This patch includes the following tests
Test mmio read
Test ufs device initialization and ufs-lu recognition
Test I/O (Performs a write followed by a read to verify)
Signed-off-by: Jeuk Kim
---
MAINTAINERS | 1 +
tests/qtest/meson.build | 1 +
tests/qtest/ufs-test.c | 575
Universal Flash Storage (UFS) is a high-performance mass storage device
with a serial interface. It is primarily used as a high-performance
data storage device for embedded applications.
This commit contains code for UFS device to be recognized
as a UFS PCI device.
Patches to handle UFS logical
This commit adds support for ufs logical unit.
The LU handles processing for the SCSI command,
unit descriptor query request.
This commit enables the UFS device to process
IO requests.
Signed-off-by: Jeuk Kim
---
hw/ufs/lu.c | 1445 ++
This commit makes the UFS device support query
and nop out transfer requests.
The next patch would be support for UFS logical
unit and scsi command transfer request.
Signed-off-by: Jeuk Kim
---
hw/ufs/trace-events | 1 +
hw/ufs/ufs.c| 980 +++-
Since v6:
- Add tests/qtest/ufs-test.c to test ufs initialisation and I/O
- Add struct UtpTaskReqDesc to include/block/ufs.h
- Fix ufs_log2() logic
- Fix ufs-lu to use 4K as default block size to match the ufs spec
Since I created a new file, tests/qtest/ufs-test.c, I added Laurent Vivier to
the
25.07.2023 22:11, Bastian Koppelmann wrote:
...
Michael Tokarev has already picked it up. See
https://lore.kernel.org/qemu-devel/20230725145829.37782-11-phi...@linaro.org/T/#u
I noticed that too, we did it almost at the same time.
But there's nothing wrong with that. It doesn't matter
how a
On Mon, Jul 24, 2023 at 05:46:11PM +0200, Stefano Garzarella wrote:
> The way the virtio-blk driver is implemented in libblkio,
> it's much easier to use blkio_set_int() instead of blkio_get_int()
> and have it fail right away to see if `fd` is supported by the
> transport. See
On Mon, Jul 24, 2023 at 05:46:10PM +0200, Stefano Garzarella wrote:
> libblkio 1.3.0 added support of "fd" property for virtio-blk-vhost-vdpa
> driver. In QEMU, starting from commit cad2ccc395 ("block/blkio: use
> qemu_open() to support fd passing for virtio-blk") we are using
>
On Tue, Jul 25, 2023 at 12:37:44PM +0200, Stefano Garzarella wrote:
> Until libblkio 1.3.0, virtio-blk drivers had completion eventfd
> notifications enabled from the start, but from the next releases
> this is no longer the case, so we have to explicitly enable them.
>
> In fact, the libblkio
Hi Phil,
On Tue, Jul 25, 2023 at 04:58:29PM +0200, Philippe Mathieu-Daudé wrote:
> From: Bastian Koppelmann
>
> this name is used by capstone and will lead to a build failure of QEMU,
> when capstone is enabled. So we rename it to tricore_has_feature(), to
> match has_feature() in translate.c.
available in the Git repository at:
>
> https://github.com/philmd/qemu.git tags/misc-fixes-20230725
>
> for you to fetch changes up to f8cfdd2038c1823301e6df753242e465b1dc8539:
>
> target/tricore: Rename tricore_feature (2023-07-25 14:42:00 +0200)
>
> --
On Tue, Jul 25, 2023 at 9:04 AM Philippe Mathieu-Daudé
wrote:
>
> Hi Niklas, John, Paolo, Kevin,
>
> On 19/7/23 12:47, Niklas Cassel wrote:
>
> >> Niklas Cassel (8):
> >>hw/ide/ahci: remove stray backslash
> >>hw/ide/core: set ERR_STAT in unsupported command completion
> >>
On Tue, Jul 25, 2023 at 02:33:36PM -0400, John Snow wrote:
> On Tue, Jul 25, 2023 at 2:26 PM Daniel P. Berrangé
> wrote:
> >
> > On Tue, Jul 25, 2023 at 02:03:37PM -0400, John Snow wrote:
> > > By using a socketpair for all of the sockets managed by the VM class and
> > > its extensions, we
On Tue, Jul 25, 2023 at 2:26 PM Daniel P. Berrangé wrote:
>
> On Tue, Jul 25, 2023 at 02:03:37PM -0400, John Snow wrote:
> > By using a socketpair for all of the sockets managed by the VM class and
> > its extensions, we don't need the sock_dir argument anymore, so remove
> > it.
> >
> > We only
On Tue, Jul 25, 2023 at 02:03:37PM -0400, John Snow wrote:
> By using a socketpair for all of the sockets managed by the VM class and
> its extensions, we don't need the sock_dir argument anymore, so remove
> it.
>
> We only added this argument so that we could specify a second, shorter
>
On Tue, Jul 25, 2023 at 02:03:36PM -0400, John Snow wrote:
> Like the QMP and console sockets, begin using socketpairs for the qtest
> connection, too. After this patch, we'll be able to remove the vestigial
> sock_dir argument, but that cleanup is best done in its own patch.
>
> Signed-off-by:
On Tue, Jul 25, 2023 at 02:03:35PM -0400, John Snow wrote:
> Create a socketpair for the console output. This should help eliminate
> race conditions around console text early in the boot process that might
> otherwise have been dropped on the floor before being able to connect to
> QEMU under
On Tue, Jul 25, 2023 at 02:03:34PM -0400, John Snow wrote:
> Useful if we want to use ConsoleSocket() for a socket created by
> socketpair().
>
> Signed-off-by: John Snow
> ---
> python/qemu/machine/console_socket.py | 29 +++
> 1 file changed, 21 insertions(+), 8
On Tue, Jul 25, 2023 at 02:03:33PM -0400, John Snow wrote:
> If everything has gone smoothly, we'll already have closed the socket we
> gave to the child during post_launch. The other half of the pair that we
> gave to the QMP connection should, likewise, be definitively closed by
> now.
>
>
On Tue, Jul 25, 2023 at 02:03:32PM -0400, John Snow wrote:
> This property isn't meant to do much else besides return a list of
> strings, so move this setup back out into _pre_launch().
>
> Signed-off-by: John Snow
> ---
> python/qemu/machine/machine.py | 5 ++---
> 1 file changed, 2
Like we did for the QMP socket, use socketpair() for the console socket
so that hopefully there isn't a race condition during early boot where
data might get dropped on the floor.
May or may not help with various race conditions where early console
output is not showing up in the logs and/or
Like the QMP and console sockets, begin using socketpairs for the qtest
connection, too. After this patch, we'll be able to remove the vestigial
sock_dir argument, but that cleanup is best done in its own patch.
Signed-off-by: John Snow
---
python/qemu/machine/qtest.py | 49
Create a socketpair for the console output. This should help eliminate
race conditions around console text early in the boot process that might
otherwise have been dropped on the floor before being able to connect to
QEMU under "server,nowait".
Signed-off-by: John Snow
---
By using a socketpair for all of the sockets managed by the VM class and
its extensions, we don't need the sock_dir argument anymore, so remove
it.
We only added this argument so that we could specify a second, shorter
temporary directory for cases where the temp/log dirs were "too long" as
a
Useful if we want to use ConsoleSocket() for a socket created by
socketpair().
Signed-off-by: John Snow
---
python/qemu/machine/console_socket.py | 29 +++
1 file changed, 21 insertions(+), 8 deletions(-)
diff --git a/python/qemu/machine/console_socket.py
If everything has gone smoothly, we'll already have closed the socket we
gave to the child during post_launch. The other half of the pair that we
gave to the QMP connection should, likewise, be definitively closed by
now.
However, in the cleanup path, it's possible we've created the socketpair
This property isn't meant to do much else besides return a list of
strings, so move this setup back out into _pre_launch().
Signed-off-by: John Snow
---
python/qemu/machine/machine.py | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/python/qemu/machine/machine.py
From: Vladimir Sementsov-Ogievskiy
Actually block job is not completed without this final flush. It's
rather unexpected to have broken target when job was successfully
completed long ago and now we fail to flush or process just
crashed/killed.
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
There seems to be an intermittent failure on the s390 host in
the qemu:block / io-qcow2-copy-before-write test:
https://gitlab.com/qemu-project/qemu/-/jobs/4737819873
The log says the test was expecting to do some reading
and writing but got an unexpected 'permission denied'
error on the read.
On 25.07.23 18:13, Denis V. Lunev wrote:
On 7/25/23 16:25, Vladimir Sementsov-Ogievskiy wrote:
On 24.07.23 10:30, Evanzhang wrote:
block_stream will not actively flush l2_table_cache,when qemu
process exception exit,causing disk data loss
Signed-off-by: Evanzhang
---
block/stream.c | 6
On 7/25/23 16:25, Vladimir Sementsov-Ogievskiy wrote:
On 24.07.23 10:30, Evanzhang wrote:
block_stream will not actively flush l2_table_cache,when qemu
process exception exit,causing disk data loss
Signed-off-by: Evanzhang
---
block/stream.c | 6 ++
1 file changed, 6 insertions(+)
From: Peter Maydell
In CPUSparcState we define the fprs field as uint64_t. However we
then refer to it in translate.c via a TCGv_i32 which we set up with
tcg_global_mem_new_ptr(). This means that on a big-endian host when
the guest does something to writo te the FPRS register this value
ends
From: Bastian Koppelmann
this name is used by capstone and will lead to a build failure of QEMU,
when capstone is enabled. So we rename it to tricore_has_feature(), to
match has_feature() in translate.c.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1774
Signed-off-by: Bastian
Coverity reports a potential overruns (CID 1517770):
Overrunning array "mxu_gpr" of 15 8-byte elements at
element index 4294967295 (byte offset 34359738367)
using index "XRb - 1U" (which evaluates to 4294967295).
Add a gen_extract_mxu_gpr() helper similar to
gen_load_mxu_gpr() to safely
From: Thomas Huth
It's possible to compile QEMU without the USB devices (e.g. when using
"--without-default-devices" as option for the "configure" script).
To be still able to run the loongson3-virt machine in default mode with
such a QEMU binary, we have to check here for the availability of
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Richard Henderson
Message-Id: <20230712060806.82323-2-phi...@linaro.org>
---
target/mips/tcg/mxu_translate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/target/mips/tcg/mxu_translate.c b/target/mips/tcg/mxu_translate.c
We already evaluated directory_shift and leaf_shift in
page_table_walk_refill(), no need to do that again: pass
as argument.
Signed-off-by: Philippe Mathieu-Daudé
Reviewed-by: Peter Maydell
Message-Id: <20230717213504.24777-2-phi...@linaro.org>
---
target/mips/tcg/sysemu/tlb_helper.c | 18
Coverity reports a potential overrun (CID 1517769):
Overrunning array "mxu_gpr" of 15 8-byte elements at
element index 4294967295 (byte offset 34359738367)
using index "XRb - 1U" (which evaluates to 4294967295).
Use gen_load_mxu_gpr() to safely load MXU registers.
Fixes: ff7936f009
From: Peter Maydell
Coverity points out that in page_table_walk_refill() we can
shift by a negative number, which is undefined behaviour
(CID 1452918, 1452920, 1452922). We already catch the
negative directory_shift and leaf_shift as being a "bail
out early" case, but not until we've already
From: Thomas Huth
The firmware of the m68k next-cube machine uses the loopback mode
for self-testing the hardware and currently fails during this step.
By implementing the loopback mode, we can make the firmware pass
to the next step.
Signed-off-by: Thomas Huth
Reviewed-by: Mark Cave-Ayland
The following changes since commit 3ee44ec72753ec0ff05ad1569dfa609203d722b2:
Merge tag 'pull-request-2023-07-24' of https://gitlab.com/thuth/qemu into
staging (2023-07-24 18:06:36 +0100)
are available in the Git repository at:
https://github.com/philmd/qemu.git tags/misc-fixes-20230725
From: Bernhard Beschow
Since commit c0a55a0c9da2 "hw/sd/sdhci: Support big endian SD host controller
interfaces" sdhci_common_realize() forces all SD card controllers to use either
sdhci_mmio_le_ops or sdhci_mmio_be_ops, depending on the "endianness" property.
However, there are device models
On 24.07.23 10:30, Evanzhang wrote:
block_stream will not actively flush l2_table_cache,when qemu
process exception exit,causing disk data loss
Signed-off-by: Evanzhang
---
block/stream.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/block/stream.c b/block/stream.c
index
The libblkio-based io_uring block driver currently has no test coverage.
Running existing qemu-iotests with io_uring is possible with just a few
tweaks.
Signed-off-by: Stefan Hajnoczi
---
tests/qemu-iotests/033 | 1 +
tests/qemu-iotests/check | 2 +-
tests/qemu-iotests/common.rc |
197 and 215 modify IMGPROTO, IMGFMT, etc temporarily while creating an
additional image file after the main test image. The backing file name
still has the old IMGPROTO, IMGFMT, etc. This means _filter_img_create
will not filter the backing_file= output from qemu-img.
Add a helper that filters
_filter_img_create_filenames fails to remove $IMGPROTO: when
$REMOTE_TEST_DIR matches. echo "$IMGPROTO:$REMOTE_TEST_DIR" |
_filter_img_create_filenames results in "$IMGPROTO:TEST_DIR" instead of
"TEST_DIR".
Substitute "s#$REMOTE_TEST_DIR#TEST_DIR#g" after
"s#$IMGPROTO:$TEST_DIR#TEST_DIR#g" since
The qemu-iotests test suite requires filename parsing because it does
not use image options syntax everywhere. Add it now so that later
patches can enable qemu-iotests for the io_uring block driver.
The blkio.c code has other libblkio-based drivers that could benefit
from filename parsing too.
Signed-off-by: Stefan Hajnoczi
---
tests/qemu-iotests/052 | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/qemu-iotests/052 b/tests/qemu-iotests/052
index 2f23ac9b65..42bf4c4aef 100755
--- a/tests/qemu-iotests/052
+++ b/tests/qemu-iotests/052
@@ -38,7 +38,7 @@ trap
Signed-off-by: Stefan Hajnoczi
---
tests/qemu-iotests/012 | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/qemu-iotests/012 b/tests/qemu-iotests/012
index 3a24d2ca8d..25dfc8ae12 100755
--- a/tests/qemu-iotests/012
+++ b/tests/qemu-iotests/012
@@ -40,7 +40,7 @@ trap
Some block drivers do not support image creation. This is inconvenient
for test cases, which usually want to create an image before testing
various I/O requests.
Use bdrv_co_create_opts_simple(), which initializes a pre-existing image
instead of creating a new image, so that tests will be able to
Add support for block/blkio.c's "io_uring" block driver to qemu-iotests:
$ (cd build && tests/qemu-iotests/check -io_uring)
I took a look at supporting ./check -qcow2 -io_uring, but there are many
failures related to test assumptions about image creation and QEMU command-line
syntax. Only
Hi Niklas, John, Paolo, Kevin,
On 19/7/23 12:47, Niklas Cassel wrote:
Niklas Cassel (8):
hw/ide/ahci: remove stray backslash
hw/ide/core: set ERR_STAT in unsupported command completion
hw/ide/ahci: write D2H FIS when processing NCQ command
hw/ide/ahci: simplify and document PxCI
On Tue, Jul 25, 2023 at 12:15:40PM +0100, Daniel P. Berrangé wrote:
On Tue, Jul 25, 2023 at 01:11:55PM +0200, Stefano Garzarella wrote:
qemu_open() in blkio_virtio_blk_common_open() is used to open the
character device (e.g. /dev/vhost-vdpa-0 or /dev/vfio/vfio) or in
the future eventually the
On Tue, Jul 25, 2023 at 01:11:55PM +0200, Stefano Garzarella wrote:
> qemu_open() in blkio_virtio_blk_common_open() is used to open the
> character device (e.g. /dev/vhost-vdpa-0 or /dev/vfio/vfio) or in
> the future eventually the unix socket.
>
> In all these cases we cannot open the path in
qemu_open() in blkio_virtio_blk_common_open() is used to open the
character device (e.g. /dev/vhost-vdpa-0 or /dev/vfio/vfio) or in
the future eventually the unix socket.
In all these cases we cannot open the path in read-only mode,
when the `read-only` option of blockdev is on, because the
> 2023年7月25日 下午6:42，Li Feng 写道：
>
> If the backend crashes and restarts, the device is broken.
> This patch adds reconnect for vhost-user-scsi.
>
> Tested with spdk backend.
>
> Signed-off-by: Li Feng
> ---
> hw/scsi/vhost-scsi-common.c | 6 +
> hw/scsi/vhost-user-scsi.c
Multile devices need this macro, move it to a common header.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c
When the vhost-user reconnect to the backend, the notifer should be
cleanup. Otherwise, the fd resource will be exhausted.
Fixes: f9a09ca3ea ("vhost: add support for configure interrupt")
Signed-off-by: Li Feng
---
hw/virtio/vhost.c | 2 ++
1 file changed, 2 insertions(+)
diff --git
Hi,
This patchset adds reconnect support for vhost-user-scsi. At the same
times, fix vhost fd leak and refactor some code.
Changes for v2:
- Split the v1 patch to small separate patchset;
- New patch for fixing fd leak, which has sent to reviewers in another
mail;
- Implement the
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 6 +
hw/scsi/vhost-user-scsi.c | 220 +++---
Get_inflight_fd is sent only once. When reconnecting to the backend,
qemu sent set_inflight_fd to the backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git
Until libblkio 1.3.0, virtio-blk drivers had completion eventfd
notifications enabled from the start, but from the next releases
this is no longer the case, so we have to explicitly enable them.
In fact, the libblkio documentation says they could be disabled,
so we should always enable them at
Thanks for your comments.
> 2023年7月25日 上午1:21，Raphael Norwitz 写道：
>
> Very excited to see this. High level looks good modulo a few small things.
>
> My major concern is around existing vhost-user-scsi backends which don’t
> support VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD. IMO we should hide the
63 matches
Mail list logo
