BACKUP_FORMAT_PBS) {
+-        if (!task->has_password) {
+-            error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'password'");
++        if (!has_password) {
++            error_set(errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'password'");
+             goto err_mutex;
+         }
+-        if (!task->has_backup_id) {
+-            error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-id'");
++        if (!has_backup_id) {
++            error_set(errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-id'");
+             goto err_mutex;
+         }
+-        if (!task->has_backup_time) {
+-            error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-time'");
++        if (!has_backup_time) {
++            error_set(errp, ERROR_CLASS_GENERIC_ERROR, "missing parameter 'backup-time'");
+             goto err_mutex;
+         }
+ 
+         int dump_cb_block_size = PROXMOX_BACKUP_DEFAULT_CHUNK_SIZE; // Hardcoded (4M)
+         firewall_name = "fw.conf";
+ 
+-        bool use_dirty_bitmap = task->has_use_dirty_bitmap && task->use_dirty_bitmap;
+-
+-
+         char *pbs_err = NULL;
+         pbs = proxmox_backup_new(
+-            task->backup_file,
+-            task->backup_id,
+-            task->backup_time,
++            backup_file,
++            backup_id,
++            backup_time,
+             dump_cb_block_size,
+-            task->has_password ? task->password : NULL,
+-            task->has_keyfile ? task->keyfile : NULL,
+-            task->has_key_password ? task->key_password : NULL,
+-            task->has_compress ? task->compress : true,
+-            task->has_encrypt ? task->encrypt : task->has_keyfile,
+-            task->has_fingerprint ? task->fingerprint : NULL,
++            has_password ? password : NULL,
++            has_keyfile ? keyfile : NULL,
++            has_key_password ? key_password : NULL,
++            has_compress ? compress : true,
++            has_encrypt ? encrypt : has_keyfile,
++            has_fingerprint ? fingerprint : NULL,
+              &pbs_err);
+ 
+         if (!pbs) {
+-            error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
++            error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                       "proxmox_backup_new failed: %s", pbs_err);
+             proxmox_backup_free_error(pbs_err);
+             goto err_mutex;
+         }
+ 
+-        int connect_result = proxmox_backup_co_connect(pbs, task->errp);
++        int connect_result = proxmox_backup_co_connect(pbs, errp);
+         if (connect_result < 0)
+             goto err_mutex;
+ 
+@@ -741,9 +710,9 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             BdrvDirtyBitmap *bitmap = bdrv_find_dirty_bitmap(di->bs, PBS_BITMAP_NAME);
+             bool expect_only_dirty = false;
+ 
+-            if (use_dirty_bitmap) {
++            if (has_use_dirty_bitmap && use_dirty_bitmap) {
+                 if (bitmap == NULL) {
+-                    bitmap = bdrv_create_dirty_bitmap(di->bs, dump_cb_block_size, PBS_BITMAP_NAME, task->errp);
++                    bitmap = bdrv_create_dirty_bitmap(di->bs, dump_cb_block_size, PBS_BITMAP_NAME, errp);
+                     if (!bitmap) {
+                         goto err_mutex;
+                     }
+@@ -773,12 +742,12 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+                 }
+             }
+ 
+-            int dev_id = proxmox_backup_co_register_image(pbs, devname, di->size, expect_only_dirty, task->errp);
++            int dev_id = proxmox_backup_co_register_image(pbs, devname, di->size, expect_only_dirty, errp);
+             if (dev_id < 0) {
+                 goto err_mutex;
+             }
+ 
+-            if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_pbs_cb, di, task->errp))) {
++            if (!(di->target = bdrv_backup_dump_create(dump_cb_block_size, di->size, pvebackup_co_dump_pbs_cb, di, errp))) {
+                 goto err_mutex;
+             }
+ 
+@@ -792,10 +761,10 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             backup_state.stat.bitmap_list = g_list_append(backup_state.stat.bitmap_list, info);
+         }
+     } else if (format == BACKUP_FORMAT_VMA) {
+-        vmaw = vma_writer_create(task->backup_file, uuid, &local_err);
++        vmaw = vma_writer_create(backup_file, uuid, &local_err);
+         if (!vmaw) {
+             if (local_err) {
+-                error_propagate(task->errp, local_err);
++                error_propagate(errp, local_err);
+             }
+             goto err_mutex;
+         }
+@@ -806,25 +775,25 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             PVEBackupDevInfo *di = (PVEBackupDevInfo *)l->data;
+             l = g_list_next(l);
+ 
+-            if (!(di->target = bdrv_backup_dump_create(VMA_CLUSTER_SIZE, di->size, pvebackup_co_dump_vma_cb, di, task->errp))) {
++            if (!(di->target = bdrv_backup_dump_create(VMA_CLUSTER_SIZE, di->size, pvebackup_co_dump_vma_cb, di, errp))) {
+                 goto err_mutex;
+             }
+ 
+             const char *devname = bdrv_get_device_name(di->bs);
+             di->dev_id = vma_writer_register_stream(vmaw, devname, di->size);
+             if (di->dev_id <= 0) {
+-                error_set(task->errp, ERROR_CLASS_GENERIC_ERROR,
++                error_set(errp, ERROR_CLASS_GENERIC_ERROR,
+                           "register_stream failed");
+                 goto err_mutex;
+             }
+         }
+     } else if (format == BACKUP_FORMAT_DIR) {
+-        if (mkdir(task->backup_file, 0640) != 0) {
+-            error_setg_errno(task->errp, errno, "can't create directory '%s'\n",
+-                             task->backup_file);
++        if (mkdir(backup_file, 0640) != 0) {
++            error_setg_errno(errp, errno, "can't create directory '%s'\n",
++                             backup_file);
+             goto err_mutex;
+         }
+-        backup_dir = task->backup_file;
++        backup_dir = backup_file;
+ 
+         l = di_list;
+         while (l) {
+@@ -838,34 +807,34 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+             bdrv_img_create(di->targetfile, "raw", NULL, NULL, NULL,
+                             di->size, flags, false, &local_err);
+             if (local_err) {
+-                error_propagate(task->errp, local_err);
++                error_propagate(errp, local_err);
+                 goto err_mutex;
+             }
+ 
+             di->target = bdrv_open(di->targetfile, NULL, NULL, flags, &local_err);
+             if (!di->target) {
+-                error_propagate(task->errp, local_err);
++                error_propagate(errp, local_err);
+                 goto err_mutex;
+             }
+         }
+     } else {
+-        error_set(task->errp, ERROR_CLASS_GENERIC_ERROR, "unknown backup format");
++        error_set(errp, ERROR_CLASS_GENERIC_ERROR, "unknown backup format");
+         goto err_mutex;
+     }
+ 
+ 
+     /* add configuration file to archive */
+-    if (task->has_config_file) {
+-        if (pvebackup_co_add_config(task->config_file, config_name, format, backup_dir,
+-                                    vmaw, pbs, task->errp) != 0) {
++    if (has_config_file) {
++        if (pvebackup_co_add_config(config_file, config_name, format, backup_dir,
++                                    vmaw, pbs, errp) != 0) {
+             goto err_mutex;
+         }
+     }
+ 
+     /* add firewall file to archive */
+-    if (task->has_firewall_file) {
+-        if (pvebackup_co_add_config(task->firewall_file, firewall_name, format, backup_dir,
+-                                    vmaw, pbs, task->errp) != 0) {
++    if (has_firewall_file) {
++        if (pvebackup_co_add_config(firewall_file, firewall_name, format, backup_dir,
++                                    vmaw, pbs, errp) != 0) {
+             goto err_mutex;
+         }
+     }
+@@ -883,7 +852,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     if (backup_state.stat.backup_file) {
+         g_free(backup_state.stat.backup_file);
+     }
+-    backup_state.stat.backup_file = g_strdup(task->backup_file);
++    backup_state.stat.backup_file = g_strdup(backup_file);
+ 
+     uuid_copy(backup_state.stat.uuid, uuid);
+     uuid_unparse_lower(uuid, backup_state.stat.uuid_str);
+@@ -898,7 +867,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+ 
+     qemu_mutex_unlock(&backup_state.stat.lock);
+ 
+-    backup_state.speed = (task->has_speed && task->speed > 0) ? task->speed : 0;
++    backup_state.speed = (has_speed && speed > 0) ? speed : 0;
+ 
+     backup_state.vmaw = vmaw;
+     backup_state.pbs = pbs;
+@@ -908,8 +877,6 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     uuid_info = g_malloc0(sizeof(*uuid_info));
+     uuid_info->UUID = uuid_str;
+ 
+-    task->result = uuid_info;
+-
+     /* Run create_backup_jobs_bh outside of coroutine (in BH) but keep
+     * backup_mutex locked. This is fine, a CoMutex can be held across yield
+     * points, and we'll release it as soon as the BH reschedules us.
+@@ -923,7 +890,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     qemu_coroutine_yield();
+ 
+     if (local_err) {
+-        error_propagate(task->errp, local_err);
++        error_propagate(errp, local_err);
+         goto err;
+     }
+ 
+@@ -936,7 +903,7 @@ static void coroutine_fn pvebackup_co_prepare(void *opaque)
+     /* start the first job in the transaction */
+     job_txn_start_seq(backup_state.txn);
+ 
+-    return;
++    return uuid_info;
+ 
+ err_mutex:
+     qemu_mutex_unlock(&backup_state.stat.lock);
+@@ -967,7 +934,7 @@ err:
+     if (vmaw) {
+         Error *err = NULL;
+         vma_writer_close(vmaw, &err);
+-        unlink(task->backup_file);
++        unlink(backup_file);
+     }
+ 
+     if (pbs) {
+@@ -978,65 +945,8 @@ err:
+         rmdir(backup_dir);
+     }
+ 
+-    task->result = NULL;
+-
+     qemu_co_mutex_unlock(&backup_state.backup_mutex);
+-    return;
+-}
+-
+-UuidInfo *qmp_backup(
+-    const char *backup_file,
+-    bool has_password, const char *password,
+-    bool has_keyfile, const char *keyfile,
+-    bool has_key_password, const char *key_password,
+-    bool has_fingerprint, const char *fingerprint,
+-    bool has_backup_id, const char *backup_id,
+-    bool has_backup_time, int64_t backup_time,
+-    bool has_use_dirty_bitmap, bool use_dirty_bitmap,
+-    bool has_compress, bool compress,
+-    bool has_encrypt, bool encrypt,
+-    bool has_format, BackupFormat format,
+-    bool has_config_file, const char *config_file,
+-    bool has_firewall_file, const char *firewall_file,
+-    bool has_devlist, const char *devlist,
+-    bool has_speed, int64_t speed, Error **errp)
+-{
+-    QmpBackupTask task = {
+-        .backup_file = backup_file,
+-        .has_password = has_password,
+-        .password = password,
+-        .has_keyfile = has_keyfile,
+-        .keyfile = keyfile,
+-        .has_key_password = has_key_password,
+-        .key_password = key_password,
+-        .has_fingerprint = has_fingerprint,
+-        .fingerprint = fingerprint,
+-        .has_backup_id = has_backup_id,
+-        .backup_id = backup_id,
+-        .has_backup_time = has_backup_time,
+-        .backup_time = backup_time,
+-        .has_use_dirty_bitmap = has_use_dirty_bitmap,
+-        .use_dirty_bitmap = use_dirty_bitmap,
+-        .has_compress = has_compress,
+-        .compress = compress,
+-        .has_encrypt = has_encrypt,
+-        .encrypt = encrypt,
+-        .has_format = has_format,
+-        .format = format,
+-        .has_config_file = has_config_file,
+-        .config_file = config_file,
+-        .has_firewall_file = has_firewall_file,
+-        .firewall_file = firewall_file,
+-        .has_devlist = has_devlist,
+-        .devlist = devlist,
+-        .has_speed = has_speed,
+-        .speed = speed,
+-        .errp = errp,
+-    };
+-
+-    block_on_coroutine_fn(pvebackup_co_prepare, &task);
+-
+-    return task.result;
++    return NULL;
+ }
+ 
+ BackupStatus *qmp_query_backup(Error **errp)
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 30eb1262ff..6ff5367383 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -847,7 +847,7 @@
+                                     '*config-file': 'str',
+                                     '*firewall-file': 'str',
+                                     '*devlist': 'str', '*speed': 'int' },
+-  'returns': 'UuidInfo' }
++  'returns': 'UuidInfo', 'coroutine': true }
+ 
+ ##
+ # @query-backup:
+@@ -869,7 +869,7 @@
+ # Notes: This command succeeds even if there is no backup process running.
+ #
+ ##
+-{ 'command': 'backup-cancel' }
++{ 'command': 'backup-cancel', 'coroutine': true }
+ 
+ ##
+ # @ProxmoxSupportStatus:

debian/patches/pve/0043-PBS-add-master-key-support.patch

@@ -0,0 +1,97 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Wed, 10 Feb 2021 11:07:06 +0100
+Subject: [PATCH] PBS: add master key support
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+this requires a new enough libproxmox-backup-qemu0, and allows querying
+from the PVE side to avoid QMP calls with unsupported parameters.
+
+Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/monitor/block-hmp-cmds.c | 1 +
+ pve-backup.c                   | 3 +++
+ qapi/block-core.json           | 7 +++++++
+ 3 files changed, 11 insertions(+)
+
+diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
+index 11c84d5508..0932deb28c 100644
+--- a/block/monitor/block-hmp-cmds.c
++++ b/block/monitor/block-hmp-cmds.c
+@@ -1036,6 +1036,7 @@ void coroutine_fn hmp_backup(Monitor *mon, const QDict *qdict)
+         false, NULL, // PBS password
+         false, NULL, // PBS keyfile
+         false, NULL, // PBS key_password
++        false, NULL, // PBS master_keyfile
+         false, NULL, // PBS fingerprint
+         false, NULL, // PBS backup-id
+         false, 0, // PBS backup-time
+diff --git a/pve-backup.c b/pve-backup.c
+index dec9c0d188..076146cc1e 100644
+--- a/pve-backup.c
++++ b/pve-backup.c
+@@ -531,6 +531,7 @@ UuidInfo coroutine_fn *qmp_backup(
+     bool has_password, const char *password,
+     bool has_keyfile, const char *keyfile,
+     bool has_key_password, const char *key_password,
++    bool has_master_keyfile, const char *master_keyfile,
+     bool has_fingerprint, const char *fingerprint,
+     bool has_backup_id, const char *backup_id,
+     bool has_backup_time, int64_t backup_time,
+@@ -679,6 +680,7 @@ UuidInfo coroutine_fn *qmp_backup(
+             has_password ? password : NULL,
+             has_keyfile ? keyfile : NULL,
+             has_key_password ? key_password : NULL,
++            has_master_keyfile ? master_keyfile : NULL,
+             has_compress ? compress : true,
+             has_encrypt ? encrypt : has_keyfile,
+             has_fingerprint ? fingerprint : NULL,
+@@ -1042,5 +1044,6 @@ ProxmoxSupportStatus *qmp_query_proxmox_support(Error **errp)
+     ret->pbs_dirty_bitmap_savevm = true;
+     ret->pbs_dirty_bitmap_migration = true;
+     ret->query_bitmap_info = true;
++    ret->pbs_masterkey = true;
+     return ret;
+ }
+diff --git a/qapi/block-core.json b/qapi/block-core.json
+index 6ff5367383..bef9b65fec 100644
+--- a/qapi/block-core.json
++++ b/qapi/block-core.json
+@@ -818,6 +818,8 @@
+ #
+ # @key-password: password for keyfile (optional for format 'pbs')
+ #
++# @master-keyfile: PEM-formatted master public keyfile (optional for format 'pbs')
++#
+ # @fingerprint: server cert fingerprint (optional for format 'pbs')
+ #
+ # @backup-id: backup ID (required for format 'pbs')
+@@ -837,6 +839,7 @@
+                                     '*password': 'str',
+                                     '*keyfile': 'str',
+                                     '*key-password': 'str',
++                                    '*master-keyfile': 'str',
+                                     '*fingerprint': 'str',
+                                     '*backup-id': 'str',
+                                     '*backup-time': 'int',
+@@ -889,6 +892,9 @@
+ #                              migration cap if this is false/unset may lead
+ #                              to crashes on migration!
+ #
++# @pbs-masterkey: True if the QMP backup call supports the 'master_keyfile'
++#                 parameter.
++#
+ # @pbs-library-version: Running version of libproxmox-backup-qemu0 library.
+ #
+ ##
+@@ -897,6 +903,7 @@
+             'query-bitmap-info': 'bool',
+             'pbs-dirty-bitmap-savevm': 'bool',
+             'pbs-dirty-bitmap-migration': 'bool',
++            'pbs-masterkey': 'bool',
+             'pbs-library-version': 'str' } }
+ 
+ ##

debian/patches/pve/0043-move-savevm-async-back-into-a-coroutine.patch

@@ -1,111 +0,0 @@
-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
-From: Wolfgang Bumiller <w.bumiller@proxmox.com>
-Date: Thu, 30 Apr 2020 15:55:37 +0200
-Subject: [PATCH] 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>
----
- savevm-async.c | 28 +++++++++-------------------
- 1 file changed, 9 insertions(+), 19 deletions(-)
-
-diff --git a/savevm-async.c b/savevm-async.c
-index a38b15d652..af865b9a0a 100644
---- a/savevm-async.c
-+++ b/savevm-async.c
-@@ -51,7 +51,7 @@ static struct SnapshotState {
-     QEMUFile *file;
-     int64_t total_time;
-     QEMUBH *cleanup_bh;
--    QemuThread thread;
-+    Coroutine *co;
- } snap_state;
- 
- SaveVMInfo *qmp_query_savevm(Error **errp)
-@@ -201,11 +201,9 @@ static void process_savevm_cleanup(void *opaque)
-     int ret;
-     qemu_bh_delete(snap_state.cleanup_bh);
-     snap_state.cleanup_bh = NULL;
-+    snap_state.co = NULL;
-     qemu_savevm_state_cleanup();
- 
--    qemu_mutex_unlock_iothread();
--    qemu_thread_join(&snap_state.thread);
--    qemu_mutex_lock_iothread();
-     ret = save_snapshot_cleanup();
-     if (ret < 0) {
-         save_snapshot_error("save_snapshot_cleanup error %d", ret);
-@@ -221,18 +219,13 @@ static void process_savevm_cleanup(void *opaque)
-     }
- }
- 
--static void *process_savevm_thread(void *opaque)
-+static void process_savevm_coro(void *opaque)
- {
-     int ret;
-     int64_t maxlen;
-     MigrationState *ms = migrate_get_current();
- 
--    rcu_register_thread();
--
--    qemu_savevm_state_header(snap_state.file);
--    qemu_savevm_state_setup(snap_state.file);
-     ret = qemu_file_get_error(snap_state.file);
--
-     if (ret < 0) {
-         save_snapshot_error("qemu_savevm_state_setup failed");
-         goto out;
-@@ -247,16 +240,13 @@ static void *process_savevm_thread(void *opaque)
-         maxlen = blk_getlength(snap_state.target) - 30*1024*1024;
- 
-         if (pending_size > 400000 && snap_state.bs_pos + pending_size < maxlen) {
--            qemu_mutex_lock_iothread();
-             ret = qemu_savevm_state_iterate(snap_state.file, false);
-             if (ret < 0) {
-                 save_snapshot_error("qemu_savevm_state_iterate error %d", ret);
-                 break;
-             }
--            qemu_mutex_unlock_iothread();
-             DPRINTF("savevm inerate pending size %lu ret %d\n", pending_size, ret);
-         } else {
--            qemu_mutex_lock_iothread();
-             qemu_system_wakeup_request(QEMU_WAKEUP_REASON_OTHER, NULL);
-             ret = global_state_store();
-             if (ret) {
-@@ -285,16 +275,12 @@ static void *process_savevm_thread(void *opaque)
-     }
- 
-     qemu_bh_schedule(snap_state.cleanup_bh);
--    qemu_mutex_unlock_iothread();
- 
- out:
-     /* set migration state accordingly and clear soon-to-be stale file */
-     migrate_set_state(&ms->state, MIGRATION_STATUS_SETUP,
-                       ret ? MIGRATION_STATUS_FAILED : MIGRATION_STATUS_COMPLETED);
-     ms->to_dst_file = NULL;
--
--    rcu_unregister_thread();
--    return NULL;
- }
- 
- void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
-@@ -373,8 +359,12 @@ void qmp_savevm_start(bool has_statefile, const char *statefile, Error **errp)
- 
-     snap_state.state = SAVE_STATE_ACTIVE;
-     snap_state.cleanup_bh = qemu_bh_new(process_savevm_cleanup, &snap_state);
--    qemu_thread_create(&snap_state.thread, "savevm-async", process_savevm_thread,
--                       NULL, QEMU_THREAD_JOINABLE);
-+    snap_state.co = qemu_coroutine_create(&process_savevm_coro, NULL);
-+    qemu_mutex_unlock_iothread();
-+    qemu_savevm_state_header(snap_state.file);
-+    qemu_savevm_state_setup(snap_state.file);
-+    qemu_mutex_lock_iothread();
-+    aio_co_schedule(iohandler_get_aio_context(), snap_state.co);
- 
-     return;
- 

debian/patches/pve/0044-PVE-block-pbs-fast-path-reads-without-allocation-if-.patch

@@ -0,0 +1,52 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Wed, 9 Dec 2020 11:46:57 +0100
+Subject: [PATCH] PVE: block/pbs: fast-path reads without allocation if
+ possible
+
+...and switch over to g_malloc/g_free while at it to align with other
+QEMU code.
+
+Tracing shows the fast-path is taken almost all the time, though not
+100% so the slow one is still necessary.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/pbs.c | 17 ++++++++++++++---
+ 1 file changed, 14 insertions(+), 3 deletions(-)
+
+diff --git a/block/pbs.c b/block/pbs.c
+index 1481a2bfd1..fbf0d8d845 100644
+--- a/block/pbs.c
++++ b/block/pbs.c
+@@ -200,7 +200,16 @@ static coroutine_fn int pbs_co_preadv(BlockDriverState *bs,
+     BDRVPBSState *s = bs->opaque;
+     int ret;
+     char *pbs_error = NULL;
+-    uint8_t *buf = malloc(bytes);
++    uint8_t *buf;
++    bool inline_buf = true;
++
++    /* for single-buffer IO vectors we can fast-path the write directly to it */
++    if (qiov->niov == 1 && qiov->iov->iov_len >= bytes) {
++        buf = qiov->iov->iov_base;
++    } else {
++        inline_buf = false;
++        buf = g_malloc(bytes);
++    }
+ 
+     ReadCallbackData rcb = {
+         .co = qemu_coroutine_self(),
+@@ -218,8 +227,10 @@ static coroutine_fn int pbs_co_preadv(BlockDriverState *bs,
+         return -EIO;
+     }
+ 
+-    qemu_iovec_from_buf(qiov, 0, buf, bytes);
+-    free(buf);
++    if (!inline_buf) {
++        qemu_iovec_from_buf(qiov, 0, buf, bytes);
++        g_free(buf);
++    }
+ 
+     return ret;
+ }

debian/patches/pve/0045-PVE-block-stream-increase-chunk-size.patch

@@ -0,0 +1,23 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Tue, 2 Mar 2021 16:34:28 +0100
+Subject: [PATCH] PVE: block/stream: increase chunk size
+
+Ceph favors bigger chunks, so increase to 4M.
+---
+ block/stream.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/block/stream.c b/block/stream.c
+index 236384f2f7..a5371420e3 100644
+--- a/block/stream.c
++++ b/block/stream.c
+@@ -26,7 +26,7 @@ enum {
+      * large enough to process multiple clusters in a single call, so
+      * that populating contiguous regions of the image is efficient.
+      */
+-    STREAM_CHUNK = 512 * 1024, /* in bytes */
++    STREAM_CHUNK = 4 * 1024 * 1024, /* in bytes */
+ };
+ 
+ typedef struct StreamBlockJob {

debian/patches/pve/0046-block-io-accept-NULL-qiov-in-bdrv_pad_request.patch

@@ -0,0 +1,42 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Tue, 2 Mar 2021 16:11:54 +0100
+Subject: [PATCH] block/io: accept NULL qiov in bdrv_pad_request
+
+Some operations, e.g. block-stream, perform reads while discarding the
+results (only copy-on-read matters). In this case they will pass NULL as
+the target QEMUIOVector, which will however trip bdrv_pad_request, since
+it wants to extend its passed vector.
+
+Simply check for NULL and do nothing, there's no reason to pad the
+target if it will be discarded anyway.
+---
+ block/io.c | 13 ++++++++-----
+ 1 file changed, 8 insertions(+), 5 deletions(-)
+
+diff --git a/block/io.c b/block/io.c
+index ec5e152bb7..08dee005ec 100644
+--- a/block/io.c
++++ b/block/io.c
+@@ -1613,13 +1613,16 @@ static bool bdrv_pad_request(BlockDriverState *bs,
+         return false;
+     }
+ 
+-    qemu_iovec_init_extended(&pad->local_qiov, pad->buf, pad->head,
+-                             *qiov, *qiov_offset, *bytes,
+-                             pad->buf + pad->buf_len - pad->tail, pad->tail);
++    if (*qiov) {
++        qemu_iovec_init_extended(&pad->local_qiov, pad->buf, pad->head,
++                                *qiov, *qiov_offset, *bytes,
++                                pad->buf + pad->buf_len - pad->tail, pad->tail);
++        *qiov = &pad->local_qiov;
++        *qiov_offset = 0;
++    }
++
+     *bytes += pad->head + pad->tail;
+     *offset -= pad->head;
+-    *qiov = &pad->local_qiov;
+-    *qiov_offset = 0;
+ 
+     return true;
+ }

debian/patches/pve/0047-block-add-alloc-track-driver.patch

@@ -0,0 +1,394 @@
+From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
+From: Stefan Reiter <s.reiter@proxmox.com>
+Date: Mon, 7 Dec 2020 15:21:03 +0100
+Subject: [PATCH] block: add alloc-track driver
+
+Add a new filter node 'alloc-track', which seperates reads and writes to
+different children, thus allowing to put a backing image behind any
+blockdev (regardless of driver support). Since we can't detect any
+pre-allocated blocks, we can only track new writes, hence the write
+target ('file') for this node must always be empty.
+
+Intended use case is for live restoring, i.e. add a backup image as a
+block device into a VM, then put an alloc-track on the restore target
+and set the backup as backing. With this, one can use a regular
+'block-stream' to restore the image, while the VM can already run in the
+background. Copy-on-read will help make progress as the VM reads as
+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'.
+
+Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
+---
+ block/alloc-track.c | 342 ++++++++++++++++++++++++++++++++++++++++++++
+ block/meson.build   |   1 +
+ 2 files changed, 343 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..b579380279
+--- /dev/null
++++ b/block/alloc-track.c
+@@ -0,0 +1,345 @@
++/*
++ * 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
++ * must never be put on a node that already contains data.
++ *
++ * Copyright (c) 2020 Proxmox Server Solutions GmbH
++ * Copyright (c) 2020 Stefan Reiter <s.reiter@proxmox.com>
++ *
++ * This work is licensed under the terms of the GNU GPL, version 2 or later.
++ * See the COPYING file in the top-level directory.
++ */
++
++#include "qemu/osdep.h"
++#include "qapi/error.h"
++#include "block/block_int.h"
++#include "qapi/qmp/qdict.h"
++#include "qapi/qmp/qstring.h"
++#include "qemu/cutils.h"
++#include "qemu/option.h"
++#include "qemu/module.h"
++#include "sysemu/block-backend.h"
++
++#define TRACK_OPT_AUTO_REMOVE "auto-remove"
++
++typedef enum DropState {
++    DropNone,
++    DropRequested,
++    DropInProgress,
++} DropState;
++
++typedef struct {
++    BdrvDirtyBitmap *bitmap;
++    DropState drop_state;
++    bool auto_remove;
++} BDRVAllocTrackState;
++
++static QemuOptsList runtime_opts = {
++    .name = "alloc-track",
++    .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head),
++    .desc = {
++        {
++            .name = TRACK_OPT_AUTO_REMOVE,
++            .type = QEMU_OPT_BOOL,
++            .help = "automatically replace this node with 'file' when 'backing'"
++                    "is detached",
++        },
++        { /* end of list */ }
++    },
++};
++
++static void track_refresh_limits(BlockDriverState *bs, Error **errp)
++{
++    BlockDriverInfo bdi;
++
++    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);
++    bs->bl.request_alignment = MAX(bs->file->bs->bl.request_alignment,
++                                   MAX(bdi.cluster_size, BDRV_SECTOR_SIZE));
++}
++
++static int track_open(BlockDriverState *bs, QDict *options, int flags,
++                      Error **errp)
++{
++    BDRVAllocTrackState *s = bs->opaque;
++    QemuOpts *opts;
++    Error *local_err = NULL;
++    int ret = 0;
++
++    opts = qemu_opts_create(&runtime_opts, NULL, 0, &error_abort);
++    qemu_opts_absorb_qdict(opts, options, &local_err);
++    if (local_err) {
++        error_propagate(errp, local_err);
++        ret = -EINVAL;
++        goto fail;
++    }
++
++    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);
++    if (local_err) {
++        ret = -EINVAL;
++        error_propagate(errp, local_err);
++        goto fail;
++    }
++
++    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);
++    if (local_err) {
++        ret = -EIO;
++        error_propagate(errp, local_err);
++        goto fail;
++    }
++
++    s->drop_state = DropNone;
++
++fail:
++    if (ret < 0) {
++        bdrv_unref_child(bs, bs->file);
++        if (s->bitmap) {
++            bdrv_release_dirty_bitmap(s->bitmap);
++        }
++    }
++    qemu_opts_del(opts);
++    return ret;
++}
++
++static void track_close(BlockDriverState *bs)
++{
++    BDRVAllocTrackState *s = bs->opaque;
++    if (s->bitmap) {
++        bdrv_release_dirty_bitmap(s->bitmap);
++    }
++}
++
++static int64_t track_getlength(BlockDriverState *bs)
++{
++    return bdrv_getlength(bs->file->bs);
++}
++
++static int coroutine_fn track_co_preadv(BlockDriverState *bs,
++    uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags)
++{
++    BDRVAllocTrackState *s = bs->opaque;
++    QEMUIOVector local_qiov;
++    int ret;
++
++    /* 'cur_offset' is relative to 'offset', 'local_offset' to image start */
++    uint64_t cur_offset, local_offset;
++    int64_t local_bytes;
++    bool alloc;
++
++    /* a read request can span multiple granularity-sized chunks, and can thus
++     * contain blocks with different allocation status - we could just iterate
++     * granularity-wise, but for better performance use bdrv_dirty_bitmap_next_X
++     * to find the next flip and consider everything up to that in one go */
++    for (cur_offset = 0; cur_offset < bytes; cur_offset += local_bytes) {
++        local_offset = offset + cur_offset;
++        alloc = bdrv_dirty_bitmap_get(s->bitmap, local_offset);
++        if (alloc) {
++            local_bytes = bdrv_dirty_bitmap_next_zero(s->bitmap, local_offset,
++                                                      bytes - cur_offset);
++        } else {
++            local_bytes = bdrv_dirty_bitmap_next_dirty(s->bitmap, local_offset,
++                                                       bytes - cur_offset);
++        }
++
++        /* _bitmap_next_X return is -1 if no end found within limit, otherwise
++         * offset of next flip (to start of image) */
++        local_bytes = local_bytes < 0 ?
++            bytes - cur_offset :
++            local_bytes - local_offset;
++
++        qemu_iovec_init_slice(&local_qiov, qiov, cur_offset, local_bytes);
++
++        if (alloc) {
++            ret = bdrv_co_preadv(bs->file, local_offset, local_bytes,
++                                 &local_qiov, flags);
++        } else if (bs->backing) {
++            ret = bdrv_co_preadv(bs->backing, local_offset, local_bytes,
++                                 &local_qiov, flags);
++        } else {
++            ret = qemu_iovec_memset(&local_qiov, cur_offset, 0, local_bytes);
++        }
++
++        if (ret != 0) {
++            break;
++        }
++    }
++
++    return ret;
++}
++
++static int coroutine_fn track_co_pwritev(BlockDriverState *bs,
++    uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags)
++{
++    return bdrv_co_pwritev(bs->file, offset, bytes, qiov, flags);
++}
++
++static int coroutine_fn track_co_pwrite_zeroes(BlockDriverState *bs,
++    int64_t offset, int count, BdrvRequestFlags flags)
++{
++    return bdrv_co_pwrite_zeroes(bs->file, offset, count, flags);
++}
++
++static int coroutine_fn track_co_pdiscard(BlockDriverState *bs,
++    int64_t offset, int count)
++{
++    return bdrv_co_pdiscard(bs->file, offset, count);
++}
++
++static coroutine_fn int 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)
++{
++    BDRVAllocTrackState *s = bs->opaque;
++
++    bool alloc = bdrv_dirty_bitmap_get(s->bitmap, offset);
++    int64_t next_flipped;
++    if (alloc) {
++        next_flipped = bdrv_dirty_bitmap_next_zero(s->bitmap, offset, bytes);
++    } else {
++        next_flipped = bdrv_dirty_bitmap_next_dirty(s->bitmap, offset, bytes);
++    }
++
++    /* in case not the entire region has the same state, we need to set pnum to
++     * indicate for how many bytes our result is valid */
++    *pnum = next_flipped == -1 ? bytes : next_flipped - offset;
++    *map = offset;
++
++    if (alloc) {
++        *file = bs->file->bs;
++        return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID;
++    } else if (bs->backing) {
++        *file = bs->backing->bs;
++    }
++    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)
++{
++    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 {
++        /* 'backing' is also a child of our BDS, but we don't expect it to be
++         * writeable, so we only forward 'consistent read' */
++        *nperm = perm & BLK_PERM_CONSISTENT_READ;
++    }
++}
++
++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)
++{
++    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);
++    }
++
++    return 0;
++}
++
++static BlockDriver bdrv_alloc_track = {
++    .format_name                      = "alloc-track",
++    .instance_size                    = sizeof(BDRVAllocTrackState),
++
++    .bdrv_file_open                   = track_open,
++    .bdrv_close                       = track_close,
++    .bdrv_getlength                   = track_getlength,
++    .bdrv_child_perm                  = track_child_perm,
++    .bdrv_refresh_limits              = track_refresh_limits,
++
++    .bdrv_co_pwrite_zeroes            = track_co_pwrite_zeroes,
++    .bdrv_co_pwritev                  = track_co_pwritev,
++    .bdrv_co_preadv                   = track_co_preadv,
++    .bdrv_co_pdiscard                 = track_co_pdiscard,
++
++    .bdrv_co_flush                    = track_co_flush,
++    .bdrv_co_flush_to_disk            = track_co_flush,
++
++    .supports_backing                 = true,
++
++    .bdrv_co_block_status             = track_co_block_status,
++    .bdrv_change_backing_file         = track_change_backing_file,
++};
++
++static void bdrv_alloc_track_init(void)
++{
++    bdrv_register(&bdrv_alloc_track);
++}
++
++block_init(bdrv_alloc_track_init);
+diff --git a/block/meson.build b/block/meson.build
+index a070060e53..e387990764 100644
+--- a/block/meson.build
++++ b/block/meson.build
+@@ -2,6 +2,7 @@ block_ss.add(genh)
+ block_ss.add(files(
+   'accounting.c',
+   'aio_task.c',
++  'alloc-track.c',
+   'amend.c',
+   'backup.c',
+   'backup-dump.c',

debian/patches/series

@@ -1,3 +1,21 @@
+extra/0001-Revert-qemu-img-convert-Don-t-pre-zero-images.patch
+extra/0002-docs-don-t-install-man-page-if-guest-agent-is-disabl.patch
+extra/0003-migration-only-check-page-size-match-if-RAM-postcopy.patch
+extra/0004-virtiofsd-extract-lo_do_open-from-lo_open.patch
+extra/0005-virtiofsd-optionally-return-inode-pointer-from-lo_do.patch
+extra/0006-virtiofsd-prevent-opening-of-special-files-CVE-2020-.patch
+extra/0007-virtiofsd-Add-_llseek-to-the-seccomp-whitelist.patch
+extra/0008-virtiofsd-Add-restart_syscall-to-the-seccomp-whiteli.patch
+extra/0009-i386-acpi-restore-device-paths-for-pre-5.1-vms.patch
+extra/0010-monitor-qmp-fix-race-on-CHR_EVENT_CLOSED-without-OOB.patch
+extra/0011-block-Fix-locking-in-qmp_block_resize.patch
+extra/0012-block-Fix-deadlock-in-bdrv_co_yield_to_drain.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
+bitmap-mirror/0004-mirror-switch-to-bdrv_dirty_bitmap_merge_internal.patch
+bitmap-mirror/0005-iotests-add-test-for-bitmap-mirror.patch
+bitmap-mirror/0006-mirror-move-some-checks-to-qmp.patch
 pve/0001-PVE-Config-block-file-change-locking-default-to-off.patch
 pve/0002-PVE-Config-Adjust-network-script-path-to-etc-kvm.patch
 pve/0003-PVE-Config-set-the-CPU-model-to-kvm64-32-instead-of-.patch
@@ -14,31 +32,35 @@ pve/0013-PVE-Up-qemu-img-dd-add-n-skip_create.patch
 pve/0014-PVE-virtio-balloon-improve-query-balloon.patch
 pve/0015-PVE-qapi-modify-query-machines.patch
 pve/0016-PVE-qapi-modify-spice-query.patch
-pve/0017-PVE-internal-snapshot-async.patch
-pve/0018-PVE-block-add-the-zeroinit-block-driver-filter.patch
-pve/0019-PVE-Add-dummy-id-command-line-parameter.patch
-pve/0020-PVE-Config-Revert-target-i386-disable-LINT0-after-re.patch
-pve/0021-PVE-Up-Config-file-posix-make-locking-optiono-on-cre.patch
-pve/0022-PVE-savevm-async-kick-AIO-wait-on-block-state-write.patch
-pve/0023-PVE-move-snapshot-cleanup-into-bottom-half.patch
-pve/0024-PVE-monitor-disable-oob-capability.patch
-pve/0025-PVE-Compat-4.0-used-balloon-qemu-4-0-config-size-fal.patch
-pve/0026-PVE-Allow-version-code-in-machine-type.patch
-pve/0027-PVE-Backup-modify-job-api.patch
-pve/0028-PVE-Backup-add-vma-backup-format-code.patch
-pve/0029-PVE-Backup-add-backup-dump-block-driver.patch
-pve/0030-PVE-Backup-proxmox-backup-patches-for-qemu.patch
-pve/0031-PVE-Backup-aquire-aio_context-before-calling-backup_.patch
-pve/0032-PVE-Backup-pbs-restore-new-command-to-restore-from-p.patch
-pve/0033-PVE-Backup-avoid-coroutines-to-fix-AIO-freeze-cleanu.patch
-pve/0034-drive-mirror-add-support-for-sync-bitmap-mode-never.patch
-pve/0035-drive-mirror-add-support-for-conditional-and-always-.patch
-pve/0036-mirror-add-check-for-bitmap-mode-without-bitmap.patch
-pve/0037-mirror-switch-to-bdrv_dirty_bitmap_merge_internal.patch
-pve/0038-iotests-add-test-for-bitmap-mirror.patch
-pve/0039-mirror-move-some-checks-to-qmp.patch
-pve/0040-PVE-savevm-async-set-up-migration-state.patch
-pve/0041-PVE-Backup-avoid-use-QemuRecMutex-inside-coroutines.patch
-pve/0042-PVE-Backup-use-QemuMutex-instead-of-QemuRecMutex.patch
-pve/0043-move-savevm-async-back-into-a-coroutine.patch
-pve/0044-add-optional-buffer-size-to-QEMUFile.patch
+pve/0017-PVE-add-savevm-async-for-background-state-snapshots.patch
+pve/0018-PVE-add-optional-buffer-size-to-QEMUFile.patch
+pve/0019-PVE-block-add-the-zeroinit-block-driver-filter.patch
+pve/0020-PVE-Add-dummy-id-command-line-parameter.patch
+pve/0021-PVE-Config-Revert-target-i386-disable-LINT0-after-re.patch
+pve/0022-PVE-Up-Config-file-posix-make-locking-optiono-on-cre.patch
+pve/0023-PVE-monitor-disable-oob-capability.patch
+pve/0024-PVE-Compat-4.0-used-balloon-qemu-4-0-config-size-fal.patch
+pve/0025-PVE-Allow-version-code-in-machine-type.patch
+pve/0026-PVE-Backup-add-vma-backup-format-code.patch
+pve/0027-PVE-Backup-add-backup-dump-block-driver.patch
+pve/0028-PVE-Backup-proxmox-backup-patches-for-qemu.patch
+pve/0029-PVE-Backup-pbs-restore-new-command-to-restore-from-p.patch
+pve/0030-PVE-Backup-Add-dirty-bitmap-tracking-for-incremental.patch
+pve/0031-PVE-various-PBS-fixes.patch
+pve/0032-PVE-Add-PBS-block-driver-to-map-backup-archives-into.patch
+pve/0033-PVE-add-query_proxmox_support-QMP-command.patch
+pve/0034-PVE-add-query-pbs-bitmap-info-QMP-call.patch
+pve/0035-PVE-redirect-stderr-to-journal-when-daemonized.patch
+pve/0036-PVE-Add-sequential-job-transaction-support.patch
+pve/0037-PVE-Backup-Use-a-transaction-to-synchronize-job-stat.patch
+pve/0038-PVE-Backup-Don-t-block-on-finishing-and-cleanup-crea.patch
+pve/0039-PVE-Migrate-dirty-bitmap-state-via-savevm.patch
+pve/0040-migration-block-dirty-bitmap-migrate-other-bitmaps-e.patch
+pve/0041-PVE-fall-back-to-open-iscsi-initiatorname.patch
+pve/0042-PVE-Use-coroutine-QMP-for-backup-cancel_backup.patch
+pve/0043-PBS-add-master-key-support.patch
+pve/0044-PVE-block-pbs-fast-path-reads-without-allocation-if-.patch
+pve/0045-PVE-block-stream-increase-chunk-size.patch
+pve/0046-block-io-accept-NULL-qiov-in-bdrv_pad_request.patch
+pve/0047-block-add-alloc-track-driver.patch
+pve-qemu-5.2-vitastor.patch

debian/pve-qemu-kvm.install

@@ -1,5 +1,4 @@
 # install the userspace utilities
-vma usr/bin/
 debian/kvm-ifup etc/kvm/
 debian/kvm-ifdown etc/kvm/
 

debian/rules

@@ -22,6 +22,10 @@ PACKAGE=pve-qemu-kvm
 destdir := $(CURDIR)/debian/$(PACKAGE)
 
 flagfile := $(destdir)/usr/share/kvm/recognized-CPUID-flags-x86_64
+machinefile := $(destdir)/usr/share/kvm/machine-versions-x86_64.json
+
+# default QEMU out-of-tree build directory is ./build
+BUILDDIR=build
 
 CFLAGS = -Wall
 
@@ -31,7 +35,7 @@ else
 	CFLAGS += -O2
 endif
 
-config.status: configure
+${BUILDDIR}/config.status: configure
 	dh_testdir
 	# Add here commands to configure the package.
 
@@ -42,10 +46,11 @@ config.status: configure
 	--prefix=/usr \
 	--sysconfdir=/etc \
 	--target-list=$(ARCH)-softmmu,aarch64-softmmu \
-	--with-confsuffix="/kvm" \
+	--with-suffix="kvm" \
 	--with-pkgversion="${DEB_SOURCE}_${DEB_VERSION_UPSTREAM}" \
 	--audio-drv-list="alsa" \
 	--datadir=/usr/share \
+	--libexecdir=/usr/lib/kvm \
 	--disable-capstone \
 	--disable-gtk \
 	--disable-guest-agent \
@@ -60,7 +65,6 @@ config.status: configure
 	--enable-docs \
 	--enable-glusterfs \
 	--enable-gnutls \
-	--enable-jemalloc \
 	--enable-libiscsi \
 	--enable-libusb \
 	--enable-linux-aio \
@@ -70,11 +74,12 @@ config.status: configure
 	--enable-spice \
 	--enable-usb-redir \
 	--enable-virtfs \
+	--enable-virtiofsd \
 	--enable-xfsctl
 
 build: build-stamp
 
-build-stamp:  config.status
+build-stamp: ${BUILDDIR}/config.status
 	dh_testdir
 
 	# Add here commands to compile the package.
@@ -120,16 +125,21 @@ install: build
 	rm $(destdir)/usr/share/kvm/qemu_vga.ndrv
 	rm $(destdir)/usr/share/kvm/slof.bin
 	rm $(destdir)/usr/share/kvm/u-boot.e500
-	# remove Aplha files
+	# remove Alpha files
 	rm $(destdir)/usr/share/kvm/palcode-clipper
+	# remove RISC-V files
+	rm $(destdir)/usr/share/kvm/opensbi-riscv32-generic-fw_dynamic.elf
+	rm $(destdir)/usr/share/kvm/opensbi-riscv64-generic-fw_dynamic.elf
 
 	# Remove things we don't package at all, would be a "kvm-dev" package
 	rm -Rf $(destdir)/usr/include/linux/
 	rm -Rf $(destdir)/usr/include
-	rm -Rf $(destdir)/usr/lib*
+	rm -f $(destdir)/usr/lib/kvm/qemu-bridge-helper
+	rm -f $(destdir)/usr/lib/kvm/virtfs-proxy-helper
 
 	# CPU flags are static for QEMU version, allows avoiding more costly checks
 	$(destdir)/usr/bin/qemu-system-x86_64 -cpu help | ./debian/parse-cpu-flags.pl > $(flagfile)
+	$(destdir)/usr/bin/qemu-system-x86_64 -machine help | ./debian/parse-machines.pl > $(machinefile)
 
 # Build architecture-independent files here.
 binary-indep: build install