* Anthony Liguori [2010-11-01 08:40:36]:
> On 10/26/2010 09:14 AM, Arun R Bharadwaj wrote:
> >+Q.7: Apart from the global pool of threads, can I have my own private Queue
> >?
> >+A.7: Yes, the threadlets framework allows subsystems to create their own
> >private
> >+ queues with associated
This patch introduce a fallback mechanism for old systems that do not
support utimensat. This will fix build failure with following warnings:
hw/virtio-9p-local.c: In function 'local_utimensat':
hw/virtio-9p-local.c:479: warning: implicit declaration of function 'utimensat'
hw/virtio-9p-local.c:4
(2010/11/05 2:03), Chris Wright wrote:
> * Hidetoshi Seto (seto.hideto...@jp.fujitsu.com) wrote:
>> (2010/10/14 4:11), Blue Swirl wrote:
>>> On Wed, Oct 13, 2010 at 8:00 AM, Hidetoshi Seto
>>> wrote:
(Add CC to k...@vger)
(2010/10/12 10:52), Hao, Xudong wrote:
> Hi,
> Curren
Hi.
The current BAR allocation doesn't check overflow and some patches
are floating around which aren't merged yet.
There are several issues.
- overflow check
This should be fixed.
Some patches are proposed. None hasn't been merged yet.
Your patch also addresses this issue.
http://www.seab
Hi,
We have seen some issues with 64bit PCI address and large BAR region support in
seabios.
On attempting to register a 64bit BAR of 1GB size we had some strange failures
in qemu. After some debugging we find out that source of the issue was in
seabios PCI enumeration code.
The issue takes
XXX: 'tracetool' could be simplified by never generating real tracing code for
events with the "disable" property (not even with the 'simple' backend).
XXX: Must be applied on top of the 'backdoor' patch series
Adds a new optional keyword ("instrument") to the syntax in "trace-events". When
s
Auto-generates file "trace-gen.h" with instrumentable wrappers to generate calls
to TCG trace helpers.
Such wrappers are named 'trace_gen_##name', also reachable as
'trace_gen_##name##_backend' when instrumented.
Events with the "gen" property are also able to use TCG types on the trace event
dec
Auto-generates file "trace-helper.h" to provide TCG helpers and "trace-helper.c"
to proxy these helpers onto trace event routines.
Only trace events with the 'gen' property are affected.
Signed-off-by: Lluís Vilanova
---
.gitignore |2 +
Makefile| 15 +++
Makefile.target
Add a '--with-instrument' configuration option pointing to user-provided
instrumentation callbacks.
Make is invoked on the user-provided directory, which must build a static
library that might contain extra code needed by the user-provided
instrumentation.
Signed-off-by: Lluís Vilanova
---
Make
Header for the user to include when compiling guest applications that want to
communicate with QEMU through backdoor instructions.
Signed-off-by: Lluís Vilanova
---
backdoor/guest.h | 33 +
1 files changed, 33 insertions(+), 0 deletions(-)
create mode 100644 ba
Provides a guest application that exercices the instruction-based backdoor
communication, as well as a backdoor callback implementation that prints the
guest requests.
Signed-off-by: Lluís Vilanova
---
.gitignore |1 +
backdoor/examples/print/README | 13
Signed-off-by: Lluís Vilanova
---
target-alpha/helper.h |2 ++
target-arm/helper.h|2 ++
target-cris/helper.h |2 ++
target-i386/helper.h |2 ++
target-m68k/helper.h |2 ++
target-microblaze/helper.h |2 ++
target-mips/helper.h |2 +
Some trace events with the "gen" property will need to use TCG arguments.
As such, modify 'tracetool' to gracefully handle these types and let the backend
code transliterate these types into their native counterpart.
Signed-off-by: Lluís Vilanova
---
tracetool | 137 +++
Add a new event keyword ("instrument") that lets the user provide her own
implementation of tracing events.
Still, tracetool's original implementation is accessible through function
'_trace_##name' instead of 'trace_##name' (in case the user only wants to wrap
around the event).
Additionally, a '
Provides the ability for the guest to communicate with user-provided code inside
QEMU itself.
This backdoor communication channel is based on using instructions, as opposed
to methods based on MMIO/PIO or virtproxy/virtagent.
Then, these instructions are redirected to a set of user-provided routi
Add a '--with-backdoor' configuration option pointing to user-provided backdoor
callback implementation.
Make is invoked on the user-provided directory, which must build a static
library containing, at least, the implementation of the backdoor helpers.
Signed-off-by: Lluís Vilanova
---
Makefile
Signed-off-by: Lluís Vilanova
---
Makefile |4 -
tracetool | 313 +
2 files changed, 192 insertions(+), 125 deletions(-)
diff --git a/Makefile b/Makefile
index 02698e9..2ba52aa 100644
--- a/Makefile
+++ b/Makefile
@@ -110,12 +110,
XXX: Must be applied on top of the 'trace-instrument' patch series
Adds a new optional keyword ("gen") to the syntax in "trace-events". When
specified, this event property will generate an extra set of functions that
provide the ability to inject trace points into TCG-generated code. Note that
the
Take the unused CPUID 0x40001xxx range as the backdoor instruction.
Signed-off-by: Lluís Vilanova
---
backdoor/guest.h| 21 +
target-i386/cpuid.c | 27 +++
target-i386/helper.h|4
target-i386/translate.c |4
4 file
This provides a consistent naming scheme across all targets.
Signed-off-by: Lluís Vilanova
---
target-arm/helper.c|2
target-arm/helper.h| 450
target-arm/helpers.h | 450
targe
These helpers must be implemented by the user on "libbackdoor.a".
Signed-off-by: Lluís Vilanova
---
backdoor/helper.h | 21 +
1 files changed, 21 insertions(+), 0 deletions(-)
create mode 100644 backdoor/helper.h
diff --git a/backdoor/helper.h b/backdoor/helper.h
new file
A via -kernel supplied x86_64 ELF image is being started in 32bit mode.
Detect and exit if a 64bit image has been supplied.
Signed-off-by: Adam Lackorzynski
Acked-by: Alexander Graf
---
hw/multiboot.c |6 ++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/hw/multiboot.c
Refine check on bkpt so that smc and undefined instruction encodings are
handled as an undefined instruction and trap.
Signed-off-by: Adam Lackorzynski
---
target-arm/translate.c |9 -
1 files changed, 8 insertions(+), 1 deletions(-)
diff --git a/target-arm/translate.c b/target-arm
On 15.06.2010, at 11:15, Hannes Reinecke wrote:
> Hi all,
>
> I've run an I/O exerciser against the megasas emulation and
> found quite some issues, both with the HBA and the SCSI
> emulation.
>
> - SG_IO returns -EDOM if the internal queue is exhausted;
> we should be mapping this to QUEUE_FU
Hello
The following patch enables simulated perf event support inside of Qemu
for x86_64 systems. It enables support for the AMD performance MSRs enough
to return values for the "retired_instructions" (both user and kernel) and
"cpu_clk_unhalted" events.
This is mostly a proof of concept, I'm
On 11/03/2010 06:44 PM, Adam Litke wrote:
You've got a lot of "objects" interacting with one another in virtproxy.
I think it would help other reviewers if you could describe the
relationships between the entities such as: VPDriver, VPConn, VPChannel,
VPIForward, VPOForward, etc. This patch seri
On 11/03/2010 08:21 PM, Adam Litke wrote:
On Wed, 2010-11-03 at 10:28 -0500, Michael Roth wrote:
reads data from client/server connections as they become readable, then
sends the data over the channel
Signed-off-by: Michael Roth
---
virtproxy.c | 80 +
On 11/03/2010 07:46 PM, Adam Litke wrote:
On Wed, 2010-11-03 at 10:28 -0500, Michael Roth wrote:
Process VPPackets coming in from channel and send them to the
appropriate server/client connections.
Signed-off-by: Michael Roth
---
virtproxy.c | 42 ++
There's no reason for tap to run when VM is stopped.
If we let it, it confuses the bridge on TX
and corrupts DMA memory on RX.
Signed-off-by: Michael S. Tsirkin
---
hw/vhost_net.c |2 +-
net/tap.c |3 ++-
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/hw/vhost_net.c
Create separate lists for system and device fd handlers.
Device handlers will not run while vm is stopped.
By default all fds are assumed system so they will
keep running as before.
Signed-off-by: Michael S. Tsirkin
---
qemu-char.h |6 +++-
qemu-kvm.c |2 +-
qemu-tool.c | 10 +
vl
Here's a patch that's much smaller: instead of trying
to solve the whole problem, it fixes the specific
bug I observe, that of net device traffic when vm is stopped
confusing the bridge.
We can then slowly convert all other devices that need to be
stopped on vmstop to this API.
Comments?
Michael
On Thu, 21 Oct 2010 08:37:19 +0200
Alon Levy wrote:
> ---
> hmp-commands.hx |1 +
> monitor.c | 13 -
> 2 files changed, 13 insertions(+), 1 deletions(-)
>
> diff --git a/hmp-commands.hx b/hmp-commands.hx
> index 3014b17..289fbcb 100644
> --- a/hmp-commands.hx
> +++ b/hm
Correct ldrexd and strexd code to always read and write the
high word of the 64-bit value from addr+4.
Also make ldrexd and strexd agree that for a 64 bit value the
address in env->exclusive_addr is that of the low word.
This fixes the issues reported in
https://bugs.launchpad.net/qemu/+bug/670883
On Thu, Nov 04, 2010 at 11:45:51AM -0500, Ryan Harper wrote:
> OK. With netdev_del and drive_unplug commands (not sure if we care to
> change the names to be similar, maybe blockdev_del) in qemu, we can then
> implement the following in libvirt:
>
> 1) detach-device invocation
> 2) issue device_d
* Hidetoshi Seto (seto.hideto...@jp.fujitsu.com) wrote:
> (2010/10/14 4:11), Blue Swirl wrote:
> > On Wed, Oct 13, 2010 at 8:00 AM, Hidetoshi Seto
> > wrote:
> >> (Add CC to k...@vger)
> >>
> >> (2010/10/12 10:52), Hao, Xudong wrote:
> >>> Hi,
> >>> Currently qemu-kvm build fail on RHEL5 with gcc
On 11/03/2010 06:38 PM, Adam Litke wrote:
On Wed, 2010-11-03 at 10:28 -0500, Michael Roth wrote:
Handle data coming in over the channel as VPPackets: Process control
messages and forward data from remote client/server connections to the
appropriate server/client FD on our end.
Signed-off-by: Mi
* Michael S. Tsirkin [2010-11-03 16:46]:
> On Wed, Nov 03, 2010 at 03:59:29PM -0500, Ryan Harper wrote:
> > * Michael S. Tsirkin [2010-11-03 13:03]:
> > > On Wed, Nov 03, 2010 at 12:29:10PM -0500, Ryan Harper wrote:
> > > > * Markus Armbruster [2010-11-03 11:42]:
> > > > > Ryan Harper writes:
>
On 11/03/2010 06:02 PM, Adam Litke wrote:
On Wed, 2010-11-03 at 10:28 -0500, Michael Roth wrote:
This accept()'s connections to the socket we told virt-proxy to listen
for the channel connection on and sets the appropriate read handler for
the resulting FD.
Signed-off-by: Michael Roth
---
vir
On Thu, Nov 04, 2010 at 03:58:03PM +0100, Markus Armbruster wrote:
> Gleb Natapov writes:
>
> > On Thu, Nov 04, 2010 at 10:20:18AM +0100, Markus Armbruster wrote:
> >> Gleb Natapov writes:
> >>
> >> > Add "deriver_name" to DeviceInfo to use in device path building. In
> >>
> >> Typo "deriver".
On Thu, Nov 04, 2010 at 03:22:50PM +0100, Markus Armbruster wrote:
> Gleb Natapov writes:
>
> > On Thu, Nov 04, 2010 at 09:46:57AM +0100, Markus Armbruster wrote:
> >> > But why order of device creation is important? It shouldn't be if we
> >> > want to move HW description into config file. We ev
Gleb Natapov writes:
> On Thu, Nov 04, 2010 at 10:20:18AM +0100, Markus Armbruster wrote:
>> Gleb Natapov writes:
>>
>> > Add "deriver_name" to DeviceInfo to use in device path building. In
>>
>> Typo "deriver". Same in subject.
>>
> Heh.
>
>> > contrast to "name" "driver_name" should refer
Gleb Natapov writes:
> On Thu, Nov 04, 2010 at 09:46:57AM +0100, Markus Armbruster wrote:
>> > But why order of device creation is important? It shouldn't be if we
>> > want to move HW description into config file. We even may allow creating
>> > piix3-ide with only second IDE bus, but not first.
On 11/03/2010 05:47 PM, Adam Litke wrote:
On Wed, 2010-11-03 at 10:27 -0500, Michael Roth wrote:
+/* mirror qemu I/O-related code for standalone daemon */
+typedef struct IOHandlerRecord {
+int fd;
+IOCanReadHandler *fd_read_poll;
+IOHandler *fd_read;
+IOHandler *fd_write;
+i
Coincidentally, I was just looking at this yesterday. There's another
error you haven't pointed out, which is in the linux-user implementation
of do_strex(), which writes both halves of the 64 bit double to the same
address, rather than to addr and addr+4.
--
ARM : ldrexd and strexd implementatio
On 11/04/2010 07:54 AM, Kevin Wolf wrote:
Am 27.10.2010 20:19, schrieb Anthony Liguori:
Signed-off-by: Anthony Liguori
diff --git a/block.c b/block.c
index 1a965b2..00b6f21 100644
--- a/block.c
+++ b/block.c
@@ -603,10 +603,16 @@ int bdrv_open(BlockDriverState *bs, const char *filename,
in
Am 04.11.2010 14:14, schrieb Anthony Liguori:
> On 11/04/2010 07:54 AM, Kevin Wolf wrote:
>> Am 27.10.2010 20:19, schrieb Anthony Liguori:
>>
>>> Signed-off-by: Anthony Liguori
>>>
>>> diff --git a/block.c b/block.c
>>> index 1a965b2..00b6f21 100644
>>> --- a/block.c
>>> +++ b/block.c
>>> @@ -6
From: Blue Swirl
When building on a 64 bit host which uses 'long' for int64_t,
GCC emits a warning:
CCblock/blkverify.o
/src/qemu/block/blkverify.c: In function `blkverify_verify_readv':
/src/qemu/block/blkverify.c:304: warning: long long int format, long
unsigned int arg (arg 3)
Rework a7
If bdrv_aio_flush returns NULL, this should be treated as an error.
Signed-off-by: Kevin Wolf
---
hw/ide/core.c | 12 +---
1 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/hw/ide/core.c b/hw/ide/core.c
index bc3e916..484e0ca 100644
--- a/hw/ide/core.c
+++ b/hw/ide/core.c
Signed-off-by: Kevin Wolf
Reviewed-by: Stefan Hajnoczi
---
hw/scsi-disk.c | 17 -
1 files changed, 16 insertions(+), 1 deletions(-)
diff --git a/hw/scsi-disk.c b/hw/scsi-disk.c
index 96acfe3..6815239 100644
--- a/hw/scsi-disk.c
+++ b/hw/scsi-disk.c
@@ -45,6 +45,7 @@ do { fprin
On 11/04/2010 08:15 AM, Kevin Wolf wrote:
The following changes since commit 5fc9cfedfa09199e10b5f9b67dcd286bfeae4f7a:
Fold send_all() wrapper unix_write() into one function (2010-11-03 12:48:09
-0500)
are available in the git repository at:
git://repo.or.cz/qemu/kevin.git for-anthony
B
Signed-off-by: Kevin Wolf
---
block/vpc.c | 21 +
1 files changed, 13 insertions(+), 8 deletions(-)
diff --git a/block/vpc.c b/block/vpc.c
index e50509e..416f489 100644
--- a/block/vpc.c
+++ b/block/vpc.c
@@ -439,6 +439,10 @@ static int vpc_write(BlockDriverState *bs, int64
This changes bdrv_flush to return 0 on success and -errno in case of failure.
It's a requirement for implementing proper error handle in users of bdrv_flush.
Signed-off-by: Kevin Wolf
Reviewed-by: Stefan Hajnoczi
---
block.c | 21 +
block.h |2 +-
b
This pulls the request completion for error cases from the caller to
scsi_disk_emulate_command. This should not change semantics, but allows to
reuse scsi_handle_write_error() for flushes in the next patch.
Signed-off-by: Kevin Wolf
Reviewed-by: Stefan Hajnoczi
---
hw/scsi-disk.c | 21 +++
Fix scsi-disk to use the usual completion paths that involve rerror/werror
handling instead of directly completing the requests in cases where
bdrv_aio_readv/writev returns NULL.
Signed-off-by: Kevin Wolf
---
hw/scsi-disk.c | 11 ++-
1 files changed, 6 insertions(+), 5 deletions(-)
di
Fix virtio-blk to use the usual completion path that involves werror handling
instead of directly completing the request in cases where bdrv_aio_flush
returns NULL.
Signed-off-by: Kevin Wolf
---
hw/virtio-blk.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/hw/virtio-b
The following changes since commit 5fc9cfedfa09199e10b5f9b67dcd286bfeae4f7a:
Fold send_all() wrapper unix_write() into one function (2010-11-03 12:48:09
-0500)
are available in the git repository at:
git://repo.or.cz/qemu/kevin.git for-anthony
Blue Swirl (1):
block: avoid a warning on
The cache content may be destroyed after a failed read, better not use it any
more.
Signed-off-by: Kevin Wolf
Reviewed-by: Stefan Hajnoczi
---
block/qcow2-cluster.c |1 +
block/qcow2-refcount.c |1 +
2 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/block/qcow2-cluster.c b
This implements the rerror option for SCSI disks.
It also includes minor changes to the write path where the same code is used
that was criticized in the review for the changes to the read path required for
rerror support.
Signed-off-by: Kevin Wolf
Reviewed-by: Stefan Hajnoczi
---
blockdev.c
Am 27.10.2010 20:19, schrieb Anthony Liguori:
> Signed-off-by: Anthony Liguori
>
> diff --git a/block.c b/block.c
> index 1a965b2..00b6f21 100644
> --- a/block.c
> +++ b/block.c
> @@ -603,10 +603,16 @@ int bdrv_open(BlockDriverState *bs, const char
> *filename, int flags,
> BlockDriver
Public bug reported:
The ldrexd / strexd instructions have a flawed implementation : it never
works properly. For example, the most simple code something like:
ldrexd r0, r1, [r2]
strexd r4, r0, r1, [r2]
which should usually have r4 as zero (successfully done the exclusive
store). However, the
The Ubuntu ARM gcc defaults to building in Thumb mode. The swp/swpb
instructions only exist in ARM mode.
In any case, swp is deprecated for more recent ARM architectures (ARMv6
and above), and on ARMv7 MP it will UNDEF and have to be emulated in the
kernel. So the testandset implementation should
On Thu, Nov 04, 2010 at 10:24:45AM +0100, Markus Armbruster wrote:
> Gleb Natapov writes:
>
> > On Sun, Oct 31, 2010 at 06:25:53PM -0400, Kevin O'Connor wrote:
> >> On Sun, Oct 31, 2010 at 01:40:01PM +0200, Gleb Natapov wrote:
> >> > This is current sate of the patch series for people to comment
On Thu, Nov 04, 2010 at 10:20:18AM +0100, Markus Armbruster wrote:
> Gleb Natapov writes:
>
> > Add "deriver_name" to DeviceInfo to use in device path building. In
>
> Typo "deriver". Same in subject.
>
Heh.
> > contrast to "name" "driver_name" should refer to functionality device
> > provide
Gleb Natapov writes:
> On Sun, Oct 31, 2010 at 06:25:53PM -0400, Kevin O'Connor wrote:
>> On Sun, Oct 31, 2010 at 01:40:01PM +0200, Gleb Natapov wrote:
>> > This is current sate of the patch series for people to comment on.
>> > I tried to use open firmware naming scheme to specify device path na
On Thu, Nov 04, 2010 at 09:46:57AM +0100, Markus Armbruster wrote:
> > But why order of device creation is important? It shouldn't be if we
> > want to move HW description into config file. We even may allow creating
> > piix3-ide with only second IDE bus, but not first.
>
> That's not how buses w
Gleb Natapov writes:
> Add "deriver_name" to DeviceInfo to use in device path building. In
Typo "deriver". Same in subject.
> contrast to "name" "driver_name" should refer to functionality device
> provides instead of particular device model like "name" does.
Why is that useful in a device pa
On Wed, Nov 03, 2010 at 06:03:50PM +, Ian Molton wrote:
> On 01/11/10 13:28, Anthony Liguori wrote:
> >On 11/01/2010 06:53 AM, Alon Levy wrote:
>
> >>While we (speaking as part of the SPICE developers) want to have the same
> >>support in our virtual GPU for 3d as we have for 2d, we just don't
Gleb Natapov writes:
> On Wed, Nov 03, 2010 at 06:22:11PM +0100, Markus Armbruster wrote:
>> Gleb Natapov writes:
>>
>> > On Wed, Nov 03, 2010 at 04:18:18PM +0100, Markus Armbruster wrote:
>> >> Gleb Natapov writes:
>> >>
>> >> > On Wed, Nov 03, 2010 at 02:39:52PM +0100, Markus Armbruster wro
On Wed, Nov 03, 2010 at 06:22:11PM +0100, Markus Armbruster wrote:
> Gleb Natapov writes:
>
> > On Wed, Nov 03, 2010 at 04:18:18PM +0100, Markus Armbruster wrote:
> >> Gleb Natapov writes:
> >>
> >> > On Wed, Nov 03, 2010 at 02:39:52PM +0100, Markus Armbruster wrote:
> >> >> Here's a generic an
69 matches
Mail list logo
