Since the repaint status of the flushed output may be reset if a output
repaint is failed, it is necessary to clear the repainted flag
immediately after output repaint flush/cancel.
Signed-off-by: Tomohito Esaki
---
libweston/compositor.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-
This code calls into EGL to see if the dmabuf import modifiers
extension is available, and if not it assumes XRGB is supported.
Rather than disabling this client doesn't get build if one disables
EGL, we can just remove this and stop lying to ourselves.
Signed-off-by: Emilio Pozuelo Monfort
This reproduces https://bugs.freedesktop.org/show_bug.cgi?id=94735.
Signed-off-by: Emilio Pozuelo Monfort
---
tests/reference/subsurface_mapped-00.png | Bin 0 -> 799 bytes
tests/reference/subsurface_mapped-01.png | Bin 0 -> 841 bytes
tests/subsurface-shot-test.c | 76 +
If a surface has no content (e.g. no buffer), then it shouldn't
be mapped, so that its subsurfaces don't get mapped either.
This works fine in the desktop-shell, but is currently broken
on the weston-test module.
https://bugs.freedesktop.org/show_bug.cgi?id
Signed-off-by: Emilio Pozuelo Monfort
I forgot to git add the test ref images. That's the only change here.
Emilio Pozuelo Monfort (2):
shot: add test for sub-surface with unmapped parent
weston-test: don't map surfaces that have no content
tests/reference/subsurface_mapped-00.png | Bin 0 -> 799 bytes
tests/reference/subsurface
Hi Pekka,
On Fri, 26 Jan 2018 at 12:45, Pekka Paalanen wrote:
> On Wed, 20 Dec 2017 12:26:51 + Daniel Stone wrote:
> > @@ -3116,9 +3136,16 @@ drm_output_propose_state(struct weston_output
> > *output_base,
> > if (next_plane == primary)
> > pixman_region3
On Mon, 9 Jul 2018 15:27:46 +0100
Daniel Stone wrote:
> Collect the fallback definitions of static_assert() from desktop-shell
> and the test shell, and move them to helpers.h. This allows code
> throughout the tree to use static_assert() for build-time assertions,
> where it is supported by the
Collect the fallback definitions of static_assert() from desktop-shell
and the test shell, and move them to helpers.h. This allows code
throughout the tree to use static_assert() for build-time assertions,
where it is supported by the compiler.
As GCC goes out of its way to only add static_assert(
Hi,
On Mon, 9 Jul 2018 at 11:09, Simon Ser wrote:
>
> Bump: this is the source of bugs in xwayland [1].
>
> [1]: https://lists.x.org/archives/xorg-devel/2018-July/057285.html
>
> On May 18, 2018 9:40 PM, Simon Ser wrote:
> > Signed-off-by: Simon Ser
> > ---
> >
> > I believe it's easy to unders
On Mon, 9 Jul 2018 14:23:12 +0100
Daniel Stone wrote:
> Now that we collect information about which modifiers are supported for
> KMS display, and are able to create KMS framebuffers with modifiers,
> begin using the modifier-aware GBM API.
>
> Client buffers from dmabuf already store multi-pla
On Mon, 9 Jul 2018 14:23:11 +0100
Daniel Stone wrote:
> From: Sergi Granell
>
> The per-plane IN_FORMATS KMS property describes the format/modifier
> combinations supported for display on this plane. Read and parse this
> format, storing the data in each plane, so we can know which
> combinati
On Mon, 9 Jul 2018 14:23:10 +0100
Daniel Stone wrote:
> Add support for the GBM_BO_IMPORT_FD_MODIFIER path, which allows us to
> import multi-plane dmabufs, as well as format modifiers.
>
> Signed-off-by: Daniel Stone
> Tested-by: Emre Ucan
> ---
> configure.ac | 6 +-
> libw
On Mon, 9 Jul 2018 14:23:09 +0100
Daniel Stone wrote:
> When creating a drm_fb from client (wl_buffer/dmabuf), gbm_surface, or
> client buffers, set fb->size to 0. The size member is only used for dumb
> buffers, where we mmap the whole buffer, and need the size recorded to
> later pass to munma
On Mon, 9 Jul 2018 14:23:08 +0100
Daniel Stone wrote:
> ARRAY_LENGTH returns a size_t; rather than casting its result to
> int so we can compare to our signed index variable, just declare the
> index as a compatible type in the first place.
>
> Signed-off-by: Daniel Stone
> ---
> libweston/co
On Mon, 9 Jul 2018 14:23:07 +0100
Daniel Stone wrote:
> Import the trivial definition of BUILD_BUG_ON() from the Linux kernel.
> This evaluates an expression such that it will call sizeof() on a
> negative-length array if the expression is false, generating a compiler
> error.
>
> This will be
Move drm_assign_planes into two functions: one which proposes a plane
configuration, and another which applies that state to the Weston
internal structures. This will be used to try multiple configurations
and see which is supported.
Signed-off-by: Daniel Stone
Reviewed-by: Pekka Paalanen
Tested
When trying to assign planes, keep track of the areas which are
already occluded, and ignore views which are completely occluded. This
allows us to build a state using planes only, when there are occluded
views which cannot go into a plane behind views which can.
Signed-off-by: Daniel Stone
Teste
The atomic API can allow us to test state before we apply it, to see if
it will be valid. Add support for this, which we will later use in
assign_planes to ensure our plane configuration is valid.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
libweston/compositor-drm.c | 176
Now that we can sensibly test proposed plane configurations with atomic,
sprites are not broken.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
libweston/compositor-drm.c | 22 +++---
1 file changed, 11 insertions(+), 11 deletions(-)
diff --git a/libweston/compositor-drm.
Return a pointer to the plane state, rather than indirecting via a
weston_plane.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
libweston/compositor-drm.c | 71 +-
1 file changed, 39 insertions(+), 32 deletions(-)
diff --git a/libweston/compositor-drm.
When creating a drm_fb from client (wl_buffer/dmabuf), gbm_surface, or
client buffers, set fb->size to 0. The size member is only used for dumb
buffers, where we mmap the whole buffer, and need the size recorded to
later pass to munmap.
Determining the full size of multi-planar buffers is difficul
Add support for multiple modes: toggling whether or not the renderer
and/or planes are acceptable. This will be used to implement a smarter
plane-placement heuristic when we have support for testing output
states.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
libweston/compositor-drm.c |
Add support for the GBM_BO_IMPORT_FD_MODIFIER path, which allows us to
import multi-plane dmabufs, as well as format modifiers.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
configure.ac | 6 +-
libweston/compositor-drm.c | 201 +
2 fil
Now that we collect information about which modifiers are supported for
KMS display, and are able to create KMS framebuffers with modifiers,
begin using the modifier-aware GBM API.
Client buffers from dmabuf already store multi-plane and modifier
information into drm_fb. Extend this to drm_fb_get_
Since we now incrementally test atomic state as we build it, we can
loosen restrictions on what we can do with planes, and let the kernel
tell us whether or not it's OK.
Signed-off-by: Daniel Stone
Tested-by: Emre Ucan
---
libweston/compositor-drm.c | 21 +
1 file changed, 1
From: Sergi Granell
The per-plane IN_FORMATS KMS property describes the format/modifier
combinations supported for display on this plane. Read and parse this
format, storing the data in each plane, so we can know which
combinations might work, and which combinations definitely will not
work.
Sim
When we come to assign_planes, try very hard to ignore views which are
only visible on other outputs, rather than forcibly moving them to the
primary plane, which causes damage all round and unnecessary repaints.
Signed-off-by: Daniel Stone
Reviewed-by: Pekka Paalanen
Tested-by: Emre Ucan
---
Hi,
Another re-send after Pekka's comments on the modifiers patches. Patches
01-12 from the last series were merged after the last round.
Patch 1 is new; patches 2 and 3 were extracted from patches 5 and 6.
Patches 4-6 have otherwise seen changes after Pekka's review; all the
comments were addres
ARRAY_LENGTH returns a size_t; rather than casting its result to
int so we can compare to our signed index variable, just declare the
index as a compatible type in the first place.
Signed-off-by: Daniel Stone
---
libweston/compositor-drm.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-
Import the trivial definition of BUILD_BUG_ON() from the Linux kernel.
This evaluates an expression such that it will call sizeof() on a
negative-length array if the expression is false, generating a compiler
error.
This will be used in future patches to ensure two array lengths don't go
out of sy
On Mon, 9 Jul 2018 13:42:45 +0100
Daniel Stone wrote:
> Hi,
> On Mon, 9 Jul 2018 at 11:08, Pekka Paalanen wrote:
> > On Thu, 5 Jul 2018 18:16:41 +0100 Daniel Stone
> > wrote:
> > > The per-plane IN_FORMATS property adds information about format
> > > modifiers; we can use these to both feed
On Mon, 9 Jul 2018 13:25:50 +0100
Daniel Stone wrote:
> Hi Pekka,
>
> On Mon, 9 Jul 2018 at 11:22, Pekka Paalanen wrote:
> > On Mon, 9 Jul 2018 12:39:30 +0300 Pekka Paalanen
> > wrote:
> > > On Thu, 5 Jul 2018 18:16:40 +0100 Daniel Stone
> > > wrote:
> > > > +static struct drm_fb *
> >
Hi,
On Mon, 9 Jul 2018 at 11:08, Pekka Paalanen wrote:
> On Thu, 5 Jul 2018 18:16:41 +0100 Daniel Stone wrote:
> > The per-plane IN_FORMATS property adds information about format
> > modifiers; we can use these to both feed GBM with the set of modifiers
> > we want to use for rendering, and also
On Fri, 6 Jul 2018 11:36:49 +0100
Daniel Stone wrote:
> The DPMS connector property is an enum property in KMS, which made our
> property handling complain at startup as we weren't defining its enums.
> Fix our definition so we parse the enum values.
>
> The only user of the property is the leg
Hi Pekka,
On Mon, 9 Jul 2018 at 11:22, Pekka Paalanen wrote:
> On Mon, 9 Jul 2018 12:39:30 +0300 Pekka Paalanen wrote:
> > On Thu, 5 Jul 2018 18:16:40 +0100 Daniel Stone
> > wrote:
> > > +static struct drm_fb *
> > > +drm_fb_get_from_dmabuf(struct linux_dmabuf_buffer *dmabuf,
> > > +
Hi Pekka,
On Mon, 9 Jul 2018 at 10:39, Pekka Paalanen wrote:
> On Thu, 5 Jul 2018 18:16:40 +0100
> Daniel Stone wrote:
> > +/* XXX: TODO:
> > + *
> > + * Currently the buffer is rejected if any dmabuf attribute
> > + * flag is set. This keeps us from passing an i
On Fri, 6 Jul 2018 16:18:39 +0530
"Nautiyal, Ankit K" wrote:
> From: Ankit Nautiyal
>
> The weston.ini.man describes the mode-formats that a user can specify
> for selecting a video mode. The DRM specific examples are already
> provided in weston-drm.man, so this inofrmation is redundant and c
On Thu, 5 Jul 2018 18:16:42 +0100
Daniel Stone wrote:
> Use the extended GBM allocator interface to support modifiers and
> multi-planar BOs.
>
> Signed-off-by: Daniel Stone
> Tested-by: Emre Ucan
> ---
> configure.ac | 3 ++
> libweston/compositor-drm.c | 61 +
On Mon, 9 Jul 2018 12:39:30 +0300
Pekka Paalanen wrote:
> On Thu, 5 Jul 2018 18:16:40 +0100
> Daniel Stone wrote:
>
> > Add support for the GBM_BO_IMPORT_FD_MODIFIER path, which allows us to
> > import multi-plane dmabufs, as well as format modifiers.
> >
> > Signed-off-by: Daniel Stone
> >
On Thu, 5 Jul 2018 18:16:41 +0100
Daniel Stone wrote:
> From: Sergi Granell
>
> The per-plane IN_FORMATS property adds information about format
> modifiers; we can use these to both feed GBM with the set of modifiers
> we want to use for rendering, and also as an early-out test when we're
> tr
On Thu, 5 Jul 2018 18:16:40 +0100
Daniel Stone wrote:
> Add support for the GBM_BO_IMPORT_FD_MODIFIER path, which allows us to
> import multi-plane dmabufs, as well as format modifiers.
>
> Signed-off-by: Daniel Stone
> Tested-by: Emre Ucan
> ---
> configure.ac | 6 +-
> libw
Bump: this is the source of bugs in xwayland [1].
[1]: https://lists.x.org/archives/xorg-devel/2018-July/057285.html
On May 18, 2018 9:40 PM, Simon Ser wrote:
> Signed-off-by: Simon Ser
> ---
>
> I believe it's easy to understand that transformations are not applied to
> the logical size.
>
>
42 matches
Mail list logo
