On Tue, Sep 23, 2014 at 2:14 PM, Matt Turner wrote:
> On Sat, Sep 20, 2014 at 10:23 AM, Jason Ekstrand wrote:
>> Signed-off-by: Jason Ekstrand
>> ---
>> src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 23 ++-
>> 1 file changed, 18 insertions(+), 5 deletions(-)
>>
>> diff --
On Wed, Aug 6, 2014 at 10:46 AM, Ian Romanick wrote:
> On 07/30/2014 04:11 AM, Tapani Pälli wrote:
>> Current version can create ir_expression where operands have
>> different base type, patch adds support for unsigned type.
>>
>> Signed-off-by: Tapani Pälli
>> https://bugs.freedesktop.org/show_b
On Sat, Mar 28, 2015 at 12:23 AM, Kenneth Graunke wrote:
> These are nir_cf_nodes, not ALU instructions.
> Also, use unreachable() to preempt said review feedback.
>
> Signed-off-by: Kenneth Graunke
> ---
> src/glsl/nir/nir_validate.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> d
These are nir_cf_nodes, not ALU instructions.
Also, use unreachable() to preempt said review feedback.
Signed-off-by: Kenneth Graunke
---
src/glsl/nir/nir_validate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/glsl/nir/nir_validate.c b/src/glsl/nir/nir_validate.c
inde
The typical case of mat4*mat4*vec4 is 80 scalar multiplications, but
mat4*(mat4*vec4) is only 32.
On HSW (with vec4 vertex shaders):
instructions in affected programs: 4420 -> 3194 (-27.74%)
On BDW (with scalar vertex shaders):
instructions in affected programs: 12756 -> 6726 (-47.27%)
I
---
src/glsl/ast_to_hir.cpp | 62 -
src/glsl/glsl_types.cpp | 67 +
src/glsl/glsl_types.h | 6 +
3 files changed, 78 insertions(+), 57 deletions(-)
diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/as
---
src/glsl/ir.cpp | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/glsl/ir.cpp b/src/glsl/ir.cpp
index 54656f8..4b8ca9b 100644
--- a/src/glsl/ir.cpp
+++ b/src/glsl/ir.cpp
@@ -380,10 +380,12 @@ ir_expression::ir_expression(int op, ir_rvalue *op0,
ir_rvalue *op1)
On Thu, Dec 11, 2014 at 11:34:12PM +0100, Eduardo Lima Mitev wrote:
> From: Samuel Iglesias Gonsalvez
>
> When a vec has more elements than row components in a matrix, the
> code could end up failing an assert inside assign_to_matrix_column().
>
> This patch makes sure that when there is still r
On Fri, Mar 27, 2015 at 11:56 AM, Eric Anholt wrote:
> Matt Turner writes:
>
>> instructions in affected programs: 2858 -> 2808 (-1.75%)
>> helped:12
>> ---
>> src/glsl/nir/nir_opt_algebraic.py | 8
>> 1 file changed, 8 insertions(+)
>>
>> diff --git
On Fri, Mar 27, 2015 at 06:46:43PM -0400, Jan Vesely wrote:
> Signed-off-by: Jan Vesely
Reviewed-by: Tom Stellard
> ---
> src/gallium/auxiliary/gallivm/lp_bld_debug.cpp | 4
> 1 file changed, 4 insertions(+)
>
> diff --git a/src/gallium/auxiliary/gallivm/lp_bld_debug.cpp
> b/src/gallium/a
Signed-off-by: Jan Vesely
---
src/gallium/auxiliary/gallivm/lp_bld_debug.cpp | 4
1 file changed, 4 insertions(+)
diff --git a/src/gallium/auxiliary/gallivm/lp_bld_debug.cpp
b/src/gallium/auxiliary/gallivm/lp_bld_debug.cpp
index bf6268b..d4d453d 100644
--- a/src/gallium/auxiliary/gallivm/l
On Saturday, March 28, 2015 12:16:25 AM Timothy Arceri wrote:
> ---
>
> Piglit tests:
>
> http://lists.freedesktop.org/archives/piglit/2015-March/015369.html
>
> src/glsl/ast_to_hir.cpp | 6 ++
> 1 file changed, 6 insertions(+)
>
> diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_h
On Friday, March 27, 2015 01:54:32 PM Eric Anholt wrote:
> This will be used by the VC4 driver for doing device-independent
> optimization, and hopefully eventually replacing its whole IR. It also
> may be useful to other drivers for the same reason.
>
> v2: Add all of the instructions I was rely
On 03/27/2015 11:56 AM, Thomas Helland wrote:
> This looks good to me, and matches a solution I found
> online the other day for reducing memory allocation overhead.
>
> Reviewed-by: Thomas Helland
>
> Did you happen to profile this? Did you identify other places that
> could need a similar trea
On 03/27/2015 11:33 AM, Kenneth Graunke wrote:
> On Thursday, March 26, 2015 08:52:33 PM Ian Romanick wrote:
>> From: Ian Romanick
>>
>> The CSE algorithm will continuously allocate new ae_entry objects. As
>> each new basic block is exited, all of the previously allocated objects
>> are dumped.
Reviewed-by: Jose Fonseca
On 27/03/15 15:53, srol...@vmware.com wrote:
From: Roland Scheidegger
These functions looked quite complicated, even though what they actually did
was trivial (ever since we dropped swizzled rendering). Also drop lookup of
format block per bytes done for each block,
On Friday, March 27, 2015 04:57:45 PM Ilia Mirkin wrote:
> On Fri, Mar 27, 2015 at 4:54 PM, Eric Anholt wrote:
[snip]
> > +#ifndef PIPELINE_H
> > +#define PIPELINE_H
> > +
> > +/**
> > + * Bitflags for system values.
> > + */
> > +#define SYSTEM_BIT_SAMPLE_ID BITFIELD64_BIT(SYSTEM_VALUE_SAMPLE_ID)
On Friday, March 27, 2015 01:54:30 PM Eric Anholt wrote:
> NIR uses these enums/#defines in nir_variables and associated intrinsics,
> but I want to be able to use them from TGSI->NIR and NIR->TGSI.
> Otherwise, we had to pull in all of mtypes.h.
>
> This doesn't cover all of the enums we might wa
On Fri, Mar 27, 2015 at 4:54 PM, Eric Anholt wrote:
> diff --git a/src/gallium/auxiliary/Makefile.sources
> b/src/gallium/auxiliary/Makefile.sources
> index 09496fa..08e4e4c 100644
> --- a/src/gallium/auxiliary/Makefile.sources
> +++ b/src/gallium/auxiliary/Makefile.sources
> @@ -69,6 +69,7 @@ C_
On Fri, Mar 27, 2015 at 4:54 PM, Eric Anholt wrote:
> NIR uses these enums/#defines in nir_variables and associated intrinsics,
> but I want to be able to use them from TGSI->NIR and NIR->TGSI.
> Otherwise, we had to pull in all of mtypes.h.
>
> This doesn't cover all of the enums we might want fr
One file that actually grubs through gl_shader_programs is left.
---
src/glsl/nir/nir_lower_system_values.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/glsl/nir/nir_lower_system_values.c
b/src/glsl/nir/nir_lower_system_values.c
index a6eec65..75db08d 100644
--- a/src/g
This will be used by the VC4 driver for doing device-independent
optimization, and hopefully eventually replacing its whole IR. It also
may be useful to other drivers for the same reason.
v2: Add all of the instructions I was relying on tgsi_lowering to remove,
and more.
v3: Rebase on SSA rew
NIR uses these enums/#defines in nir_variables and associated intrinsics,
but I want to be able to use them from TGSI->NIR and NIR->TGSI.
Otherwise, we had to pull in all of mtypes.h.
This doesn't cover all of the enums we might want from a shared compiler
core (like varying slots or vert attribs)
Reviewed-by: Mark Janes
Timothy Arceri writes:
> ---
>
> Piglit tests:
>
> http://lists.freedesktop.org/archives/piglit/2015-March/015369.html
>
> src/glsl/ast_to_hir.cpp | 6 ++
> 1 file changed, 6 insertions(+)
>
> diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
> index
On Fri, Mar 27, 2015 at 3:00 PM, Brian Paul wrote:
> On 03/27/2015 12:49 PM, Ilia Mirkin wrote:
>>
>> On Fri, Mar 27, 2015 at 2:44 PM, Brian Paul wrote:
>>>
>>>
>>> The Geeks3d GpuTest GiMark test for Windows has several shaders with
>>> #version 330 that don't actually use any 3.30 features. Ot
Kenneth Graunke writes:
> On Thursday, March 26, 2015 08:52:33 PM Ian Romanick wrote:
>> From: Ian Romanick
>>
>> The CSE algorithm will continuously allocate new ae_entry objects. As
>> each new basic block is exited, all of the previously allocated objects
>> are dumped. Instead, put them i
Matt Turner writes:
> Doesn't work for analogous && cases, because of NaNs.
>
> total instructions in shared programs: 6195712 -> 6194829 (-0.01%)
> instructions in affected programs: 42000 -> 41117 (-2.10%)
> helped:403
> ---
> src/glsl/nir/nir_opt_algebraic.
On 03/27/2015 12:49 PM, Ilia Mirkin wrote:
On Fri, Mar 27, 2015 at 2:44 PM, Brian Paul wrote:
The Geeks3d GpuTest GiMark test for Windows has several shaders with
#version 330 that don't actually use any 3.30 features. Other shaders in
the test use #version 130. Furthermore, when a context i
Matt Turner writes:
> instructions in affected programs: 2858 -> 2808 (-1.75%)
> helped:12
> ---
> src/glsl/nir/nir_opt_algebraic.py | 8
> 1 file changed, 8 insertions(+)
>
> diff --git a/src/glsl/nir/nir_opt_algebraic.py
> b/src/glsl/nir/nir_opt_al
This looks good to me, and matches a solution I found
online the other day for reducing memory allocation overhead.
Reviewed-by: Thomas Helland
Did you happen to profile this? Did you identify other places that
could need a similar treatment?
I'm about to do a last round of micro-adjustments and
On Fri, Mar 27, 2015 at 2:44 PM, Brian Paul wrote:
>
> The Geeks3d GpuTest GiMark test for Windows has several shaders with
> #version 330 that don't actually use any 3.30 features. Other shaders in
> the test use #version 130. Furthermore, when a context is created, only a
> 3.0 context is requ
The Geeks3d GpuTest GiMark test for Windows has several shaders with
#version 330 that don't actually use any 3.30 features. Other shaders
in the test use #version 130. Furthermore, when a context is created,
only a 3.0 context is requested. We don't support GL 3.0 w/ 3.30
shaders but if I
Patches 11-15 are all
Reviewed-by: Ilia Mirkin
but please fix up patch #13 with that small bit I mentioned.
On Sun, Mar 22, 2015 at 11:35 AM, Samuel Pitoiset
wrote:
> Signed-off-by: Samuel Pitoiset
> ---
> src/gallium/drivers/nouveau/nvc0/nvc0_query.c | 12 +++-
> 1 file changed, 7 i
It's really confusing to have MP_COUNTER_GROUP being 0 or 1 depending
on the setting of DEBUG. Please make them the same.
On Sun, Mar 22, 2015 at 11:35 AM, Samuel Pitoiset
wrote:
> This patch exposes "Driver statistics" and "MP counters" groups.
>
> Signed-off-by: Samuel Pitoiset
> ---
> src/ga
On Thursday, March 26, 2015 08:52:33 PM Ian Romanick wrote:
> From: Ian Romanick
>
> The CSE algorithm will continuously allocate new ae_entry objects. As
> each new basic block is exited, all of the previously allocated objects
> are dumped. Instead, put them in a free list and re-use them in
What's the use-case for applying a different version to a shader with
an explicit #version?
On Fri, Mar 27, 2015 at 2:13 PM, Brian Paul wrote:
> Previously, the ctx->Const.ForceGLSLVersion setting only worked if
> the shader lacked a #version directive. Now, the ForceGLSLVersion
> setting will o
Previously, the ctx->Const.ForceGLSLVersion setting only worked if
the shader lacked a #version directive. Now, the ForceGLSLVersion
setting will override the #version directive too.
This change should be safe since it should be rare to have an app
that has a mix of shader versions and we only wa
From: Roland Scheidegger
These functions looked quite complicated, even though what they actually did
was trivial (ever since we dropped swizzled rendering). Also drop lookup of
format block per bytes done for each block, and do it once per scene instead.
This improves everybody's favorite "bench
On Fri, Mar 27, 2015 at 4:19 AM, Martin Peres
wrote:
> On 27/03/15 07:13, Ilia Mirkin wrote:
>>
>> On Fri, Mar 27, 2015 at 1:06 AM, Vinson Lee wrote:
>>>
>>> On Mon, Feb 16, 2015 at 6:14 AM, Martin Peres
>>> wrote:
@@ -1404,14 +1405,36 @@ _mesa_GenRenderbuffers(GLsizei n, GLuint
*
On Fri, Mar 27, 2015 at 4:58 AM, Michel Dänzer wrote:
> From: Flora Cui
>
> Signed-off-by: Flora Cui
> Reviewed-by: Jammy Zhou
> Reviewed-by: Michel Dänzer
> Signed-off-by: Michel Dänzer
Reviewed-by: Alex Deucher
> ---
> src/gbm/backends/dri/gbm_dri.c | 2 ++
> src/gbm/main/gbm.h
On Fri, Mar 27, 2015 at 10:05:06AM +0900, Michel Dänzer wrote:
> From: Michel Dänzer
>
> Fixes a crash in genymotion with several threads compiling shaders
> concurrently.
>
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89746
>
> Cc: 10.5
> Signed-off-by: Michel Dänzer
Reviewed-by
Hi Bernd,
On 26 March 2015 at 22:44, Bernd Kuhls wrote:
> Emil Velikov wrote in news:55136BE1.6090001
> @gmail.com:
>> and no longer requires python/mako in order to build.
>
> Hi,
>
> I am a bit confused, quoting the commit
> http://cgit.freedesktop.org/mesa/mesa/commit/?h=10.5
> &id=e98909b056
---
Piglit tests:
http://lists.freedesktop.org/archives/piglit/2015-March/015369.html
src/glsl/ast_to_hir.cpp | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
index d9ce6c9..49bddc4 100644
--- a/src/glsl/ast_to_hir.cpp
+++ b/src/glsl/
On Mar 26, 2015 6:58 PM, "Matt Turner" wrote:
>
> On Thu, Mar 26, 2015 at 1:35 PM, Jason Ekstrand
wrote:
> > What about other datatypes such as D or UW? Perhaps we should have
> > some sort of is_integer function and just do is_integer(a) !=
> > is_integer(b)?
>
> I can't come up with any case w
On Fri, 2015-03-27 at 18:07 +1100, Timothy Arceri wrote:
> On Thu, 2015-03-26 at 23:10 -0700, Ben Widawsky wrote:
> > On Mon, Dec 01, 2014 at 02:04:50PM +0100, Eduardo Lima Mitev wrote:
> > > From: Samuel Iglesias Gonsalvez
> > >
> > > GLSL ES 3.00 spec, 4.3.10 (Linking of Vertex Outputs and Frag
On 25/03/15 20:12, Emil Velikov wrote:
Plus nuke final note of osmesa from README.WIN32
Signed-off-by: Emil Velikov
---
docs/README.WIN32 | 4
docs/relnotes/10.6.0.html | 2 ++
2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/docs/README.WIN32 b/docs/README.WIN32
From: Flora Cui
Signed-off-by: Flora Cui
Reviewed-by: Jammy Zhou
Reviewed-by: Michel Dänzer
Signed-off-by: Michel Dänzer
---
src/gbm/backends/dri/gbm_dri.c | 2 ++
src/gbm/main/gbm.h | 4
2 files changed, 6 insertions(+)
diff --git a/src/gbm/backends/dri/gbm_dri.c b/src/gbm
On 27/03/15 07:13, Ilia Mirkin wrote:
On Fri, Mar 27, 2015 at 1:06 AM, Vinson Lee wrote:
On Mon, Feb 16, 2015 at 6:14 AM, Martin Peres
wrote:
@@ -1404,14 +1405,36 @@ _mesa_GenRenderbuffers(GLsizei n, GLuint *renderbuffers)
for (i = 0; i < n; i++) {
GLuint name = first + i;
On 2015-03-26 20:52:33, Ian Romanick wrote:
> From: Ian Romanick
>
> The CSE algorithm will continuously allocate new ae_entry objects. As
> each new basic block is exited, all of the previously allocated objects
> are dumped. Instead, put them in a free list and re-use them in the
> next basic
On Thursday 26 March 2015 23:10:53 Ben Widawsky wrote:
> On Mon, Dec 01, 2014 at 02:04:50PM +0100, Eduardo Lima Mitev wrote:
> > From: Samuel Iglesias Gonsalvez
> >
> > GLSL ES 3.00 spec, 4.3.10 (Linking of Vertex Outputs and Fragment Inputs),
> > page 45 says the following:
> >
> > "The type of
On Thu, 2015-03-26 at 23:10 -0700, Ben Widawsky wrote:
> On Mon, Dec 01, 2014 at 02:04:50PM +0100, Eduardo Lima Mitev wrote:
> > From: Samuel Iglesias Gonsalvez
> >
> > GLSL ES 3.00 spec, 4.3.10 (Linking of Vertex Outputs and Fragment Inputs),
> > page 45 says the following:
> >
> > "The type of
51 matches
Mail list logo
