On 2025/07/18 17:52, Paolo Abeni wrote:
Tap devices support GSO over UDP tunnel offload. Probe for such
feature in a similar manner to other offloads.
GSO over UDP tunnel needs to be enabled in addition to a "plain"
offload (TSO or USO).
No need to check separately for the outer header checksum
On 2025/07/21 22:29, Daniel P. Berrangé wrote:
On Mon, Jul 21, 2025 at 10:14:30PM +0900, Akihiko Odaki wrote:
On 2025/07/21 20:29, Arun Menon wrote:
- We need to have good error reporting in the callbacks in
VMStateDescription struct. Specifically pre_save, post_save,
pre_load and
On 2025/07/21 20:29, Arun Menon wrote:
- We need to have good error reporting in the callbacks in
VMStateDescription struct. Specifically pre_save, post_save,
pre_load and post_load callbacks.
- It is not possible to change these functions everywhere in one
patch, therefore, we introduce
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that qemu_loadvm_state() must report an error
in errp, in case of failure.
Signed-off-by: Arun
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
Reviewed-by: Daniel P. Berrangé
Signed-off-by: Arun Menon
---
migration/savevm.c | 16 +---
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that loadvm_postcopy_handle_advise() must report an error
in errp, in case of failure.
Reviewe
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that loadvm_process_command() must report an error
in errp, in case of failure.
Signed-off-by:
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that vmstate_load_state() must report an error
in errp, in case of failure.
Signed-off-by: Aru
On 2025/07/21 20:29, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that vmstate_subsection_load() must report an error
in errp, in case of failure.
Signed-off-by
On 2025/07/21 16:51, Paolo Abeni wrote:
7/20/25 12:44 PM, Akihiko Odaki wrote:
On 2025/07/18 17:52, Paolo Abeni wrote:
@@ -,26 +3362,23 @@ virtio_load(VirtIODevice *vdev, QEMUFile *f, int
version_id)
vdev->device_endian = virtio_default_endian();
}
-
On 2025/07/21 16:33, Paolo Abeni wrote:
On 7/20/25 12:41 PM, Akihiko Odaki wrote:
On 2025/07/18 17:52, Paolo Abeni wrote:
diff --git a/include/hw/virtio/virtio-features.h
b/include/hw/virtio/virtio-features.h
new file mode 100644
index 00..68e326e3e8
--- /dev/null
+++ b/include/hw
On 2025/07/18 17:52, Paolo Abeni wrote:
Extend the VirtioDeviceFeatures struct with an additional u64
to track unknown features in the 64-127 bit range and decode
the full virtio features spaces for vhost and virtio devices.
Also add entries for the soon-to-be-supported virtio net GSO over
UDP f
On 2025/07/18 17:52, Paolo Abeni wrote:
If the driver uses any of the extended features (i.e. 64 or above),
store the extended features range (64-127 bits).
At load time, let legacy features initialize the full features range
and pass it to the set helper; sub-states loading will have filled-up
On 2025/07/18 17:52, Paolo Abeni wrote:
The virtio specifications allows for up to 128 bits for the
device features. Soon we are going to use some of the 'extended'
bits features (bit 64 and above) for the virtio net driver.
Represent the virtio features bitmask with a fixes size array, and
intr
On 2025/07/18 17:52, Paolo Abeni wrote:
The set_offload() argument list is already pretty long and
we are going to introduce soon a bunch of additional offloads.
Replace the offload arguments with a single struct and update
all the relevant call-sites.
No functional changes intended.
Signed-of
On 2025/07/16 0:49, Paolo Abeni wrote:
On 7/15/25 10:05 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
diff --git a/net/tap-bsd.c b/net/tap-bsd.c
index 86b6edee94..e7de0672f4 100644
--- a/net/tap-bsd.c
+++ b/net/tap-bsd.c
@@ -217,6 +217,11 @@ int tap_probe_has_uso(int fd
On 2025/07/18 2:02, Daniel P. Berrangé wrote:
On Thu, Jul 17, 2025 at 12:34:21PM +0900, Akihiko Odaki wrote:
On 2025/07/17 9:37, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor
On 2025/07/17 9:37, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that ram_postcopy_incoming_init() must report an error
in errp, in case of failure.
Signed-off-
On 2025/07/17 9:37, Arun Menon wrote:
This is an incremental step in converting vmstate loading
code to report error via Error objects instead of directly
printing it to console/monitor.
It is ensured that loadvm_process_enable_colo() must report an error
in errp, in case of failure.
Signed-off-
On 2025/07/16 19:13, Paolo Abeni wrote:
On 7/15/25 10:07 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
When any host or guest GSO over UDP tunnel offload is enabled the
virtio net header includes the additional tunnel-related fields,
update the size accordingly.
Push the
On 2025/07/16 0:43, Paolo Abeni wrote:
On 7/15/25 9:59 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
@@ -785,11 +821,12 @@ VirtioStatus *qmp_x_query_virtio_status(const char *path,
Error **errp)
status->vhost_dev->nvqs = hdev->nvqs;
status-&
On 2025/07/16 18:14, Paolo Abeni wrote:
On 7/15/25 6:21 PM, Paolo Abeni wrote:
On 7/15/25 9:42 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
@@ -158,7 +159,10 @@ struct VirtIOPCIProxy {
uint32_t nvectors;
uint32_t dfselect;
uint32_t gfselect
On 2025/07/16 0:40, Paolo Abeni wrote:
On 7/15/25 9:24 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
+ */
+QEMU_BUILD_BUG_ON(VIRTIO_FEATURES_DWORDS != 2);
+if (virtio_128bit_features_needed(vdev)) {
There is no need to distinguish virtio_128bit_features_needed
On 2025/07/15 23:52, Paolo Abeni wrote:
On 7/15/25 8:36 AM, Akihiko Odaki wrote:
On 2025/07/11 22:02, Paolo Abeni wrote:
The set_offload() argument list is already pretty long and
we are going to introduce soon a bunch of additional offloads.
Replace the offload arguments with a single struct
On 2025/07/11 22:02, Paolo Abeni wrote:
When any host or guest GSO over UDP tunnel offload is enabled the
virtio net header includes the additional tunnel-related fields,
update the size accordingly.
Push the GSO over UDP tunnel offloads all the way down to the tap
device extending the newly int
On 2025/07/11 22:02, Paolo Abeni wrote:
Extend the VirtioDeviceFeatures struct with an additional u64
to track unknown features in the 65-128 bit range and decode
the full virtio features spaces for vhost and virtio devices.
Signed-off-by: Paolo Abeni
---
I'm unsure if it's actually legit to up
On 2025/07/11 22:02, Paolo Abeni wrote:
Tap devices support GSO over UDP tunnel offload. Probe for such
feature in a similar manner to other offloads.
GSO over UDP tunnel needs to be enabled in addition to a "plain"
offload (TSO or USO).
No need to check separately for the outer header checksu
On 2025/07/11 22:02, Paolo Abeni wrote:
Extend the features configuration space to 128 bits, and allow the
common read/write operation to access all of it.
On migration, save the 128 bit version of the features only if the
upper bits are non zero; after load zero the upper bits if the extended
f
On 2025/07/11 22:02, Paolo Abeni wrote:
If the driver uses any of the extended features (i.e. above 64),
serialize the full features range (128 bits).
This is one of the few spots that need explicitly to know and set
in stone the extended features array size; add a build bug to prevent
breaking
On 2025/07/11 22:02, Paolo Abeni wrote:
The virtio specifications allows for up to 128 bits for the
device features. Soon we are going to use some of the 'extended'
bits features (above 64) for the virtio net driver.
Represent the virtio features bitmask with a fixes size array, and
introduce a
On 2025/07/11 22:02, Paolo Abeni wrote:
The set_offload() argument list is already pretty long and
we are going to introduce soon a bunch of additional offloads.
Replace the offload arguments with a single struct and update
all the relevant call-sites.
No functional changes intended.
Signed-of
pen at all because we always install
a placeholder surface to the console when there is nothing to display.
Resolves: Coverity CID 1610328
Signed-off-by: Weifeng Liu
Reviewed-by: Akihiko Odaki
do not reset vlan filtering at set_features")
Cc: qemu-sta...@nongnu.org
Reported-by: Konstantin Shkolnyy
Signed-off-by: Akihiko Odaki
---
Not tested.
Konstantin, I would also want you to test this new version. Please also
give it Tested-by, and, if possible, Reviewed-by.
---
Changes in v2:
feature negotiation sequence.
Fixes: 06b636a1e2ad ("virtio-net: do not reset vlan filtering at set_features")
Reported-by: Konstantin Shkolnyy
Signed-off-by: Akihiko Odaki
---
Not tested.
Konstantin, please see if this patch fixes your workload.
---
hw/net/virtio-net.c | 11 ++-
d-by: Corentin BAYET
Signed-off-by: Akihiko Odaki
---
hw/pci/pcie_sriov.c | 42 +++---
1 file changed, 23 insertions(+), 19 deletions(-)
diff --git a/hw/pci/pcie_sriov.c b/hw/pci/pcie_sriov.c
index
3ad18744f4a8ed2b35144fafcdc8e7e00fec3672..a08b5258275f51876c18
ht. Can you submit a patch to fix it?
Regards,
Akihiko Odaki
On 2025/05/23 18:43, Paolo Abeni wrote:
On 5/23/25 9:19 AM, Akihiko Odaki wrote:
On 2025/05/21 20:33, Paolo Abeni wrote:
Some virtualized deployments use UDP tunnel pervasively and are impacted
negatively by the lack of GSO support for such kind of traffic in the
virtual NIC driver.
The
a: Restore vlan filtering state")
Signed-off-by: Konstantin Shkolnyy
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Reviewed-by: Akihiko Odaki
On 2025/06/21 4:47, Yiwei Zhang wrote:
On Thu, Jun 19, 2025 at 11:45 PM Alex Bennée wrote:
Yiwei Zhang writes:
On Sun, Jun 8, 2025 at 1:24 AM Akihiko Odaki
wrote:
On 2025/06/06 1:26, Alex Bennée wrote:
From: Yiwei Zhang
Venus and later native contexts have their own fence context
On 2025/06/13 5:54, John Snow wrote:
\{ is an illegal escape sequence, to get a literal backslash we need to
use \\.
This is being fixed because of an issue spotted when attempting to use
the pyupgrade script.
Signed-off-by: John Snow
Reviewed-by: Akihiko Odaki
easily.)
Signed-off-by: John Snow
Reviewed-by: Akihiko Odaki
erent repositories.
Signed-off-by: John Snow
Reviewed-by: Akihiko Odaki
On 2025/06/13 5:54, John Snow wrote:
The next patch will synchronize the qemu.qmp library with the external,
standalone version. That synchronization will require an extra ignore
for pylint, so do that now.
Signed-off-by: John Snow
Reviewed-by: Akihiko Odaki
On 2025/06/14 1:53, Sean Wei wrote:
The LGPLv2.1 boiler-plate in pdb.c file still contained
the obsolete "51 Franklin Street" postal address.
Replace it with the canonical GNU licenses URL recommended by the FSF:
https://www.gnu.org/licenses/
Signed-off-by: Sean Wei
Reviewed-b
ny time. I figured now was a good time as any to
get rid of them before they become a problem randomly some day in the
future.
I guess you meant that the old style type hints are deprecated.
Regards,
Akihiko Odaki
n't use Linux kernel internal types like u32, __u32 or __le32.
It's unfortunate that QEMU lacks endian types and a checker for them;
such a checker could not have caught this particular case, but can catch
other similar bugs.
Regards,
Akihiko Odaki
const struct iovec data
rrangement where we are allowed to use versions of
Meson that otherwise would break our build platform guarantee.)
Lately, we've had some issues with the wide range of Sphinx versions we
support presenting various cross-platform difficulties. In particular,
Akihiko Odaki has sent patches to
On 2025/06/07 0:02, Akihiko Odaki wrote:
On 2025/06/06 19:16, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/05 20:57, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/03 20:01, Alex Bennée wrote:
QOM objects can be embedded in other QOM objects and managed as part
of their
ec76ce0-c3ca-48ed-befe-e0930d4a3...@linaro.org/
+assert(is_power_of_2(size) && size >= 1 && size <= (1 << MO_SIZE));
size >= 1 is unnecessary because being power of 2 implies that. The
comment doesn't mention this condition either.
Regards,
Akihiko Odaki
#endif
return (MemOp)ctz32(size);
}
7] ui/gtk-gl-area: Remove extra draw call in
refresh" requires Cc: qemu-sta...@nongnu.org. Fixing -display gtk,gl=on
for Wayland sounds as important as this patch.
Regards,
Akihiko Odaki
Signed-off-by: Yiwei Zhang
Reviewed-by: Dmitry Osipenko
Message-Id: <20250518152651.334115-1-zzyi..
> subregion
(c) FlatView -> subregion
Therefore, the subregion will be freed with (c). So the unmapping
operation will never finish until the BQL gets unlocked.
Regards,
Akihiko Odaki
Signed-off-by: Manos Pitsidianakis
Reviewed-by: Philippe Mathieu-Daudé
Message-Id: &l
no extra changes are needed for the tests to continue
passing.
Signed-off-by: John Snow
Reviewed-by: Akihiko Odaki
I didn't notice these files. Perhaps they are better to be documented
with pythondeps.toml.
Regards,
Akihiko Odaki
On 2025/06/07 5:38, Paolo Bonzini wrote:
On Fri, Jun 6, 2025 at 11:46 AM Akihiko Odaki wrote:
This conditional is unnecessary; docs/about/build-platforms.rst says we
only supports MinGW.
I failed to catch this problem because I ran MinGW on Windows, which is
case-insensitive. Since it is case
On 2025/06/06 20:31, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/06 18:54, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/06 1:26, Alex Bennée wrote:
QOM objects can be embedded in other QOM objects and managed as part
of their lifetime but this isn't the cas
On 2025/06/06 19:16, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/05 20:57, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/03 20:01, Alex Bennée wrote:
QOM objects can be embedded in other QOM objects and managed as part
of their lifetime but this isn't the cas
On 2025/06/06 18:54, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/06 1:26, Alex Bennée wrote:
QOM objects can be embedded in other QOM objects and managed as part
of their lifetime but this isn't the case for
virtio_gpu_virgl_hostmem_region. However before we can split it out we
On 2025/06/05 17:51, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/03 20:01, Alex Bennée wrote:
Currently the boot.S code assumes everything starts at EL1. This will
break things like the memory test which will barf on unaligned memory
access when run at a higher level.
Adapt the boot
On 2025/06/05 22:24, Paolo Bonzini wrote:
From: Akihiko Odaki
In a recent discussion, Phil Dennis-Jordan pointed out a quirk in
QemuEvent destruction due to futex-like abstraction, which prevented
the usage of QemuEvent in new and existing code[1]. With some more
thoughts after this
On 2025/06/05 20:57, Alex Bennée wrote:
Akihiko Odaki writes:
On 2025/06/03 20:01, Alex Bennée wrote:
QOM objects can be embedded in other QOM objects and managed as part
of their lifetime but this isn't the case for
virtio_gpu_virgl_hostmem_region. However before we can split it out we
od to me, so:
Reviewed-by: Akihiko Odaki
Regards,
Akihiko Odaki
diff --git a/tests/qtest/libqos/igb.c b/tests/qtest/libqos/igb.c
index f40c4ec4cd..ab3ef6f0c3 100644
--- a/tests/qtest/libqos/igb.c
+++ b/tests/qtest/libqos/igb.c
@@ -104,10 +104,10 @@ static void igb_pci_start_
call
in refresh" has an undressed comment[2][7].
I would like to see improvements in how comments are addressed before a
series is resent.
Regards,
Akihiko Odaki
[1]
https://lore.kernel.org/qemu-devel/e6af12bd-1c36-4e50-8bae-d8d80cad1...@daynix.com
[2]
https://lore.kernel.org/qemu-deve
his patch can
be simply dropped.
Regards,
Akihiko Odaki
-mr = &vmr->mr;
memory_region_init_ram_ptr(mr, OBJECT(mr), "blob", size, data);
memory_region_add_subregion(&b->hostmem, offset, mr);
memory_region_set_enabled(mr, true);
@@ -131,7 +
On 2025/06/03 20:01, Alex Bennée wrote:
Thanks for volunteering to help.
Cc: Akihiko Odaki
Cc: Dmitry Osipenko
Reviewed-by: Markus Armbruster
Reviewed-by: Philippe Mathieu-Daudé
Signed-off-by: Alex Bennée
Acked-by: Michael S. Tsirkin
---
MAINTAINERS | 2 ++
1 file changed, 2 insertions
CH v4 11/17] ui/gtk-gl-area: Remove extra draw call
in refresh".
Regards,
Akihiko Odaki
+ .quad cmdline
+ .quad 128
.text
.align 4
.global __start
__start:
+ /*
+ * Initialise the stack for whatever EL we are in before
+* anything e
state to fix the memory leak.
Signed-off-by: Akihiko Odaki
---
Changes in v3:
- Rebased.
- Reordered the patches.
- Removed the extra allocation of VncWorker.
- Removed the worker field from VncState.
- Dropped the Reviewed-by tags.
- Link to v2:
https://lore.kernel.org/qemu-devel/20250531-zlib-v2
copying.
It allows allocating and freeing all shared data at once and shows that
the race with the worker thread needs to be taken care of when
accessing them.
Signed-off-by: Akihiko Odaki
---
ui/vnc.h | 49 --
ui/vnc-enc-tight.c| 456
state to fix the memory leak.
Fixes: bd023f953e5e ("vnc: threaded VNC server")
Signed-off-by: Akihiko Odaki
---
ui/vnc.h | 2 +-
ui/vnc-enc-zlib.c | 30 +++---
ui/vnc.c | 13 ++---
3 files changed, 26 insertions(+), 19 deletions(-)
di
copying.
It allows allocating and freeing all shared data at once and shows that
the race with the worker thread needs to be taken care of when
accessing them.
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
ui/vnc.h | 12 +-
ui/vnc-enc-tight.c| 341
state to fix the memory leak.
Fixes: bd023f953e5e ("vnc: threaded VNC server")
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
ui/vnc.h | 2 +-
ui/vnc-enc-zlib.c | 30 +++---
ui/vnc-jobs.c | 2 --
ui/vnc.c | 2 +
state to fix the memory leak.
Signed-off-by: Akihiko Odaki
---
Changes in v2:
- Rebased.
- Link to v1:
https://lore.kernel.org/qemu-devel/20250417-zlib-v1-0-34fad73b8...@daynix.com
---
Akihiko Odaki (2):
ui/vnc: Introduce the VncWorker type
ui/vnc: Do not copy z_stream
ui/vnc.h
d
when the user sets the property to true while it is defined as an
an error for the "host" CPU when the host doesn't have a PMU.
This fixes qtest-aarch64/arm-cpu-features on the hosts that supports
KVM but doesn't support PMU emulation.
Signed-off-by: Akihiko Odaki
---
kvm-stea
exposes all these four registers
to GDB.
It is not necessary to add ARM_CP_ALIAS to these registers because the
flag is already set.
Signed-off-by: Akihiko Odaki
---
Supersedes: <20250317-raw-v1-0-09e2dfff0...@daynix.com>
("[PATCH 0/4] target/arm: Flag PMCNTENCLR with ARM_CP_NO_RAW"
ps up.
Signed-off-by: Akihiko Odaki
---
My email address at Daynix will remain active at least for the next
month.
I added the submitter of the patches I recently reviewed to Cc. Please
Cc my new email address for future reviews.
I have a number of patches that are not pulled yet. Colleagues at D
Report hashing capability so that virtio-net can deliver the correct
capability information to the guest.
Signed-off-by: Akihiko Odaki
---
include/net/net.h | 3 +++
net/net.c | 9 +
net/vhost-vdpa.c | 28
3 files changed, 40 insertions(+)
diff
DEFINE_PROP_ON_OFF_AUTO_BIT64() corresponds to DEFINE_PROP_ON_OFF_AUTO()
as DEFINE_PROP_BIT64() corresponds to DEFINE_PROP_BOOL(). The difference
is that DEFINE_PROP_ON_OFF_AUTO_BIT64() exposes OnOffAuto instead of
bool.
Signed-off-by: Akihiko Odaki
---
include/hw/qdev-properties.h | 18
ing")
Signed-off-by: Akihiko Odaki
---
Changes in v2:
- Updated a documentation comment of pre_load_queues() for clarity.
- Link to v1:
https://lore.kernel.org/qemu-devel/20250510-n-v1-1-19ee26ac3...@daynix.com
---
include/hw/virtio/virtio.h | 10 --
hw/net/virtio-net.c| 1
It is no longer used.
Signed-off-by: Akihiko Odaki
---
net/vhost-vdpa.c | 8
1 file changed, 8 deletions(-)
diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
index 4f63ded40d47..149c0f7f1766 100644
--- a/net/vhost-vdpa.c
+++ b/net/vhost-vdpa.c
@@ -238,12 +238,6 @@ static void
This allows offloading hash reporting and RSS to tap.
Signed-off-by: Akihiko Odaki
---
hw/net/virtio-net.c | 69 +++--
1 file changed, 56 insertions(+), 13 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 52fe404b3431
eBPF RSS virtio-net support was written in assumption that there is only
one alternative RSS implementation: 'in-qemu' RSS. It is no longer true,
and we now have yet another implementation; namely the peer RSS.
Signed-off-by: Akihiko Odaki
---
docs/devel/ebpf_rs
v4:
- Rebased.
- Added a reference to the documentation to the cover letter.
- Link to v3:
https://lore.kernel.org/r/20240915-hash-v3-0-79cb08d28...@daynix.com
---
Akihiko Odaki (5):
net: Allow configuring virtio hashing
virtio-net: Offload hashing to peer
virtio-net: Offload hashin
This allows offloading virtio-net hashing to tap on Linux.
Signed-off-by: Akihiko Odaki
---
net/tap-linux.h | 1 +
net/tap_int.h | 1 +
net/tap-bsd.c | 5 +
net/tap-linux.c | 5 +
net/tap-solaris.c | 5 +
net/tap-stub.c| 5 +
net/tap.c | 8
7
This is necessary to offload hashing to tap.
Signed-off-by: Akihiko Odaki
---
hw/net/virtio-net.c | 15 +--
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 0a333d560d7b..3469c211b13a 100644
--- a/hw/net/virtio-net.c
+++ b
This adds functions to configure virtio hashing and implements it
for Linux's tap. vDPA will have empty functions as configuring virtio
hashing is done with the load().
Signed-off-by: Akihiko Odaki
---
include/net/net.h | 13 +
net/tap-linux.h | 3 +++
net/tap_int.h
ot;on" for the corresponding properties.
Signed-off-by: Akihiko Odaki
---
include/hw/virtio/virtio-net.h | 1 +
hw/net/virtio-net.c| 45 --
2 files changed, 44 insertions(+), 2 deletions(-)
diff --git a/include/hw/virtio/virtio-net.h b/
Retrieve peer hashing capability instead of hardcoding.
Signed-off-by: Akihiko Odaki
---
include/hw/virtio/virtio-net.h | 5 ++-
hw/net/virtio-net.c| 71 ++
net/vhost-vdpa.c | 4 +--
3 files changed, 64 insertions(+), 16
Move virtio_net_get_features() to the later part of the file so that
it can call other functions.
Signed-off-by: Akihiko Odaki
---
hw/net/virtio-net.c | 146 ++--
1 file changed, 73 insertions(+), 73 deletions(-)
diff --git a/hw/net/virtio-net.c
ypes. This also helps add another
hashing mechanism proposed by the patch series mentioned earlier.
Signed-off-by: Akihiko Odaki
---
Akihiko Odaki (6):
qdev-properties: Add DEFINE_PROP_ON_OFF_AUTO_BIT64()
net/vhost-vdpa: Report hashing capability
virtio-net: Move virtio_net_ge
the alignment of holes.
So stop asserting the assumption on the holes and handle unaligned holes
properly.
Signed-off-by: Akihiko Odaki
---
Changes in v2:
- Changed to round the number also when the specified offset in a hole.
- Changed to iterate until finding an aligned location.
- Link to v1:
intended to satisfy.
Signed-off-by: Akihiko Odaki
---
include/qemu/thread.h | 19
util/event.c | 84 ---
2 files changed, 66 insertions(+), 37 deletions(-)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index
pause_event can utilize qemu_event_reset() to discard events.
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
migration/migration.h | 2 +-
migration/migration.c | 21 +
2 files changed, 10 insertions(+), 13 deletions(-)
diff --git a/migration
atomic
operations to ensure ordering between them.
Signed-off-by: Akihiko Odaki
Tested-by: Phil Dennis-Jordan
Reviewed-by: Phil Dennis-Jordan
---
util/qemu-thread-posix.c | 50 +++-
1 file changed, 24 insertions(+), 26 deletions(-)
diff --git a/util
pair of it and qatomic_read()
is cheaper than qatomic_xchg(). Therefore remove the pair and simplify
the code instead.
Signed-off-by: Akihiko Odaki
---
util/event.c | 19 +--
1 file changed, 5 insertions(+), 14 deletions(-)
diff --git a/util/event.c b/util/event.c
index e937804a92a8..df6d6
Use the futex-based implementation of QemuEvent on Windows to
remove code duplication and remove the overhead of event object
construction and destruction.
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
include/qemu/thread-posix.h | 9 ---
include/qemu/thread-win32.h
thread_sync_sem is an one-shot event so it can be converted into
QemuEvent, which is more lightweight.
Signed-off-by: Akihiko Odaki
Reviewed-by: Fabiano Rosas
---
migration/migration.h| 4 ++--
migration/postcopy-ram.c | 10 +-
migration/savevm.c | 2 +-
3 files changed, 8
scripts/checkpatch.pl warns for __linux__ saying "architecture specific
defines should be avoided".
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
include/qemu/thread-posix.h | 2 +-
util/qemu-thread-posix.c| 6 +++---
2 files changed, 4 insertions(+), 4
Document QemuEvent to help choose an appropriate synchronization
primitive.
Signed-off-by: Akihiko Odaki
---
include/qemu/thread.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/include/qemu/thread.h b/include/qemu/thread.h
index 573f8c9ede20..f0302ed01fdb 100644
--- a/include
Windows supports futex-like APIs since Windows 8 and Windows Server
2012.
Signed-off-by: Akihiko Odaki
---
meson.build | 2 ++
include/qemu/futex.h | 53 ++-
tests/unit/test-aio-multithread.c | 2 +-
util/lockcnt.c
colo_exit_sem and colo_incoming_sem represent one-shot events so they
can be converted into QemuEvent, which is more lightweight.
Signed-off-by: Akihiko Odaki
Reviewed-by: Fabiano Rosas
Reviewed-by: Philippe Mathieu-Daudé
---
migration/migration.h | 6 +++---
migration/colo.c | 20
sem in AppleGFXReadMemoryJob is an one-shot event so it can be converted
into QemuEvent, which is more specialized for such a use case.
Signed-off-by: Akihiko Odaki
Reviewed-by: Philippe Mathieu-Daudé
---
hw/display/apple-gfx.m | 10 +-
1 file changed, 5 insertions(+), 5 deletions
inue
> to block or not.
Signed-off-by: Akihiko Odaki
---
include/qemu/futex.h | 9 +
tests/unit/test-aio-multithread.c | 4 +++-
util/qemu-thread-posix.c | 35 +--
3 files changed, 25 insertions(+), 23 deletions(-)
diff --git a/i
1 - 100 of 1952 matches
Mail list logo
