Both XIVE and XICS interrupt mode are advertised for the moment.
Signed-off-by: Cédric Le Goater
---
hw/ppc/spapr.c | 25 -
1 file changed, 20 insertions(+), 5 deletions(-)
diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index e3567543e6e6..d05c83cdb322 100644
--- a/hw/ppc/
The XIVE object has its own set of qirqs which is to be used when the
XIVE exploitation interrupt mode is activated.
Signed-off-by: Cédric Le Goater
---
hw/intc/spapr_xive.c| 13 +
hw/ppc/spapr.c | 4
include/hw/ppc/spapr_xive.h | 1 +
include/hw/ppc/xive.
The XiveSource and sPAPRXive device models will be shared between the
emulated and the KVM mode. The difference will reside in the way the
memory regions are initialized and in the qemu_irq handler. Introduce
common realize routines to share some code.
Signed-off-by: Cédric Le Goater
---
hw/intc
These define a new capability and a new KVM device for the XIVE native
exploitation interrupt mode. New ioctls are also introduced to
initialize the KVM device and handle VM migration.
Signed-off-by: Cédric Le Goater
---
linux-headers/asm-powerpc/kvm.h | 18 ++
linux-headers/linu
On sPAPR, the creation of the interrupt presenter depends on some of
the machine attributes. When the XIVE exploitation interrupt mode is
available, this will get more complex. So provide a machine-level
helper to isolate the process and hide the details to the sPAPR core
realize function.
Signed-
Currently we mishandle emulation of the getdents syscall for the
case of a 64 bit guest on a 32 bit host -- it defaults into
the 'host and guest same size' codepath and generates incorrect
structures in the guest buffer.
We can't easily handle the 64-on-32 case using the host getdents
syscall, bec
As the VM is connected to the KVM interrupt device at the init of the
machine, KVM support is added for XICS or for XIVE but not both at the
same time. This should change later on when KVM resets are supported.
Signed-off-by: Cédric Le Goater
---
hw/intc/xics_kvm.c | 22 +-
h
This introduces a set of XIVE models specific to KVM. They handle the
synchronization of the state with KVM, for the monitor usage and for
the migration.
The TIMA and ESB MMIO regions are initialized differently under KVM.
'ram device' memory mappings, similarly to VFIO, are exposed to the
guest a
The XIVE models for the emulated and the KVM mode will have a lot in
common. Introduce some classes to handle the differences, mostly to
synchronize the state with KVM for the monitor and migration. This is
very much like XICS is doing.
Signed-off-by: Cédric Le Goater
---
hw/intc/spapr_xive.c
The interrupt controller KVM device needs to be destroyed and then
recreated accordingly with the interrupt mode negociated at CAS time.
A new KVM_DESTROY_DEVICE is required for this purpose along with the
necessary support in Linux/KVM.
This won't work without the vpcus being first disconnected f
Based on a patch by Martin Schrodt.
Gerd Hoffmann (4):
audio/hda: create millisecond timers that handle IO
audio/hda: turn some dprintfs into trace points
audio/hda: tweak timer adjust logic
audio/hda: detect output buffer overruns and underruns
include/hw/compat.h | 7 ++
hw/audio/h
The vCPUs are disconnected from the KVM device using a 'disable=1' as
last argument of the KVM_ENABLE_CAP ioctl. This is a bit hacky, we
should probably introduce a KVM_DISABLE_CAP ioctl.
Signed-off-by: Cédric Le Goater
---
hw/intc/spapr_xive_kvm.c | 55 ++
On the POWER9 processor, the XIVE interrupt controller can control
interrupt sources using MMIO to trigger events, to EOI or to turn off
the sources. Priority management and interrupt acknowledgment is also
controlled by MMIO in the presenter sub-engine.
These MMIO regions are exposed to guests in
This is simple model of the POWER9 XIVE interrupt controller for the
PowerNV machine. XIVE for baremetal is a complex controller and the
model only addresses the needs of the skiboot firmware. Support is
provided for :
* virtual structure descriptor tables describing the XIVE
internal tables sto
The XIVE MMIO regions should be set on the destination before the XIVE
sources are restored. This is currently handled at the machine level
because it depends on the KVM initialization to be done before
anything else.
But it has ugly consequences on MMU, which seems broken after migration :
Oop
which we will use in the sPAPR machine to reset the interrupt
controller of each CPU at the KVM level.
Signed-off-by: Cédric Le Goater
---
hw/intc/intc.c | 26 ++
include/hw/intc/intc.h | 21 +
2 files changed, 47 insertions(+)
diff --git a/hw
Drop the option-less display types (egl-headless, curses, none) from
parse_display(), so they'll be handled by parse_display_qapi().
Signed-off-by: Gerd Hoffmann
---
vl.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/vl.c b/vl.c
index 784c3fb795..39303d6b55 100644
--- a/vl.c
+++ b/vl.
The type of the interrupt presenter depends on the processor family,
POWER8 uses XICS and POWER9 uses XIVE. Provide a machine-level helper
to isolate the process and hide the details to the pnv core realize
function.
Signed-off-by: Cédric Le Goater
---
hw/ppc/pnv.c | 14 ++
h
Currently, the HDA device tries to sync itself with the QEMU audio
backend by waiting for the guest driver to handle buffer completion
interrupts. This causes the backend to often read too much data from the
device, as well as running out of data whenever the guest takes too long
to handle the inte
We will use it to get the CPU interrupt presenter in XIVE.
Signed-off-by: Cédric Le Goater
---
hw/ppc/pnv.c | 16
hw/ppc/ppc.c | 16
include/hw/ppc/ppc.h | 1 +
3 files changed, 17 insertions(+), 16 deletions(-)
diff --git a/hw/ppc/pnv.c b/hw/p
On Thu, 19 Apr 2018 14:33:18 +0200
Igor Mammedov wrote:
> On Thu, 19 Apr 2018 17:21:23 +1000
> David Gibson wrote:
>
> > If the -mem-path option is set, we attempt to map the guest's RAM from a
> > file in the given path; it's usually used to back guest RAM with hugepages.
> > If we're unable t
We have some jitter in the audio timer call frequency and buffer sizes.
So it is rather pointless trying to be very exact, effect is a constant
up+down adjustment. So adjust only in case we are off too much.
Signed-off-by: Gerd Hoffmann
---
hw/audio/hda-codec.c | 20 +---
hw/au
On 04/16/2018 06:29 AM, David Gibson wrote:
> On Thu, Apr 12, 2018 at 10:36:10AM +0200, Cédric Le Goater wrote:
>> On 04/12/2018 07:16 AM, David Gibson wrote:
>>> On Mon, Feb 12, 2018 at 09:55:17AM +1100, Benjamin Herrenschmidt wrote:
On Sun, 2018-02-11 at 19:08 +1100, David Gibson wrote:
On 19/04/2018 14:43, Liang, Cunming wrote:
>> 2. Memory barriers. Right now after updating the avail idx,
>> virtio does smp_wmb() and then the MMIO write. Normal hardware
>> drivers do wmb() which is an sfence. Can a PCI device read bypass
>> index write and see a stale index value?
>
> A compile
If some event caused some larger playback hickup the fine-grained timer
adjust isn't able to recover. Use a buffer overruns and underruns as
indicator for that. Reset timer adjust logic in case we detected one.
Signed-off-by: Gerd Hoffmann
---
hw/audio/hda-codec.c | 22 ++
On Wed 11 Apr 2018 08:54:13 PM CEST, Max Reitz wrote:
> When converting mirror's I/O to coroutines, we are going to need a point
> where these coroutines are created. mirror_perform() is going to be
> that point.
>
> Signed-off-by: Max Reitz
> Reviewed-by: Fam Zheng
> Reviewed-by: Vladimir Semen
Signed-off-by: Gerd Hoffmann
---
hw/audio/hda-codec.c | 9 -
hw/audio/trace-events | 4
2 files changed, 8 insertions(+), 5 deletions(-)
diff --git a/hw/audio/hda-codec.c b/hw/audio/hda-codec.c
index 9ee2f3d55f..13d4473f18 100644
--- a/hw/audio/hda-codec.c
+++ b/hw/audio/hda-codec.
Gerd Hoffmann (4):
ui: add qapi parser for -display
ui: switch trivial displays to qapi parser
ui: switch gtk display to qapi parser
ui: document non-qapi parser cases.
vl.c | 81 ++--
1 file changed, 41 insertions(+), 40 d
On 04/19/2018 02:58 PM, Cornelia Huck wrote:
> On Thu, 19 Apr 2018 14:33:18 +0200
> Igor Mammedov wrote:
>
>> On Thu, 19 Apr 2018 17:21:23 +1000
>> David Gibson wrote:
>>
>>> If the -mem-path option is set, we attempt to map the guest's RAM from a
>>> file in the given path; it's usually used
Marc-André Lureau writes:
> Hi
>
> On Thu, Apr 19, 2018 at 12:09 PM, Marc-André Lureau
> wrote:
>>> On the one hand, we provide safe conversion macros QOBJECT() and
>>> qobject_to(). By the way, shouting one but not the other is a bit ugly.
>>
>> QOBJECT is static upcast, the compiler will shou
Hi
On Thu, Apr 19, 2018 at 8:18 AM, Markus Armbruster wrote:
> Marc-André Lureau writes:
>
>> While it may be convenient to accept NULL value in
>> qobject_unref() (for similar reasons as free() accepts NULL), it is
>> not such a good idea for qobject_ref(), assert() on NULL. One place
>> relied
Add parse_display_qapi() function which parses the -display command line
using a qapi visitor for DisplayOptions. Wire up as default catch in
parse_display().
Improves the error message for unknown display types.
Also enables json as -display argument, i.e. -display "{ 'type': 'gtk' }"
Signed-o
Drop the gtk option parser from parse_display(), so parse_display_qapi()
will handle it instead.
Signed-off-by: Gerd Hoffmann
---
vl.c | 32
1 file changed, 32 deletions(-)
diff --git a/vl.c b/vl.c
index 39303d6b55..1123cd792e 100644
--- a/vl.c
+++ b/vl.c
@@ -21
Add comments to the cases not (yet) switched
over to parse_display_qapi().
Signed-off-by: Gerd Hoffmann
---
vl.c | 14 ++
1 file changed, 14 insertions(+)
diff --git a/vl.c b/vl.c
index 1123cd792e..d35c443b8c 100644
--- a/vl.c
+++ b/vl.c
@@ -2121,6 +2121,16 @@ static void parse_disp
On Tue, 17 Apr 2018 17:17:13 +1000
David Gibson wrote:
> af81cf323c1 "spapr: CPU hotplug support" added a direct call to
> spapr_cpu_reset() in spapr_cpu_init(), as well as registering it as a
> reset callback. That was in order to make sure that the reset function
> got called for a newly hotpl
Also dont assume x86_64 compiler can build i386 binaries.
Signed-off-by: Alex Bennée
Tested-by: Philippe Mathieu-Daudé
---
v2
- drop using system x86_64 compiler for i386 compiles
---
configure | 24
1 file changed, 16 insertions(+), 8 deletions(-)
diff --git a/conf
As an individual compiler may be able to support several targets with
the appropriate flags we need to expose this to the user as well.
Signed-off-by: Alex Bennée
---
configure | 8
1 file changed, 8 insertions(+)
diff --git a/configure b/configure
index 038fb9db7b..29af33716d 100755
-
This tests for a working docker installation without sudo and sets up
config-host.mak accordingly. This will be useful from cross compiling
things in the future.
Signed-off-by: Alex Bennée
---
configure | 17 +
1 file changed, 17 insertions(+)
diff --git a/configure b/configure
From: Peter Maydell
This is a helper function for the configure script. It replies yes,
sudo or no to inform the user if non-interactive docker support is
available. We trap the Exception to fail gracefully.
Signed-off-by: Alex Bennée
---
tests/docker/docker.py | 18 ++
1 file
This allows us to specify cross compilers for our guests. This is
useful for building test images/programs. Currently we re-run the
compile test for each target. I couldn't think of a way to cache the
value for a given arch without getting messier configure code.
The cross compiler for the guest i
Hi,
This is the second revision of my attempt to revive the tests/tcg
directory. You can find the first iteration here:
https://lists.gnu.org/archive/html/qemu-devel/2018-04/msg01361.html
Changes:
I've slightly re-jigged the Makefile inclusion to make
tests/tcg/Makefile very simple. All the T
We can build tests for the host system with the compiler that we have
selected.
Signed-off-by: Alex Bennée
---
configure | 16 +++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/configure b/configure
index 29af33716d..44d310a692 100755
--- a/configure
+++ b/configure
When calling our cross-compilation images we want to call something
other than the default cc.
Signed-off-by: Alex Bennée
---
v2
- use arg.cc default to simplify logic
---
tests/docker/docker.py | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/tests/docker/docker.py b/te
Un-comment the remaining tests. I removed the itimer value tests
because I'm fairly sure a re-arming timer will always have a different
value in it when you grab it.
Signed-off-by: Alex Bennée
---
tests/tcg/multiarch/linux-test.c | 20
1 file changed, 8 insertions(+), 12 del
Signed-off-by: Alex Bennée
Reviewed-by: Fam Zheng
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
tests/docker/docker.py | 4
1 file changed, 4 insertions(+)
diff --git a/tests/docker/docker.py b/tests/docker/docker.py
index 59bce9f19a..37bfa98005 100755
--- a/t
From: Fam Zheng
We have -Werror=missing-prototype, add a dummy prototype to avoid that
warning.
Signed-off-by: Fam Zheng
Reviewed-by: Thomas Huth
---
tests/tcg/i386/hello-i386.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/tcg/i386/hello-i386.c b/tests/tcg/i386/hello-i386.c
index
This doesn't add any additional tests but enables building the
multiarch tests for s390x.
Signed-off-by: Alex Bennée
Acked-by: Cornelia Huck
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
tests/tcg/s390x/Makefile.include | 2 ++
1 file changed, 2 insertions(+)
crea
This make is now invoked from each individual target make with the
appropriate CC and EXTRA_CFLAGS set for each guest. It then includes
additional Makefile.targets from:
- tests/tcg/multiarch (always)
- tests/tcg/$(TARGET_BASE_ARCH) (if available)
- tests/tcg/$(TARGET_NAME)
The order is imp
Signed-off-by: Fam Zheng
[AJB: add if args.paths check]
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
v2
- add if args.paths check to avoid iterating null argument
---
tests/docker/docker.py | 25 +
1 file changed
From: Fam Zheng
To be more accurate on its purpose and make code that looks for a certain
target out of this variable more readable.
Signed-off-by: Fam Zheng
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
Makefile | 20 ++--
configure
Remove dependencies on QEMU's source tree and build directly.
Signed-off-by: Alex Bennée
---
tests/tcg/i386/test-i386-fprem.c | 12 +---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/tests/tcg/i386/test-i386-fprem.c b/tests/tcg/i386/test-i386-fprem.c
index 1a71623204..66f
The sources for x86_64 are shared in the i386 directory which will be
included thanks to TARGET_BASE_ARCH. However not all sources build so
we need to filter out the ones we can't build in the 64 bit world.
Signed-off-by: Alex Bennée
---
tests/tcg/i386/Makefile.target | 2 +-
tests/tcg/x86_64
While you can construct a compile command that does work using the
x86_64 host compiler that most people use this is flakey. Different
distros handle this is different ways so we default to using a known
good i386 compiler via docker.
Signed-off-by: Alex Bennée
---
tests/tcg/i386/Makefile.includ
This prevents us trying to build fp-test for 32 bit guests as they
would fail due to pollution from config-host.h when including osdep.h.
As keeping softfloat.o would be useful maybe we can conditionally
disabled it?
Signed-off-by: Alex Bennée
---
tests/tcg/multiarch/Makefile.target | 2 ++
1 fi
Nothing much here yet.
Signed-off-by: Alex Bennée
---
tests/tcg/aarch64/Makefile.target | 5 +
1 file changed, 5 insertions(+)
create mode 100644 tests/tcg/aarch64/Makefile.target
diff --git a/tests/tcg/aarch64/Makefile.target
b/tests/tcg/aarch64/Makefile.target
new file mode 100644
index
From: Fam Zheng
To keep the compiler happy, and to fit in our buildsys flags:
- Make local functions "static"
- #ifdef out unused functions
- drop cutils/osdep dependencies
Signed-off-by: Fam Zheng
[AJB: drop cutils/osdep dependencies]
Signed-off-by: Alex Bennée
Reviewed-by: Thomas Huth
---
Signed-off-by: Alex Bennée
---
tests/tcg/i386/test-i386.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/tcg/i386/test-i386.c b/tests/tcg/i386/test-i386.c
index caef4da176..a29b41e764 100644
--- a/tests/tcg/i386/test-i386.c
+++ b/tests/tcg/i386/test-i386.c
@@ -2258,7 +2
These only need to be built for i386 guests. This includes a stub
tests/tcg/i386/Makfile.target which absorbs some of what was in
tests/tcg/Makefile.
Signed-off-by: Alex Bennée
Reviewed-by: Thomas Huth
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
v2
- move VPATH
Currently this just enables building the multiarch tests.
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
tests/tcg/ppc64le/Makefile.include | 2 ++
1 file changed, 2 insertions(+)
create mode 100644 tests/tcg/ppc64le/Makefile.include
diff
From: Fam Zheng
It has some basic *-devel.i686 packages to be used with "gcc -m32" as a
32 bit cross build environment.
Signed-off-by: Fam Zheng
[AJB: add glibc-static]
Signed-off-by: Alex Bennée
---
v2
- include static glibc package as well
---
tests/docker/dockerfiles/fedora-i386-cross.d
It doesn't build for i386 easily and we probably need a better more
methodical test for SSE et all in QEMU.
Signed-off-by: Alex Bennée
---
tests/tcg/i386/Makefile.target | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/tcg/i386/Makefile.target b/tests/tcg/i386/Makefile.t
From: "Emilio G. Cota"
This will allow us to run correctness tests against our
FP implementation. The test can be run in two modes (called
"testers"): host and soft. With the former we check the results
and FP flags on the host machine against the model.
With the latter we check QEMU's fpu primit
This will ensure all linux-user targets build their guest test
programs.
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
v2
- use -include instead of complex macro stuff
- also include TARGET_BASE_ARCH/Makefile
---
tests/Makefile.include
On 19.04.2018 15:34, Christian Borntraeger wrote:
>
>
> On 04/19/2018 02:58 PM, Cornelia Huck wrote:
>> On Thu, 19 Apr 2018 14:33:18 +0200
>> Igor Mammedov wrote:
>>
>>> On Thu, 19 Apr 2018 17:21:23 +1000
>>> David Gibson wrote:
>>>
If the -mem-path option is set, we attempt to map the gue
This doesn't add any additional tests but enables building the
multiarch tests for MIPS using docker cross compilers. We don't have a
cross compiler for mips64 big endian though.
Signed-off-by: Alex Bennée
---
tests/tcg/mips/Makefile.include | 17 +
1 file changed, 17 insertions(
We will want to build these for all supported guest architectures so
lets move them all into one place. We also drop test_path at this
point because it needs qemu utils and glib bits which is hard to
support for cross compiling.
Signed-off-by: Alex Bennée
Reviewed-by: Thomas Huth
---
v2
- mov
Define this in one place to make it easy to re-use.
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
---
tests/docker/Makefile.include | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/tests/docker/Makefile.include b/tests/docker/Makefile.include
index
These only need to be built for MIPS guests.
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
tests/tcg/README | 11 ---
tests/tcg/mips/README | 7 +++
tests/tcg/{ => mips}/hello-mips.c | 0
3 files
We need to rename the source file to a .S so we can do a single-line
assemble and link invocation. We also specify the additional CFLAGS
for the compile as it's a non-standard ARM binary.
Signed-off-by: Alex Bennée
---
tests/tcg/arm/Makefile.target | 5 +
tests/tcg/a
Signed-off-by: Alex Bennée
---
tests/tcg/mips/Makefile.target | 19 +++
1 file changed, 19 insertions(+)
create mode 100644 tests/tcg/mips/Makefile.target
diff --git a/tests/tcg/mips/Makefile.target b/tests/tcg/mips/Makefile.target
new file mode 100644
index 00..71f0c2dd
This runs through the usual float to float conversions and crucially
also runs with ARM Alternative Half Precision Format.
Signed-off-by: Alex Bennée
---
tests/tcg/aarch64/Makefile.target |8 +
tests/tcg/aarch64/fcvt.c | 260 ++
tests/tcg/aarch64/fcvt.out| 1305
The compiler complains about the old __mode__ style attributes.
Signed-off-by: Alex Bennée
---
tests/tcg/i386/test-i386.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/tcg/i386/test-i386.c b/tests/tcg/i386/test-i386.c
index cae6a7773a..caef4da176 100644
--- a/test
In commit 95695effe8caa552b8f2 we changed the v7M/v8M stack
pop code to use a new v7m_stack_read() function that checks
whether the read should fail due to an MPU or bus abort.
We missed one call though, the one which reads the signature
word for the callee-saved register part of the frame.
Correc
The test-i386 test case is a little special as it includes assembler
files. Add the additional compile magic to assemble these bits and
link them to the final binary.
Signed-off-by: Alex Bennée
Reviewed-by: Thomas Huth
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
Signed-off-by: Alex Bennée
---
tests/tcg/aarch64/Makefile.target | 2 +-
tests/tcg/aarch64/sysregs.c | 14 ++
2 files changed, 15 insertions(+), 1 deletion(-)
create mode 100644 tests/tcg/aarch64/sysregs.c
diff --git a/tests/tcg/aarch64/Makefile.target
b/tests/tcg/aarch64/Ma
The compile for fp-test is a little funky as we need to include a
build of QEMU's softfloat as well. There is still a problem caused by
the inclusion of config-host.h by osdep.h as it will trigger problems
like:
In file included from
/home/alex/lsrc/qemu/qemu.git/include/qemu/bitops.h:16:0,
Now all the build infrastructure is in place we can build tests for
each guest that we support. That support mainly depends on having
cross compilers installed or docker setup. To keep all the logic for
that together we put the rules in tests/tcg/Makefile.include and
include it from the main Makefi
This adds support for the various FOO-to-BAR conversions to fp-test.
It also defines PREC_HALF although currently that is not used and will
need compile time detection for _Float16 support.
I've added a small test file for testing against regressions.
Signed-off-by: Alex Bennée
---
tests/fp/fp-
Remove a stale TODO comment -- we have now made the arm_ldl_ptw()
and arm_ldq_ptw() functions propagate physical memory read errors
out to their callers.
Signed-off-by: Peter Maydell
---
target/arm/helper.c | 8 +---
1 file changed, 1 insertion(+), 7 deletions(-)
diff --git a/target/arm/hel
From: "Emilio G. Cota"
To ease the cross-compilation of tests that do not use glib.
Signed-off-by: Emilio G. Cota
---
include/qemu/osdep.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index 41658060a7..fabaf73007 100644
--- a/include/qemu/os
On 04/19/2018 12:51 PM, KONRAD Frederic wrote:
When compiling on a machine without libfdt installed the configure script
should try to get libfdt from the git or should die because otherwise
CONFIG_LIBFDT is not set and the build process end in an error in the link
phase.. eg:
hw/riscv/virt.o:
From: "Emilio G. Cota"
To ease the cross-compilation of softfloat.o.
Signed-off-by: Emilio G. Cota
---
fpu/softfloat.c | 4
1 file changed, 4 insertions(+)
diff --git a/fpu/softfloat.c b/fpu/softfloat.c
index 70e0c40a1c..a2a5d4ea33 100644
--- a/fpu/softfloat.c
+++ b/fpu/softfloat.c
@@ -8
These only need to be built for ARM guests.
Signed-off-by: Alex Bennée
Reviewed-by: Thomas Huth
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
v2
- do VPATH manipulation in tests/tcg/arm/Makefile.target
- merge with fix hello-arm test
---
tests/tcg/README
On Wed 11 Apr 2018 08:54:17 PM CEST, Max Reitz wrote:
> With this, the mirror_top_bs is no longer just a technically required
> node in the BDS graph but actually represents the block job operation.
>
> Also, drop MirrorBlockJob.source, as we can reach it through
> mirror_top_bs->backing.
>
> Signe
Hi
On Thu, Apr 19, 2018 at 8:16 AM, Markus Armbruster wrote:
> Marc-André Lureau writes:
>
>> Now that we can safely call QOBJECT() on QObject * as well as its
>> subtypes, we can have macros qobject_ref() / qobject_unref() that work
>> everywhere instead of having to use QINCREF() / QDECREF() f
On Wed 11 Apr 2018 08:54:24 PM CEST, Max Reitz wrote:
> This patch allows the user to specify whether to use active or only
> background mode for mirror block jobs. Currently, this setting will
> remain constant for the duration of the entire block job.
>
> Signed-off-by: Max Reitz
Reviewed-by:
On Wed 11 Apr 2018 08:54:25 PM CEST, Max Reitz wrote:
> Signed-off-by: Max Reitz
> Reviewed-by: Fam Zheng
Reviewed-by: Alberto Garcia
Berto
Hi,
This series aims to get rid of the distinction between QObject, that
must use qobject_incref/qobject_decref and its various derived types
that have to use QINCREF/QDECREF. Instead, replace it with
qobject_ref/qobject_unref for all types.
v6: after Eric and Markus reviews
- remove double-under
All QObject types have the base QObject as their first field. This
allows the simplification of qobject_to().
This explicitly guarantees that existing casts work correctly (even
though we'd prefer to get rid of such casts in any location except the
qobject.h macros)
Signed-off-by: Marc-André Lure
While it may be convenient to accept NULL value in
qobject_unref() (for similar reasons as free() accepts NULL), it is
not such a good idea for qobject_ref(): now assert() on NULL.
Some places relied on that behaviour (the monitor request id for
example), and it's best to be explicit that NULL is
For convenience and clarity, make it possible to call qobject_ref() at
the time when the reference is associated with a variable, or
argument, by making qobject_ref() return the same pointer as given.
Signed-off-by: Marc-André Lureau
Reviewed-by: Eric Blake
---
include/qapi/qmp/qnull.h |
By moving the base fields to a QObjectBase_, QObject can be a type
which also has a 'base' field. This allows writing a generic QOBJECT()
macro that will work with any QObject type, including QObject
itself. The container_of() macro ensures that the object to cast has a
QObjectBase_ base field, giv
On Thu, Apr 19, 2018 at 03:02:40PM +0200, Paolo Bonzini wrote:
> On 19/04/2018 14:43, Liang, Cunming wrote:
> >> 2. Memory barriers. Right now after updating the avail idx,
> >> virtio does smp_wmb() and then the MMIO write. Normal hardware
> >> drivers do wmb() which is an sfence. Can a PCI devic
On 04/19/2018 10:01 AM, Marc-André Lureau wrote:
> All QObject types have the base QObject as their first field. This
> allows the simplification of qobject_to().
>
> This explicitly guarantees that existing casts work correctly (even
> though we'd prefer to get rid of such casts in any location e
On Wed 11 Apr 2018 08:54:18 PM CEST, Max Reitz wrote:
> Currently, bdrv_replace_node() refuses to create loops from one BDS to
> itself if the BDS to be replaced is the backing node of the BDS to
> replace it: Say there is a node A and a node B. Replacing B by A means
> making all references to B
On Thu, Apr 19, 2018 at 5:20 PM, Eric Blake wrote:
> On 04/19/2018 10:01 AM, Marc-André Lureau wrote:
>> All QObject types have the base QObject as their first field. This
>> allows the simplification of qobject_to().
>>
>> This explicitly guarantees that existing casts work correctly (even
>> tho
On Wed, Apr 18, 2018 at 7:39 PM, Viktor Prutyanov
wrote:
> This patch adds Windows crashdumping feature. Now QEMU can produce ELF-dump
> containing Windows crashdump header, which can help to convert to a valid
> WinDbg-understandable crashdump file, or immediately create such file.
> The crashdum
On 04/19/2018 10:01 AM, Marc-André Lureau wrote:
> Now that we can safely call QOBJECT() on QObject * as well as its
> subtypes, we can have macros qobject_ref() / qobject_unref() that work
> everywhere instead of having to use QINCREF() / QDECREF() for QObject
> and qobject_incref() / qobject_decr
On Thu, 2018-04-19 at 16:29 +1000, David Gibson wrote:
> Currently the "pseries" machine type will (usually) advertise
> different pagesizes to the guest when running under KVM and TCG, which
> is not how things are supposed to work.
>
> This comes from poor handling of hardware limitations which
On 04/19/2018 10:01 AM, Marc-André Lureau wrote:
> For convenience and clarity, make it possible to call qobject_ref() at
> the time when the reference is associated with a variable, or
> argument, by making qobject_ref() return the same pointer as given.
>
> Signed-off-by: Marc-André Lureau
> Re
1 - 100 of 414 matches
Mail list logo
