Otherwise the guest can abuse that control to cause e.g. PCIe
Unsupported Request responses (by disabling memory and/or I/O decoding
and subsequently causing [CPU side] accesses to the respective address
ranges), which (depending on system configuration) may be fatal to the
host.
This is CVE-2015-2756 / XSA-126.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Message-id: alpine.DEB.2.02.1503311510300.7690@kaball.uk.xensource.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
libxseg has changed license to GPLv3. QEMU includes GPL "v2 only" code
which is not compatible with GPLv3. This means the resulting binaries
may not be redistributable!
Disable Archipelago (libxseg) by default to prevent accidental license
violations. Also warn if linking against libxseg is enabled to remind
the user.
Note that this commit does not constitute any advice about software
licensing. If you have doubts you should consult a lawyer.
Cc: Chrysostomos Nanakos <cnanakos@grnet.gr>
Suggested-by: Kevin Wolf <kwolf@redhat.com>
Reported-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Message-id: 1428587538-8765-1-git-send-email-stefanha@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Commit 951c6300f7 out-of-lined the 32-bit-host versions of
tcg_gen_{ld,st}_i64, but in the process it inadvertently changed
an #ifdef HOST_WORDS_BIGENDIAN to #ifdef TCG_TARGET_WORDS_BIGENDIAN.
Since the latter doesn't get defined anywhere this meant we always
took the "LE host" codepath, and stored the two halves of the value
in the wrong order on BE hosts. This typically breaks any 64-bit
guest on a 32-bit BE host completely, and will have possibly more
subtle effects even for 32-bit guests.
Switch the ifdef back to HOST_WORDS_BIGENDIAN.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Tested-by: Andreas Färber <afaerber@suse.de>
Message-id: 1428523029-13620-1-git-send-email-peter.maydell@linaro.org
newer libiscsi versions may return zero events from iscsi_which_events.
In this case iscsi_service will return immediately without any progress.
To avoid busy waiting for iscsi_which_events to change we deregister all
read and write handlers in this case and schedule a timer to periodically
check iscsi_which_events for changed events.
Next libiscsi version will introduce async reconnects and zero events
are returned while libiscsi is waiting for a reconnect retry.
Signed-off-by: Peter Lieven <pl@kamp.de>
Message-id: 1428437295-29577-1-git-send-email-pl@kamp.de
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
There are two problems with memory barriers in async.c. The fix is
to use atomic_xchg in order to achieve sequential consistency between
the scheduling of a bottom half and the corresponding execution.
First, if bh->scheduled is already 1 in qemu_bh_schedule, QEMU does
not execute a memory barrier to order any writes needed by the callback
before the read of bh->scheduled. If the other side sees req->state as
THREAD_ACTIVE, the callback is not invoked and you get deadlock.
Second, the memory barrier in aio_bh_poll is too weak. Without this
patch, it is possible that bh->scheduled = 0 is not "published" until
after the callback has returned. Another thread wants to schedule the
bottom half, but it sees bh->scheduled = 1 and does nothing. This causes
a lost wakeup. The memory barrier should have been changed to smp_mb()
in commit 924fe12 (aio: fix qemu_bh_schedule() bh->ctx race condition,
2014-06-03) together with qemu_bh_schedule()'s. Guess who reviewed
that patch?
Both of these involve a store and a load, so they are reproducible on
x86_64 as well. It is however much easier on aarch64, where the
libguestfs test suite triggers the bug fairly easily. Even there the
failure can go away or appear depending on compiler optimization level,
tracing options, or even kernel debugging options.
Paul Leveille however reported how to trigger the problem within 15
minutes on x86_64 as well. His (untested) recipe, reproduced here
for reference, is the following:
1) Qcow2 (or 3) is critical – raw files alone seem to avoid the problem.
2) Use “cache=directsync” rather than the default of
“cache=none” to make it happen easier.
3) Use a server with a write-back RAID controller to allow for rapid
IO rates.
4) Run a random-access load that (mostly) writes chunks to various
files on the virtual block device.
a. I use ‘diskload.exe c:25’, a Microsoft HCT load
generator, on Windows VMs.
b. Iometer can probably be configured to generate a similar load.
5) Run multiple VMs in parallel, against the same storage device,
to shake the failure out sooner.
6) IvyBridge and Haswell processors for certain; not sure about others.
A similar patch survived over 12 hours of testing, where an unpatched
QEMU would fail within 15 minutes.
This bug is, most likely, also the cause of failures in the libguestfs
testsuite on AArch64.
Thanks to Laszlo Ersek for initially reporting this bug, to Stefan
Hajnoczi for suggesting closer examination of qemu_bh_schedule, and to
Paul for providing test input and a prototype patch.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Reported-by: Paul Leveille <Paul.Leveille@stratus.com>
Reported-by: John Snow <jsnow@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1428419779-26062-1-git-send-email-pbonzini@redhat.com
Suggested-by: Paul Leveille <Paul.Leveille@stratus.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Commit 0b183fc871:"memory: move mem_path handling to
memory_region_allocate_system_memory" split memory_region_init_ram and
memory_region_init_ram_from_file. Also it moved mem-path handling a step
up from memory_region_init_ram to memory_region_allocate_system_memory.
Therefore for any board that uses memory_region_init_ram directly,
-mem-path is not supported.
Fix this by replacing memory_region_init_ram with
memory_region_allocate_system_memory.
Signed-off-by: Dirk Mueller <dmueller@suse.com>
Message-id: CAL5wTH4UHYKpJF=dLJfFzxpufjY189chnCow47-ySuLf8GLbug@mail.gmail.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
After qemu_iovec_destroy, the QEMUIOVector's size is zeroed and
the zero size ultimately is used to compute virtqueue_push's len
argument. Therefore, reads from virtio-blk devices did not
migrate their results correctly. (Writes were okay).
Save the size in virtio_blk_handle_request, and use it when the request
is completed.
Based on a patch by Wen Congyang.
Signed-off-by: Wen Congyang <wency@cn.fujitsu.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Tested-by: Li Zhijian <lizhijian@cn.fujitsu.com>
Message-id: 1427997044-392-1-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In recent qemu versions, it is possible to override the backing file
name and format that is stored in the image file with values given at
runtime. In such cases, the temporary override could end up in the
image header if the qcow2 header was updated, while obviously correct
behaviour would be to leave the on-disk backing file path/format
unchanged.
Fix this and add a test case for it.
Reported-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1428411796-2852-1-git-send-email-kwolf@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
kvm_stat.{1,pod} started showing up as untracked files in my
directory, and I nearly accidentally merged them into a commit
with my usual habit of 'git add .'. Rather than spelling out
each such file, just ignore the entire pattern.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
With a mask value of 0x00400000, the result will never be 1.
This fixes a Coverity warning.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Probably a copy&paste bug. Fixing it helps identifying the device model
behind port 0x61.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: Gonglei <arei.gonglei@huawei.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Commit 89b516d8b9 ("glib: add
compatibility interface for g_get_monotonic_time()") aimed
at making qemu build with old glib versions. At least SLES11SP3,
however, contains a backport of g_get_monotonic_time() while
keeping the reported glib version at 2.22.
Let's work around this by a strategically placed #define.
Signed-off-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Message-id: 1427987865-433-2-git-send-email-cornelia.huck@de.ibm.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
virtio-blk fix, because Wen only posted a prototype and the changes
I made were pretty large. It definitely needs another pair of eyes
(but it is a 2.3 regression and a blocker).
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAABCAAGBQJVHUtcAAoJEL/70l94x66Dr1IIAIq+2i+79CoOW1VgoqdN9hBM
co7s8MzVwPzSPxMzicv94nAMBcPWfmoGzi8JJebxUmQDmZ7L33bOa64cbVgmQ5wU
gfJ+yea8TRkyDH/psv8jcvesdEgwRUnpOTCrEv4GD+haTHWZB4t6HW4KpolxbclJ
igs5MUxcH/cM//L+nyadr/ZuJvPy4sakquvnMqr5Zzr/xjwLwzbJIksoZthMdxW5
oZAY1g0T2f5sSpt0kl8lYid1njbTBX9z8ZkoftQSjRJFZ253vl+CxXFuJevke1hE
ayRhRKg5VfAyPY4txEK0+27E/Jo4X51y+6Et0N5yzrQsSZ1FeBtxzngm14gEQZA=
=CfCj
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
Another round of small fixes. I am not including the
virtio-blk fix, because Wen only posted a prototype and the changes
I made were pretty large. It definitely needs another pair of eyes
(but it is a 2.3 regression and a blocker).
# gpg: Signature made Thu Apr 2 14:59:56 2015 BST using RSA key ID 78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4 E2F7 7E15 100C CD36 69B1
# Subkey fingerprint: F133 3857 4B66 2389 866C 7682 BFFB D25F 78C7 AE83
* remotes/bonzini/tags/for-upstream:
Use $(MAKE) for recursive make
kvm-all: Sync dirty-bitmap from kvm before kvm destroy the corresponding dirty_bitmap
util/qemu-config: fix regression of qmp_query_command_line_options
target-i386: clear bsp bit when designating bsp
qga: fitering out -fstack-protector-strong
target-i386: save 64-bit CR3 in 64-bit SMM state save area
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
On BSDs "make" is typically BSD make, while "gmake" is GNU make.
Signed-off-by: Ed Maste <emaste@freebsd.org>
Message-Id: <1427911118-21905-1-git-send-email-emaste@freebsd.org>
[Fix $(INSTALLER) too as reported by Fam Zheng. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Sometimes, we destroy the dirty_bitmap in kvm_memory_slot before any sync action
occur, this bit in dirty_bitmap will be missed, and which will lead the corresponding
dirty pages to be missed in migration.
This usually happens when do migration during VM's Start-up or Reboot.
Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com>
[Use s->migration_log instead of exec.c's in_migration. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Commit 49d2e64 (machine: remove qemu_machine_opts global list)
made machine options specific to machine sub-type, leaving
the qemu_machine_opts desc array empty. Sadly this is the place
qmp_query_command_line_options is looking for supported options.
As a fix for for 2.3 the machine_qemu_opts (the generic ones)
are restored only for qemu-config scope.
We need to find a better fix for 2.4.
Reported-by: Tony Krowiak <akrowiak@linux.vnet.ibm.com>
Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
Message-Id: <1427906841-1576-1-git-send-email-marcel@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Since the BSP bit is writable on real hardware, during reset all the CPUs which
were not chosen to be the BSP should have their BSP bit cleared. This fix is
required for KVM to work correctly when it changes the BSP bit.
An additional fix is required for QEMU tcg to allow software to change the BSP
bit.
Signed-off-by: Nadav Amit <namit@cs.technion.ac.il>
Message-Id: <1427932716-11800-1-git-send-email-namit@cs.technion.ac.il>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
configure script may add -fstack-protector-strong option instead
of -fstack-protector-all, depending on availability ( see
commit 63678e17c ). Both options have to by filtered out for
qga-vss.dll, otherwise MinGW cross-compilation fails at linking
stage.
Signed-off-by: Joseph Hindin <jhindin@daynix.com>
Message-Id: <1427906337-20805-2-git-send-email-jhindin@daynix.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Device models aren't supposed to go on fishing expeditions for
backends. They should expose suitable properties for the user to set.
For onboard devices, board code sets them.
A number of sysbus devices pick up block backends in their init() /
instance_init() methods with drive_get_next() instead: sl-nand,
milkymist-memcard, pl181, generic-sdhci.
Likewise, a number of sysbus devices pick up character backends in
their init() / realize() methods with qemu_char_get_next_serial():
cadence_uart, digic-uart, etraxfs,serial, lm32-juart, lm32-uart,
milkymist-uart, pl011, stm32f2xx-usart, xlnx.xps-uartlite.
All these mistakes are already marked FIXME. See the commit that
added these FIXMEs for a more detailed explanation of what's wrong.
Fortunately, only machines ppce500 and pseries-* support -device with
sysbus devices, and none of the devices above is supported with these
machines.
Set cannot_instantiate_with_device_add_yet to preserve our luck.
Cc: Andrzej Zaborowski <balrogg@gmail.com>
Cc: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Cc: Antony Pavlov <antonynpavlov@gmail.com>
Cc: "Edgar E. Iglesias" <edgar.iglesias@gmail.com>
Cc: Michael Walle <michael@walle.cc>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Device models aren't supposed to go on fishing expeditions for
backends. They should expose suitable properties for the user to set.
For onboard devices, board code sets them.
"sdhci-pci" picks up its block backend in its realize() method with
drive_get_next() instead. Already marked FIXME. See the commit that
added the FIXME for a more detailed explanation of what's wrong.
We can't fix this in time for the release, but since the device is new
in 2.3, we can set cannot_instantiate_with_device_add_yet to disable
it before this mistake becomes ABI, and we have to support command
lines like
$ qemu -drive if=sd -drive if=sd,file=sd.img -device sdhci-pci -device sdhci-pci
forever.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
NICs defined with -net nic are for board initialization to wire up.
Board code examines nd_table[] to find them, and creates devices with
their qdev NIC properties set accordingly.
Except "allwinner-a10" goes on a fishing expedition for NIC
configuration instead of exposing the usual NIC properties for board
code to set: it uses nd_table[0] in its instance_init() method.
Picking up the first -net nic option's configuration that way works
when the device is created by board code. But it's inappropriate for
-device and device_add. Not only is it inconsistent with how the
other block device models work (they get their configuration from
properties "mac", "vlan", "netdev"), it breaks when nd_table[0] has
been picked up by the board or a previous -device / device_add
already.
Example:
$ qemu-system-arm -S -M cubieboard -device allwinner-a10
qemu-system-arm: -device allwinner-a10: Property 'allwinner-emac.netdev' can't take value 'hub0port0', it's in use
Aborted (core dumped)
It also breaks in other entertaining ways:
$ qemu-system-arm -M highbank -device allwinner-a10
qemu-system-arm: -device allwinner-a10: Unsupported NIC model: xgmac
$ qemu-system-arm -M highbank -net nic,model=allwinner-emac -device allwinner-a10
qemu-system-arm: Unsupported NIC model: allwinner-emac
Mark the mistake with a FIXME comment.
Cc: Li Guang <lig.fnst@cn.fujitsu.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Character devices defined with -serial and -parallel are for board
initialization to wire up. Board code examines serial_hds[] and
parallel_hds[] to find them, and creates devices with their qdev
chardev properties set accordingly.
Except a few devices go on a fishing expedition for a suitable backend
instead of exposing a chardev property for board code to set: they use
serial_hds[] (often via qemu_char_get_next_serial()) or parallel_hds[]
in their realize() or init() method to connect to a backend.
Picking up backends that way works when the devices are created by
board code. But it's inappropriate for -device or device_add. Not
only is it inconsistent with how the other characrer device models
work (they connect to a backend explicitly identified by a "chardev"
property), it breaks when the backend has been picked up by the board
or a previous -device / device_add already.
Example:
$ qemu-system-ppc64 -M bamboo -S -device i82378 -device pc87312 -device pc87312
qemu-system-ppc64: -device pc87312: Property 'isa-parallel.chardev' can't take value 'parallel0', it's in use
Mark them with suitable FIXME comments.
Cc: Li Guang <lig.fnst@cn.fujitsu.com>
Cc: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Cc: Antony Pavlov <antonynpavlov@gmail.com>
Cc: "Edgar E. Iglesias" <edgar.iglesias@gmail.com>
Cc: Michael Walle <michael@walle.cc>
Cc: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Cc: "Andreas Färber" <andreas.faerber@web.de>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Drives defined with if!=none are for board initialization to wire up.
Board code calls drive_get() or similar to find them, and creates
devices with their qdev drive properties set accordingly.
Except a few devices go on a fishing expedition for a suitable backend
instead of exposing a drive property for board code to set: they call
driver_get() or drive_get_next() in their realize() or init() method
to implicitly connect to the "next" backend with a certain interface
type.
Picking up backends that way works when the devices are created by
board code. But it's inappropriate for -device or device_add. Not
only is this inconsistent with how the other block device models work
(they connect to a backend explicitly identified by a "drive"
property), it breaks when the "next" backend has been picked up by the
board already.
Example:
$ qemu-system-arm -S -M connex -pflash flash.img -device ssi-sd
Aborted (core dumped)
Mark them with suitable FIXME comments.
Cc: Andrzej Zaborowski <balrogg@gmail.com>
Cc: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Cc: "Andreas Färber" <andreas.faerber@web.de>
Cc: Michael Walle <michael@walle.cc>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Commits 6e05a12f8f and db25a1581 both attempt to fix the
same "failed to free memory containing flash filename" bug,
with the effect that when they were both applied we ended
up freeing the memory twice. Delete the spurious extra free.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1427968334-14527-1-git-send-email-peter.maydell@linaro.org
As 4de9a88(hw/arm/virt: Fix memory leak reported by Coverity)
and 6e05a12(arm: fix memory leak) both handle the memory leak
reported by Coverity, this cause qemu corruption due to
double free.
Signed-off-by: Shannon Zhao <zhaoshenglong@huawei.com>
Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1427944026-8968-1-git-send-email-zhaoshenglong@huawei.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The current code was negatively indexing the cpu state array and not
synchronizing banked spsr register state with the current mode's spsr
state, causing occasional failures with migration.
Some munging is done to take care of the aarch64 mapping and also to
ensure the most current value of the spsr is updated to the banked
registers (relevant for KVM<->TCG migration).
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
For migration to work we need to sync all of the register state. This is
especially noticeable when GCC starts using FP registers as spill
registers even with integer programs.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
As there is logic to deal with the difference between edge and level
triggered interrupts in the kernel we must ensure it knows the
configuration of the IRQs before we restore the pending state.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This adds the saving and restore of the current Multi-Processing state
of the machine. While the KVM_GET/SET_MP_STATE API exposes a number of
potential states for x86 we only use two for ARM. Either the process is
running or not. We then save this state into the cpu_powered TCG state
to avoid changing the serialisation format.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The AArch64 SPSR_EL1 register is architecturally mandated to
be mapped to the AArch32 SPSR_svc register. This means its
state should live in QEMU's env->banked_spsr[1] field.
Correct the various places in the code that incorrectly
put it in banked_spsr[0].
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
As the conditional statement had to be split anyway, we can also
add a better error report message.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1426877982-3603-1-git-send-email-sw@weilnetz.de
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
As the conditional statement had to be split anyway, we can also
add a better error report message.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1426877963-3556-1-git-send-email-sw@weilnetz.de
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Coverity reports a resource leak for sysboot_filename which is allocated
by qemu_find_file.
In addition, that name is used to get the size of the image, but a
different image name was used to load it.
In addition, instead of passing the maximum allowed image size the actual
image size was passed to load_image_targphys.
Fix all three issues.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Message-id: 1426326781-2488-1-git-send-email-sw@weilnetz.de
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The VNC server websockets decoder will read and buffer data from
websockets clients until it sees the end of the HTTP headers,
as indicated by \r\n\r\n. In theory this allows a malicious to
trick QEMU into consuming an arbitrary amount of RAM. In practice,
because QEMU runs g_strstr_len() across the buffered header data,
it will spend increasingly long burning CPU time searching for
the substring match and less & less time reading data. So while
this does cause arbitrary memory growth, the bigger problem is
that QEMU will be burning 100% of available CPU time.
A novnc websockets client typically sends headers of around
512 bytes in length. As such it is reasonable to place a 4096
byte limit on the amount of data buffered while searching for
the end of HTTP headers.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The logic for decoding websocket frames wants to fully
decode the frame header and payload, before allowing the
VNC server to see any of the payload data. There is no
size limit on websocket payloads, so this allows a
malicious network client to consume 2^64 bytes in memory
in QEMU. It can trigger this denial of service before
the VNC server even performs any authentication.
The fix is to decode the header, and then incrementally
decode the payload data as it is needed. With this fix
the websocket decoder will allow at most 4k of data to
be buffered before decoding and processing payload.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
[ kraxel: fix frequent spurious disconnects, suggested by Peter Maydell ]
@@ -361,7 +361,7 @@ int vncws_decode_frame_payload(Buffer *input,
- *payload_size = input->offset;
+ *payload_size = *payload_remain;
[ kraxel: fix 32bit build ]
@@ -306,7 +306,7 @@ struct VncState
- uint64_t ws_payload_remain;
+ size_t ws_payload_remain;
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This reverts commit c3c1bb99d1.
It causes problems with boards that declare memory regions shorter
than the registers they contain.
Reported-by: Zoltan Balaton <balaton@eik.bme.hu>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If QEMU forks after the CPU threads have been created, qemu_mutex_lock_iothread
will not be able to do qemu_cpu_kick_thread. There is no solution other than
assuming that forks after the CPU threads have been created will end up in an
exec. Forks before the CPU threads have been created (such as -daemonize)
have to call rcu_after_fork manually.
Notably, the oxygen theme for GTK+ forks and shows a "No such process" error
without this patch.
This patch can be reverted once the iothread loses the "kick the TCG thread"
magic.
User-mode emulation does not use the iothread, so it can also call
rcu_after_fork.
Reported by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Tested by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Commit cd61cb2 pc: acpi-build: generate pvpanic device description dynamically
introduced regression changing pvpanic device HID from
QEMU0001 to QEMU0002.
Fix AML generated code so that pvpanic device
would keep its original HID. i.e. QEMU0001
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reported-by: Gal Hammer <ghammer@redhat.com>
Message-Id: <1427717907-25027-1-git-send-email-imammedo@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
showing a memory device whose memdev is removed leads an assert:
(qemu) object_add memory-backend-ram,id=ram0,size=128M
(qemu) device_add pc-dimm,id=d0,memdev=ram0
(qemu) object_del ram0
(qemu) info memory-devices
**
ERROR:qom/object.c:1274:object_get_canonical_path_component:\
assertion failed: (obj->parent != NULL)
Aborted
The patch prevents removing an in-use mem backend and error out.
Signed-off-by: Lin Ma <lma@suse.com>
Message-Id: <1427704589-7688-3-git-send-email-lma@suse.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
If backends implement the can_be_deleted and it returns false,
Then the qmp_object_del won't delete the given backends.
Signed-off-by: Lin Ma <lma@suse.com>
Message-Id: <1427704589-7688-2-git-send-email-lma@suse.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
