On Fri, Feb 6, 2015 at 4:21 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> This lowers double dot product and lrp to fma.
>>
>> Signed-off-by: Dave Airlie
>> ---
>> src/glsl/lower_instructions.cpp | 83
>>
On Fri, Feb 6, 2015 at 4:15 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Tapani Pälli
>>
>> Patch fixes Piglit test:
>>arb_gpu_shader_fp64/preprocessor/fs-output-double.frag
>>
>> and adds additional validation for shader outputs.
>>
>> Signed-off-by: Tapani P
On Fri, Feb 6, 2015 at 3:58 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> Signed-off-by: Dave Airlie
>> ---
>> src/glsl/loop_controls.cpp | 16
>> 1 file changed, 12 insertions(+), 4 deletions(-)
>>
>> diff --git a/src/glsl/loop
On Fri, Feb 6, 2015 at 3:40 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> Signed-off-by: Dave Airlie
>> ---
>> src/glsl/glsl_lexer.ll | 29 +
>> 1 file changed, 25 insertions(+), 4 deletions(-)
>>
>> diff --git a/src/
On Fri, Feb 6, 2015 at 3:32 AM, Ian Romanick wrote:
> Three comments near the middle / bottom...
>
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> Signed-off-by: Dave Airlie
>> ---
>> src/glsl/ast.h | 2 ++
>> src/glsl/ast_function.cpp | 67
>> ++
On Fri, Feb 6, 2015 at 3:08 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> Signed-off-by: Dave Airlie
>> Reviewed-by: Matt Turner
>> ---
>> src/glsl/lower_ubo_reference.cpp | 12 +---
>> 1 file changed, 9 insertions(+), 3 deletions(-)
>>
Hm... yeah. Looks like ir_binop_dot could use some help. I think
I'm just going to guard it with a if (!float) for now. Most of the
code is pretty good about checking it.
On Fri, Feb 6, 2015 at 3:05 AM, Ian Romanick wrote:
> Does opt_algebraic also need updates?
>
> On 02/06/2015 06:56 AM, Il
On Fri, Feb 6, 2015 at 3:02 AM, Ian Romanick wrote:
>> @@ -708,6 +763,9 @@ ir_expression::constant_expression_value(struct
>> hash_table *variable_context)
>>case GLSL_TYPE_FLOAT:
>> data.f[c] = op[0]->value.f[c] - floor(op[0]->value.f[c]);
>> break;
>> + case GLS
On Fri, Feb 6, 2015 at 3:02 AM, Ian Romanick wrote:
>> + case ir_unop_pack_double_2x32: {
>> + /* XXX needs to be checked on big-endian */
>
> I think this should just work. The spec says that the 32-bit values are
> little-endian:
>
> "The first vector component specifies the 32 least
On Sat, Feb 7, 2015 at 8:25 PM, Matt Turner wrote:
> On Sat, Feb 7, 2015 at 5:16 PM, Ilia Mirkin wrote:
>> On Sat, Feb 7, 2015 at 8:10 PM, Ilia Mirkin wrote:
> diff --git a/src/glsl/ir.h b/src/glsl/ir.h
> index a0f48b2..6e7c654 100644
> --- a/src/glsl/ir.h
> +++ b/src/glsl/ir.h
>
On Sat, Feb 7, 2015 at 5:16 PM, Ilia Mirkin wrote:
> On Sat, Feb 7, 2015 at 8:10 PM, Ilia Mirkin wrote:
diff --git a/src/glsl/ir.h b/src/glsl/ir.h
index a0f48b2..6e7c654 100644
--- a/src/glsl/ir.h
+++ b/src/glsl/ir.h
@@ -1275,6 +1275,13 @@ enum ir_expression_operation {
>
On Sat, Feb 7, 2015 at 8:10 PM, Ilia Mirkin wrote:
>>> diff --git a/src/glsl/ir.h b/src/glsl/ir.h
>>> index a0f48b2..6e7c654 100644
>>> --- a/src/glsl/ir.h
>>> +++ b/src/glsl/ir.h
>>> @@ -1275,6 +1275,13 @@ enum ir_expression_operation {
>>> ir_unop_bitcast_f2u, /**< Bit-identical float-to-uin
On Fri, Feb 6, 2015 at 2:47 AM, Ian Romanick wrote:
> There's a bunch of tabs that should be replaced with spaces.
S... this whole mixed tab/space situation is quite annoying. A lot
of lines come from copying other stuff, and having to fix up tabs on
those can be quite a bother. I'd be happy
These will optimize any tree of a commutative and associative operation
that contains more than one constant to have only one constant in the
end. For example, it will optimize
((4 * a) * b) * (5 * c)
into
20 * ((a * b) * c)
Note that when we port Matt's tree balancing pass, we'll have to think
This is useful since later rules can recognize the constant without
needing to go through a constant folding pass. This is especially
important for the upcoming constant reassociation rules, since without
this it would take many runs of the optimization loop, one rule at a
time, to optimize one exp
These are similar to normal variables. They let us match trees with
arbitrary opcodes, as long as the opcodes match and possibly only if
they have certain algebraic properties, like being associative or
commutatative. The syntax is '!name' (so that it uses a different
character from # and @ which w
This series extends nir_algebraic/nir_search to support matching on arbitrary
opcodes, and then uses it to implement constant reassociation similar to what
we already do in opt_algebraic. I haven't piglit tested this yet, although I
did verify that it did the right thing in a few simple tests. Note
https://bugs.freedesktop.org/show_bug.cgi?id=89026
Chris Forbes changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
Apologies -- I'd not noticed that there was an equivalent patch
already attached to the bug report.
In that case, James Legg's patch is:
Reviewed-by: Chris Forbes
(and I'll push it shortly)
On Sun, Feb 8, 2015 at 1:47 PM, Chris Forbes wrote:
> Renderbuffers are never layered. Previously we'd
On Fri, Feb 6, 2015 at 2:38 AM, Ian Romanick wrote:
> On 02/06/2015 06:56 AM, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> Signed-off-by: Dave Airlie
>> ---
>> src/glsl/link_uniforms.cpp | 7 ++-
>> 1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/src/glsl/link_uniforms.cp
Renderbuffers are never layered. Previously we'd just leave whatever
value was present from the previous attachment.
Signed-off-by: Chris Forbes
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89026
---
src/mesa/main/fbobject.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/mesa/
On Fri, Feb 6, 2015 at 3:29 AM, Pohjolainen, Topi
wrote:
> On Fri, Feb 06, 2015 at 10:04:13AM +0200, Pohjolainen, Topi wrote:
>> On Thu, Feb 05, 2015 at 11:56:28PM -0500, Ilia Mirkin wrote:
>> > From: Dave Airlie
>> >
>> > This causes a lot of warnings about unchecked type in
>> > switch statemen
On Fri, Feb 6, 2015 at 3:04 AM, Pohjolainen, Topi
wrote:
> On Thu, Feb 05, 2015 at 11:56:28PM -0500, Ilia Mirkin wrote:
>> From: Dave Airlie
>>
>> This causes a lot of warnings about unchecked type in
>> switch statements - fix them later.
>
> The rest of the series fixes things in the compiler f
https://bugs.freedesktop.org/show_bug.cgi?id=89026
--- Comment #2 from James Legg ---
Created attachment 113257
--> https://bugs.freedesktop.org/attachment.cgi?id=113257&action=edit
Apitrace trace from reproducer
--
You are receiving this mail because:
You are the QA Contact for the bug.
You
https://bugs.freedesktop.org/show_bug.cgi?id=89026
--- Comment #1 from James Legg ---
Created attachment 113256
--> https://bugs.freedesktop.org/attachment.cgi?id=113256&action=edit
Reproducer
--
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for
https://bugs.freedesktop.org/show_bug.cgi?id=89026
Bug ID: 89026
Summary: Renderbuffer layered state used for framebuffer
completeness test
Product: Mesa
Version: git
Hardware: Other
OS: All
Stat
I think we decided not to support unreleased LLVM builds on stable releases.
This is because building without errors is not enough -- there are often
other changes that need to go with this. Furthermore it's often a moving
target.
In short, if you want to use bleeding edge LLVM, you must use
Reviewed-by: Chris Forbes
On Sat, Feb 7, 2015 at 1:32 AM, Kenneth Graunke wrote:
> The hardware's integer luminance formats are completely unusable;
> currently we fall back to RGBA. This means we need to override
> the texture swizzle to obtain the XXX1 values expected for luminance
> formats.
On 07/02/15 00:10, Matt Turner wrote:
On Fri, Feb 6, 2015 at 3:58 PM, Emil Velikov wrote:
"util" is meant to be for shared utility across the entire code base -
both Mesa and Gallium. It's been growing slowly as people move things
there. It might make sense to move a lot of src/gallium/auxili
On Sat, Feb 7, 2015 at 6:16 AM, Eric Anholt wrote:
> No change on shader-db on i965.
> ---
> src/glsl/nir/nir_opt_algebraic.py | 9 +
> 1 file changed, 9 insertions(+)
>
> diff --git a/src/glsl/nir/nir_opt_algebraic.py
> b/src/glsl/nir/nir_opt_algebraic.py
> index a5fe19a..0512a8f 100644
On Fri, Feb 6, 2015 at 1:11 PM, Iago Toral wrote:
> Hi,
>
> Eduardo and I have been looking into a few dEQP test failures that deal
> with wide line rendering. There are a few of them that fail because of
> how clipping is implemented for this case.
>
> The problem in these cases seems to be that
I have pushed the series. Thanks.
Marek
On Sat, Feb 7, 2015 at 9:53 AM, Park, Jeongmin wrote:
> Since only pp_jimenezmlaa uses depth buffer, we can make it optional.
> ---
> src/gallium/state_trackers/dri/dri_drawable.c | 2 +-
> src/gallium/state_trackers/dri/drisw.c| 2 +-
> 2 files c
https://bugs.freedesktop.org/show_bug.cgi?id=69101
Russ Dill changed:
What|Removed |Added
CC||russ.d...@gmail.com
--
You are receiving th
I'm not sure if I fully understand, but it seems the problem is that
the start/index_bias is copied to SQ_VTX_BASE_VTX and not
VGT_INDX_OFFSET. I guess you can avoid using BASE_VTX by setting
fetch_type=NO_INDEX_OFFSET, right? If yes, then there is a simple
solution: The COPY_DW packet. It can be u
---
src/gallium/state_trackers/osmesa/osmesa.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/gallium/state_trackers/osmesa/osmesa.c
b/src/gallium/state_trackers/osmesa/osmesa.c
index 4e2b2e0..b634d55 100644
--- a/src/gallium/state_trackers/osmesa/osmesa.c
+++ b/src/ga
Since only pp_jimenezmlaa uses depth buffer, we can make it optional.
---
src/gallium/state_trackers/dri/dri_drawable.c | 2 +-
src/gallium/state_trackers/dri/drisw.c| 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/gallium/state_trackers/dri/dri_drawable.c
b/src/
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88962
---
src/gallium/auxiliary/postprocess/pp_mlaa.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/gallium/auxiliary/postprocess/pp_mlaa.c
b/src/gallium/auxiliary/postprocess/pp_mlaa.c
index 9827b8c..147d14d 100644
--- a/src/gal
37 matches
Mail list logo
