Hi Veluri,
glad to see you're interested.
There have been several attempts at a DriConf replacement over the
years. I believe the most advanced one currently is by Jean Hertel at
https://github.com/jlHertel/adriconf -- so your best bet is probably to
check that out, see what if anything is st
https://bugs.freedesktop.org/show_bug.cgi?id=106833
Bug ID: 106833
Summary: glLinkProgram is expected to fail when vertex
attribute aliasing happens on ES3.0 context or later
Product: Mesa
Version: git
Hardware: x86-64 (AMD
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Eduardo Lima Mitev
This function will be the entry point for linking the uniforms from
the nir_shader objects associated with the gl_linked_shaders of a
program.
This patch includes initial support for linking uniforms from NIR
shaders. It is
Seems reasonable.
Reviewed-by: Timothy Arceri
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Neil Roberts
When SpvDecorationBinding is encountered in the SPIR-V source it now
sets explicit_binding on the nir_variable. This will be used to
determine whether to initialise sampler and image
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Eduardo Lima Mitev
Previous patch 'spirv: Get rid of vtn_variable_mode_image/sampler'
made possible that interface_type of image/sampler nir_variable may be
null. This patch accounts for it and make use of the type of the
variable instead of t
Looks ok to me.
Reviewed-by: Timothy Arceri
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Neil Roberts
vtn_variable_mode_image and _sampler are instead replaced with
vtn_variable_mode_uniform which encompasses both of them. In the few
places where it was neccessary to distinguish between
Reviewed-by: Timothy Arceri
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Eduardo Lima Mitev
I think it is more accurate to call it a sampler target (?).
---
src/compiler/nir_types.cpp | 7 +++
src/compiler/nir_types.h | 1 +
2 files changed, 8 insertions(+)
diff --git a/src/co
I guess this will need to be rewritten with all the deref stuff almost
ready to land. If you manage to get this stuff in before then you can
have a:
Reviewed-by: Timothy Arceri
On 12/05/18 19:40, Alejandro Piñeiro wrote:
From: Eduardo Lima Mitev
Vulkan has the concept of separate image and
I still dislike the idea of this bool but in the name of progress.
Patches 1-2 are:
Reviewed-by: Timothy Arceri
On 12/05/18 19:40, Alejandro Piñeiro wrote:
---
src/mesa/drivers/dri/i965/brw_program.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i
There aren't to many people familiar with this part of the compiler.
CCing Ian and Ken in case they missed this patch.
On 29/05/18 04:21, zhaowei yuan wrote:
Line number of a predefined macro should be set
as where it is referenced rather than declared
Signed-off-by: zhaowei yuan
Bugzilla: h
Reviewed-by: Ian Romanick
On 06/04/2018 02:33 PM, zhaowei yuan wrote:
> GLSL ES 1.0.17 specifies that "double" is a keyword reserved
>
> Signed-off-by: zhaowei yuan
> ---
> src/compiler/glsl/glsl_lexer.ll | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/compiler/gl
https://bugs.freedesktop.org/show_bug.cgi?id=106823
--- Comment #5 from Zhaowei Yuan ---
patch has been posted here:
https://patchwork.freedesktop.org/patch/227593/
please review it
--
You are receiving this mail because:
You are the assignee for the bug.
Reviewed-by: Matt Turner
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
https://bugs.freedesktop.org/show_bug.cgi?id=106810
Timothy Arceri changed:
What|Removed |Added
Component|Drivers/DRI/i965|Mesa core
QA Contact|intel-3d-
The LLVM 6 code reduced it to a non-array call. We need to do that
with the new code too.
This fixes dEQP-VK.glsl.texture_functions.query.texturequerylod.*array* for
radv.
Fixes: a9a79934412 "amd/common: use the dimension-aware image intrinsics on
LLVM 7+"
---
src/amd/common/ac_llvm_build.c |
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #9 from Jason Ekstrand ---
(In reply to Ian Romanick from comment #8)
> (In reply to Ian Romanick from comment #7)
> > (In reply to Jason Ekstrand from comment #6)
> > > Also, please turn your test case there into a piglit test. Ext
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #8 from Ian Romanick ---
(In reply to Ian Romanick from comment #7)
> (In reply to Jason Ekstrand from comment #6)
> > Also, please turn your test case there into a piglit test. Extra points if
> > it's a CTS test so we force everyo
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #7 from Ian Romanick ---
(In reply to Jason Ekstrand from comment #6)
> Also, please turn your test case there into a piglit test. Extra points if
> it's a CTS test so we force everyone else to be correct too. :-)
I'm not sure if y
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #6 from Jason Ekstrand ---
Also, please turn your test case there into a piglit test. Extra points if
it's a CTS test so we force everyone else to be correct too. :-)
--
You are receiving this mail because:
You are the assignee fo
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #5 from Jason Ekstrand ---
I'd be a fan of just making GLSL IR bail in copy propagation of anything that
isn't a local, global, or uniform. We could try to make it smart enough to
handle barriers correctly with SSBOs and shared vari
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #4 from Ilia Mirkin ---
(In reply to Ian Romanick from comment #3)
> (In reply to Ilia Mirkin from comment #2)
> > Are you sure that's illegal? I'd imagine that you'd need that variable to be
> > marked as coherent, or to introduce a
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #3 from Ian Romanick ---
(In reply to Ilia Mirkin from comment #2)
> Are you sure that's illegal? I'd imagine that you'd need that variable to be
> marked as coherent, or to introduce a compiler barrier somehow.
Legality isn't reall
https://bugs.freedesktop.org/show_bug.cgi?id=106687
Bas Nieuwenhuizen changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
The memcpy had the wrong size and this was causing crashes on 32-bit
builds of the driver.
Fixes: 6a9525bf6729a8 "intel/eu: Switch to a logical state stack"
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106830
---
src/intel/compiler/brw_eu.c | 2 +-
1 file changed, 1 insertion(+), 1 dele
https://bugs.freedesktop.org/show_bug.cgi?id=106823
--- Comment #4 from Kenneth Graunke ---
(In reply to Mark Janes from comment #3)
> Mesa passes the GLES3 variant of this test. Also, the test is listed in the
> following dEQP source file:
>
> android/cts/master/src/gles2-failures.txt
>
> A
Signed-off-by: Gert Wollny
---
.../tests/test_glsl_to_tgsi_array_merge.cpp| 211 +
1 file changed, 211 insertions(+)
diff --git a/src/mesa/state_tracker/tests/test_glsl_to_tgsi_array_merge.cpp
b/src/mesa/state_tracker/tests/test_glsl_to_tgsi_array_merge.cpp
index ca4
v4: Also track the register given in inst->resource. (thanks: Benedikt Schemmer
for testing the patches on radeonsi, which revealed that I was missing
tracking this)
Signed-off-by: Gert Wollny
---
.../state_tracker/st_glsl_to_tgsi_temprename.cpp | 62 +-
1 file cha
Signed-off-by: Gert Wollny
---
src/mesa/state_tracker/tests/st_tests_common.cpp | 181 ---
src/mesa/state_tracker/tests/st_tests_common.h | 32 +++-
2 files changed, 186 insertions(+), 27 deletions(-)
diff --git a/src/mesa/state_tracker/tests/st_tests_common.cpp
b/src/mes
In preparation of the array live range tracking the evaluation of the read
mask is moved out the register live range tracking to the enclosing call
of the generalized read access tracking.
Signed-off-by: Gert Wollny
---
src/mesa/state_tracker/st_glsl_to_tgsi_temprename.cpp | 15 ---
This patch ties in the array split, merge, and interleave code.
shader-db changes in the TGSI code are:
original code | array-merge | change
mean max | meanmax | best mean % worst
---
In preparartion of adding the tracking of the live range the classes that refer
to temporary registers are renamed.
Reviewed-by: Nicolai Hähnle
Signed-off-by: Gert Wollny
---
.../state_tracker/st_glsl_to_tgsi_temprename.cpp | 61 +++---
.../state_tracker/st_glsl_to_tgsi_tempre
Because of the indirect access it is impossible to obtain an accurate per
component and array element tracking. Therefore, the tracking is simplified
to only track whether any element was accessed, whether this happend
conditionally in a loop. In addition, while tracking of temporaries requires
a p
v4: - Remove logic for evaluation of swizzles and merges since this
was moved to array_live_range. This class now only handles the
actual remapping.
Signed-off-by: Gert Wollny
---
.../state_tracker/st_glsl_to_tgsi_array_merge.cpp | 100 +
.../state_tracker/
v2: - Define tests also in the meson.build file.
v4: - Check no-op mapping of all bits.
- Convert tests to the new class layout used in the merge evaulation.
- remove dependency on llvm in meson build (Thanks Dylan Baker for pointing
out that this might not needed)
Signed-off-by: Ge
v4: - Update the code to use the new merge logic.
- Use a cleaner, class-based approach for the evaluation of merges.
Signed-off-by: Gert Wollny
---
.../state_tracker/st_glsl_to_tgsi_array_merge.cpp | 383 -
.../state_tracker/st_glsl_to_tgsi_array_merge.h| 26 +-
2 f
This class holds the array length, live range, and accessed components, and
it implements the logic for evaluating how arrays are merged and interleaved.
v4: - Add logic to evaluate merge and interleave of a pair of arrays to
the class array_live_range.
- document class
- update comm
in constructs like below, currently the live range estimation extends the live
range
of t unecessarily to the whole loop because it was not detected that t is
unconditional written and later read only in the "if (a)" scope.
while (foo) {
...
if (a) {
...
if (b)
t
When mesa is compiled in debug mode then this adds the possibility
to print out some statistics about the translated and optimized TGSI
shaders to a file.
The functionality is enabled by setting the environment variable
GLSL_TO_TGSI_PRINT_STATS
to the file name where the statistics should be
On one hand "live range" is the term used in the literature, and on the
other hand a distinction is needed from the array live ranges.
v4: Fix indentions and white spaces
Reviewed-by: Nicolai Hähnle (v3)
Signed-off-by: Gert Wollny
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 16 ++-
Array whose elements are only accessed directly are replaced by the
according number of temporary registers. By doing so the otherwise
reserved register range becomes subject to further optimizations like
copy propagation and register merging.
Thanks to the resulting reduced register pressure this
taking the comments of Nicolai Hähnle into account I've updated the series. I'd
also
like to thank Benedikt Schemmer for going out on the adventure to these these
patches
on radeonsi for which this code path is actually not relevant. It made me
realize,
that I was not tracking the inst->resour
Having this if statement here prevented the next if statement from being
reached in the case of image stores, which is needed for instructions with
indirect bindless handles like "STORE TEMP[ADDR[2].x+1](1) ...".
Signed-off-by: Rhys Perry
---
src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi
https://bugs.freedesktop.org/show_bug.cgi?id=106774
--- Comment #2 from Ilia Mirkin ---
Are you sure that's illegal? I'd imagine that you'd need that variable to be
marked as coherent, or to introduce a compiler barrier somehow.
One could also argue that an atomic should introduce such a barrier
Hopefully this reduces the number of fixup commits we need for the
automake build.
---
.travis.yml | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/.travis.yml b/.travis.yml
index 3a88b24bdbec..c53d74ea44ca 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -159,7 +159,7 @@ matrix:
This will catch many automake bugs that end-users get to experience first,
otherwise.
---
.travis.yml | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/.travis.yml b/.travis.yml
index 1f83f5b8d74d..3a88b24bdbec 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -591,7 +591,9 @@ scr
https://bugs.freedesktop.org/show_bug.cgi?id=106806
Bas Nieuwenhuizen changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://bugs.freedesktop.org/show_bug.cgi?id=106774
Ian Romanick changed:
What|Removed |Added
Summary|NIR "copy propagates" loads |GLSL IR copy propagates
Hi everyone,
Throwing a small suggestion here:
Once a consensus is reached, do add some beefy comments inline. Mostly
of any parts that are different from the X11/GBM/other backends.
A reference to this thread might also be a good idea.
Thanks
Emil
___
On 5 June 2018 at 18:13, Matt Turner wrote:
> On Tue, Jun 5, 2018 at 8:24 AM, Emil Velikov wrote:
>> On 5 June 2018 at 16:07, Eric Engestrom wrote:
>>> On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
From: Emil Velikov
Recently the wayland-egl library and pkg-config fi
We have no corresponding ISL format for XBGR2101010.
Signed-off-by: Lionel Landwerlin
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106642
---
src/mesa/drivers/dri/i965/intel_screen.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/intel_screen.c
b/src/m
An invalid format could be computed through the
modifier_is_supported() helper. Better verify each step.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106642
Signed-off-by: Lionel Landwerlin
---
src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 4
src/mesa/drivers/dri/i965/intel_scr
https://bugs.freedesktop.org/show_bug.cgi?id=106801
Scott D Phillips changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://bugs.freedesktop.org/show_bug.cgi?id=106776
--- Comment #4 from Scott D Phillips ---
oops, I meant this one:
commit db9cec5689bfc4031e0ddcc5b177c09f2deabb07
Author: Scott D Phillips
Date: Fri Jun 1 16:26:56 2018 -0700
util: tests: vma test depends on C++11 support
Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=106776
Scott D Phillips changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
https://bugs.freedesktop.org/show_bug.cgi?id=106778
Scott D Phillips changed:
What|Removed |Added
Resolution|--- |FIXED
Status|NEW
On Tue, 2018-06-05 at 12:41 +0100, Daniel Stone wrote:
> Hi Juan,
>
> On 5 June 2018 at 09:51, Juan A. Suarez Romero wrote:
> > On Mon, 2018-06-04 at 13:22 +0100, Daniel Stone wrote:
> > > The first query will correctly return (w1,h1). The second query will
> > > incorrectly also return (w1,h1),
Hi Dan,
Sharing a couple of small ideas.
On 5 June 2018 at 18:02, Daniel Stone wrote:
>>> drm-gralloc.git
Empty - please nuke, alongside bugzilla & other infra.
>>> drm.git
Out of curiosity - this and others (say igt) projects are accessible
as mesa/$foo and drm/$foo.
I'd image the same approa
On Tue, Jun 5, 2018 at 8:24 AM, Emil Velikov wrote:
> On 5 June 2018 at 16:07, Eric Engestrom wrote:
>> On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
>>> From: Emil Velikov
>>>
>>> Recently the wayland-egl library and pkg-config file were moved to the
>>> Wayland repository. With th
On MacOS, the build fails because of a compiler complaint about
a downcast:
vma_random_test.cpp:239:18: error: non-constant-expression cannot be narrowed
from type 'unsigned long' to 'uint_fast32_t' (aka 'unsigned int') in
initializer list [-Wc++11-narrowing]
random_test r{seed};
Signed-off-by: Pierre Moreau
---
Notes:
Changes in:
* v8: Add DEFINES to libclspirv and libclover, in autotools, as they would
otherwise never know whether CLOVER_ALLOW_SPIRV has been defined
(Dave Airlie)
* v7: Update the dependency name (meson) and the libs variable (Make
Reviewed-by: Karol Herbst
Reviewed-by: Dylan Baker
Signed-off-by: Pierre Moreau
---
Notes:
Changes in:
* v8: * Properly align LLVMSPIRVLib comment (Dylan Baker)
* Only define CLOVER_ALLOW_SPIRV when **both** dependencies are found:
autotools was only requiring one
On 5 June 2018 at 17:55, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 17:52:17 +0100, Daniel Stone wrote:
>> > I assume that's now mesa, libdrm, piglit, shader-db, crucible, kmscube,
>> > mesa-demos; what about igt? Anything else?
>>
>> We currently have under /git/mesa:
>> clover.git
>> crucibl
https://bugs.freedesktop.org/show_bug.cgi?id=106823
--- Comment #3 from Mark Janes ---
Mesa passes the GLES3 variant of this test. Also, the test is listed in the
following dEQP source file:
android/cts/master/src/gles2-failures.txt
Are we sure that this test should pass on GLES2?
--
You a
On Tuesday, 2018-06-05 17:52:17 +0100, Daniel Stone wrote:
> Hi Eric,
>
> On 5 June 2018 at 17:49, Eric Engestrom wrote:
> > On Tuesday, 2018-06-05 09:11:58 -0700, Jason Ekstrand wrote:
> >> Given the discussion below, I think we'll make piglit a sub-project of
> >> mesa. Those who need commit a
On Tuesday, 2018-06-05 17:17:03 +0100, Emil Velikov wrote:
> On 5 June 2018 at 16:38, Eric Engestrom wrote:
> > On Tuesday, 2018-06-05 16:24:59 +0100, Emil Velikov wrote:
> >> On 5 June 2018 at 16:07, Eric Engestrom wrote:
> >> > On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
> >> >>
Hi Eric,
On 5 June 2018 at 17:49, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 09:11:58 -0700, Jason Ekstrand wrote:
>> Given the discussion below, I think we'll make piglit a sub-project of
>> mesa. Those who need commit access to piglit but not mesa can be added
>> directly to the piglit pro
On Tue, Jun 5, 2018 at 7:14 AM, Emil Velikov wrote:
> From: Emil Velikov
>
> Recently the wayland-egl library and pkg-config file were moved to the
> Wayland repository. With that a strange conflict came to be - which one
> should be used and when.
>
> The long term goal is to remove the Mesa cop
On Tuesday, 2018-06-05 09:11:58 -0700, Jason Ekstrand wrote:
> Given the discussion below, I think we'll make piglit a sub-project of
> mesa. Those who need commit access to piglit but not mesa can be added
> directly to the piglit project.
>
> Unless someone objects strongly, I'll plan to migrat
An invalid format could be computed through the
modifier_is_supported() helper. Better verify each step.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=10664
Signed-off-by: Lionel Landwerlin
---
src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 4
src/mesa/drivers/dri/i965/intel_scre
Tested-by: Gurchetan Singh
Reviewed-by: Gurchetan Singh
On Mon, Jun 4, 2018 at 9:32 PM Dave Airlie wrote:
>
> From: Dave Airlie
>
> This enables ARB_gpu_shader_fp64 if the host provides it.
> ---
> src/gallium/drivers/virgl/virgl_screen.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(
On 5 June 2018 at 16:38, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 16:24:59 +0100, Emil Velikov wrote:
>> On 5 June 2018 at 16:07, Eric Engestrom wrote:
>> > On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
>> >> From: Emil Velikov
>> >>
>> >> Recently the wayland-egl library and
Given the discussion below, I think we'll make piglit a sub-project of
mesa. Those who need commit access to piglit but not mesa can be added
directly to the piglit project.
Unless someone objects strongly, I'll plan to migrate piglit on Thursday
along with all the other mesa repos.
On Mon, Jun
https://bugs.freedesktop.org/show_bug.cgi?id=106784
Eric Engestrom changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
On Tue, Jun 5, 2018 at 10:38 AM, Philipp Zabel wrote:
> The temporary input/output register 0 is reserved for position in the
> fragment shader. Hardwire it to 0 and start other input/output variables
> at 1. The intrinsic input load and output store base corresponds to the
> temporary register nu
On Tuesday, 2018-06-05 08:37:55 -0700, Dylan Baker wrote:
> You might also be interested in this:
> https://github.com/mesonbuild/meson/pull/3696
I am, thanks for doing this :)
>
> Quoting Eric Engestrom (2018-06-05 03:59:40)
> > I'd like to eventually drop support for the confusing "an array of
Am Di., 5. Juni 2018 um 16:40 Uhr schrieb Philipp Zabel
:
>
> Hi!
>
> we have been interested in NIR support for etnaviv for a while, for the
> obvious reasons: gaining access to common optimizations, better support
> for non-trivial code transformations, better register allocation, and
> the promi
On Tuesday, 2018-06-05 16:24:59 +0100, Emil Velikov wrote:
> On 5 June 2018 at 16:07, Eric Engestrom wrote:
> > On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
> >> From: Emil Velikov
> >>
> >> Recently the wayland-egl library and pkg-config file were moved to the
> >> Wayland reposito
On Tue, Jun 5, 2018 at 10:38 AM, Philipp Zabel wrote:
> Since all threads share a global temporary vec4 register file, it is
> important to reduce temporary register use of shaders.
> Using source swizzles and destination write mask of ALU operations we
> can layer smaller virtual registers on top
You might also be interested in this:
https://github.com/mesonbuild/meson/pull/3696
Quoting Eric Engestrom (2018-06-05 03:59:40)
> I'd like to eventually drop support for the confusing "an array of
> a single empty string is meant to be interpreted as an empty array", so
> let's start by not using
---
Forgot this one ...
src/amd/vulkan/radv_pipeline.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/src/amd/vulkan/radv_pipeline.c b/src/amd/vulkan/radv_pipeline.c
index 34249bf843d..375f7c357d3 100644
--- a/src/amd/vulkan/radv_pipeline.c
+++ b/src/amd/vulkan/radv_pipeline.c
@@ -2039,8 +
On 5 June 2018 at 16:15, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 16:03:14 +0100, Emil Velikov wrote:
>> On 5 June 2018 at 15:36, Eric Engestrom wrote:
>> > On Tuesday, 2018-06-05 15:14:34 +0100, Emil Velikov wrote:
>> >> From: Emil Velikov
>> >>
>> >> As covered earlier - this is the vers
On 5 June 2018 at 16:07, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
>> From: Emil Velikov
>>
>> Recently the wayland-egl library and pkg-config file were moved to the
>> Wayland repository. With that a strange conflict came to be - which one
>> should be us
This seems fine to me,
Reviewed-by: Dylan Baker
Quoting Eric Engestrom (2018-06-05 03:59:40)
> I'd like to eventually drop support for the confusing "an array of
> a single empty string is meant to be interpreted as an empty array", so
> let's start by not using it anymore.
>
> Signed-off-by: E
On Tuesday, 2018-06-05 16:03:14 +0100, Emil Velikov wrote:
> On 5 June 2018 at 15:36, Eric Engestrom wrote:
> > On Tuesday, 2018-06-05 15:14:34 +0100, Emil Velikov wrote:
> >> From: Emil Velikov
> >>
> >> As covered earlier - this is the version which ships
> >> wayland-egl{,-backend} and associa
On Tuesday, 2018-06-05 15:14:30 +0100, Emil Velikov wrote:
> From: Emil Velikov
>
> Recently the wayland-egl library and pkg-config file were moved to the
> Wayland repository. With that a strange conflict came to be - which one
> should be used and when.
>
> The long term goal is to remove the
On 5 June 2018 at 15:36, Eric Engestrom wrote:
> On Tuesday, 2018-06-05 15:14:34 +0100, Emil Velikov wrote:
>> From: Emil Velikov
>>
>> As covered earlier - this is the version which ships
>> wayland-egl{,-backend} and associated files.
>>
>> Signed-off-by: Emil Velikov
>> ---
>> .travis.yml |
On Sun, Jun 3, 2018 at 4:18 PM, Bas Nieuwenhuizen
wrote:
> On Sat, Jun 2, 2018 at 2:48 AM, Rob Clark wrote:
> > On Fri, Jun 1, 2018 at 1:01 AM, Jason Ekstrand
> wrote:
> >> This is something that Connor and I have been talking about for some
> time
> >> now. The basic idea is to replace the cu
From: Michael Tretter
Ignore ssa_undef nir instructions when generating code, because all
users of the undefined values are removed and undefined values are not
used. The instructions should be removed while rewriting the users of
undefined variables, but are not removed yet.
Signed-off-by: Mich
We represent allocated temporaries as NIR global registers. Turn local
registers left over by the nir_convert_from_ssa pass into global
registers with the correct temporary number.
Signed-off-by: Philipp Zabel
---
src/gallium/drivers/etnaviv/etnaviv_nir.c | 21 +
1 file chang
The temporary input/output register 0 is reserved for position in the
fragment shader. Hardwire it to 0 and start other input/output variables
at 1. The intrinsic input load and output store base corresponds to the
temporary register number.
Signed-off-by: Philipp Zabel
Signed-off-by: Michael Tre
From: Michael Tretter
If a uniform is source to an output store operation, we have to emit an
actual mov instruction that copies from the uniform register into the
temporary register that is used as a shader output.
Signed-off-by: Michael Tretter
Signed-off-by: Philipp Zabel
---
src/gallium/d
From: Michael Tretter
The hardware does not allow two different uniform registers to be used
as sources in the same ALU instruction. Emit mov instructions to
temporary registers for all but one uniform register in this case.
Signed-off-by: Michael Tretter
Signed-off-by: Philipp Zabel
---
src/
From: Michael Tretter
Start emitting ALU instructions for nir_alu_instr structures from the
NIR shader's main function implementation.
Signed-off-by: Michael Tretter
Signed-off-by: Philipp Zabel
---
.../drivers/etnaviv/etnaviv_compiler.c| 290 ++
1 file changed, 290 in
Emit TEXLD, TEXLDB, and TEXLDL instructions from nir_texop_tex,
nir_texop_txb, and nir_texop_txl texture loads, respectively.
Signed-off-by: Philipp Zabel
Signed-off-by: Michael Tretter
---
.../drivers/etnaviv/etnaviv_compiler.c| 65 +++
1 file changed, 65 insertions(+)
From: Michael Tretter
flog2 is implemented as 2 instructions in the Vivante machine code. The
log calculates x and y, which have to be multiplied to get the actual
log result.
We need to allocate a temporary register for this, which will fail if to
many flog2 instructions are used (e.g. in the g
Add a debug option ETNA_MESA_DEBUG="nir" that will cause the etnaviv
gallium pipe driver to advertise support and preference for NIR shaders.
Signed-off-by: Philipp Zabel
Signed-off-by: Michael Tretter
---
src/gallium/drivers/etnaviv/etnaviv_debug.h | 3 +++
src/gallium/drivers/etnaviv/etnavi
Emit conditional branches for nir_cf_code_if blocks following a
comparison operation. The NIR compiler does not assign registers to
the comparison operations, and the emitter does not emit any
instructions for them.
Instead, the nir_cf_code_if blocks cause emission of a conditional
branch instructi
The texture instructions expect the source register to have the correct number
of components. Since we only have vec4 hardware registers, rewrite the number
of components to 4 and insert a mov instruction to a new vec4 SSA right before
the texture load instruction.
TODO: To support destination reg
From: Michael Tretter
If the source to a mov instruction is undefined, the result is undefined
as well. In that case it is valid to drop the mov instruction.
Signed-off-by: Michael Tretter
Signed-off-by: Philipp Zabel
---
src/gallium/drivers/etnaviv/etnaviv_nir.c | 35 +++
The Vivante instruction set does not allow empty shaders.
If the shader is empty, add a nop to allow execution.
Signed-off-by: Philipp Zabel
Signed-off-by: Michael Tretter
---
src/gallium/drivers/etnaviv/etnaviv_nir.c | 16
1 file changed, 16 insertions(+)
diff --git a/src/gal
1 - 100 of 159 matches
Mail list logo
