Some devices need to distinguish cold start reset from waking up from a
suspended state. This patch adds new value to the enum, and updates the
i386 wakeup method to use this new reset type.
Signed-off-by: Juraj Marcin
---
docs/devel/reset.rst| 7 +++
hw/i386/pc.c| 2
device can access the reset type in the hold
phase. The last patch of the series implements the final support in the
hold phase of the virtio-mem reset callback and exposes
VIRTIO_MEM_F_PERSISTENT_SUSPEND to the kernel.
[1]: https://lore.kernel.org/all/20240318120645.105664-1-da...@redhat.com/
Juraj
ype instead.
Signed-off-by: Juraj Marcin
---
hw/arm/aspeed.c| 4 ++--
hw/arm/mps2-tz.c | 4 ++--
hw/core/reset.c| 7 ++-
hw/hppa/machine.c | 4 ++--
hw/i386/microvm.c | 4 ++--
hw/i386/pc.c | 6 +++---
hw/ppc/pegasos
feature is not exposed by QEMU.
This patch adds the code to skip the reset on wake-up and exposes
theVIRTIO_MEM_F_PERSISTENT_SUSPEND feature to the guest kernel driver
when suspending is possible in QEMU (currently only x86).
Signed-off-by: Juraj Marcin
---
hw/virtio/virtio-mem.c | 10
list, implements the necessary methods and replaces
qemu_[un]register_reset() calls with qemu_[un]register_resettable().
Signed-off-by: Juraj Marcin
---
hw/virtio/virtio-mem.c | 39 ++
include/hw/virtio/virtio-mem.h | 4
2 files changed, 29 insertions
On Tue, Aug 6, 2024 at 6:40 PM David Hildenbrand wrote:
>
> On 06.08.24 18:07, Juraj Marcin wrote:
> > Currently, both qemu_devices_reset() and MachineClass->reset() use
> > ShutdownCause for the reason of the reset. However, the Resettable
> > interface uses ResetState
On Thu, Aug 8, 2024 at 2:18 PM Peter Maydell wrote:
>
> On Tue, 6 Aug 2024 at 17:08, Juraj Marcin wrote:
> >
> > Some devices need to distinguish cold start reset from waking up from a
> > suspended state. This patch adds new value to the enum, and updates the
> >
On Thu, Aug 8, 2024 at 5:47 PM Peter Maydell wrote:
>
> On Thu, 8 Aug 2024 at 13:37, David Hildenbrand wrote:
> >
> > On 08.08.24 14:25, Peter Maydell wrote:
> > > On Tue, 6 Aug 2024 at 17:08, Juraj Marcin wrote:
> > >>
> > >> LegacyReset d
instead.
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
Reviewed-by: Peter Maydell
---
hw/arm/aspeed.c| 4 ++--
hw/arm/mps2-tz.c | 4 ++--
hw/core/reset.c| 5 +
hw/hppa/machine.c | 4 ++--
hw/i386/microvm.c | 4 ++--
hw/i386
feature is not exposed by QEMU.
This patch adds the code to skip the reset on wake-up and exposes
theVIRTIO_MEM_F_PERSISTENT_SUSPEND feature to the guest kernel driver
when suspending is possible in QEMU (currently only x86).
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
hw
:
- Removed unnecessary include directives changes
- Updated RESET_TYPE_WAKEUP documentation
- Removed unnecessary interface from VirtIOMEMClass type info
Juraj Marcin (4):
reset: Use ResetType for qemu_devices_reset() and
MachineClass::reset()
reset: Add RESET_TYPE_WAKEUP
virtio-mem: Use new
Some devices need to distinguish cold start reset from waking up from a
suspended state. This patch adds new value to the enum, and updates the
i386 wakeup method to use this new reset type.
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
docs/devel/reset.rst| 8
hw
to use the new Resettable framework and replaces
qemu_[un]register_reset() calls with qemu_[un]register_resettable().
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
hw/virtio/virtio-mem.c | 38 +-
include/hw/virtio/virtio-mem.h | 4
On Tue, Aug 13, 2024 at 6:37 PM Peter Maydell wrote:
>
> On Tue, 13 Aug 2024 at 16:39, Juraj Marcin wrote:
> >
> > Some devices need to distinguish cold start reset from waking up from a
> > suspended state. This patch adds new value to the enum, and updates the
>
Hi Peter,
On Tue, Aug 20, 2024 at 1:56 PM Peter Maydell wrote:
>
> On Tue, 20 Aug 2024 at 12:40, David Hildenbrand wrote:
> >
> > On 14.08.24 14:32, Juraj Marcin wrote:
> > > On Tue, Aug 13, 2024 at 6:37 PM Peter Maydell
> > > wrote:
> > >>
&
VirtIOMEMClass type info
Juraj Marcin (4):
reset: Use ResetType for qemu_devices_reset() and
MachineClass::reset()
reset: Add RESET_TYPE_WAKEUP
virtio-mem: Use new Resettable framework instead of LegacyReset
virtio-mem: Add support for suspend+wake-up with plugged memory
docs/devel/reset.rst
Some devices need to distinguish cold start reset from waking up from a
suspended state. This patch adds new value to the enum, and updates the
i386 wakeup method to use this new reset type.
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
docs/devel/reset.rst| 11
instead.
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
Reviewed-by: Peter Maydell
---
hw/arm/aspeed.c| 4 ++--
hw/arm/mps2-tz.c | 4 ++--
hw/core/reset.c| 5 +
hw/hppa/machine.c | 4 ++--
hw/i386/microvm.c | 4 ++--
hw/i386
to use the new Resettable framework and replaces
qemu_[un]register_reset() calls with qemu_[un]register_resettable().
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
hw/virtio/virtio-mem.c | 38 +-
include/hw/virtio/virtio-mem.h | 4
feature is not exposed by QEMU.
This patch adds the code to skip the reset on wake-up and exposes
theVIRTIO_MEM_F_PERSISTENT_SUSPEND feature to the guest kernel driver
when suspending is possible in QEMU (currently only x86).
Signed-off-by: Juraj Marcin
Reviewed-by: David Hildenbrand
---
hw
ill initially allocate.
> + */
> +if (s->nr_slots < KVM_MEMSLOTS_NR_ALLOC_DEFAULT) {
> +ret = -EINVAL;
> +fprintf(stderr, "KVM max supported number of slots (%d) too small\n",
> +s->nr_slots);
> +goto err;
> +}
> +
> s->nr_as = kvm_check_extension(s, KVM_CAP_MULTI_ADDRESS_SPACE);
> if (s->nr_as <= 1) {
> s->nr_as = 1;
> diff --git a/accel/kvm/trace-events b/accel/kvm/trace-events
> index 37626c1ac5..ad2ae6fca5 100644
> --- a/accel/kvm/trace-events
> +++ b/accel/kvm/trace-events
> @@ -36,3 +36,4 @@ kvm_io_window_exit(void) ""
> kvm_run_exit_system_event(int cpu_index, uint32_t event_type) "cpu_index %d,
> system_even_type %"PRIu32
> kvm_convert_memory(uint64_t start, uint64_t size, const char *msg) "start
> 0x%" PRIx64 " size 0x%" PRIx64 " %s"
> kvm_memory_fault(uint64_t start, uint64_t size, uint64_t flags) "start 0x%"
> PRIx64 " size 0x%" PRIx64 " flags 0x%" PRIx64
> +kvm_slots_grow(unsigned int old, unsigned int new) "%u -> %u"
> --
> 2.45.0
>
--
Juraj Marcin
Hi Peter,
On Thu, Sep 5, 2024 at 6:00 PM Peter Xu wrote:
>
> On Thu, Sep 05, 2024 at 05:32:46PM +0200, Juraj Marcin wrote:
> > Hi Peter,
>
> Hi, Juraj,
>
> [...]
>
> > > unsigned int kvm_get_max_memslots(void)
> > > {
> > > KVMS
, setting capabilities when the migration is still
running leads to an error.
By moving the wait before the setup, we ensure this does not happen.
Cc: Peter Xu
Signed-off-by: Juraj Marcin
---
tests/qtest/migration-test.c | 12 ++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff
This patch adds a new version of qtest_init_with_env() that allows
specifying QMP capabilities that should be enabled during handshake.
This is useful for example if a test needs out-of-band execution of QMP
commands, it can initialize with the oob capability.
Signed-off-by: Juraj Marcin
recovery.
Signed-off-by: Juraj Marcin
---
tests/qtest/migration/framework.c | 23 +--
tests/qtest/migration/framework.h | 2 ++
tests/qtest/migration/migration-qmp.c | 2 +-
3 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/tests/qtest/migration
series, adds a new qtest initialization function
which accepts capabilities the test wants to enable. This allows a test
to enable the oob capability.
The second patch then enables the oob capability in migration tests and
changes the execution of the migrate-recover command.
Juraj Marcin (2
On 2025-01-30 16:04, Peter Xu wrote:
> On Thu, Jan 30, 2025 at 05:11:36PM +0100, Juraj Marcin wrote:
> > When there are no page requests from the destination side and the
> > connection breaks, the destination might not be aware of it. This is
> > the case for example with a
On 2025-01-31 10:03, Peter Xu wrote:
> On Fri, Jan 31, 2025 at 02:42:41PM +0100, Juraj Marcin wrote:
> > On 2025-01-30 16:04, Peter Xu wrote:
> > > On Thu, Jan 30, 2025 at 05:11:36PM +0100, Juraj Marcin wrote:
> > > > When there are no page requests fro
on parameter, by default it is
6 ms. By setting this parameter to 0, sending keepalive messages
can be disabled completely.
Signed-off-by: Juraj Marcin
---
migration/migration.c| 12
migration/migration.h| 1 +
migration/options.c | 35 ++
| 1 +
> migration/savevm.h | 6 +-
> migration/migration.c | 209 +++-
> migration/savevm.c | 116
> migration/vmstate.c | 6 +-
> tests/qtest/libqos/libqos.c | 3 +-
> migration/trace-events
On 2025-03-24 11:12, Daniel P. Berrangé wrote:
> On Wed, Mar 19, 2025 at 05:36:20PM +0100, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > The default idle period for TCP connection could be even 2 hours.
> > However, in some cases, the application needs to be awar
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
option, but only on client-side sockets. However, this option is also
useful for server-side sockets, so they can check if a client is still
tuozzo.com/
---
V2:
- moved socket options setting into a common function for both server
and client sockets (suggested by Vladimir)
Juraj Marcin (2):
util/qemu-sockets: Add support for keep-alive flag to passive sockets
utils/qemu-sockets: Introduce keep-alive-idle-period inet socket
optio
From: Juraj Marcin
The default idle period for TCP connection could be even 2 hours.
However, in some cases, the application needs to be aware of a
connection issue much sooner.
This is the case, for example, for postcopy live migration. If there is
no traffic from the migration destination
Hi Daniel,
On 2025-03-24 10:30, Daniel P. Berrangé wrote:
> On Wed, Mar 19, 2025 at 05:36:19PM +0100, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
> > introduces the keep-alive flag, which enab
From: Juraj Marcin
This is done in preparation for enabling the SO_KEEPALIVE support for
server sockets and adding settings for more TCP keep-alive socket
options.
Signed-off-by: Juraj Marcin
---
util/qemu-sockets.c | 29 +++--
1 file changed, 19 insertions(+), 10
From: Juraj Marcin
With the default TCP stack configuration, it could be even 2 hours
before the connection times out due to the other side not being
reachable. However, in some cases, the application needs to be aware of
a connection issue much sooner.
This is the case, for example, for
t copy fix and the common function
setting socket options into a separate commit
- refactored inet_listen_saddr()
V2:
- moved socket options setting into a common function for both server
and client sockets (suggested by Vladimir)
Juraj Marcin (5):
io: Fix partial stru
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
option, but only on client-side sockets. However, this option is also
useful for server-side sockets, so they can check if a client is still
On 2025-03-14 20:24, Vladimir Sementsov-Ogievskiy wrote:
> On 03.03.25 17:33, Juraj Marcin wrote:
> > Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
> > introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
> > option, but only on client-
Hi Daniel
On 2025-04-11 16:49, Daniel P. Berrangé wrote:
> On Fri, Apr 11, 2025 at 02:54:29PM +0100, Daniel P. Berrangé wrote:
> > On Tue, Apr 08, 2025 at 01:25:04PM +0200, Juraj Marcin wrote:
> > > From: Juraj Marcin
> > >
> > > With the default TCP sta
cleanups for postcopy
>
> migration/migration.c | 28 ++--
> migration/multifd.c| 6 +-
> migration/trace-events | 4 ++--
> 3 files changed, 13 insertions(+), 25 deletions(-)
>
> --
> 2.35.3
>
Both patches look good to me.
Reviewed-by: Juraj Marcin
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, but this flag is not copied together
with other options in qio_dns_resolver_lookup_sync_inet().
This patch fixes this issue and also prevents future ones by copying the
entire
From: Juraj Marcin
To get a listening socket, we need to first create a socket, try binding
it to a certain port, and lastly starting listening to it. Each of these
operations can fail due to various reasons, one of them being that the
requested address/port is already in use. In such case, the
ram_page_hint_collect(rs, &next_block, &next_page);
> +} else {
> +next_block = rs->last_seen_block;
> +next_page = rs->last_page;
> +}
> +
> +pss_init(pss, next_block, next_page);
>
> while (true){
> if (!get_queued_page(rs, pss)) {
> @@ -2339,6 +2425,13 @@ static void ram_save_cleanup(void *opaque)
> ram_state_cleanup(rsp);
> }
>
> +static void ram_page_hint_reset(PageLocationHint *hint)
> +{
> +hint->location.block = NULL;
> +hint->location.offset = 0;
> +hint->valid = false;
> +}
> +
> static void ram_state_reset(RAMState *rs)
> {
> int i;
> @@ -2351,6 +2444,8 @@ static void ram_state_reset(RAMState *rs)
> rs->last_page = 0;
> rs->last_version = ram_list.version;
> rs->xbzrle_started = false;
> +
> +ram_page_hint_reset(&rs->page_hint);
> }
>
> #define MAX_WAIT 50 /* ms, half buffered_file limit */
> --
> 2.48.1
>
Reviewed-by: Juraj Marcin
Hi Prasad,
On 2025-02-15 18:01, Prasad Pandit wrote:
> From: Prasad Pandit
>
> Enable Multifd and Postcopy migration together.
> The migration_ioc_process_incoming() routine
> checks magic value sent on each channel and
> helps to properly setup multifd and postcopy
> channels.
>
> The Precopy
tuozzo.com/
Juraj Marcin (2):
util/qemu-sockets: Add support for keep-alive flag to passive sockets
utils/qemu-sockets: Introduce keep-alive-idle-period inet socket
option
io/dns-resolver.c | 6 +
meson.build | 2 ++
qapi/sockets.json | 9 +--
util/qemu-sock
'qio_dns_resolver_lookup_sync_inet()'
where the keep-alive flag is not copied together with other attributes.
Signed-off-by: Juraj Marcin
---
io/dns-resolver.c | 2 ++
qapi/sockets.json | 4 ++--
util/qemu-sockets.c | 19 +--
3 files changed, 17 insertions(+), 8 deletions(-)
diff
ault value is 0, which means system configuration is used, and no
custom value is set.
Signed-off-by: Juraj Marcin
---
io/dns-resolver.c | 4
meson.build | 2 ++
qapi/sockets.json | 5 +
util/qemu-sockets.c | 39 +++
4 files changed, 50
On 2025-02-20 11:06, Fabiano Rosas wrote:
> Peter Xu writes:
>
> > On the incoming migration side, QEMU uses a coroutine to load all the VM
> > states. Inside, it may reference MigrationState on global states like
> > migration capabilities, parameters, error state, shared mutexes and more.
> >
t
> before incoming")
> Signed-off-by: Peter Xu
> ---
> migration/options.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Reviewed-by: Juraj Marcin
suitable for a parameter dump instead. Hence
> move it over.
>
> Signed-off-by: Peter Xu
> ---
> migration/migration-hmp-cmds.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Reviewed-by: Juraj Marcin
s: [\n");
> +
> +for (addr = info->socket_address; addr; addr = addr->next) {
> +char *s = socket_uri(addr->value);
> + monitor_printf(mon, "\t%s\n", s);
> +g_free(s);
> }
> +monitor_pri
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, but this flag is not copied together
with other options in qio_dns_resolver_lookup_sync_inet().
This patch fixes this issue and also prevents future ones by copying the
entire
From: Juraj Marcin
With the default TCP stack configuration, it could be even 2 hours
before the connection times out due to the other side not being
reachable. However, in some cases, the application needs to be aware of
a connection issue much sooner.
This is the case, for example, for
From: Juraj Marcin
Before refactoring the inet_parse() function, this patch adds tests
verifying the inet address parsing works correctly.
Signed-off-by: Juraj Marcin
---
tests/unit/test-util-sockets.c | 161 +
1 file changed, 161 insertions(+)
diff --git a
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
option, but only on client-side sockets. However, this option is also
useful for server-side sockets, so they can check if a client is still
From: Juraj Marcin
To get a listening socket, we need to first create a socket, try binding
it to a certain port, and lastly starting listening to it. Each of these
operations can fail due to various reasons, one of them being that the
requested address/port is already in use. In such case, the
From: Juraj Marcin
This is done in preparation for enabling the SO_KEEPALIVE support for
server sockets and adding settings for more TCP keep-alive socket
options.
Signed-off-by: Juraj Marcin
---
util/qemu-sockets.c | 29 +++--
1 file changed, 19 insertions(+), 10
From: Juraj Marcin
Currently, the inet address parser cannot handle multiple options where
one is prefixed with the name of the other. For example, with the
'keep-alive-idle' option added, the current parser cannot parse
'127.0.0.1:5000,keep-alive-idle=60,keep-alive' correctl
uggested by Vladimir)
V1:
https://lore.kernel.org/all/20250303143312.640909-1-jmar...@redhat.com/
Juraj Marcin (7):
io: Fix partial struct copy in qio_dns_resolver_lookup_sync_inet()
util/qemu-sockets: Refactor setting client sockopts into a separate
function
util/qemu-sockets: Refactor
Hi Michael,
On 2025-05-25 20:15, Michael Tokarev wrote:
> On 21.05.2025 16:52, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
> > introduces the keep-alive flag, but this flag is not copied togethe
catch. Thanks!
Reviewed-by: Juraj Marcin
>
> Fixes 316e8ee8d6 ("util/qemu-sockets: Refactor inet_parse() to use QemuOpts")
> Signed-off-by: Matheus Tavares Bernardino
>
> ---
> tests/unit/test-util-sockets.c | 4
> 1 file changed, 4 insertions(+)
>
From: Juraj Marcin
Currently, the inet address parser cannot handle multiple options where
one is prefixed with the name of the other. For example, with the
'keep-alive-idle' option added, the current parser cannot parse
'127.0.0.1:5000,keep-alive-idle=60,keep-alive' correctl
From: Juraj Marcin
With the default TCP stack configuration, it could be even 2 hours
before the connection times out due to the other side not being
reachable. However, in some cases, the application needs to be aware of
a connection issue much sooner.
This is the case, for example, for
From: Juraj Marcin
To get a listening socket, we need to first create a socket, try binding
it to a certain port, and lastly starting listening to it. Each of these
operations can fail due to various reasons, one of them being that the
requested address/port is already in use. In such case, the
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, but this flag is not copied together
with other options in qio_dns_resolver_lookup_sync_inet().
This patch fixes this issue and also prevents future ones by copying the
entire
From: Juraj Marcin
This is done in preparation for enabling the SO_KEEPALIVE support for
server sockets and adding settings for more TCP keep-alive socket
options.
Signed-off-by: Juraj Marcin
Reviewed-by: Daniel P. Berrangé
---
util/qemu-sockets.c | 29 +++--
1 file
From: Juraj Marcin
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
option, but only on client-side sockets. However, this option is also
useful for server-side sockets, so they can check if a client is still
tions setting into a common function for both server
and client sockets (suggested by Vladimir)
V1:
https://lore.kernel.org/all/20250303143312.640909-1-jmar...@redhat.com/
Juraj Marcin (6):
io: Fix partial struct copy in qio_dns_resolver_lookup_sync_inet()
util/qemu-sockets: Refactor
Hi Peter
On 2025-05-21 12:04, Peter Xu wrote:
> On Wed, May 21, 2025 at 05:16:13PM +0200, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > If a virtual machine is paused for an extended period time, for example,
> > due to an incoming migration, there are al
From: Juraj Marcin
If a virtual machine is paused for an extended period time, for example,
due to an incoming migration, there are also no changes on the screen.
VNC in such case increases the display update interval by
VNC_REFRESH_INTERVAL_INC (50 ms). The update interval can then grow up
to
On 2025-05-29 16:42, Juraj Marcin wrote:
> Hi Peter
>
> On 2025-05-27 17:58, Peter Xu wrote:
> > The hook is only defined in two vmstate users ("ram" and "block dirty
> > bitmap"), meanwhile both of them define the hook exactly the same as the
> >
On 2025-05-27 17:58, Peter Xu wrote:
> It's not possible to happen in bg-snapshot case.
>
> Signed-off-by: Peter Xu
> ---
> migration/migration.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
Reviewed-by: Juraj Marcin
>
> No functional change intended.
Could be some future users, that would benefit from separate hooks for
precopy and postcopy?
In case we are going to drop it, I think the '_precopy' suffix could be
dropped too, as the handler would be used for postcopy too.
Best regards
Juraj Marc
On 2025-05-27 17:58, Peter Xu wrote:
> The check over PAGE_DIRTY_FOUND isn't necessary. We could indent one less
> and assert that instead.
>
> Signed-off-by: Peter Xu
> ---
> migration/ram.c | 20 +++-
> 1 file changed, 11 insertions(+), 9 deleti
On 2025-05-27 17:58, Peter Xu wrote:
> Take notes on start/end state of dirty pages for the whole system.
>
> Signed-off-by: Peter Xu
> ---
> migration/ram.c| 5 +
> migration/trace-events | 1 +
> 2 files changed, 6 insertions(+)
Reviewed-by: Juraj Marcin
the
> two completely differnt use cases (precopy v.s. postcopy).
>
> (4) Trivial touch up on threshold_size comparision
>
> Which changes:
>
> "(!pending_size || pending_size < s->threshold_size)"
>
> into:
>
> "(pending_size <= s->threshold_size)"
>
> Signed-off-by: Peter Xu
> ---
> migration/migration.c | 56 +++
> 1 file changed, 41 insertions(+), 15 deletions(-)
Reviewed-by: Juraj Marcin
age Types: normal=367713, zero=41195
> Page Rates (pps): transfer=40900, dirty=4
> Others: dirty_syncs=2, postcopy_req=57503
>
> Cc: Zhijian Li (Fujitsu)
> Suggested-by: Dr. David Alan Gilbert
> Signed-off-by: Peter Xu
> ---
> migration/migration-hmp-cmds.c | 59 ++
> 1 file changed, 31 insertions(+), 28 deletions(-)
Reviewed-by: Juraj Marcin
.
>
> Signed-off-by: Peter Xu
> ---
> docs/devel/migration/postcopy.rst | 36 +++
> 1 file changed, 17 insertions(+), 19 deletions(-)
Reviewed-by: Juraj Marcin
erevalov
> Cc: Markus Armbruster
> Tested-by: Mario Casquero
> Signed-off-by: Peter Xu
> ---
Reviewed-by: Juraj Marcin
save_section_footer(f, se);
> if (ret < 0) {
> --
> 2.49.0
>
There are still mentions of 'save_live_complete_precopy' in:
include/migration/register.h:100
* parallel with @save_live_complete_precopy handlers.
docs/devel/migration/vfio.rst:78
* A ``save_live_complete_precopy`` function that sets the VFIO device in
docs/devel/migration/vfio.rst:198
.save_live_complete_precopy() is called for each active device
docs/devel/migration/vfio.rst:200
.save_live_complete_precopy() until
docs/devel/migration/main.rst:511
- A ``save_live_complete_precopy`` function that must transmit the
Also, should we also drop "live" from
'save_live_complete_precopy_thread' as well? IIUC they are called
together with (now) 'save_complete()' during precopy.
Best regards
Juraj Marcin
On 2025-06-13 10:07, Peter Xu wrote:
> Since we use the same save_complete() hook for both precopy and postcopy,
> add a set of helpers to invoke the hook() to dedup the code.
>
> Signed-off-by: Peter Xu
> ---
Reviewed-by: Juraj Marcin
ctional change intended.
>
> Signed-off-by: Peter Xu
> ---
Reviewed-by: Juraj Marcin
On 2025-06-24 11:41, Peter Xu wrote:
> On Tue, Jun 24, 2025 at 04:36:39PM +0200, Juraj Marcin wrote:
>
> [...]
>
> > There are still mentions of 'save_live_complete_precopy' in:
> >
> > include/migration/register.h:100
> > * para
From: Juraj Marcin
Recent patch [1] renames the save_live_complete_precopy handler to
save_complete, as the machine is not live in most cases when this
handler is executed. The same is true also for
save_live_complete_precopy_thread, therefore this patch removes the
"live" keywor
On 2025-07-10 13:17, Peter Maydell wrote:
> On Thu, 22 May 2025 at 11:33, Daniel P. Berrangé wrote:
> >
> > From: Juraj Marcin
> >
> > To get a listening socket, we need to first create a socket, try binding
> > it to a certain port, and lastly starti
Hi Daniel,
On 2025-08-01 18:02, Daniel P. Berrangé wrote:
> This is a followup to previously merged patches that claimed to
> workaround the gnutls bug impacting migration, but in fact were
> essentially non-functional. Juraj Marcin pointed this out, and
> this new patch tweaks the w
Hi all!
On 2025-08-05 10:44, Fabiano Rosas wrote:
> Daniel P. Berrangé writes:
>
> > On Mon, Aug 04, 2025 at 04:27:45PM -0300, Fabiano Rosas wrote:
> >> Juraj Marcin writes:
> >>
> >> > Hi Daniel,
> >> >
> >> > On 2025-08
py-run command
is actually sent (in this series boundary of "postcopy-setup" and
"postcopy-active"), however, I am not sure if such change would not have
any unwanted implications.
Juraj Marcin (4):
qemu-thread: Introduce qemu_thread_detach()
migration: Fix state transition in pos
From: Juraj Marcin
During switchover at the start of postcopy the source side sends a
package containing all non-postcopiable device state and a postcopy-run
command and transitions to a "postcopy-active" state. However, if the
destination side fails to load the device states or cras
From: Juraj Marcin
Currently, QEMU threads abstraction supports both joinable and detached
threads, but once a thread is marked as joinable it must be joined using
qemu_thread_join() and cannot be detached later.
For POSIX implementation, pthread_detach() is used. For Windows, marking
the
From: Juraj Marcin
Depending on where an error during postcopy_start() happens, the state
can be either "active", "device" or "cancelling", but never
"postcopy-active". Migration state is transitioned to "postcopy-active"
only just before a succ
From: Juraj Marcin
This patch allows joining the migration listen thread. This is done in
preparation for the introduction of "postcopy-setup" state at the
beginning of a postcopy migration, when destination can fail gracefully
and source side then resume to a running state.
In ca
Hi Peter,
On 2025-08-07 16:57, Peter Xu wrote:
> On Thu, Aug 07, 2025 at 01:49:11PM +0200, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > This patch allows joining the migration listen thread. This is done in
> > preparation for the introduction of &q
Hi Peter,
On 2025-08-07 16:54, Peter Xu wrote:
> On Thu, Aug 07, 2025 at 01:49:10PM +0200, Juraj Marcin wrote:
> > From: Juraj Marcin
> >
> > Depending on where an error during postcopy_start() happens, the state
> > can be either "active", "device
96 matches
Mail list logo
