Acked-by: Jonathon Jongsma
On Mon, 2018-07-02 at 08:43 +0100, Frediano Ziglio wrote:
> Also add some comments.
>
> Signed-off-by: Frediano Ziglio
> ---
> spice-protocol | 2 +-
> vdagent/vdagent.cpp | 45 ++-
> --
> 2 files changed, 27 insertions(
On Thu, 2018-06-28 at 14:23 +0100, Frediano Ziglio wrote:
> QUIC is implemented using 2 C template files.
> One for single channel and one for multiple (RGB) channels.
> Unify the 2 templates to have a single source.
>
> Frediano Ziglio (6):
> quic: Call encode from golomb_coding
> quic: Conti
Looks like it was only ever called from the old client code that was
removed from the spice-server repository.
Acked-by: Jonathon Jongsma
On Tue, 2018-07-03 at 16:10 +0200, Christophe Fergeau wrote:
> Nothing calls it in spice-gtk or spice-server
>
> Signed-off-by: Christophe Fergeau
> ---
>
On Tue, 2018-07-03 at 17:10 +0200, Christophe de Dinechin wrote:
> There is a lot of debate, distributed across a large number of
> patches, regarding similar SPICE and glib facilities. For a number of
> things, there are two sets of parallel APIs with slightly different
> behaviour. This is undesi
>
> On 03/07/18 12:04, Uri Lublin wrote:
> > On 07/03/2018 05:35 PM, Eduardo Lima (Etrunko) wrote:
> >> This test was timing out when running in gitlab-ci, so decreasing the
> >> loop count in order to make it run faster. Example:
> >>
> >> https://gitlab.freedesktop.org/etrunko/spice-common/-/job
On 03/07/18 12:04, Uri Lublin wrote:
> On 07/03/2018 05:35 PM, Eduardo Lima (Etrunko) wrote:
>> This test was timing out when running in gitlab-ci, so decreasing the
>> loop count in order to make it run faster. Example:
>>
>> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
>
> Hi,
>
> On Thu, Jun 28, 2018 at 09:43:47PM +0100, Frediano Ziglio wrote:
> > This avoids compilers detect misaligned access.
>
> Maybe "This helps some compilers to realize QXLRam is correctly aligned"?
>
Well, the current code tells the compiler that this structure
is NOT aligned.
But we actually
On Thu, Jun 28, 2018 at 09:43:47PM +0100, Frediano Ziglio wrote:
> This avoids compilers detect misaligned access.
Maybe "This helps some compilers to realize QXLRam is correctly aligned"?
> The structure is allocated at the beginning of a page so surely
> will be 4 bytes aligned.
Not immediatl
Looks simpler to me, though I don't know the history of why these were
separated like this. Unless somebody else has objections:
Acked-by: Jonathon Jongsma
On Thu, 2018-06-28 at 14:23 +0100, Frediano Ziglio wrote:
> golomb_coding is always followed by a encode call.
> Simplify code calling dire
>
> On Thu, Jun 28, 2018 at 07:40:19AM +0100, Frediano Ziglio wrote:
> > Make code faster on platforms not supporting unaligned access
> > by default.
>
> How does this work exactly? Rather than silencing the warning with
> SPICE_UNALIGNED_CAST, and then potentially dereferencing a non-aligned
>
Apologies beforehand if it's a bit offtopic; not sure what place would be
more suitable...
Regarding storage devices, Proxmox wiki mentions that SCSI virtio driver is
in its way to replace current virtio block driver.
https://pve.proxmox.com/wiki/Paravirtualized_Block_Drivers_for_Windows
However t
>
> Hi,
>
>
> On 05/25/2018 03:55 PM, Frediano Ziglio wrote:
> > display_queue is queued with decoded frames ready to be displayed.
> > However current code can insert a timeout before displaying and
> > removing the queued frames. As the frames are not compressed the
> > usage of memory by this
On Thu, Jun 28, 2018 at 07:40:19AM +0100, Frediano Ziglio wrote:
> Make code faster on platforms not supporting unaligned access
> by default.
How does this work exactly? Rather than silencing the warning with
SPICE_UNALIGNED_CAST, and then potentially dereferencing a non-aligned
32 bit value, and
Mr. Cole Robinson? His name indeed appears in the Windows VirtIO ISO
changelog from the fedoraproject site...
Do you know if he lurkes this mailing list?
2018-07-03 10:08 GMT-05:00 Christophe Fergeau :
> On Tue, Jul 03, 2018 at 09:47:38AM -0500, Carlos González wrote:
> > Thanks yet again.
> >
>
There is a lot of debate, distributed across a large number of patches,
regarding similar SPICE and glib facilities. For a number of things, there are
two sets of parallel APIs with slightly different behaviour. This is
undesirable, as it introduces confusion.
Pros: this lets us customize the b
On Tue, Jul 03, 2018 at 09:47:38AM -0500, Carlos González wrote:
> Thanks yet again.
>
> https://lists.freedesktop.org/archives/spice-devel/2018-June/044236.html
> If this patch makes it to release, which win-virtio ISO branch is it
> normally put in? "Latest"?
I don't know how Cole handles these
On 07/03/2018 05:35 PM, Eduardo Lima (Etrunko) wrote:
This test was timing out when running in gitlab-ci, so decreasing the
loop count in order to make it run faster. Example:
https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
Hi,
Just curious, why the test failed with meson but
On Tue, Jul 03, 2018 at 04:31:10PM +0200, Christophe de Dinechin wrote:
> I think we never had a discussion of what we really want in each case, and
> that’s causing the confusion.
>
> First, a meta-rule. Like you, there is a lot in SPICE code I don’t
> like. When in doubt, I try to use consisten
Hi,
On 05/25/2018 03:55 PM, Frediano Ziglio wrote:
display_queue is queued with decoded frames ready to be displayed.
However current code can insert a timeout before displaying and
removing the queued frames. As the frames are not compressed the
usage of memory by this queue can became in some
On 03/07/18 11:40, Frediano Ziglio wrote:
>>
>> This test was timing out when running in gitlab-ci, so decreasing the
>> loop count in order to make it run faster. Example:
>>
>> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
>>
>> Signed-off-by: Eduardo Lima (Etrunko)
>
> Acked-
Thanks yet again.
https://lists.freedesktop.org/archives/spice-devel/2018-June/044236.html
If this patch makes it to release, which win-virtio ISO branch is it
normally put in? "Latest"?
Where can one check current spice-server changelog or version in order to
see whether patch has been released
>
> This test was timing out when running in gitlab-ci, so decreasing the
> loop count in order to make it run faster. Example:
>
> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
>
> Signed-off-by: Eduardo Lima (Etrunko)
Acked-by: Frediano Ziglio
Frediano
> ---
> tests/tes
On 03/07/18 10:35, Frediano Ziglio wrote:
>>
>> On 02/07/18 07:31, Frediano Ziglio wrote:
>
> Test-region timeouts, maybe we should increase the timeout for all tests?
>
> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
>
Maybe just for this test w
This test was timing out when running in gitlab-ci, so decreasing the
loop count in order to make it run faster. Example:
https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
Signed-off-by: Eduardo Lima (Etrunko)
---
tests/test-region.c | 2 +-
1 file changed, 1 insertion(+), 1 delet
> On 3 Jul 2018, at 15:51, Christophe Fergeau wrote:
>
> On Tue, Jul 03, 2018 at 03:16:19PM +0200, Christophe de Dinechin wrote:
diff --git a/common/canvas_base.c b/common/canvas_base.c
index 6bf6e5d..bbaaf96 100644
--- a/common/canvas_base.c
+++ b/common/canvas_base.c
Nothing calls it in spice-gtk or spice-server
Signed-off-by: Christophe Fergeau
---
common/sw_canvas.c | 26 --
common/sw_canvas.h | 11 ---
2 files changed, 37 deletions(-)
diff --git a/common/sw_canvas.c b/common/sw_canvas.c
index c41edb1..c5528c2 100644
--- a/
> On 3 Jul 2018, at 15:29, Frediano Ziglio wrote:
>
>>
>> From: Christophe de Dinechin
>>
>> The spice_ variants simply mirror the glib API,
>> without adding significant value.
>>
>> Signed-off-by: Christophe de Dinechin
>> --
>> Following up on Christophe Fergeau's comment:
>>
>>> I'd f
On Tue, Jul 03, 2018 at 03:51:58PM +0200, Christophe Fergeau wrote:
> > Depending on context, it may. So what? Are you suggesting to do something
> > else?
>
> if (d == NULL) {
> abort();
> free(d);
(obviously not 'd' here, but free(something);)
Christophe
signature.asc
Description:
On Tue, Jul 03, 2018 at 03:16:19PM +0200, Christophe de Dinechin wrote:
> >> diff --git a/common/canvas_base.c b/common/canvas_base.c
> >> index 6bf6e5d..bbaaf96 100644
> >> --- a/common/canvas_base.c
> >> +++ b/common/canvas_base.c
> >> @@ -3072,6 +3072,7 @@ static void canvas_draw_stroke(SpiceCan
On Mon, Jun 25, 2018 at 10:50:27AM +0100, Frediano Ziglio wrote:
> self_bitmap flag is used for some complex drawing not possible
> by QXL_DRAW_COPY commands. Having this flag set causes
> spice-server do draw part of the screen, copy that part on new
> allocated image and reduce network optimisati
>
> On 02/07/18 07:31, Frediano Ziglio wrote:
> >>
> >>>
> >>> Test-region timeouts, maybe we should increase the timeout for all tests?
> >>>
> >>> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
> >>>
> >>
> >> Maybe just for this test we could decrease the for loop counter
> >>
>
> From: Christophe de Dinechin
>
> The spice_ variants simply mirror the glib API,
> without adding significant value.
>
> Signed-off-by: Christophe de Dinechin
> --
> Following up on Christophe Fergeau's comment:
>
> > I'd favour g_return_if_fail(), I'd really like to kill these various
>
> On 3 Jul 2018, at 12:11, Frediano Ziglio wrote:
>
>>
>> On Fri, Jun 29, 2018 at 05:21:22PM +0200, Christophe de Dinechin wrote:
>>> From: Christophe de Dinechin
>>>
>>> In some error cases, canvas_get_image may return NULL.
>>> When this happens, calls like pixman_image_get_width(s)
>>> wi
Hey,
On Tue, Jun 26, 2018 at 10:31:00AM +0100, Frediano Ziglio wrote:
> The type of the image is just copied from network without
> any check and later used for array indexing.
Have you considered something like this? This way we can be sure no code
can overflow the IS_IMAGE_TYPE_PLT[] array. Sim
Acked-by: Christophe de Dinechin
> On 2 Jul 2018, at 09:43, Frediano Ziglio wrote:
>
> Signed-off-by: Frediano Ziglio
> ---
> vdagent/vdagent.cpp | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/vdagent/vdagent.cpp b/vdagent/vdagent.cpp
> index 1e8f27c..78c42d1 100644
> On 3 Jul 2018, at 11:47, Christophe Fergeau wrote:
>
> On Fri, Jun 29, 2018 at 05:21:22PM +0200, Christophe de Dinechin wrote:
>> From: Christophe de Dinechin
>>
>> In some error cases, canvas_get_image may return NULL.
>> When this happens, calls like pixman_image_get_width(s)
>> will cras
From: Christophe de Dinechin
The spice_ variants simply mirror the glib API,
without adding significant value.
Signed-off-by: Christophe de Dinechin
--
Follow up a comment from Christophe Fergeau:
>I'd favour g_return_if_fail(), I'd really like to kill these various
> spice_* calls mirroring g
From: Christophe de Dinechin
The spice_ variants simply mirror the glib API,
without adding significant value.
Signed-off-by: Christophe de Dinechin
--
Following up on Christophe Fergeau's comment:
> I'd favour g_return_if_fail(), I'd really like to kill these various
> spice_* calls mirroring
On 02/07/18 07:31, Frediano Ziglio wrote:
>>
>>>
>>> Test-region timeouts, maybe we should increase the timeout for all tests?
>>>
>>> https://gitlab.freedesktop.org/etrunko/spice-common/-/jobs/6546
>>>
>>
>> Maybe just for this test we could decrease the for loop counter
>> near the end of this te
On Tue, 2018-07-03 at 06:56 -0400, Frediano Ziglio wrote:
> >
> > Hi all,
> >
> > after posting the PoC patch series for the monitors_config rework and
> > the discussion there getting a bit hard to follow, I thought I'd try to
> > summarize the proposed protocol. There are also spice-gtk API cha
Changes since v2:
-Adding support in h265
-printing and logging changes
-Fixes according to previous comments
-Rebase
*Switching between capturing method is still using the XLIB CAPTURE
macro, hopefully the whole xlib capturing system could be dropped soon
*Logging and printing are currently do
Gstreamer based plugin utilizing gstreamer elements to capture
screen from X, convert and encode into h264/h265/vp8/vp9/mjpeg stream
Configure with --enable-gstreamer, will be built as a separate plugin.
The plugin was made for testing purposes, it was mainly tested with
the x264enc (h264 is the d
Acked-by: Christophe Fergeau
(when reverting this commit, compilation of the test case fails, with
this commit in, the test case succeeds)
Christophe
On Thu, Jun 28, 2018 at 07:40:20AM +0100, Frediano Ziglio wrote:
> Check the previous fix for "zero" attribute works correctly
> (commit bc9df581
Acked-by: Christophe Fergeau
On Tue, Jul 03, 2018 at 11:24:03AM +0100, Frediano Ziglio wrote:
> Pointers to memory allocated in user space are never NULL.
> The only exception can be if you explicitly map memory at zero.
> There is however no reasons for such requirement and this practise
> was
>
> Hi all,
>
> after posting the PoC patch series for the monitors_config rework and
> the discussion there getting a bit hard to follow, I thought I'd try to
> summarize the proposed protocol. There are also spice-gtk API changes
> to consider, which I've decided to leave for a later email.
>
Most users of spice_common.h don't need it, or only need log.h. It only
has a few users outside of spice-common. It's not very well defined
which headers it should contain. This commit removes spice_common.h in
favour of direct inclusion of the needed headers.
Signed-off-by: Christophe Fergeau
--
spice-gtk was the last user, and stopped using it on Jun 14th 2017
in 040090ccba34 "build-sys: remove -DSPICE_DISABLE_ABORT"
Signed-off-by: Christophe Fergeau
---
common/log.c | 4
common/spice_common.h | 4
tests/test-logging.c | 14 ++
3 files changed, 2 inser
There are only 2 users in spice-common, and none in spice-gtk/spice
Signed-off-by: Christophe Fergeau
---
common/pixman_utils.c | 4 ++--
common/spice_common.h | 2 --
2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/common/pixman_utils.c b/common/pixman_utils.c
index 81806c9..5d1c
Signed-off-by: Christophe Fergeau
---
server/dispatcher.c | 1 -
server/red-common.h | 1 -
2 files changed, 2 deletions(-)
diff --git a/server/dispatcher.c b/server/dispatcher.c
index 0953db875..48dc915a6 100644
--- a/server/dispatcher.c
+++ b/server/dispatcher.c
@@ -26,7 +26,6 @@
#include
#
>
> On Tue, Jun 19, 2018 at 11:05:26AM +0100, Frediano Ziglio wrote:
> > Pointers to memory allocated in user space are never NULL.
> > The only exception can be if you explicitly map memory at zero.
> > There is however no reasons for such requirement and this practise
> > was also removed from L
Pointers to memory allocated in user space are never NULL.
The only exception can be if you explicitly map memory at zero.
There is however no reasons for such requirement and this practise
was also removed from Linux due to security reasons.
This API looks copied from a kernel environment where va
ping
>
> Check the previous fix for "zero" attribute works correctly
> (commit bc9df5816210efbe15ca03e82510f34174784082, "marshal: Fix a bug
> with zero attribute").
>
> Signed-off-by: Frediano Ziglio
> ---
> tests/test-marshallers.c | 13 +
> tests/test-marshallers.h | 4
>
> On Fri, Jun 29, 2018 at 05:21:22PM +0200, Christophe de Dinechin wrote:
> > From: Christophe de Dinechin
> >
> > In some error cases, canvas_get_image may return NULL.
> > When this happens, calls like pixman_image_get_width(s)
> > will crash. Add additional error paths to deal with
> > thes
>
> spice-gtk was the last user, and stopped using it in 90ccba34
> "build-sys: remove -DSPICE_DISABLE_ABORT"
>
I would put the date of the commit (not that important).
Looking at 90ccba34 it states "but since it works fine without this flag
since 2012 (commit 05a203af120016)", not clear why wo
On Fri, Jun 29, 2018 at 05:21:22PM +0200, Christophe de Dinechin wrote:
> From: Christophe de Dinechin
>
> In some error cases, canvas_get_image may return NULL.
> When this happens, calls like pixman_image_get_width(s)
> will crash. Add additional error paths to deal with
> these cases.
I assum
spice-gtk was the last user, and stopped using it in 90ccba34
"build-sys: remove -DSPICE_DISABLE_ABORT"
Signed-off-by: Christophe Fergeau
---
common/log.c | 4
common/spice_common.h | 4
tests/test-logging.c | 14 ++
3 files changed, 2 insertions(+), 20 deletion
Acked-by: Christophe Fergeau
On Wed, Jun 27, 2018 at 08:00:36AM +0100, Frediano Ziglio wrote:
> Check if encoder contains an invalid character.
>
> Signed-off-by: Frediano Ziglio
> ---
> server/tests/test-codecs-parsing.c | 6 ++
> 1 file changed, 6 insertions(+)
>
> diff --git a/server/
Acked-by: Christophe Fergeau
On Wed, Jun 27, 2018 at 08:00:37AM +0100, Frediano Ziglio wrote:
> In case we pass something like "spice:mjpeg$%*" the last part is
> ignore making the string parse correctly.
> A single pair should end by either string terminator or pair terminator.
>
> Signed-off-
> On 2 Jul 2018, at 18:23, Frediano Ziglio wrote:
>
>
>
> - Original Message -
>> From: "Christophe de Dinechin"
>> To: "Eduardo Lima (Etrunko)"
>> Cc: spice-devel@lists.freedesktop.org
>> Sent: Monday, 2 July, 2018 3:01:41 PM
>> Subject: Re: [Spice-devel] [PATCH spice-common] Updat
On Mon, Jul 02, 2018 at 01:20:39PM -0500, Carlos González wrote:
> https://lists.freedesktop.org/archives/spice-devel/2018-June/044237.html
> I just read spice-server thing is actually a hypervisor part. So if this
> patch makes it to release, what would I need to update in order to test?
> All lib
60 matches
Mail list logo
