Implement the new virtio sockets device for host<->guest communication
using the Sockets API. Most of the work is done in a vhost kernel
driver so that virtio-vsock can hook into the AF_VSOCK address family.
The QEMU vhost-vsock device handles configuration and live migration
while the rx/tx happens in the vhost_vsock.ko Linux kernel driver.
The vsock device must be given a CID (host-wide unique address):
# qemu -device vhost-vsock-pci,id=vhost-vsock-pci0,guest-cid=3 ...
For more information see:
http://qemu-project.org/Features/VirtioVsock
[Endianness fixes and virtio-ccw support by Claudio Imbrenda
<imbrenda@linux.vnet.ibm.com>]
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
[mst: rebase to master]
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Use kvm acceleration if available.
Disable kernel-irqchip and use qemu64 cpu
for both kvm and tcg cases.
Using kvm acceleration saves about a second
and disabling kernel-irqchip has no visible
performance impact.
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
!legacy && !modern is shorter than !(legacy || modern).
I also perfer this (less ()s) as a matter of taste.
Cc: Greg Kurz <gkurz@linux.vnet.ibm.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
We used to set vring call fd unconditionally even if guest driver does
not use MSIX for this vritqueue at all. This will cause lots of
unnecessary userspace access and other checks for drivers does not use
interrupt at all (e.g virtio-net pmd). So check and clean vring call
fd if guest does not use any vector for this virtqueue at
all.
Perf diffs (on rx) shows lots of cpus wasted on vhost_signal() were saved:
#
28.12% -27.82% [vhost] [k] vhost_signal
14.44% -1.69% [kernel.vmlinux] [k] copy_user_generic_string
7.05% +1.53% [kernel.vmlinux] [k] __free_page_frag
6.51% +5.53% [vhost] [k] vhost_get_vq_desc
...
Pktgen tests shows 15.8% improvement on rx pps and 6.5% on tx pps.
Before: RX 2.08Mpps TX 1.35Mpps
After: RX 2.41Mpps TX 1.44Mpps
Signed-off-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Without presuming if we got there because of a user mistake or some
more subtle bug in the tooling, it really does not make sense to
implement a non-functional device.
Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
The statistics virtqueue is not migrated properly because virtio-balloon
does not include s->stats_vq_elem in the migration stream.
After migration the statistics virtqueue hangs because the host never
completes the last element (s->stats_vq_elem is NULL on the destination
QEMU). Therefore the guest never submits new elements and the virtqueue
is hung.
Instead of changing the migration stream format in an incompatible way,
detect the migration case and rewind the virtqueue so the last element
can be completed.
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Roman Kagan <rkagan@virtuozzo.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Suggested-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Ladi Prosek <lprosek@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
virtqueue_discard() requires a VirtQueueElement but virtio-balloon does
not migrate its in-use element. Introduce a new function that is
similar to virtqueue_discard() but doesn't require a VirtQueueElement.
This will allow virtio-balloon to access element again after migration
with the usual proviso that the guest may have modified the vring since
last time.
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Roman Kagan <rkagan@virtuozzo.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Ladi Prosek <lprosek@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
The one pending element is being freed but not discarded on device
reset, which causes svq->inuse to creep up, eventually hitting the
"Virtqueue size exceeded" error.
Properly discarding the element on device reset makes sure that its
buffers are unmapped and the inuse counter stays balanced.
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Roman Kagan <rkagan@virtuozzo.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Ladi Prosek <lprosek@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
vq->inuse must be zeroed upon device reset like most other virtqueue
fields.
In theory, virtio_reset() just needs assert(vq->inuse == 0) since
devices must clean up in-flight requests during reset (requests cannot
not be leaked!).
In practice, it is difficult to achieve vq->inuse == 0 across reset
because balloon, blk, 9p, etc implement various different strategies for
cleaning up requests. Most devices call g_free(elem) directly without
telling virtio.c that the VirtQueueElement is cleaned up. Therefore
vq->inuse is not decremented during reset.
This patch zeroes vq->inuse and trusts that devices are not leaking
VirtQueueElements across reset.
I will send a follow-up series that refactors request life-cycle across
all devices and converts vq->inuse = 0 into assert(vq->inuse == 0) but
this more invasive approach is not appropriate for stable trees.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Cc: qemu-stable <qemu-stable@nongnu.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Ladi Prosek <lprosek@redhat.com>
Currently each VQ Notification Virtio Capability is allocated
on a different page. The idea is to enable split drivers within
guests, however there are no known plans to do that.
The allocation will result in a 8MB BAR, more than various
guest firmwares pre-allocates for PCI Bridges hotplug process.
Reserve 4 bytes per VQ by default and add a new parameter
"page-per-vq" to be used with split drivers.
Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Some software algorithms are based on the hardware's cache info, for example,
for x86 linux kernel, when cpu1 want to wakeup a task on cpu2, cpu1 will trigger
a resched IPI and told cpu2 to do the wakeup if they don't share low level
cache. Oppositely, cpu1 will access cpu2's runqueue directly if they share llc.
The relevant linux-kernel code as bellow:
static void ttwu_queue(struct task_struct *p, int cpu)
{
struct rq *rq = cpu_rq(cpu);
......
if (... && !cpus_share_cache(smp_processor_id(), cpu)) {
......
ttwu_queue_remote(p, cpu); /* will trigger RES IPI */
return;
}
......
ttwu_do_activate(rq, p, 0); /* access target's rq directly */
......
}
In real hardware, the cpus on the same socket share L3 cache, so one won't
trigger a resched IPIs when wakeup a task on others. But QEMU doesn't present a
virtual L3 cache info for VM, then the linux guest will trigger lots of RES IPIs
under some workloads even if the virtual cpus belongs to the same virtual socket.
For KVM, there will be lots of vmexit due to guest send IPIs.
The workload is a SAP HANA's testsuite, we run it one round(about 40 minuates)
and observe the (Suse11sp3)Guest's amounts of RES IPIs which triggering during
the period:
No-L3 With-L3(applied this patch)
cpu0: 363890 44582
cpu1: 373405 43109
cpu2: 340783 43797
cpu3: 333854 43409
cpu4: 327170 40038
cpu5: 325491 39922
cpu6: 319129 42391
cpu7: 306480 41035
cpu8: 161139 32188
cpu9: 164649 31024
cpu10: 149823 30398
cpu11: 149823 32455
cpu12: 164830 35143
cpu13: 172269 35805
cpu14: 179979 33898
cpu15: 194505 32754
avg: 268963.6 40129.8
The VM's topology is "1*socket 8*cores 2*threads".
After present virtual L3 cache info for VM, the amounts of RES IPIs in guest
reduce 85%.
For KVM, vcpus send IPIs will cause vmexit which is expensive, so it can cause
severe performance degradation. We had tested the overall system performance if
vcpus actually run on sparate physical socket. With L3 cache, the performance
improves 7.2%~33.1%(avg:15.7%).
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This will used by the next patch.
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
PIO MR registration should use size from the correct notify struct.
Doesn't affect any visible behaviour because the field values are the
same (both are 4).
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
event_notifier_init() can fail in real life, for example when there
are not enough open file handles available (EMFILE) when using a lot
of devices. So instead of leaving the average user with a cryptic
error number only, print out a proper error message with strerror()
instead, so that the user has a better way to figure out what is
going on and that using "ulimit -n" might help here for example.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Version: GnuPG v2
iQEcBAABCAAGBQJX0kCLAAoJEMo1YkxqkXHGt2oH/j9t5iSsGzbjOAmhLKW7t/hS
r1+B6gEvyRVzFBI2hCwcauHXKMM4v8qEMeZ9n0QeMk1BSjl1PGGg7BRrp2MnotC7
7ELT/CqLbmXqILZp+x2rVEF+y1f5GcQdgA3bQCuQ6zM34b3Zuv1OnjuOoYVKHaeJ
YnZopSCgIzaNq0noovG76l0TTYddUEa3Nbt97XtVhTmie+1z+Fyup9gdfHLF427z
gUIhzGzfYLN8lL1eQX70+xUF7EpLwz/aySMsp63EaZU3DQ3m4EqrR9jH8xpbC/+k
XwwFwH/C0284pxgZStkeHlvMRChzrIvKPYcAPg7wYnILMh61LKe84w5kRwMKRBU=
=vq7v
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/famz/tags/docker-pull-request' into staging
# gpg: Signature made Fri 09 Sep 2016 05:54:35 BST
# gpg: using RSA key 0xCA35624C6A9171C6
# gpg: Good signature from "Fam Zheng <famz@redhat.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 5003 7CB7 9706 0F76 F021 AD56 CA35 624C 6A91 71C6
* remotes/famz/tags/docker-pull-request:
docker: silence debootstrap when --quiet is given
docker: build debootstrap after cloning
docker: make sure debootstrap is at least 1.0.67
docker: print warning if EXECUTABLE is not set when building debootstrap image
docker: debian-bootstrap.pre: print helpful message if DEB_ARCH/DEB_TYPE unset
docker: debian-bootstrap.pre: print error messages to stderr
docker: avoid dependency on 'realpath' package
docker.py: don't hang on large docker output
docker: Add a glib2-2.22 image
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Some tests use the qtest protocol "memset" command with a zero
size, expecting it to do nothing. However in the current code this
will result in calling memset() with a NULL pointer, which is
undefined behaviour. Detect and specially handle zero sizes to
avoid this.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1470393800-7882-1-git-send-email-peter.maydell@linaro.org
A few strings are allocated and never freed.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
In all cases, even when the dict doesn't contain 'ram', the qmp response
must be unref.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The path is allocated and should be freed.
The qmp response should be unref, but then 'machine' must be duplicated.
Use a destroy function for the PCTestData.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Allows one to specify a destroy function for the test data.
Add a fallback using glib g_test_add_vtable() internal function, whose
signature changed over time. Tested with glib 2.22, 2.26 and 2.48, which
according to git log should be enough to cover all variations.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Simplify a bit the code by using g_strdup_printf() and store it in a
non-const value so casting is no longer needed, and ownership is
clearer.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Free the timer allocated in instance_init.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Andrew Baumann <Andrew.Baumann@microsoft.com>
Further cleanup would need to call qemu_free_irq() at the appropriate
time, but for now this silences ASAN about direct leaks.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
The free_ranges array is used as a temporary pointer array, the segment
should still be freed, however, it shouldn't free the elements themself.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Tested-by: Marcel Apfelbaum <marcel@redhat.com>
Reviewed-by: Marcel Apfelbaum <marcel@redhat.com>
machine_class_base_init() member name is allocated by
machine_class_base_init(), but not freed by
machine_class_finalize(). Simply freeing there doesn't work,
because DEFINE_PC_MACHINE() overwrites it with a literal string.
Fix DEFINE_PC_MACHINE() not to overwrite it, and add the missing
free to machine_class_finalize().
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
qemu_irq is already a pointer, no need to have an extra pointer level.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The isa_register_portio_list() function allocates ioports
data/state. Let's keep the reference to this data on some owner. This
isn't enough to fix leaks, but at least, ASAN stops complaining of
direct leaks. Further cleanup would require calling
portio_list_del/destroy().
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Free the list returned by visit_type_intList().
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Free the config blacklist list, not just the elements. Do it so in the
more appropriate function config_free().
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Those functions are only available since glib 2.28.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
If we silence docker when --quiet is given, we should also silence the
.pre script (i.e. debootstrap).
Only discards stdout, so some diagnostics (e.g. from git clone) are
still printed. Most of the verbose output is gone however and this way
we still have a chance to see error messages.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-9-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
When using the git version of debootstrap (because no usable version
of debootstrap was installed on the host), we need to run 'make' so
that devices.tar.gz gets built. Otherwise the first debootstrap stage
will fail without printing any error message.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-8-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
debootstrap prior to 1.0.67 generated an empty sources.list during
foreign bootstraps (Debian#732255 [1]). Fall back to the git checkout
if the installed debootstrap version is too old.
[1] https://bugs.debian.org/732255
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-7-git-send-email-silbe@linux.vnet.ibm.com>
[Update 'sort -C' to 'sorc -c &>/dev/null' - Fam]
Signed-off-by: Fam Zheng <famz@redhat.com>
Building the debian-debootstrap image will usually fail if EXECUTABLE
isn't set (when using the Makefile). Warn the user in this case so
they know why it's failing.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-6-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
The debian-bootstrap image doesn't choose a default architecture and
distribution version, instead the user has to set both DEB_ARCH and
DEB_TYPE in the environment. Print a reasonably helpful message if
either of them isn't set instead of complaining about "qemu-" being
missing or erroring out because we cannot cd to the mirror URL.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-5-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Send error messages where they belong so they're seen even if stdout
is redirected to /dev/null.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-4-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
The 'realpath' executable is shipped in a separate package that isn't
installed by default on some distros.
We already use 'readlink -e' (provided by GNU coreutils) in some other
part of the code, so let's settle for that instead.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Message-Id: <1473192351-601-3-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Unlike Popen.communicate(), subprocess.call() doesn't read from the
stdout file descriptor. If the child process produces more output than
fits into the pipe buffer, it will block indefinitely.
If we don't intend to consume the output, just send it straight to
/dev/null to avoid this issue.
Signed-off-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
Reviewed-by: Janosch Frank <frankja@linux.vnet.ibm.com>
Message-Id: <1473192351-601-2-git-send-email-silbe@linux.vnet.ibm.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
