Hi everybody,
these check some corner cases in how GL_TEXTURE_MAX_LEVEL is initialized
for immutable textures. Mesa does not do this yet, which among other things
leads to a user-visible bug in mipmap generation as exposed in
https://bugs.freedesktop.org/show_bug.cgi?id=91993 .
Please review. I
This test is a bit pedantic and the required behavior is perhaps
surprising, but it's what the spec says and it's compatible with
the clamping that Mesa does when you try to explicitly set a large
MAX_LEVEL via glTexParameteri.
---
tests/spec/arb_texture_view/max-level.c | 27
and associated documentation files (the "Software"),
@@ -26,6 +27,9 @@
* Tests GL_ARB_texture_storage
* Note: only the glTexStorage2D() function is tested with actual rendering.
* Brian Paul
+ *
+ * Additional author(s):
+ * Nicolai Hähnle <nicolai.haeh...@amd.com>
*/
#include &
person obtaining a
* copy of this software and associated documentation files (the "Software"),
@@ -26,6 +27,9 @@
* Tests GL_ARB_texture_storage
* Note: only the glTexStorage2D() function is tested with actual rendering.
* Brian Paul
+ *
+ * Additional author(s):
+ * Nico
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Since no-op is a valid implementation, this only checks the error conditions.
Mesa currently fails this, I have a patch in the pipeline.
Signed-off-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Signed-off-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
I lost track of this one, it's quite old already. I'll push it out soon.
tests/spec/arb_texture_storage/texture-storage.c | 88
1 file changed, 8
Based on the test in https://bugs.freedesktop.org/show_bug.cgi?id=95130
which exposed an error in radeonsi non-monolithic fragment shaders.
---
.../execution/fs-fwidth-color.shader_test | 44 ++
1 file changed, 44 insertions(+)
create mode 100644
Hi Alejandro,
my understanding of the spec is that the test should be correct as
originally written, i.e. with glTextureBarrier only between full
rectangle draws.
FWIW, radeonsi passes the test as-is (though I reduced the GLSL version
requirement, which I'd ask you to change as well before
On 24.06.2016 16:42, Alejandro Piñeiro wrote:
On 24/06/16 12:17, Nicolai Hähnle wrote:
Hi Alejandro,
Hi, thanks for the quick answer, more comments below.
my understanding of the spec is that the test should be correct as
originally written, i.e. with glTextureBarrier only between full
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test polygon stipple and its interaction with a simple fragment program.
Exposes a bug that lived in Gallium's stipple emulation for a very long
time until it was hit by Kodi (reported by Christian König).
---
tests/
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This is similar to the texsubimage test, but checks various type and format
conversions as well as a different set of pixel unpack settings.
---
tests/all.py | 4 +-
tests/texturing/CMakeLists.gl.txt| 1 +
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/all.py | 3 ++
tests/spec/ext_texture_array/compressed.c | 51 +--
2 files changed, 44 insertions(+), 10 deletions(-)
diff --git a/tests/all.py b/tests/all.py
index 5
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/spec/gl-2.1/pbo.c | 46 +-
1 file changed, 41 insertions(+), 5 deletions(-)
diff --git a/tests/spec/gl-2.1/pbo.c b/tests/spec/gl-2.1/pbo.c
index 7149ab5..bb4c035 100644
--- a/tests/spec/
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/all.py | 1 +
tests/spec/arb_texture_view/texsubimage-levels.c | 31 ++--
2 files changed, 30 insertions(+), 2 deletions(-)
diff --git a/tests/all.py b/tests/all.py
index a
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Also test the GL_RGBA internal format because GL_RGB has less hardware
support, so GL_RGBA is likelier to actually hit the fast path.
---
tests/all.py | 2 ++
tests/texturing/cubemap.
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/all.py | 1 +
tests/spec/arb_texture_view/texsubimage-layers.c | 31 ++--
2 files changed, 30 insertions(+), 2 deletions(-)
diff --git a/tests/all.py b/tests/all.py
index 0
On 19.01.2016 15:25, Ilia Mirkin wrote:
On Tue, Jan 19, 2016 at 3:22 PM, Ilia Mirkin <imir...@alum.mit.edu> wrote:
On Tue, Jan 19, 2016 at 3:20 PM, Nicolai Hähnle <nhaeh...@gmail.com> wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test a pure depth write with c
On 20.01.2016 20:43, Ian Romanick wrote:
On 01/18/2016 01:40 PM, Nicolai Hähnle wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This is similar to the texsubimage test, but checks various type and format
conversions as well as a different set of pixel unpack settings.
I haven't
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This commit adds an optional 'depthbuffer' line to the [require]
section as well as the possibility of enabling/disabling GL_DEPTH_TEST
and probing depth values.
---
tests/shaders/shader_runner.c | 21 -
1 file chang
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test a pure depth write with conditional discard in the fragment shader.
This currently fails in radeonsi.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93761
---
tests/shaders/glsl-fs-discard-only.shader_tes
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test a pure depth write with conditional discard in the fragment shader.
This currently fails in radeonsi.
v2: use vertex shader passthrough and move to tests/spec/glsl-1.10 (Ilia Mirkin)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This is a slightly extended variant of equal_images from texsubimage
which will also be used in the new texsubimage-unpack test.
I've decided to call it "equal" instead of "compare", since I feel that
makes the meanin
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
The idea is that tests can get an idea of how much precision to expect when
using these formats.
v2: clarify that min_bits is for the depth component (Ian Romanick)
---
tests/fbo/fbo-formats.h
Hi,
this is a restructuring of the first part of the PBO coverage series I sent
recently, based on Ian's feedback.
There's a bit more code sharing, the run_concurrent business is cleared up,
and bunch of other cleanups.
Cheers,
Nicolai
--
tests/all.py | 4 +-
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
All other variants (array, cube_map-array, and pbo variants) are already
running concurrently
---
tests/all.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/all.py b/tests/all.py
index bb174cb..28aba13 100644
--- a
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/util/piglit-util-gl.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/tests/util/piglit-util-gl.c b/tests/util/piglit-util-gl.c
index bbc174d..61764fb 100644
--- a/tests/util/piglit-util-gl.c
+++ b/tests/util/piglit-uti
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This is similar to the texsubimage test, but checks various type and format
conversions as well as a different set of pixel unpack settings.
v2: cleanups and share some code with texsubimage
---
tests/all.py | 2 +
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/shaders/shader_runner.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tests/shaders/shader_runner.c b/tests/shaders/shader_runner.c
index 206258c..2d127e2 100644
--- a/tests/shaders/shader_runner.c
+++ b/tests/s
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/shaders/shader_runner.c | 17 +
tests/util/piglit-util-gl.c | 35 +++
tests/util/piglit-util-gl.h | 1 +
3 files changed, 53 insertions(+)
diff --git a/tests/shaders/shader_runn
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This exercises a required DCC decompression path for Radeon VI.
---
.../execution/write-to-rendered-image.shader_test | 73 ++
1 file changed, 73 insertions(+)
create mode 100644
tests/spec/arb_shader_image_load
Hi,
this is a cleanup of a bunch of tests that I used during development of
those extensions, including a number of shader_runner extensions I find
useful.
I ended up changing the tests so that 'depthfunc' currently isn't used
anymore, but it might be useful in the future, so...
Cheers,
Nicolai
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
.../execution/indirect.shader_test | 60 ++
1 file changed, 60 insertions(+)
create mode 100644
tests/spec/arb_shader_storage_buffer_object/execution/indirect.shader_test
diff --git
a/test
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Core profile 4.20 does not have the gl_FragColor built-in, and atomic
counters need a binding.
---
tests/spec/arb_arrays_of_arrays/compiler/glsl-4.20-basic-types.frag | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git
a
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
... so just use an explicit out declaration unconditionally for such versions.
---
tests/shaders/version-mixing.c | 25 -
1 file changed, 24 insertions(+), 1 deletion(-)
diff --git a/tests/shaders/version-mixing.c b
From a quick skim, the tests look good to me.
However, for the synchronization/barrier one, I'd appreciate if you'd
add a "sanity check" mode similar to the barrier tests for shader
images. That is, a mode in which the barriers *aren't* emitted, so that
one can easily see whether/how the
Thanks!
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
On 18.03.2016 10:58, Ilia Mirkin wrote:
---
tests/shaders/shader_runner.c | 43 +++
1 file changed, 43 insertions(+)
diff --git a/tests/shaders/shader_runner.c b/tests/s
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This exhibits an MSAA-specific bug in the radeonsi driver with fast stencil
clear and decompress.
The test can optionally disable MSAA and change the number of samples, though
I've only added the default variant to the all run - that's
On 25.04.2016 13:08, Matt Turner wrote:
On Wed, Apr 6, 2016 at 6:10 PM, Nicolai Hähnle <nhaeh...@gmail.com> wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/shaders/shader_runner.c | 17 +
tests/util/piglit-util
Hi Chad,
I saw Matt's email and went ahead and just did this before seeing your
email. Sorry about that!
Nicolai
On 25.04.2016 16:49, Chad Versace wrote:
The function always returned texture name 0. Fix to to return the
actual texture name.
Found by GCC warning:
On 14.04.2016 23:43, Francisco Jerez wrote:
Nicolai Hähnle <nicolai.haeh...@amd.com> writes:
On 14.04.2016 22:37, Francisco Jerez wrote:
This test seems bogus, ARB_shader_image_load_store doesn't give you any
ordering or concurrency guarantees for independent shader invocation
On 15.04.2016 17:12, Francisco Jerez wrote:
For a test doing almost the same thing but not relying on unspecified
invocation ordering, see
"tests/spec/arb_shader_image_load_store/shader-mem-barrier.c" -- It
would be interesting to see whether you can get it to reproduce the GCN
coherency bug
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
By transposing the loop and the writer/reader branching, the effective level
of parallelism in the test is increased for a typical wave-based architecture.
Increasing the window size on top of that helps to reliably expose different
implemen
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Extracted from a bug in the Mesa GLSL compiler that was triggered by
Firewatch. I will also send out a patch for Mesa.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96762
---
.../execution/fs-discard-deep-branch.shader_test
icolai Hähnle wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Currently fails due to an error in Gallium st_glsl_to_tgsi. Observed
in the
game Overlord.
---
I'm also going to send out a candidate fix to mesa-dev in a moment.
.../execution/vsfs-unused-array-member.shad
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Currently fails due to an error in Gallium st_glsl_to_tgsi. Observed in the
game Overlord.
---
I'm also going to send out a candidate fix to mesa-dev in a moment.
.../execution/vsfs-unused-array-member.shader_test | 41 +++
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
I've found this to be rather useful with longer shader_test scripts.
---
tests/shaders/shader_runner.c | 22 +++---
1 file changed, 19 insertions(+), 3 deletions(-)
diff --git a/tests/shaders/shader_runner.c b/tests/s
From: "Yang, Kefeng"
With OpenGL 4.5, the spec was changed to remove primitive restart from
glDrawArrays and friends.
---
tests/general/primitive-restart.c | 12 +++-
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/tests/general/primitive-restart.c
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test polygon stipple and its interaction with a simple fragment shader.
Exposes a bug that lived in Gallium's stipple emulation for a very long
time until it was hit by Kodi (reported by Christian König).
---
I've had this on my disk for quit
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Compared to the other array bounds tests, this one exercises a larger array
size.
The motivation is for Radeon, where arrays can be lowered either into the
register file or into "real" memory. The larger array in this test will be
On 11.08.2016 07:56, Tapani Pälli wrote:
Hi;
On 08/10/2016 09:34 PM, Nicolai Hähnle wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Compared to the other array bounds tests, this one exercises a larger
array
size.
The motivation is for Radeon, where arrays can be lowered
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
On 01.07.2016 19:05, Alejandro Piñeiro wrote:
This patch adapts the test to the last change on mesa. Copying
the rationale on that patch here:
For the case (both src or dst) where we had a texobject, but the
texobject
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This tests that indirect data is picked up correctly when it's produced by a
transform feedback operation.
---
tests/all.py | 1 +
tests/spec/arb_draw_indirect/CMakeLists.gl.txt| 1 +
test
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Test polygon stipple and its interaction with a simple fragment shader.
Exposes a bug that lived in Gallium's stipple emulation for a very long
time until it was hit by Kodi (reported by Christian König).
v2:
- require only OpenGL 2.1 (Bria
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
On 05.08.2016 01:14, Kenneth Graunke wrote:
This reverts commit 4e549ddb500cf677b6fa16d9ebdfa67cc23da097,
dropping the hack from Gallium that I just deleted from i965.
See the previous commit for rationale.
---
src/mesa/state_t
I'm glad your sanity is still hanging on :) I particularly like the "Yes
& No" in the comparison table.
Anyway, this patch and the piglit patches are
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
On 05.08.2016 01:14, Kenneth Graunke wrote:
OpenGL 4.4 specifies
correctly) to check that the only
change is an addition of new generated tests. All the previously
generated tests are unchanged (in file names and contents).
Cheers,
Nicolai
Dylan
Quoting Nicolai Hähnle (2017-01-26 10:59:21)
Hi all,
increase the test coverage of GL_ARB_gpu_shader_int64
On 31.01.2017 19:23, Eric Anholt wrote:
Nicolai Hähnle <nhaeh...@gmail.com> writes:
Hi all,
increase the test coverage of GL_ARB_gpu_shader_int64 by testing various
conversions. To avoid reinventing the wheel or duplicating code, I've
extended the existing fp64 conversion test gene
On 01.02.2017 19:30, Dylan Baker wrote:
Can you modify 'unittests/generators/test_generators.py' to know about the
rename, and run the tests? 'pytest unittest/generators' or "tox -e
'py{27,35}-generators'" will run the tests, depending on what you have
installed.
FWIW, there are some warnings
On 05.02.2017 19:12, Ilia Mirkin wrote:
Signed-off-by: Ilia Mirkin <imir...@alum.mit.edu>
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
generated_tests/builtin_function.py | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/gen
On 04.02.2017 22:10, Ilia Mirkin wrote:
Signed-off-by: Ilia Mirkin <imir...@alum.mit.edu>
Looks reasonable.
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
generated_tests/builtin_function.py | 24 +++-
1 file changed, 19 insertions(+), 5 deleti
On 05.02.2017 20:28, Marek Olšák wrote:
From: Marek Olšák <marek.ol...@amd.com>
this happened to me once
Thanks, I vaguely remember seeing something like this as well. Both patches:
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
... though you might want to wait
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/all.py | 5 +++
tests/spec/CMakeLists.txt | 1 +
tests/spec/arb_sparse_buffer/CMakeLists.gl.txt | 12 ++
tests/spec/arb_sparse_buffer/CMakeLists.txt| 1 +
test
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
tests/all.py | 1 +
tests/spec/arb_sparse_buffer/CMakeLists.gl.txt | 1 +
tests/spec/arb_sparse_buffer/basic.c | 190 +
3 files changed, 192 insertions(+)
creat
The GLCTS has tests for the extension as well, but perhaps other people also
find simple tests like this one easier to work with...
Cheers,
Nicolai
___
Piglit mailing list
Piglit@lists.freedesktop.org
On 26.01.2017 20:27, Dylan Baker wrote:
Quoting Nicolai Hähnle (2017-01-26 10:59:23)
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Currently only 'd' for double, but will be extended for 64-bit integer
types.
---
generated_tests/gen_conversion.py | 21 +
1 file c
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
We will extend it with int64 types.
---
generated_tests/CMakeLists.txt | 44 +-
generated_tests/gen_conversion.py | 605 +
generated_tests/gen_conversion_fp64.py
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Use generic names to reflect the upcoming dual use for 64-bit int types.
---
generated_tests/gen_conversion.py | 44 +++
1 file changed, 22 insertions(+), 22 deletions(-)
diff --git a/generated
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
generated_tests/gen_conversion.py | 302 ++---
generated_tests/templates/gen_conversion/base.mako | 2 +-
2 files changed, 264 insertions(+), 40 deletions(-)
diff --git a/generated_tests/gen_conversion
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Currently only 'd' for double, but will be extended for 64-bit integer
types.
---
generated_tests/gen_conversion.py | 21 +
1 file changed, 13 insertions(+), 8 deletions(-)
diff --git a/generated_tests/gen_conversion
Hi all,
increase the test coverage of GL_ARB_gpu_shader_int64 by testing various
conversions. To avoid reinventing the wheel or duplicating code, I've
extended the existing fp64 conversion test generation. The first couple
of patches simply move code around and rename things. The last patch adds
I don't see anything in the GLX spec about GLXPixmaps being destroyed
automatically. Without this patch, glx-visuals-depth/stencil -pixmap fail
with an X protocol error after some number of FBConfigs.
---
tests/util/piglit-glx-util.c | 17 +
1 file changed, 17 insertions(+)
diff
Thanks for doing this.
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
On 17.02.2017 13:08, Marek Olšák wrote:
From: Marek Olšák <marek.ol...@amd.com>
this was broken in radeonsi
---
tests/general/draw-elements.c | 34 +++---
1 file changed, 3
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Found by ASAN.
---
tests/spec/glsl-1.50/execution/get-active-attrib-array.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/spec/glsl-1.50/execution/get-active-attrib-array.c
b/tests/spec/glsl-1.50/execution/get-
ts/shaders/glsl-cache-fallback-shader-source.c
b/tests/shaders/glsl-cache-fallback-shader-source.c
new file mode 100644
index 000..f764617
--- /dev/null
+++ b/tests/shaders/glsl-cache-fallback-shader-source.c
@@ -0,0 +1,187 @@
+/*
+ * Copyright (c) 2009 Nicolai Hähnle
I suspect your copy
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This reduces the chance of false positives caused by system load.
---
tests/all.py | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/tests/all.py b/tests/all.py
index 3961656..6e07125 100644
--- a/tests/all.py
+++ b
Thanks for doing that, the numbers are impressive.
Two small comments on the C parts; it's been too long since I looked
into the Python parts to be able to properly comment.
Cheers,
Nicolai
On 09.09.2016 21:18, Dylan Baker wrote:
This allows shader_runner to consume more than one test at a
On 09.09.2016 21:18, Dylan Baker wrote:
Bug fixes from me for the above patch.
I've split these out to make them easier to review. Before merging to
master this will be squashed into the previous patch.
Changes:
- Fix the assertion of subuniform_locations
- Add teardown function for UBOs
On 09.09.2016 21:18, Dylan Baker wrote:
From: Marek Olšák
piglit_report_result prevents running multiple tests per process.
---
tests/shaders/shader_runner.c | 231 +--
1 file changed, 140 insertions(+), 91 deletions(-)
diff --git
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
A bunch of additions to the ARB_query_buffer_object test:
1. test different result types (including 64 bit types)
2. compare the written result with equality to the result obtained by
querying on the CPU
3. add an additional mode to catch
On 13.09.2016 15:38, Emil Velikov wrote:
On 12 September 2016 at 16:31, Nicolai Hähnle <nhaeh...@gmail.com> wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This reduces the chance of false positives caused by system load.
From my experience the {arb,ext}-timer-query tests
On 16.09.2016 10:44, Alejandro Piñeiro wrote:
On 15/09/16 18:02, Nicolai Hähnle wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
A bunch of additions to the ARB_query_buffer_object test:
1. test different result types (including 64 bit types)
2. compare the written result with eq
Hi Samuel,
Thank you for doing this!
There should also be a test that exercises a range of different work
group sizes with the same shader, up to the implementation maximums.
On 08.09.2016 21:00, Samuel Pitoiset wrote:
Signed-off-by: Samuel Pitoiset
---
er_shader / the OpenGL >= 4.3 spec applies. With the comment
updated accordingly, this patch is
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
+ prog = piglit_build_simple_program_multiple_shaders(
+ GL_COMPUTE_SHADER, variable_work_group_size_shader
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
.../vs-fs-array-interleave-range.shader_test | 90 ++
1 file changed, 90 insertions(+)
create mode 100644
tests/spec/arb_enhanced_layouts/execution/component-layout/vs-fs-array-interleave-range.shader_test
diff
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
.../component-layout/vs-fs-array-dvec3.shader_test | 69 ++
1 file changed, 69 insertions(+)
create mode 100644
tests/spec/arb_enhanced_layouts/execution/component-layout/vs-fs-array-dvec3.shader_test
diff --git
a
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This is a regression test for a very specific compiler bug in radeonsi.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97887
---
.../execution/fs-loop-cont-else-bug.shader_test| 44 ++
1 file changed, 44 inse
((x * y * z) > max_local_x)
+ break;
... and analogously here.
With those two minor comments addressed, the series is
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
No need to send a v3 as far as I'm concerned :)
Cheers,
Nicolai
+
On 10.09.2016 17:19, Samuel Pitoiset wrote:
This allows to dispatch compute shaders with a variable local size
using the new "compute group size" command.
Signed-off-by: Samuel Pitoiset
---
tests/shaders/shader_runner.c | 7 +++
1 file changed, 7 insertions(+)
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Exposes a subtle bug in radeonsi on VI when 3-component data is precisely
at the end of a buffer object.
---
tests/spec/gl-2.0/CMakeLists.gl.txt| 1 +
tests/spec/gl-2.0/vertexattribpointer-size-3.c | 268 ++
On 17.10.2016 22:33, Matt Turner wrote:
On Mon, Oct 17, 2016 at 12:54 PM, Samuel Pitoiset
wrote:
On 10/17/2016 09:45 PM, Samuel Pitoiset wrote:
Thanks for fixing this.
Reviewed-by: Samuel Pitoiset
Actually, we need to check for
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
Cc: Ian Romanick <i...@freedesktop.org>
---
.../fs-inout-index-two-level.shader_test | 49 ++
1 file changed, 49 insertions(+)
create mode 100644
tests/shaders/out-parameter-indexing/fs-ino
zes being a problem on
Windows. Did the exact window size matter for the driver bug? The rest
of the test looks like it should be window size agnostic. (The same for
the second test.)
Apart from that:
Reviewed-by: Nicolai Hähnle <nicolai.haeh...@amd.com>
+
Hi all,
I found the 'regressions' page of the HTML summary to be pretty useless
if you're comparing more than two runs, since it ends up being very
similar to 'changes'.
The first patch makes 'regressions' more useful for me when comparing many
chronologically sorted runs. The second patch just
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
The idea is that 'regressions' can be useful when a summary is generated
for a sequence of chronologically sorted runs of the same configuration.
In that case, what's interesting is the state of the driver in the latest
run. With this
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
---
framework/programs/summary.py | 5 +
framework/summary/console_.py | 4 +++-
2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/framework/programs/summary.py b/framework/programs/summary.py
index e400d9a..037bb56
, the same questions for the fixes page,
because I have no use for that at all :)
Nicolai
Now regressions is some weird other thing not part of that taxonomy.
I'd rather that invariant remain - if you change one, you change all.
On Mon, Oct 24, 2016 at 10:02 AM, Nicolai Hähnle <nhaeh...@
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This test exposes a subtle bug in st_glsl_to_tgsi. I found this bug while
investigating the referenced bug report.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98664
---
.../geometry/gs-conditional-emit.shader_test
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
This shows an error in UBO load lowering.
---
.../fs-ubo-named-block-explicit-offset.shader_test | 30 ++
1 file changed, 30 insertions(+)
create mode 100644
tests/spec/arb_enhanced_layouts/execution/fs-ubo-named-block-ex
On 31.10.2016 22:52, Nicolai Hähnle wrote:
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
The first two vertices are swapped for triangle strips with adjacency; see
Table 10.1 of the OpenGL 4.5 (Compatibility Profile) spec.
FWIW, people report that the test passes as-is on NVidia ha
From: Nicolai Hähnle <nicolai.haeh...@amd.com>
The first two vertices are swapped for triangle strips with adjacency; see
Table 10.1 of the OpenGL 4.5 (Compatibility Profile) spec.
---
tests/spec/gl-3.2/adj-prims.c | 18 +-
1 file changed, 13 insertions(+), 5 deletions(-)
1 - 100 of 204 matches
Mail list logo