On Fri, 28 Oct 2016, Jonathon Jongsma wrote:
> If this is temporary, what's the long-term solution? I assume that the
> proper solution is not simple or obvious since then you'd have just
> fixed it properly. But I think that if we're committing a 'temporary'
> fix, the commit log should have an e
If this is temporary, what's the long-term solution? I assume that the
proper solution is not simple or obvious since then you'd have just
fixed it properly. But I think that if we're committing a 'temporary'
fix, the commit log should have an explanation of why it's temporary
and what the long-ter
Acked-by: Jonathon Jongsma
On Thu, 2016-10-27 at 18:42 +0200, Francois Gouget wrote:
> Signed-off-by: Francois Gouget
> ---
> src/spice-channel.h | 5 +++--
> src/spice-util.h | 2 --
> src/usb-device-manager.h | 2 +-
> 3 files changed, 4 insertions(+), 5 deletions(-)
>
> diff
Better, thanks.
Acked-by: Jonathon Jongsma
On Fri, 2016-10-28 at 11:39 +0200, Francois Gouget wrote:
> The new policy is that deprecation warnings should not be disabled
> globally or even per-file with SPICE_NO_DEPRECATED. Instead they
> should
> either be fixed, or ignored locally with
> G
On Fri, 2016-10-28 at 06:46 -0400, Frediano Ziglio wrote:
> >
> >
> > In commit beec1b41, we manually limited this property value in
> > _set_property(). But there's a simpler way to do it: via the param
> > spec
> > for the property.
> >
> > This also means that we can remove the warning log in
By the way, it's nice to know which repository the patch applies to by
looking at the email subject. It just helps get into the proper frame
of mind quickly before starting to review.
You can set that using the format.subjectprefix configuration
parameter. For example, I use the following configur
Keep a weak references to all registered channels, and free the list of
channels upon destruction.
---
server/reds.c | 14 ++
1 file changed, 14 insertions(+)
diff --git a/server/reds.c b/server/reds.c
index c2ac4d9..9034197 100644
--- a/server/reds.c
+++ b/server/reds.c
@@ -452,9 +45
RedsState is only freed at application exit, so this leak didn't matter
much in practice.
---
server/reds.c | 4
1 file changed, 4 insertions(+)
diff --git a/server/reds.c b/server/reds.c
index c03ed18..c2ac4d9 100644
--- a/server/reds.c
+++ b/server/reds.c
@@ -3681,8 +3681,12 @@ SPICE_GNUC_
---
server/display-channel.c | 3 +++
server/sound.c | 1 +
2 files changed, 4 insertions(+)
diff --git a/server/display-channel.c b/server/display-channel.c
index b483b57..e46d067 100644
--- a/server/display-channel.c
+++ b/server/display-channel.c
@@ -1904,6 +1904,9 @@ void display_ch
Add a 'self' parameter to all of the char device virtual functions so
that we don't have to play games with the 'opaque' pointer.
---
server/char-device.c | 8
server/char-device.h | 13 +
server/reds.c| 17 -
server/smartcard.c | 29
From: Christophe Fergeau
It's more natural to do things this way with glib, and this allows to
remove some coupling between Reds and RedCharDeviceVDIPort. Before this
commit, the RedCharDeviceVDIPort has to provide a on_free_self_token()
because Reds needs to know about it. With a signal, RedChar
Rather than having a single global channel for smartcard devices, store
the channel as a private member of the device it is associated with.
---
server/smartcard.c | 15 +--
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git a/server/smartcard.c b/server/smartcard.c
index 5391
The destructor function was calling reds_cleanup() for each server at
application exit. But reds_cleanup() only cleans up a couple of
statistics-related variables. We should actually free the servers by
calling spice_server_destroy(). This left reds_cleanup() unused, so the
contents of this functio
This is the final series of patches from the refactory branch. The RedClient
GObject patch series is still under review, but since these patches are not
dependant on that series, I decided to send these out while waiting for those
patches to get reviewed and re-worked.
Christophe Fergeau (1):
Re
Store the channel in the char device object, and unref it in dispose.
Previously, we were just creating a channel and allowing it to leak.
There may be better long-term solutions, but this works for now.
---
server/spicevmc.c | 25 -
1 file changed, 24 insertions(+), 1 dele
Internal types use the 'Red' prefix by convention, rather than the
'Spice' prefix. In addition, this type inherits from RedChannel, so
makes the code a lot clearer to call it a 'channel' rather than a
'state'.
---
server/spicevmc.c | 161 +++---
1 fi
On Fri, 2016-10-28 at 11:59 +0100, Frediano Ziglio wrote:
> A bit more similar to previous
The patch looks fine to me, but it's not clear to me why it's an
improvement. Can you expand on that in the commit log at least?
>
> Signed-off-by: Frediano Ziglio
> ---
> server/red-channel-client.c | 2
Signed-off-by: Francois Gouget
---
src/spiceqxl_io_port.c | 11 ---
1 file changed, 11 deletions(-)
diff --git a/src/spiceqxl_io_port.c b/src/spiceqxl_io_port.c
index 165b8a9..6721d3f 100644
--- a/src/spiceqxl_io_port.c
+++ b/src/spiceqxl_io_port.c
@@ -44,17 +44,6 @@
ret = &m_it
On Thu, 1 Sep 2016, Christophe Fergeau wrote:
[...]
> > You'll notice that this patch does not directly check for
> > create_xxx_decoder() errors. Instead it relies on the previous patchset
> > [2] deleting broken streams so that the following messages will run into
> > an unknown stream.
>
> I
On Fri, Oct 28, 2016 at 11:59:56AM +0100, Frediano Ziglio wrote:
> This was introduced with 96e94c6f32d4345ea25ef7a474fbd92de03b38ad
> (Convert RedChannel hierarchy to GObject).
> The id of CursorChannel/DisplayChannel were always 0 causing
> a wrong assertion on the code.
>
> Signed-off-by: Fredi
On Fri, Oct 28, 2016 at 11:59:53AM +0100, Frediano Ziglio wrote:
> There was a chance that on error GQueue were not
> initialized but an attempt to destroy it is made.
> This assure GQueue is initialized as soon as
s/assure/ensures
> possible. Note that red_channel_client_initable_init
> is calle
On Fri, Oct 28, 2016 at 11:59:52AM +0100, Frediano Ziglio wrote:
> Base object should be constructed first.
This sounds a bit too affirmative, glib/gio usually chains up to the
parents 'constructed' at the end of the vfunc. This sentence probably
can be dropped.
> In this case there is not much c
I believe I've acked this already
Acked-by: Christophe Fergeau
On Fri, Oct 28, 2016 at 11:59:51AM +0100, Frediano Ziglio wrote:
> This was introduced with 96e94c6f32d4345ea25ef7a474fbd92de03b38ad
> (Convert RedChannel hierarchy to GObject).
> The handle-acks settings was TRUE for CursorChannel a
On Fri, Oct 28, 2016 at 11:59:54AM +0100, Frediano Ziglio wrote:
> These fields need just channel to be set to be initialized.
> Move their initialization to constructor to make sure
> they are initialized as soon as possible.
Why?
Not related to this patch, but I'm not sure initializing pos and
The message has to be terminated by a newline as it's not going to get
added automatically.
Signed-off-by: Christophe Fergeau
---
drivers/gpu/drm/qxl/qxl_fb.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c
index 2cd
When the QXL driver receives a QXL_INTERRUPT_CLIENT_MONITORS_CONFIG interrupt,
we currently always notify userspace that there was some hotplug event.
However, gnome-shell/mutter is reacting to this event by attempting a
resolution change, which it does by issueing drmModeRmFB, drmModeAddFB,
and t
The use of drm_cvt_mode() in qxl_add_monitors_config_modes() means that
the resolutions we are going to present to user-space are going to be
rounded down to a multiple of 8. In the QXL arbitrary resolution case,
this is not useful.
This commit forces the actual width/height that was requested by t
qxl_crtc_set_from_monitors_config() is defined in qxl_drv.h but never
implemented.
Signed-off-by: Christophe Fergeau
---
drivers/gpu/drm/qxl/qxl_drv.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index da41e1f..590ba25 100644
Hey,
This patch series starts by doing some various small cleanups (patch 1 to 4),
and
then there are 2 fixes for issues which were noticed in fedora 25, the more
annoying one
being the one fixed by patch 5 where the VM screen would constantly flicker
when wayland is used. Patch 6 is very hacky,
They are not used outside of their respective source file
Signed-off-by: Christophe Fergeau
---
drivers/gpu/drm/qxl/qxl_cmd.c | 2 +-
drivers/gpu/drm/qxl/qxl_display.c | 4 ++--
drivers/gpu/drm/qxl/qxl_drv.h | 3 ---
3 files changed, 3 insertions(+), 6 deletions(-)
diff --git a/drivers/
qdev->gem.objects was initialized directly in qxl_device_init() rather
than going through qxl_gem_init(), and qxl_gem_fini() was never called.
Signed-off-by: Christophe Fergeau
---
drivers/gpu/drm/qxl/qxl_kms.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm
On Fri, 28 Oct 2016, Frediano Ziglio wrote:
[...]
> > In both cases I did the test by setting the min and max quantizer to the
> > same value, tracing the frame size in encode_frame() and playing the
> > same test video. For instance:
> >
> > - gstenc_opts = g_strdup("byte-stream=true aud=
On Thu, 27 Oct 2016, Francois Gouget wrote:
[...]
> Testing this you'll notice that reducing the available bandwidth results
> in freezes and recovery times that take a dozen second or more. This is
> because it's only after a sufficient backlog has been created (possibly
> a single large video
A bit more similar to previous
Signed-off-by: Frediano Ziglio
---
server/red-channel-client.c | 28 +++-
1 file changed, 15 insertions(+), 13 deletions(-)
diff --git a/server/red-channel-client.c b/server/red-channel-client.c
index 6c78237..42f1d2c 100644
--- a/server/re
Base object should be constructed first.
In this case there is not much change but better
to follow that style as all other constructors
do.
Signed-off-by: Frediano Ziglio
---
server/red-channel.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/server/red-channel.c b/serv
This was introduced with 96e94c6f32d4345ea25ef7a474fbd92de03b38ad
(Convert RedChannel hierarchy to GObject).
The id of CursorChannel/DisplayChannel were always 0 causing
a wrong assertion on the code.
Signed-off-by: Frediano Ziglio
---
server/cursor-channel.c | 1 +
server/display-channel.c | 1
This was introduced with 96e94c6f32d4345ea25ef7a474fbd92de03b38ad
(Convert RedChannel hierarchy to GObject).
The handle-acks settings was TRUE for CursorChannel and DisplayChannel.
Signed-off-by: Frediano Ziglio
---
server/cursor-channel.c | 1 +
server/display-channel.c | 1 +
2 files changed,
There was a chance that on error GQueue were not
initialized but an attempt to destroy it is made.
This assure GQueue is initialized as soon as
possible. Note that red_channel_client_initable_init
is called after all init and construction callbacks.
Signed-off-by: Frediano Ziglio
---
server/red-
Some regressions from former code and some style changes
to prevent some maintainance issues.
Frediano Ziglio (6):
Fix handle-acks regression
Call parent constructor always at the beginning
red-channel-client: Init pipe field during init
red-channel-client: Move incoming/ougoing initializa
These fields need just channel to be set to be initialized.
Move their initialization to constructor to make sure
they are initialized as soon as possible.
Signed-off-by: Frediano Ziglio
---
server/dummy-channel-client.c | 2 --
server/red-channel-client.c | 18 +-
2 files cha
>
> On Thu, 27 Oct 2016, Francois Gouget wrote:
>
> > When uncapped x264enc can compress the frames beyond recognition in low
> > bitrate situation. Beyond the set limit the gains are modest and it is
> > better to drop frames to reduce the bit rate further.
>
> In case anyone is interested, her
>
> In commit beec1b41, we manually limited this property value in
> _set_property(). But there's a simpler way to do it: via the param spec
> for the property.
>
> This also means that we can remove the warning log in red_worker_new()
> since GObject will automatically warn if a property is assi
>
> ---
> NEWS | 8
> 1 file changed, 8 insertions(+)
>
> diff --git a/NEWS b/NEWS
> index 6300f52..57877ea 100644
> --- a/NEWS
> +++ b/NEWS
> @@ -1,3 +1,11 @@
> +v0.8.0
> +==
> +- add multi monitor/dynamic resolution support on Win8+
> + when the qxl-wddm-dod driver is used
> +- d
This one is from Dave Airlie, not me, I've amended the author in my
local copy.
Christophe
On Fri, Oct 28, 2016 at 12:18:21PM +0200, Christophe Fergeau wrote:
> This is not working properly at the moment.
> ---
> src/qxl_uxa.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/src/qxl_u
This is not working properly at the moment.
---
src/qxl_uxa.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/qxl_uxa.c b/src/qxl_uxa.c
index 875d663..694db27 100644
--- a/src/qxl_uxa.c
+++ b/src/qxl_uxa.c
@@ -414,6 +414,8 @@ qxl_create_pixmap (ScreenPtr screen, int w, int h, int
depth,
---
src/spiceqxl_main_loop.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/spiceqxl_main_loop.c b/src/spiceqxl_main_loop.c
index db89b6d..49b715a 100644
--- a/src/spiceqxl_main_loop.c
+++ b/src/spiceqxl_main_loop.c
@@ -274,7 +274,7 @@ static void xspice_block_handler(poin
Hey,
It has been a long time since the last QXL Xorg driver release, would be nice
to have one soon :)
Given that Xspice is probably not working with Xorg 1.19, I've added a
configure.ac patch erroring out when both are in use.
I've also added two patches that we have been carrying for a long tim
---
NEWS | 13 +
configure.ac | 2 +-
2 files changed, 14 insertions(+), 1 deletion(-)
diff --git a/NEWS b/NEWS
index 59e0e4b..237d121 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,16 @@
+Major changes in 0.1.5
+==
+
+Fix X restarts when kms is in use with multi
xspice needs to be updated to cope with some X.Org 1.19 API changes,
better to make that explicit at configure time rather than letting
people discover the hard way (it builds with warnings but will not work)
that it's broken.
---
configure.ac | 8 ++--
1 file changed, 6 insertions(+), 2 delet
From: Dave Airlie
This should help with bug #974198
---
src/qxl_uxa.c | 8
1 file changed, 8 insertions(+)
diff --git a/src/qxl_uxa.c b/src/qxl_uxa.c
index 694db27..fa7d85f 100644
--- a/src/qxl_uxa.c
+++ b/src/qxl_uxa.c
@@ -221,12 +221,16 @@ qxl_has_composite (qxl_screen_t *qxl)
{
#i
Signed-off-by: Francois Gouget
---
This is an alternative take on the previous patch: instead of reusing
the GLib macro it just duplicates its logic.
GLib approach
-
Pros: - Automatic support for whichever platforms GLib supports.
- Simple.
Cons: - If the project uses GLib it
The new policy is that deprecation warnings should not be disabled
globally or even per-file with SPICE_NO_DEPRECATED. Instead they should
either be fixed, or ignored locally with
G_GNUC_{BEGIN,END}_IGNORE_DEPRECATIONS.
Signed-off-by: Francois Gouget
---
This is a proposed rewording for this
This gains us automatic support for whichever compilers GLib supports in
this macro when used in projects that use GLib.
Signed-off-by: Francois Gouget
---
This can make sense if we consider GLib to be a platform similar to gcc,
MSVC: it just provides a G_XXX API instead of an __attribute((xxx)
Il 28/10/2016 10:59, Christophe Fergeau ha scritto:
> On Thu, Oct 27, 2016 at 04:52:57PM +0200, Fabio Fantoni wrote:
>> Il 27/10/2016 12:03, Christophe Fergeau ha scritto:
>>> On Thu, Oct 27, 2016 at 10:40:47AM +0200, Fabio Fantoni wrote:
Is there any vdagent testing build with all newer patch
From: Jeremy Bicha
Ubuntu's Unity still ships the Startup Applications app
(gnome-session-properties).
Setting NoDisplay=true means this autostart file won't show up in there
since this isn't something we really want users to unknowingly disable
that easily.
Resolves: https://bugs.freedesktop.o
On Thu, Oct 27, 2016 at 04:52:57PM +0200, Fabio Fantoni wrote:
> Il 27/10/2016 12:03, Christophe Fergeau ha scritto:
> > On Thu, Oct 27, 2016 at 10:40:47AM +0200, Fabio Fantoni wrote:
> >> Is there any vdagent testing build with all newer patches available
> >> somewhere?
> > https://copr.fedorainf
---
NEWS | 8
1 file changed, 8 insertions(+)
diff --git a/NEWS b/NEWS
index 6300f52..57877ea 100644
--- a/NEWS
+++ b/NEWS
@@ -1,3 +1,11 @@
+v0.8.0
+==
+- add multi monitor/dynamic resolution support on Win8+
+ when the qxl-wddm-dod driver is used
+- disable drag and drop when the s
On Thu, Oct 27, 2016 at 12:39:02PM -0400, Frediano Ziglio wrote:
> >
> > Hey,
> >
> > On Wed, Oct 26, 2016 at 09:45:15AM +0100, Frediano Ziglio wrote:
> > > These functions were implementing the same stuff as empty
> > > messages functions provided by RedChannel so reuse them.
> > >
> > > The im
On Thu, Oct 27, 2016 at 12:41:33PM -0400, Frediano Ziglio wrote:
> >
> > Hey,
> >
> > On Wed, Oct 26, 2016 at 09:45:15AM +0100, Frediano Ziglio wrote:
> > > These functions were implementing the same stuff as empty
> > > messages functions provided by RedChannel so reuse them.
> > >
> > > The im
59 matches
Mail list logo
