Looks OK to me.
Reviewed-by: Chris Forbes
On Thu, Jan 15, 2015 at 5:50 PM, Jason Ekstrand wrote:
> v2: Add a comment as to why we are using variables instead of lowering to
> an index like the rest of the input intrinsics.
>
> Cc: Chris Forbes
> Cc: Matt Turner
>
> ---
> Sorry for the res
v2: Add a comment as to why we are using variables instead of lowering to
an index like the rest of the input intrinsics.
Cc: Chris Forbes
Cc: Matt Turner
---
Sorry for the resend, but Connor didn't feel as if he could review this.
This is the last patch that needs review before I push.
s
On 01/15/2015 10:29 AM, Matt Turner wrote:
> On Wed, Jan 14, 2015 at 1:05 PM, Ian Romanick wrote:
>> On 01/09/2015 07:59 PM, Matt Turner wrote:
>>> total instructions in shared programs: 5886767 -> 5881642 (-0.09%)
>>> instructions in affected programs: 532199 -> 527074 (-0.96%)
>>> GAINED:
On Wed, Jan 14, 2015 at 7:27 PM, Connor Abbott wrote:
> Reviewed-by: Connor Abbott
>
> Just curious, how did this come about?
>
I was running shader-db and it was segfaulting. I'm not sure why I didn't
notice it before. Perhaps it's because of the new optimizations you asked
me to add? In an
Reviewed-by: Connor Abbott
Just curious, how did this come about?
On Wed, Jan 14, 2015 at 10:15 PM, Jason Ekstrand wrote:
> Without the break, it was possible that an instruction would match multiple
> expressions. If this happened, you could end up trying to replace it
> multiple times and ge
Without the break, it was possible that an instruction would match multiple
expressions. If this happened, you could end up trying to replace it
multiple times and get a segfault. This makes it so that, after a
successful replacement, it moves on to the next instruction.
---
src/glsl/nir/nir_alg
On Thu, Jan 15, 2015 at 3:50 AM, Michel Dänzer wrote:
> On 14.01.2015 20:22, Marek Olšák wrote:
>> On Wed, Jan 14, 2015 at 3:33 AM, Michel Dänzer wrote:
>>> On 13.01.2015 19:11, Marek Olšák wrote:
This is weird. We only enable this optimization if the shader code
doesn't use CENTROID at
On 14.01.2015 20:22, Marek Olšák wrote:
> On Wed, Jan 14, 2015 at 3:33 AM, Michel Dänzer wrote:
>> On 13.01.2015 19:11, Marek Olšák wrote:
>>> This is weird. We only enable this optimization if the shader code
>>> doesn't use CENTROID at all. This is the main flag:
>>> sctx->ps_shader->info.uses_c
On Wed, Jan 14, 2015 at 6:53 PM, Jason Ekstrand wrote:
>
>
> On Sun, Jan 11, 2015 at 7:52 PM, Connor Abbott wrote:
>>
>> Patches without my Reviewed-by on them on your branch are:
>>
>> "i965/fs_nir: Add support for sample_pos and sample_id"
>>
>> This one gets rewritten later on so I think it's
On Wed, Jan 14, 2015 at 8:54 AM, Kenneth Graunke wrote:
> In the common case, this is redundant - we set it a few lines later.
>
> However, when irb->mt->fast_clear_state == INTEL_FAST_CLEAR_STATE_CLEAR,
> this causes irb->need_downsample to be set, even though we skip altering
> the buffer at all
On Wed, Jan 14, 2015 at 1:52 PM, Matt Turner wrote:
> On Wed, Jan 14, 2015 at 1:29 PM, Matt Turner wrote:
>> glsl: Optimize certain if-statements to just casts from the condition
>
> Cherry-picked to master, the shader-db results are
>
> total instructions in shared programs: 5965630 -> 5952789 (
In the common case, this is redundant - we set it a few lines later.
However, when irb->mt->fast_clear_state == INTEL_FAST_CLEAR_STATE_CLEAR,
this causes irb->need_downsample to be set, even though we skip altering
the buffer at all. We don't want to do that.
Signed-off-by: Kenneth Graunke
---
On Tuesday, January 13, 2015 11:37:47 PM Ben Widawsky wrote:
> An upcoming patch is going to introduce some code here, and having this code
> organized as the patch does makes it a bit easier to read later.
>
> There should be no functional change here.
>
> Signed-off-by: Ben Widawsky
> ---
> s
Reviewed-by: Matt Turner
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
On Sun, Jan 11, 2015 at 7:52 PM, Connor Abbott wrote:
> Patches without my Reviewed-by on them on your branch are:
>
> "i965/fs_nir: Add support for sample_pos and sample_id"
>
> This one gets rewritten later on so I think it's ok to not review it.
>
> "i965/fs: Allow reinterpretation in constant
Together with the patches to lower all wildcard copies,
Reviewed-by: Connor Abbott
On Tue, Dec 16, 2014 at 1:11 AM, Jason Ekstrand wrote:
> ---
> src/glsl/Makefile.sources | 1 +
> src/glsl/nir/nir.h | 2 +
> src/glsl/nir/nir_lower_locals_to_regs.c | 313
This, patch 157, and patch 158 are
Reviewed-by: Connor Abbott
On Wed, Jan 14, 2015 at 6:28 PM, Jason Ekstrand wrote:
> ---
> src/glsl/Makefile.sources | 1 +
> src/glsl/nir/nir.h | 3 +
> src/glsl/nir/nir_lower_var_copies.c | 223
> ++
On Wed, Jan 14, 2015 at 2:34 PM, Connor Abbott wrote:
> On Wed, Jan 14, 2015 at 5:15 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Sun, Jan 11, 2015 at 7:08 PM, Connor Abbott
> wrote:
> >>
> >> On Tue, Jan 6, 2015 at 5:21 PM, Jason Ekstrand
> >> wrote:
> >> >
> >> >
> >> > On Mon, Jan 5, 2015 at
---
src/glsl/nir/nir_lower_vars_to_ssa.c | 187 +--
1 file changed, 46 insertions(+), 141 deletions(-)
diff --git a/src/glsl/nir/nir_lower_vars_to_ssa.c
b/src/glsl/nir/nir_lower_vars_to_ssa.c
index 222ddcd..7fe10e9 100644
--- a/src/glsl/nir/nir_lower_vars_to_ssa.c
---
src/mesa/drivers/dri/i965/brw_fs_nir.cpp | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
b/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
index 218cd5c..e8f398a 100644
--- a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
+++ b/src/mesa/drivers/dri/i965/brw_
---
src/glsl/Makefile.sources | 1 +
src/glsl/nir/nir.h | 3 +
src/glsl/nir/nir_lower_var_copies.c | 223
3 files changed, 227 insertions(+)
create mode 100644 src/glsl/nir/nir_lower_var_copies.c
diff --git a/src/glsl/Makefile.s
On Wed, Jan 14, 2015 at 5:15 PM, Jason Ekstrand wrote:
>
>
> On Sun, Jan 11, 2015 at 7:08 PM, Connor Abbott wrote:
>>
>> On Tue, Jan 6, 2015 at 5:21 PM, Jason Ekstrand
>> wrote:
>> >
>> >
>> > On Mon, Jan 5, 2015 at 10:00 PM, Jason Ekstrand
>> > wrote:
>> >>
>> >>
>> >>
>> >> On Mon, Jan 5, 201
On Wed, Jan 14, 2015 at 12:12 PM, wrote:
> From: Mike Mason
>
> Changes the initial internal format of a render buffer
> to GL_RGBA4 in GLES 3. This fixes a failure in the following
> DrawElements test:
>
> dEQP-GLES3.functional.state_query.rbo.renderbuffer_internal_format
> ---
> src/mesa/main
On Sun, Jan 11, 2015 at 7:08 PM, Connor Abbott wrote:
> On Tue, Jan 6, 2015 at 5:21 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Mon, Jan 5, 2015 at 10:00 PM, Jason Ekstrand
> > wrote:
> >>
> >>
> >>
> >> On Mon, Jan 5, 2015 at 9:12 PM, Connor Abbott
> wrote:
> >>>
> >>> Hi,
> >>>
> >>> Was it y
On Wed, Jan 14, 2015 at 1:10 PM, Connor Abbott wrote:
> On Fri, Jan 9, 2015 at 8:27 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Fri, Jan 9, 2015 at 4:38 PM, Connor Abbott
> wrote:
> >>
> >> + case nir_intrinsic_copy_var:
> >> + unreachable("There should be no copies whatso
On Wed, Jan 14, 2015 at 5:05 PM, Jason Ekstrand wrote:
>
>
> On Wed, Jan 14, 2015 at 1:05 PM, Connor Abbott wrote:
>>
>> On Wed, Jan 14, 2015 at 3:36 PM, Jason Ekstrand
>> wrote:
>> >
>> >
>> > On Fri, Jan 9, 2015 at 11:31 AM, Connor Abbott
>> > wrote:
>> >>
>> >> On Tue, Jan 6, 2015 at 7:34 PM
Reviewed-by: Connor Abbott
On Wed, Jan 14, 2015 at 5:02 PM, Jason Ekstrand wrote:
> This refactor allows you to more easily get the deref node associated with
> a given variable. We then use that new functionality in the
> deref_may_be_aliased function instead of creating a 1-element deref chai
On Wed, 2015-01-14 at 22:12 +0100, David Heidelberg wrote:
> On 01/14/2015 09:53 PM, Jan Vesely wrote:
> > Signed-off-by: Jan Vesely
> > ---
> > src/mesa/main/bufferobj.c | 18 +-
> > src/mesa/main/buffers.c | 2 +-
> > src/mesa/main/clear.c
On Wed, Jan 14, 2015 at 1:05 PM, Connor Abbott wrote:
> On Wed, Jan 14, 2015 at 3:36 PM, Jason Ekstrand
> wrote:
> >
> >
> > On Fri, Jan 9, 2015 at 11:31 AM, Connor Abbott
> wrote:
> >>
> >> On Tue, Jan 6, 2015 at 7:34 PM, Jason Ekstrand
> >> wrote:
> >> > Additional description was added to a
On Fri, Jan 9, 2015 at 10:58 AM, Connor Abbott wrote:
> On Sun, Jan 4, 2015 at 4:01 PM, Connor Abbott wrote:
> [...]
> >> static bool
> >> deref_may_be_aliased(nir_deref_var *deref,
> >> struct lower_variables_state *state)
> >> {
> >>nir_deref_var var_deref = *deref;
>
This refactor allows you to more easily get the deref node associated with
a given variable. We then use that new functionality in the
deref_may_be_aliased function instead of creating a 1-element deref chain.
---
src/glsl/nir/nir_lower_vars_to_ssa.c | 48 +++-
1 f
On Wed, Jan 14, 2015 at 1:29 PM, Matt Turner wrote:
> glsl: Optimize certain if-statements to just casts from the condition
Cherry-picked to master, the shader-db results are
total instructions in shared programs: 5965630 -> 5952789 (-0.22%)
instructions in affected programs: 737228 -> 72438
With this squashed in, patch 90 is
Reviewed-by: Connor Abbott
On Wed, Jan 14, 2015 at 2:57 PM, Jason Ekstrand wrote:
> ---
> src/glsl/nir/nir_split_var_copies.c | 96
> ++---
> 1 file changed, 78 insertions(+), 18 deletions(-)
>
> diff --git a/src/glsl/nir/nir_
Reviewed-by: Connor Abbott
On Wed, Jan 14, 2015 at 3:43 PM, Jason Ekstrand wrote:
> The original name wasn't particularly descriptive. This one indicates that
> it actually gives you SSA values as opposed to the old pass which lowered
> variables to registers.
> ---
> src/glsl/Makefile.sources
On Wed, Jan 14, 2015 at 1:05 PM, Ian Romanick wrote:
> On 01/09/2015 07:59 PM, Matt Turner wrote:
>> total instructions in shared programs: 5886767 -> 5881642 (-0.09%)
>> instructions in affected programs: 532199 -> 527074 (-0.96%)
>> GAINED:9
>> LOST:
On 01/14/2015 09:53 PM, Jan Vesely wrote:
Signed-off-by: Jan Vesely
---
src/mesa/main/bufferobj.c | 18 +-
src/mesa/main/buffers.c | 2 +-
src/mesa/main/clear.c | 2 +-
src/mesa/main/dlist.c | 2 +-
The below code crashes when vector_elements <= 0
Fixes Warray-bounds warnings
Signed-off-by: Jan Vesely
---
src/mesa/program/ir_to_mesa.cpp| 1 +
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 1 +
2 files changed, 2 insertions(+)
diff --git a/src/mesa/program/ir_to_mesa.cpp b/src/mes
On Fri, Jan 9, 2015 at 8:27 PM, Jason Ekstrand wrote:
>
>
> On Fri, Jan 9, 2015 at 4:38 PM, Connor Abbott wrote:
>>
>> + case nir_intrinsic_copy_var:
>> + unreachable("There should be no copies whatsoever at this
>> point");
>> + break;
>> >>>
>> >>>
>> >>
On Wed, Jan 14, 2015 at 3:36 PM, Jason Ekstrand wrote:
>
>
> On Fri, Jan 9, 2015 at 11:31 AM, Connor Abbott wrote:
>>
>> On Tue, Jan 6, 2015 at 7:34 PM, Jason Ekstrand
>> wrote:
>> > Additional description was added to a variety of places. Also, we no
>> > longer use the term "leaf" to describe
On 01/09/2015 07:59 PM, Matt Turner wrote:
> total instructions in shared programs: 5886767 -> 5881642 (-0.09%)
> instructions in affected programs: 532199 -> 527074 (-0.96%)
> GAINED:9
> LOST: 6
I'll did out the rest of the data
these two patches fix cca 90 warnings when compiling src/mesa.
The idea is to reduce warning bloat so that the remaining stuff
can be looked into to check whether the warning hints at bigger issue
gcc 4.9.2 with Wextra:
from
2 -Warray-bounds
29 -Wmaybe-uninitialized
38 -Wmissing-fi
Signed-off-by: Jan Vesely
---
src/mesa/main/bufferobj.c | 18 +-
src/mesa/main/buffers.c | 2 +-
src/mesa/main/clear.c | 2 +-
src/mesa/main/dlist.c | 2 +-
src/mesa/main/errors.c| 8
Signed-off-by: Jan Vesely
---
src/mesa/program/prog_execute.c | 1 -
src/mesa/program/prog_statevars.c | 2 --
2 files changed, 3 deletions(-)
diff --git a/src/mesa/program/prog_execute.c b/src/mesa/program/prog_execute.c
index 33c1751..85aba00 100644
--- a/src/mesa/program/prog_execute.c
+++
On 01/12/2015 11:09 PM, Iago Toral wrote:
> Hi,
>
> Mesa fails the following dEQP tests:
>
> dEQP-GLES3.functional.shaders.linkage.varying.rules.invalid_type_struct_array
> dEQP-GLES3.functional.shaders.linkage.varying.rules.invalid_type_struct_struct
> dEQP-GLES3.functional.shaders.linkage.varyi
The original name wasn't particularly descriptive. This one indicates that
it actually gives you SSA values as opposed to the old pass which lowered
variables to registers.
---
src/glsl/Makefile.sources|2 +-
src/glsl/nir/nir.h |2 +-
src/glsl/nir/nir
On Fri, Jan 9, 2015 at 11:31 AM, Connor Abbott wrote:
> On Tue, Jan 6, 2015 at 7:34 PM, Jason Ekstrand
> wrote:
> > Additional description was added to a variety of places. Also, we no
> > longer use the term "leaf" to describe fully-qualified direct derefs.
> > Instead, we simply use the term
This looks fine to me. We should probably also do this for snorm formats.
I don't care if that's part of this or in a separate patch.
--Jason
Reviewed-by: Jason Ekstrand
On Wed, Jan 14, 2015 at 11:46 AM, Neil Roberts wrote:
> When converting to a format that has fewer bits the previous code w
From: Mike Mason
Changes the initial internal format of a render buffer
to GL_RGBA4 in GLES 3. This fixes a failure in the following
DrawElements test:
dEQP-GLES3.functional.state_query.rbo.renderbuffer_internal_format
---
src/mesa/main/renderbuffer.c | 19 ++-
1 file changed, 1
---
src/glsl/nir/nir_split_var_copies.c | 96 ++---
1 file changed, 78 insertions(+), 18 deletions(-)
diff --git a/src/glsl/nir/nir_split_var_copies.c
b/src/glsl/nir/nir_split_var_copies.c
index 70ce30a..4d663b5 100644
--- a/src/glsl/nir/nir_split_var_copies.c
+++
When converting to a format that has fewer bits the previous code was just
shifting off the bits. This doesn't provide very accurate results. For example
when converting from 8 bits to 5 bits it is equivalent to doing this:
x * 32 / 256
This works as if it's taking a value from a range where 256
To fix build when libdrm is not found,
commit a594cec7e3ef275c386054127a357110a19dd823 did put several
parts of egl code under #ifdef HAVE_DRM_PLATFORM.
HAVE_DRM_PLATFORM means the egl drm platform is being built.
What should have been used instead is HAVE_LIBDRM.
At a few locations, the HAVE_DRM
First two are
Reviewed-by: Jason Ekstrand
On Tue, Jan 13, 2015 at 11:37 PM, Ben Widawsky
wrote:
> Blits to or from a y-tiled surface must always be a multiple of the tile
> size.
> From page 16 of the HSW PRM
> (
> https://01.org/linuxgraphics/sites/default/files/documentation/intel-gfx-prm-osr
On Tue, Jan 13, 2015 at 11:37 PM, Ben Widawsky
wrote:
> The blit engine is limited to 32Kx32K transfer. In cases where we have to
> fall
> back to the blitter, and when trying to blit a slice of a 2d texture
> array, or
> face of a cube map, we don't need to transfer the entire texture.
>
> I dou
On Tue, Jan 13, 2015 at 11:37 PM, Ben Widawsky
wrote:
> This patch will use a new calculation to determine if a surface can be
> blitted
> from or to. Previously, the "total_height" member was used. Total_height
> in the
> case of 2d, 3d, and cube map arrays is the height of each slice/layer/face
On Tue, Jan 13, 2015 at 8:21 PM, Kenneth Graunke
wrote:
> On Tuesday, January 13, 2015 08:02:54 PM Jason Ekstrand wrote:
> > On Jan 13, 2015 6:38 PM, "Kenneth Graunke"
> wrote:
> > >
> > > On Tuesday, January 13, 2015 10:28:20 AM Jason Ekstrand wrote:
> > > > Going through the for loop every tim
Neil Roberts writes:
> + assert(src_bits + dst_bits <= sizeof(x) * 8);
Erm, actually I didn't realise there were places calling this with
dst_bits set to 32, so this isn't going to work. I probably should have
waited for Piglit to finish before sending the patch. Sorry.
- Neil
When converting to a format that has fewer bits the previous code was just
shifting off the bits. This doesn't provide very accurate results. For example
when converting from 8 bits to 5 bits it is equivalent to doing this:
x * 32 / 256
This works as if it's taking a value from a range where 256
Hello,
Mesa fails the following dEQP GLES 3 test:
dEQP-GLES3.functional.fbo.api.attachment_query_default_fbo
This test calls glGetFramebufferAttachmentParameteriv() to get
FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE information from COLOR_ATTACHMENT0.
with the default framebuffer bound to target.
Then
El 2015-01-14 16:46, Tapani escribió:
On 01/14/2015 03:16 PM, Iago Toral Quiroga wrote:
Hi Tapani,
thanks for looking into this one, I certainly missed that extension...
On mié, 2015-01-14 at 14:04 +0200, Tapani Pälli wrote:
Commit 8ec6534 changed texture upload path and the way how texture
f
On 01/14/2015 03:16 PM, Iago Toral Quiroga wrote:
Hi Tapani,
thanks for looking into this one, I certainly missed that extension...
On mié, 2015-01-14 at 14:04 +0200, Tapani Pälli wrote:
Commit 8ec6534 changed texture upload path and the way how texture
format is being checked, this commit add
On Jan 14, 2015 5:16 AM, "Iago Toral Quiroga" wrote:
>
> Hi Tapani,
>
> thanks for looking into this one, I certainly missed that extension...
>
> On mié, 2015-01-14 at 14:04 +0200, Tapani Pälli wrote:
> > Commit 8ec6534 changed texture upload path and the way how texture
> > format is being check
On 01/13/2015 06:09 PM, Chad Versace wrote:
Brian and Ian, me and Michael have a Mesa TLS question at the message's bottom.
On 01/13/2015 04:41 PM, Mason, Michael W wrote:
-Original Message-
From: Versace, Chad
Sent: Tuesday, January 13, 2015 3:42 PM
To: Mason, Michael W; mesa-dev@list
Hi Tapani,
thanks for looking into this one, I certainly missed that extension...
On mié, 2015-01-14 at 14:04 +0200, Tapani Pälli wrote:
> Commit 8ec6534 changed texture upload path and the way how texture
> format is being checked, this commit adds support for GL_RGB with
> GL_UNSIGNED_INT_2_10_
https://bugs.freedesktop.org/show_bug.cgi?id=88335
Samuel Iglesias changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
Commit 8ec6534 changed texture upload path and the way how texture
format is being checked, this commit adds support for GL_RGB with
GL_UNSIGNED_INT_2_10_10_10_REV as specified by the extension
EXT_texture_type_2_10_10_10_REV specification.
This fixes regression in ES3 conformance test
ES3-CTS.
I don't know. I couldn't find any useful info about it either.
Marek
On Wed, Jan 14, 2015 at 3:33 AM, Michel Dänzer wrote:
> On 13.01.2015 19:11, Marek Olšák wrote:
>> This is weird. We only enable this optimization if the shader code
>> doesn't use CENTROID at all. This is the main flag:
>> sct
On Wednesday, January 14, 2015 04:27:30 PM Timothy Arceri wrote:
> Hi all,
>
> Not sure how many of you have come across the webpage "Major Linux
> Problems on the Desktop 2015 edition" [1]. Accurate or not it seems to
> be getting a bit of attention. Anyway Mesa gets a bit of a bashing in
> there
On Tue, Jan 13, 2015 at 11:37:50PM -0800, Ben Widawsky wrote:
> This patch will use a new calculation to determine if a surface can be blitted
> from or to. Previously, the "total_height" member was used. Total_height in
> the
> case of 2d, 3d, and cube map arrays is the height of each slice/layer
68 matches
Mail list logo
