On Fri, 09/01 16:29, Philippe Mathieu-Daudé wrote:
> On 08/30/2017 12:34 AM, Fam Zheng wrote:
> > On Tue, 08/29 14:34, Philippe Mathieu-Daudé wrote:
> > > > +self._args = [ \
> > > > +"-nodefaults", "-m", "2G",
> > > > +"-cpu", "host",
> > > > +"-netdev",
On Fri, Sep 01, 2017 at 09:38:12AM +0200, Thomas Huth wrote:
> The target/ppc/STATUS file has seen its last real update 10 years
> ago - so the information in there is not up to date anymore. Since
> nobody seems to care about this file, let's simply remove it.
>
> Signed-off-by: Thomas Huth
App
Hi Halil,
On 2017/9/1 19:45, Halil Pasic wrote:
>
>
> On 09/01/2017 02:47 AM, Longpeng (Mike) wrote:
>> Ping...
>>
>> Stefan, Halil, do you have any suggestion ?
>>
>
> Hi Longpeng,
>
> I've ran trough your patch, and it reads much better that
> what I recall v18 used to read like. Because it
On Tue, Aug 29, 2017 at 02:33:12 -0400, Pranith Kumar wrote:
> Currently, we cannot use mttcg for running strong memory model guests
> on weak memory model hosts due to missing ordering semantics.
>
> We implicitly generate fence instructions for stronger guests if an
This confused me. By "We imp
On 09/01/2017 03:03 PM, Eric Blake wrote:
There was only one caller; it's easier to inline things.
Signed-off-by: Eric Blake
Acked-by: Philippe Mathieu-Daudé
---
tests/libqtest.c | 22 ++
1 file changed, 6 insertions(+), 16 deletions(-)
diff --git a/tests/libqtest.c
On 09/01/2017 03:03 PM, Eric Blake wrote:
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. It turns out that with the
recent changes to libqos, fw_cfg-test was not even using
global_qtest an
Quoting Michael Roth (2017-08-31 13:00:00)
> Quoting Michael Roth (2017-08-28 19:13:35)
> > Hi everyone,
> >
> > The following new patches are queued for QEMU stable v2.9.1:
> >
> >
> > https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_mdroth_qemu_commits_stable-2D2.9-2Dstaging&d
On 09/01/2017 02:03 PM, Eric Blake wrote:
> We already have another qtest_init() in the tree, for the
> top-level qtest.c device; having two functions with different
> signatures is confusing. Rename the libqtest version to
> qtest_start() to eliminate the duplication.
>
> Signed-off-by: Eric B
On 09/01/2017 02:03 PM, Eric Blake wrote:
> Remove the trivial wrappers qtest_start() and qtest_end(), to make
> it obvious in the rest of the testsuite where we are still relying on
> global_qtest. Doing this makes it easier to see what remaining
> cleanups will be needed if we don't want an im
On 09/01/2017 02:03 PM, Eric Blake wrote:
> Managing parallel connections to two different monitors via
> the implicit global_qtest makes it hard to copy-and-paste code
> to tests that are not aware of the implicit state; the
> management of global_qtest is even harder to follow because
> it was
On 09/01/2017 02:03 PM, Eric Blake wrote:
> When initializing a QPCIBus, track which QTestState the bus is
> associated with (so that a later patch can then explicitly use
> that test state for all communication on the bus, rather than
> blindly relying on global_qtest). Update the initializatio
On 09/01/2017 02:03 PM, Eric Blake wrote:
> Drop one more client of global_qtest by teaching all fw_cfg test
> functionality (invoked through alloc-pc) to pass in an explicit
> QTestState, adjusting all callers. In particular, fw_cfg-test
> had to reorder things to create the test state prior to
On 08/30/2017 10:02 PM, Fam Zheng wrote:
> On Thu, 08/31 00:45, Ishani Chugh wrote:
>> This patch is the test for incremental backup implementation in Backup tool.
>> The test employs two basic subtests:
>> 1) Backing up an empty guest and comparing it with base image.
>> 2) Writing a pattern to
I suggest as your commit message here, something like:
"backup: Add QEMU backup tool"
A good commit message should say what applying the patch will do:
"What will happen if I apply this patch?"
"It will 'Add QEMU backup tool'"
On 08/30/2017 01:14 PM, Ishani Chugh wrote:
> qemu-backup will be a
On Fri, Sep 1, 2017 at 3:17 PM, Eric Blake wrote:
> On 09/01/2017 05:13 PM, Alistair Francis wrote:
>> 'On Fri, Sep 1, 2017 at 3:50 AM, Gerd Hoffmann wrote:
>>> Drop pixman submodule and support for the "internal" pixman build.
>>> pixman should be reasonable well established meanwhile that we do
On 09/01/2017 05:13 PM, Alistair Francis wrote:
> 'On Fri, Sep 1, 2017 at 3:50 AM, Gerd Hoffmann wrote:
>> Drop pixman submodule and support for the "internal" pixman build.
>> pixman should be reasonable well established meanwhile that we don't
>
>
> Is there any use case for '--with-system-pix
'On Fri, Sep 1, 2017 at 3:50 AM, Gerd Hoffmann wrote:
> Drop pixman submodule and support for the "internal" pixman build.
> pixman should be reasonable well established meanwhile that we don't
s/reasonable/reasonably/g
s/meanwhile/meaning/g
> need the fallback submodule any more. While being a
Public bug reported:
The raspberry pi 3 B uses a 64-bit instruction set. It would be useful
if qemu could emulate the boardn
** Affects: qemu
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to
Linux uses SRAT to determine the maximum memory in a system, which is
used to determine whether to use the swiotlb for IOMMU or not for a
device that supports only 32 bits of addresses.
When there is no NUMA configuration, qemu will not build SRAT. And when
memory hotplug is done, some Linux devic
Hi,
I am having trouble using the simpleparser.py script in QEMU. Hopefully, I
am doing something incorrect and you can quickly set me back on course.
I have generated a trace file as per the instructions in
docs/devel/tracing.txt. However, whenever I try to run the simpleparser.py
script I get
This is the initial version of the Inter Processor Interrupt device.
Signed-off-by: Alistair Francis
---
default-configs/microblaze-softmmu.mak | 1 +
hw/intc/Makefile.objs | 1 +
hw/intc/xlnx-zynqmp-ipi.c | 377 +
include/hw/int
The Xilinx ZynqMP SoC has two main processing systems in it. The ARM
processing system (which is already modeled in QEMU) and the MicroBlaze
Power Management Unit (PMU). This is the inital work for adding support
for the PMU.
The PMU susbsystem runs along side the ARM system on hardware, but due
t
This series adds the ZynqMP Power Management Unit (PMU) machine with basic
functionality.
The machine only has the CPU and memory connected, but that is enough
to run some of the ROM code on the machine.
The series also adds the IPI device and connects it to the ZynqMP ARM
side and the ZynqMP PM
Connect the MicroBlaze CPU and the ROM and RAM memory regions.
Signed-off-by: Alistair Francis
---
hw/microblaze/xlnx-zynqmp-pmu.c | 65 +++--
1 file changed, 63 insertions(+), 2 deletions(-)
diff --git a/hw/microblaze/xlnx-zynqmp-pmu.c b/hw/microblaze/xlnx-
In preperation for having an ARM and MicroBlaze ZynqMP machine let's
split out the current ARM specific config options.
Signed-off-by: Alistair Francis
---
default-configs/aarch64-softmmu.mak | 1 +
hw/arm/Makefile.objs| 2 +-
hw/display/Makefile.objs| 2 +-
hw/dma/M
Signed-off-by: Alistair Francis
---
hw/arm/xlnx-zynqmp.c | 14 ++
include/hw/arm/xlnx-zynqmp.h | 2 ++
2 files changed, 16 insertions(+)
diff --git a/hw/arm/xlnx-zynqmp.c b/hw/arm/xlnx-zynqmp.c
index 9eceadbdc8..1988cb47f4 100644
--- a/hw/arm/xlnx-zynqmp.c
+++ b/hw/arm/xlnx
Signed-off-by: Alistair Francis
---
hw/microblaze/xlnx-zynqmp-pmu.c | 28
1 file changed, 28 insertions(+)
diff --git a/hw/microblaze/xlnx-zynqmp-pmu.c b/hw/microblaze/xlnx-zynqmp-pmu.c
index 33584cfa4d..a6c14495ec 100644
--- a/hw/microblaze/xlnx-zynqmp-pmu.c
+++ b/
On Fri, Sep 01, 2017 at 09:51:15AM -0700, Alistair Francis wrote:
> diff --git a/target/mips/kvm.c b/target/mips/kvm.c
> index 3317905e71..a23aa438d2 100644
> --- a/target/mips/kvm.c
> +++ b/target/mips/kvm.c
> @@ -95,11 +95,11 @@ void kvm_mips_reset_vcpu(MIPSCPU *cpu)
> CPUMIPSState *env = &c
On 08/29/2017 06:31 PM, Philippe Mathieu-Daudé wrote:
> Hi Richard,
>
> I can't find anything to say about this patch... Hardcore stuff.
> Some part could be more a bit more verbose but after a while focusing it makes
> sens.
> I wonder how long it took you to write this :) "roughly 2h"
Not quite
On 08/31/2017 03:42 AM, Fam Zheng wrote:
Reviewed-by: Stefan Hajnoczi
Reviewed-by: Philippe Mathieu-Daudé
Signed-off-by: Fam Zheng
---
.gitignore | 2 ++
1 file changed, 2 insertions(+)
diff --git a/.gitignore b/.gitignore
index cf65316863..643e23e515 100644
--- a/.gitignore
+++ b/.gitigno
On 08/31/2017 03:42 AM, Fam Zheng wrote:
Signed-off-by: Fam Zheng
---
scripts/archive-source.sh | 29 +
1 file changed, 29 insertions(+)
create mode 100755 scripts/archive-source.sh
diff --git a/scripts/archive-source.sh b/scripts/archive-source.sh
new file mode
On 09/01/2017 03:10 PM, Brian Steffens wrote:
> This adds a boolean option called 'shared' to block devices. It defaults
> to off/false. When enabled for a particular block device, the 'shared' option
> causes the block migration code to skip over syncing of that device. This
> allows controlling e
On 09/01/2017 10:21 AM, Fam Zheng wrote:
> A few tweaks to the docker images and running commands allow us to run
> qemu-iotests in the fedora container. This will be added to the patchew test
> cases later.
>
> v3: Add nbd, luks and limit to quick group.
>
> v2: Add nettle-devel because a few io
This adds a boolean option called 'shared' to block devices. It defaults
to off/false. When enabled for a particular block device, the 'shared' option
causes the block migration code to skip over syncing of that device. This
allows controlling exactly which block devices get synced during a migrati
On 09/01/2017 09:33 AM, Daniel P. Berrange wrote:
> On Fri, Sep 01, 2017 at 09:26:30AM -0500, Eric Blake wrote:
>> On 09/01/2017 05:54 AM, Daniel P. Berrange wrote:
>>> Two recently added iotests are currently breaking 'check -luks',
>>> fix the easy one and disable the hard one.
>>>
>>> Daniel P.
Hi Fam,
On 08/31/2017 03:42 AM, Fam Zheng wrote:
This adds a 32bit guest.
The official LTS cloud image is downloaded and initialized with
cloud-init.
Signed-off-by: Fam Zheng
---
tests/vm/ubuntu.i386 | 88
1 file changed, 88 insertions(+
On 09/01/2017 05:39 AM, Thomas Huth wrote:
/* Start with a drive used by a device that unplugs instantaneously */
-qtest_start("-drive if=none,id=drive0,file=null-co://,format=raw"
-" -device virtio-scsi-pci"
-" -device scsi-hd,drive=driv
On 08/30/2017 12:34 AM, Fam Zheng wrote:
On Tue, 08/29 14:34, Philippe Mathieu-Daudé wrote:
+self._args = [ \
+"-nodefaults", "-m", "2G",
+"-cpu", "host",
+"-netdev", "user,id=vnet,hostfwd=:0.0.0.0:0-:22",
+"-device", "virtio-net-pci,netdev
On 09/01/2017 07:07 AM, Eduardo Otubo wrote:
> When not available, isa-fdc falls into assert instead of proper error
> exit. This patch fixes this behavior.
>
> Signed-off-by: Eduardo Otubo
> ---
> hw/block/fdc.c | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/hw
On 2017-08-31 15:14, Daniel P. Berrange wrote:
[...]
> NB, some distros have already changed /usr/bin/python to point to py3
> so all the files with #!/usr/bin/python in QEMU are already broken on
> those distros.
For what it's worth, I (at least used to) build qemu regularly on Arch
Linux and t
On 09/01/2017 03:03 PM, Eric Blake wrote:
Drop one more client of global_qtest by teaching all fw_cfg test
functionality (invoked through alloc-pc) to pass in an explicit
QTestState, adjusting all callers. In particular, fw_cfg-test
had to reorder things to create the test state prior to creatin
On 09/01/2017 03:03 PM, Eric Blake wrote:
Now that the QPCIBus tracks QTestState, use that state instead
of an implicit reliance on global_qtest.
Signed-off-by: Eric Blake
Reviewed-by: Philippe Mathieu-Daudé
---
tests/libqos/pci-pc.c| 47 --
Signed-off-by: Eric Blake
---
Fix older gcc duplicate typedef warning [patchew]
---
tests/libqos/fw_cfg.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/libqos/fw_cfg.h b/tests/libqos/fw_cfg.h
index 396dd4ee1e..0353416af0 100644
--- a/tests/libqos/fw_cfg.h
+++ b/tests/
Hi Eric,
On 09/01/2017 03:03 PM, Eric Blake wrote:
When initializing a QPCIBus, track which QTestState the bus is
associated with (so that a later patch can then explicitly use
that test state for all communication on the bus, rather than
blindly relying on global_qtest). Update the initializat
For s390x, the hang definitely still occurs using a chroot with qemu-
s390x-static copied in and no QEMU_LD_PREFIX. I'm not in a good position
to test other big-endian architectures though.
--
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QE
>>> On 8/28/2017 at 6:13 PM, wrote:
> Hi everyone,
>
> The following new patches are queued for QEMU stable v2.9.1:
>
> https://github.com/mdroth/qemu/commits/stable-2.9-staging
>
> The release is planned for 2017-09-07:
>
> http://wiki.qemu.org/Planning/2.9
>
> Please respond here or C
On 09/01/2017 06:56 AM, Stefan Hajnoczi wrote:
> On Thu, Aug 31, 2017 at 08:14:53PM -0400, John Snow wrote:
>> Wherever possible, replace all printfs with proper tracing.
>> In most places I've tried to do a straight replacement, but
>> forthcoming patches may calibrate the tracing to be a little
On 09/01/2017 01:31 PM, no-re...@patchew.org wrote:
> Hi,
>
> This series failed automatic build test. Please find the testing commands and
> their output below. If you have docker installed, you can probably reproduce
> it
> locally.
>
> CC tests/libqos/pci.o
> CC tests/libqos/fw_
On 09/01/2017 10:22 AM, Fam Zheng wrote:
> Signed-off-by: Fam Zheng
> ---
> tests/docker/test-block | 22 ++
> 1 file changed, 22 insertions(+)
> create mode 100755 tests/docker/test-block
>
> diff --git a/tests/docker/test-block b/tests/docker/test-block
> new file mode 100
Hi,
This series failed automatic build test. Please find the testing commands and
their output below. If you have docker installed, you can probably reproduce it
locally.
Message-id: 20170901180340.30009-1-ebl...@redhat.com
Subject: [Qemu-devel] [PATCH v6 00/29] Preliminary libqtest cleanups
Type
On Fri, Sep 01, 2017 at 10:25:42AM -0600, Jan Beulich wrote:
> Xen and qemu having identical #define-s (with different names) is a
> strong hint that these should have been part of the public interface
> from the very start. Use them if they're available, falling back to
> privately defined values
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. Adjust the helper code to
use explicit state instead, and update all callers.
Signed-off-by: Eric Blake
---
tests/acpi-utils.h | 65 +
Remove the trivial wrappers qtest_start() and qtest_end(), to make
it obvious in the rest of the testsuite where we are still relying on
global_qtest. Doing this makes it easier to see what remaining
cleanups will be needed if we don't want an implicit dependency
on global state. Many tests can a
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. Adjust the helper code to
use explicit state instead, and update all callers.
Fix some trailing whitespace while touching the file.
Signed-of
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. Although qmp-test does not
maintain parallel qtest connections, it was the last test
assigning to global_qtest. It's just as easy to be explic
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. It turns out that with the
recent changes to libqos, fw_cfg-test was not even using
global_qtest any more. Avoid a pointless strdup while at i
Several callers were formatting a string into a temporary
variable before calling qtest_init(); factor that into the
common code.
Signed-off-by: Eric Blake
---
tests/libqtest.h | 7 ---
tests/libqtest.c | 13 ++--
tests/bios-tables-test.c | 7 +--
tests/postcopy
Managing parallel connections to two different monitors via
the implicit global_qtest makes it hard to copy-and-paste code
to tests that are not aware of the implicit state. Now that
libqos is fully-explicit, it is easy to expand the rest of
vhost-user-test to also use explicit state everywhere.
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. Although bios-tables-test does
not maintain parallel qtest connections, it's just as easy to be
explicit about the state; once all tests have b
As a general rule, we prefer avoiding implicit global state
because it makes code harder to safely copy and paste without
thinking about the global state. Improve this test to be
explicit about the state.
Signed-off-by: Eric Blake
---
tests/wdt_ib700-test.c | 63 +++-
When initializing a QPCIBus, track which QTestState the bus is
associated with (so that a later patch can then explicitly use
that test state for all communication on the bus, rather than
blindly relying on global_qtest). Update the initialization
functions to take another parameter, and update al
Managing parallel connections to two different monitors via
the implicit global_qtest makes it hard to copy-and-paste code
to tests that are not aware of the implicit state. We are
already half-using the explicit APIs, so fix the remaining
spots to also be explicit about which connection is in use
Managing parallel connections to two different monitors via
the implicit global_qtest makes it hard to copy-and-paste code
to tests that are not aware of the implicit state. Since we
have already fixed qpci to avoid global_qtest, we can now
simplify by not using global_qtest anywhere in ivshmem-te
Now that the QPCIBus tracks QTestState, use that state instead
of an implicit reliance on global_qtest.
Signed-off-by: Eric Blake
---
tests/libqos/pci-pc.c| 47 ---
tests/libqos/pci-spapr.c | 20 ++--
2 files changed, 34 insertions(
There was only one caller; it's easier to inline things.
Signed-off-by: Eric Blake
---
tests/libqtest.c | 22 ++
1 file changed, 6 insertions(+), 16 deletions(-)
diff --git a/tests/libqtest.c b/tests/libqtest.c
index 5d16351e24..b6dd26e54a 100644
--- a/tests/libqtest.c
+++ b
We already have another qtest_init() in the tree, for the
top-level qtest.c device; having two functions with different
signatures is confusing. Rename the libqtest version to
qtest_start() to eliminate the duplication.
Signed-off-by: Eric Blake
---
tests/libqtest.h | 8
Drop one more client of global_qtest by teaching all fw_cfg test
functionality (invoked through alloc-pc) to pass in an explicit
QTestState, adjusting all callers. In particular, fw_cfg-test
had to reorder things to create the test state prior to creating
the fw_cfg.
Signed-off-by: Eric Blake
--
Don't open-code something that has a convenient helper available.
Signed-off-by: Eric Blake
Reviewed-by: Markus Armbruster
Reviewed-by: Philippe Mathieu-Daudé
---
tests/numa-test.c | 21 +++--
1 file changed, 3 insertions(+), 18 deletions(-)
diff --git a/tests/numa-test.c b/t
From: Markus Armbruster
We still use hacks like qmp("") to wait for an event, even though we
have qmp_eventwait() since commit 8fe941f, and qmp_eventwait_ref()
since commit 7ffe312. Both commits neglected to convert all the
existing hacks. Make up what they missed.
Bonus: gets rid of empty for
Drop one more client of global_qtest by teaching all virtio test
functionality to pass in an explicit QTestState in constructors,
where it is then reused for later access. Adjust all callers.
This gets us one step closer to eliminating implicit use of
global_qtest.
Signed-off-by: Eric Blake
---
Put static functions prior to public ones, in part so that
improvements to qtest_start() can benefit from the static
helpers without needing forward references. Code motion, with
no semantic change.
Signed-off-by: Eric Blake
---
tests/libqtest.c | 263 +++
Drop one more client of global_qtest by teaching all rtas test
functionality to pass in an explicit QTestState, adjusting all
callers.
Signed-off-by: Eric Blake
---
tests/libqos/rtas.h | 9 ++---
tests/libqos/pci-spapr.c | 18 --
tests/libqos/rtas.c | 21 ++
Markus gave some good advice on my 'v4: Clean up around qmp() and hmp()'
series [1], but Paolo made the suggestion on v5 [2] that we are probably
better off avoiding dependence on a magic global variable, rather than
making such dependence even more prominent. So, this round of patches
starts to h
Managing parallel connections to two different monitors via
the implicit global_qtest makes it hard to copy-and-paste code
to tests that are not aware of the implicit state; the
management of global_qtest is even harder to follow because
it was masked behind set_context().
Instead, explicitly pass
Rather than make multiple callers call strlen(), it's easier if
socket_send() itself can compute a length via strlen() if none
was provided (caller passes -1). Callers that can get at the
length more efficiently are left that way.
Signed-off-by: Eric Blake
---
tests/libqtest.c | 10 ++
Back when the test was introduced, in commit 62c39b307, the
test was set up to run qemu-ga directly on the host performing
the test, and defaults to limiting itself to safe commands. At
the time, it was envisioned that setting QGA_TEST_SIDE_EFFECTING
in the environment could cover a few more comma
Prior to commit 063c23d9, we were tracking a list of parallel
qtest objects, in order to safely clean up a SIGABRT handler
only after the last connection quits. But when we switched to
more of glib's infrastructure, the list became dead code that
is never assigned to.
Signed-off-by: Eric Blake
R
This will keep checkpatch happy when the next patch does code motion.
Fix the include order to match HACKING when adding the needed header.
Signed-off-by: Eric Blake
Reviewed-by: Philippe Mathieu-Daudé
---
tests/libqtest.c | 19 +--
1 file changed, 13 insertions(+), 6 deletions
Right now, tests/multiboot is not run by default during 'make check';
but when it is run in-tree, it was leaving some junk behind that could
accidentally get checked in with a careless 'git add tests'.
Signed-off-by: Eric Blake
---
tests/.gitignore | 2 ++
1 file changed, 2 insertions(+)
diff -
Assertions should be separate from the side effects, since in
theory, g_assert() can be disabled (in practice, we can't really
ever do that).
Signed-off-by: Eric Blake
---
qtest.c | 80 ++---
1 file changed, 57 insertions(+), 23 deletio
It doesn't matter if things are unsorted, but finding stuff in a list
is easier when it is sorted. (Sorted under LC_ALL=C rules, rather than
en_US.UTF-8).
Signed-off-by: Eric Blake
---
tests/.gitignore | 30 +++---
1 file changed, 15 insertions(+), 15 deletions(-)
diff
On Fri, Sep 1, 2017 at 10:17 AM, wrote:
> Hi,
>
> This series failed automatic build test. Please find the testing commands and
> their output below. If you have docker installed, you can probably reproduce
> it
> locally.
>
> Message-id: cover.1504228916.git.alistair.fran...@xilinx.com
> Subjec
Hi,
Sergio, I was trying to applying patch 1/13 and 2/13 and then I ran:
./configure and saw: 'HVF support yes'
after that 'make' was happy
and:
./x86_64-softmmu/qemu-system-x86_64 --help | grep accel
\property accel=accel1[:accel2[:...]] selects accelerator
Hi,
> > i.e. something like this:
> >
> > https://www.kraxel.org/cgit/qemu-firmware/
> >
> > (only seabios is there for now)
>
> So presumably we'd want a 'make install' target there too, to place
> the
> various ROMS in the location QEMU expects ?
Yes.
Or place them in $prefix/share/qemu-f
From: Prem Mallappa
This patch builds the smmuv3 node in the ACPI IORT table. As
the smmu is dynamically instantianted using the platform bus,
the dt node creation function fills the information used by
the IORT table generation function (base address, base irq,
type of the smmu).
The RID space
This patch adds a node creation function for the smmuv3. Using
"-device smmuv3" it is now possible to get the iommu instantiated
on ARM VIRT machine. The node creation function handles the creation
of the smmuv3 node and also add the iommu-map property on the generic
PCI host controller node.
Sign
In VFIO use cases, the virtual smmu translates IOVA->IPA (stage 1)
whereas the physical SMMU translates IPA -> host PA (stage 2).
The 2 stages of the physical SMMU are currently not used.
Instead both stage 1 and stage2 mappings are combined together
and programmed in a single stage (S1) in the ph
The VirtMachineState contains some dt phandles that will be used
in some node creation functions. For instance we plan to use the
PCI host controller phandle in the smmu node creation function. So
let's pass the VirtMachineState handle down to the node creation
functions by enhancing the involved d
Let's allocate a phandle for the PCI host controller dt
node and store this latter in the Virt Machine State. This
will simplify fdt operations when we bind smmu and PCI host
controller.
Signed-off-by: Eric Auger
---
hw/arm/virt.c | 5 -
include/hw/arm/virt.h | 1 +
2 files changed,
SMMUV3 does not support any IOVA range TLBI command:
SMMU_CMD_TLBI_NH_VA invalidates TLB entries by page.
That's an issue when running DPDK on guest. DPDK uses
hugepages but each time a hugepage is mapped on guest side,
a storm of SMMU_CMD_TLBI_NH_VA commands get sent by the
guest smmuv3 driver and
Let's introduce a helper function aiming at recording an
event in the event queue.
Signed-off-by: Eric Auger
---
At the moment, for some events we do not fill all the fields.
Typically filling the FetchAddr field resulting of an abort
on page table walk would require to return more information
memory_region_iommu_replay() is used for VFIO integration.
However its default implementation is not adapted to SMMUv3
IOMMU memory region. Indeed the input address range is too
huge and its execution is too slow as it calls the translate()
callback on each granule.
Let's implement the replay cal
This patch implements the IOMMU Memory Region translate()
callback. Most of the code relates to the translation
configuration decoding and check (STE, CD).
Signed-off-by: Eric Auger
---
hw/arm/smmuv3-internal.h | 182 +++-
hw/arm/smmuv3.c | 351 ++
The Base address of the platform bus may be useful for device node
creation function. This is typically the case if the node creation
function also prepares data for ACPI table generation.
Signed-off-by: Eric Auger
---
hw/arm/sysbus-fdt.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/hw/
Now we have relevant helpers for queue and irq
management, let's implement MMIO write operations
Signed-off-by: Eric Auger
---
hw/arm/smmuv3-internal.h | 103 +++-
hw/arm/smmuv3.c | 204 ++-
hw/arm/trace-events | 15 +
We introduce helpers to read/write into the circular queues.
smmuv3_read_cmdq and smmuv3_write_evtq will become static
later on.
Signed-off-by: Eric Auger
---
hw/arm/smmuv3-internal.h | 48 ++-
hw/arm/smmuv3.c | 75
We introduce some helpers to handle wired IRQs and especially
GERROR interrupt. SMMU writes GERROR register on GERROR event
and SW acks GERROR interrupts by setting GERRORn.
The Wired interrupts are edge sensitive.
Signed-off-by: Eric Auger
---
Is CMD_SYNC interrupt enabled somewhere?
---
hw/
When the guest invalidates data structure (STE, CD) or TLB
we need to notify the IOMMU region notifiers. This allows
vhost integration and also prepares for VFIO integration.
Signed-off-by: Eric Auger
---
v6 -> v7:
- move SMMU_CMD_TLBI_NH_VA_AM in a separate patch
- rationalize names and add som
This patch implements the page table walk for VMSAv8-64.
The page table walk function is devised to walk the tables
for a range of IOVAs and to call a callback for each valid
leaf entry (frame or block).
smmu_page_walk_level_64() handles the walk from a specific level.
The advantage of using recu
From: Prem Mallappa
This patch implements a skeleton for the smmuv3 device.
Datatypes and register definitions are introduced. The MMIO
region, the interrupts and the queue are initialized (PRI is
not supported).
Only the MMIO read operation is implemented here.
Signed-off-by: Prem Mallappa
Si
1 - 100 of 381 matches
Mail list logo
