Dead code elimination would get rid of the extra instructions, but
skipping this saves iterations through the optimization loop.
>From shader-db:
N Min MaxMedian AvgStddev
x 14672 3 16 3 3.13345150.59904168
+ 14672 1
As per the discussion on IRC, this is trying to fix an issue with a
GLSL compiler pass lower_output_reads that always generates 8 output
writes for any shader which writes gl_FragData. I'll fix this in the
GLSL compiler. NAK.
Marek
On Sun, Oct 20, 2013 at 2:33 AM, Jay Cornwall wrote:
> This patc
This patch identifies shader exports to unbound CBs and removes them during
TGSI to LLVM IR lowering. The method is identical to the one used in the
gallium/r600 driver.
The GLSL lower_output_reads pass generates temporary copies for writes to
shader outputs. In the case of gl_FragData, this resul
On 14 October 2013 10:12, Anuj Phogat wrote:
> Implement the FS backend for new builtins added by the extension:
> in vec2 gl_SamplePosition
> in int gl_SampleID
> in int gl_NumSamples
> out int gl_SampleMask[]
>
There is a lot going on in this one patch, and it's getting hard to follow
all the
On 18 October 2013 10:30, Anuj Phogat wrote:
> I know we can specify stride if we have a brw_reg :-
> fs_reg (stride(brw_vec1_grf(0, 0), 2, 4, 0));
>
> But I could not find a reliable way to use stride if we have a fs_reg.
> That's why I used OR to get the desired result.
>
The right way to do
On 10/19/2013 06:18 AM, Vinson Lee wrote:
Fixes "Uninitialized scalar field" defect reported by Coverity.
Signed-off-by: Vinson Lee
Reviewed-by: Vadim Girlin
---
src/gallium/drivers/r600/sb/sb_shader.cpp | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/gallium/d
On 10/17/2013 02:35 PM, Carl Worth wrote:
There is nothing in the OpenGL specification which prevents the user from
calling glGenQueries to generate a new query object while another object is
active. Neither is there anything in the Mesa implementation which prevents
this. So remove the INVALID_O
On 18 October 2013 11:07, Ian Romanick wrote:
> On 10/17/2013 08:07 PM, Paul Berry wrote:
> > Previously, Mesa followed the linkage rules outlined in the GLSL
> > 1.20-1.40 specs, which (collectively) said that GLSL versions 1.10 and
> > 1.20 could be linked together, but no other versions could
On 18 October 2013 16:46, Eric Anholt wrote:
> Paul Berry writes:
> > Since most geometry shaders used in piglit testing are small,
> > DUAL_INSTANCED mode won't get exercised very much in a normal piglit
> > run. To force DUAL_INSTANCED mode to be used for all geometry
> > shaders, set INTEL_D
On 18 October 2013 17:04, Eric Anholt wrote:
> Paul Berry writes:
>
> > Geometry shaders that run in "DUAL_INSTANCED" mode store their inputs
> > in vec4's. This means that when compiling gl_PointSize input
> > swizzling (a MOV instruction which uses a geometry shader input as
> > both source a
On 10/18/2013 01:47 PM, srol...@vmware.com wrote:
From: Roland Scheidegger
For seamless cube filtering it is necessary to determine new faces and new
coords per sample. The logic for this is _seriously_ complex (what needs
to happen is very "asymmetric" wrt face, x/y under/overflow), further
co
Silence a bunch of MSVC type conversion warnings.
Changed return type of S_FIXED to int32_t (signed). The result
is the same. It just seems more intuitive that a signed conversion
function should return a signed value.
---
src/mesa/main/macros.h |6 +++---
1 file changed, 3 insertions(+), 3
The code for this was removed from Mesa some time ago.
---
include/GL/gl.h | 21 -
1 file changed, 21 deletions(-)
diff --git a/include/GL/gl.h b/include/GL/gl.h
index 968032c..6b94e3f 100644
--- a/include/GL/gl.h
+++ b/include/GL/gl.h
@@ -2102,27 +2102,6 @@ typedef void (AP
This extension never saw any real use so remove it.
---
include/GL/gl.h | 20
src/mapi/glapi/gen/gl_API.xml | 32
2 files changed, 52 deletions(-)
diff --git a/include/GL/gl.h b/include/GL/gl.h
index babb746..968032c 100644
-
From: Christian König
Using atomic function for ncs is superfluous since it is
protected by a mutex anyway. Also lock the mutex only once
while retrieving the next CS for submission.
Signed-off-by: Christian König
---
src/gallium/winsys/radeon/drm/radeon_drm_winsys.c | 31 -
15 matches
Mail list logo
