On 2/11/25 22:43, Kevin Wolf wrote:
+/// Use QEMU's event loops to run a Rust [`Future`] to completion and return
its result.
+///
+/// This function must be called in coroutine context. If the future isn't
ready yet, it yields.
+pub fn qemu_co_run_future(future: F) -> F::Output {
+let wake
On Tue, Feb 11, 2025 at 10:43:27PM +0100, Kevin Wolf wrote:
> This adds a separate block driver for the bochs image format called
> 'bochs-rs' so that for the moment both the C implementation and the Rust
> implementation can be present in the same build. The intention is to
> remove the C implemen
John Snow writes:
> This is being done primarily to ensure consistency between the source
> documents and the final, rendered HTML output. Because
> member/feature/returns sections will always appear in a visually grouped
> element in the HTML output, prohibiting free paragraphs between those
> s
On 12/2/25 09:27, Thomas Huth wrote:
On 10/02/2025 21.41, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the p
John Snow writes:
> This clarifies sections that are mistaken by the parser as "intro"
> sections to be "details" sections instead.
Impact on output? See notes inline.
>
> Signed-off-by: John Snow
> ---
> qapi/machine.json | 2 ++
> qapi/migration.json| 4
> qapi/qom.json
This is an initial, minimal part of the chardev bindings. It is a
common convention in QEMU to return a positive value in case of success,
and a negated errno value in case of error. Unfortunately, using errno
portably in Rust is a bit complicated; on Unix the errno values are
supported natively
On Wed, Feb 12, 2025 at 09:14:57AM +, Daniel P. Berrangé wrote:
> On Tue, Feb 11, 2025 at 10:43:27PM +0100, Kevin Wolf wrote:
> > This adds a separate block driver for the bochs image format called
> > 'bochs-rs' so that for the moment both the C implementation and the Rust
> > implementation c
On 1/30/25 11:08, Maciej S. Szmigiero wrote:
From: "Maciej S. Szmigiero"
Add support for VFIOMultifd data structure that will contain most of the
receive-side data together with its init/cleanup methods.
Signed-off-by: Maciej S. Szmigiero
---
hw/vfio/migration.c | 52 +
This device is only used by the x86 targets, which are only
built as little-endian. Therefore the DEVICE_NATIVE_ENDIAN
definition expand to DEVICE_LITTLE_ENDIAN (besides, the
DEVICE_BIG_ENDIAN case isn't tested). Simplify directly
using DEVICE_LITTLE_ENDIAN.
Signed-off-by: Philippe Mathieu-Daudé
This device is only used by the ARM targets, which are only
built as little-endian. Therefore the DEVICE_NATIVE_ENDIAN
definition expand to DEVICE_LITTLE_ENDIAN (besides, the
DEVICE_BIG_ENDIAN case isn't tested). Simplify directly
using DEVICE_LITTLE_ENDIAN.
Signed-off-by: Philippe Mathieu-Daudé
When devices are only built for little-endian targets,
DEVICE_NATIVE_ENDIAN expand to DEVICE_LITTLE_ENDIAN.
Replace by the latter to simplify.
Philippe Mathieu-Daudé (5):
hw/arm: Mark Allwinner Technology devices as little-endian
hw/mips: Mark Boston machine devices as little-endian
hw/mips:
The Loonson3 Virt machine is only built as little-endian.
Therefore the DEVICE_NATIVE_ENDIAN definition expand to
DEVICE_LITTLE_ENDIAN (besides, the DEVICE_BIG_ENDIAN case
isn't tested). Simplify directly using DEVICE_LITTLE_ENDIAN.
Signed-off-by: Philippe Mathieu-Daudé
---
hw/mips/loongson3_vir
The Boston machine is only built as little-endian.
Therefore the DEVICE_NATIVE_ENDIAN definition expand to
DEVICE_LITTLE_ENDIAN (besides, the DEVICE_BIG_ENDIAN case
isn't tested). Simplify directly using DEVICE_LITTLE_ENDIAN.
Signed-off-by: Philippe Mathieu-Daudé
---
hw/mips/boston.c | 6 +++---
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Have the MicroblazeMachine class being common to both
MicroblazeBigEndianMachine and MicroblazeLittleEndianMachine
classes. Move the xmaton and ballerina tests to the parent class.
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Thomas Huth
These devices are only used by the ARM targets, which are
only built as little-endian. Therefore the DEVICE_NATIVE_ENDIAN
definition expand to DEVICE_LITTLE_ENDIAN (besides, the
DEVICE_BIG_ENDIAN case isn't tested). Simplify directly using
DEVICE_LITTLE_ENDIAN.
Signed-off-by: Philippe Mathieu-Daud
On 12/2/25 12:37, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (default).
Signed-off-by: Philippe Mathieu-Daudé
---
qapi/common.json | 1
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine usin
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Avoid fetching assets from www.qemu-advent-calendar.org
website, prefer fetching microblaze assets from GitLab servers.
Suggested-by: Thomas Huth
Signed-off-by: Philippe Mathieu-Daudé
---
tests/functional/test_microblazeel_s3adsp1800.py | 3
On 12/2/25 12:43, Philippe Mathieu-Daudé wrote:
On 12/2/25 12:37, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (default).
Signed-off-by: Philippe Mathieu-Dau
On 12/2/25 12:46, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Have the MicroblazeMachine class being common to both
MicroblazeBigEndianMachine and MicroblazeLittleEndianMachine
classes. Move the xmaton and ballerina tests to the parent class.
Signed-off-by: Philippe Ma
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (default).
Signed-off-by: Philippe Mathieu-Daudé
---
qapi/common.json| 16
include/hw/qdev-p
On 12/2/25 11:24, Thomas Huth wrote:
On 12/02/2025 10.19, Philippe Mathieu-Daudé wrote:
On 12/2/25 09:27, Thomas Huth wrote:
On 10/02/2025 21.41, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "littl
On 12/2/25 12:49, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Avoid fetching assets from www.qemu-advent-calendar.org
website, prefer fetching microblaze assets from GitLab servers.
Suggested-by: Thomas Huth
Signed-off-by: Philippe Mathieu-Daudé
---
tests/functiona
On 12/2/25 10:10, Daniel P. Berrangé wrote:
On Tue, Feb 11, 2025 at 07:53:32PM +0100, Philippe Mathieu-Daudé wrote:
On 11/2/25 19:48, Philippe Mathieu-Daudé wrote:
On 11/2/25 19:26, Stefan Hajnoczi wrote:
On Mon, Feb 10, 2025 at 3:43 PM Philippe Mathieu-Daudé
wrote:
The following changes si
On Wed, Feb 12, 2025 at 12:30:20PM +0100, Alexander Graf wrote:
>
> On 12.02.25 11:24, Gerd Hoffmann wrote:
> >
> > Why do you use confidential computing in the first place if you trust
> > the host with your EFI variables? I'd rather see something simliar
> > running under guest control, in svs
On Wed, Feb 12, 2025 at 01:02:18PM +0100, Philippe Mathieu-Daudé wrote:
> On 12/2/25 12:43, Philippe Mathieu-Daudé wrote:
> > On 12/2/25 12:37, Thomas Huth wrote:
> > > On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
> > > > Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
> >
CTR entries are accessed using ctrsource, ctrtarget and ctrdata
registers using smcsrind/sscsrind extension. This commits extends
the csrind extension to support CTR registers.
ctrsource is accessible through xireg CSR, ctrtarget is accessible
through xireg1 and ctrdata is accessible through xireg
On 12/02/2025 10.19, Philippe Mathieu-Daudé wrote:
On 12/2/25 09:27, Thomas Huth wrote:
On 10/02/2025 21.41, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the devic
On 12/2/25 10:19, Philippe Mathieu-Daudé wrote:
On 12/2/25 09:27, Thomas Huth wrote:
On 10/02/2025 21.41, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
e
RBD schedules the request completion code (qemu_rbd_finish_bh()) to run
in the BDS's AioContext. The intent seems to be to run it in the same
context that the original request coroutine ran in, i.e. the thread on
whose stack the RBDTask object exists (see qemu_rbd_start_co()).
However, with multi
On Tue, Feb 11, 2025 at 07:53:32PM +0100, Philippe Mathieu-Daudé wrote:
> On 11/2/25 19:48, Philippe Mathieu-Daudé wrote:
> > On 11/2/25 19:26, Stefan Hajnoczi wrote:
> > > On Mon, Feb 10, 2025 at 3:43 PM Philippe Mathieu-Daudé
> > > wrote:
> > > >
> > > > The following changes since commit
> > >
Hi,
> > +/* read header */
> > +dma_memory_read(&address_space_memory, dma,
> > +uv->buffer, sizeof(*mhdr),
> > +MEMTXATTRS_UNSPECIFIED);
>
> Depending on DMA sounds appealing at first, but can fall apart in corner
> cases. I know of 2 cases where
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine using the
device.
Signed-off-by: Philippe Mathieu-D
The archive used in test_microblaze_s3adsp1800.py (testing a
big-endian target) contains a big-endian kernel. Rename using
the _BE suffix.
Similarly, the archive in test_microblazeel_s3adsp1800 (testing
a little-endian target) contains a little-endian kernel. Rename
using _LE suffix.
These change
Commit f0ec14c78c4 ("tests/avocado: Fix console data loss") fixed
QEMUMachine's problem with console, we don't need to use the sleep()
kludges.
Suggested-by: Thomas Huth
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Thomas Huth
Reviewed-by: Richard Henderson
Message-Id: <20250206131052.30
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine using the device.
Signed-off-by: Philippe Mathieu-Daudé
Have the MicroblazeMachine class being common to both
MicroblazeBigEndianMachine and MicroblazeLittleEndianMachine
classes. Move the xmaton and ballerina tests to the parent class.
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Thomas Huth
Message-Id: <20250206131052.30207-16-phi...@linaro.o
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (default).
Reviewed-by: Thomas Huth
Signed-off-by: Philippe Mathieu-Daudé
---
qapi/common.json| 16
include/hw/qdev-properties-system.h | 7 +
Since v6:
- Simplify MemoryRegionOps indexing (Thomas)
Since v5:
- Introduce QAPI EndianMode
- Update RISCV machine while rebasing
- Fixed INTC use on PPC (Thomas)
- Dropped patch adding more machines (Daniel)
Since v4 & v3:
- Addressed Thomas review comments
Since v2:
- Addressed Richard's revi
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine using the device.
Signed-off-by: Philippe Mathieu-Daudé
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine using the
device.
Signed-off-by: Philippe Mathieu-D
Make microblaze tests a bit more generic.
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Thomas Huth
Message-Id: <20250206131052.30207-14-phi...@linaro.org>
---
tests/functional/test_microblaze_s3adsp1800.py | 7 +--
tests/functional/test_microblazeel_s3adsp1800.py | 7 +--
2 file
On 12/2/25 13:36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Grr, please read:
Add the "endianness" proper
Am 12.02.2025 um 10:28 hat Paolo Bonzini geschrieben:
> On 2/11/25 22:43, Kevin Wolf wrote:
> > +/// Use QEMU's event loops to run a Rust [`Future`] to completion and
> > return its result.
> > +///
> > +/// This function must be called in coroutine context. If the future isn't
> > ready yet, it
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine using the
device.
Signed-off-by: Philippe Mathieu-D
On 12/02/2025 13.36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine usin
Am 12.02.2025 um 10:41 hat Daniel P. Berrangé geschrieben:
> On Wed, Feb 12, 2025 at 09:14:57AM +, Daniel P. Berrangé wrote:
> > On Tue, Feb 11, 2025 at 10:43:27PM +0100, Kevin Wolf wrote:
> > > This adds a separate block driver for the bochs image format called
> > > 'bochs-rs' so that for the
On 04/02/25 10:53AM, Jonathan Cameron wrote:
On Mon, 3 Feb 2025 11:29:49 +0530
Arpit Kumar wrote:
Add some description of what is being added here.
Key issue in here is that clearing the CEL doesn't make
sense. It is a description of what the device supports, there
is no state to clear in it.
On 12/2/25 13:56, BALATON Zoltan wrote:
On Wed, 12 Feb 2025, Philippe Mathieu-Daudé wrote:
On 12/2/25 12:37, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (de
On 12.02.25 13:28, Gerd Hoffmann wrote:
On Wed, Feb 12, 2025 at 12:30:20PM +0100, Alexander Graf wrote:
On 12.02.25 11:24, Gerd Hoffmann wrote:
Why do you use confidential computing in the first place if you trust
the host with your EFI variables? I'd rather see something simliar
running und
On 12/2/25 12:42, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the p
On 04/02/25 10:28AM, Jonathan Cameron wrote:
On Mon, 3 Feb 2025 11:29:48 +0530
Arpit Kumar wrote:
Please add some descriptive teext here.
Sure, will append here in V2 patch.
Signed-off-by: Arpit Kumar
Reviewed-by: Alok Rathore
Reviewed-by: Krishna Kanth Reddy
Hi Arpit,
Whilst it is g
On 04/02/25 10:58AM, Jonathan Cameron wrote:
On Mon, 3 Feb 2025 11:29:50 +0530
Arpit Kumar wrote:
Signed-off-by: Arpit Kumar
Reviewed-by: Alok Rathore
Reviewed-by: Krishna Kanth Reddy
Likewise, the CEL isn't a dynamic thing. Asking to populate
it makes little sense so the log capability
On 12/02/2025 13.36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine usin
On Wed, Feb 12, 2025 at 01:58:15PM +0100, Kevin Wolf wrote:
> Am 12.02.2025 um 10:41 hat Daniel P. Berrangé geschrieben:
> > On Wed, Feb 12, 2025 at 09:14:57AM +, Daniel P. Berrangé wrote:
> > > On Tue, Feb 11, 2025 at 10:43:27PM +0100, Kevin Wolf wrote:
> > > > This adds a separate block drive
On 12/02/2025 13.36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine
On 12/02/2025 13.36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine
Am 12.02.2025 um 10:32 hat Hanna Czenczek geschrieben:
> RBD schedules the request completion code (qemu_rbd_finish_bh()) to run
> in the BDS's AioContext. The intent seems to be to run it in the same
> context that the original request coroutine ran in, i.e. the thread on
> whose stack the RBDTas
Hi,
On Tue, 11 Feb 2025 at 20:50, Peter Xu wrote:
> > * Yes. AFAIU, tls/file channels don't send magic values.
> Please double check whether TLS will send magics. AFAICT, they should.
===
* ... Also tls live migration already does
* tls handshake while initializing main channel so with tls t
Applied, thanks.
Please update the changelog at https://wiki.qemu.org/ChangeLog/10.0 for any
user-visible changes.
signature.asc
Description: PGP signature
On 12/2/25 14:53, Philippe Mathieu-Daudé wrote:
On 12/2/25 13:56, BALATON Zoltan wrote:
On Wed, 12 Feb 2025, Philippe Mathieu-Daudé wrote:
On 12/2/25 12:37, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
On Tue, Feb 11, 2025 at 5:52 PM Peter Xu wrote:
>
> From: Daniil Tatianin
>
> This will be used in the following commits to make it possible to only
> lock memory on fault instead of right away.
Hi Peter and Daniil,
Please take a look at this CI failure:
https://gitlab.com/qemu-project/qemu/-/jo
Hi Daniil,
On 12/2/25 15:39, Daniil Tatianin wrote:
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
---
include/system/os-posix.h | 2 +-
include/system/os-win32.h | 3 ++-
meson.build
On 12.02.25 17:39, Daniil Tatianin wrote:
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
Reviewed-by: Vladimir Sementsov-Ogievskiy
--
Best regards,
Vladimir
On 2/12/2025 07:38, Eugenio Perez Martin wrote:
On Tue, Feb 11, 2025 at 5:20 PM Konstantin Shkolnyy wrote:
Add .set_vnet_le() function that always returns success, assuming that
vDPA h/w always implements LE data format. Otherwise, QEMU disables vDPA and
outputs the message:
"backend does not
On 2/12/25 5:48 PM, Philippe Mathieu-Daudé wrote:
Hi Daniil,
On 12/2/25 15:39, Daniil Tatianin wrote:
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
---
include/system/os-posix.h | 2 +-
in
On 11/2/25 17:19, Konstantin Shkolnyy wrote:
Add .set_vnet_le() function that always returns success, assuming that
vDPA h/w always implements LE data format. Otherwise, QEMU disables vDPA and
outputs the message:
"backend does not support LE vnet headers; falling back on userspace virtio"
Signe
On 12/2/25 15:51, Daniil Tatianin wrote:
On 2/12/25 5:48 PM, Philippe Mathieu-Daudé wrote:
Hi Daniil,
On 12/2/25 15:39, Daniil Tatianin wrote:
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
--
On 30/1/25 19:36, Peter Maydell wrote:
On Thu, 30 Jan 2025 at 18:25, Philippe Mathieu-Daudé wrote:
When not specified, Cortex-A9MP configures its GIC with 64 external
IRQs (see commit a32134aad89 "arm:make the number of GIC interrupts
configurable"). Add the GIC_EXT_IRQS definition (with a com
Am 12.02.2025 um 10:29 hat Paolo Bonzini geschrieben:
> On 2/11/25 22:43, Kevin Wolf wrote:
> > Signed-off-by: Kevin Wolf
> > ---
> > rust/wrapper.h| 4
> > meson.build | 1 +
> > rust/qemu-api/src/zeroable.rs | 5 +++--
> > 3 files changed, 8 insertions
On Wed, 12 Feb 2025, Philippe Mathieu-Daudé wrote:
On 12/2/25 12:37, Thomas Huth wrote:
On 12/02/2025 12.24, Philippe Mathieu-Daudé wrote:
Introduce the EndianMode type and the DEFINE_PROP_ENDIAN() macros.
Endianness can be BIG, LITTLE or unspecified (default).
Signed-off-by: Philippe Mathieu-
On Wed, Feb 12, 2025 at 1:47 PM Kevin Wolf wrote:
> > > +pub fn qemu_co_run_future(future: F) -> F::Output {
> > > +let waker = Arc::new(RunFutureWaker {
> > > +co: unsafe { bindings::qemu_coroutine_self() },
> > > +})
> > > +.into();
> >
> > into what? :) Maybe you can add th
Am 12.02.2025 um 08:45 hat Philippe Mathieu-Daudé geschrieben:
> On 11/2/25 22:43, Kevin Wolf wrote:
> > This adds a separate block driver for the bochs image format called
> > 'bochs-rs' so that for the moment both the C implementation and the Rust
> > implementation can be present in the same bui
On 12/02/2025 13.36, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness on the single machine
On 1/30/25 11:08, Maciej S. Szmigiero wrote:
From: "Maciej S. Szmigiero"
The multifd received data needs to be reassembled since device state
packets sent via different multifd channels can arrive out-of-order.
Therefore, each VFIO device state packet carries a header indicating its
position i
On Wed, Feb 12, 2025 at 2:13 PM Kevin Wolf wrote:
> Yes, we definitely need some proper bindings there. I'm already tired of
> writing things like this:
>
> return -(bindings::EINVAL as std::os::raw::c_int)
>
> Or even:
>
> return e
> .raw_os_error()
> .unwrap_or(-(bindings
On 12/2/25 13:59, Kevin Wolf wrote:
Am 12.02.2025 um 08:45 hat Philippe Mathieu-Daudé geschrieben:
On 11/2/25 22:43, Kevin Wolf wrote:
This adds a separate block driver for the bochs image format called
'bochs-rs' so that for the moment both the C implementation and the Rust
implementation can
On Tue, Feb 11, 2025 at 5:20 PM Konstantin Shkolnyy wrote:
>
> Add .set_vnet_le() function that always returns success, assuming that
> vDPA h/w always implements LE data format. Otherwise, QEMU disables vDPA and
> outputs the message:
> "backend does not support LE vnet headers; falling back on u
On 2/12/25 5:13 PM, Stefan Hajnoczi wrote:
On Tue, Feb 11, 2025 at 5:52 PM Peter Xu wrote:
From: Daniil Tatianin
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Hi Peter and Daniil,
Please take a look at this CI failure:
ht
On 12.02.25 14:26, Kevin Wolf wrote:
Am 12.02.2025 um 10:32 hat Hanna Czenczek geschrieben:
RBD schedules the request completion code (qemu_rbd_finish_bh()) to run
in the BDS's AioContext. The intent seems to be to run it in the same
context that the original request coroutine ran in, i.e. the
Instead of comment
"Keep this type consistent with the nbd-server-start arguments", we
can simply merge these things.
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
No problem for me to rebase on top of
[PATCH 0/2] nbd: Allow debugging tuning of handshake limit
if it goes earlier.
Also, not th
On 11.02.25 00:46, Eric Blake wrote:
On Thu, Feb 06, 2025 at 10:20:09AM +0300, Vladimir Sementsov-Ogievskiy wrote:
---
qapi/block-export.json | 10 ++
include/block/nbd.h| 6 +++---
[..]
@@ -52,6 +57,10 @@
#
# @addr: Address on which to listen.
#
+#
On Wed, Feb 12, 2025 at 06:57:48PM +0530, Prasad Pandit wrote:
> Hi,
>
> On Tue, 11 Feb 2025 at 20:50, Peter Xu wrote:
> > > * Yes. AFAIU, tls/file channels don't send magic values.
> > Please double check whether TLS will send magics. AFAICT, they should.
> ===
> * ... Also tls live migration
Currently, passing mem-lock=on to QEMU causes memory usage to grow by
huge amounts:
no memlock:
$ ./qemu-system-x86_64 -overcommit mem-lock=off
$ ps -p $(pidof ./qemu-system-x86_64) -o rss=
45652
$ ./qemu-system-x86_64 -overcommit mem-lock=off -enable-kvm
$ ps -p $(pidof ./qem
This will be extended in the future commits, let's move it out of line
right away so that it's easier to read.
Reviewed-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Peter Xu
Signed-off-by: Daniil Tatianin
---
system/vl.c | 21 ++---
1 file changed, 14 insertions(+), 7 deletion
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
---
include/system/os-posix.h | 2 +-
include/system/os-win32.h | 3 ++-
meson.build | 6 ++
migration/postcopy-ram.c | 2 +-
o
Replace the boolean value enable_mlock with an enum and add a helper to
decide whether we should be calling os_mlock.
This is a stepping stone towards introducing a new mlock mode, which
will be the third possible state of this enum.
Reviewed-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Peter X
Locking the memory without MCL_ONFAULT instantly prefaults any mmaped
anonymous memory with a write-fault, which introduces a lot of extra
overhead in terms of memory usage when all you want to do is to prevent
kcompactd from migrating and compacting QEMU pages. Add an option to
only lock pages laz
The following changes since commit f9edf32ea2e18a56de5d92f57e9d10565c822367:
Merge tag 'pull-request-2025-02-11' of https://gitlab.com/thuth/qemu into
staging (2025-02-11 13:27:32 -0500)
are available in the Git repository at:
https://gitlab.com/stefanha/qemu.git tags/tracing-pull-request
Hi,
> > Yes. Knowing both physical and virtual address works only for memory
> > you allocated yourself before ExitBootServices. So you can't pass on
> > pointers from the OS, you have to copy the data to a buffer where you
> > know the physical address instead. Yes, some overhead. Should st
From: Daniel P. Berrangé
If the 'stap' binary is missing in $PATH, a huge trace is thrown
$ qemu-trace-stap list /usr/bin/qemu-system-x86_64
Traceback (most recent call last):
File "/usr/bin/qemu-trace-stap", line 169, in
main()
File "/usr/bin/qemu-trace-stap", line 165, in main
arg
On Mon, Feb 10, 2025 at 09:46:46AM +0100, David Hildenbrand wrote:
> We want to pass another flag that will be stored in MemTxAttrs. So pass
> MemTxAttrs directly.
>
> Reviewed-by: Peter Xu
> Reviewed-by: Philippe Mathieu-Daudé
> Signed-off-by: David Hildenbrand
> ---
> hw/core/loader.c
On 2/12/25 6:23 PM, Peter Xu wrote:
On Wed, Feb 12, 2025 at 05:39:17PM +0300, Daniil Tatianin wrote:
This will be used in the following commits to make it possible to only
lock memory on fault instead of right away.
Signed-off-by: Daniil Tatianin
---
include/system/os-posix.h | 2 +-
incl
On 12/2/25 10:34, Philippe Mathieu-Daudé wrote:
On 12/2/25 10:19, Philippe Mathieu-Daudé wrote:
On 12/2/25 09:27, Thomas Huth wrote:
On 10/02/2025 21.41, Philippe Mathieu-Daudé wrote:
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add th
On 2/11/25 22:43, Kevin Wolf wrote:
Signed-off-by: Kevin Wolf
---
rust/wrapper.h| 4
meson.build | 1 +
rust/qemu-api/src/zeroable.rs | 5 +++--
3 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/rust/wrapper.h b/rust/wrapper.h
index 41be87
Acked-by: zhenwei pi
On 2/11/25 21:55, Stefano Garzarella wrote:
The function `vhost_dev_init()` expects the `struct vhost_dev`
(passed as a parameter) to be fully initialized. This is important
because some parts of the code check whether `vhost_dev->config_ops`
is NULL to determine if it has
Hi Stefan,
Are you ok with queuing this patch ?
On Fri, Dec 06, 2024 at 11:45:24AM +, Daniel P. Berrangé wrote:
> If the 'stap' binary is missing in $PATH, a huge trace is thrown
>
> $ qemu-trace-stap list /usr/bin/qemu-system-x86_64
> Traceback (most recent call last):
> File "/usr/bi
On 2/11/25 22:43, Kevin Wolf wrote:
The existing qemu_api library can't be linked into tools because it
contains a few bindings for things that only exist in the system
emulator.
This adds a new "system" feature to the qemu_api crate that enables the
system emulator parts in it, and build the cr
Replace the DEVICE_NATIVE_ENDIAN MemoryRegionOps by a pair
of DEVICE_LITTLE_ENDIAN / DEVICE_BIG_ENDIAN.
Add the "little-endian" property to select the device
endianness, defaulting to little endian.
Set the proper endianness for each machine using the device.
Signed-off-by: Philippe Mathieu-Daudé
Since v5:
- Introduce QAPI EndianMode
- Update RISCV machine while rebasing
- Fixed INTC use on PPC (Thomas)
- Dropped patch adding more machines (Daniel)
Since v4 & v3:
- Addressed Thomas review comments
Since v2:
- Addressed Richard's review comments
Since v1:
- Make device endianness configur
1 - 100 of 273 matches
Mail list logo
