Reviewed-by: Tapani Pälli
On 30.01.2018 17:11, Juan A. Suarez Romero wrote:
According with OpenGL GLSL 3.20 spec, section 4.3.9:
"It is a link-time error if any particular shader interface
contains:
- two different blocks, each having no instance name, and each
having a me
Reviewed-by: Tapani Pälli
(I verified the 1st one and I trust you on the 2nd one.)
BTW I witnessed pahole crashing when processing visit() methods of
ir_print_visitor class, did you experience that? My pahole version is
1.9, it dies in /lib64/libdwarves.so.1 after some prints like:
--- 8< -
On Tue, Jan 30, 2018 at 12:25 PM, Michel Dänzer wrote:
> On 2018-01-29 10:34 AM, Ian Kumlien wrote:
>> Hi, I know that this is not exactly the right place but I'm kinda
>> dumbfounded...
>>
>> I run wayland with mesa on a radeon rx 480 graphics card - but it
>> stopped working a while ago (been fo
On 2018-01-30 23:17:04, Tapani Pälli wrote:
> This patch makes disk_cache initialize path and index lazily so
> that we can utilize disk_cache without a path using callback
> functionality introduced by next patch.
>
> v2: unmap mmap and destroy queue only if index_mmap exists
>
> Signed-off-by:
On 2018-01-30 23:17:05, Tapani Pälli wrote:
> v2: add disk_cache_has_key, disk_cache_put_key support
> using blob cache (Nicolai, Jordan)
>
> v3: rename set_cb as put_cb to match existing naming (Timothy)
>
> Signed-off-by: Tapani Pälli
> ---
> src/util/disk_cache.c | 49 +++
On 2018-01-30 23:17:06, Tapani Pälli wrote:
> v2: adjust to change that moved cache from ctx to screen
>
> Signed-off-by: Tapani Pälli
> ---
> src/mesa/drivers/dri/i965/intel_screen.c | 21 +
> 1 file changed, 21 insertions(+)
>
> diff --git a/src/mesa/drivers/dri/i965/intel
Am Dienstag, 30. Januar 2018, 00:34:18 CET schrieb Dylan Baker:
> Quoting Marc Dietrich (2018-01-27 06:36:51)
>
> > Hi Dylan,
> >
> > Am Donnerstag, 25. Januar 2018, 20:32:23 CET schrieb Dylan Baker:
> > > Currently meson implements the same logic as SCons for translations,
> > > namely it doesn'
Series is:
Reviewed-by: Samuel Pitoiset
On 01/31/2018 12:41 AM, Dave Airlie wrote:
From: Dave Airlie
Bas's pipeline rework made me relook at the struct packing:
radv_cmd_state: 984->968
radv_cmd_buffer: 2910->2896
radv_image: 1008->1000
radv_pipeline: 1640->1632
Signed-off-by: Dave Airlie
You also need to update docs/relnotes.
On 01/31/2018 03:46 AM, Andres Rodriguez wrote:
Support for these extensions is available in radeonsi.
Signed-off-by: Andres Rodriguez
---
docs/features.txt | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/docs/features.txt b/doc
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_nir_to_llvm.c | 23 ---
1 file changed, 16 insertions(+), 7 deletions(-)
diff --git a/src/amd/common/ac_nir_to_llvm.c b/src/amd/common/ac_nir_to_llvm.c
index fd5989389b..04fe51935a 100644
--- a/src/amd/common/ac_nir_to_llvm
That's quite useless and that pollutes the output.
Signed-off-by: Samuel Pitoiset
---
src/amd/vulkan/radv_pipeline.c | 30 +-
src/amd/vulkan/radv_shader.h | 9 +
2 files changed, 18 insertions(+), 21 deletions(-)
diff --git a/src/amd/vulkan/radv_pipeline.c
Reviewed-by: Juan A. Suarez
On Fri, 2018-01-26 at 12:10 +0100, Antia Puentes wrote:
> This reverts commit 513c2263cbff45edb105c7b46e58f316e06746ab.
>
> _mesa_base_fbo_format_ is used to validate the internalformat
> passed to RenderbufferStorage, which in the OpenGL 4.6 is said:
>
> "An INVALI
Reviewed-by: Bas Nieuwenhuizen
On Wed, Jan 31, 2018 at 11:23 AM, Samuel Pitoiset wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> src/amd/common/ac_nir_to_llvm.c | 23 ---
> 1 file changed, 16 insertions(+), 7 deletions(-)
>
> diff --git a/src/amd/common/ac_nir_to_llvm.c b/
On Thursday, 2018-01-25 16:14:45 -0800, Dylan Baker wrote:
> Signed-off-by: Dylan Baker
Reviewed-by: Eric Engestrom
> ---
>
> I have tested building every mesa driver against this (with and without udev!)
> so I'm pretty sure that this is the last pkgbuild problem.
>
> I'm sure I'll be sad in
deqp does not allow any KHX extensions, and since deqp is included
in android-cts, android does not allow any khx extensions.
So disable VK_KHX_multiview on android.
---
src/amd/vulkan/radv_extensions.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/amd/vulkan/radv_exten
Reviewed-by: Samuel Pitoiset
On 01/31/2018 12:31 PM, Bas Nieuwenhuizen wrote:
deqp does not allow any KHX extensions, and since deqp is included
in android-cts, android does not allow any khx extensions.
So disable VK_KHX_multiview on android.
---
src/amd/vulkan/radv_extensions.py | 2 +-
1
On 30 January 2018 at 14:22, Lucas Stach wrote:
> Take into account the resource format, instead of applying a hardcoded
> 32bpp. This not only over-allocates 16bpp formats, but also results in
> a wrong stride being filled into the handle.
Bikeshed: just use util_format_getblocksizebits()? getbl
On 30 January 2018 at 21:31, Dylan Baker wrote:
> Quoting Emil Velikov (2018-01-30 10:43:06)
>> On 29 January 2018 at 18:57, Dylan Baker wrote:
>> > Quoting Eric Engestrom (2018-01-29 10:15:50)
>> >> The big comment is taken from the equivalent block in configure.ac
>> >>
>> >> Signed-off-by: Eri
On 30 January 2018 at 23:41, Dave Airlie wrote:
> From: Dave Airlie
>
> This repacks to reduce the usage from 72->64 bytes, but also
> makes the descriptions static const so they don't just stay in
> the same unit.
>
> Signed-off-by: Dave Airlie
> ---
> src/amd/vulkan/vk_format.h| 3 ++-
On Tuesday, 2018-01-30 13:31:09 -0800, Dylan Baker wrote:
> Quoting Emil Velikov (2018-01-30 10:43:06)
> > On 29 January 2018 at 18:57, Dylan Baker wrote:
> > > Quoting Eric Engestrom (2018-01-29 10:15:50)
> > >> The big comment is taken from the equivalent block in configure.ac
> > >>
> > >> Sign
The big comment is taken from the equivalent block in configure.ac
Signed-off-by: Eric Engestrom
---
meson.build | 30 +
src/gallium/targets/d3dadapter9/meson.build | 2 +-
src/mesa/drivers/dri/meson.build| 2 +-
3 files c
Signed-off-by: Eric Engestrom
---
meson.build | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/meson.build b/meson.build
index 30c2198d77dde383d7ac..a98410133d462028d670 100644
--- a/meson.build
+++ b/meson.build
@@ -211,11 +211,6 @@ if with_gallium_pl111 and not with_
On Wednesday, 2018-01-31 11:44:11 +, Emil Velikov wrote:
> On 30 January 2018 at 23:41, Dave Airlie wrote:
> > From: Dave Airlie
> >
> > This repacks to reduce the usage from 72->64 bytes, but also
> > makes the descriptions static const so they don't just stay in
> > the same unit.
> >
> > S
Signed-off-by: Eric Engestrom
---
src/gallium/auxiliary/util/u_format_table.py | 98 ++--
1 file changed, 50 insertions(+), 48 deletions(-)
diff --git a/src/gallium/auxiliary/util/u_format_table.py
b/src/gallium/auxiliary/util/u_format_table.py
index a09ae53cbc8966ba5bb1
Signed-off-by: Eric Engestrom
---
src/amd/vulkan/vk_format_table.py | 22 --
1 file changed, 12 insertions(+), 10 deletions(-)
diff --git a/src/amd/vulkan/vk_format_table.py
b/src/amd/vulkan/vk_format_table.py
index 36352b108d0b5220a901..f903e21f697dc42981ed 100644
--- a/src
https://bugs.freedesktop.org/show_bug.cgi?id=7391
yanhua <78666...@qq.com> changed:
What|Removed |Added
CC||78666...@qq.com
--
You are rec
On 31 January 2018 at 07:17, Tapani Pälli wrote:
> v2: cleanup, move callbacks to _egl_display struct (Emil Velikov)
> adapt to earlier ctx->screen changes
>
> Signed-off-by: Tapani Pälli
> ---
> src/egl/drivers/dri2/egl_dri2.c | 25 +
> src/egl/drivers/dri2/egl_dri2.
On 31 January 2018 at 07:17, Tapani Pälli wrote:
> Patch moves functions higher so that we can utilize them from
> test_disk_cache_create which is modified by next patch.
>
> Signed-off-by: Tapani Pälli
> ---
I realise you're just moving stuff, so feel free to ignore the nits.
> src/compiler/gl
Reviewed-by: Bas Nieuwenhuizen
On Wed, Jan 31, 2018 at 11:40 AM, Samuel Pitoiset wrote:
> That's quite useless and that pollutes the output.
>
> Signed-off-by: Samuel Pitoiset
> ---
> src/amd/vulkan/radv_pipeline.c | 30 +-
> src/amd/vulkan/radv_shader.h | 9 +++
On 31.01.2018 11:27, Jordan Justen wrote:
On 2018-01-30 23:17:06, Tapani Pälli wrote:
v2: adjust to change that moved cache from ctx to screen
Signed-off-by: Tapani Pälli
---
src/mesa/drivers/dri/i965/intel_screen.c | 21 +
1 file changed, 21 insertions(+)
diff --git
On 31.01.2018 11:18, Jordan Justen wrote:
On 2018-01-30 23:17:05, Tapani Pälli wrote:
v2: add disk_cache_has_key, disk_cache_put_key support
using blob cache (Nicolai, Jordan)
v3: rename set_cb as put_cb to match existing naming (Timothy)
Signed-off-by: Tapani Pälli
---
src/util/disk
On 31.01.2018 13:58, Emil Velikov wrote:
On 31 January 2018 at 07:17, Tapani Pälli wrote:
v2: cleanup, move callbacks to _egl_display struct (Emil Velikov)
adapt to earlier ctx->screen changes
Signed-off-by: Tapani Pälli
---
src/egl/drivers/dri2/egl_dri2.c | 25 ++
Reviewed-by: Marek Olšák
Marek
On Wed, Jan 31, 2018 at 3:35 AM, Brian Paul wrote:
> This reverts part of the patch which introduced the GLenum16 change.
> Fixes a conform regression found by Roland.
>
> Fixes: f96a69f916aed405 ("mesa: replace GLenum with GLenum16 in
> common structures (v4)")
>
On 31 January 2018 at 07:17, Tapani Pälli wrote:
> This patch makes disk_cache initialize path and index lazily so
> that we can utilize disk_cache without a path using callback
> functionality introduced by next patch.
>
> v2: unmap mmap and destroy queue only if index_mmap exists
>
> Signed-off-
On Wed, Jan 31, 2018 at 4:34 AM, Ian Romanick wrote:
> On 01/30/2018 07:48 AM, Marek Olšák wrote:
>> From: Marek Olšák
>>
>> This only affects drivers that set DriverFlags.NewBlend.
>> ---
>> src/mesa/main/blend.c | 6 --
>> src/mesa/main/blend.h | 41
>> +++
On 31.01.2018 15:07, Tapani Pälli wrote:
On 31.01.2018 13:58, Emil Velikov wrote:
On 31 January 2018 at 07:17, Tapani Pälli wrote:
v2: cleanup, move callbacks to _egl_display struct (Emil Velikov)
adapt to earlier ctx->screen changes
Signed-off-by: Tapani Pälli
---
src/egl/drivers
On 31.01.2018 15:18, Emil Velikov wrote:
On 31 January 2018 at 07:17, Tapani Pälli wrote:
This patch makes disk_cache initialize path and index lazily so
that we can utilize disk_cache without a path using callback
functionality introduced by next patch.
v2: unmap mmap and destroy queue only
On 31 January 2018 at 07:17, Tapani Pälli wrote:
> - /* Before doing anything else, ensure that with
> -* MESA_GLSL_CACHE_DISABLE set to true, that disk_cache_create returns
> NULL.
> -*/
> - setenv("MESA_GLSL_CACHE_DISABLE", "true", 1);
> - cache = disk_cache_create("test", "make_
On 31.01.2018 15:41, Emil Velikov wrote:
On 31 January 2018 at 07:17, Tapani Pälli wrote:
- /* Before doing anything else, ensure that with
-* MESA_GLSL_CACHE_DISABLE set to true, that disk_cache_create returns NULL.
-*/
- setenv("MESA_GLSL_CACHE_DISABLE", "true", 1);
- cache =
On 31.01.2018 15:51, Tapani Pälli wrote:
On 31.01.2018 15:41, Emil Velikov wrote:
On 31 January 2018 at 07:17, Tapani Pälli wrote:
- /* Before doing anything else, ensure that with
- * MESA_GLSL_CACHE_DISABLE set to true, that disk_cache_create
returns NULL.
- */
- setenv("MES
On 31 January 2018 at 13:39, Tapani Pälli wrote:
>
>
> On 31.01.2018 15:18, Emil Velikov wrote:
>>
>> On 31 January 2018 at 07:17, Tapani Pälli wrote:
>>>
>>> This patch makes disk_cache initialize path and index lazily so
>>> that we can utilize disk_cache without a path using callback
>>> funct
When the application doesn't provide its own pipeline cache,
the driver uses a in-memory cache but it shouldn't insert any
entries when the cache is explicitely disabled by the user.
Found while running my experimental pipeline-db tool with a
ton of shaders, the memory footprint was just huge, and
On 31 January 2018 at 13:55, Tapani Pälli wrote:
>
>
> On 31.01.2018 15:51, Tapani Pälli wrote:
>>
>>
>>
>> On 31.01.2018 15:41, Emil Velikov wrote:
>>>
>>> On 31 January 2018 at 07:17, Tapani Pälli wrote:
>>>
- /* Before doing anything else, ensure that with
-* MESA_GLSL_CACHE_DI
On 23/01/2018 18:05, Andres Rodriguez wrote:
Make sure memory is accessible to the external client, for the specified
memory object, before the signal/after the wait.
v2: fixed flush order with respect to wait/signal emission
Signed-off-by: Andres Rodriguez
---
src/mesa/main/dd.h
On 30 January 2018 at 20:27, Dylan Baker wrote:
> Quoting Emil Velikov (2018-01-30 10:56:42)
>> Hi Jon,
>>
>> On 28 January 2018 at 14:24, Jon Turney wrote:
>> > ---
>> > meson.build | 2 +-
>> > 1 file changed, 1 insertion(+), 1 deletion(-)
>> >
>> > diff --git a/meson.build b/meson.build
>> >
https://bugs.freedesktop.org/show_bug.cgi?id=104777
Anton Sudak changed:
What|Removed |Added
CC||anton.su...@gmail.com
--
You are receivi
On 01/31/2018 01:48 AM, Tapani Pälli wrote:
Reviewed-by: Tapani Pälli
(I verified the 1st one and I trust you on the 2nd one.)
BTW I witnessed pahole crashing when processing visit() methods of
ir_print_visitor class, did you experience that? My pahole version is
1.9, it dies in /lib64/libdw
On 31/01/2018 15:21, Emil Velikov wrote:
On 30 January 2018 at 20:27, Dylan Baker wrote:
Quoting Emil Velikov (2018-01-30 10:56:42)
Hi Jon,
On 28 January 2018 at 14:24, Jon Turney wrote:
---
meson.build | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/meson.build b/mes
On 31 January 2018 at 16:07, Jon Turney wrote:
> On 31/01/2018 15:21, Emil Velikov wrote:
>>
>> On 30 January 2018 at 20:27, Dylan Baker wrote:
>>>
>>> Quoting Emil Velikov (2018-01-30 10:56:42)
Hi Jon,
On 28 January 2018 at 14:24, Jon Turney
wrote:
>
> ---
>
https://bugs.freedesktop.org/show_bug.cgi?id=104836
--- Comment #3 from Emil Velikov ---
Having played with this at bit - it's a massive yak exercise.
It's not a new bug by any means, so I'll lower it, slightly, on my priority
list.
Will CC you on the series when it's finished.
--
You are rece
On 01/31/2018 12:55 AM, mathias.froehl...@gmx.net wrote:
From: Mathias Fröhlich
The materials are now moved to the end of the
generic attributes block to the range 4-15.
Before, the way the position and generic 0 attribute
is handled was dependent on the presence and kind of
the currently atta
Hi Matthias,
Nice work! It's nice to get rid of some of those attribute loops.
The series looks good to me, just assorted nit-picks and comments on a
few patches.
-Brian
On 01/31/2018 12:55 AM, mathias.froehl...@gmx.net wrote:
From: Mathias Fröhlich
When executing a display list draw, fo
On 01/31/2018 12:55 AM, mathias.froehl...@gmx.net wrote:
From: Mathias Fröhlich
Since the first material attribute no longer aliases with
the generic0 attribute, only aliasing between generic0 and
position is left and entirely dependent on the enabled
state of the VAO. So introduce a gl_attribu
On 01/31/2018 12:55 AM, mathias.froehl...@gmx.net wrote:
From: Mathias Fröhlich
Instead of each context having its own map instance for
this purpose, use a global static const map.
Signed-off-by: Mathias Fröhlich
---
src/mesa/vbo/vbo_context.c| 23 ++
src/mesa/vbo/vbo_exec.
On 01/31/2018 09:54 AM, Brian Paul wrote:
Hi Matthias,
Nice work! It's nice to get rid of some of those attribute loops.
The series looks good to me, just assorted nit-picks and comments on a
few patches.
Oh, and it might be good to write a new piglit test or two to exercise
the POS/GENERIC
Support for these extensions is available in radeonsi.
v2: also updated relnotes
Signed-off-by: Andres Rodriguez
---
Let me know if the formatting for the relnotes is what is expected.
I based it on the previous versions.
docs/features.txt | 4 ++--
docs/relnotes/18.1.0.html | 3 ++-
On 31 January 2018 at 15:20, Jon Turney wrote:
> On 23/01/2018 18:05, Andres Rodriguez wrote:
>>
>> Make sure memory is accessible to the external client, for the specified
>> memory object, before the signal/after the wait.
>>
>> v2: fixed flush order with respect to wait/signal emission
>>
>> Si
Patches 1 & 2:
Reviewed-by: Nicolai Hähnle
On 31.01.2018 00:41, Dave Airlie wrote:
From: Dave Airlie
This reduces the size from 28->24 bytes.
Signed-off-by: Dave Airlie
---
src/amd/common/ac_shader_info.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/amd/commo
Reviewed-by: Samuel Pitoiset
On 01/31/2018 06:04 PM, Andres Rodriguez wrote:
Support for these extensions is available in radeonsi.
v2: also updated relnotes
Signed-off-by: Andres Rodriguez
---
Let me know if the formatting for the relnotes is what is expected.
I based it on the previous ve
Don't want an overly large numBufferBarriers/numTextureBarriers to blow
up the stack.
Suggested-by: Emil Velikov
Signed-off-by: Andres Rodriguez
---
src/mesa/main/externalobjects.c | 14 ++
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/src/mesa/main/externalobjects.
Signed-off-by: Eric Engestrom
---
v2: I missed a few lines in v1 for some reason; converted as well
---
src/gallium/auxiliary/util/u_format_table.py | 108 ++-
1 file changed, 55 insertions(+), 53 deletions(-)
diff --git a/src/gallium/auxiliary/util/u_format_table.py
b/s
On Tue, Jan 30, 2018 at 05:14:39PM -0800, Jason Ekstrand wrote:
> On Tue, Jan 30, 2018 at 5:03 PM, Nanley Chery wrote:
>
> > On Tue, Jan 30, 2018 at 04:25:59PM -0800, Jason Ekstrand wrote:
> > > On Tue, Jan 30, 2018 at 2:54 PM, Nanley Chery
> > wrote:
> > >
> > > > On Fri, Jan 19, 2018 at 03:47:
On 31 January 2018 at 17:30, Andres Rodriguez wrote:
> Don't want an overly large numBufferBarriers/numTextureBarriers to blow
> up the stack.
>
> Suggested-by: Emil Velikov
> Signed-off-by: Andres Rodriguez
> ---
Thanks for sorting this Andres.
> src/mesa/main/externalobjects.c | 14 +
Don't want an overly large numBufferBarriers/numTextureBarriers to blow
up the stack.
v2: handle malloc errors
Suggested-by: Emil Velikov
Signed-off-by: Andres Rodriguez
---
src/mesa/main/externalobjects.c | 14 ++
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/src/
Ugh, I had to read this change many times to understand it, but I do
think it makes sense. The comments in the code helped a lot too.
Reviewed-by: Rafael Antognolli
On Mon, Jan 29, 2018 at 06:25:30PM +0200, Andres Gomez wrote:
> The emission of vertex attributes corresponding to dvec3 and dvec4
On 2018-01-31 01:25 PM, Andres Rodriguez wrote:
Don't want an overly large numBufferBarriers/numTextureBarriers to blow
up the stack.
v2: handle malloc errors
Someone forgot to update his patch correctly before sending it out...
Suggested-by: Emil Velikov
Signed-off-by: Andres Rodriguez
Don't want an overly large numBufferBarriers/numTextureBarriers to blow
up the stack.
v2: handle malloc errors
v3: fix patch
Suggested-by: Emil Velikov
Signed-off-by: Andres Rodriguez
---
src/mesa/main/externalobjects.c | 48 +++--
1 file changed, 42 inserti
This is a very simple pass that just shrinks load_push_constant
intrinsics when some components are unused. For now, it can just
shrink vec4 to vec3, vec3 to vec2 and so on.
Signed-off-by: Samuel Pitoiset
---
src/compiler/Makefile.sources | 1 +
src/compiler/nir/meson.build
LLVM can't shrink loads.
Totals from affected shaders:
SGPRS: 62528 -> 59955 (-4.11 %)
VGPRS: 44708 -> 44616 (-0.21 %)
Spilled SGPRs: 16 -> 8 (-50.00 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 1355528 -> 135517
In commit 3f353342a6b6744773c26ed66b12afed42bd57af (present in 17.3.0)
we started unconditionally using I915_EXEC_NO_RELOC, which was
introduced in Linux v3.9. ChromeOS kernel 3.8 has backported this,
so it should work too.
Running on older kernels would likely result in every single batch
being
From: Marek Olšák
This only affects drivers that set DriverFlags.NewBlend.
v2: - fix typo advanded -> advanced
- return "enum gl_advanced_blend_mode" from
_mesa_get_advanced_blend_sh_constant
- don't call FLUSH_VERTICES twice
---
src/mesa/main/blend.c | 6 --
src/
From: Marek Olšák
---
src/mesa/state_tracker/st_atom_blend.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/mesa/state_tracker/st_atom_blend.c
b/src/mesa/state_tracker/st_atom_blend.c
index 8f644ba..62042a6 100644
--- a/src/mesa/state_tracker/st_atom_blend.c
+++ b/sr
From: Marek Olšák
If the GL API is missing, different blend functions can't be set
through GL.
---
src/mesa/state_tracker/st_atom_blend.c | 20
1 file changed, 8 insertions(+), 12 deletions(-)
diff --git a/src/mesa/state_tracker/st_atom_blend.c
b/src/mesa/state_tracker/st_
From: Marek Olšák
Non-MRT cases always translate blend state for 1 color buffer only.
MRT cases only check and translate blend state for enabled color buffers.
This also avoids an assertion failure in translate_blend for:
dEQP-GLES31.functional.draw_buffers_indexed.overwrite_common.common_adv
From: Marek Olšák
4 bits per draw buffer, 8 draw buffers in total --> 32 bits.
This is easier to work with.
---
src/mesa/drivers/common/driverfuncs.c| 8 ++--
src/mesa/drivers/common/meta.c | 41 +++---
src/mesa/drivers/common/meta.h
From: Marek Olšák
---
src/mesa/state_tracker/st_atom_blend.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/mesa/state_tracker/st_atom_blend.c
b/src/mesa/state_tracker/st_atom_blend.c
index 62042a6..24ef09b 100644
--- a/src/mesa/state_tracker/st_atom_blend.c
From: Marek Olšák
very few drivers actually implement it.
---
src/gallium/auxiliary/util/u_dump_state.c | 1 -
src/gallium/auxiliary/vl/vl_compositor.c | 1 -
src/gallium/auxiliary/vl/vl_idct.c | 1 -
src/gallium/auxiliary/vl/vl_mc.c | 1 -
src/gallium/au
nine uses it, GL uses it, a bunch of drivers implement it ... why is
it being removed?
On Wed, Jan 31, 2018 at 2:55 PM, Marek Olšák wrote:
> From: Marek Olšák
>
> very few drivers actually implement it.
> ---
> src/gallium/auxiliary/util/u_dump_state.c | 1 -
> src/gallium/auxiliary/vl/vl
On Wed, Jan 31, 2018 at 8:58 PM, Ilia Mirkin wrote:
> nine uses it, GL uses it, a bunch of drivers implement it ... why is
> it being removed?
very few drivers actually implement it.
Marek
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https:
On Wed, Jan 31, 2018 at 2:59 PM, Marek Olšák wrote:
> On Wed, Jan 31, 2018 at 8:58 PM, Ilia Mirkin wrote:
>> nine uses it, GL uses it, a bunch of drivers implement it ... why is
>> it being removed?
>
> very few drivers actually implement it.
nv30, a2xx-a4xx, etnaviv, i915, virgl.
None of the m
Acked-by: Marek Olšák
Marek
On Wed, Jan 31, 2018 at 6:37 AM, Timothy Arceri wrote:
> On 31/01/18 15:05, Timothy Arceri wrote:
>>
>> This has been unused since 100796c15c3a.
>> ---
>>
>> Please note this is not even compile tested as I don't have clover >
>> 7.0.0 repo to go with my current ll
Reviewed-by: Marek Olšák
Marek
On Wed, Jan 31, 2018 at 3:00 AM, Timothy Arceri wrote:
> I forgot to add that adding the opts also required some of the lowering
> passes to be called slightly earlier.
>
>
> On 31/01/18 12:58, Timothy Arceri wrote:
>>
>> All of the current gallium nir driver use
Am 31.01.2018 um 20:55 schrieb Marek Olšák:
> From: Marek Olšák
>
> 4 bits per draw buffer, 8 draw buffers in total --> 32 bits.
>
> This is easier to work with.
> ---
> src/mesa/drivers/common/driverfuncs.c| 8 ++--
> src/mesa/drivers/common/meta.c | 41 +++--
Quoting Kenneth Graunke (2018-01-31 19:33:13)
> In commit 3f353342a6b6744773c26ed66b12afed42bd57af (present in 17.3.0)
> we started unconditionally using I915_EXEC_NO_RELOC, which was
> introduced in Linux v3.9. ChromeOS kernel 3.8 has backported this,
> so it should work too.
>
> Running on olde
On Wed, Jan 31, 2018 at 9:31 PM, Roland Scheidegger wrote:
> Am 31.01.2018 um 20:55 schrieb Marek Olšák:
>> From: Marek Olšák
>>
>> 4 bits per draw buffer, 8 draw buffers in total --> 32 bits.
>>
>> This is easier to work with.
>> ---
>> src/mesa/drivers/common/driverfuncs.c| 8 ++--
Timothy Arceri writes:
> All of the current gallium nir driver use these optimisations but
> they do so in their backends. Having these called in the backend
> only can cause a number of problems:
>
> - Shader compile times are greater because the opts need to do
> significant passes over all s
Marek Olšák writes:
> From: Marek Olšák
>
> very few drivers actually implement it.
I disagree. If the hardware supports it and the API supports it, then
we should support it, too.
I've got a branch around somewhere for vc4 dithering.
signature.asc
Description: PGP signature
___
On Jan 31, 2018 11:14 PM, "Eric Anholt" wrote:
Marek Olšák writes:
> From: Marek Olšák
>
> very few drivers actually implement it.
I disagree. If the hardware supports it and the API supports it, then
we should support it, too.
I've got a branch around somewhere for vc4 dithering.
Fair en
From: Dave Airlie
The vtx operations never got translated, so things worked by
0 being equal to 0, translate them so we can use the proper buffer
resinfo code.
Signed-off-by: Dave Airlie
---
src/gallium/drivers/r600/r600_asm.c| 2 +-
src/gallium/drivers/r600/r600_shader.c | 2 +-
2 files c
On Mon, Jan 22, 2018 at 12:19:33AM -0800, Jason Ekstrand wrote:
> v2 (Jason Ekstrand):
> - Return an enum instead of a boolean
>
> v3 (Jason Ekstrand):
> - Return ANV_FAST_CLEAR_NONE instead of false (Topi)
> - Rename ANV_FAST_CLEAR_ANY to ANV_FAST_CLEAR_DEFAULT_VALUE
> - Add documentation for
Ah I see now how that's supposed to work...
Previous to adding GET_BUFFER_RESINFO the op was just a fixed zero, and
the op for this is the same on eg/cm (and we should not hit it with
r600). But indeed that looks more like the code elsewhere...
Reviewed-by: Roland Scheidegger
Am 01.02.2018 um 01
This fixes some rendering corruption in a couple of Android apps that
use window-system MSAA.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104741
Cc: mesa-sta...@lists.freedesktop.org
Cc: Chad Versace
---
src/mesa/drivers/dri/i965/brw_context.c | 15 +++
1 file changed, 15
From: Dave Airlie
the vpm bit wasn't being applied to the push/pop instructions.
Signed-off-by: Dave Airlie
---
src/gallium/drivers/r600/eg_asm.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/gallium/drivers/r600/eg_asm.c
b/src/gallium/drivers/r600/eg_asm.c
index f8651bd..c03a9d8 10
From: Dave Airlie
This is a cayman only patch, it doesn't appear that
evergreen supports the ALU on VPM. I'll try and figure it out later.
All I can say for this patch is it passes the piglit test and
the CTS tests.
This also disable sb for helper invocations until it can handle
the special ALU
From: Dave Airlie
This picks the correct double packing function.
Signed-off-by: Dave Airlie
---
src/compiler/glsl/lower_64bit.cpp | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/compiler/glsl/lower_64bit.cpp
b/src/compiler/glsl/lower_64bit.cpp
index c7c6d1cb31..
From: Dave Airlie
In order to deal with conversions properly we need to extract
non-64bit sources from vectors instead of expanding them as
the 64-bit code does.
We need non-64bit sources for the 32->64 conversion functions.
Signed-off-by: Dave Airlie
---
src/compiler/glsl/lower_64bit.cpp | 3
From: Dave Airlie
If this function saw a 32-bit source it would just return the
IR without doing any conversion, this adds the ability to denote
where 32-bit sources are expected and will be used in subsequent
patches to add 32->64 conversions.
Signed-off-by: Dave Airlie
---
src/compiler/glsl/
From: Dave Airlie
This just enables the lowering if requested.
Signed-off-by: Dave Airlie
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
b/src/mesa/state_tracker/st_glsl_to_t
From: Dave Airlie
Some hardware has no conversion for these (cayman), so we want
to lower them early using the common code. This adds a flag to allow
the lowering pass to take these conversions into consideration.
Signed-off-by: Dave Airlie
---
src/compiler/glsl/ir_optimization.h | 2 ++
src/
Elie has been working on soft-fp64 code in a branch for a while,
and I've been slowly helping trying to get it in place so we can
enable GL4.x on the evergreen GPUs.
However I've discovered on cayman we don't have int->double or
uint->double support, and my conversion functions which did
(int->f32
1 - 100 of 132 matches
Mail list logo
