Sergey Bugaev
85830c059e
Plug GTask leaks
...
The error-prone pattern seems to be:
GTask *task = g_task_new (...);
if (condition)
{
g_task_return_... (task, ...);
/* need g_object_unref (task) here! */
return;
}
Signed-off-by: Sergey Bugaev <bugaevc@gmail.com>
2024-08-05 14:38:23 +03:00
Matthias Clasen
1c92f2d963
Merge branch 'matthiasc/for-main' into 'main'
...
Disable blur-performance build
See merge request GNOME/gtk!7549
2024-08-04 02:52:53 +00:00
Matthias Clasen
12e0ec0b48
Add gtk_css_parser_has_percentage
2024-08-03 22:31:32 -04:00
Matthias Clasen
5574bf59f6
Disable blur-performance build
...
This is an old test that isn't very relevant anymore, and it has
some linking problems because it includes private headers that
have inlined functions.
2024-08-03 22:31:21 -04:00
Matthias Clasen
7ee27cc0ad
Merge branch 'matthiasc/for-main' into 'main'
...
gpu: Rename function
See merge request GNOME/gtk!7548
2024-08-04 01:23:41 +00:00
Benjamin Otte
126d689086
gpu: Rename function
...
We want to reuse gsk_gpu_color_to_float() for use with GdkColor and this
function will be replaced. But until that's fully done, we need 2
different names.
So rename this one to something else
2024-08-03 20:59:40 -04:00
Matthias Clasen
3fca84f974
Merge branch 'matthiasc/for-main' into 'main'
...
Fix a copy-paste error
See merge request GNOME/gtk!7543
2024-08-02 13:45:40 +00:00
Matthias Clasen
2be48a8f86
Pass ccs to the cairo blur code
...
This is needed to set the color properly.
2024-08-02 09:13:41 -04:00
Matthias Clasen
389d246af8
Fix up the blur-performance test
...
Add a hack to work aroundt he fact that the blur performance test
includes private headers which inline accesses to non-exported
variables.
2024-08-02 09:12:37 -04:00
Matthias Clasen
9e0f41bd05
Fix a copy-paste error
...
We were using the wrong api to get the color of an outset shadow
node. This just worked by luck.
2024-08-02 09:00:51 -04:00
Matthias Clasen
564a237454
Merge branch 'matthiasc/for-main' into 'main'
...
Cosmetics: Remove a leftover line
See merge request GNOME/gtk!7540
2024-08-01 22:24:55 +00:00
Matthias Clasen
86b59ba643
Cosmetics: Remove a leftover line
...
This looks just like a copy-paste leftover. Suprisingly, it compiled
just fine and worked too.
2024-08-01 17:19:50 -04:00
Matthias Clasen
487e568722
Merge branch 'matthiasc/for-main' into 'main'
...
vulkan: Don't try srgb for the atlas
Closes #6887
See merge request GNOME/gtk!7539
2024-08-01 19:40:00 +00:00
Matthias Clasen
67080dab21
vulkan: Don't try srgb for the atlas
...
This makes the Emoji rendering come out wrong.
Fixes : #6887
2024-08-01 14:43:49 -04:00
Benjamin Otte
0809e6585d
Merge branch 'wip/otte/for-main' into 'main'
...
vulkan: Add error checking in a place
See merge request GNOME/gtk!7538
2024-08-01 11:14:12 +00:00
Benjamin Otte
809b98c96e
vulkan: Don't leak fds
...
It makes sense to close the fds before doing an early return.
D'oh.
2024-08-01 12:51:46 +02:00
Benjamin Otte
059af1365a
vulkan: Add error checking in a place
...
Doesn't hurt to have, but I just added it while debugging something
unrelated.
2024-08-01 12:51:46 +02:00
Matthias Clasen
fbcffb36c8
Merge branch 'matthiasc/for-main' into 'main'
...
Drop debug code
See merge request GNOME/gtk!7536
2024-08-01 01:31:16 +00:00
Matthias Clasen
3af0cfb53b
Drop debug code
...
The dmabuf download code for P010 had a hack left in it that
was preventing it from working.
2024-07-31 21:08:14 -04:00
Matthias Clasen
96422fcf52
Merge branch 'no-cursor-viewport' into 'main'
...
wayland: No viewporter for cursors
See merge request GNOME/gtk!7531
2024-07-31 20:23:34 +00:00
Matthias Clasen
2754f2be2d
Merge branch 'test-cal-31' into 'main'
...
ci: avoid another date-dependent failure
See merge request GNOME/gtk!7535
2024-07-31 20:12:16 +00:00
Matthias Clasen
3d802177be
wayland: No viewporter for cursors
...
The outlook for mutter supporting this in GNOME 47 are cloudy,
so lets flip the switch back. You can still set
USE_POINTER_VIEWPORT in the environment to try this code.
2024-07-31 07:44:48 -04:00
Matthias Clasen
a02d4ba2a1
Merge branch 'matthiasc/for-main' into 'main'
...
dmabuf: Fix the disjointness check
See merge request GNOME/gtk!7529
2024-07-31 05:04:02 +00:00
Matthias Clasen
249cf0833c
dmabuf: Fix the disjointness check
...
We were checking the wrong fds here.
This came up while toying with udmabuf and creating multi-plane
NV12 buffers.
2024-07-30 23:44:07 -04:00
Jeremy Bícha
20b00d3947
ci: avoid another date-dependent failure
...
Set the day before setting the month to avoid
failure when running the test on the 31st day of a month
2024-07-30 15:50:12 -04:00
Matthias Clasen
269dcaaadf
Post-release version bump
2024-07-30 15:14:25 -04:00
Matthias Clasen
322bbd2376
4.15.4
2024-07-30 12:50:22 -04:00
Benjamin Otte
52bec248ce
Merge branch 'wip/otte/for-main' into 'main'
...
gpu: Fix shadows even more
See merge request GNOME/gtk!7527
2024-07-30 16:29:59 +00:00
Benjamin Otte
9f71528a05
gpu: Fix shadows even more
...
Math is hard.
But this time, it comes with a test!
2024-07-30 18:01:45 +02:00
Benjamin Otte
65e02a3bae
Merge branch 'wip/otte/for-main' into 'main'
...
gpu: Fix shadows some more
Closes #6896
See merge request GNOME/gtk!7526
2024-07-30 14:48:34 +00:00
Benjamin Otte
defc4d335e
gpu: Fix shadows some more
...
It turns out the "step" variable could up as 0 when p.y ~= 3.0 ||
p.y ~= r.y - 3.0
That was not enough to trigger it though because if "start" and "end"
were the same value, the "y <= end" check in the loop would immediately
terminate it.
However, if start + epsilon == end so that end != start but (end - start)
/ 7 == 0, then step would end up as 0 and the loop would never
terminate.
And if that happened, it would bring down GPUs.
So recode this whole machinery to make it impossible to infloop.
Fixes #6896
2024-07-30 16:09:38 +02:00
Benjamin Otte
bdcfcfa5b9
Merge branch 'wip/otte/for-main' into 'main'
...
gpu: Draw proper shadows again
Closes #6888
See merge request GNOME/gtk!7518
2024-07-30 00:19:37 +00:00
Benjamin Otte
4a94c91772
gpu: Don't blur tiny blur radii
...
We get those wrong, and there's not really a visual effect.
And since we do a check anyway, just disallow them and treat them as
unblurred.
2024-07-30 01:50:56 +02:00
Benjamin Otte
cc3ed89e34
gpu: Draw proper shadows again
...
The fix in commit 5e7f227d
broke shadows while trying to make them
faster.
So use a better way to make them faster.
With the normalized blur radius, we can now conclude that all the values
too far from p.y will cause the gauss() call to return close to 0, so we
can skip any y value that is too far from p.y.
And that allows us to put an upper limit on the loop iterations.
Tests included
Fixes #6888
2024-07-30 01:50:56 +02:00
Benjamin Otte
87e8265a2d
Merge branch 'wip/otte/colomplications' into 'main'
...
wayland: Do not use compositor-provided image descriptions
See merge request GNOME/gtk!7522
2024-07-29 17:56:41 +00:00
Matthias Clasen
9fd6d431e3
Merge branch 'fix-ephy-crash' into 'main'
...
subsurface: Avoid a crash
Closes #6891
See merge request GNOME/gtk!7521
2024-07-29 17:39:05 +00:00
Benjamin Otte
5059ae1d7b
gpu: Simplify box-shadow shader
...
Instead of doing complicated math, normalize the values to a sigma
of 1.0, and then use that.
This should also be beneficial for shader performance, because 1.0 is a
constant and constant-elimination can kick in on the inlined functions.
2024-07-29 19:14:10 +02:00
Benjamin Otte
50bb42f0f3
gl: Don't crash on box shadows with negative scales
...
Some places didn't fabs() the scale values properly, leading to
assertions when offscreens with negative sizes were created.
2024-07-29 19:14:10 +02:00
Benjamin Otte
d71a7c901d
wayland: Do not use compositor-provided image descriptions
...
When the compositor sends us an image description, we currently happily
reuse it.
However, those image descriptions may contain optional properties that
we do not handle - example: reference white level. So if we were to
reuse that image description, we would set a wrong reference white
level.
To avoid issues like that, never use compositor-provided image
descriptions.
However, query those image descriptions and map them to the closest
GdkColorState, so that we can quickly look up *our* version of that
image description and use that one.
2024-07-29 18:45:41 +02:00
Matthias Clasen
65f727731c
subsurface: Avoid a crash
...
When finalizing a subsurface, we need to make sure it is removed
from the sibling lists in its parent, or bad things will happen.
This should crashes seen in Epiphany nightly.
Fixes : #6891
2024-07-29 11:54:10 -04:00
Matthias Clasen
ff60b19e4b
Merge branch 'matthiasc/for-main' into 'main'
...
image-tool: Add an --undecorated option
See merge request GNOME/gtk!7520
2024-07-29 15:52:42 +00:00
Matthias Clasen
b8f12c5d28
cicp: Improve the docs
...
Fix some formatting issues, and document GdkCicpRange.
2024-07-29 11:17:13 -04:00
Matthias Clasen
bfb779ac2f
gsk: Add missing cicp transfer functions
...
This was overlooked in 50ea9450ea
.
2024-07-29 10:35:32 -04:00
Matthias Clasen
84c28d2efb
image-tool: Add an --undecorated option
...
This is useful if you want to debug things and don't want to end
up in the decoration rendering code paths.
2024-07-29 08:13:21 -04:00
Matthias Clasen
06372c5f2a
Merge branch 'matthiasc/for-main' into 'main'
...
Fix indirect color state conversions
See merge request GNOME/gtk!7519
2024-07-29 11:11:28 +00:00
Matthias Clasen
3d75292c4d
Fix indirect color state conversions
...
convert_func2 is a 'from' conversion function, ie it expects to
be passed the target color state. This was wrong both in
gdk_memory_convert and gdk_memory_convert_color_state.
2024-07-28 23:40:59 -04:00
Matthias Clasen
541ffbe4b5
Merge branch 'matthiasc/for-main' into 'main'
...
gst: Fix a wrong unref function
See merge request GNOME/gtk!7517
2024-07-28 23:35:16 +00:00
Matthias Clasen
17524f2794
gst: Fix a wrong unref function
...
Use gdk_color_state_unref on a color state.
2024-07-28 19:09:05 -04:00
Matthias Clasen
1040d7c7e4
Merge branch 'matthiasc/for-main' into 'main'
...
cicp: Fill in some more transfer functions
See merge request GNOME/gtk!7515
2024-07-28 20:51:40 +00:00
Matthias Clasen
1b53359cd3
Merge branch 'wip/smcv/ubsan' into 'main'
...
Avoid undefined behaviour in a few simple cases
See merge request GNOME/gtk!7508
2024-07-28 20:00:06 +00:00