sdk/emulator/qemu.git
12 years agomain loop: use msec-based timeout in glib_select_fill
Paolo Bonzini [Tue, 20 Mar 2012 09:49:17 +0000 (10:49 +0100)]
main loop: use msec-based timeout in glib_select_fill

The timeval-based timeout is not needed until we actually invoke select,
so compute it only then.  Also group the two calls that modify the
timeout, glib_select_fill and os_host_main_loop_wait.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotarget-sparc: QOM'ify CPU
Andreas Färber [Wed, 4 Apr 2012 23:29:40 +0000 (01:29 +0200)]
target-sparc: QOM'ify CPU

Embed CPUSPARCState as first member of SPARCCPU.
Drop cpu_sparc_close() in favor of object_delete() and a finalizer.
Let cpu_state_reset() call cpu_reset().

Make TYPE_SPARC_CPU non-abstract for now.
Distinguish between "sparc-cpu" and "sparc64-cpu".

Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotarget-sparc: Rename cpu_init.c
Andreas Färber [Wed, 4 Apr 2012 23:29:39 +0000 (01:29 +0200)]
target-sparc: Rename cpu_init.c

Align QOM'ified targets, with a view to simplify Makefile.target.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoMerge branch 's390-for-upstream' of git://repo.or.cz/qemu/agraf
Blue Swirl [Wed, 4 Apr 2012 20:45:03 +0000 (20:45 +0000)]
Merge branch 's390-for-upstream' of git://repo.or.cz/qemu/agraf

* 's390-for-upstream' of git://repo.or.cz/qemu/agraf:
  target-s390x: Update s390x_{tod,cpu}_timer() to use S390CPU
  target-s390x: QOM'ify CPU init
  target-s390x: QOM'ify CPU reset
  target-s390x: QOM'ify CPU

12 years agoImprove interrupt handling priority
Artyom Tarasenko [Tue, 3 Apr 2012 15:49:05 +0000 (17:49 +0200)]
Improve interrupt handling priority

The vector interrupt has higher priority than interrupt_level_n.
Also check only interrupt_level_n concurency when TL > 0, the traps of
other types may be nested.

Signed-off-by: Artyom Tarasenko <atar4qemu@gmail.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoFix vector interrupt handling
Artyom Tarasenko [Tue, 3 Apr 2012 15:49:04 +0000 (17:49 +0200)]
Fix vector interrupt handling

Don't produce stray irq 5, don't overwrite ivec_data if still busy with
processing of the previous interrupt.

Signed-off-by: Artyom Tarasenko <atar4qemu@gmail.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotarget-s390x: Update s390x_{tod,cpu}_timer() to use S390CPU
Andreas Färber [Mon, 2 Apr 2012 12:00:43 +0000 (14:00 +0200)]
target-s390x: Update s390x_{tod,cpu}_timer() to use S390CPU

In place of CPUS390XState pass S390CPU as opaque from the new initfn.
cpu_interrupt() is anticipated to take a CPUState in the future.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
12 years agotarget-s390x: QOM'ify CPU init
Andreas Färber [Mon, 2 Apr 2012 11:56:29 +0000 (13:56 +0200)]
target-s390x: QOM'ify CPU init

Move code from cpu_s390x_init() into an initfn.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
12 years agotarget-s390x: QOM'ify CPU reset
Andreas Färber [Mon, 2 Apr 2012 11:31:59 +0000 (13:31 +0200)]
target-s390x: QOM'ify CPU reset

Move code from cpu_state_reset() to s390_cpu_reset().

Signed-off-by: Andreas Färber <afaerber@suse.de>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
12 years agotarget-s390x: QOM'ify CPU
Andreas Färber [Mon, 2 Apr 2012 09:39:23 +0000 (11:39 +0200)]
target-s390x: QOM'ify CPU

Embed CPUS390XState as first member of S390CPU.
Since -cpu is being ignored, make TYPE_S390_CPU non-abstract.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
12 years agoqdev: put all devices under /machine
Paolo Bonzini [Wed, 28 Mar 2012 14:34:12 +0000 (16:34 +0200)]
qdev: put all devices under /machine

Avoid cluttering too much the QOM root.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqdev: give all devices a canonical path
Paolo Bonzini [Tue, 27 Mar 2012 16:38:47 +0000 (18:38 +0200)]
qdev: give all devices a canonical path

A strong limitation of QOM right now is that unconverted ports
(e.g. all...) do not give a canonical path to devices that are
part of the board.  This in turn makes it impossible to replace
PROP_PTR with a QOM link for example.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqdev: add children before qdev_init
Paolo Bonzini [Tue, 27 Mar 2012 16:38:46 +0000 (18:38 +0200)]
qdev: add children before qdev_init

We want the composition tree to to be in order by the time we call
qdev_init, so that a single set of the toplevel realize property can
propagate all the way down the composition tree.

This is not the case so far.  Unfortunately, this is incompatible
with calling qdev_init in the constructor wrappers for devices,
so for now we need to unattach some devices that are created through
those wrappers.  This will be fixed by removing qdev_init and instead
setting the toplevel realize property after machine init.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqom: add container_get
Paolo Bonzini [Tue, 27 Mar 2012 16:38:45 +0000 (18:38 +0200)]
qom: add container_get

This is QOM "mkdir -p".  It is useful when referring to
container objects such as "/machine".

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoMerge remote-tracking branch 'kiszka/queues/slirp' into staging
Anthony Liguori [Mon, 2 Apr 2012 14:46:45 +0000 (09:46 -0500)]
Merge remote-tracking branch 'kiszka/queues/slirp' into staging

* kiszka/queues/slirp:
  slirp: Signal free input buffer space to io-thread
  w32/slirp: Undefine error constants before their redefinition
  slirp: use socket_set_nonblock
  slirp: clean up conflicts with system headers

12 years agoMerge remote-tracking branch 'stefanha/tracing' into staging
Anthony Liguori [Mon, 2 Apr 2012 14:44:35 +0000 (09:44 -0500)]
Merge remote-tracking branch 'stefanha/tracing' into staging

* stefanha/tracing:
  tracetool: dtrace: handle in and next reserved words
  tracetool: dtrace disabled-events fix
  Makefile.target: code stp dependency on trace-events

12 years agoserial: clear LSR.TEMT when populating the TSR
Anthony Liguori [Sun, 1 Apr 2012 19:18:30 +0000 (14:18 -0500)]
serial: clear LSR.TEMT when populating the TSR

We never actually clear the TEMT (transmit sending register empty) flag when
populating the TSR.  We set the flag, but since it's never cleared, setting it
is sort of pointless..

I found this with a unit test case.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoserial: fix retry logic
Anthony Liguori [Sun, 1 Apr 2012 19:03:21 +0000 (14:03 -0500)]
serial: fix retry logic

I'm not sure if the retry logic has ever worked when not using FIFO mode.  I
found this while writing a test case although code inspection confirms it is
definitely broken.

The TSR retry logic will never actually happen because it is guarded by an
'if (s->tsr_rety > 0)' but this is the only place that can ever make the
variable greater than zero.  That effectively makes the retry logic an 'if (0)'.

I believe this is a typo and the intention was >= 0.  Once this is fixed though,
I see double transmits with my test case.  This is because in the non FIFO
case, serial_xmit may get invoked while LSR.THRE is still high because the
character was processed but the retransmit timer was still active.

We can handle this by simply checking for LSR.THRE and returning early.  It's
possible that the FIFO paths also need some attention.

Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqtest: use qemu_gettimeofday()
Anthony Liguori [Fri, 30 Mar 2012 19:04:04 +0000 (14:04 -0500)]
qtest: use qemu_gettimeofday()

On linux, qemu_timeval will always be two long ints.  On windows, we use our
own struct definition.  This should fix win64.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agomilkymist-vgafb: add missing register
Michael Walle [Sat, 31 Mar 2012 17:55:25 +0000 (19:55 +0200)]
milkymist-vgafb: add missing register

This bug existed since the first commit. Fortunately, the affected
registers have no functionality in qemu. This will only prevent the
following warning:
  milkymist_vgafb: write access to unknown register 0x00000034

Signed-off-by: Michael Walle <michael@walle.cc>
12 years agotarget-lm32: add simple disassembler
Michael Walle [Sat, 31 Mar 2012 17:54:20 +0000 (19:54 +0200)]
target-lm32: add simple disassembler

Because binutils disassembler is based on libopcode, this is a rewrite from
scratch.

Signed-off-by: Michael Walle <michael@walle.cc>
12 years agomilkymist-sysctl: support for new core version
Michael Walle [Sat, 31 Mar 2012 17:54:09 +0000 (19:54 +0200)]
milkymist-sysctl: support for new core version

The new version introduces the following new registers:
 - SoC clock frequency: read-only of system clock used on the SoC
 - debug scratchpad: 8 bit scratchpad register
 - debug write lock: write once register, without any function on QEMU

Signed-off-by: Michael Walle <michael@walle.cc>
12 years agotarget-lm32: init tcg only if available
Michael Walle [Sat, 31 Mar 2012 17:54:05 +0000 (19:54 +0200)]
target-lm32: init tcg only if available

Once qtest support for target-lm32 arrives, tcg may be disabled.

Signed-off-by: Michael Walle <michael@walle.cc>
12 years agotests: fix out-of-tree building for lm32 target
Michael Walle [Sat, 31 Mar 2012 17:53:51 +0000 (19:53 +0200)]
tests: fix out-of-tree building for lm32 target

Signed-off-by: Michael Walle <michael@walle.cc>
12 years agoMerge branch 'arm-devs.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm
Blue Swirl [Sat, 31 Mar 2012 12:10:07 +0000 (12:10 +0000)]
Merge branch 'arm-devs.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm

* 'arm-devs.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm:
  pl031: switch clock base to rtc_clock
  pl031: rearm alarm timer upon load
  arm: switch real-time clocks to rtc_clock
  omap: switch omap_lpg to vm_clock
  rtc: add -rtc clock=rt

12 years agoMerge branch 'qom-cpu-unicore32.v3' of git://github.com/afaerber/qemu-cpu
Blue Swirl [Sat, 31 Mar 2012 12:09:06 +0000 (12:09 +0000)]
Merge branch 'qom-cpu-unicore32.v3' of git://github.com/afaerber/qemu-cpu

* 'qom-cpu-unicore32.v3' of git://github.com/afaerber/qemu-cpu:
  target-unicore32: Move CPU-dependent init into initfn
  target-unicore32: QOM'ify CPU
  target-unicore32: License future contributions under GPLv2+
  target-unicore32: Relicense to GPLv2+
  MAINTAINERS: Add entry for UniCore32

12 years agoMerge branch 'target-arm.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm
Blue Swirl [Sat, 31 Mar 2012 12:06:10 +0000 (12:06 +0000)]
Merge branch 'target-arm.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm

* 'target-arm.for-upstream' of git://git.linaro.org/people/pmaydell/qemu-arm:
  ARM: Permit any ARMv6K CPU to read the MVFR0 and MVFR1 VFP registers.
  target-arm: Minimalistic CPU QOM'ification
  target-arm: Drop cpu_arm_close()

12 years agobuildfix: check for old pod2man versions
Gerd Hoffmann [Thu, 29 Mar 2012 08:55:18 +0000 (10:55 +0200)]
buildfix: check for old pod2man versions

Older pod2man don't have a --utf8 switch, check for this in conffigure
and use it only when present.  Fixes build on RHEL-5.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoFix conversion from lower to upper case with Turkish locale
Stefan Weil [Tue, 27 Mar 2012 17:23:53 +0000 (19:23 +0200)]
Fix conversion from lower to upper case with Turkish locale

Some locale settings let make fail or create wrong results
because tr '[:lower:]' '[:upper:]' which is used to convert
from lower to upper case depends on the locale.

With locale tr_TR.UTF-8, lower case 'i' is not converted to 'I'.
This results in wrong entries in config-host.h like these ones:

  #define CONFIG_QEMU_PREFiX "/usr/local"
  #define CONFIG_QEMU_BiNDiR "/usr/local/bin"

This problem was reported by Emre Ersin.

The same problem occurs when configure creates the target specific
files config-target.mak. They get wrong declarations:

TARGET_CRiS=y
TARGET_i386=y
TARGET_MiCROBLAZE=y
TARGET_MiPS64=y
TARGET_MiPS=y
TARGET_UNiCORE32=y

It is sufficient to restrict the conversion to the characters a-z.

Using this explicit range avoids the dependency on the locale
settings and is also shorter.

v2:
POSIX says that 'tr a-z' is unspecified outside of the POSIX
locale, so we must set LC_ALL=C to make sure that we are using
POSIX (hint from Eric Blake, thanks).

Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoqtest: avoid a warning with RTC test
Blue Swirl [Fri, 30 Mar 2012 17:42:27 +0000 (17:42 +0000)]
qtest: avoid a warning with RTC test

Avoid this warning on OpenBSD:
  CC    tests/rtc-test.o
/src/qemu/tests/rtc-test.c: In function 'check_time':
/src/qemu/tests/rtc-test.c:171: warning: format '%ld' expects type 'long int', but argument 2 has type 'time_t'
/src/qemu/tests/rtc-test.c:173: warning: format '%ld' expects type 'long int', but argument 2 has type 'time_t'

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoqtest: fix 32-bit build
Anthony Liguori [Fri, 30 Mar 2012 17:53:54 +0000 (12:53 -0500)]
qtest: fix 32-bit build

time_t appears to be an unsigned long so use %ld.

Reported-by: Stefan Weil <weil@mail.berlios.de>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoslirp: Signal free input buffer space to io-thread
Jan Kiszka [Fri, 30 Mar 2012 17:29:08 +0000 (19:29 +0200)]
slirp: Signal free input buffer space to io-thread

This massively accelerates slirp reception speed: If data arrives
faster than the guest can read it from the input buffer, the file
descriptor for the corresponding socket was taken out of the fdset for
select. However, the event of the guest reading enough data from the
buffer was not signaled. Thus, the io-thread only noticed this change
on the next time-driven poll. Fix this by kicking the io-thread as
required.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
12 years agoqtest: fix out of tree build
Blue Swirl [Fri, 30 Mar 2012 17:39:33 +0000 (17:39 +0000)]
qtest: fix out of tree build

Trace objects are also needed if tracing is enabled.

Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoqtest: add rtc-test test-case
Anthony Liguori [Wed, 28 Mar 2012 13:42:07 +0000 (15:42 +0200)]
qtest: add rtc-test test-case

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agortc: split out macros into a header file and use in test case
Anthony Liguori [Wed, 28 Mar 2012 13:42:06 +0000 (15:42 +0200)]
rtc: split out macros into a header file and use in test case

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqtest: add C version of test infrastructure
Anthony Liguori [Wed, 28 Mar 2012 13:42:05 +0000 (15:42 +0200)]
qtest: add C version of test infrastructure

This also includes a qtest wrapper script to make it easier to launch qtest
tests directly.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqtest: add clock management
Paolo Bonzini [Wed, 28 Mar 2012 13:42:04 +0000 (15:42 +0200)]
qtest: add clock management

This patch combines qtest and -icount together to turn the vm_clock
into a source that can be fully managed by the client.  To this end new
commands clock_step and clock_set are added.  Hooking them with libqtest
is left as an exercise to the reader.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqtest: IRQ interception infrastructure
Paolo Bonzini [Wed, 28 Mar 2012 13:42:03 +0000 (15:42 +0200)]
qtest: IRQ interception infrastructure

Since /i440fx/piix3 is being removed from the composition tree, the
IO-APIC is placed under /i440fx.  This is wrong and should be changed
as soon as the /i440fx/piix3 path is put back.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqtest: add test framework
Anthony Liguori [Wed, 28 Mar 2012 13:42:02 +0000 (15:42 +0200)]
qtest: add test framework

The idea behind qtest is pretty simple.  Instead of executing a CPU via TCG or
KVM, rely on an external process to send events to the device model that the CPU
would normally generate.

qtest presents itself as an accelerator.  In addition, a new option is added to
establish a qtest server (-qtest) that takes a character device.  This is what
allows the external process to send CPU events to the device model.

qtest uses a simple line based protocol to send the events.  Documentation of
that protocol is in qtest.c.

I considered reusing the monitor for this job.  Adding interrupts would be a bit
difficult.  In addition, logging would also be difficult.

qtest has extensive logging support.  All protocol commands are logged with
time stamps using a new command line option (-qtest-log).  Logging is important
since ultimately, this is a feature for debugging.

Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agotest makefile overhaul
Paolo Bonzini [Wed, 28 Mar 2012 13:42:01 +0000 (15:42 +0200)]
test makefile overhaul

This introduces new test reporting infrastructure based on
gtester and gtester-report.

Also, all existing tests are moved to tests/, and tests/Makefile
is reorganized to factor out the commonalities in the rules.

Signed-off-by: Anthony Liguori <aliguori@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agotracetool: dtrace: handle in and next reserved words
Alon Levy [Thu, 29 Mar 2012 21:35:36 +0000 (23:35 +0200)]
tracetool: dtrace: handle in and next reserved words

Signed-off-by: Alon Levy <alevy@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agotracetool: dtrace disabled-events fix
Lee Essen [Tue, 20 Mar 2012 17:02:40 +0000 (17:02 +0000)]
tracetool: dtrace disabled-events fix

If there are "disabled" entries in the trace-events file then
linetod_nop() is called if the backend is dtrace, it's currently
not present. Also equivalent fix for stap.

Signed-off-by: Lee Essen <lee.essen@nowonline.co.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agoMakefile.target: code stp dependency on trace-events
Alon Levy [Sun, 18 Mar 2012 07:47:15 +0000 (08:47 +0100)]
Makefile.target: code stp dependency on trace-events

Signed-off-by: Alon Levy <alevy@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agopl031: switch clock base to rtc_clock
Paolo Bonzini [Fri, 30 Mar 2012 10:31:23 +0000 (10:31 +0000)]
pl031: switch clock base to rtc_clock

This lets the user specify the desired semantics.  By default, the RTC
will follow adjustments from the host's NTP client, and will remain in
sync when the virtual machine is stopped.  The previous behavior, which
provides determinism with both icount and qtest, remains available with
"-rtc clock=vm".

pl031 supports migration, so we need to convert the time base from
rtc_clock to vm_clock and back for backwards compatibility.  (The
rtc_clock may not be synchronized on the two machines, especially with
savevm/loadvm, so the conversion is needed anyway.  And since any time
base will do, why not pick the one base that is backwards compatible).

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agopl031: rearm alarm timer upon load
Paolo Bonzini [Fri, 30 Mar 2012 10:31:22 +0000 (10:31 +0000)]
pl031: rearm alarm timer upon load

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agoarm: switch real-time clocks to rtc_clock
Paolo Bonzini [Fri, 30 Mar 2012 10:31:22 +0000 (10:31 +0000)]
arm: switch real-time clocks to rtc_clock

This lets the user specify the desired semantics.  By default, the RTC
will follow adjustments from the host's NTP client.  "-rtc clock=vm" will
improve determinism with both icount and qtest.  Finally, the previous
behavior is available with "-rtc clock=rt".

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agoomap: switch omap_lpg to vm_clock
Paolo Bonzini [Fri, 30 Mar 2012 10:31:22 +0000 (10:31 +0000)]
omap: switch omap_lpg to vm_clock

The output of the pulse generator needs to be deterministic when
running in -icount mode, and to remain constant whenever the VM is
stopped.  So the right clock to use is vm_clock.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agortc: add -rtc clock=rt
Paolo Bonzini [Fri, 30 Mar 2012 10:31:21 +0000 (10:31 +0000)]
rtc: add -rtc clock=rt

This will let people use backwards-compatible semantics for devices that
will be affected by the following patches.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agoARM: Permit any ARMv6K CPU to read the MVFR0 and MVFR1 VFP registers.
Andrew Towers [Thu, 29 Mar 2012 02:41:08 +0000 (02:41 +0000)]
ARM: Permit any ARMv6K CPU to read the MVFR0 and MVFR1 VFP registers.

This patch replaces the ARM_FEATURE_VFP3 test when reading MVFR registers
with a test for a new feature flag ARM_FEATURE_MVFR, and sets this feature
for all ARMv6K cores (ARM1156 is not a v6K core, yet supports MVFR; qemu
does not support ARM1156 at this time.)

MVFR0 and MVFR1 were introduced in ARM1136JF-S r1p0 (ARMv6K, VFPv2) and are
present in ARM1156T2F-S (non-v6K), ARM1176JZF-S, ARM11MPCore and newer cores.
Reference: ARM DDI 0211H, 0290G, 0301H, 0360E.
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0211h/Ffbefjag.html

Without this change, the linux kernel will not boot with VFP support enabled
under ARM1176 system emulation, due to the unconditional use of MVFR1 at the
end of vfp_init() in arch/arm/vfp/vfpmodule.c:

  VFP support v0.3: implemetor 41 architecture 1 part 20 variant b rev 5
  Internal error: Oops - undefined instruction: 0 [#1]

Signed-off-by: Andrew Towers <atowers@gmail.com>
Reviewed-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agotarget-unicore32: Move CPU-dependent init into initfn
Andreas Färber [Thu, 29 Mar 2012 16:03:18 +0000 (18:03 +0200)]
target-unicore32: Move CPU-dependent init into initfn

Instead of setting values in a CPUID switch, do so in initfn functions.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
12 years agotarget-unicore32: QOM'ify CPU
Andreas Färber [Tue, 14 Feb 2012 00:16:17 +0000 (01:16 +0100)]
target-unicore32: QOM'ify CPU

Embed CPUUniCore32State as first member of UniCore32CPU.

Contributed under GPLv2+.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
12 years agotarget-unicore32: License future contributions under GPLv2+
Andreas Färber [Thu, 15 Mar 2012 14:02:14 +0000 (15:02 +0100)]
target-unicore32: License future contributions under GPLv2+

This is to limit relicensing obstacles to the pending IBM investigation.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
12 years agotarget-unicore32: Relicense to GPLv2+
Andreas Färber [Tue, 13 Mar 2012 15:48:19 +0000 (16:48 +0100)]
target-unicore32: Relicense to GPLv2+

Adopt the license text suggested by Guan Xue-tao (with a minor
simplification) for all target-unicore/ files except helper.c.

To helper.c Anthony Liguori contributed a qemu_malloc() -> g_malloc()
conversion, still pending IBM relicensing approval, so that remains
GPLv2 for now.

By relicensing all possible parts now, we avoid having to formally
relicense new, e.g., QOM code.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
Signed-off-by: Dor Laor <dlaor@redhat.com>
12 years agoMAINTAINERS: Add entry for UniCore32
Andreas Färber [Tue, 13 Mar 2012 23:45:43 +0000 (00:45 +0100)]
MAINTAINERS: Add entry for UniCore32

Signed-off-by: Andreas Färber <afaerber@suse.de>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
12 years agoMerge remote-tracking branch 'origin/master' into staging
Anthony Liguori [Thu, 29 Mar 2012 18:32:41 +0000 (13:32 -0500)]
Merge remote-tracking branch 'origin/master' into staging

* origin/master:
  qemu tcg: Remove one entry of INDEX_op_ld_i64 from ppc_op_defs

12 years agotarget-arm: Minimalistic CPU QOM'ification
Andreas Färber [Thu, 29 Mar 2012 04:50:31 +0000 (04:50 +0000)]
target-arm: Minimalistic CPU QOM'ification

Introduce only one non-abstract type TYPE_ARM_CPU and do not touch
cp15 registers to not interfere with Peter's ongoing remodelling.
Embed CPUARMState as first (additional) field of ARMCPU.

Let CPUClass::reset() call cpu_state_reset() for now.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agotarget-arm: Drop cpu_arm_close()
Andreas Färber [Thu, 29 Mar 2012 04:50:30 +0000 (04:50 +0000)]
target-arm: Drop cpu_arm_close()

It's unused, so no need to QOM'ify it later.

Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
12 years agoqemu tcg: Remove one entry of INDEX_op_ld_i64 from ppc_op_defs
Li Zhang [Thu, 29 Mar 2012 03:29:02 +0000 (14:29 +1100)]
qemu tcg: Remove one entry of INDEX_op_ld_i64 from ppc_op_defs

There two entries of INDEX_op_ld_i64 in the ppc_op_defs.  That causes an
assertion failure in tcg_add_target_add_op_defs() when --enable-debug is
used on a ppc64 backend (that's ppc64 host, not target).

Signed-off-by: Li Zhang <zhlcindy@linux.vnet.ibm.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: malc <av1474@comtv.ru>
12 years agow32/slirp: Undefine error constants before their redefinition
Jan Kiszka [Wed, 28 Mar 2012 19:03:42 +0000 (21:03 +0200)]
w32/slirp: Undefine error constants before their redefinition

Less warnings for your console.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
12 years agoslirp: use socket_set_nonblock
Paolo Bonzini [Thu, 22 Mar 2012 00:02:52 +0000 (01:02 +0100)]
slirp: use socket_set_nonblock

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
12 years agoslirp: clean up conflicts with system headers
Paolo Bonzini [Thu, 22 Mar 2012 00:02:51 +0000 (01:02 +0100)]
slirp: clean up conflicts with system headers

Right now, slirp/slirp.h cannot include some system headers and,
indirectly, qemu_socket.h.  Clean this up, and remove a duplicate
prototype that was introduced because of that.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
12 years agoMerge remote-tracking branch 'sstabellini/disk_io' into staging
Anthony Liguori [Wed, 28 Mar 2012 17:46:39 +0000 (12:46 -0500)]
Merge remote-tracking branch 'sstabellini/disk_io' into staging

* sstabellini/disk_io:
  xen_disk: when using AIO flush after the operation is completed
  xen_disk: open disk with BDRV_O_NOCACHE | BDRV_O_CACHE_WB | BDRV_O_NATIVE_AIO

12 years agoMerge remote-tracking branch 'sstabellini/xen-fixes' into staging
Anthony Liguori [Wed, 28 Mar 2012 17:46:11 +0000 (12:46 -0500)]
Merge remote-tracking branch 'sstabellini/xen-fixes' into staging

* sstabellini/xen-fixes:
  xen_disk: detach the blkdev before bdrv_delete
  xen_console: ignore console disconnect events from console/0

12 years agoMerge remote-tracking branch 'qmp/queue/qmp' into staging
Anthony Liguori [Wed, 28 Mar 2012 17:45:22 +0000 (12:45 -0500)]
Merge remote-tracking branch 'qmp/queue/qmp' into staging

* qmp/queue/qmp:
  qmp: document strict parsing
  qmp: parse commands in strict mode
  qmp: add and use q type specifier
  qapi: add strict mode to input visitor
  qapi: place outermost object on qiv stack
  qapi: untangle next_list
  qapi: allow freeing partially-allocated objects
  qapi: shortcut visits on errors
  qapi: fix memory leak on error
  qapi: fail hard on stack imbalance
  qapi: add a test case for type errors
  qapi: add struct-errors test case to test-qmp-output-visitor
  qapi: fix double free in qmp_output_visitor_cleanup()

12 years agoMerge remote-tracking branch 'alon/libcacard' into staging
Anthony Liguori [Wed, 28 Mar 2012 17:44:04 +0000 (12:44 -0500)]
Merge remote-tracking branch 'alon/libcacard' into staging

* alon/libcacard:
  libcacard/vcard_emul_nss: add warning for old coolkey
  libcacard/vcard_emul_nss: handle no readers at startup
  libcacard/vcard_emul_nss: don't stop thread when there are no slots

12 years agosparc: pass page aligned addresses to tlb_set_page
Blue Swirl [Sun, 18 Mar 2012 11:31:23 +0000 (11:31 +0000)]
sparc: pass page aligned addresses to tlb_set_page

Mask incoming page address early so that resolved addresses
are page aligned. Remove further address masking.

Tested-by: Artyom Tarasenko <atar4qemu@gmail.com>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agoxen_disk: detach the blkdev before bdrv_delete
Stefano Stabellini [Tue, 27 Mar 2012 16:03:07 +0000 (16:03 +0000)]
xen_disk: detach the blkdev before bdrv_delete

We need to detach the blkdev from the BlockDriverState before calling
bdrv_delete.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
12 years agoxen_console: ignore console disconnect events from console/0
Stefano Stabellini [Tue, 27 Mar 2012 16:02:21 +0000 (16:02 +0000)]
xen_console: ignore console disconnect events from console/0

The first console has a different location compared to other PV devices
(console, rather than device/console/0) and doesn't obey the xenstore
state protocol. We already special case the first console in con_init
and con_initialise, we should also do it in con_disconnect.

Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
12 years agoqmp: document strict parsing
Paolo Bonzini [Thu, 22 Mar 2012 21:39:39 +0000 (22:39 +0100)]
qmp: document strict parsing

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqmp: parse commands in strict mode
Paolo Bonzini [Thu, 22 Mar 2012 11:51:12 +0000 (12:51 +0100)]
qmp: parse commands in strict mode

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqmp: add and use q type specifier
Paolo Bonzini [Thu, 22 Mar 2012 11:51:11 +0000 (12:51 +0100)]
qmp: add and use q type specifier

"O" is being used by the transaction and qom-set commands to mean "any
QObject", but it really means "do not validate the argument list".
Add a new specifier with the correct meaning.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: add strict mode to input visitor
Paolo Bonzini [Thu, 22 Mar 2012 11:51:10 +0000 (12:51 +0100)]
qapi: add strict mode to input visitor

While QMP in general is designed so that it is possible to ignore
unknown arguments, in the case of the QMP server it is better to
reject them to detect bad clients.  In fact, we're already doing
this at the top level in the argument checker.  To extend this to
complex structures, add a mode to the input visitor where it checks
for unvisited keys and raises an error if it finds one.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: place outermost object on qiv stack
Paolo Bonzini [Thu, 22 Mar 2012 11:51:09 +0000 (12:51 +0100)]
qapi: place outermost object on qiv stack

This is a slight change in the implementation of QMPInputVisitor
that helps when adding strict mode.

Const QObjects cannot be inc/decref-ed, and that's why QMPInputVisitor
relies heavily on weak references to inner objects.  I'm not removing
the weak references now, but since refcount+const is a lost battle in C
(C++ has "mutable") I think removing const is fine in this case.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: untangle next_list
Paolo Bonzini [Thu, 22 Mar 2012 21:38:40 +0000 (22:38 +0100)]
qapi: untangle next_list

Right now, the semantics of next_list are complicated.  The caller must:

* call start_list

* call next_list for each element *including the first*

* on the first call to next_list, the second argument should point to
NULL and the result is the head of the list.  On subsequent calls,
the second argument should point to the last node (last result of
next_list) and next_list itself tacks the element at the tail of the
list.

This works for both input and output visitor, but having the visitor
write memory when it is only reading the list is ugly.  Plus, relying
on *list to detect the first call is tricky and undocumented.

We can initialize so->entry in next_list instead of start_list, leaving
it NULL in start_list.  This way next_list sees clearly whether it is
on the first call---as a bonus, it discriminates the cases based on
internal state of the visitor rather than external state.  We can
also pull the assignment of the list head from generated code up to
next_list.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: allow freeing partially-allocated objects
Paolo Bonzini [Thu, 22 Mar 2012 11:51:07 +0000 (12:51 +0100)]
qapi: allow freeing partially-allocated objects

Objects going through the dealloc visitor can be only partially allocated.
Detect the situation and avoid a segfault.  This also helps with the
input visitor, when there are errors.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: shortcut visits on errors
Paolo Bonzini [Thu, 22 Mar 2012 11:51:06 +0000 (12:51 +0100)]
qapi: shortcut visits on errors

We can exit very soon if we enter a visitor with a preexisting error.
This simplifies some cases because we will not have to deal with
obj being non-NULL while *obj is NULL.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: fix memory leak on error
Paolo Bonzini [Thu, 22 Mar 2012 11:51:05 +0000 (12:51 +0100)]
qapi: fix memory leak on error

QmpInputVisitor would leak the malloced struct if the stack was
overflowed.  This can be easily fixed using error_propagate.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: fail hard on stack imbalance
Paolo Bonzini [Thu, 22 Mar 2012 11:51:04 +0000 (12:51 +0100)]
qapi: fail hard on stack imbalance

QmpOutputVisitor will segfault if an imbalanced end function is
called.  So we can abort in QmpInputVisitor too.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: add a test case for type errors
Paolo Bonzini [Thu, 22 Mar 2012 11:51:03 +0000 (12:51 +0100)]
qapi: add a test case for type errors

There is no test case for parse errors, add one.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: add struct-errors test case to test-qmp-output-visitor
Paolo Bonzini [Tue, 20 Mar 2012 10:22:49 +0000 (11:22 +0100)]
qapi: add struct-errors test case to test-qmp-output-visitor

This test case verifies that invalid native enums are caught, and causes
qapi to tear down the QObject tree under construction, exercising the
previous patch.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoqapi: fix double free in qmp_output_visitor_cleanup()
Laszlo Ersek [Tue, 20 Mar 2012 10:22:48 +0000 (11:22 +0100)]
qapi: fix double free in qmp_output_visitor_cleanup()

Stack entries in QmpOutputVisitor are navigation links (weak references),
except the bottom (ie. least recently added) entry, which owns the root
QObject [1]. Make qmp_output_visitor_cleanup() drop the stack entries,
then release the QObject tree by the root.

Attempting to serialize an invalid enum inside a dictionary is an example
for triggering the double free.

[1] http://lists.nongnu.org/archive/html/qemu-devel/2012-03/msg03276.html

Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
12 years agoMerge remote-tracking branch 'mdroth/qga-pull-3-26-12' into staging
Anthony Liguori [Mon, 26 Mar 2012 20:17:51 +0000 (15:17 -0500)]
Merge remote-tracking branch 'mdroth/qga-pull-3-26-12' into staging

* mdroth/qga-pull-3-26-12:
  qemu-ga: fix bsd build, and re-org linux-specific implementations

12 years agoMerge remote-tracking branch 'stefanha/trivial-patches' into staging
Anthony Liguori [Mon, 26 Mar 2012 20:08:26 +0000 (15:08 -0500)]
Merge remote-tracking branch 'stefanha/trivial-patches' into staging

* stefanha/trivial-patches:
  trace-events: Fix broken build caused by wrong format specifier
  test: add test-qmp-commands to make check
  qapi: remove print statements from test-qmp-commands
  test: remove qemu-ga reference
  vl.c: fix '-cpu ?' segfault

12 years agoMerge remote-tracking branch 'spice/spice.v51' into staging
Anthony Liguori [Mon, 26 Mar 2012 20:08:15 +0000 (15:08 -0500)]
Merge remote-tracking branch 'spice/spice.v51' into staging

* spice/spice.v51:
  ui/spice-display: use uintptr_t when casting qxl physical addresses
  ui/spice-display.c: Fix compilation warnings on 32 bit hosts

12 years agotest: remove qemu-ga reference
Michael Roth [Wed, 21 Mar 2012 15:10:46 +0000 (10:10 -0500)]
test: remove qemu-ga reference

This was added by mistake a while back.

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoMan page: Add -global description
Miroslav Rezanina [Wed, 21 Mar 2012 12:46:54 +0000 (13:46 +0100)]
Man page: Add -global description

There's only TODO information in qemu man page for -global option. This is a basic description of this option with simple example.

Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
v4:
 - break long line

v3:
 - add use case description
 - use prop instead of property

v2:
 - Use better value in example
Patch:
--
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agoqapi: add c_fun to escape function names
Federico Simoncelli [Tue, 20 Mar 2012 13:54:35 +0000 (13:54 +0000)]
qapi: add c_fun to escape function names

Signed-off-by: Federico Simoncelli <fsimonce@redhat.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
12 years agolibcacard/vcard_emul_nss: add warning for old coolkey
Alon Levy [Thu, 22 Mar 2012 18:00:36 +0000 (20:00 +0200)]
libcacard/vcard_emul_nss: add warning for old coolkey

Older coolkey versions (before the future fix of RHBZ 802435) have
a fake card reader created if no reader is detected during module
initialization. Warn libcacard users if the faulty coolkey is detected
by checking for the fake reader name "E-Gate 0 0".

Signed-off-by: Alon Levy <alevy@redhat.com>
12 years agolibcacard/vcard_emul_nss: handle no readers at startup
Alon Levy [Thu, 22 Mar 2012 17:58:58 +0000 (19:58 +0200)]
libcacard/vcard_emul_nss: handle no readers at startup

When starting with no readers, coolkey should show no slots (with
RHBZ 806038 fixed). Fix initialization to launch the event handling
thread for each module that isn't the internal module regardless of the
number of slots detected for it at initialization time, since slot
number may start as 0 and is dynamic.

RHBZ: 802435
Signed-off-by: Alon Levy <alevy@redhat.com>
12 years agolibcacard/vcard_emul_nss: don't stop thread when there are no slots
Alon Levy [Thu, 22 Mar 2012 17:44:49 +0000 (19:44 +0200)]
libcacard/vcard_emul_nss: don't stop thread when there are no slots

Signed-off-by: Alon Levy <alevy@redhat.com>
12 years agotrace-events: Fix broken build caused by wrong format specifier
Stefan Weil [Sat, 24 Mar 2012 07:13:34 +0000 (08:13 +0100)]
trace-events: Fix broken build caused by wrong format specifier

mem is an uint64_t value, so %lx was wrong.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agotest: add test-qmp-commands to make check
Michael Roth [Tue, 20 Mar 2012 00:18:26 +0000 (19:18 -0500)]
test: add test-qmp-commands to make check

All the deps are here but the test was never added to the list of tests
for make check

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agoqapi: remove print statements from test-qmp-commands
Michael Roth [Tue, 20 Mar 2012 00:18:25 +0000 (19:18 -0500)]
qapi: remove print statements from test-qmp-commands

This is necessary for nicer make check integration.

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agotest: remove qemu-ga reference
Michael Roth [Wed, 21 Mar 2012 15:10:46 +0000 (10:10 -0500)]
test: remove qemu-ga reference

This was added by mistake a while back.

Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agovl.c: fix '-cpu ?' segfault
Eduardo Habkost [Wed, 21 Mar 2012 12:33:40 +0000 (09:33 -0300)]
vl.c: fix '-cpu ?' segfault

Fix stupid copy&paste mistake at commit
ecf40beae7dcbb057d4f115207f9d8276832a774: I moved code around but kept
"optarg" on the cpu_list() call.

Reported-by: Jiri Denemark <jdenemar@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
12 years agoqemu-ga: fix bsd build, and re-org linux-specific implementations
Michael Roth [Sun, 25 Mar 2012 18:59:41 +0000 (13:59 -0500)]
qemu-ga: fix bsd build, and re-org linux-specific implementations

12 years agotcg-sparc: Add debug_frame support.
Richard Henderson [Sat, 24 Mar 2012 17:47:39 +0000 (10:47 -0700)]
tcg-sparc: Add debug_frame support.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotcg-hppa: Add debug_frame support.
Richard Henderson [Sat, 24 Mar 2012 17:47:38 +0000 (10:47 -0700)]
tcg-hppa: Add debug_frame support.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotcg: Allow ELF_HOST_FLAGS and ELF_OSABI overrides in gdb-jit.
Richard Henderson [Sat, 24 Mar 2012 17:47:37 +0000 (10:47 -0700)]
tcg: Allow ELF_HOST_FLAGS and ELF_OSABI overrides in gdb-jit.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
12 years agotcg: Add debug_info to JIT ELF image.
Richard Henderson [Sat, 24 Mar 2012 17:47:36 +0000 (10:47 -0700)]
tcg: Add debug_info to JIT ELF image.

This allows us to actually supply a function name in softmmu builds;
gdb doesn't pick up the minimal symbol table otherwise.  Also add a
bit of documentation and statically generate more of the ELF image.

Signed-off-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Blue Swirl <blauwirbel@gmail.com>