Age | Commit message (Collapse) | Author |
|
Currently migration_tls_get_creds() adds the reference of creds
but there was no place to unref it. So the OBJECT(creds) will
never be freed and result in memory leak.
The leak stack:
Direct leak of 104 byte(s) in 1 object(s) allocated from:
#0 0xffffa88bd20b in __interceptor_malloc (/usr/lib64/libasan.so.4+0xd320b)
#1 0xffffa7f0cb1b in g_malloc (/usr/lib64/libglib-2.0.so.0+0x58b1b)
#2 0x14b58cb in object_new_with_type qom/object.c:634
#3 0x14b597b in object_new qom/object.c:645
#4 0x14c0e4f in user_creatable_add_type qom/object_interfaces.c:59
#5 0x141c78b in qmp_object_add qom/qom-qmp-cmds.c:312
#6 0x140e513 in qmp_marshal_object_add qapi/qapi-commands-qom.c:279
#7 0x176ba97 in do_qmp_dispatch qapi/qmp-dispatch.c:165
#8 0x176bee7 in qmp_dispatch qapi/qmp-dispatch.c:208
#9 0x136e337 in monitor_qmp_dispatch monitor/qmp.c:150
#10 0x136eae3 in monitor_qmp_bh_dispatcher monitor/qmp.c:239
#11 0x1852e93 in aio_bh_call util/async.c:89
#12 0x18531b7 in aio_bh_poll util/async.c:117
#13 0x18616bf in aio_dispatch util/aio-posix.c:459
#14 0x1853f37 in aio_ctx_dispatch util/async.c:268
#15 0xffffa7f06a7b in g_main_context_dispatch (/usr/lib64/libglib-2.0.so.0+0x52a7b)
Since we're fine to use the borrowed reference when using the creds,
so just remove the object_ref() in migration_tls_get_creds().
Signed-off-by: Zhenyu Ye <yezhenyu2@huawei.com>
Message-Id: <20200722033228.71-1-yezhenyu2@huawei.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
With blockdev, a BlockDriverState may not have a device name,
so using a node name is required as an alternative.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20200827111606.1408275-2-berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
The vsock channel is more widely use in some new features, for example,
the Nitro/Enclave. It can also be used as the migration channel.
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Message-Id: <20200806074030.174-3-longpeng2@huawei.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
Currently, the only difference of tcp channel and unix channel in
migration/socket.c is the way to build SocketAddress, but socket_parse()
can handle these two types, so use it to instead of tcp_build_address()
and unix_build_address().
The socket-type channel can be further unified based on the up, this
would be helpful for us to add other socket-type channels.
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Message-Id: <20200806074030.174-2-longpeng2@huawei.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
Rename the macros to make them consistent with the MIGRATION_OBJ
macro name.
This will make future conversion to OBJECT_DECLARE* easier.
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Tested-By: Roman Bolshakov <r.bolshakov@yadro.com>
Message-Id: <20200825192110.3528606-51-ehabkost@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
|
|
This migration parameter allows mapping block node names and bitmap
names to aliases for the purpose of block dirty bitmap migration.
This way, management tools can use different node and bitmap names on
the source and destination and pass the mapping of how bitmaps are to be
transferred to qemu (on the source, the destination, or even both with
arbitrary aliases in the migration stream).
While touching this code, fix a bug where bitmap names longer than 255
bytes would fail an assertion in qemu_put_counted_string().
Suggested-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20200820150725.68687-2-mreitz@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Meson doesn't enjoy the same flexibility we have with Make in choosing
the include path. In particular the tracing headers are using
$(build_root)/$(<D).
In order to keep the include directives unchanged,
the simplest solution is to generate headers with patterns like
"trace/trace-audio.h" and place forwarding headers in the source tree
such that for example "audio/trace.h" includes "trace/trace-audio.h".
This patch is too ugly to be applied to the Makefiles now. It's only
a way to separate the changes to the tracing header files from the
Meson rewrite of the tracing logic.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Noticed while reviewing the file for newer patches.
Fixes: b35ebdf076
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200727203206.134996-1-eblake@redhat.com>
|
|
First, if only bitmaps postcopy is enabled (and not ram postcopy)
postcopy_pause_incoming crashes on an assertion
assert(mis->to_src_file).
And anyway, bitmaps postcopy is not prepared to be somehow recovered.
The original idea instead is that if bitmaps postcopy failed, we just
lose some bitmaps, which is not critical. So, on failure we just need
to remove unfinished bitmaps and guest should continue execution on
destination.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200727194236.19551-18-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
If target is turned off prior to postcopy finished, target crashes
because busy bitmaps are found at shutdown.
Canceling incoming migration helps, as it removes all unfinished (and
therefore busy) bitmaps.
Similarly on source we crash in bdrv_close_all which asserts that all
bdrv states are removed, because bdrv states involved into dirty bitmap
migration are referenced by it. So, we need to cancel outgoing
migration as well.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200727194236.19551-17-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Bitmaps data is not critical, and we should not fail the migration (or
use postcopy recovering) because of dirty-bitmaps migration failure.
Instead we should just lose unfinished bitmaps.
Still we have to report io stream violation errors, as they affect the
whole migration stream.
While touching this, tighten code that was previously blindly calling
malloc on a size read from the migration stream, as a corrupted stream
(perhaps from a malicious user) should not be able to convince us to
allocate an inordinate amount of memory.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200727194236.19551-16-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: typo fixes, enhance commit message]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Keep bitmap state for disabled bitmaps too. Keep the state until the
end of the process. It's needed for the following commit to implement
bitmap postcopy canceling.
To clean-up the new list the following logic is used:
We need two events to consider bitmap migration finished:
1. chunk with DIRTY_BITMAP_MIG_FLAG_COMPLETE flag should be received
2. dirty_bitmap_mig_before_vm_start should be called
These two events may come in any order, so we understand which one is
last, and on the last of them we remove bitmap migration state from the
list.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200727194236.19551-15-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
bdrv_enable_dirty_bitmap_locked() call does nothing, as if we are in
postcopy, bitmap successor must be enabled, and reclaim operation will
enable the bitmap.
So, actually we need just call _reclaim_ in both if branches, and
making differences only to add an assertion seems not really good. The
logic becomes simple: on load complete we do reclaim and that's all.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200727194236.19551-14-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
finish_lock is bad name, as lock used not only on process end.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200727194236.19551-13-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Move all state variables into one global struct. Reduce global
variable usage, utilizing opaque pointer where possible.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Message-Id: <20200727194236.19551-12-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
No reasons to keep two public init functions.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20200727194236.19551-11-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Rename dirty_bitmap_mig_cleanup to dirty_bitmap_do_save_cleanup, to
stress that it is on save part.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200727194236.19551-10-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Rename types to be symmetrical for load/save part and shorter.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200727194236.19551-9-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Using the _locked version of bdrv_enable_dirty_bitmap to bypass locking
is wrong as we do not already own the mutex. Moreover, the adjacent
call to bdrv_dirty_bitmap_enable_successor grabs the mutex.
Fixes: 58f72b965e9e1q
Cc: qemu-stable@nongnu.org # v3.0
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200727194236.19551-8-vsementsov@virtuozzo.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
We shouldn't fail when finding an unnamed bitmap in a unnamed node or
node with auto-generated node name, as bitmap migration ignores such
bitmaps in the first place.
Fixes: 82640edb88faa
Fixes: 4ff5cc121b089
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200626130658.76498-1-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: commit message grammar tweaks]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Remove superfluous breaks, as there is a "return" before them.
Signed-off-by: Liao Pingfang <liao.pingfang@zte.com.cn>
Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
|
|
qemu_fclose() could return error, f.e. if bdrv_co_flush() will return
the error.
This validation will become more important once we will start waiting of
asynchronous IO operations, started from bdrv_write_vmstate(), which are
coming soon.
Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Fam Zheng <fam@euphon.net>
CC: Juan Quintela <quintela@redhat.com>
CC: Denis Plotnikov <dplotnikov@virtuozzo.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
|
|
"tmp.tls_hostname" and "tmp.tls_creds" allocated by migrate_params_test_apply()
is forgot to free at the end of qmp_migrate_set_parameters(). Fix that.
The leak stack:
Direct leak of 2 byte(s) in 2 object(s) allocated from:
#0 0xffffb597c20b in __interceptor_malloc (/usr/lib64/libasan.so.4+0xd320b)
#1 0xffffb52dcb1b in g_malloc (/usr/lib64/libglib-2.0.so.0+0x58b1b)
#2 0xffffb52f8143 in g_strdup (/usr/lib64/libglib-2.0.so.0+0x74143)
#3 0xaaaac52447fb in migrate_params_test_apply (/usr/src/debug/qemu-4.1.0/migration/migration.c:1377)
#4 0xaaaac52fdca7 in qmp_migrate_set_parameters (/usr/src/debug/qemu-4.1.0/qapi/qapi-commands-migration.c:192)
#5 0xaaaac551d543 in qmp_dispatch (/usr/src/debug/qemu-4.1.0/qapi/qmp-dispatch.c:165)
#6 0xaaaac52a0a8f in qmp_dispatch (/usr/src/debug/qemu-4.1.0/monitor/qmp.c:125)
#7 0xaaaac52a1c7f in monitor_qmp_dispatch (/usr/src/debug/qemu-4.1.0/monitor/qmp.c:214)
#8 0xaaaac55cb0cf in aio_bh_call (/usr/src/debug/qemu-4.1.0/util/async.c:117)
#9 0xaaaac55d4543 in aio_bh_poll (/usr/src/debug/qemu-4.1.0/util/aio-posix.c:459)
#10 0xaaaac55cae0f in aio_dispatch (/usr/src/debug/qemu-4.1.0/util/async.c:268)
#11 0xffffb52d6a7b in g_main_context_dispatch (/usr/lib64/libglib-2.0.so.0+0x52a7b)
#12 0xaaaac55d1e3b(/usr/bin/qemu-kvm-4.1.0+0x1622e3b)
#13 0xaaaac4e314bb(/usr/bin/qemu-kvm-4.1.0+0xe824bb)
#14 0xaaaac47f45ef(/usr/bin/qemu-kvm-4.1.0+0x8455ef)
#15 0xffffb4bfef3f in __libc_start_main (/usr/lib64/libc.so.6+0x23f3f)
#16 0xaaaac47ffacb(/usr/bin/qemu-kvm-4.1.0+0x850acb)
Direct leak of 2 byte(s) in 2 object(s) allocated from:
#0 0xffffb597c20b in __interceptor_malloc (/usr/lib64/libasan.so.4+0xd320b)
#1 0xffffb52dcb1b in g_malloc (/usr/lib64/libglib-2.0.so.0+0x58b1b)
#2 0xffffb52f8143 in g_strdup (/usr/lib64/libglib-2.0.so.0+0x74143)
#3 0xaaaac5244893 in migrate_params_test_apply (/usr/src/debug/qemu-4.1.0/migration/migration.c:1382)
#4 0xaaaac52fdca7 in qmp_migrate_set_parameters (/usr/src/debug/qemu-4.1.0/qapi/qapi-commands-migration.c:192)
#5 0xaaaac551d543 in qmp_dispatch (/usr/src/debug/qemu-4.1.0/qapi/qmp-dispatch.c)
#6 0xaaaac52a0a8f in qmp_dispatch (/usr/src/debug/qemu-4.1.0/monitor/qmp.c:125)
#7 0xaaaac52a1c7f in monitor_qmp_dispatch (/usr/src/debug/qemu-4.1.0/monitor/qmp.c:214)
#8 0xaaaac55cb0cf in aio_bh_call (/usr/src/debug/qemu-4.1.0/util/async.c:117)
#9 0xaaaac55d4543 in aio_bh_poll (/usr/src/debug/qemu-4.1.0/util/aio-posix.c:459)
#10 0xaaaac55cae0f in in aio_dispatch (/usr/src/debug/qemu-4.1.0/util/async.c:268)
#11 0xffffb52d6a7b in g_main_context_dispatch (/usr/lib64/libglib-2.0.so.0+0x52a7b)
#12 0xaaaac55d1e3b(/usr/bin/qemu-kvm-4.1.0+0x1622e3b)
#13 0xaaaac4e314bb(/usr/bin/qemu-kvm-4.1.0+0xe824bb)
#14 0xaaaac47f45ef (/usr/bin/qemu-kvm-4.1.0+0x8455ef)
#15 0xffffb4bfef3f in __libc_start_main (/usr/lib64/libc.so.6+0x23f3f)
#16 0xaaaac47ffacb(/usr/bin/qemu-kvm-4.1.0+0x850acb)
Signed-off-by: Chuan Zheng <zhengchuan@huawei.com>
Reviewed-by: KeQian Zhu <zhukeqian1@huawei.com>
Reviewed-by: HaiLiang <zhang.zhanghailiang@huawei.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
|
|
move the vcpu throttling functionality into its own module.
This functionality is not specific to any accelerator,
and it is used currently by migration to slow down guests to try to
have migrations converge, and by the cocoa MacOS UI to throttle speed.
cpu-throttle contains the controls to adjust and inspect throttle
settings, start (set) and stop vcpu throttling, and the throttling
function itself that is run periodically on vcpus to make them take a nap.
Execution of the throttling function on all vcpus is triggered by a timer,
registered at module initialization.
No functionality change.
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20200629093504.3228-3-cfontana@suse.de>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
virtio,acpi: features, fixes, cleanups.
vdpa support
virtio-mem support
a handy script for disassembling acpi tables
misc fixes and cleanups
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Tue 07 Jul 2020 13:00:35 BST
# gpg: using RSA key 5D09FD0871C8F85B94CA8A0D281F0DB8D28D5469
# gpg: issuer "mst@redhat.com"
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>" [full]
# gpg: aka "Michael S. Tsirkin <mst@redhat.com>" [full]
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17 0970 C350 3912 AFBE 8E67
# Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA 8A0D 281F 0DB8 D28D 5469
* remotes/mst/tags/for_upstream: (41 commits)
vhost-vdpa: introduce vhost-vdpa net client
vhost-vdpa: introduce vhost-vdpa backend
vhost_net: introduce set_config & get_config
vhost: implement vhost_force_iommu method
vhost: introduce new VhostOps vhost_force_iommu
vhost: implement vhost_vq_get_addr method
vhost: introduce new VhostOps vhost_vq_get_addr
vhost: implement vhost_dev_start method
vhost: introduce new VhostOps vhost_dev_start
vhost: check the existence of vhost_set_iotlb_callback
virtio-pci: implement queue_enabled method
virtio-bus: introduce queue_enabled method
vhost_net: use the function qemu_get_peer
net: introduce qemu_get_peer
MAINTAINERS: add VT-d entry
docs: vhost-user: add Virtio status protocol feature
tests/acpi: remove stale allowed tables
numa: Auto-enable NUMA when any memory devices are possible
virtio-mem: Exclude unplugged memory during migration
virtio-mem: Add trace events
...
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
# Conflicts:
# hw/arm/virt.c
# hw/virtio/trace-events
|
|
real_dirty_pages becomes equal to total ram size after dirty log sync
in ram_init_bitmaps, the reason is that the bitmap of ramblock is
initialized to be all set, so old path counts them as "real dirty" at
beginning.
This causes wrong dirty rate and false positive throttling.
Signed-off-by: Keqian Zhu <zhukeqian1@huawei.com>
Message-Id: <20200622032037.31112-1-zhukeqian1@huawei.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
This function returns a boolean success and we're returning -1;
lets just use the 'out' error path.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Fixes: 58b7c17e226 ("Disable mlock around incoming postcopy")
Buglink: https://bugs.launchpad.net/qemu/+bug/1885720
Message-Id: <20200701093557.130096-1-dgilbert@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
COLO will copy all memory in a RAM block, disable discarding of RAM.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Tested-by: Lukas Straub <lukasstraub2@web.de>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Hailiang Zhang <zhang.zhanghailiang@huawei.com>
Cc: Juan Quintela <quintela@redhat.com>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20200626072248.78761-10-david@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
RDMA will pin all guest memory (as documented in docs/rdma.txt). We want
to disable RAM block discards - however, to keep it simple use
ram_block_discard_is_required() instead of inhibiting.
Note: It is not sufficient to limit disabling to pin_all. Even when only
conditionally pinning 1 MB chunks, as soon as one page within such a
chunk was discarded and one page not, the discarded pages will be pinned
as well.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Juan Quintela <quintela@redhat.com>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20200626072248.78761-9-david@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
The only remaining special case is postcopy. It cannot handle
concurrent discards yet, which would result in requesting already sent
pages from the source. Special-case it in virtio-balloon instead.
Introduce migration_in_incoming_postcopy(), to find out if incoming
postcopy is active.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Juan Quintela <quintela@redhat.com>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20200626072248.78761-7-david@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
|
|
qemu_rdma_registration_stop() uses the ERROR() macro to create, report
to stderr, and store an Error object. The stored Error object is
never used, and its memory is leaked.
Even where ERROR() doesn't leak, it is ill-advised. The whole point
of passing an Error to the caller is letting the caller handle the
error. Error handling may report to stderr, to somewhere else, or not
at all. Also reporting in the callee mixes up concerns that should be
kept separate. Since I don't know what reporting to stderr is
supposed to accomplish, I'm not touching it.
Commit 2a1bc8bde7 "migration/rdma: rdma_accept_incoming_migration fix
error handling" plugged the same leak in
rdma_accept_incoming_migration().
Plug the memory leak the same way: keep the report part, delete the
store part.
The report part uses fprintf(). If it's truly an error, it should use
error_report() instead. But I don't know, so I leave it alone, just
like commit 2a1bc8bde7 did.
Fixes: 2da776db4846eadcb808598a5d3484d149773c05
Cc: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Juan Quintela <quintela@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20200630090351.1247703-27-armbru@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
I'm not aware of any immediate bugs in qemu where a second runtime
evaluation of the arguments to MIN() or MAX() causes a problem, but
proactively preventing such abuse is easier than falling prey to an
unintended case down the road. At any rate, here's the conversation
that sparked the current patch:
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg05718.html
Update the MIN/MAX macros to only evaluate their argument once at
runtime; this uses typeof(1 ? (a) : (b)) to ensure that we are
promoting the temporaries to the same type as the final comparison (we
have to trigger type promotion, as typeof(bitfield) won't compile; and
we can't use typeof((a) + (b)) or even typeof((a) + 0), as some of our
uses of MAX are on void* pointers where such addition is undefined).
However, we are unable to work around gcc refusing to compile ({}) in
a constant context (such as the array length of a static variable),
even when only used in the dead branch of a __builtin_choose_expr(),
so we have to provide a second macro pair MIN_CONST and MAX_CONST for
use when both arguments are known to be compile-time constants and
where the result must also be usable as a constant; this second form
evaluates arguments multiple times but that doesn't matter for
constants. By using a void expression as the expansion if a
non-constant is presented to this second form, we can enlist the
compiler to ensure the double evaluation is not attempted on
non-constants.
Alas, as both macros now rely on compiler intrinsics, they are no
longer usable in preprocessor #if conditions; those will just have to
be open-coded or the logic rewritten into #define or runtime 'if'
conditions (but where the compiler dead-code-elimination will probably
still apply).
I tested that both gcc 10.1.1 and clang 10.0.0 produce errors for all
forms of macro mis-use. As the errors can sometimes be cryptic, I'm
demonstrating the gcc output:
Use of MIN when MIN_CONST is needed:
In file included from /home/eblake/qemu/qemu-img.c:25:
/home/eblake/qemu/include/qemu/osdep.h:249:5: error: braced-group within expression allowed only inside a function
249 | ({ \
| ^
/home/eblake/qemu/qemu-img.c:92:12: note: in expansion of macro ‘MIN’
92 | char array[MIN(1, 2)] = "";
| ^~~
Use of MIN_CONST when MIN is needed:
/home/eblake/qemu/qemu-img.c: In function ‘is_allocated_sectors’:
/home/eblake/qemu/qemu-img.c:1225:15: error: void value not ignored as it ought to be
1225 | i = MIN_CONST(i, n);
| ^
Use of MIN in the preprocessor:
In file included from /home/eblake/qemu/accel/tcg/translate-all.c:20:
/home/eblake/qemu/accel/tcg/translate-all.c: In function ‘page_check_range’:
/home/eblake/qemu/include/qemu/osdep.h:249:6: error: token "{" is not valid in preprocessor expressions
249 | ({ \
| ^
Fix the resulting callsites that used #if or computed a compile-time
constant min or max to use the new macros. cpu-defs.h is interesting,
as CPU_TLB_DYN_MAX_BITS is sometimes used as a constant and sometimes
dynamic.
It may be worth improving glib's MIN/MAX definitions to be saner, but
that is a task for another day.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200625162602.700741-1-eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
'remotes/dgilbert/tags/pull-migration-20200617a' into staging
Migration (and HMP and virtiofs) pull 2020-06-17
Migration:
HMP/migration and test changes from Mao Zhongyi
multifd fix from Laurent Vivier
HMP
qom-set partial reversion/change from David Hildenbrand
now you need -j to pass json format, but it's regained the
old 100M type format.
Memory leak fix from Pan Nengyuan
Virtiofs
fchmod seccomp fix from Max Reitz
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
# gpg: Signature made Wed 17 Jun 2020 19:34:58 BST
# gpg: using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A 9FA9 0516 331E BC5B FDE7
* remotes/dgilbert/tags/pull-migration-20200617a:
migration: fix multifd_send_pages() next channel
docs/xbzrle: update 'cache miss rate' and 'encoding rate' to docs
monitor/hmp-cmds: improvements for the 'info migrate'
monitor/hmp-cmds: add 'goto end' to reduce duplicate code.
monitor/hmp-cmds: delete redundant Error check before invoke hmp_handle_error()
monitor/hmp-cmds: don't silently output when running 'migrate_set_downtime' fails
monitor/hmp-cmds: add units for migrate_parameters
tests/migration: fix unreachable path in stress test
tests/migration: mem leak fix
hmp: Make json format optional for qom-set
qom-hmp-cmds: fix a memleak in hmp_qom_get
virtiofsd: Whitelist fchmod
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
It's reported an error of implicit conversion from "unsigned long" to
"double" when compiling with Clang 10. Simply make the encoding rate 0
when the encoded_size is 0.
Fixes: e460a4b1a4
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reported-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Wei Wang <wei.w.wang@intel.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200617201309.1640952-3-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
|
|
multifd_send_pages() loops around the available channels,
the next channel to use between two calls to multifd_send_pages() is stored
inside a local static variable, next_channel.
It works well, except if the number of channels decreases between two calls
to multifd_send_pages(). In this case, the loop can try to access the
data of a channel that doesn't exist anymore.
The problem can be triggered if we start a migration with a given number of
channels and then we cancel the migration to restart it with a lower number.
This ends generally with an error like:
qemu-system-ppc64: .../util/qemu-thread-posix.c:77: qemu_mutex_lock_impl: Assertion `mutex->initialized' failed.
This patch fixes the error by capping next_channel with the current number
of channels before using it.
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20200617113154.593233-1-lvivier@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-31-armbru@redhat.com>
|
|
This code is not related to hardware emulation.
Move it under accel/ with the other hypervisors.
Reviewed-by: Paul Durrant <paul@xen.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200508100222.7112-1-philmd@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
migration_rate_limit will erroneously ratelimit a shutdown socket,
which causes the migration thread to hang in ram_save_host_page
if the socket is shutdown.
Fix this by explicitly testing if the socket has errors or was
shutdown in migration_rate_limit.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <e79085bbe2d46dfa007dd41820194d5e2d4fcd80.1590007004.git.lukasstraub2@web.de>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
If the secondary has to failover during checkpointing, it still is
in the old state (i.e. different state than primary). Thus we can't
expose the primary state until after the checkpoint is sent.
This fixes sporadic connection reset of client connections during
failover.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <d4555dd5146a54518c4d9d4efd996b7c745c6687.1589193382.git.lukasstraub2@web.de>
Reviewed-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
If vmstate_loading is true, secondary_vm_do_failover will set failover
status to FAILOVER_STATUS_RELAUNCH and return success without initiating
failover. However, if there is an error during the vmstate_loading
section, failover isn't relaunched. Instead we then wait for
failover on colo_incoming_sem.
Fix this by relaunching failover even if there was an error. Also,
to make this work properly, set vmstate_loading to false when
returning during the vmstate_loading section.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <f60b0a8e2fadaaec792e04819dfc46951842d6ba.1589193382.git.lukasstraub2@web.de>
Reviewed-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
If we suceed in receiving ram state, but fail receiving the device
state, there will be a mismatch between the two.
Fix this by flushing the ram cache only after the vmstate has been
received.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <3289d007d494cb0e2f05b1cf4ae6a78d300fede3.1589193382.git.lukasstraub2@web.de>
Reviewed-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
cpu_synchronize_all_pre_loadvm() marks all vcpus as dirty, so the
registers are loaded from CPUState before we continue running
the vm. However if we failover during checkpoint, CPUState is not
initialized and the registers are loaded with garbage. This causes
guest hangs and crashes.
Fix this by using cpu_synchronize_all_states(), which initializes
CPUState from the current cpu registers additionally to marking
the vcpus as dirty.
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <9675031ce557b73ebd10e7bd20ebbf57f30b177c.1589193382.git.lukasstraub2@web.de>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
If multiple packets miscompare in a short timeframe, the semaphore
value will be increased multiple times. This causes multiple
checkpoints even if one would be sufficient.
Fix this by using a event instead of a semaphore for triggering
checkpoints. Now, checkpoint requests will be ignored until the
checkpoint event is sent to colo-compare (which releases the
miscompared packets).
Benchmark results (iperf3):
Client-to-server tcp:
without patch: ~66 Mbit/s
with patch: ~61 Mbit/s
Server-to-client tcp:
without patch: ~702 Kbit/s
with patch: ~16 Mbit/s
Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Message-Id: <fd601ba1beb524aada54ba66e87ebfc12cf4574b.1589193382.git.lukasstraub2@web.de>
Reviewed-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
When error happen in initializing 'rdma_return_path', we should cleanup rdma context
before g_free(rdma) to avoid some memleaks. This patch fix that.
Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-Id: <20200508100755.7875-3-pannengyuan@huawei.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
|
|
'rdma' is NULL when taking the first error branch in rdma_start_incoming_migration.
And it will cause a null pointer access in label 'err'. Fix that.
Fixes: 59c59c67ee6b0327ae932deb303caa47919aeb1e
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-Id: <20200508100755.7875-2-pannengyuan@huawei.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Note this is CID 1428762
|
|
It actually never worked with libvirt, as auto-generated names are
different on source and destination.
It's unsafe and useless to migrate by auto-generated node-names, so
let's forbid it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200521220648.3255-7-vsementsov@virtuozzo.com>
[eblake: shorter subject line]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Andrey Shinkevich <andrey.shinkevich@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200521220648.3255-6-vsementsov@virtuozzo.com>
[eblake: shorter subject line]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Important thing for bitmap migration is to select destination block
node to obtain the migrated bitmap.
Prepatch, on source we use bdrv_get_device_or_node_name() to identify
the node, and on target we do bdrv_lookup_bs.
bdrv_get_device_or_node_name() returns blk name only for direct
children of blk. So, bitmaps of direct children of blks are migrated by
blk name and others - by node name.
Old libvirt is unprepared to bitmap migration by node-name,
node-names are mostly auto-generated. So actually only migration by blk
name works for it.
Newer libvirt will use new interface (which will be added soon) to
specify node-mapping for bitmaps migration explicitly. Still, let's
improve the current behavior a bit.
Now, consider classic libvirt migrations assisted by mirror block job:
mirror block job inserts filter, so our source is not a direct child of
blk, and bitmaps are migrated by node-names. And this just doesn't work
with auto-generated node names.
Let's fix it by using blk-name even if some implicit filters are
inserted.
Note2: we, of course, can't skip filters and use blk name to migrate
bitmaps in filtered node by blk name for this blk if these filters have
named bitmaps which should be migrated.
Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1652424
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200521220648.3255-4-vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
[eblake: comment typo fix, shorter subject line]
Signed-off-by: Eric Blake <eblake@redhat.com>
|
|
Split out handling one bs, it is needed for the following commit, which
will handle BlockBackends separately.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200521220648.3255-2-vsementsov@virtuozzo.com>
[eblake: shorter subject line]
Signed-off-by: Eric Blake <eblake@redhat.com>
|