Between the submission of a request and the unplug, other devices
with larger limits may have been queued new requests without flushing
the batch.
Using the new `dev_max_batch` parameter, laio_io_unplug() can check
if the batch exceeds the device limit to flush the current batch.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20211026162346.253081-4-sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This new parameter can be used by block devices to limit the
Linux AIO batch size more than the limit set by the AIO context.
file-posix backend supports this, passing its `aio-max-batch` option
previously added.
Add an helper function to calculate the maximum batch size.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20211026162346.253081-3-sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Commit d7ddd0a161 ("linux-aio: limit the batch size using
`aio-max-batch` parameter") added a way to limit the batch size
of Linux AIO backend for the entire AIO context.
The same AIO context can be shared by multiple devices, so
latency-sensitive devices may want to limit the batch size even
more to avoid increasing latency.
For this reason we add the `aio-max-batch` option to the file
backend, which will be used by the next commits to limit the size of
batches including requests generated by this device.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20211026162346.253081-2-sgarzare@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Commit 0445409d74 ("iothread: generalize
iothread_set_param/iothread_get_param") moved common code to set and
get IOThread parameters in two new functions.
These functions are called inside callbacks, so we don't need to use an
opaque pointer. Let's replace `void *opaque` parameter with
`IOThreadParamInfo *info`.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210727145936.147032-3-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Commit 1793ad0247 ("iothread: add aio-max-batch parameter") added
a new parameter (aio-max-batch) to IOThread and used PollParamInfo
structure to handle it.
Since it is not a parameter of the polling mechanism, we rename the
structure to a more generic IOThreadParamInfo.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20210727145936.147032-2-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
virtio-vsock features, like VIRTIO_VSOCK_F_SEQPACKET, can be handled
by vhost-vsock-common parent class. In this way, we can reuse the
same code for all virtio-vsock backends (i.e. vhost-vsock,
vhost-user-vsock).
Let's move `seqpacket` property to vhost-vsock-common class, add
vhost_vsock_common_get_features() used by children, and disable
`seqpacket` for vhost-user-vsock device for machine types < 6.2.
The behavior of vhost-vsock device doesn't change; vhost-user-vsock
device now supports `seqpacket` property.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210921161642.206461-3-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Commit 1e08fd0a46 ("vhost-vsock: SOCK_SEQPACKET feature bit support")
enabled the SEQPACKET feature bit.
This commit is released with QEMU 6.1, so if we try to migrate a VM where
the host kernel supports SEQPACKET but machine type version is less than
6.1, we get the following errors:
Features 0x130000002 unsupported. Allowed features: 0x179000000
Failed to load virtio-vhost_vsock:virtio
error while loading state for instance 0x0 of device '0000:00:05.0/virtio-vhost_vsock'
load of migration failed: Operation not permitted
Let's disable the feature bit for machine types < 6.1.
We add a new OnOffAuto property for this, called `seqpacket`.
When it is `auto` (default), QEMU behaves as before, trying to enable the
feature, when it is `on` QEMU will fail if the backend (vhost-vsock
kernel module) doesn't support it.
Fixes: 1e08fd0a46 ("vhost-vsock: SOCK_SEQPACKET feature bit support")
Cc: qemu-stable@nongnu.org
Reported-by: Jiang Wang <jiang.wang@bytedance.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210921161642.206461-2-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
QAPI generates `struct BackupPerf` where `max-workers` value is stored
in an `int64_t` variable.
But block_copy_async(), and the underlying code, uses an `int` parameter.
At the end that variable is used to initialize `max_busy_tasks` in
block/aio_task.c causing the following assertion failure if a value
greater than INT_MAX(2147483647) is used:
../block/aio_task.c:63: aio_task_pool_wait_one: Assertion `pool->busy_tasks > 0' failed.
Let's check that `max-workers` doesn't exceed INT_MAX and print an
error in that case.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2009310
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20211005161157.282396-2-sgarzare@redhat.com>
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
In mirror_iteration() we call mirror_wait_on_conflicts() with
`self` parameter set to NULL.
Starting from commit d44dae1a7c we dereference `self` pointer in
mirror_wait_on_conflicts() without checks if it is not NULL.
Backtrace:
Program terminated with signal SIGSEGV, Segmentation fault.
#0 mirror_wait_on_conflicts (self=0x0, s=<optimized out>, offset=<optimized out>, bytes=<optimized out>)
at ../block/mirror.c:172
172 self->waiting_for_op = op;
[Current thread is 1 (Thread 0x7f0908931ec0 (LWP 380249))]
(gdb) bt
#0 mirror_wait_on_conflicts (self=0x0, s=<optimized out>, offset=<optimized out>, bytes=<optimized out>)
at ../block/mirror.c:172
#1 0x00005610c5d9d631 in mirror_run (job=0x5610c76a2c00, errp=<optimized out>) at ../block/mirror.c:491
#2 0x00005610c5d58726 in job_co_entry (opaque=0x5610c76a2c00) at ../job.c:917
#3 0x00005610c5f046c6 in coroutine_trampoline (i0=<optimized out>, i1=<optimized out>)
at ../util/coroutine-ucontext.c:173
#4 0x00007f0909975820 in ?? () at ../sysdeps/unix/sysv/linux/x86_64/__start_context.S:91
from /usr/lib64/libc.so.6
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=2001404
Fixes: d44dae1a7c ("block/mirror: fix active mirror dead-lock in mirror_wait_on_conflicts")
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210910124533.288318-1-sgarzare@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
When there are multiple queues attached to the same AIO context,
some requests may experience high latency, since in the worst case
the AIO engine queue is only flushed when it is full (MAX_EVENTS) or
there are no more queues plugged.
Commit 2558cb8dd4 ("linux-aio: increasing MAX_EVENTS to a larger
hardcoded value") changed MAX_EVENTS from 128 to 1024, to increase
the number of in-flight requests. But this change also increased
the potential maximum batch to 1024 elements.
When there is a single queue attached to the AIO context, the issue
is mitigated from laio_io_unplug() that will flush the queue every
time is invoked since there can't be others queue plugged.
Let's use the new `aio-max-batch` IOThread parameter to mitigate
this issue, limiting the number of requests in a batch.
We also define a default value (32): this value is obtained running
some benchmarks and it represents a good tradeoff between the latency
increase while a request is queued and the cost of the io_submit(2)
system call.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-4-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The `aio-max-batch` parameter will be propagated to AIO engines
and it will be used to control the maximum number of queued requests.
When there are in queue a number of requests equal to `aio-max-batch`,
the engine invokes the system call to forward the requests to the kernel.
This parameter allows us to control the maximum batch size to reduce
the latency that requests might accumulate while queued in the AIO
engine queue.
If `aio-max-batch` is equal to 0 (default value), the AIO engine will
use its default maximum batch size value.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-3-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Changes in preparation for next patches where we add a new
parameter not related to the poll mechanism.
Let's add two new generic functions (iothread_set_param and
iothread_get_param) that we use to set and get IOThread
parameters.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-2-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Commit cc3d15a5ea ("docs: rstfy s390 dasd ipl documentation")
converted docs/devel/s390-dasd-ipl.txt to docs/devel/s390-dasd-ipl.rst.
We still have several references to the old file, so let's fix them
with the following command:
sed -i s/s390-dasd-ipl.txt/s390-dasd-ipl.rst/ \
$(git grep -l docs/devel/s390-dasd-ipl.txt)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20210517151702.109066-6-sgarzare@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Commit 6e8a3ff6ed ("docs/specs/tpm: reST-ify TPM documentation")
converted docs/specs/tpm.txt to docs/specs/tpm.rst.
We still have several references to the old file, so let's fix them
with the following command:
sed -i s/tpm.txt/tpm.rst/ $(git grep -l docs/specs/tpm.txt)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210517151702.109066-5-sgarzare@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Commit a14f0bf165 ("docs: convert build system documentation to rST")
converted docs/devel/build-system.txt to docs/devel/build-system.rst.
We still have several references to the old file, so let's fix them
with the following command:
sed -i s/build-system.txt/build-system.rst/ \
$(git grep -l docs/devel/build-system.txt)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210517151702.109066-4-sgarzare@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Commit 15e8699f00 ("atomics: convert to reStructuredText") converted
docs/devel/atomics.txt to docs/devel/atomics.rst.
We still have several references to the old file, so let's fix them
with the following command:
sed -i s/atomics.txt/atomics.rst/ $(git grep -l docs/devel/atomics.txt)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210517151702.109066-3-sgarzare@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Commit e50caf4a5c ("tracing: convert documentation to rST")
converted docs/devel/tracing.txt to docs/devel/tracing.rst.
We still have several references to the old file, so let's fix them
with the following command:
sed -i s/tracing.txt/tracing.rst/ $(git grep -l docs/devel/tracing.txt)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210517151702.109066-2-sgarzare@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
get_relocated_path() allocates a GString object and returns the
character data (C string) to the caller without freeing the memory
allocated for that object as reported by valgrind:
24 bytes in 1 blocks are definitely lost in loss record 2,805 of 6,532
at 0x4839809: malloc (vg_replace_malloc.c:307)
by 0x55AABB8: g_malloc (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x55C2481: g_slice_alloc (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x55C4827: g_string_sized_new (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x55C4CEA: g_string_new (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x906314: get_relocated_path (cutils.c:1036)
by 0x6E1F77: qemu_read_default_config_file (vl.c:2122)
by 0x6E1F77: qemu_init (vl.c:2687)
by 0x3E3AF8: main (main.c:49)
Let's use g_string_free(gstring, false) to free only the GString object
and transfer the ownership of the character data to the caller.
Fixes: f4f5ed2cbd ("cutils: introduce get_relocated_path")
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210412170255.231406-1-sgarzare@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
When we allocate 'q_namespace', we forgot to set 'has_q_namespace'
to true. This can cause several issues, including a memory leak,
since qapi_free_BlockdevCreateOptions() does not deallocate that
memory, as reported by valgrind:
13 bytes in 1 blocks are definitely lost in loss record 7 of 96
at 0x4839809: malloc (vg_replace_malloc.c:307)
by 0x48CEBB8: g_malloc (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x48E3FE3: g_strdup (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x180010: qemu_rbd_co_create_opts (rbd.c:446)
by 0x1AE72C: bdrv_create_co_entry (block.c:492)
by 0x241902: coroutine_trampoline (coroutine-ucontext.c:173)
by 0x57530AF: ??? (in /usr/lib64/libc-2.32.so)
by 0x1FFEFFFA6F: ???
Fix setting 'has_q_namespace' to true when we allocate 'q_namespace'.
Fixes: 19ae9ae014 ("block/rbd: Add support for ceph namespaces")
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210329150129.121182-3-sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In qemu_rbd_connect(), 'mon_host' is allocated by qemu_rbd_mon_host()
using g_strjoinv(), but it's only freed in the error path, leaking
memory in the success path as reported by valgrind:
80 bytes in 4 blocks are definitely lost in loss record 5,028 of 6,516
at 0x4839809: malloc (vg_replace_malloc.c:307)
by 0x5315BB8: g_malloc (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x532B6FF: g_strjoinv (in /usr/lib64/libglib-2.0.so.0.6600.8)
by 0x87D07E: qemu_rbd_mon_host (rbd.c:538)
by 0x87D07E: qemu_rbd_connect (rbd.c:562)
by 0x87E1CE: qemu_rbd_open (rbd.c:740)
by 0x840EB1: bdrv_open_driver (block.c:1528)
by 0x8453A9: bdrv_open_common (block.c:1802)
by 0x8453A9: bdrv_open_inherit (block.c:3444)
by 0x8464C2: bdrv_open (block.c:3537)
by 0x8108CD: qmp_blockdev_add (blockdev.c:3569)
by 0x8EA61B: qmp_marshal_blockdev_add (qapi-commands-block-core.c:1086)
by 0x90B528: do_qmp_dispatch_bh (qmp-dispatch.c:131)
by 0x907EA4: aio_bh_poll (async.c:164)
Fix freeing 'mon_host' also when qemu_rbd_connect() ends correctly.
Fixes: 0a55679b4a
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210329150129.121182-2-sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
QemuOpts is usually created merging the QemuOptsList of format
and protocol. So, when the format calls bdr_create_file(), the 'opts'
parameter contains a QemuOptsList with a combination of format and
protocol default values.
The format properly removes its options before calling
bdr_create_file(), but the default values remain in 'opts->list'.
So if the protocol has options with the same name (e.g. rbd has
'cluster_size' as qcow2), it will see the default values of the format,
since for overlapping options, the format wins.
To avoid this issue, lets convert QemuOpts to QDict, in this way we take
only the set options, and then convert it back to QemuOpts, using the
'create_opts' of the protocol. So the new QemuOpts, will contain only the
protocol defaults.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210308161232.248833-1-sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When a block job fails, we report strerror(-job->job.ret) error
message, also if the job set an error object.
Let's report a better error message using error_get_pretty(job->job.err).
If an error object was not set, strerror(-job->ret) is used as fallback,
as explained in include/qemu/job.h:
typedef struct Job {
...
/**
* Error object for a failed job.
* If job->ret is nonzero and an error object was not set, it will be set
* to strerror(-job->ret) during job_completed.
*/
Error *err;
}
In block_job_query() there can be a transient where 'job.err' is not set
by a scheduled bottom half. In that case we use strerror(-job->ret) as it
was before.
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210225103633.76746-1-sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Commit 9d7bd0826f introduced a new 'use-disabled-flag' property
set to true by default.
To allow the migration, we set this property to false in the hw_compat,
but in the wrong place (hw_compat_4_1).
Since commit 9d7bd0826f was released with QEMU 5.0, we move
'use-disabled-flag' property to hw_compat_4_2, so 4.2 machine types
will have the pre-patch behavior and the migration can work.
The issue was discovered with vhost-vsock device and 4.2 machine
type without running any kernel in the VM:
$ qemu-4.2 -M pc-q35-4.2,accel=kvm \
-device vhost-vsock-pci,guest-cid=4 \
-monitor stdio -incoming tcp:0:3333
$ qemu-5.2 -M pc-q35-4.2,accel=kvm \
-device vhost-vsock-pci,guest-cid=3 \
-monitor stdio
(qemu) migrate -d tcp:0:3333
# qemu-4.2 output
qemu-system-x86_64: Failed to load virtio-vhost_vsock:virtio
qemu-system-x86_64: error while loading state for instance 0x0 of device '0000:00:03.0/virtio-vhost_vsock'
qemu-system-x86_64: load of migration failed: No such file or directory
Reported-by: Jing Zhao <jinzhao@redhat.com>
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1907255
Fixes: 9d7bd0826f ("virtio-pci: disable vring processing when bus-mastering is disabled")
Cc: mdroth@linux.vnet.ibm.com
CC: qemu-stable@nongnu.org
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20210108171252.209502-1-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
vhost IOTLB API uses read()/write() to exchange iotlb messages with
the kernel module.
The QEMU implementation expects a non-blocking fd, indeed commit
c471ad0e9b ("vhost_net: device IOTLB support") set it for vhost-net.
Without this patch, if we enable iommu for the vhost-vsock device,
QEMU can hang when exchanging IOTLB messages.
As commit 894022e616 ("net: check if the file descriptor is valid
before using it") did for tap, let's use qemu_try_set_nonblock()
when fd is provided by the user.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20201029144849.70958-1-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Running checkpatch on a directory that contains a cover letter reports
this error:
Checking /tmp/tmpbnngauy3/0000-cover-letter.patch...
ERROR: Does not appear to be a unified-diff format patch
total: 1 errors, 0 warnings, 0 lines checked
Let's skip cover letter as it is already done in the Linux kernel
commits 06330fc40e3f ("checkpatch: avoid NOT_UNIFIED_DIFF errors
on cover-letter.patch files") and a08ffbef4ab7 ("checkpatch: fix
ignoring cover-letter logic").
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200917170212.92672-1-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
virtio-vsock was introduced after the release of VIRTIO 1.0
specifications, so it should be 'modern-only'.
This patch forces virtio version 1 as done for vhost-vsock-pci.
To avoid migration issues, we force virtio version 1 only when
legacy check is enabled in the new machine types (>= 5.1).
Cc: qemu-stable@nongnu.org
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200921122506.82515-5-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Commit 9b3a35ec82 ("virtio: verify that legacy support is not
accidentally on") added a safety check that requires to set
'disable-legacy=on' on vhost-user-vsock-pci device:
$ ./qemu-system-x86_64 ... \
-chardev socket,id=char0,reconnect=0,path=/tmp/vhost4.socket \
-device vhost-user-vsock-pci,chardev=char0
qemu-system-x86_64: -device vhost-user-vsock-pci,chardev=char0:
device is modern-only, use disable-legacy=on
virtio-vsock was introduced after the release of VIRTIO 1.0
specifications, so it should be 'modern-only'.
This patch forces virtio version 1 and removes the 'transitional_name'
property, as done for vhost-vsock-pci, removing the need to specify
'disable-legacy=on' on vhost-user-vsock-pci device.
Cc: qemu-stable@nongnu.org
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200921122506.82515-4-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Commit 9b3a35ec82 ("virtio: verify that legacy support is not
accidentally on") added a safety check that requires to set
'disable-legacy=on' on vhost-vsock-pci device:
$ ./qemu-system-x86_64 ... -device vhost-vsock-pci,guest-cid=5
qemu-system-x86_64: -device vhost-vsock-pci,guest-cid=5:
device is modern-only, use disable-legacy=on
virtio-vsock was introduced after the release of VIRTIO 1.0
specifications, so it should be 'modern-only'.
In addition Cornelia verified that forcing a legacy mode on
vhost-vsock-pci device using x86-64 host and s390x guest, so with
different endianness, produces strange behaviours.
This patch forces virtio version 1 and removes the 'transitional_name'
property removing the need to specify 'disable-legacy=on' on
vhost-vsock-pci device.
To avoid migration issues, we force virtio version 1 only when
legacy check is enabled in the new machine types (>= 5.1).
As the transitional device name is not commonly used, we do not
provide compatibility handling for it.
Cc: qemu-stable@nongnu.org
Reported-by: Qian Cai <caiqian@redhat.com>
Reported-by: Qinghua Cheng <qcheng@redhat.com>
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1868449
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200921122506.82515-3-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Commit 9b3a35ec82 ("virtio: verify that legacy support is not accidentally
on") added a check that returns an error if legacy support is on, but the
device does not support legacy.
Unfortunately some devices were wrongly declared legacy capable even if
they were not (e.g vhost-vsock).
To avoid migration issues, we add a virtio-device property
(x-disable-legacy-check) to skip the legacy error, printing a warning
instead, for machine types < 5.1.
Cc: qemu-stable@nongnu.org
Fixes: 9b3a35ec82 ("virtio: verify that legacy support is not accidentally on")
Suggested-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200921122506.82515-2-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
If there are less than 2 arguments in version_ge(), the second
'shift' prints this error:
../configure: line 232: shift: shift count out of range
As Eric suggested, we can use 'shift ${2:+2}' which works out to
'shift 2' if $2 is set, or 'shift' (implicitly shift 1) if $2
is not set.
This patch replaces both 'shift; shift' occurrences in version_ge()
with 'shift ${2:+2}'.
Suggested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200821203558.10338-1-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
libqemuutil.a build fails with this error:
/usr/bin/ld: libqemuutil.a(util_fdmon-io_uring.c.o): in function `get_sqe':
qemu/build/../util/fdmon-io_uring.c:83: undefined reference to `io_uring_get_sqe'
/usr/bin/ld: qemu/build/../util/fdmon-io_uring.c:92: undefined reference to `io_uring_submit'
/usr/bin/ld: qemu/build/../util/fdmon-io_uring.c:96: undefined reference to `io_uring_get_sqe'
/usr/bin/ld: libqemuutil.a(util_fdmon-io_uring.c.o): in function `fdmon_io_uring_wait':
qemu/build/../util/fdmon-io_uring.c:289: undefined reference to `io_uring_submit_and_wait'
/usr/bin/ld: libqemuutil.a(util_fdmon-io_uring.c.o): in function `fdmon_io_uring_setup':
qemu/build/../util/fdmon-io_uring.c:328: undefined reference to `io_uring_queue_init'
/usr/bin/ld: libqemuutil.a(util_fdmon-io_uring.c.o): in function `fdmon_io_uring_destroy':
qemu/build/../util/fdmon-io_uring.c:343: undefined reference to `io_uring_queue_exit'
collect2: error: ld returned 1 exit status
This patch fix the issue adding 'linux_io_uring' dependency for
fdmon-io_uring.c
Fixes: a81df1b68b ("libqemuutil, qapi, trace: convert to meson")
Cc: pbonzini@redhat.com
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200821154853.94379-1-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add the PCI version of vhost-user-vsock
Launch QEMU like this:
qemu -chardev socket,path=/tmp/vm.vsock,id=chr0 \
-device vhost-user-vsock-pci,chardev=chr0
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200522122512.87413-4-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch introduces a vhost-user device for vsock, using the
vhost-vsock-common parent class.
The vhost-user-vsock device can be used to implement the virtio-vsock
device emulation in user-space.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200522122512.87413-3-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch prepares the introduction of vhost-user-vsock, moving
the common code usable for both vhost-vsock and vhost-user-vsock
devices, in the new vhost-vsock-common parent class.
While moving the code, fixed checkpatch warnings about block comments.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20200522122512.87413-2-sgarzare@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
In qemu_luring_poll_cb() we are not using the cqe peeked from the
CQ ring. We are using io_uring_peek_cqe() only to see if there
are cqes ready, so we can replace it with io_uring_cq_ready().
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20200519134942.118178-1-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
As recently documented [1], io_uring_enter(2) syscall can return an
error (errno=EINTR) if the operation was interrupted by a delivery
of a signal before it could complete.
This should happen when IORING_ENTER_GETEVENTS flag is used, for
example during io_uring_submit_and_wait() or during io_uring_submit()
when IORING_SETUP_IOPOLL is enabled.
We shouldn't have this problem for now, but it's better to prevent it.
[1] https://github.com/axboe/liburing/commit/344355ec6619de8f4e64584c9736530b5346e4f4
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20200519133041.112138-1-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In order to reduce the memory footprint we map into memory
the initrd using g_mapped_file_new() instead of reading it.
In this way we can share the initrd pages between multiple
instances of QEMU.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20190724143105.307042-4-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In order to reduce the memory footprint we map into memory
the ELF to load using g_mapped_file_new_from_fd() instead of
reading each sections. In this way we can share the ELF pages
between multiple instances of QEMU.
Suggested-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20190724143105.307042-3-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This patch allows handling an ELF memory-mapped, taking care
the reference count of the GMappedFile* passed through
rom_add_elf_program().
In this case, the 'data' pointer is not heap-allocated, so
we cannot free it.
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20190724143105.307042-2-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Since the get_str_sep() function is no longer used in
net/net.c, we can remove it.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Use the glib function to split host address and port in
the parse_host_port() function.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
net_client_init() uses a variable length array to store the prefix
of 'ipv6-net' parameter (e.g. if ipv6-net=fec0::0/64, the prefix
is 'fec0::0').
This patch introduces g_strsplit() to split the 'ipv6-net' parameter,
so we can remove the variable length array.
Suggested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
If 'ipv6-prefixlen' is not a number, the current behaviour
produces an assertion failure:
$ qemu-system-x86_64 -net user,ipv6-net=feca::0/a
qemu-system-x86_64: qemu/util/qemu-option.c:1175: qemu_opts_foreach:
Assertion `!errp || !*errp' failed.
Aborted (core dumped)
This patch fixes it, jumping to the end of the function when
'ipv6-prefixlen' is not a number, and printing the more friendly
message:
$ qemu-system-x86_64 -net user,ipv6-net=feca::0/a
qemu-system-x86_64: Parameter 'ipv6-prefixlen' expects a number
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
RBD APIs don't allow us to write more than the size set with
rbd_create() or rbd_resize().
In order to support growing images (eg. qcow2), we resize the
image before write operations that exceed the current size.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190509145927.293369-1-sgarzare@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Add missing 'falloc' among the allowed values of 'preallocation'
option; show it and 'full' only when they are supported.
('falloc' is supported if defined CONFIG_GLUSTERFS_FALLOCATE,
'full' is supported if defined CONFIG_GLUSTERFS_ZEROFILL)
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190524075848.23781-4-sgarzare@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Using IEC binary prefixes in order to make the code more readable.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Using IEC binary prefixes in order to make the code more readable,
with the exception of DEFAULT_LOG_SIZE because it's passed to
stringify().
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
IEC binary prefixes are already defined in "qemu/units.h",
so we can remove redundant definitions in "block/vhdx.h".
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Some multiboot images could be in the ELF format. In the current
implementation QEMU fails because we try to load these images
as a PVH image.
In order to fix this issue, we should try multiboot first (we
already check the multiboot magic header before to load it).
If it is not a multiboot image, we can try the PVH loader.
Fixes: ab969087da ("pvh: Boot uncompressed kernel using direct boot ABI", 2019-01-15)
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20190214180216.246707-1-sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This patch adds the support of DISCARD and WRITE_ZEROES commands,
that have been introduced in the virtio-blk protocol to have
better performance when using SSD backend.
We support only one segment per request since multiple segments
are not widely used and there are no userspace APIs that allow
applications to submit multiple segments in a single call.
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-7-sgarzare@redhat.com
Message-Id: <20190221103314.58500-7-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In order to use VirtIOFeature also in other virtio devices, we move
its declaration and the endof() macro (renamed in virtio_endof())
in virtio.h.
We add virtio_feature_get_config_size() function to iterate the array
of VirtIOFeature and to return the config size depending on the
features enabled. (as virtio_net_set_config_size() did)
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-5-sgarzare@redhat.com
Message-Id: <20190221103314.58500-5-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Since configurable features for virtio-blk are growing, this patch
adds host_features field in the struct VirtIOBlock. (as in virtio-net)
In this way, we can avoid to add new fields for new properties and
we can directly set VIRTIO_BLK_F* flags in the host_features.
We update "config-wce" and "scsi" property definition to use the new
host_features field without change the behaviour.
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20190221103314.58500-3-sgarzare@redhat.com
Message-Id: <20190221103314.58500-3-sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In order to avoid migration issues, we enable PVH only for
machine type >= 4.0
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The new pvh.bin option rom can be used with SeaBIOS to boot
uncompressed kernel using the x86/HVM direct boot ABI.
pvh.S contains the entry point of the option rom. It runs
in real mode, loads the e820 table querying the BIOS, and
then it switches to 32bit protected mode and jumps to the
pvh_load_kernel() written in pvh_main.c.
pvh_load_kernel() loads the cmdline and kernel entry_point
using fw_cfg, then it looks for RSDP, fills the
hvm_start_info required by x86/HVM ABI, and finally jumps
to the kernel entry_point.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
In order to allow other option roms to use these common
useful functions and definitions, this patch put them
in two new C header files called optrom.h and
optrom_fw_cfg.h. We also add useful out*() in*()
functions for different size, and new fw_cfg functions
to use when DMA feature is not available.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
VirtIOBlock *s is already defined and initialized with req->dev
on top of virtio_blk_handle_request(), so we can remove it from
the code block of VIRTIO_BLK_T_GET_ID case.
Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20190130095231.42081-1-sgarzare@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Stefano Garzarella