Currently, VFIO bytes_transferred is not reset properly:
1. bytes_transferred is not reset after a VM snapshot (so a migration
following a snapshot will report incorrect value).
2. bytes_transferred is a single counter for all VFIO devices, however
upon migration failure it is reset multiple times, by each VFIO
device.
Fix it by introducing a new function vfio_reset_bytes_transferred() and
calling it during migration and snapshot start.
Remove existing bytes_transferred reset in VFIO migration state
notifier, which is not needed anymore.
Fixes: 3710586caa5d ("qapi: Add VFIO devices migration stats in Migration stats")
Signed-off-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Reviewed-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Implement switchover ack logic. This prevents the source from stopping
the VM and completing the migration until an ACK is received from the
destination that it's OK to do so.
To achieve this, a new SaveVMHandlers handler switchover_ack_needed()
and a new return path message MIG_RP_MSG_SWITCHOVER_ACK are added.
The switchover_ack_needed() handler is called during migration setup in
the destination to check if switchover ack is used by the migrated
device.
When switchover is approved by all migrated devices in the destination
that support this capability, the MIG_RP_MSG_SWITCHOVER_ACK return path
message is sent to the source to notify it that it's OK to do
switchover.
Signed-off-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Tested-by: YangHang Liu <yanghliu@redhat.com>
Acked-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Actually global_state_store() can never fail. Let's get rid of extra
error paths.
To make things clear, use new runstate_get() and use same approach for
global_state_store() and global_state_store_running().
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20230517123752.21615-3-vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
These way we can make them atomic and use this functions from any
place. I also moved all functions that use rate_limit to
migration-stats.
Functions got renamed, they are not qemu_file anymore.
qemu_file_rate_limit -> migration_rate_exceeded
qemu_file_set_rate_limit -> migration_rate_set
qemu_file_get_rate_limit -> migration_rate_get
qemu_file_reset_rate_limit -> migration_rate_reset
qemu_file_acct_rate_limit -> migration_rate_account.
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20230515195709.63843-6-quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Function is already quite long.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20230508130909.65420-7-quintela@redhat.com>
Change all the functions that use it. It was already passed as
uint64_t.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20230504113841.23130-8-quintela@redhat.com>
migration_stats is just too long, and it is going to have more than
ram counters in the near future.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
There is already include/qemu/stats.h, so stats.h was a bad idea.
We want this file to not depend on anything else, we will move all the
migration counters/stats to this struct.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
For VMS_ARRAY typed vmsd fields, also dump the number of entries in the
array in -vmstate-dump.
Without such information, vmstate static checker can report false negatives
of incompatible vmsd on VMS_ARRAY typed fields, when the src/dst do not
have the same type of array defined. It's because in the checker we only
check against size of fields within a VMSD field.
One example: e1000e used to have a field defined as a boolean array with 5
entries, then removed it and replaced it with UNUSED (in 31e3f318c8b535):
- VMSTATE_BOOL_ARRAY(core.eitr_intr_pending, E1000EState,
- E1000E_MSIX_VEC_NUM),
+ VMSTATE_UNUSED(E1000E_MSIX_VEC_NUM),
It's a legal replacement but vmstate static checker is not happy with it,
because it checks only against the "size" field between the two
fields (here one is BOOL_ARRAY, the other is UNUSED):
For BOOL_ARRAY:
{
"field": "core.eitr_intr_pending",
"version_id": 0,
"field_exists": false,
"size": 1
},
For UNUSED:
{
"field": "unused",
"version_id": 0,
"field_exists": false,
"size": 5
},
It's not the script to blame because there's just not enough information
dumped to show the total size of the entry for an array. Add it.
Note that this will not break old vmstate checker because the field will
just be ignored.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Instead of print it to STDERR, bring the error upwards so that it can be
reported via QMP responses.
E.g.:
{ "execute": "migrate-set-capabilities" ,
"arguments": { "capabilities":
[ { "capability": "postcopy-ram", "state": true } ] } }
{ "error":
{ "class": "GenericError",
"desc": "Postcopy is not supported: Host backend files need to be TMPFS
or HUGETLBFS only" } }
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Once that we are there, we rename the function to migrate_block()
to be consistent with all other capabilities.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
We move there all capabilities helpers from migration.c.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
---
Following David advise:
- looked through the history, capabilities are newer than 2012, so we
can remove that bit of the header.
- This part is posterior to Anthony.
Original Author is Orit. Once there,
I put myself. Peter Xu also did quite a bit of work here.
Anyone else wants/needs to be there? I didn't search too hard
because nobody asked before to be added.
What do you think?
It is clear from the context what that means, and such a long name
with the extra long names of the capabilities make very difficilut to
stay inside the 80 columns limit.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Once that res_compatible is removed, they don't make sense anymore.
We remove the _only preffix. And to make things clearer we rename
them to must_precopy and can_postcopy.
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Nothing assigns to it after previous commit.
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Postcopy with preempt-mode enabled needs two channels to communicate. The
order of channel establishment is not guaranteed. It can happen that the
dest QEMU got the preempt channel connection request before the main
channel is established, then the migration may make no progress even during
precopy due to the wrong order.
To fix it, create the preempt channel only if we know the main channel is
established.
For a general postcopy migration, we delay it until postcopy_start(),
that's where we already went through some part of precopy on the main
channel. To make sure dest QEMU has already established the channel, we
wait until we got the first PONG received. That's something we do at the
start of precopy when postcopy enabled so it's guaranteed to happen sooner
or later.
For a postcopy recovery, we delay it to qemu_savevm_state_resume_prepare()
where we'll have round trips of data on bitmap synchronizations, which
means the main channel must have been established.
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
I called the helper function from the wrong top level function.
This code was introduced in:
commit c8df4a7aeffcb46020f610526eea621fa5b0cd47
Author: Juan Quintela <quintela@redhat.com>
Date: Mon Oct 3 02:00:03 2022 +0200
migration: Split save_live_pending() into state_pending_*
We split the function into to:
- state_pending_estimate: We estimate the remaining state size without
stopping the machine.
- state pending_exact: We calculate the exact amount of remaining
state.
Thanks to Avihai Horon <avihaih@nvidia.com> for finding it.
Fixes:c8df4a7aeffcb46020f610526eea621fa5b0cd47
When we introduced that patch, we enden calling
state_pending_estimate() helper from qemu_savevm_statepending_exact()
and
state_pending_exact() helper from qemu_savevm_statepending_estimate()
This patch fixes it.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Perform a check on vmsd structures during test runs in the hope
of catching any missing terminators and other simple screwups.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
We fairly regularly forget VMSTATE_END_OF_LIST markers off descriptions;
given that the current check is only for ->name being NULL, sometimes
we get unlucky and the code apparently works and no one spots the error.
Explicitly add a flag, VMS_END that should be set, and assert it is
set during the traversal.
Note: This can't go in until we update the copy of vmstate.h in slirp.
Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
For virtio-mem, we want to have the plugged/unplugged state of memory
blocks available before migrating any actual RAM content, and perform
sanity checks before touching anything on the destination. This
information is immutable on the migration source while migration is active,
We want to use this information for proper preallocation support with
migration: currently, we don't preallocate memory on the migration target,
and especially with hugetlb, we can easily run out of hugetlb pages during
RAM migration and will crash (SIGBUS) instead of catching this gracefully
via preallocation.
Migrating device state via a VMSD before we start iterating is currently
impossible: the only approach that would be possible is avoiding a VMSD
and migrating state manually during save_setup(), to be restored during
load_state().
Let's allow for migrating device state via a VMSD early, during the
setup phase in qemu_savevm_state_setup(). To keep it simple, we
indicate applicable VMSD's using an "early_setup" flag.
Note that only very selected devices (i.e., ones seriously messing with
RAM setup) are supposed to make use of such early state migration.
While at it, also use a bool for the "unmigratable" member.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>S
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
... and store it in the migration state. This is a preparation for
storing selected vmds's already in qemu_savevm_state_setup().
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Let's move more code into vmstate_save(), reducing code duplication and
preparing for reuse of vmstate_save() in qemu_savevm_state_setup(). We
have to move vmstate_save() to make the compiler happy.
We'll now also trace from qemu_save_device_state(), triggering the same
tracepoints as previously called from
qemu_savevm_state_complete_precopy_non_iterable() only. Note that
qemu_save_device_state() ignores iterable device state, such as RAM,
and consequently doesn't trigger some other trace points (e.g.,
trace_savevm_state_setup()).
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Until previous commit, save_live_pending() was used for ram. Now with
the split into state_pending_estimate() and state_pending_exact() it
is not needed anymore, so remove them.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We split the function into to:
- state_pending_estimate: We estimate the remaining state size without
stopping the machine.
- state pending_exact: We calculate the exact amount of remaining
state.
The only "device" that implements different functions for _estimate()
and _exact() is ram.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We have two inclusion loops:
block/block.h
-> block/block-global-state.h
-> block/block-common.h
-> block/blockjob.h
-> block/block.h
block/block.h
-> block/block-io.h
-> block/block-common.h
-> block/blockjob.h
-> block/block.h
I believe these go back to Emanuele's reorganization of the block API,
merged a few months ago in commit d7e2fe4aac8.
Fortunately, breaking them is merely a matter of deleting unnecessary
includes from headers, and adding them back in places where they are
now missing.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20221221133551.3967339-2-armbru@redhat.com>
* Run docker probe only if docker or podman are available
The docker probe uses "sudo -n" which can cause an e-mail with a security warning
each time when configure is run. Therefore run docker probe only if either docker
or podman are available.
That avoids the problematic "sudo -n" on build environments which have neither
docker nor podman installed.
Fixes: c4575b59155e2e00 ("configure: store container engine in config-host.mak")
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-Id: <20221030083510.310584-1-sw@weilnetz.de>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20221117172532.538149-2-alex.bennee@linaro.org>
* tests/avocado/machine_aspeed.py: Reduce noise on the console for SDK tests
The Aspeed SDK images are based on OpenBMC which starts a lot of
services. The output noise on the console can break from time to time
the test waiting for the logging prompt.
Change the U-Boot bootargs variable to add "quiet" to the kernel
command line and reduce the output volume. This also drops the test on
the CPU id which was nice to have but not essential.
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20221104075347.370503-1-clg@kaod.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20221117172532.538149-3-alex.bennee@linaro.org>
* tests/docker: allow user to override check target
This is useful when trying to bisect a particular failing test behind
a docker run. For example:
make docker-test-clang@fedora \
TARGET_LIST=arm-softmmu \
TEST_COMMAND="meson test qtest-arm/qos-test" \
J=9 V=1
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-4-alex.bennee@linaro.org>
* docs/devel: add a maintainers section to development process
We don't currently have a clear place in the documentation to describe
the roles and responsibilities of a maintainer. Lets create one so we
can. I've moved a few small bits out of other files to try and keep
everything in one place.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-5-alex.bennee@linaro.org>
* docs/devel: make language a little less code centric
We welcome all sorts of patches.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-6-alex.bennee@linaro.org>
* docs/devel: simplify the minimal checklist
The bullet points are quite long and contain process tips. Move those
bits of the bullet to the relevant sections and link to them. Use a
table for nicer formatting of the checklist.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-7-alex.bennee@linaro.org>
* docs/devel: try and improve the language around patch review
It is important that contributors take the review process seriously
and we collaborate in a respectful way while avoiding personal
attacks. Try and make this clear in the language.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-8-alex.bennee@linaro.org>
* tests/avocado: Raise timeout for boot_linux.py:BootLinuxPPC64.test_pseries_tcg
On my machine, a debug build of QEMU takes about 260 seconds to
complete this test, so with the current timeout value of 180 seconds
it always times out. Double the timeout value to 360 so the test
definitely has enough time to complete.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20221110142901.3832318-1-peter.maydell@linaro.org>
Message-Id: <20221117172532.538149-9-alex.bennee@linaro.org>
* tests/avocado: introduce alpine virt test for CI
The boot_linux tests download and run a full cloud image boot and
start a full distro. While the ability to test the full boot chain is
worthwhile it is perhaps a little too heavy weight and causes issues
in CI. Fix this by introducing a new alpine linux ISO boot in
machine_aarch64_virt.
This boots a fully loaded -cpu max with all the bells and whistles in
31s on my machine. A full debug build takes around 180s on my machine
so we set a more generous timeout to cover that.
We don't add a test for lesser GIC versions although there is some
coverage for that already in the boot_xen.py tests. If we want to
introduce more comprehensive testing we can do it with a custom kernel
and initrd rather than a full distro boot.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-10-alex.bennee@linaro.org>
* tests/avocado: skip aarch64 cloud TCG tests in CI
We now have a much lighter weight test in machine_aarch64_virt which
tests the full boot chain in less time. Rename the tests while we are
at it to make it clear it is a Fedora cloud image.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221117172532.538149-11-alex.bennee@linaro.org>
* gitlab: integrate coverage report
This should hopefully give is nice coverage information about what our
tests (or at least the subset we are running) have hit. Ideally we
would want a way to trigger coverage on tests likely to be affected by
the current commit.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221117172532.538149-12-alex.bennee@linaro.org>
* vhost: mask VIRTIO_F_RING_RESET for vhost and vhost-user devices
Commit 69e1c14aa2 ("virtio: core: vq reset feature negotation support")
enabled VIRTIO_F_RING_RESET by default for all virtio devices.
This feature is not currently emulated by QEMU, so for vhost and
vhost-user devices we need to make sure it is supported by the offloaded
device emulation (in-kernel or in another process).
To do this we need to add VIRTIO_F_RING_RESET to the features bitmap
passed to vhost_get_features(). This way it will be masked if the device
does not support it.
This issue was initially discovered with vhost-vsock and vhost-user-vsock,
and then also tested with vhost-user-rng which confirmed the same issue.
They fail when sending features through VHOST_SET_FEATURES ioctl or
VHOST_USER_SET_FEATURES message, since VIRTIO_F_RING_RESET is negotiated
by the guest (Linux >= v6.0), but not supported by the device.
Fixes: 69e1c14aa2 ("virtio: core: vq reset feature negotation support")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1318
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20221121101101.29400-1-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Acked-by: Jason Wang <jasowang@redhat.com>
* tests: acpi: whitelist DSDT before moving PRQx to _SB scope
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20221121153613.3972225-2-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* acpi: x86: move RPQx field back to _SB scope
Commit 47a373faa6b2 (acpi: pc/q35: drop ad-hoc PCI-ISA bridge AML routines and let bus ennumeration generate AML)
moved ISA bridge AML generation to respective devices and was using
aml_alias() to provide PRQx fields in _SB. scope. However, it turned
out that SeaBIOS was not able to process Alias opcode when parsing DSDT,
resulting in lack of keyboard during boot (SeaBIOS console, grub, FreeDOS).
While fix for SeaBIOS is posted
https://mail.coreboot.org/hyperkitty/list/seabios@seabios.org/thread/RGPL7HESH5U5JRLEO6FP77CZVHZK5J65/
fixed SeaBIOS might not make into QEMU-7.2 in time.
Hence this workaround that puts PRQx back into _SB scope
and gets rid of aliases in ISA bridge description, so
DSDT will be parsable by broken SeaBIOS.
That brings back hardcoded references to ISA bridge
PCI0.S08.P40C/PCI0.SF8.PIRQ
where middle part now is auto generated based on slot it's
plugged in, but it should be fine as bridge initialization
also hardcodes PCI address of the bridge so it can't ever
move. Once QEMU tree has fixed SeaBIOS blob, we should be able
to drop this part and revert back to alias based approach
Reported-by: Volker Rümelin <vr_qemu@t-online.de>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20221121153613.3972225-3-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* tests: acpi: x86: update expected DSDT after moving PRQx fields in _SB scope
Expected DSDT changes,
pc:
- Field (P40C, ByteAcc, NoLock, Preserve)
+ Scope (\_SB)
{
- PRQ0, 8,
- PRQ1, 8,
- PRQ2, 8,
- PRQ3, 8
+ Field (PCI0.S08.P40C, ByteAcc, NoLock, Preserve)
+ {
+ PRQ0, 8,
+ PRQ1, 8,
+ PRQ2, 8,
+ PRQ3, 8
+ }
}
- Alias (PRQ0, \_SB.PRQ0)
- Alias (PRQ1, \_SB.PRQ1)
- Alias (PRQ2, \_SB.PRQ2)
- Alias (PRQ3, \_SB.PRQ3)
q35:
- Field (PIRQ, ByteAcc, NoLock, Preserve)
- {
- PRQA, 8,
- PRQB, 8,
- PRQC, 8,
- PRQD, 8,
- Offset (0x08),
- PRQE, 8,
- PRQF, 8,
- PRQG, 8,
- PRQH, 8
+ Scope (\_SB)
+ {
+ Field (PCI0.SF8.PIRQ, ByteAcc, NoLock, Preserve)
+ {
+ PRQA, 8,
+ PRQB, 8,
+ PRQC, 8,
+ PRQD, 8,
+ Offset (0x08),
+ PRQE, 8,
+ PRQF, 8,
+ PRQG, 8,
+ PRQH, 8
+ }
}
- Alias (PRQA, \_SB.PRQA)
- Alias (PRQB, \_SB.PRQB)
- Alias (PRQC, \_SB.PRQC)
- Alias (PRQD, \_SB.PRQD)
- Alias (PRQE, \_SB.PRQE)
- Alias (PRQF, \_SB.PRQF)
- Alias (PRQG, \_SB.PRQG)
- Alias (PRQH, \_SB.PRQH)
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20221121153613.3972225-4-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* MAINTAINERS: add mst to list of biosbits maintainers
Adding Michael's name to the list of bios bits maintainers so that all changes
and fixes into biosbits framework can go through his tree and he is notified.
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Ani Sinha <ani@anisinha.ca>
Message-Id: <20221111151138.36988-1-ani@anisinha.ca>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* tests/avocado: configure acpi-bits to use avocado timeout
Instead of using a hardcoded timeout, just rely on Avocado's built-in
test case timeout. This helps avoid timeout issues on machines where 60
seconds is not sufficient.
Signed-off-by: John Snow <jsnow@redhat.com>
Message-Id: <20221115212759.3095751-1-jsnow@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Ani Sinha <ani@anisinha.ca>
* acpi/tests/avocado/bits: keep the work directory when BITS_DEBUG is set in env
Debugging bits issue often involves running the QEMU command line manually
outside of the avocado environment with the generated ISO. Hence, its
inconvenient if the iso gets cleaned up after the test has finished. This change
makes sure that the work directory is kept after the test finishes if the test
is run with BITS_DEBUG=1 in the environment so that the iso is available for use
with the QEMU command line.
CC: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Ani Sinha <ani@anisinha.ca>
Message-Id: <20221117113630.543495-1-ani@anisinha.ca>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* virtio: disable error for out of spec queue-enable
Virtio 1.0 is pretty clear that features have to be
negotiated before enabling VQs. Unfortunately Seabios
ignored this ever since gaining 1.0 support (UEFI is ok).
Comment the error out for now, and add a TODO.
Fixes: 3c37f8b8d1 ("virtio: introduce virtio_queue_enable()")
Cc: "Kangjie Xu" <kangjie.xu@linux.alibaba.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20221121200339.362452-1-mst@redhat.com>
* hw/loongarch: Add default stdout uart in fdt
Add "chosen" subnode into LoongArch fdt, and set it's
"stdout-path" prop to uart node.
Signed-off-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Message-Id: <20221115114923.3372414-1-yangxiaojuan@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
* hw/loongarch: Fix setprop_sized method in fdt rtc node.
Fix setprop_sized method in fdt rtc node.
Signed-off-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Message-Id: <20221116040300.3459818-1-yangxiaojuan@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
* hw/loongarch: Replace the value of uart info with macro
Using macro to replace the value of uart info such as addr, size
in acpi_build method.
Signed-off-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Message-Id: <20221115115008.3372489-1-yangxiaojuan@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
* target/arm: Don't do two-stage lookup if stage 2 is disabled
In get_phys_addr_with_struct(), we call get_phys_addr_twostage() if
the CPU supports EL2. However, we don't check here that stage 2 is
actually enabled. Instead we only check that inside
get_phys_addr_twostage() to skip stage 2 translation. This means
that even if stage 2 is disabled we still tell the stage 1 lookup to
do its page table walks via stage 2.
This works by luck for normal CPU accesses, but it breaks for debug
accesses, which are used by the disassembler and also by semihosting
file reads and writes, because the debug case takes a different code
path inside S1_ptw_translate().
This means that setups that use semihosting for file loads are broken
(a regression since 7.1, introduced in recent ptw refactoring), and
that sometimes disassembly in debug logs reports "unable to read
memory" rather than showing the guest insns.
Fix the bug by hoisting the "is stage 2 enabled?" check up to
get_phys_addr_with_struct(), so that we handle S2 disabled the same
way we do the "no EL2" case, with a simple single stage lookup.
Reported-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20221121212404.1450382-1-peter.maydell@linaro.org
* target/arm: Use signed quantity to represent VMSAv8-64 translation level
The LPA2 extension implements 52-bit virtual addressing for 4k and 16k
translation granules, and for the former, this means an additional level
of translation is needed. This means we start counting at -1 instead of
0 when doing a walk, and so 'level' is now a signed quantity, and should
be typed as such. So turn it from uint32_t into int32_t.
This avoids a level of -1 getting misinterpreted as being >= 3, and
terminating a page table walk prematurely with a bogus output address.
Cc: Peter Maydell <peter.maydell@linaro.org>
Cc: Philippe Mathieu-Daudé <f4bug@amsat.org>
Cc: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
* Update VERSION for v7.2.0-rc2
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
* tests/avocado: Update the URLs of the advent calendar images
The qemu-advent-calendar.org server will be decommissioned soon.
I've mirrored the images that we use for the QEMU CI to gitlab,
so update their URLs to point to the new location.
Message-Id: <20221121102436.78635-1-thuth@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* tests/qtest: Decrease the amount of output from the qom-test
The logs in the gitlab-CI have a size constraint, and sometimes
we already hit this limit. The biggest part of the log then seems
to be filled by the qom-test, so we should decrease the size of
the output - which can be done easily by not printing the path
for each property, since the path has already been logged at the
beginning of each node that we handle here.
However, if we omit the path, we should make sure to not recurse
into child nodes in between, so that it is clear to which node
each property belongs. Thus store the children and links in a
temporary list and recurse only at the end of each node, when
all properties have already been printed.
Message-Id: <20221121194240.149268-1-thuth@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* tests/avocado: use new rootfs for orangepi test
The old URL wasn't stable. I suspect the current URL will only be
stable for a few months so maybe we need another strategy for hosting
rootfs snapshots?
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20221118113309.1057790-1-alex.bennee@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* Revert "usbredir: avoid queuing hello packet on snapshot restore"
Run state is also in RUN_STATE_PRELAUNCH while "-S" is used.
This reverts commit 0631d4b448454ae8a1ab091c447e3f71ab6e088a
Signed-off-by: Joelle van Dyne <j@getutm.app>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The original commit broke the usage of usbredir with libvirt, which
starts every domain with "-S".
This workaround is no longer needed because the usbredir behavior
has been fixed in the meantime:
https://gitlab.freedesktop.org/spice/usbredir/-/merge_requests/61
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Message-Id: <1689cec3eadcea87255e390cb236033aca72e168.1669193161.git.jtomko@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* gtk: disable GTK Clipboard with a new meson option
The GTK Clipboard implementation may cause guest hangs.
Therefore implement new configure switch: --enable-gtk-clipboard,
as a meson option disabled by default, which warns in the help
text about the experimental nature of the feature.
Regenerate the meson build options to include it.
The initialization of the clipboard is gtk.c, as well as the
compilation of gtk-clipboard.c are now conditional on this new
option to be set.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1150
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
Message-Id: <20221121135538.14625-1-cfontana@suse.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* hw/usb/hcd-xhci.c: spelling: tranfer
Fixes: effaf5a240e03020f4ae953e10b764622c3e87cc
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Message-Id: <20221105114851.306206-1-mjt@msgid.tls.msk.ru>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* ui/gtk: prevent ui lock up when dpy_gl_update called again before current draw event occurs
A warning, "qemu: warning: console: no gl-unblock within" followed by
guest scanout lockup can happen if dpy_gl_update is called in a row
and the second call is made before gd_draw_event scheduled by the first
call is taking place. This is because draw call returns without decrementing
gl_block ref count if the dmabuf was already submitted as shown below.
(gd_gl_area_draw/gd_egl_draw)
if (dmabuf) {
if (!dmabuf->draw_submitted) {
return;
} else {
dmabuf->draw_submitted = false;
}
}
So it should not schedule any redundant draw event in case draw_submitted is
already set in gd_egl_fluch/gd_gl_area_scanout_flush.
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Vivek Kasireddy <vivek.kasireddy@intel.com>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20221021192315.9110-1-dongwon.kim@intel.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* hw/usb/hcd-xhci: Reset the XHCIState with device_cold_reset()
Currently the hcd-xhci-pci and hcd-xhci-sysbus devices, which are
mostly wrappers around the TYPE_XHCI device, which is a direct
subclass of TYPE_DEVICE. Since TYPE_DEVICE devices are not on any
qbus and do not get automatically reset, the wrapper devices both
reset the TYPE_XHCI device in their own reset functions. However,
they do this using device_legacy_reset(), which will reset the device
itself but not any bus it has.
Switch to device_cold_reset(), which avoids using a deprecated
function and also propagates reset along any child buses.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20221014145423.2102706-1-peter.maydell@linaro.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* hw/audio/intel-hda: don't reset codecs twice
Currently the intel-hda device has a reset method which manually
resets all the codecs by calling device_legacy_reset() on them. This
means they get reset twice, once because child devices on a qbus get
reset before the parent device's reset method is called, and then
again because we're manually resetting them.
Drop the manual reset call, and ensure that codecs are still reset
when the guest does a reset via ICH6_GCTL_RESET by using
device_cold_reset() (which resets all the devices on the qbus as well
as the device itself) instead of a direct call to the reset function.
This is a slight ordering change because the (only) codec reset now
happens before the controller registers etc are reset, rather than
once before and then once after, but the codec reset function
hda_audio_reset() doesn't care.
This lets us drop a use of device_legacy_reset(), which is
deprecated.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221014142632.2092404-2-peter.maydell@linaro.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* hw/audio/intel-hda: Drop unnecessary prototype
The only use of intel_hda_reset() is after its definition, so we
don't need to separately declare its prototype at the top of the
file; drop the unnecessary line.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221014142632.2092404-3-peter.maydell@linaro.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* add syx snapshot extras
* it compiles!
* virtiofsd: Add `sigreturn` to the seccomp whitelist
The virtiofsd currently crashes on s390x. This is because of a
`sigreturn` system call. See audit log below:
type=SECCOMP msg=audit(1669382477.611:459): auid=4294967295 uid=0 gid=0 ses=4294967295 subj=system_u:system_r:virtd_t:s0-s0:c0.c1023 pid=6649 comm="virtiofsd" exe="/usr/libexec/virtiofsd" sig=31 arch=80000016 syscall=119 compat=0 ip=0x3fff15f748a code=0x80000000AUID="unset" UID="root" GID="root" ARCH=s390x SYSCALL=sigreturn
Signed-off-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Reviewed-by: German Maglione <gmaglione@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221125143946.27717-1-mhartmay@linux.ibm.com>
* libvhost-user: Fix wrong type of argument to formatting function (reported by LGTM)
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-Id: <20220422070144.1043697-2-sw@weilnetz.de>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-2-sw@weilnetz.de>
* libvhost-user: Fix format strings
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220422070144.1043697-3-sw@weilnetz.de>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-3-sw@weilnetz.de>
* libvhost-user: Fix two more format strings
This fix is required for 32 bit hosts. The bug was detected by CI
for arm-linux, but is also relevant for i386-linux.
Reported-by: Stefan Hajnoczi <stefanha@gmail.com>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-4-sw@weilnetz.de>
* libvhost-user: Add format attribute to local function vu_panic
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220422070144.1043697-4-sw@weilnetz.de>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-5-sw@weilnetz.de>
* MAINTAINERS: Add subprojects/libvhost-user to section "vhost"
Signed-off-by: Stefan Weil <sw@weilnetz.de>
[Michael agreed to act as maintainer for libvhost-user via email in
https://lore.kernel.org/qemu-devel/20221123015218-mutt-send-email-mst@kernel.org/.
--Stefan]
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-6-sw@weilnetz.de>
* Add G_GNUC_PRINTF to function qemu_set_info_str and fix related issues
With the G_GNUC_PRINTF function attribute the compiler detects
two potential insecure format strings:
../../../net/stream.c:248:31: warning: format string is not a string literal (potentially insecure) [-Wformat-security]
qemu_set_info_str(&s->nc, uri);
^~~
../../../net/stream.c:322:31: warning: format string is not a string literal (potentially insecure) [-Wformat-security]
qemu_set_info_str(&s->nc, uri);
^~~
There are also two other warnings:
../../../net/socket.c:182:35: warning: zero-length gnu_printf format string [-Wformat-zero-length]
182 | qemu_set_info_str(&s->nc, "");
| ^~
../../../net/stream.c:170:35: warning: zero-length gnu_printf format string [-Wformat-zero-length]
170 | qemu_set_info_str(&s->nc, "");
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221126152507.283271-7-sw@weilnetz.de>
* del ramfile
* update seabios source from 1.16.0 to 1.16.1
git shortlog rel-1.16.0..rel-1.16.1
===================================
Gerd Hoffmann (3):
malloc: use variable for ZoneHigh size
malloc: use large ZoneHigh when there is enough memory
virtio-blk: use larger default request size
Igor Mammedov (1):
acpi: parse Alias object
Volker Rümelin (2):
pci: refactor the pci_config_*() functions
reset: force standard PCI configuration access
Xiaofei Lee (1):
virtio-blk: Fix incorrect type conversion in virtio_blk_op()
Xuan Zhuo (2):
virtio-mmio: read/write the hi 32 features for mmio
virtio: finalize features before using device
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* update seabios binaries to 1.16.1
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
* fix for non i386 archs
* replay: Fix declaration of replay_read_next_clock
Fixes the build with gcc 13:
replay/replay-time.c:34:6: error: conflicting types for \
'replay_read_next_clock' due to enum/integer mismatch; \
have 'void(ReplayClockKind)' [-Werror=enum-int-mismatch]
34 | void replay_read_next_clock(ReplayClockKind kind)
| ^~~~~~~~~~~~~~~~~~~~~~
In file included from ../qemu/replay/replay-time.c:14:
replay/replay-internal.h:139:6: note: previous declaration of \
'replay_read_next_clock' with type 'void(unsigned int)'
139 | void replay_read_next_clock(unsigned int kind);
| ^~~~~~~~~~~~~~~~~~~~~~
Fixes: 8eda206e090 ("replay: recording and replaying clock ticks")
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Wilfred Mallawa <wilfred.mallawa@wdc.com>
Reviewed-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221129010547.284051-1-richard.henderson@linaro.org>
* hw/display/qxl: Have qxl_log_command Return early if no log_cmd handler
Only 3 command types are logged: no need to call qxl_phys2virt()
for the other types. Using different cases will help to pass
different structure sizes to qxl_phys2virt() in a pair of commits.
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221128202741.4945-2-philmd@linaro.org>
* hw/display/qxl: Document qxl_phys2virt()
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221128202741.4945-3-philmd@linaro.org>
* hw/display/qxl: Pass requested buffer size to qxl_phys2virt()
Currently qxl_phys2virt() doesn't check for buffer overrun.
In order to do so in the next commit, pass the buffer size
as argument.
For QXLCursor in qxl_render_cursor() -> qxl_cursor() we
verify the size of the chunked data ahead, checking we can
access 'sizeof(QXLCursor) + chunk->data_size' bytes.
Since in the SPICE_CURSOR_TYPE_MONO case the cursor is
assumed to fit in one chunk, no change are required.
In SPICE_CURSOR_TYPE_ALPHA the ahead read is handled in
qxl_unpack_chunks().
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221128202741.4945-4-philmd@linaro.org>
* hw/display/qxl: Avoid buffer overrun in qxl_phys2virt (CVE-2022-4144)
Have qxl_get_check_slot_offset() return false if the requested
buffer size does not fit within the slot memory region.
Similarly qxl_phys2virt() now returns NULL in such case, and
qxl_dirty_one_surface() aborts.
This avoids buffer overrun in the host pointer returned by
memory_region_get_ram_ptr().
Fixes: CVE-2022-4144 (out-of-bounds read)
Reported-by: Wenxu Yin (@awxylitol)
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1336
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221128202741.4945-5-philmd@linaro.org>
* hw/display/qxl: Assert memory slot fits in preallocated MemoryRegion
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221128202741.4945-6-philmd@linaro.org>
* block-backend: avoid bdrv_unregister_buf() NULL pointer deref
bdrv_*() APIs expect a valid BlockDriverState. Calling them with bs=NULL
leads to undefined behavior.
Jonathan Cameron reported this following NULL pointer dereference when a
VM with a virtio-blk device and a memory-backend-file object is
terminated:
1. qemu_cleanup() closes all drives, setting blk->root to NULL
2. qemu_cleanup() calls user_creatable_cleanup(), which results in a RAM
block notifier callback because the memory-backend-file is destroyed.
3. blk_unregister_buf() is called by virtio-blk's BlockRamRegistrar
notifier callback and undefined behavior occurs.
Fixes: baf422684d73 ("virtio-blk: use BDRV_REQ_REGISTERED_BUF optimization hint")
Co-authored-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221121211923.1993171-1-stefanha@redhat.com>
* target/arm: Set TCGCPUOps.restore_state_to_opc for v7m
This setting got missed, breaking v7m.
Fixes: 56c6c98df85c ("target/arm: Convert to tcg_ops restore_state_to_opc")
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1347
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Evgeny Ermakov <evgeny.v.ermakov@gmail.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221129204146.550394-1-richard.henderson@linaro.org>
* Update VERSION for v7.2.0-rc3
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
* hooks are now post mem access
* tests/qtests: override "force-legacy" for gpio virtio-mmio tests
The GPIO device is a VIRTIO_F_VERSION_1 devices but running with a
legacy MMIO interface we miss out that feature bit causing confusion.
For the GPIO test force the mmio bus to support non-legacy so we can
properly test it.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1333
Message-Id: <20221130112439.2527228-2-alex.bennee@linaro.org>
Acked-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* vhost: enable vrings in vhost_dev_start() for vhost-user devices
Commit 02b61f38d3 ("hw/virtio: incorporate backend features in features")
properly negotiates VHOST_USER_F_PROTOCOL_FEATURES with the vhost-user
backend, but we forgot to enable vrings as specified in
docs/interop/vhost-user.rst:
If ``VHOST_USER_F_PROTOCOL_FEATURES`` has not been negotiated, the
ring starts directly in the enabled state.
If ``VHOST_USER_F_PROTOCOL_FEATURES`` has been negotiated, the ring is
initialized in a disabled state and is enabled by
``VHOST_USER_SET_VRING_ENABLE`` with parameter 1.
Some vhost-user front-ends already did this by calling
vhost_ops.vhost_set_vring_enable() directly:
- backends/cryptodev-vhost.c
- hw/net/virtio-net.c
- hw/virtio/vhost-user-gpio.c
But most didn't do that, so we would leave the vrings disabled and some
backends would not work. We observed this issue with the rust version of
virtiofsd [1], which uses the event loop [2] provided by the
vhost-user-backend crate where requests are not processed if vring is
not enabled.
Let's fix this issue by enabling the vrings in vhost_dev_start() for
vhost-user front-ends that don't already do this directly. Same thing
also in vhost_dev_stop() where we disable vrings.
[1] https://gitlab.com/virtio-fs/virtiofsd
[2] https://github.com/rust-vmm/vhost/blob/240fc2966/crates/vhost-user-backend/src/event_loop.rs#L217
Fixes: 02b61f38d3 ("hw/virtio: incorporate backend features in features")
Reported-by: German Maglione <gmaglione@redhat.com>
Tested-by: German Maglione <gmaglione@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Acked-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20221123131630.52020-1-sgarzare@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20221130112439.2527228-3-alex.bennee@linaro.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* hw/virtio: add started_vu status field to vhost-user-gpio
As per the fix to vhost-user-blk in f5b22d06fb (vhost: recheck dev
state in the vhost_migration_log routine) we really should track the
connection and starting separately.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20221130112439.2527228-4-alex.bennee@linaro.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* hw/virtio: generalise CHR_EVENT_CLOSED handling
..and use for both virtio-user-blk and virtio-user-gpio. This avoids
the circular close by deferring shutdown due to disconnection until a
later point. virtio-user-blk already had this mechanism in place so
generalise it as a vhost-user helper function and use for both blk and
gpio devices.
While we are at it we also fix up vhost-user-gpio to re-establish the
event handler after close down so we can reconnect later.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20221130112439.2527228-5-alex.bennee@linaro.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* include/hw: VM state takes precedence in virtio_device_should_start
The VM status should always preempt the device status for these
checks. This ensures the device is in the correct state when we
suspend the VM prior to migrations. This restores the checks to the
order they where in before the refactoring moved things around.
While we are at it lets improve our documentation of the various
fields involved and document the two functions.
Fixes: 9f6bcfd99f (hw/virtio: move vm_running check to virtio_device_started)
Fixes: 259d69c00b (hw/virtio: introduce virtio_device_should_start)
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Christian Borntraeger <borntraeger@linux.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20221130112439.2527228-6-alex.bennee@linaro.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
* hw/nvme: fix aio cancel in format
There are several bugs in the async cancel code for the Format command.
Firstly, cancelling a format operation neglects to set iocb->ret as well
as clearing the iocb->aiocb after cancelling the underlying aiocb which
causes the aio callback to ignore the cancellation. Trivial fix.
Secondly, and worse, because the request is queued up for posting to the
CQ in a bottom half, if the cancellation is due to the submission queue
being deleted (which calls blk_aio_cancel), the req structure is
deallocated in nvme_del_sq prior to the bottom half being schedulued.
Fix this by simply removing the bottom half, there is no reason to defer
it anyway.
Fixes: 3bcf26d3d619 ("hw/nvme: reimplement format nvm to allow cancellation")
Reported-by: Jonathan Derrick <jonathan.derrick@linux.dev>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
* hw/nvme: fix aio cancel in flush
Make sure that iocb->aiocb is NULL'ed when cancelling.
Fix a potential use-after-free by removing the bottom half and enqueuing
the completion directly.
Fixes: 38f4ac65ac88 ("hw/nvme: reimplement flush to allow cancellation")
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
* hw/nvme: fix aio cancel in zone reset
If the zone reset operation is cancelled but the block unmap operation
completes normally, the callback will continue resetting the next zone
since it neglects to check iocb->ret which will have been set to
-ECANCELED. Make sure that this is checked and bail out if an error is
present.
Secondly, fix a potential use-after-free by removing the bottom half and
enqueuing the completion directly.
Fixes: 63d96e4ffd71 ("hw/nvme: reimplement zone reset to allow cancellation")
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
* hw/nvme: fix aio cancel in dsm
When the DSM operation is cancelled asynchronously, we set iocb->ret to
-ECANCELED. However, the callback function only checks the return value
of the completed aio, which may have completed succesfully prior to the
cancellation and thus the callback ends up continuing the dsm operation
instead of bailing out. Fix this.
Secondly, fix a potential use-after-free by removing the bottom half and
enqueuing the completion directly.
Fixes: d7d1474fd85d ("hw/nvme: reimplement dsm to allow cancellation")
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
* hw/nvme: remove copy bh scheduling
Fix a potential use-after-free by removing the bottom half and enqueuing
the completion directly.
Fixes: 796d20681d9b ("hw/nvme: reimplement the copy command to allow aio cancellation")
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
* target/i386: allow MMX instructions with CR4.OSFXSR=0
MMX state is saved/restored by FSAVE/FRSTOR so the instructions are
not illegal opcodes even if CR4.OSFXSR=0. Make sure that validate_vex
takes into account the prefix and only checks HF_OSFXSR_MASK in the
presence of an SSE instruction.
Fixes: 20581aadec5e ("target/i386: validate VEX prefixes via the instructions' exception classes", 2022-10-18)
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1350
Reported-by: Helge Konetzka (@hejko on gitlab.com)
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
* target/i386: Always completely initialize TranslateFault
In get_physical_address, the canonical address check failed to
set TranslateFault.stage2, which resulted in an uninitialized
read from the struct when reporting the fault in x86_cpu_tlb_fill.
Adjust all error paths to use structure assignment so that the
entire struct is always initialized.
Reported-by: Daniel Hoffman <dhoff749@gmail.com>
Fixes: 9bbcf372193a ("target/i386: Reorg GET_HPHYS")
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20221201074522.178498-1-richard.henderson@linaro.org>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1324
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
* hw/loongarch/virt: Add cfi01 pflash device
Add cfi01 pflash device for LoongArch virt machine
Signed-off-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20221130100647.398565-1-yangxiaojuan@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
* Sync pc on breakpoints
* tests/qtest/migration-test: Fix unlink error and memory leaks
When running the migration test compiled with Clang from Fedora 37
and sanitizers enabled, there is an error complaining about unlink():
../tests/qtest/migration-test.c:1072:12: runtime error: null pointer
passed as argument 1, which is declared to never be null
/usr/include/unistd.h:858:48: note: nonnull attribute specified here
SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior
../tests/qtest/migration-test.c:1072:12 in
(test program exited with status code 1)
TAP parsing error: Too few tests run (expected 33, got 20)
The data->clientcert and data->clientkey pointers can indeed be unset
in some tests, so we have to check them before calling unlink() with
those.
While we're at it, I also noticed that the code is only freeing
some but not all of the allocated strings in this function, and
indeed, valgrind is also complaining about memory leaks here.
So let's call g_free() on all allocated strings to avoid leaking
memory here.
Message-Id: <20221125083054.117504-1-thuth@redhat.com>
Tested-by: Bin Meng <bmeng@tinylab.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* target/s390x/tcg: Fix and improve the SACF instruction
The SET ADDRESS SPACE CONTROL FAST instruction is not privileged, it can be
used from problem space, too. Just the switching to the home address space
is privileged and should still generate a privilege exception. This bug is
e.g. causing programs like Java that use the "getcpu" vdso kernel function
to crash (see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=990417#26 ).
While we're at it, also check if DAT is not enabled. In that case the
instruction is supposed to generate a special operation exception.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/655
Message-Id: <20221201184443.136355-1-thuth@redhat.com>
Reviewed-by: Ilya Leoshkevich <iii@linux.ibm.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* hw/display/next-fb: Fix comment typo
Signed-off-by: Evgeny Ermakov <evgeny.v.ermakov@gmail.com>
Message-Id: <20221125160849.23711-1-evgeny.v.ermakov@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
* fix dev snapshots
* working syx snaps
* Revert "hw/loongarch/virt: Add cfi01 pflash device"
This reverts commit 14dccc8ea6ece7ee63273144fb55e4770a05e0fd.
Signed-off-by: Song Gao <gaosong@loongson.cn>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20221205113007.683505-1-gaosong@loongson.cn>
* Update VERSION for v7.2.0-rc4
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Ani Sinha <ani@anisinha.ca>
Signed-off-by: John Snow <jsnow@redhat.com>
Signed-off-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Claudio Fontana <cfontana@suse.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Dongwon Kim <dongwon.kim@intel.com>
Signed-off-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Evgeny Ermakov <evgeny.v.ermakov@gmail.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Co-authored-by: Stefan Weil <sw@weilnetz.de>
Co-authored-by: Cédric Le Goater <clg@kaod.org>
Co-authored-by: Alex Bennée <alex.bennee@linaro.org>
Co-authored-by: Peter Maydell <peter.maydell@linaro.org>
Co-authored-by: Stefano Garzarella <sgarzare@redhat.com>
Co-authored-by: Igor Mammedov <imammedo@redhat.com>
Co-authored-by: Ani Sinha <ani@anisinha.ca>
Co-authored-by: John Snow <jsnow@redhat.com>
Co-authored-by: Michael S. Tsirkin <mst@redhat.com>
Co-authored-by: Xiaojuan Yang <yangxiaojuan@loongson.cn>
Co-authored-by: Stefan Hajnoczi <stefanha@redhat.com>
Co-authored-by: Ard Biesheuvel <ardb@kernel.org>
Co-authored-by: Thomas Huth <thuth@redhat.com>
Co-authored-by: Joelle van Dyne <j@getutm.app>
Co-authored-by: Claudio Fontana <cfontana@suse.de>
Co-authored-by: Michael Tokarev <mjt@tls.msk.ru>
Co-authored-by: Dongwon Kim <dongwon.kim@intel.com>
Co-authored-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Co-authored-by: Stefan Weil via <qemu-devel@nongnu.org>
Co-authored-by: Gerd Hoffmann <kraxel@redhat.com>
Co-authored-by: Richard Henderson <richard.henderson@linaro.org>
Co-authored-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Co-authored-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Co-authored-by: Evgeny Ermakov <evgeny.v.ermakov@gmail.com>
Co-authored-by: Klaus Jensen <k.jensen@samsung.com>
Co-authored-by: Paolo Bonzini <pbonzini@redhat.com>
Co-authored-by: Song Gao <gaosong@loongson.cn>
Snapshot loading only expects to call deterministic handlers, not
non-deterministic ones. So introduce a way of registering handlers that
won't be called when reseting for snapshots.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Message-id: 20221025004327.568476-2-Jason@zx2c4.com
[PMM: updated json doc comment with Markus' text; fixed
checkpatch style nit]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
To allow postcopy recovery, the ram fast load (preempt-only) dest QEMU thread
needs similar handling on fault tolerance. When ram_load_postcopy() fails,
instead of stopping the thread it halts with a semaphore, preparing to be
kicked again when recovery is detected.
A mutex is introduced to make sure there's no concurrent operation upon the
socket. To make it simple, the fast ram load thread will take the mutex during
its whole procedure, and only release it if it's paused. The fast-path socket
will be properly released by the main loading thread safely when there's
network failures during postcopy with that mutex held.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185506.27257-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Create a new socket for postcopy to be prepared to send postcopy requested
pages via this specific channel, so as to not get blocked by precopy pages.
A new thread is also created on dest qemu to receive data from this new channel
based on the ram_load_postcopy() routine.
The ram_load_postcopy(POSTCOPY) branch and the thread has not started to
function, and that'll be done in follow up patches.
Cleanup the new sockets on both src/dst QEMUs, meanwhile look after the new
thread too to make sure it'll be recycled properly.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220707185502.27149-1-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: With Peter's fix to quieten compiler warning on
start_migration
Now that all QEMUFile callbacks are removed, the entire concept can be
deleted.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
With this change, all QEMUFile usage is backed by QIOChannel at
last.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Wrap long lines
The name 'ftell' gives the misleading impression that the QEMUFile
objects are seekable. This is not the case, as in general we just
have an opaque stream. The users of this method are only interested
in the total bytes processed. This switches to a new name that
reflects the intended usage.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Wrapped long line
Previously migration didn't have an easy way to cleanup the listening
transport, migrate recovery only allows to execute once. That's done with a
trick flag in postcopy_recover_triggered.
Now the facility is already there.
Drop postcopy_recover_triggered and instead allows a new migrate-recover to
release the previous listener transport.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220331150857.74406-8-peterx@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Snapshots run also under the BQL, so they all are
in the global state API. The aiocontext lock that they hold
is currently an overkill and in future could be removed.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220303151616.325444-23-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Following the bdrv_activate renaming, change also the name
of the respective callers.
bdrv_invalidate_cache_all -> bdrv_activate_all
blk_invalidate_cache -> blk_activate
test_sync_op_invalidate_cache -> test_sync_op_activate
No functional change intended.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220209105452.1694545-5-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Postcopy create threads. A common manner is we init a sem and use it to sync
with the thread. Namely, we have fault_thread_sem and listen_thread_sem and
they're only used for this.
Make it a shared infrastructure so it's easier to create yet another thread.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-7-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Postcopy handles huge pages in a special way that currently we can only have
one "channel" to transfer the page.
It's because when we install pages using UFFDIO_COPY, we need to have the whole
huge page ready, it also means we need to have a temp huge page when trying to
receive the whole content of the page.
Currently all maintainance around this tmp page is global: firstly we'll
allocate a temp huge page, then we maintain its status mostly within
ram_load_postcopy().
To enable multiple channels for postcopy, the first thing we need to do is to
prepare N temp huge pages as caching, one for each channel.
Meanwhile we need to maintain the tmp huge page status per-channel too.
To give some example, some local variables maintained in ram_load_postcopy()
are listed; they are responsible for maintaining temp huge page status:
- all_zero: this keeps whether this huge page contains all zeros
- target_pages: this counts how many target pages have been copied
- host_page: this keeps the host ptr for the page to install
Move all these fields to be together with the temp huge pages to form a new
structure called PostcopyTmpPage. Then for each (future) postcopy channel, we
need one structure to keep the state around.
For vanilla postcopy, obviously there's only one channel. It contains both
precopy and postcopy pages.
This patch teaches the dest migration node to start realize the possible number
of postcopy channels by introducing the "postcopy_channels" variable. Its
value is calculated when setup postcopy on dest node (during POSTCOPY_LISTEN
phase).
Vanilla postcopy will have channels=1, but when postcopy-preempt capability is
enabled (in the future), we will boost it to 2 because even during partial
sending of a precopy huge page we still want to preempt it and start sending
the postcopy requested page right away (so we start to keep two temp huge
pages; more if we want to enable multifd). In this patch there's a TODO marked
for that; so far the channels is always set to 1.
We need to send one "host huge page" on one channel only and we cannot split
them, because otherwise the data upon the same huge page can locate on more
than one channel so we need more complicated logic to manage. One temp host
huge page for each channel will be enough for us for now.
Postcopy will still always use the index=0 huge page even after this patch.
However it prepares for the latter patches where it can start to use multiple
channels (which needs src intervention, because only src knows which channel we
should use).
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-5-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Fixed up long line
Remove the old two tracepoints and they're even near each other:
trace_loadvm_postcopy_handle_run_cpu_sync()
trace_loadvm_postcopy_handle_run_vmstart()
Add trace_loadvm_postcopy_handle_run_bh() with a finer granule trace.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-4-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
The enablement of postcopy listening has a few steps, add a few tracepoints to
be there ready for some basic measurements for them.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-3-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
It'll be easier to read the name rather than index of sub-cmd when debugging.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220301083925.33483-2-peterx@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Should qemu_savevm_state_iterate() encounter a failure when calling a
particular save_live_iterate function, report the error code returned
by the function.
Signed-off-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
It's efficient, but hackish to call yank unregister calls in channel_close(),
especially it'll be hard to debug when qemu crashed with some yank function
leaked.
Remove that hack, but instead explicitly unregister yank functions at the
places where needed, they are:
(on src)
- migrate_fd_cleanup
- postcopy_pause
(on dst)
- migration_incoming_state_destroy
- postcopy_pause_incoming
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20210722175841.938739-6-peterx@redhat.com>
Reviewed-by: Lukas Straub <lukasstraub2@web.de>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
