Fix some of the nasty TCG race conditions and crashes by implementing
cpu_exit() as setting a flag which is checked at the start of each TB.
This avoids crashes if a thread or signal handler calls cpu_exit()
while the execution thread is itself modifying the TB graph (which
may happen in system emulation mode as well as in linux-user mode
with a multithreaded guest binary).
This fixes the crashes seen in LP:668799; however there are another
class of crashes described in LP:1098729 which stem from the fact
that in linux-user with a multithreaded guest all threads will
use and modify the same global TCG date structures (including the
generated code buffer) without any kind of locking. This means that
multithreaded guest binaries are still in the "unsupported"
category.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
If tcg_qemu_tb_exec() returns a value whose low bits don't indicate a
link to an indexed next TB, this means that the TB execution never
started (eg because the instruction counter hit zero). In this case the
guest PC has to be reset to the address of the start of the TB.
Refactor the cpu-exec code to make all tcg_qemu_tb_exec() calls pass
through a wrapper function which does this restoration if necessary.
Note that the apparent change in cpu_exec_nocache() from calling
cpu_pc_from_tb() with the old TB to calling it with the TB returned by
do_tcg_qemu_tb_exec() is safe, because in the nocache case we can
guarantee that the TB we try to execute is not linked to any others,
so the only possible returned TB is the one we started at. That is,
we should arguably previously have included in cpu_exec_nocache() an
assert(next_tb & ~TB_EXIT_MASK) == tb), since the API requires restore
from next_tb but we were using tb.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Introduce ENV_OFFSET macros which can be used in non-target-specific
code that needs to generate TCG instructions which reference CPUState
fields given the cpu_env register that TCG targets set up with a
pointer to the CPUArchState struct.
Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Document tcg_qemu_tb_exec(). In particular, its return value is a
combination of a pointer to the next translation block and some
extra information in the low two bits. Provide some #defines for
the values passed in these bits to improve code clarity.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Fix build breakage by 803d805bcef4ea7b7d6ef0b4929263e1160d6b3c:
make tcg_out_addsub2() always available.
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
This doesn't happen in the real hardware. The Zynq TRM explicitly states that
this bit has no effect on the rx descriptor pointer ("The receive queue
pointer register is unaffected").
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: 06fdf92b78ee62d8965779bafd29c8df1a5d2718.1360901435.git.peter.crosthwaite@xilinx.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Bits in the ISR were continually mirroring their corresponding TX/RX SR bits.
This is incorrect. The ISR bits are only ever set at the time their
corresponding event occurs.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: cedfb6d108318846480b416a6041023ea5a353d6.1360901435.git.peter.crosthwaite@xilinx.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The gem_receive() function replicates the logic for whether or not the device
can rx. Just call the actual gem_can_receive() function in place.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: bf7f93969f3e01fbc76d68d2955307fdbad11bb1.1360901435.git.peter.crosthwaite@xilinx.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Currently, M25P80 uses an object property to differentiate between flash parts.
Changed this over to use QOM sub-classes - the actual names of the different parts
are used to create a set of dynamic classes which passes the part info as class
data. The object no longer needs to search the known_devices table for itself,
instead it just gets its info from its own class.
Kept the intermediate class definition private to m25p80.c for the moment, as
the expectation is parts will only be added as new entries in the table. We can
factor out the TYPE_M25P80 abstraction into a header on a demand basis.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: e24e156d-ff96-4901-997a-e31178b08bee@VA3EHSMHS021.ehs.local
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The Xilinx Zynq device has two SDHCI controllers. Added to the machine model.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Allows for repeating of -sd arguments in the same way as -pflash and -mtdblock.
Acked-by: Igor Mitsyanko <i.mitsyanko@samsung.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Device model for standard SD Host Controller Interface (SDHCI) compliant with
version 2.00 of SD association specification.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Igor Mitsyanko <i.mitsyanko@samsung.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Split the SCU in a9mpcore out into its own object definition. mpcore is now
just a container for the mpcore components.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This field was write only and thus unused. Removed.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
In QEMU emulation, there is no functional difference between the ARM mpcore
private timers and watchdogs. Removed all the distinction between the two from
arm_mptimer.c and converted it to be just the mptimer. a9mpcore and arm11mpcore
just instantiate the same mptimer object twice to get both timer and WDT.
If in the future we want to make the WDT functionally different then we can use
either QOM hierarchy to derive WDT from from mptimer, or we can add a property
"is-wdt" or some such.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
To conform with QEMU coding style.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Trivial find replace on type names "timerblock" and "arm_mptimer_state" to
conform with QEMU coding style.
Signed-off-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The code for handling the default "unknown command state" case in
pflash_read in pflash_cfi01.c comments "reset state & treat it as
a read". However the code doesn't actually do this. Moving the
default case to the top of the switch so it can fall through into
the read case brings this file into line with pflash_cfi02 and
makes the code behave as the comments suggest.
The pflash_cfi01 code has always had this bug -- it was presumably
introduced when the original author copied the cfi02 code and
rearranged the order of the switch statement without noticing
that the default case relied on the fall-through.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Tested-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: 1358777318-7579-3-git-send-email-peter.maydell@linaro.org
The Intel flash command set requires that a read operation after
doing a 'single byte write' command returns the status register;
add this case to pflash_read() so we return the correct information.
Similarly, the case for the 0x28 flavour of block erase was missing.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Tested-by: Peter Crosthwaite <peter.crosthwaite@xilinx.com>
Message-id: 1358777318-7579-2-git-send-email-peter.maydell@linaro.org
The proper mapping between 24 hours and 12 hours modes is:
0 12 AM
1-11 1-11 AM
12 12 PM
13-23 1-11 PM
Fix code accordingly.
Signed-off-by: Antoine Mathys <barsamin@gmail.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Alex Williamson (3):
seabios q35: Enable all PIRQn IRQs at startup
seabios q35: Add new PCI slot to irq routing function
seabios: Add a dummy PCI slot to irq mapping function
Avik Sil (1):
USB-EHCI: Fix null pointer assignment
Kevin O'Connor (4):
Update tools/acpi_extract.py to handle iasl 20130117 release.
Fix Makefile - don't reference "out/" directly, instead use "$(OUT)".
build: Don't require $(OUT) to be a sub-directory of the main
directory.
Verify CC is valid during build tests.
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
These correspond very closely to the insns that we're emulating.
Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
The virtio-serial device is expected to use 2 MSI vectors, one for
control queue and a second shared for all queues.
Signed-off-by: Gal Hammer <ghammer@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This commit does the same for the ICH9 LPC as commit 1ec4ba74 for the
PIIX3. For the present we're ignoring the Full Reset (FULL_RST) and System
Reset (SYS_RST) bits; the guest can read them back but that's it.
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch addresses the issue fully described here:
http://lists.nongnu.org/archive/html/qemu-devel/2013-02/msg01804.html
Linux kernels prior to 2.6.36 do not disable the PCI device during
enumeration process. Since lower and higher parts of a 64bit BAR
are programmed separately this leads to qemu receiving a request to occupy
a completely wrong address region for a short period of time.
We have found that the boot process screws up completely if kvm-apic range
is overlapped even for a short period of time (it is fine for other
regions though).
This patch raises the priority of the kvm-apic memory region, so it is
never pushed out by PCI devices. The patch is quite safe as it does not
touch memory manager.
Signed-off-by: Alexey Korolev <akorolex@gmail.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
QEMU 1.3 does not emulate the link auto negotiation, so if migrate to a
1.3 machine during link auto negotiation, the guest link will be set to down.
Fix this by just disabling auto negotiation for 1.3 and older.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Edivaldo reports a problem that the array of NetClientState in NICState is too
large - MAX_QUEUE_NUM(1024) which will wastes memory even if multiqueue is not
used.
Instead of static arrays, solving this issue by allocating the queues on demand
for both the NetClientState array in NICState and VirtIONetQueue array in
VirtIONet.
Tested by myself, with single virtio-net-pci device. The memory allocation is
almost the same as when multiqueue is not merged.
Cc: Edivaldo de Araujo Pereira <edivaldoapereira@yahoo.com.br>
Cc: qemu-stable@nongnu.org
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
historically the kernel queues packets two times. once
at the device and second in qdisc. this is believed to cause
interface stalls if one of these queues overruns.
setting IFF_ONE_QUEUE is the default in kernels >= 3.8. the
flag is ignored since then. see kernel commit
5d097109257c03a71845729f8db6b5770c4bbedc
Signed-off-by: Peter Lieven <pl@kamp.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Obviously, hub does not support multiqueue tap. So this patch forbids creating
multiple queue tap when hub is used to prevent the crash when command line such
as "-net tap,queues=2" is used.
Cc: qemu-stable@nongnu.org
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In the current implementation of qemu, running without a network
backend will cause the queue to grow unbounded when the guest is
transmitting traffic.
This patch fixes the problem by implementing bounded size NetQueue,
used with an arbitrary limit of 10000 packets, and dropping packets
when the queue is full _and_ the sender does not pass a callback.
The second condition makes sure that we never drop packets that
contains a callback (which would be tricky, because the producer
expects the callback to be run when all previous packets have been
consumed; so we cannot run it when the packet is dropped).
If documentation is correct, producers that submit a callback should
stop sending when their packet is queued, so there is no real risk
that the queue exceeds the max size by large values.
Signed-off-by: Luigi Rizzo <rizzo@iet.unipi.it>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
When frontend and backend are connected through a hub as below
(showing only one direction), and the frontend (or in general, all
output ports of the hub) cannot accept more traffic, the backend
queues packets in queue-A.
When the frontend (or in general, one output port) becomes ready again,
quemu tries to flush packets from queue-B, which is unfortunately empty.
e1000.0 <--[queue B]-- hub0port0(hub)hub0port1 <--[queue A]-- tap.0
To fix this i propose to introduce a new function net_hub_flush()
which is called when trying to flush a queue connected to a hub.
Signed-off-by: Luigi Rizzo <rizzo@iet.unipi.it>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This fixes two bugs related to memory sync during
migration:
- ram address calculation was missing the chunk
address, so the wrong page was dirtied
- one after last was used instead of the
end address of a region, which might overflow to 0
and cause us to skip the region when the region ends at
~0x0ull.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Jason Wang <jasowang@redhat.com>
Older glib doesn't implement g_poll(). Most notably the glib version in use
on SLE11 is on 2.18 which is hit by this.
We do want to use g_poll() in the source however. So on older systems, just
wrap it with functions that do exist on older versions.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
Message-id: 1361835970-2889-1-git-send-email-agraf@suse.de
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Switch the default for qemu_log logging output from "/tmp/qemu.log"
to stderr. This is an incompatible change in some sense, but logging
is mostly used for debugging purposes so it shouldn't affect production
use. The previous behaviour can be obtained by adding "-D /tmp/qemu.log"
to the command line.
This change requires us to:
* update all the documentation/help text (we take the opportunity
to smooth out minor inconsistencies between the phrasing in
linux-user/bsd-user/system help messages)
* make linux-user and bsd-user defer to qemu-log for the default
logging destination rather than overriding it themselves
* ensure that all logfile closing is done via qemu_log_close()
and that that function doesn't close stderr
as well as the obvious change to the behaviour of do_qemu_set_log()
when no logfile name has been specified.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-id: 1361901160-28729-1-git-send-email-peter.maydell@linaro.org
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The GDK_KEY_XXX symbols are new in GTK3 and only the most
recent GTK2 releases. Most versions of GTK2 have simply
used GDK_XXX
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-14-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The gtk_widget_get_realized method only arrived in GTK 2.20,
so defined a compat macro for earlier GTK
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-13-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Add a arg to configure to switch from GTK2 (default) to
GTK3 (optional) build for QEMU.
./configure --with-gtkabi=3.0
will choose GTK3, while
./configure --with-gtkabi=2.0
will choose GTK2 (and remains the current default)
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-12-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The x_keymap.o file is required by both GTK and SDL builds,
so it must be explicitly listed as a GTK dep to ensure the
linker works when SDL is disabled
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-11-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In GTK3 the 'expose-event' signal has been replaced by a new
'draw' signal. The only difference is that the latter will
pre-create the cairo drawing context & set the clip mask.
Since the drawing code is already structured in a nice way,
we can just wire up the 'gd_draw_event' method to the 'draw'
signal in GTK3
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-10-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The gtk_widget_size_request method has been replaced by
the gtk_widget_get_preferred_size method in GTK3. Conditionally
call the new method in GTK3
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-9-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
In GTK3 the gdk_display_warp_pointer method is deprecated.
Instead we should use gdk_device_warp on the GdkDevice
instead associated with the event being processed.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-8-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The gtk_menu_append method has long been deprecated in favour
of the gtk_menu_shell_append method. The former is now entirely
gone in GTK3, so switch all code to the latter which works on
both GTK2 and GTK3
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-7-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
The GtkVBox class is deprecated, in favour of just using the
GtkBox class directly. Eventually even GtkBox will be
deprecated in favour of GtkGrid, but that is a bigger fix
which can wait.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-6-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
On GTK3 there is support for multiple pointer devices, so
rather than using gdk_pointer_grab / gdk_pointer_ungrab
we should iterate over all devices, grabbing each one in
turn
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 1361805646-6425-5-git-send-email-berrange@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>