Right now it could be used on destination side to
enable vCPU blocktime calculation for postcopy live migration.
vCPU blocktime - it's time since vCPU thread was put into
interruptible sleep, till memory page was copied and thread awake.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Alexey Perevalov <a.perevalov@samsung.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Since commit 3a38429748 ("Add a "no HPT" encoding to HTAB migration stream")
the HTAB migration stream contains a header set to "-1", meaning there
is no HPT. Teach analyze-migration.py to ignore the section in this case.
Without this fix, the script fails with a dump from a POWER9 guest:
Traceback (most recent call last):
File "./qemu/scripts/analyze-migration.py", line 602, in <module>
dump.read(dump_memory = args.memory)
File "./qemu/scripts/analyze-migration.py", line 539, in read
section.read()
File "./qemu/scripts/analyze-migration.py", line 250, in read
self.file.readvar(n_valid * self.HASH_PTE_SIZE_64)
File "./qemu/scripts/analyze-migration.py", line 64, in readvar
raise Exception("Unexpected end of %s at 0x%x" % (self.filename, self.file.tell()))
Exception: Unexpected end of migrate.dump at 0x1d4763ba
Fixes: 3a38429748 ("Add a "no HPT" encoding to HTAB migration stream")
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Mostly just manual conversion with very minor fixes.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Kashyap Chamarthy <kchamart@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Otherwise, we can't use it after calling socket_start_incoming_migration
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Once there, do one thing for line
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
We use int for everything (int64_t), and then we check that value is
between 0 and 255. Change it to the valid types.
This change only happens for HMP. QMP always use bytes and similar.
Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Host: Mac OS 10.12.5
Compiler: Apple LLVM version 8.1.0 (clang-802.0.42)
slirp/ip6_icmp.c:80:38: warning: taking address of packed member 'ip_src' of class or
structure 'ip6' may result in an unaligned pointer value
[-Waddress-of-packed-member]
IN6_IS_ADDR_UNSPECIFIED(&ip->ip_src)) {
^~~~~~~~~~
/usr/include/netinet6/in6.h:238:42: note: expanded from macro 'IN6_IS_ADDR_UNSPECIFIED'
((*(const __uint32_t *)(const void *)(&(a)->s6_addr[0]) == 0) && \
^
Reported-by: John Arbuckle <programmingkidx@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
When SDL2 windows change focus while a key is held, the window that
receives the focus also receives a new KeyDown event, without an
autorepeat flag. This means that if a WM places the qemu console
over the main window after Ctrl-Alt-2, the console closes immediately
after opening. Then, the main window receives the KeyDown event again
and the whole process repeats.
This patch makes the SDL2 UI ignore the KeyDown events on a window that
just received the focus, if the GUI modifier was held. The ignore flag
is reset on a first KeyUp event. This effectively works around the issue
above.
Signed-off-by: Jindrich Makovicka <makovick@gmail.com>
Message-Id: <20171117112258.5888-4-makovick@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Some older versions of gcc complain if a typedef is defined twice:
target/xtensa/translate.c:81: error: redefinition of typedef 'DisasContext'
target/xtensa/cpu.h:339: note: previous declaration of 'DisasContext' was here
Remove the now-redundant typedef from the definition of the struct in
translate.c.
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1515762528-22818-1-git-send-email-peter.maydell@linaro.org
This fields points to an old interface that is no more
used in the current code.
Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Message-id: 20171122135625.16625-1-fziglio@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The GTK 3.0 release was made in Feb, 2011:
https://blog.gtk.org/2011/02/10/gtk-3-0-released/
That will soon be 7 years ago, which is enough time to consider
the 3.x series widely supported.
Thus we deprecate the GTK 2.x support, which will allow us to
delete it in the last release of 2018. By this time, GTK 3.x
will be almost 8 years old.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20171212113440.16483-1-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
If kbd_queue is not empty and queue_count >= queue_limit,
we should free evt.
Change-Id: Ieeacf90d5e7e370a40452ec79031912d8b864d83
Signed-off-by: linzhecheng <linzhecheng@huawei.com>
Message-id: 20171225023730.5512-1-linzhecheng@huawei.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
While the QIOChannel APIs for reading/writing data return ssize_t, with negative
value indicating an error, the VNC code passes this return value through the
vnc_client_io_error() method. This detects the error condition, disconnects the
client and returns 0 to indicate error. Thus all the VNC helper methods should
return size_t (unsigned), and misleading comments which refer to the possibility
of negative return values need fixing.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-14-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The VNC client throttling is quite subtle so will benefit from having trace
points available for live debugging.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-13-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The previous patches fix problems with throttling of forced framebuffer updates
and audio data capture that would cause the QEMU output buffer size to grow
without bound. Those fixes are graceful in that once the client catches up with
reading data from the server, everything continues operating normally.
There is some data which the server sends to the client that is impractical to
throttle. Specifically there are various pseudo framebuffer update encodings to
inform the client of things like desktop resizes, pointer changes, audio
playback start/stop, LED state and so on. These generally only involve sending
a very small amount of data to the client, but a malicious guest might be able
to do things that trigger these changes at a very high rate. Throttling them is
not practical as missed or delayed events would cause broken behaviour for the
client.
This patch thus takes a more forceful approach of setting an absolute upper
bound on the amount of data we permit to be present in the output buffer at
any time. The previous patch set a threshold for throttling the output buffer
by allowing an amount of data equivalent to one complete framebuffer update and
one seconds worth of audio data. On top of this it allowed for one further
forced framebuffer update to be queued.
To be conservative, we thus take that throttling threshold and multiply it by
5 to form an absolute upper bound. If this bound is hit during vnc_write() we
forceably disconnect the client, refusing to queue further data. This limit is
high enough that it should never be hit unless a malicious client is trying to
exploit the sever, or the network is completely saturated preventing any sending
of data on the socket.
This completes the fix for CVE-2017-15124 started in the previous patches.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-12-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The VNC server must throttle data sent to the client to prevent the 'output'
buffer size growing without bound, if the client stops reading data off the
socket (either maliciously or due to stalled/slow network connection).
The current throttling is very crude because it simply checks whether the
output buffer offset is zero. This check is disabled if the client has requested
a forced update, because we want to send these as soon as possible.
As a result, the VNC client can cause QEMU to allocate arbitrary amounts of RAM.
They can first start something in the guest that triggers lots of framebuffer
updates eg play a youtube video. Then repeatedly send full framebuffer update
requests, but never read data back from the server. This can easily make QEMU's
VNC server send buffer consume 100MB of RAM per second, until the OOM killer
starts reaping processes (hopefully the rogue QEMU process, but it might pick
others...).
To address this we make the throttling more intelligent, so we can throttle
full updates. When we get a forced update request, we keep track of exactly how
much data we put on the output buffer. We will not process a subsequent forced
update request until this data has been fully sent on the wire. We always allow
one forced update request to be in flight, regardless of what data is queued
for incremental updates or audio data. The slight complication is that we do
not initially know how much data an update will send, as this is done in the
background by the VNC job thread. So we must track the fact that the job thread
has an update pending, and not process any further updates until this job is
has been completed & put data on the output buffer.
This unbounded memory growth affects all VNC server configurations supported by
QEMU, with no workaround possible. The mitigating factor is that it can only be
triggered by a client that has authenticated with the VNC server, and who is
able to trigger a large quantity of framebuffer updates or audio samples from
the guest OS. Mostly they'll just succeed in getting the OOM killer to kill
their own QEMU process, but its possible other processes can get taken out as
collateral damage.
This is a more general variant of the similar unbounded memory usage flaw in
the websockets server, that was previously assigned CVE-2017-15268, and fixed
in 2.11 by:
commit a7b20a8efa
Author: Daniel P. Berrange <berrange@redhat.com>
Date: Mon Oct 9 14:43:42 2017 +0100
io: monitor encoutput buffer size from websocket GSource
This new general memory usage flaw has been assigned CVE-2017-15124, and is
partially fixed by this patch.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-11-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The VNC server must throttle data sent to the client to prevent the 'output'
buffer size growing without bound, if the client stops reading data off the
socket (either maliciously or due to stalled/slow network connection).
The current throttling is very crude because it simply checks whether the
output buffer offset is zero. This check must be disabled if audio capture is
enabled, because when streaming audio the output buffer offset will rarely be
zero due to queued audio data, and so this would starve framebuffer updates.
As a result, the VNC client can cause QEMU to allocate arbitrary amounts of RAM.
They can first start something in the guest that triggers lots of framebuffer
updates eg play a youtube video. Then enable audio capture, and simply never
read data back from the server. This can easily make QEMU's VNC server send
buffer consume 100MB of RAM per second, until the OOM killer starts reaping
processes (hopefully the rogue QEMU process, but it might pick others...).
To address this we make the throttling more intelligent, so we can throttle
when audio capture is active too. To determine how to throttle incremental
updates or audio data, we calculate a size threshold. Normally the threshold is
the approximate number of bytes associated with a single complete framebuffer
update. ie width * height * bytes per pixel. We'll send incremental updates
until we hit this threshold, at which point we'll stop sending updates until
data has been written to the wire, causing the output buffer offset to fall
back below the threshold.
If audio capture is enabled, we increase the size of the threshold to also
allow for upto 1 seconds worth of audio data samples. ie nchannels * bytes
per sample * frequency. This allows the output buffer to have a mixture of
incremental framebuffer updates and audio data queued, but once the threshold
is exceeded, audio data will be dropped and incremental updates will be
throttled.
This unbounded memory growth affects all VNC server configurations supported by
QEMU, with no workaround possible. The mitigating factor is that it can only be
triggered by a client that has authenticated with the VNC server, and who is
able to trigger a large quantity of framebuffer updates or audio samples from
the guest OS. Mostly they'll just succeed in getting the OOM killer to kill
their own QEMU process, but its possible other processes can get taken out as
collateral damage.
This is a more general variant of the similar unbounded memory usage flaw in
the websockets server, that was previously assigned CVE-2017-15268, and fixed
in 2.11 by:
commit a7b20a8efa
Author: Daniel P. Berrange <berrange@redhat.com>
Date: Mon Oct 9 14:43:42 2017 +0100
io: monitor encoutput buffer size from websocket GSource
This new general memory usage flaw has been assigned CVE-2017-15124, and is
partially fixed by this patch.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-10-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The logic for determining if it is possible to send an update to the client
will become more complicated shortly, so pull it out into a separate method
for easier extension later.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-9-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
According to the RFB protocol, a client sends one or more framebuffer update
requests to the server. The server can reply with a single framebuffer update
response, that covers all previously received requests. Once the client has
read this update from the server, it may send further framebuffer update
requests to monitor future changes. The client is free to delay sending the
framebuffer update request if it needs to throttle the amount of data it is
reading from the server.
The QEMU VNC server, however, has never correctly handled the framebuffer
update requests. Once QEMU has received an update request, it will continue to
send client updates forever, even if the client hasn't asked for further
updates. This prevents the client from throttling back data it gets from the
server. This change fixes the flawed logic such that after a set of updates are
sent out, QEMU waits for a further update request before sending more data.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-8-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Currently the VNC servers tracks whether a client has requested an incremental
or forced update with two boolean flags. There are only really 3 distinct
states to track, so create an enum to more accurately reflect permitted states.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-7-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
When we encode data for writing with SASL, we encode the entire pending output
buffer. The subsequent write, however, may not be able to send the full encoded
data in one go though, particularly with a slow network. So we delay setting the
output buffer offset back to zero until all the SASL encoded data is sent.
Between encoding the data and completing sending of the SASL encoded data,
however, more data might have been placed on the pending output buffer. So it
is not valid to set offset back to zero. Instead we must keep track of how much
data we consumed during encoding and subtract only that amount.
With the current bug we would be throwing away some pending data without having
sent it at all. By sheer luck this did not previously cause any serious problem
because appending data to the send buffer is always an atomic action, so we
only ever throw away complete RFB protocol messages. In the case of frame buffer
updates we'd catch up fairly quickly, so no obvious problem was visible.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-6-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The vnc_update_client() method checks the 'has_dirty' flag to see if there are
dirty regions that are pending to send to the client. Regardless of this flag,
if a forced update is requested, updates must be sent. For unknown reasons
though, the code also tries to sent updates if audio capture is enabled. This
makes no sense as audio capture state does not impact framebuffer contents, so
this check is removed.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-5-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Now that previous dead / unreachable code has been removed, we can simplify
the indentation in the vnc_client_update method.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-4-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
A previous commit:
commit 5a8be0f73d
Author: Gerd Hoffmann <kraxel@redhat.com>
Date: Wed Jul 13 12:21:20 2016 +0200
vnc: make sure we finish disconnect
Added a check for vs->disconnecting at the very start of the
vnc_update_client method. This means that the very next "if"
statement check for !vs->disconnecting always evaluates true,
and is thus redundant. This in turn means the vs->disconnecting
check at the very end of the method never evaluates true, and
is thus unreachable code.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-3-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
There is only one caller of vnc_update_client and that always passes false
for the 'sync' parameter.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20171218191228.31018-2-berrange@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add a new test with --enable-debug using clang/asan/ubsan, remove
--enable-debug from test-clang & test-mingw.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20180104160523.22995-7-marcandre.lureau@redhat.com>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When --enable-debug is turned on, configure doesn't set -O level, and
uses default compiler -O0 level, which is slow.
Instead, use -Og if supported by the compiler (optimize debugging
experience), or -O1 (keeps code somewhat debuggable and works around
compiler bugs).
Unfortunately, gcc has many false-positive maybe-uninitialized
errors with Og and O1 (f27 gcc 7.2.1 20170915):
/home/elmarco/src/qemu/hw/ipmi/isa_ipmi_kcs.c: In function ‘ipmi_kcs_ioport_read’:
/home/elmarco/src/qemu/hw/ipmi/isa_ipmi_kcs.c:279:12: error: ‘ret’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
return ret;
^~~
cc1: all warnings being treated as errors
make: *** [/home/elmarco/src/qemu/rules.mak:66: hw/ipmi/isa_ipmi_kcs.o] Error 1
make: *** Waiting for unfinished jobs....
/home/elmarco/src/qemu/hw/ide/ahci.c: In function ‘ahci_populate_sglist’:
/home/elmarco/src/qemu/hw/ide/ahci.c:903:58: error: ‘tbl_entry_size’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
if ((off_idx == -1) || (off_pos < 0) || (off_pos > tbl_entry_size)) {
~~~~~~~~~^~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make: *** [/home/elmarco/src/qemu/rules.mak:66: hw/ide/ahci.o] Error 1
/home/elmarco/src/qemu/hw/display/qxl.c: In function ‘qxl_add_memslot’:
/home/elmarco/src/qemu/hw/display/qxl.c:1397:52: error: ‘pci_start’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
memslot.virt_end = virt_start + (guest_end - pci_start);
~~~~~~~~~~~~~^~~~~~~~~~~~
/home/elmarco/src/qemu/hw/display/qxl.c:1389:9: error: ‘pci_region’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
qxl_set_guest_bug(d, "%s: pci_region = %d", __func__, pci_region);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
There seems to be a long list of related bugs in upstream GCC, some of
them are being fixed very recently:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24639
For now, let's workaround it by using Wno-maybe-uninitialized (gcc-only).
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20180104160523.22995-5-marcandre.lureau@redhat.com>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Move generic make flags in MAKEFLAGS (SUBDIR_MAKEFLAGS is more qemu specific).
Use --quiet to silence make 'is up to date' message.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20180104160523.22995-3-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
When linking qemu-ga under some configuration (when gthread-2.0.pc
doesn't have -pthread, as happening atm with meson build), you may
have this linking issue:
/usr/bin/ld: libqemuutil.a(qemu-thread-posix.o): undefined reference to symbol 'pthread_setname_np@@GLIBC_2.12'
/usr/lib64/libpthread.so.0: error adding symbols: DSO missing from command line
Make sure qemu-ga links with the pthread library, by adding correct
flags to libs_qga.
This is really a QEMU bug, because it's QEMU code that's using pthread
functions, and so we must explicitly link against pthreads. The bug
was just masked by the fact that often some pkg-config or another for
one of our dependencies will add -pthread to the link line anyway.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20180104160523.22995-2-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
It's a replacement of g_timeout_add[_seconds]() for chardevs. Chardevs
now can have dedicated gcontext, we should always bind chardev tasks
onto those gcontext rather than the default main context. Since there
are quite a few of g_timeout_add[_seconds]() callers, a new function
qemu_chr_timeout_add_ms() is introduced.
One thing to mention is that, terminal3270 is still always running on
main gcontext. However let's convert that as well since it's still part
of chardev codes and in case one day we'll miss that when we move it out
of main gcontext too.
Also, convert all the timers from GSource tags into GSource pointers.
Gsource tag IDs and g_source_remove()s can only work with default
gcontext, while now these GSources can logically be attached to other
contexts. So let's use explicit g_source_destroy() plus another
g_source_unref() to remove a timer.
Note: when in the timer handler, we don't need the g_source_destroy()
any more since that'll be done automatically if the timer handler
returns false (and that's what all the current handlers do).
Yet another note: in pty_chr_rearm_timer() we take special care for
ms=1000. This patch merged the two cases into one.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180104141835.17987-4-peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The idle task will be attached to main gcontext even if the chardev
backend is running in another gcontext. Fix the only caller by
extending the g_idle_add() logic into the more powerful
g_source_attach(). It's basically g_idle_add_full() implementation, but
with the chardev's gcontext passed in.
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180104141835.17987-3-peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
In commit 6bbb6c0644 ("chardev: use per-dev context for
io_add_watch_poll", 2017-09-22) all the chardev watches are converted to
use per-chardev gcontext to support chardev to be run outside default
main thread. However that's still missing one call from the frontend
code. Touch that up.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20180104141835.17987-2-peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Certain PMU-related MSRs are not supported for CPUs with PMU
architecture below version 2. KVM rejects any access to them (see
intel_is_valid_msr_idx routine in KVM), and QEMU fails on the following
assertion:
kvm_put_msrs: Assertion `ret == cpu->kvm_msr_buf->nmsrs' failed.
QEMU also could fail if KVM exposes less fixed counters then 3. It could
happen if host system run inside another hypervisor, which is tweaking
PMU-related CPUID. To prevent possible fail, number of fixed counters now is
obtained in the same way as number of GP counters.
Reviewed-by: Roman Kagan <rkagan@virtuozzo.com>
Signed-off-by: Jan Dakinevich <jan.dakinevich@virtuozzo.com>
Message-Id: <1514383466-7257-1-git-send-email-jan.dakinevich@virtuozzo.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
HPET saves its state by calculating the current time and recovers timer
offset using this calculated value. But these calculations include
divisions and multiplications. Therefore the timer state cannot be recovered
precise enough.
This patch introduces saving of the original value of the offset to
preserve the determinism of the timer.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Signed-off-by: Maria Klimushenkova <maria.klimushenkova@ispras.ru>
Reviewed-by: Juan Quintela <quintela@redhat.com>
--
v3: Added compat property for correct migration.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
