When an output is resized (permanent mode switch), we should also notify the
shell client so that the panel and background fits to the new screen dimensions.
Signed-off-by: David Fort
---
desktop-shell/shell.c | 93 +++
desktop-shell/shell.h | 8 +
When an output permanently changes its resolution, the output on the right
should be moved accordingly. We also add an event for output resizing so that
plugins can react when an output is resized.
Signed-off-by: David Fort
Reviewed-by: Pekka Paalanen
---
src/compositor.c | 35 +
So, the RDP compositor were given some love (2 patches integrated).
I have made the changes for seat releasing.
And I have also modified the handling of output resize and it should be better
now.
David Fort (3):
rdp: don't release the seat until it is safe v2
compositor: recompute output pos
Releasing a seat is not safe, so let's just announce it without keyboard
and mouse until this is fixed. Without this patch we just can't reconnect on
the RDP compositor as it crashes.
v2: fixed the leak of the xkb_keymap
Signed-off-by: David Fort
Acked-by: Pekka Paalanen
Reviewed-by: Sam Spilsb
Hi Bill,
> On May 17, 2016, at 1:18 AM, spit...@gmail.com wrote:
>
> From: Bill Spitzak
>
> Signed-off-by: Bill Spitzak
> ---
> src/scanner.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/scanner.c b/src/scanner.c
> index 037ebdb..33271bc 100644
> --- a/src/scann
I've inlined some replies below.
On Fri, May 27, 2016 at 1:13 PM Yong Bakos wrote:
> On May 27, 2016, at 10:29 AM, Mike Blumenkrantz
> wrote:
> >
> > this adds a method for compositors to change various draw attributes
> > for a surface
> >
> > Signed-off-by: Mike Blumenkrantz
> > Signed-off-b
On May 27, 2016, at 10:29 AM, Mike Blumenkrantz wrote:
>
> this adds a method for compositors to change various draw attributes
> for a surface
>
> Signed-off-by: Mike Blumenkrantz
> Signed-off-by: Jonas Ådahl
Hi Mike & Jonas,
A question about communicating default state, and some
minor nits
Hi guys,
Currently during cross compilation of the wayland library the
wayland-scanner utility is built by cross-gcc and then used on the
developer machine which lead build error.
I suggest to build development version, for example,
$(top_builddir)/build-wayland-scanner and to use this utility du
From: Bill Spitzak
Signed-off-by: Bill Spitzak
---
src/scanner.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/scanner.c b/src/scanner.c
index 037ebdb..33271bc 100644
--- a/src/scanner.c
+++ b/src/scanner.c
@@ -1350,7 +1350,7 @@ emit_structs(struct wl_list *message_lis
From: Bill Spitzak
Splits the allocation of wl_proxy from the assignment of the id. This removes
some code such as the malloc from the lock, possibly improving multithreaded
performance.
Removes unnecessary lock in wl_proxy_create_wrapper.
This does not change the public api.
Signed-off-by: Bi
From: Bill Spitzak
This function assigns the id into an existing wl_proxy structure. Allows
setting of the queue before binding so that events are sent to the correct
thread.
The arguments are purposely swapped around so that just changing
wl_registry_bind
to wl_registry_bind_id will not compil
this adds a method for compositors to change various draw attributes
for a surface
Signed-off-by: Mike Blumenkrantz
Signed-off-by: Jonas Ådahl
---
unstable/xdg-shell/xdg-shell-unstable-v6.xml | 69
1 file changed, 69 insertions(+)
diff --git a/unstable/xdg-shell/xd
From: Bill Spitzak
I think this makes it less confusing. "factory" was already used for internal
functions.
Signed-off-by: Bill Spitzak
---
src/wayland-client-core.h | 10 +-
src/wayland-client.c | 48 +++
2 files changed, 29 insertions(
From: Bill Spitzak
Scanner produces a wl_xyz_new() for each object, and for each constructor
request it produces an "_id" version that takes one of these, instead of
creating one. This allows the queue and listener to be set before creation of
the object so events always go to the correct queue.
On Sun, May 22, 2016 at 02:02:57PM +0500, Alexander E. Patrakov wrote:
> 22.05.2016 13:33, Alexander E. Patrakov пишет:
> >22.05.2016 03:51, Zbigniew Jędrzejewski-Szmek пишет:
> >>Hi,
> >>
> >>systemd v230 has been tagged. Enjoy!
> >>
> >>CHANGES WITH 230:
> >
> >
> >
> >>* Framebuffer devi
On Mon, May 16, 2016 at 3:43 PM, Yong Bakos wrote:
> On May 7, 2016, at 5:57 AM, Otavio Salvador wrote:
>>
>> From: Tom Hochstein
>>
>> Exit the program if an unrecognized command line option is found.
>>
>> Signed-off-by; Tom Hochstein
>> Signed-off-by: Otavio Salvador
>
> Simple enough of a
The first three patches are cleanup that does not change the api. In particular
this change moves the malloc of wl_proxy outside the lock.
Patch 4 adds an "_id" version of all the constructors that takes an existing
wl_proxy, and a "_new" function to create these proxies. This is an alternative
to
Hello Giulio,
With this patch, and mine to fix MODULEDIR of patch #2, this patch does
not build due to incomplete type:
src/screenshooter.c:46:24: error: field 'process' has incomplete type
struct weston_process process;
The patch look fine, if I ignore this building error, I also saw a lot
of
Hi David,
On Fri, May 27, 2016 at 6:48 AM, Hardening wrote:
> Le 26/04/2016 23:34, David Fort a écrit :
>> Compared to the first version:
>> compositor: recompute output position on mode switch
>> desktop-shell: resize background, panel and lock surface on output
>> resize
>>
I checked m
Hello Giulio,
The idea of the patch is fine, but the patch doesn't work. See following
comments. (@Quentin: no it's does not work as it is ;) )
On 24/05/2016 18:59, Giulio Camuffo wrote:
> The backends are now installed in lib/libweston-0, and the include
> files that will be used by libweston in
Hello Giulio,
The patch do what it describe.
Reviewed-by: Benoit Gschwind
On 24/05/2016 18:59, Giulio Camuffo wrote:
> ---
> src/input.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/src/input.c b/src/input.c
> index 8fe898c..c6cdedd 100644
> --- a/src/input.c
>
Multiplanar formats are supported by using drmModeAddFB2 and bypassing
gbm. If drmModeAddFB2 isn't available, the existing gbm bo import path
is used and multiplanar formats are unsupported.
Signed-off-by: Tomohito Esaki
---
src/compositor-drm.c | 47 +++--
This implementations bypasses gbm and passes the dmabuf handles directly
to libdrm for composition.
Signed-off-by: Tomohito Esaki
---
src/compositor-drm.c | 128 +++
1 file changed, 108 insertions(+), 20 deletions(-)
diff --git a/src/compositor-dr
These patches add support multiplanar formats by using linux_dmabuf instead of
gbm for scanout/sprite plane importing.
When rendering linux_dmabuf buffers, gbm isn't required, and multiplanar
formats can be supported by using drmModeAddFB2() instead of gbm_import().
The first patch adapts the fun
On Tue, 26 Apr 2016 23:34:07 +0200
David Fort wrote:
> When an output permanently changes its resolution, the output on the right
> should be moved accordingly. We also add an event for output resizing so that
> plugins can react when an output is resized.
>
> Signed-off-by: David Fort
> ---
>
On Tue, 26 Apr 2016 23:34:08 +0200
David Fort wrote:
> When an output is resized (permanent mode switch), we should also notify the
> shell client so that the panel, background and the lock screen fits to the
> new screen dimensions.
>
> Signed-off-by: David Fort
> ---
> desktop-shell/shell.c
On Fri, 20 May 2016 15:02:02 +0200
Hardening wrote:
> Le 20/05/2016 11:58, Sam Spilsbury a écrit :
> > On Fri, May 20, 2016 at 5:33 PM, David Fort wrote:
> >> Releasing a seat is not safe, so let's just announce it without keyboard
> >> and mouse until this is fixed. Without this patch we just
27 matches
Mail list logo