On 12/14/2017 08:32 PM, Bas Nieuwenhuizen wrote:
On Thu, Dec 14, 2017 at 4:48 PM, Samuel Pitoiset
wrote:
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_shader_info.c | 8
src/amd/common/ac_shader_info.h | 1 +
2 files
Mesa 17.2.7 is now available.
In this release we have:
The current queue consists of a variety of fixes, with a sizeable hunk in the
shared GLSL codebase.
Whereas for individual drivers - i965 has a crash fix for when playing various
Valve games, r600 and nouveau have tweaks in their compiler
Ported from RadeonSI.
Signed-off-by: Samuel Pitoiset
---
src/amd/vulkan/radv_cmd_buffer.c | 14 ++
1 file changed, 14 insertions(+)
diff --git a/src/amd/vulkan/radv_cmd_buffer.c b/src/amd/vulkan/radv_cmd_buffer.c
index d6aaff707b..4a048485c8 100644
---
Ported from RadeonSI.
Signed-off-by: Samuel Pitoiset
---
src/amd/vulkan/radv_cmd_buffer.c | 15 ++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/src/amd/vulkan/radv_cmd_buffer.c b/src/amd/vulkan/radv_cmd_buffer.c
index
Signed-off-by: Samuel Pitoiset
---
src/amd/vulkan/radv_cmd_buffer.c | 12 +---
src/amd/vulkan/radv_device.c | 10 ++
src/amd/vulkan/radv_private.h| 1 +
3 files changed, 12 insertions(+), 11 deletions(-)
diff --git
Use 16_ABGR instead of 32_ABGR if Z isn't written.
Ported from RadeonSI.
No CTS regressions on Polaris.
v2: - make use of ac_get_spi_shader_z_format()
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_nir_to_llvm.c | 46 +++--
Signed-off-by: Samuel Pitoiset
---
src/gallium/drivers/radeonsi/si_shader.c| 22 ++
src/gallium/drivers/radeonsi/si_shader.h| 2 --
src/gallium/drivers/radeonsi/si_state_shaders.c | 3 ++-
3 files changed, 4 insertions(+), 23
ac_shader_util.c will contain shader helpers for RadeonSI
and RADV.
Signed-off-by: Samuel Pitoiset
---
src/amd/Makefile.sources| 5 -
src/amd/common/ac_shader_util.c | 45 +
src/amd/common/ac_shader_util.h | 33
On 12/13/2017 01:59 PM, boyuan.zh...@amd.com wrote:
From: Boyuan Zhang
Signed-off-by: Boyuan Zhang
Reviewed-by: Leo Liu
---
src/gallium/drivers/radeon/radeon_vce_52.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
On 12/13/2017 12:53 AM, Marek Olšák wrote:
From: Marek Olšák
Increase the limit and handle non-square images better.
This makes glxgears 20% faster on APUs, and a little more on dGPUs.
We all use and love glxgears.
We love it. :)
Reviewed-by: Samuel Pitoiset
Hi,
As expected, this series fixes the perf regression in GfxBench when fast
clears were disabled. On SKL GT2:
* 2-5% Manhattan 3.1
* 1% AztecRuins & CarChase (on top of Francisco's large improvement
between the perf regression and this fix)
On 14.12.2017 03:54, Jason Ekstrand wrote:
We should also not load the input SGPRs and VGPRS, but
let's start with this for now.
Signed-off-by: Samuel Pitoiset
---
src/amd/vulkan/radv_shader.c | 11 ---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/src/amd/vulkan/radv_shader.c
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_shader_info.c | 7 ++-
src/amd/common/ac_shader_info.h | 1 +
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/src/amd/common/ac_shader_info.c b/src/amd/common/ac_shader_info.c
index
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_shader_info.c | 8
src/amd/common/ac_shader_info.h | 1 +
2 files changed, 9 insertions(+)
diff --git a/src/amd/common/ac_shader_info.c b/src/amd/common/ac_shader_info.c
index 09dd4bbd55..01949770d6 100644
On 12/13/2017 07:46 AM, Lvzhihong (ReJohn) wrote:
> Hi,
>
> We met a problem on ubuntu17.10 for arm server with amdgpu(AMD
> RADEON PRO WX7100), we use open source driver which are integrated in
> ubuntu17.10. And the architecture is AArch64-linux-gnu.
>
> we install :
>
>
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_nir_to_llvm.c | 3 ++-
src/amd/common/ac_shader_info.c | 3 +++
src/amd/common/ac_shader_info.h | 1 +
src/amd/vulkan/radv_shader.c| 2 +-
4 files changed, 7 insertions(+), 2 deletions(-)
diff --git
Since you remembered to modify dispatch_sanity.cpp in patch 2, I'm going
to assume that 'make check' still passes. If that's the case, the series is
Reviewed-by: Ian Romanick
On 12/14/2017 04:03 AM, Tapani Pälli wrote:
> Hi;
>
> Here's a revisited
Reviewed-by: Bas Nieuwenhuizen
for the series.
On Thu, Dec 14, 2017 at 4:48 PM, Samuel Pitoiset
wrote:
> We should also not load the input SGPRs and VGPRS, but
> let's start with this for now.
>
> Signed-off-by: Samuel Pitoiset
From: Emil Velikov
Cc:
Fixes: 513d7ffa23d ("util: Add a SHA1 unit test program")
Signed-off-by: Emil Velikov
---
We want this and the original commit for stable, to catch any
breakage that may happen.
On Wed, 2017-12-13 at 14:56 -0800, Dylan Baker wrote:
> Quoting Jan Vesely (2017-12-13 14:23:21)
> > On Wed, 2017-12-13 at 13:54 -0800, Dylan Baker wrote:
> > > Quoting Jan Vesely (2017-12-13 12:53:25)
> > > > On Wed, 2017-12-13 at 09:47 -0800, Dylan Baker wrote:
> > > > > +if (with_gallium_va or
Reviewed-by: Bas Nieuwenhuizen
On Thu, Dec 14, 2017 at 5:32 PM, Samuel Pitoiset
wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> src/amd/common/ac_nir_to_llvm.c | 3 ++-
> src/amd/common/ac_shader_info.c | 3 +++
>
Reviewed-by: Bas Nieuwenhuizen
Would it make sense to move the compute_resource_limits calculation to
pipeline creation time?
On Thu, Dec 14, 2017 at 3:51 PM, Samuel Pitoiset
wrote:
> Ported from RadeonSI.
>
> Signed-off-by: Samuel Pitoiset
On Thu, Dec 14, 2017 at 4:48 PM, Samuel Pitoiset
wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> src/amd/common/ac_shader_info.c | 8
> src/amd/common/ac_shader_info.h | 1 +
> 2 files changed, 9 insertions(+)
>
> diff --git
Reviewed-by: Bas Nieuwenhuizen
for the series.
On Thu, Dec 14, 2017 at 1:51 PM, Samuel Pitoiset
wrote:
> ac_shader_util.c will contain shader helpers for RadeonSI
> and RADV.
>
> Signed-off-by: Samuel Pitoiset
>
From: Kevin Rogovin
This patch series implements a needed workaround for Gen9 for ASTC5x5
sampler reads. The crux of the work around is to make sure that the
sampler does not read an ASTC5x5 texture and a surface with an auxilary
buffer without having a texture cache
From: Kevin Rogovin
Signed-off-by: Kevin Rogovin
---
src/mesa/drivers/dri/i965/intel_tex_validate.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/src/mesa/drivers/dri/i965/intel_tex_validate.c
From: Kevin Rogovin
Signed-off-by: Kevin Rogovin
---
src/mesa/drivers/dri/i965/Makefile.sources| 1 +
src/mesa/drivers/dri/i965/brw_context.c | 6 +
src/mesa/drivers/dri/i965/brw_context.h | 24 ++
From: Kevin Rogovin
Signed-off-by: Kevin Rogovin
---
src/mesa/drivers/dri/i965/brw_compute.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/mesa/drivers/dri/i965/brw_compute.c
b/src/mesa/drivers/dri/i965/brw_compute.c
index
From: Kevin Rogovin
Signed-off-by: Kevin Rogovin
---
src/mesa/drivers/dri/i965/brw_draw.c | 16 ++--
src/mesa/drivers/dri/i965/brw_wm_surface_state.c | 5 +
2 files changed, 19 insertions(+), 2 deletions(-)
diff
From: Kevin Rogovin
Signed-off-by: Kevin Rogovin
---
src/mesa/drivers/dri/i965/genX_blorp_exec.c | 5 +
src/mesa/drivers/dri/i965/intel_tex_image.c | 16
2 files changed, 17 insertions(+), 4 deletions(-)
diff --git
---
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 55 +++---
1 file changed, 48 insertions(+), 7 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
b/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
index 2065db3475..c960dc77fb 100644
binner's GatherScissors() will be turned into a real gather in the not
too distant future.
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 176 ++---
1 file changed, 88 insertions(+), 88 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/binner.cpp
---
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 22 +-
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 80 ++
2 files changed, 23 insertions(+), 79 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/jitter/builder_misc.cpp
Move out of binner/clipper; hand them down from the frontend code instead.
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 124 ++---
src/gallium/drivers/swr/rasterizer/core/clip.cpp | 25 ++---
src/gallium/drivers/swr/rasterizer/core/clip.h | 58 +++---
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 9 -
src/gallium/drivers/swr/rasterizer/core/clip.h | 5 -
2 files changed, 12 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/binner.cpp
Highlights include simd16 work, thread pool initialization rework,
and code cleanup.
Tim Rowley (20):
swr/rast: Remove unneeded copy of gather mask
swr/rast: Binner fixes for viewport index offset handling
swr/rast: Corrections to multi-scissor handling
swr/rast: WIP - Widen fetch shader
Widen vertex gather/storage to SIMD16 for all component types.
---
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 716 -
1 file changed, 689 insertions(+), 27 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
---
.../swr/rasterizer/codegen/gen_llvm_ir_macros.py | 3 +-
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 41 -
.../drivers/swr/rasterizer/jitter/builder_misc.h | 7 +-
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 175 ++---
4 files changed, 194
Simplifies calling code, gets gather function interface closer to llvm's
masked_gather.
---
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 20 +
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 34 +-
2 files changed, 14 insertions(+), 40 deletions(-)
Ease future code maintenance, prepare for folding simd8 and simd16 versions.
---
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 244 ++---
1 file changed, 62 insertions(+), 182 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 127 -
src/gallium/drivers/swr/rasterizer/core/binner.h | 127 +
2 files changed, 127 insertions(+), 127 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/binner.cpp
---
src/gallium/drivers/swr/rasterizer/core/clip.h | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/clip.h
b/src/gallium/drivers/swr/rasterizer/core/clip.h
index 148f661ab4..8b947668d3 100644
---
Reviewed-by: Bas Nieuwenhuizen
On Thu, Dec 14, 2017 at 12:51 PM, Samuel Pitoiset
wrote:
> The number of grid components is always 3 when gl_NumWorkGroups
> is declared, because it relies on the number of components of
>
On 12/14/2017 08:35 PM, Bas Nieuwenhuizen wrote:
Reviewed-by: Bas Nieuwenhuizen
Would it make sense to move the compute_resource_limits calculation to
pipeline creation time?
Yeah, possibly.
On Thu, Dec 14, 2017 at 3:51 PM, Samuel Pitoiset
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 118 +++-
src/gallium/drivers/swr/rasterizer/core/clip.cpp | 30 ++--
src/gallium/drivers/swr/rasterizer/core/clip.h | 35 +++--
src/gallium/drivers/swr/rasterizer/core/context.h | 4 +-
Add BASE_NUMA_NODE, BASE_CORE, BASE_THREAD parameters to
SwrCreateContext.
Add optional SWR_API_THREADING_INFO parameter to SwrCreateContext to
control reservation of API threads.
Add SwrBindApiThread() function to allow binding of API threads to
reserved HW threads.
---
---
src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
b/src/gallium/drivers/swr/rasterizer/jitter/fetch_jit.cpp
index ec3b5eafcc..1312ac0009 100644
---
---
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 38 ++---
.../drivers/swr/rasterizer/jitter/builder_misc.h | 5 +-
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 92 ++
3 files changed, 30 insertions(+), 105 deletions(-)
diff --git
---
src/gallium/drivers/swr/rasterizer/core/binner.cpp | 13 +
src/gallium/drivers/swr/rasterizer/core/clip.h | 1 +
2 files changed, 2 insertions(+), 12 deletions(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/binner.cpp
Also widen the 16-bit a 8-bit integer vertex component gathers to SIMD16.
---
.../swr/rasterizer/codegen/gen_llvm_ir_macros.py | 1 +
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 36 +
.../drivers/swr/rasterizer/jitter/builder_misc.h | 3 +
---
src/gallium/drivers/swr/rasterizer/core/api.cpp | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/drivers/swr/rasterizer/core/api.cpp
b/src/gallium/drivers/swr/rasterizer/core/api.cpp
index 25a3f34841..09b482dcc0 100644
---
---
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 60 ++
.../drivers/swr/rasterizer/jitter/builder_misc.h | 3 +-
.../drivers/swr/rasterizer/jitter/fetch_jit.cpp| 30 +--
3 files changed, 32 insertions(+), 61 deletions(-)
diff --git
Replace use of x86 intrinsic with general llvm IR instruction.
Generates the same final assembly.
---
.../swr/rasterizer/codegen/gen_llvm_ir_macros.py | 2 --
.../drivers/swr/rasterizer/jitter/builder_misc.cpp | 30 --
.../drivers/swr/rasterizer/jitter/builder_misc.h |
On Wed, Dec 6, 2017 at 3:31 PM, Ian Romanick wrote:
> On 12/05/2017 08:25 AM, Ilia Mirkin wrote:
>> On Tue, Dec 5, 2017 at 8:18 AM, Emil Velikov
>> wrote:
>>> Hi Rob,
>>>
>>> On 5 December 2017 at 12:54, Rob Clark wrote:
Any chance to wrap up this review?
Thanks!
Eduardo
On 11/30/2017 06:28 PM, Eduardo Lima Mitev wrote:
> Hello,
>
> This is the second version of the series providing initial support for
> ARB_gl_spirv and ARB_spirv_extensions in Mesa and i965.
>
> First version of the series can be found at
>
Fine with me
Reviewed-by: Jason Ekstrand
On Thu, Dec 14, 2017 at 4:56 PM, Kenneth Graunke
wrote:
> According to the RENDER_SURFACE_STATE internal documentation, the
> R32G32B32_FLOAT restriction is marked "IVB" only. We choose to apply
> it to
Any chance to wrap up this review?
Thanks!
Eduardo
On 12/13/2017 08:32 PM, Eduardo Lima Mitev wrote:
> Hi,
>
> This is the 3rd version of the series adding initial support for ARB_gl_spirv.
>
> Previous versions of this series included also support for
> ARB_spirv_extensions, but we have
Tested-by: Dieter Nützel
Dieter
Am 14.12.2017 06:02, schrieb Timothy Arceri:
Fixes 56 crashes in radeonsi.
---
src/mesa/state_tracker/st_glsl_to_nir.cpp | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/mesa/state_tracker/st_glsl_to_nir.cpp
Tested-by: Dieter Nützel
Dieter
Am 14.12.2017 00:14, schrieb Timothy Arceri:
---
src/mesa/state_tracker/st_glsl_to_nir.cpp | 22 +-
1 file changed, 17 insertions(+), 5 deletions(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_nir.cpp
Tested-by: Dieter Nützel
Dieter
Am 14.12.2017 04:48, schrieb Timothy Arceri:
We need to move this to a separate loop because
nir_compact_varyings() can alter the IR of a previous stage.
Fixes: 6648bd68fd27 "st/glsl_to_nir: enable NIR link time opts"
---
It turns out there's already a glslang bug for this and it was closed in
March:
https://github.com/KhronosGroup/glslang/issues/809
Unfortunately, there are applications shipping with these shaders so
failure isn't really an option.
--Jason
On Thu, Dec 14, 2017 at 7:56 PM, Jason Ekstrand
On 14.12.2017 21:20, Ian Romanick wrote:
Since you remembered to modify dispatch_sanity.cpp in patch 2, I'm going
to assume that 'make check' still passes. If that's the case, the series is
Reviewed-by: Ian Romanick
Yes, 'make check' passes;
Thanks for the
Oops, sorry, wrong thread.
This is version 2 of the series and there is a version 3 which is the
one that needs review.
Eduardo
On 12/15/2017 08:13 AM, Eduardo Lima Mitev wrote:
> Any chance to wrap up this review?
>
> Thanks!
>
> Eduardo
>
> On 11/30/2017 06:28 PM, Eduardo Lima Mitev wrote:
The Talos Principle contains shaders with an OpSelect between two
vectors where the condition is a scalar boolean. This is technically
against the spec bout nir_builder gracefully handles it by splatting
out the condition to all the channels. So long as the condition is a
boolean, just emit a
This one do not apply anylonger after Samuel's commit
amd/common: add ac_build_waitcnt()
#225b19880204024a805cc54b1001d09ef3b58054
For your motivation:
I've tested V1 and V2 of the whole series (before the latest master
commits) and could ran _all_ my 'normal' stuff.
Even UH run with GREAT
According to the RENDER_SURFACE_STATE internal documentation, the
R32G32B32_FLOAT restriction is marked "IVB" only. We choose to apply
it to Ivybridge and Baytrail, but not Haswell.
Fixes KHR-GL46.texture_size_promotion.functional on Haswell.
Changes these tests from crashing to skipping on
This allows dclose()'ing this code in dynamically-linked library without
leaking memory.
---
src/gallium/state_trackers/osmesa/osmesa.c | 77 +-
1 file changed, 55 insertions(+), 22 deletions(-)
diff --git a/src/gallium/state_trackers/osmesa/osmesa.c
This looks really good :)
I can't find anything to nitpick :
Reviewed-by: Lionel Landwerlin
On 13/12/17 20:05, Jason Ekstrand wrote:
Both aubinator and aubinator_error_decode try and do the same task of
decoding batches. They both have code to try and decode
The number of grid components is always 3 when gl_NumWorkGroups
is declared, because it relies on the number of components of
nir_instrinsic_load_num_work_groups.
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_nir_to_llvm.c | 9 ++---
Use a boolean instead because the number of needed SGPRs
is always 3.
Signed-off-by: Samuel Pitoiset
---
src/amd/common/ac_nir_to_llvm.c | 7 ---
src/amd/common/ac_shader_info.c | 2 +-
src/amd/common/ac_shader_info.h | 2 +-
3 files changed, 6 insertions(+), 5
On 12/13/2017 09:21 PM, Bas Nieuwenhuizen wrote:
On Tue, Dec 12, 2017 at 6:08 PM, Samuel Pitoiset
wrote:
Use 16_ABGR instead of 32_ABGR if Z isn't written.
Ported from RadeonSI.
No CTS regressions on Polaris.
Signed-off-by: Samuel Pitoiset
On 13/12/17 18:52, Timothy Arceri wrote:
V2: drop type param and just use ctx->i32
I forgot to add that this drops the ctx->nctx check. Both driver now
just follow the same path, the strangeness I had been seeing is no
longer present it was probably just a bug during development.
---
Hi;
Here's a revisited GL_EXT_disjoint_timer_query series. One patch got
dropped (as discussed with Lionel) and enabling is now via
EXT_disjoint_timer_query boolean as was intended (Ian).
Thanks;
Tapani Pälli (4):
mesa: add DisjointOperation to gl_shared_state
glapi: add
Most entrypoints already available via other extensions like
GL_EXT_occlusion_query_boolean, GL_EXT_timer_query.
Signed-off-by: Tapani Pälli
Reviewed-by: Lionel Landwerlin
---
src/mapi/glapi/gen/es_EXT.xml | 16
Following dEQP cases pass:
dEQP-EGL.functional.get_proc_address.extension.gl_ext_disjoint_timer_query
dEQP-EGL.functional.client_extensions.disjoint
Piglit test 'ext_disjoint_timer_query-simple' passes with these changes.
No changes/regression observed in Intel CI.
Signed-off-by: Tapani
This state will be used by EXT_disjoint_timer_query. As first
usage, patch sets DisjointOperation true when gpu reset happens.
Signed-off-by: Tapani Pälli
Reviewed-by: Lionel Landwerlin
---
src/mesa/main/mtypes.h | 8
Patch adds GL_GPU_DISJOINT_EXT and enables to use timer queries when
EXT_disjoint_timer_query is enabled.
v2: enable extension only when EXT_disjoint_timer_query set
Signed-off-by: Tapani Pälli
Reviewed-by: Lionel Landwerlin (v1)
---
77 matches
Mail list logo