SPARC is the last target that uses legacy way of parsing
and initializing cpu features, drop legacy approach and
convert features to properties so that SPARC could as minimum
benefit from generic cpu_generic_init(), common with
x86 +-feat parser
PS:
the main purpose is to remove legacy way of cpu
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Philippe Mathieu-Daudé
Tested-by: Philippe Mathieu-Daudé
---
CC: Fabien Chouteau
CC: Mark Cave-Ayland
CC: Artyom Tarasenko
---
target/sparc/cpu.h | 3 +--
hw/sparc/leon3.c | 2 +-
cpu_s390x_init() is used only *-user targets indirectly
via cpu_init() macro and has a hack to assign ids to created
cpus (I'm not sure if 'id' really matters to *-user emulation).
So to on safe side, instead of having custom wrapper to do numbering
replace it with cpu_generic_init() and use S390C
drop custom cpu_hppa_init() in favor of cpu_generic_init(),
to make cpu_generic_init() work all we need is to provide
cc->class_by_name callback that would resolve any cpu_model
to the sole TYPE_HPPA_CPU to match current behaviour.
Signed-off-by: Igor Mammedov
Acked-by: Richard Henderson
---
te
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/sh4/cpu.h | 3 +--
hw/sh4/r2d.c | 2 +-
hw/sh4/shix.c| 2 +-
target/sh4/cpu.c | 5 -
4 files changed, 3 insertions(+), 9 deletions(-)
diff --git a/t
cpu_tilegx_init() always falls back to TYPE_TILEGX_CPU object
regardless of cpu_model. Put fallback logic into
tilegx_cpu_class_by_name() which would translate any cpu_model
into TYPE_TILEGX_CPU class and replace cpu_tilegx_init()
with cpu_generic_init().
Signed-off-by: Igor Mammedov
Reviewed-by:
cpu_mb_init() always falls back to TYPE_MICROBLAZE_CPU object
regardless of cpu_model. Put fallback logic into
mb_cpu_class_by_name() which would translate any cpu_model
into TYPE_MICROBLAZE_CPU class and replace cpu_mb_init()
with cpu_generic_init().
Signed-off-by: Igor Mammedov
Reviewed-by: Phi
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Andrew Jones
Tested-by: Philippe Mathieu-Daudé
---
target/arm/cpu.h| 3 +--
hw/arm/musicpal.c | 2 +-
hw/arm/omap1.c | 2 +-
hw/arm/omap2.c | 2 +-
hw/arm/pxa2xx.c | 4
cpu_nios2_init() always falls back to TYPE_NIOS2_CPU object
regardless of cpu_model. Put fallback logic into
nios2_cpu_class_by_name() which would translate any cpu_model
into TYPE_NIOS2_CPU class and replace cpu_nios2_init()
with cpu_generic_init()
Signed-off-by: Igor Mammedov
Reviewed-by: Phili
call register_m68k_insns() at realize time which makes
cpu_m68k_init() typical object creation function.
As result we can replace it with cpu_generic_init()
which does the same job, reducing code duplication a bit.
Signed-off-by: Igor Mammedov
Tested-by: Thomas Huth
Acked-by: Laurent Vivier
---
call xtensa_irq_init() at realize time which makes
cpu_xtensa_init() like generic cpu creation function.
As result we can replace it with cpu_generic_init()
which does the same job, reducing code duplication a bit.
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
Reviewed-by: Philippe M
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/tricore/cpu.h | 4 +---
hw/tricore/tricore_testboard.c | 2 +-
target/tricore/helper.c| 5 -
3 files changed, 2 insertions(+), 9 deletions(
On 08/24/2017 03:45 AM, Markus Armbruster wrote:
> The lookup tables have a sentinel, no need to make callers pass their
> size.
>
> Fun: the header has it in the wrong position. Good riddance.
>
> Signed-off-by: Markus Armbruster
> ---
> block.c | 1 -
> block/file-posix.c
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/moxie/cpu.h | 3 +--
hw/moxie/moxiesim.c | 2 +-
target/moxie/cpu.c | 5 -
3 files changed, 2 insertions(+), 8 deletions(-)
diff --git a/target/moxie/cpu.
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/cris/cpu.h| 3 +--
hw/cris/axis_dev88.c | 2 +-
target/cris/cpu.c| 5 -
3 files changed, 2 insertions(+), 8 deletions(-)
diff --git a/target/cris/cp
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/i386/cpu.h | 3 +--
target/i386/cpu.c | 5 -
2 files changed, 1 insertion(+), 7 deletions(-)
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 051867
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Acked-by: Michael Walle
Reviewed-by: Eduardo Habkost
---
target/lm32/cpu.h | 3 +--
hw/lm32/lm32_boards.c | 4 ++--
hw/lm32/milkymist.c | 2 +-
target/lm32/helper.c | 5 -
4 files changed,
On 08/24/2017 07:59 AM, meeho yuen wrote:
> Public bug reported:
>
> The vm guest runs a iotest program, and i migrate it with virsh --copy-
> storage-all,then the qemu process on the source host happens to crash
> with the following message:
>
> kvm: block/io.c:1543: bdrv_co_pwritev: Assertion `
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Acked-by: Stafford Horne
Reviewed-by: Eduardo Habkost
---
target/openrisc/cpu.h | 4 +---
hw/openrisc/openrisc_sim.c | 2 +-
target/openrisc/cpu.c | 5 -
3 files changed, 2 insertions(+)
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Reviewed-by: Eduardo Habkost
---
target/unicore32/cpu.h| 4 +---
hw/unicore32/puv3.c | 2 +-
target/unicore32/helper.c | 5 -
3 files changed, 2 insertions(+), 9 deletions(-)
diff --git
it's just a wrapper, drop it and use cpu_generic_init() directly
Signed-off-by: Igor Mammedov
Acked-by: David Gibson
Reviewed-by: Hervé Poussineau
Reviewed-by: Eduardo Habkost
---
target/ppc/cpu.h| 3 +--
hw/ppc/e500.c | 3 ++-
hw/ppc/mac_newworld.c | 3 ++-
hw
On 08/24/2017 01:31 PM, Igor Mammedov wrote:
with features converted to properties we can use the same
approach as x86 for features parsing and drop legacy
approach that manipulated CPU instance directly.
New sparc_cpu_parse_features() will allow only +-feat
and explicitly disable feat=on|off syn
On 08/24/2017 03:45 AM, Markus Armbruster wrote:
> Signed-off-by: Markus Armbruster
> ---
> docs/devel/qapi-code-gen.txt | 47
> +++-
> 1 file changed, 38 insertions(+), 9 deletions(-)
Might be nice to call out which commit(s) failed to update the docs wh
* Peter Xu (pet...@redhat.com) wrote:
> On Wed, Jun 28, 2017 at 08:00:20PM +0100, Dr. David Alan Gilbert (git) wrote:
> > From: "Dr. David Alan Gilbert"
> >
> > The choice of call to discard a block is getting more complicated
> > for other cases. We use fallocate PUNCH_HOLE in any file cases;
Hi Igor,
On 08/24/2017 01:31 PM, Igor Mammedov wrote:
since commit ( 9262685b cpu: Factor out cpu_generic_init() )
features parsed by it were truncated only to the 1st feature
after CPU name due to fact that
featurestr = strtok(NULL, ",");
cc->parse_features(cpu, featurestr, &err);
wou
On 08/24/2017 11:21 AM, Paolo Bonzini wrote:
> On 24/08/2017 17:33, Stefan Hajnoczi wrote:
>> This patch enters read_reply_co directly in
>> nbd_client_attach_aio_context(). This is safe because new_context is
>> acquired by the caller. This ensures that read_reply_co reaches its
>> first yield p
On 08/24/2017 12:07 PM, Peter Maydell wrote:
On 24 August 2017 at 16:53, Farhan Ali wrote:
On 08/24/2017 11:50 AM, Thomas Huth wrote:
True, so that could still be an issue. Looking at the cio.h in the
kernel, they define the struct like this:
struct ccw1 {
__u8 cmd_code;
On 08/24/2017 10:33 AM, Stefan Hajnoczi wrote:
> Currently 083 waits for the nbd-fault-injector.py server to start up by
> looping until netstat shows the TCP listen socket.
>
> The startup protocol can be simplified by passing a 0 port number to
> nbd-fault-injector.py. The kernel will allocate
On 24/08/2017 19:37, Eric Blake wrote:
> On 08/24/2017 11:21 AM, Paolo Bonzini wrote:
>> On 24/08/2017 17:33, Stefan Hajnoczi wrote:
>>> This patch enters read_reply_co directly in
>>> nbd_client_attach_aio_context(). This is safe because new_context is
>>> acquired by the caller. This ensures th
On 08/24/2017 10:33 AM, Stefan Hajnoczi wrote:
> 083 only tests TCP. Some failures might be specific to UNIX domain
> sockets.
>
> A few adjustments are necessary:
>
> 1. Generating a port number and waiting for server startup is
>TCP-specific. Use the new nbd-fault-injector.py startup prot
On Thu, Aug 24, 2017 at 04:38:43PM +0800, Fam Zheng wrote:
> On Thu, 08/24 08:21, Stefan Hajnoczi wrote:
> > Tests should declare resources upfront in a with statement. Resources are
> > automatically cleaned up whether the test passes or fails:
> >
> > with FilePath('test.img') as img_path,
>
On 08/24/2017 07:38 PM, Farhan Ali wrote:
>
>
> On 08/24/2017 12:07 PM, Peter Maydell wrote:
>> On 24 August 2017 at 16:53, Farhan Ali wrote:
>>>
>>>
>>> On 08/24/2017 11:50 AM, Thomas Huth wrote:
True, so that could still be an issue. Looking at the cio.h in the
kernel, they define
On 08/24/2017 06:02 PM, Halil Pasic wrote:
>
>
> On 08/24/2017 05:53 PM, Farhan Ali wrote:
>>
>>
>> On 08/24/2017 11:50 AM, Thomas Huth wrote:
>>> On 24.08.2017 17:47, Halil Pasic wrote:
On 08/24/2017 05:35 PM, Thomas Huth wrote:
> On 24.08.2017 17:13, Cornelia Huck wrote:
>>
On Thu, Aug 24, 2017 at 12:54:48PM +1000, Paul Mackerras wrote:
> On Mon, Aug 21, 2017 at 05:00:36PM -0300, Thiago Jung Bauermann wrote:
> > LoPAPR says:
> >
> > “ibm,processor-storage-keys”
> >
> > property name indicating the number of virtual storage keys supported
> > by the proce
Eric Blake writes:
> On 08/22/2017 01:52 AM, Markus Armbruster wrote:
>> Suggested-by: Max Reitz
>> Signed-off-by: Markus Armbruster
>> ---
>> include/qapi/qmp/qnum.h | 21 +
>> 1 file changed, 21 insertions(+)
>>
>> diff --git a/include/qapi/qmp/qnum.h b/include/qapi/qmp/
Eric Blake writes:
> On 08/24/2017 03:45 AM, Markus Armbruster wrote:
>> Signed-off-by: Markus Armbruster
>> ---
>> docs/devel/qapi-code-gen.txt | 47
>> +++-
>> 1 file changed, 38 insertions(+), 9 deletions(-)
>
> Might be nice to call out which commit(
On Tue, Aug 22, 2017 at 01:15:33PM +0300, Manos Pitsidianakis wrote:
> /* Increments the reference count of a ThrottleGroup given its name.
> *
> * If no ThrottleGroup is found with the given name a new one is
> * created.
> *
> + * This function edits throttle_groups and must be called un
Just to understand the urgency. Is the problem happening with the BIOS that is
shipped in the
upstream git tree, or a self-built one?
On 08/24/2017 05:05 PM, Farhan Ali wrote:
> Hi,
>
> There is an issue in QEMU bios which is exposed by commit
>
> commit 198c0d1f9df8c429502cb744fc26b6ba6e71db7
Eric Blake writes:
> On 08/24/2017 03:45 AM, Markus Armbruster wrote:
>> The lookup tables have a sentinel, no need to make callers pass their
>> size.
>>
>> Fun: the header has it in the wrong position. Good riddance.
>>
>> Signed-off-by: Markus Armbruster
>> ---
>> block.c
Marc-André Lureau writes:
> Hi
>
> On Thu, Aug 24, 2017 at 10:45 AM, Markus Armbruster wrote:
>> We have a strict separation between enum TpmModel and tpm_models[]:
>>
>> * TpmModel may have any number of members. It just happens to have one.
>>
>> * tpm_register_model() uses the first empty sl
On Tue, Aug 22, 2017 at 01:15:34PM +0300, Manos Pitsidianakis wrote:
> @@ -3095,6 +3096,20 @@
> '*tls-creds': 'str' } }
>
> ##
> +# @BlockdevOptionsThrottle:
> +#
> +# Driver specific block device options for the throttle driver
> +#
> +# @throttle-group: the name of the throttle-
On Tue, Aug 22, 2017 at 01:15:35PM +0300, Manos Pitsidianakis wrote:
> +_supported_fmt qcow2
What makes this test qcow2-specific? With additional filtering for
IMGFMT it should be possible to run this on any image format.
Hello,
On behalf of the QEMU Team, I'd like to announce the availability of the
fifth release candidate for the QEMU 2.10 release. This release is meant
for testing purposes and should not be used in a production environment.
http://download.qemu-project.org/qemu-2.10.0-rc4.tar.xz
http://dow
On 08/24/2017 01:35 PM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> On 08/24/2017 03:45 AM, Markus Armbruster wrote:
>>> The lookup tables have a sentinel, no need to make callers pass their
>>> size.
>>>
>>> Fun: the header has it in the wrong position. Good riddance.
>>>
>>> +++ b/incl
On 08/24/17 20:53, Michael Roth wrote:
> Hello,
>
> On behalf of the QEMU Team, I'd like to announce the availability of the
> fifth release candidate for the QEMU 2.10 release. This release is meant
> for testing purposes and should not be used in a production environment.
>
> http://download
I think 'make test' is an abandoned testing option for QEMU so this report
might be unimportant. Here is my results for running 'make test':
$ make test
make -C tests/tcg test
LINKtest_path
Undefined symbols for architecture x86_64:
"_buffer_is_zero", referenced from:
_qemu_iovec_is
Bug: introspection documentation is in section "Tracing commands".
Cause: sub-schema qapi/introspect.json lacks a section header, and
therefore goes into whatever section precedes its include.
Fix: add a section header.
Signed-off-by: Markus Armbruster
Reviewed-by: Marc-André Lureau
---
qapi/
Except for block-core.json, the sub-schemas are self-contained: if
they use a symbol defined in another sub-schema, they include that
sub-schema. To check, feed the sub-schema to qapi2texi (or any other
QAPI generator) along with the pragma from qapi-schema.json.
Fix up things to make block-core.
Cc: "Daniel P. Berrange"
Cc: Alberto Garcia
Cc: Dr. David Alan Gilbert
Cc: Gerd Hoffmann
Cc: Jason Wang
Cc: Juan Quintela
Cc: Marc-André Lureau
Cc: Paolo Bonzini
Much of the QAPI schema really belongs to a subsystem, but MAINTAINERS
can't tell when it's all in a big ball of mud (qapi-schem
The generated QEMU QMP reference is now structured as follows:
1.1 Introduction
1.2 Stability Considerations
1.3 Common data types
1.4 Socket data types
1.5 VM run state
1.6 Cryptography
1.7 Block devices
1.7.1 Block core (VM unrelated)
1.7.2 QAPI block definiti
Signed-off-by: Markus Armbruster
---
Makefile | 2 +-
qapi-schema.json | 134 -
qapi/event.json | 138 ---
3 files changed, 134 insertions(+), 140 deletions(-)
delete mode 100644 qa
Bug: section "Rocker switch device" starts with the rocker stuff, but
then has unrelated stuff, like ReplayMode, xen-load-devices-state, ...
Cause: rocker.json is included in the middle of section "QMP commands".
Fix: include it in a sane place, namely next to the other sub-schemas.
Signed-off-b
Cc: "Daniel P. Berrange"
Cc: Gerd Hoffmann
Cc: Paolo Bonzini
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 1 +
Makefile | 4 +-
qapi-schema.json | 152 +--
qapi/block-core.json | 2 +-
qapi/common.json |
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 1 +
Makefile | 1 +
qapi-schema.json | 151 +--
qapi/transaction.json | 158 ++
4 files changed, 161 insertions(+), 150 d
query-version and query-commands are in common.json for no good
reason. Several similar queries are in qapi-schema.json. Move them
there.
Signed-off-by: Markus Armbruster
---
qapi-schema.json | 103 +++
qapi/common.json | 103
Documentation generated with qapi2texi.py is in source order, with
included sub-schemas inserted at the first include directive
(subsequent include directives have no effect). To get a sane and
stable order, it's best to include each sub-schema just once, or
include it first in qapi-schema.json.
Cc: Paolo Bonzini
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 1 +
Makefile| 1 +
qapi-schema.json| 165 +---
qapi/event.json | 182 ---
qapi/run-state.json | 352 ++
Sadly, we don't have a TPM maintainer, not even a MAINTAINERS entry.
Create one, and mark it orphaned.
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 8
Makefile | 1 +
qapi-schema.json | 132 +
qapi/tpm.json| 137
Cc: Paolo Bonzini
Cc: Marc-André Lureau
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 1 +
Makefile | 1 +
qapi-schema.json | 511 +---
qapi/char.json | 538 +++
qapi/even
Cc: Alberto Garcia
Signed-off-by: Markus Armbruster
---
qapi/block.json | 68 +
qapi/event.json | 68 -
2 files changed, 68 insertions(+), 68 deletions(-)
diff --git a/qapi/block.json
Cc: Jason Wang
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 1 +
Makefile | 1 +
qapi-schema.json | 675 +---
qapi/event.json | 24 --
qapi/net.json| 706 +++
5 files
UI stuff is remote desktop stuff (Spice, VNC) and input stuff (mouse,
keyboard).
Cc: Gerd Hoffmann
Signed-off-by: Markus Armbruster
---
MAINTAINERS | 2 +
Makefile | 3 +-
qapi-schema.json | 784 +---
qapi/event.json | 175 --
qa
This series adds a new "discard-data" option to
memory-backend-file. The new option will be useful if somebody
is sharing RAM contents on a pre-existing file using share=on,
but don't need data to be flushed to disk when QEMU exits.
Internally, it will trigger a madvise(MADV_REMOVE) call when the
From: "Dr. David Alan Gilbert"
Hi,
This is a RFC/WIP series that enables postcopy migration
with shared memory to a vhost-user process.
It's based off current-head + Alexey's bitmap series
It's tested with vhost-user-bridge and a dpdk (modified by Maxime
that will get posted separately) - both
Cc: Juan Quintela
Cc: Dr. David Alan Gilbert
Signed-off-by: Markus Armbruster
---
MAINTAINERS |1 +
Makefile|1 +
qapi-schema.json| 1056 +
qapi/common.json| 16 +
qapi/event.json | 38 --
qapi/migration
From: "Dr. David Alan Gilbert"
Utility to give the offset of a host pointer within a RAMBlock
(assuming we already know it's in that RAMBlock)
Signed-off-by: Dr. David Alan Gilbert
---
exec.c| 10 ++
include/exec/cpu-common.h | 1 +
2 files changed, 11 insertions(+
Define QEMU_MADV_REMOVE, so we can use it with qemu_madvise().
Signed-off-by: Eduardo Habkost
---
Changes v1 -> v2:
* New patch added to series
---
include/qemu/osdep.h | 7 +++
1 file changed, 7 insertions(+)
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index 6855b94..e9fa217 1
From: "Dr. David Alan Gilbert"
Only process received packets if the queue has been started.
Signed-off-by: Dr. David Alan Gilbert
---
tests/vhost-user-bridge.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/tests/vhost-user-bridge.c b/tests/vhost-user-bridge.c
index 1e5b5ca3da..324abee53d
From: "Dr. David Alan Gilbert"
Use a flag on the RAMBlock to state whether it has the
UFFDIO_ZEROPAGE capability, use it when it's available.
This allows the use of postcopy on tmpfs as well as hugepage
backed files.
Signed-off-by: Dr. David Alan Gilbert
---
exec.c| 15 +++
Delete all user-creatable objects in /objects when exiting QEMU, so they
can perform cleanup actions.
Signed-off-by: Eduardo Habkost
---
Changes v1 -> v2:
* (none)
---
include/qom/object_interfaces.h | 8
qom/object_interfaces.c | 5 +
vl.c| 1 +
The new option can be used to indicate that the file contents can
be destroyed and don't need to be flushed to disk when QEMU exits
or when the memory backend object is removed.
Internally, it will trigger a madvise(MADV_REMOVE) call when the
memory backend is removed.
Signed-off-by: Eduardo Habk
From: "Dr. David Alan Gilbert"
Utility for testing the map when you already know the offset
in the RAMBlock.
Signed-off-by: Dr. David Alan Gilbert
---
migration/ram.c | 5 +
migration/ram.h | 1 +
2 files changed, 6 insertions(+)
diff --git a/migration/ram.c b/migration/ram.c
index affb20
From: "Dr. David Alan Gilbert"
Resolve fault addresses read off the clients UFD into RAMBlock
and offset, and call back to the postcopy code to ask for the page.
Signed-off-by: Dr. David Alan Gilbert
---
hw/virtio/trace-events | 3 +++
hw/virtio/vhost-user.c | 30 +
Eric Blake writes:
> On 08/24/2017 01:35 PM, Markus Armbruster wrote:
>> Eric Blake writes:
>>
>>> On 08/24/2017 03:45 AM, Markus Armbruster wrote:
The lookup tables have a sentinel, no need to make callers pass their
size.
Fun: the header has it in the wrong position. Good
From: "Dr. David Alan Gilbert"
Add a vhost feature flag for postcopy support, and
use the postcopy notifier to check it before allowing postcopy.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.h | 1 +
docs/interop/vhost-user.txt | 10 +
hw/vir
From: "Dr. David Alan Gilbert"
Cause the vhost-user client to be woken up whenever:
a) We place a page in postcopy mode
b) We get a fault and the page has already been received
Signed-off-by: Dr. David Alan Gilbert
---
migration/postcopy-ram.c | 14 ++
migration/trace-events
From: "Dr. David Alan Gilbert"
Allow other userfaultfd's to be registered into the fault thread
so that handlers for shared memory can get responses.
Signed-off-by: Dr. David Alan Gilbert
---
migration/migration.c| 3 +
migration/migration.h| 2 +
migration/postcopy-ram.c | 212 +++
From: "Dr. David Alan Gilbert"
Wire up a notifier to send a VHOST_USER_POSTCOPY_ADVISE
message on an incoming advise.
Later patches will fill in the behaviour/contents of the
message.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 21 ---
contrib
From: "Dr. David Alan Gilbert"
Add a vu_queue_started method to complement vu_queue_enabled.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 6 ++
contrib/libvhost-user/libvhost-user.h | 9 +
2 files changed, 15 insertions(+)
diff --git a/contrib/
From: "Dr. David Alan Gilbert"
**HACK - better solution needed **
We have the situation where:
qemu bridge
send set_mem_table
map memory
a) mark area with UFD
send reply with ma
From: "Dr. David Alan Gilbert"
Register the UFD that comes in as the response to the 'advise' method
with the postcopy code.
Signed-off-by: Dr. David Alan Gilbert
---
hw/virtio/vhost-user.c | 21 -
migration/postcopy-ram.h | 2 +-
2 files changed, 21 insertions(+), 2 del
From: "Dr. David Alan Gilbert"
When new regions are sent to the client using SET_MEM_TABLE, register
them with the userfaultfd.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 32
1 file changed, 32 insertions(+)
diff --git a/
From: "Dr. David Alan Gilbert"
The choice of call to discard a block is getting more complicated
for other cases. We use fallocate PUNCH_HOLE in any file cases;
it works for both hugepage and for tmpfs.
We use the DONTNEED for non-hugepage cases either where they're
anonymous or where they're p
From: "Dr. David Alan Gilbert"
Now that we have the mechanisms in here, allow shared memory in a
postcopy.
Note that QEMU can't tell who all the users of shared regions are
and thus can't tell whether all the users of the shared regions
have appropriate support for postcopy. Those devices that
From: "Dr. David Alan Gilbert"
Notify the vhost-user client on reception of the 'postcopy-listen'
event from the source.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 21 +
contrib/libvhost-user/libvhost-user.h | 2 ++
docs/interop/vhost
From: "Dr. David Alan Gilbert"
Send requests back to the source for shared page requests.
Signed-off-by: Dr. David Alan Gilbert
---
migration/migration.h| 2 ++
migration/postcopy-ram.c | 31 ---
migration/postcopy-ram.h | 3 +++
migration/trace-events | 2
From: "Dr. David Alan Gilbert"
Tell QEMU we understand the protocol features needed for postcopy.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 33 +
1 file changed, 33 insertions(+)
diff --git a/contrib/libvhost-user/libvhos
From: "Dr. David Alan Gilbert"
We need a better way, but at the moment we need the address of the
mappings sent back to qemu so it can interpret the messages on the
userfaultfd it reads.
Note: We don't ask for the default 'ack' reply since we've got our own.
Signed-off-by: Dr. David Alan Gilber
From: "Dr. David Alan Gilbert"
Add a notifier chain for postcopy with a 'reason' flag
and an opportunity for a notifier member to return an error.
Call it when enabling postcopy.
This will initially used to enable devices to declare they're unable
to postcopy and later to notify of devices of s
From: "Dr. David Alan Gilbert"
Clear the area and turn off THP.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 32 ++--
1 file changed, 30 insertions(+), 2 deletions(-)
diff --git a/contrib/libvhost-user/libvhost-user.c
b/contrib
From: "Dr. David Alan Gilbert"
Where two regions are created with a gap such that when aligned
to hugepage boundaries, the two regions overlap, merge them.
I also add quite a few trace events to see what's going on.
Note: This doesn't handle all the cases, but does handle the common
case on a P
From: "Dr. David Alan Gilbert"
Send a 'wake' request on a userfaultfd for a shared process.
The address in the clients address space is specified together
with the RAMBlock it was resolved to.
Signed-off-by: Dr. David Alan Gilbert
---
migration/postcopy-ram.c | 26 ++
m
From: "Dr. David Alan Gilbert"
Allow replies with fds (for postcopy)
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 29 -
1 file changed, 28 insertions(+), 1 deletion(-)
diff --git a/contrib/libvhost-user/libvhost-user.c
b/contri
From: "Dr. David Alan Gilbert"
This message is sent just before the end of postcopy to get the
client to stop using userfault since we wont respond to any more
requests. It should close userfaultfd so that any other pages
get mapped to the backing file automatically by the kernel, since
at this
From: "Dr. David Alan Gilbert"
The previous patch merges hugepage regions with small gaps
back into one; this patch stops some of the cases that split
them up again.
vhost_set_memory avoids adding small regions that are being
dirty-monitored (typically VGA regions), but vhost_set_memory
does rem
From: "Dr. David Alan Gilbert"
Register a waker function in vhost-user code to be notified when
pages arrive or requests to previously mapped pages get requested.
Signed-off-by: Dr. David Alan Gilbert
---
hw/virtio/trace-events | 3 +++
hw/virtio/vhost-user.c | 26 ++
From: "Dr. David Alan Gilbert"
Open a userfaultfd (on a postcopy_advise) and send it back in
the reply to the qemu for it to monitor.
Signed-off-by: Dr. David Alan Gilbert
---
contrib/libvhost-user/libvhost-user.c | 26 +++---
contrib/libvhost-user/libvhost-user.h | 3 +++
From: "Dr. David Alan Gilbert"
Add some notes to the migration documentation for shared memory
postcopy.
Signed-off-by: Dr. David Alan Gilbert
---
docs/devel/migration.txt | 39 +++
1 file changed, 39 insertions(+)
diff --git a/docs/devel/migration.txt b/do
From: "Dr. David Alan Gilbert"
Wire up a call to VHOST_USER_POSTCOPY_END message to the vhost clients
right before we ask the listener thread to shutdown.
Signed-off-by: Dr. David Alan Gilbert
---
hw/virtio/trace-events | 2 ++
hw/virtio/vhost-user.c | 30 ++
m
From: "Dr. David Alan Gilbert"
Stash the RAMBlock and offset for later use looking up
addresses.
Signed-off-by: Dr. David Alan Gilbert
---
hw/virtio/trace-events | 1 +
hw/virtio/vhost-user.c | 30 ++
2 files changed, 31 insertions(+)
diff --git a/hw/virtio/trace-
201 - 300 of 360 matches
Mail list logo
