This series lives on my gles3-glsl-packing branch.
I tested this series against my piglit gles3-glsl-packing branch on Ivybridge
and Sandybridge. No regressions relative to the Mesa gles3-d72ba278e.
v2:
- Do all of idr's minor fixes.
- Improve the rounding bias in _mesa_float_to_half.
- Fo
s/num_operands()/get_num_operands()/
Discovered because Eclipse failed to resolve the false reference.
Signed-off-by: Chad Versace
---
src/glsl/ir.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/glsl/ir.h b/src/glsl/ir.h
index 85fc5ce..c9c91e8 100644
--- a/src/glsl/ir.
For each of the following functions, add a declaration to
builtins/profiles/300es.glsl and create new file
builtins/ir/${funcname}.ir:
packSnorm2x16 unpackSnorm2x16
packUnorm2x16 unpackUnorm2x16
packHalf2x16 unpackHalf2x16
Signed-off-by: Chad Versace
---
src/glsl/builtins/ir/packHalf2
For each function {pack,unpack}{Snorm,Unorm,Half}2x16, add a corresponding
opcode to enum ir_expression_operation. Validate the new opcodes in
ir_validate.cpp.
Also, add opcodes for scalarized variants of the Half2x16 functions. (The
code generator for the i965 fragment shader requires that all
Move round_to_even's definition to mesa/main so that _mesa_float_to_half()
can use it in order to eliminate rounding bias.
In additon to moving the fuction definition, prefix its name with "_mesa",
just as all other functions in mesa/main are prefixed.
Signed-off-by: Chad Versace
---
src/glsl/M
A subsequent patch will add mesa/main/imports.c as a dependency to the
compiler, which in turn requires that _mesa_warning() be defined.
The real definition of _mesa_warning() is in mesa/main/errors.c, but to
pull that file into the standalone scaffolding would require transitively
pulling in the
Not all float32 values can be exactly represented as a float16.
_mesa_float_to_half() rounded such intermediate float32 values to zero by
truncating unrepresentable bits in the mantissa. This behavior is bia
This patch improves _mesa_float_to_half() by rounding intermediate float32
values to the n
That is, evaluate constant expressions of the following functions:
packSnorm2x16 unpackSnorm2x16
packUnorm2x16 unpackUnorm2x16
packHalf2x16 unpackHalf2x16
v2: Reuse _mesa_pack_float_to_half and its inverse to evaluate
pack/unpackHalf2x16 [for idr].
Signed-off-by: Chad Versace
---
This eliminates unexpected behavior due to unitialized values.
Signed-off-by: Chad Versace
---
src/glsl/ir_builder.h | 7 +++
1 file changed, 7 insertions(+)
diff --git a/src/glsl/ir_builder.h b/src/glsl/ir_builder.h
index 067858d..278294f 100644
--- a/src/glsl/ir_builder.h
+++ b/src/glsl/i
Add the following functions, each of which construct the similarly named
ir expression:
div, round_even, clamp
equal, less, greater, lequal, gequal
logic_not, logic_and, logic_or
bit_not, bit_or, bit_and, lshift, rshift
f2i, i2f, f2u, u2f, i2u, u2i
Signed-off-by: Chad Versa
Add method ir_factory::constant. This little method constructs an
ir_constant using the factory's mem_ctx.
Signed-off-by: Chad Versace
---
src/glsl/ir_builder.h | 24
1 file changed, 24 insertions(+)
diff --git a/src/glsl/ir_builder.h b/src/glsl/ir_builder.h
index 01c7
Using this enum improves the readibility of calls to assign(), whose third
argument is a writemask.
Signed-off-by: Chad Versace
---
src/glsl/ir_builder.h | 9 +
1 file changed, 9 insertions(+)
diff --git a/src/glsl/ir_builder.h b/src/glsl/ir_builder.h
index f3aa0d7..82e3762 100644
--- a
Add two overloaded variants of
ir_if *if_tree()
The new functions allow one to chain together if-trees within a single C++
expression that resembles a real if-statement.
Signed-off-by: Chad Versace
---
src/glsl/ir_builder.cpp | 29 +
src/glsl/ir_builder.h | 6
Replace tabs with spaces. According to docs/devinfo.html, Mesa's
indetation style is:
indent -br -i3 -npcs --no-tabs infile.c -o outfile.c
This patch prevents whitespace weirdness in the next patch.
Signed-off-by: Chad Versace
---
src/glsl/ir.cpp | 4 ++--
1 file changed, 2 insertions(+), 2 d
In ir_expression's constructor, the cases for {bit,logic}_{and,or,xor}
failed to handle the case when both operands were vectors.
Note: This is a candidate for the stable branches.
Signed-off-by: Chad Versace
---
src/glsl/ir.cpp | 5 +
1 file changed, 5 insertions(+)
diff --git a/src/glsl/i
Lower them to arithmetic and bit manipulation expressions.
v2:
- Rewrite using ir_builder. [for idr]
- In lowering packHalf2x16, don't truncate subnormal float16 values to zero.
And round to even rather than to zero. [for stereotype441]
CC: Ian Romanick
CC: Paul Berry
Signed-off-by: Ch
On gen < 7, we fully lower all operations to arithmetic and bitwise
operations.
On gen >= 7, we fully lower the Snorm2x16 and Unorm2x16 operations, and
partially lower the Half2x16 operations.
v2:
- Comment that scalarization is needed only for SOA code [for idr].
- Replace switch-statement w
The GLSL ES 3.00 operations packHalf2x16 and unpackHalf2x16 will emit
these opcodes.
- Define the opcodes BRW_OPCODE_{F32TO16,F16TO32}.
- Add the opcodes to the brw_disasm table.
- Define convenience functions brw_{F32TO16,F16TO32}.
Signed-off-by: Chad Versace
---
src/mesa/drivers/dri/i965/brw_
Signed-off-by: Chad Versace
---
src/mesa/drivers/dri/i965/brw_vec4.h | 3 +
src/mesa/drivers/dri/i965/brw_vec4_emit.cpp| 8 ++
src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp | 155 +
3 files changed, 166 insertions(+)
diff --git a/src/mesa/drivers/dri/i9
v2: Remove lewd comment [for idr].
Signed-off-by: Chad Versace
---
src/mesa/drivers/dri/i965/brw_defines.h| 1 +
src/mesa/drivers/dri/i965/brw_fs.h | 7 ++
.../dri/i965/brw_fs_channel_expressions.cpp| 12
src/mesa/drivers/dri/i965/brw_fs_emit.cpp
Signed-off-by: Chad Versace
---
src/mesa/drivers/dri/i965/brw_eu_emit.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/brw_eu_emit.c
b/src/mesa/drivers/dri/i965/brw_eu_emit.c
index fecbff1..b34754a 100644
--- a/src/mesa/drivers/dri/i965/brw_eu_e
Paul Berry writes:
> On 19 January 2013 11:06, Kenneth Graunke wrote:
>> + //intel_renderbuffer_set_needs_downsample(dst_irb);
>>
>
> I'm ok leaving this out--at the moment there's no way this code can be hit
> for a multisampled destination image. However, if we ever get around to
> implemen
Am 18.01.2013 20:29, schrieb Marek Olšák:
On Fri, Jan 18, 2013 at 5:20 PM, Christian König
wrote:
Am 18.01.2013 16:40, schrieb Marek Olšák:
[SNIP]
Having common base classes for all three radeon drivers sounds like a good
idea to me and sharing the texture code between r600g and SI also.
Well
Am 19.01.2013 23:31, schrieb Maarten Lankhorst:
This reduces jitter slightly in a cleaner way, without desynchronizing mplayer2
as badly
when falling behind.
Signed-off-by: Maarten Lankhorst
Reviewed-by: Christian König
---
diff --git a/src/gallium/auxiliary/vl/vl_winsys_dri.c
b/src/gall
From: Christian König
KMS support is out and stable for a couple of years now and
the userspace code has deprecated or abandoned the old UMS interface.
So make the KMS interface the default and deprecate the UMS interface
in the kernel as well.
v2: rebased on alex/drm-next-3.9-wip
Signed-off-b
On 01/20/2013 09:16 PM, Kenneth Graunke wrote:
On 01/20/2013 05:49 PM, Ian Romanick wrote:
From: Ian Romanick
There really isn't any point. There is no resource savings, and we have
to do gymnastics in the driver to make it work.
There are also bad interactions with multisampling and OpenGL
On 21 January 2013 00:55, Eric Anholt wrote:
> Paul Berry writes:
>
> > On 19 January 2013 11:06, Kenneth Graunke wrote:
> >> + //intel_renderbuffer_set_needs_downsample(dst_irb);
> >>
> >
> > I'm ok leaving this out--at the moment there's no way this code can be
> hit
> > for a multisampled
---
src/gallium/state_trackers/dri/common/dri_drawable.c |9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/gallium/state_trackers/dri/common/dri_drawable.c
b/src/gallium/state_trackers/dri/common/dri_drawable.c
index 2ea5478..d817a8c 100644
--- a/src/gallium/stat
ST_FLUSH_FRONT may call driThrottle, which is implemented with dri_flush.
This prevents double flush as well as fence leaks caused by a recursion
in the middle of throttling.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=58839
---
src/gallium/state_trackers/dri/common/dri_drawable.c |
From: Ander Conselvan de Oliveira
Otherwise, we crash when the callback is executed, since the dri2_surf
pointer may point to invalid data.
---
The wayland EGL platform code was heavily changed in master for the
implementation of the EGL_EXT_buffer_age extension, so this bug does
not exist there
From: Michel Dänzer
Signed-off-by: Michel Dänzer
---
lib/Target/R600/SIInstructions.td | 10 ++
1 file changed, 10 insertions(+)
diff --git a/lib/Target/R600/SIInstructions.td
b/lib/Target/R600/SIInstructions.td
index 4164c55..01b61f7 100644
--- a/lib/Target/R600/SIInstructions.td
The resampleRow setup depends on pixelSize. For color buffers,
we don't know the pixelSize until we're in the buffer loop. Move
that code inside the loop.
Fixes: http://bugs.freedesktop.org/show_bug.cgi?id=59541
---
src/mesa/swrast/s_blit.c | 46 +++---
On Mon, 2013-01-21 at 16:35 +0100, Marek Olšák wrote:
> ST_FLUSH_FRONT may call driThrottle, which is implemented with dri_flush.
> This prevents double flush as well as fence leaks caused by a recursion
> in the middle of throttling.
>
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=588
---
src/mesa/swrast/s_blit.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/src/mesa/swrast/s_blit.c b/src/mesa/swrast/s_blit.c
index 605c80a..08ec5e2 100644
--- a/src/mesa/swrast/s_blit.c
+++ b/src/mesa/swrast/s_blit.c
@@ -347,7 +347,7 @@ blit_nearest(struct gl_context
Fixes piglit's fbo-blit-stretch test.
---
src/mesa/swrast/s_blit.c | 23 ---
1 files changed, 8 insertions(+), 15 deletions(-)
diff --git a/src/mesa/swrast/s_blit.c b/src/mesa/swrast/s_blit.c
index 08ec5e2..82fa43f 100644
--- a/src/mesa/swrast/s_blit.c
+++ b/src/mesa/swrast/
This is mostly a re-spin of the patch that I sent out many months ago.
The main change is that when on-line compression is not available, only
a subset of extension strings is advertised. This was based on feedback
from a developer whose application does submit uncompressed data with
the expectati
From: Ian Romanick
Signed-off-by: Ian Romanick
---
src/mesa/drivers/dri/intel/intel_extensions.c | 1 -
src/mesa/drivers/dri/nouveau/nv10_context.c | 2 +-
src/mesa/drivers/dri/nouveau/nv20_context.c | 2 +-
src/mesa/drivers/dri/r200/r200_context.c | 3 ++-
src/mesa/drivers/dri/radeon/
From: Ian Romanick
This is technically outside the ANGLE spec, but it seems unlikely to
cause any harm.
Signed-off-by: Ian Romanick
---
src/mesa/main/extensions.c | 4 ++--
src/mesa/main/glformats.c | 4 +---
src/mesa/main/texformat.c | 14 ++
src/mesa/main/teximage.c | 15 ++
From: Ian Romanick
Always enable the use of pre-compressed texture data. The ability to
perform on-line compression still requires the presence of libtxc_dxtn
or an explicit driconf over-ride. Applications that just want to submit
precompessed data when an on-line compressor is not available ca
---
src/mesa/state_tracker/st_extensions.c |3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/mesa/state_tracker/st_extensions.c
b/src/mesa/state_tracker/st_extensions.c
index d52c840..7ab8d59 100644
--- a/src/mesa/state_tracker/st_extensions.c
+++ b/src/mesa/state_track
Am Freitag, 18. Januar 2013, 17:58:27 schrieben Sie:
> From: Tom Stellard
>
> If we build clover with LLVM static libraries, then clover and also each
> pipe_*.so driver that is built will contain their own static copy of
> LLVM. The recent automake changes have uncovered a problem where
> the p
On Mon, Jan 21, 2013 at 06:32:04PM +0100, Johannes Obermayr wrote:
> Am Freitag, 18. Januar 2013, 17:58:27 schrieben Sie:
> > From: Tom Stellard
> >
> > If we build clover with LLVM static libraries, then clover and also each
> > pipe_*.so driver that is built will contain their own static copy o
From: José Fonseca
It appears that scons implicit dependency scanners fail to chain
dependencies of generated headers when these are outside the build tree.
This patch ensures generated source files are _always_ put in the build
tree. I'm not 100% this will fix all depency issues, but from my
ex
https://bugs.freedesktop.org/show_bug.cgi?id=59238
--- Comment #4 from Tobias Jakobi ---
Concerning visibility for the DRI drivers: Does anyone know which symbols have
to be exported so that libGL can load the drivers?
I've reduced the exported symbols via ld's --version-script to the ones
conta
On Mon, Jan 21, 2013 at 06:36:00PM +0100, Tom Stellard wrote:
> On Mon, Jan 21, 2013 at 06:32:04PM +0100, Johannes Obermayr wrote:
> > Am Freitag, 18. Januar 2013, 17:58:27 schrieben Sie:
> > > From: Tom Stellard
> > >
> > > If we build clover with LLVM static libraries, then clover and also each
On 01/18/2013 10:00 AM, Matt Turner wrote:
> Squashed with two reverts:
>
> Revert "android: Update for builtin_stubs.cpp move"
>
> This reverts commit c0def90ede1e939173041b8785303de90f8fdc6c.
>
> Partially revert "scons: Update for builtin_stubs.cpp"
>
> This reverts commit 8ac4b82699ad0a59ae
Paul Berry writes:
> On 21 January 2013 00:55, Eric Anholt wrote:
>
>> Paul Berry writes:
>>
>> > On 19 January 2013 11:06, Kenneth Graunke wrote:
>> >> + //intel_renderbuffer_set_needs_downsample(dst_irb);
>> >>
>> >
>> > I'm ok leaving this out--at the moment there's no way this code can b
---
lib/Target/R600/AMDILISelDAGToDAG.cpp | 3 +++
lib/Target/R600/R600ISelLowering.cpp | 6 ++
2 files changed, 9 insertions(+)
diff --git a/lib/Target/R600/AMDILISelDAGToDAG.cpp
b/lib/Target/R600/AMDILISelDAGToDAG.cpp
index 567b3e2..ece26ef 100644
--- a/lib/Target/R600/AMDILISelDAGToDAG.c
---
lib/Target/R600/AMDILISelDAGToDAG.cpp | 51 ---
1 file changed, 48 insertions(+), 3 deletions(-)
diff --git a/lib/Target/R600/AMDILISelDAGToDAG.cpp
b/lib/Target/R600/AMDILISelDAGToDAG.cpp
index ece26ef..84223f6 100644
--- a/lib/Target/R600/AMDILISelDAGToDAG.cp
Ian Romanick writes:
> From: Ian Romanick
>
> There really isn't any point. There is no resource savings, and we have
> to do gymnastics in the driver to make it work.
Acked-by: Eric Anholt
pgpAcAocqZlm9.pgp
Description: PGP signature
___
mesa-dev
Left behind by a8ab7e33.
Cc: Paul Berry
---
src/mesa/main/APIspec.dtd | 52 -
1 files changed, 0 insertions(+), 52 deletions(-)
delete mode 100644 src/mesa/main/APIspec.dtd
diff --git a/src/mesa/main/APIspec.dtd b/src/mesa/main/APIspec.dtd
deleted
---
src/mesa/state_tracker/st_cb_texture.c |1 +
src/mesa/state_tracker/st_extensions.c |1 +
src/mesa/state_tracker/st_format.c | 34
src/mesa/state_tracker/st_format.h |3 +++
4 files changed, 39 insertions(+)
diff --git a/src/mesa/state_tr
configure.ac sets up a PYTHON2 variable, which is what we want
AX_PYTHON_MODULE to use (since we only use Python 2 for now).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31598
Cc: Matt Turner
Cc: zephi...@gmail.com
---
m4/ax_python_module.m4 | 8
1 file changed, 4 insertions(+
On Mon, Jan 21, 2013 at 10:46 AM, Kenneth Graunke wrote:
> configure.ac sets up a PYTHON2 variable, which is what we want
> AX_PYTHON_MODULE to use (since we only use Python 2 for now).
>
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31598
> Cc: Matt Turner
> Cc: zephi...@gmail.com
> -
In debug builds, set clipped vertex window coordinates to NaN values
to help debugging. Otherwise, we're just leaving the coordinate in clip
space and it's invalid to use it later expecting it to be a window coord.
---
src/gallium/auxiliary/draw/draw_cliptest_tmp.h | 15 +++
1 files
---
src/gallium/auxiliary/draw/draw_pipe_clip.c | 21 -
1 files changed, 20 insertions(+), 1 deletions(-)
diff --git a/src/gallium/auxiliary/draw/draw_pipe_clip.c
b/src/gallium/auxiliary/draw/draw_pipe_clip.c
index c02d0ef..878d779 100644
--- a/src/gallium/auxiliary/draw/dr
I don't see how this could have ever worked right.
The screen-space interpolation code uses the vertex->data[pos_attr]
position which contain window coords. But window coords are only
computed for the unclipped vertices; the clipped vertices have
undefined window coords (see draw_cliptest_tmp.h).
https://bugs.freedesktop.org/show_bug.cgi?id=55476
--- Comment #3 from Brian Paul ---
I just posted a patch for this. The problem was in the clipper's vertex
interpolation code. The screen-space linear-interpolation code was broken so
we were getting bogus texcoords in the fragment shader used
On 01/21/2013 03:49 AM, Chad Versace wrote:
This series lives on my gles3-glsl-packing branch.
The series is
Reviewed-by: Ian Romanick
Hopefully Eric, Ken, or Paul can comment about the i965 back-end bits.
We should also expose the GL_ARB_shading_language_packing. It's the
same set of fun
On 01/21/2013 10:57 AM, jfons...@vmware.com wrote:
From: José Fonseca
It appears that scons implicit dependency scanners fail to chain
dependencies of generated headers when these are outside the build tree.
This patch ensures generated source files are _always_ put in the build
tree. I'm not 1
On 01/21/2013 10:20 AM, Marek Olšák wrote:
---
src/mesa/state_tracker/st_extensions.c |3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/mesa/state_tracker/st_extensions.c
b/src/mesa/state_tracker/st_extensions.c
index d52c840..7ab8d59 100644
--- a/src/mesa/state_tra
On 01/21/2013 11:45 AM, Marek Olšák wrote:
---
src/mesa/state_tracker/st_cb_texture.c |1 +
src/mesa/state_tracker/st_extensions.c |1 +
src/mesa/state_tracker/st_format.c | 34
src/mesa/state_tracker/st_format.h |3 +++
4 files changed
On Mon, Jan 21, 2013 at 10:56 AM, Ian Romanick wrote:
> We should also expose the GL_ARB_shading_language_packing. It's the same
> set of functions, but it's desktop GL instead of ES. That should be a
> follow up patch... we want to get these patches in as soon as reasonably
> possible.
The ARB
On 20 January 2013 14:15, Paul Berry wrote:
> On 19 January 2013 11:06, Kenneth Graunke wrote:
>
>> + /* CopyTexSubImage should happen even in conditional rendering. We
>> could
>> +* turn it off and back on again. For now, just bail instead.
>> +*/
>> + if (ctx->Query.CondRenderQu
According to page 163 of the ES 3.0 spec:
"Texture lookups involving textures with depth component data generate
a texture base color C_b either using depth data directly or by
performing a comparison with the D_ref value used to perform the
lookup, as described in section 3.8.15. The resultin
https://bugs.freedesktop.org/show_bug.cgi?id=31598
--- Comment #3 from Jan Vesely ---
Created attachment 73402
--> https://bugs.freedesktop.org/attachment.cgi?id=73402&action=edit
use correct version of python to make the check
The attached patch fixes the check on systems where python2 is not
Series looks good to me.
Reviewed-by: Brian Paul
Jose
- Original Message -
> I don't see how this could have ever worked right.
>
> The screen-space interpolation code uses the vertex->data[pos_attr]
> position which contain window coords. But window coords are only
> computed for the
s/one/single/ in the subject?
pgpXudzZrYmN1.pgp
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Ian Romanick writes:
> From: Ian Romanick
>
> This is technically outside the ANGLE spec, but it seems unlikely to
> cause any harm.
Seems like there's some cleanup available if one asserts that the ANGLE
extension is present if the old EXT S3TC extension is present.
> + if (_mesa_is_desktop
Ian Romanick writes:
> From: Ian Romanick
>
> Always enable the use of pre-compressed texture data. The ability to
> perform on-line compression still requires the presence of libtxc_dxtn
> or an explicit driconf over-ride. Applications that just want to submit
> precompessed data when an on-l
On 01/21/2013 02:13 PM, Matt Turner wrote:
On Mon, Jan 21, 2013 at 10:56 AM, Ian Romanick wrote:
We should also expose the GL_ARB_shading_language_packing. It's the same
set of functions, but it's desktop GL instead of ES. That should be a
follow up patch... we want to get these patches in as
Ian Romanick writes:
> From: Ian Romanick
>
> This error was added in the 3.0.1 update to the OpenGL ES 3.0 spec.
> Fixes the updated gles3conform packed_depth_stencil_parameters test.
I was wondering if this wasn't a spec correction in general, but it's
not present in the desktop specs posted
Ian Romanick writes:
> From: Ian Romanick
>
> Fixes gles3conform
> framebuffer_blit_error_blitframebuffer_multisampled_read_buffer_different_origins.
Took me a moment to decide that "bounds" in the spec citation didn't
mean that they wanted to allow flipping during the resolve.
Reviewed-by: Er
Kenneth Graunke writes:
> Thanks to Fredrik Höglund, all the hard work was already done.
>
> Tested using a modified oglconform (that actually runs these tests on
> our driver); it looks like there may be some bugs when using client
> arrays. All applicable non-compatibility tests passed.
>
> Fo
On 01/18/2013 02:11 PM, Kenneth Graunke wrote:
Thanks to Fredrik Höglund, all the hard work was already done.
Tested using a modified oglconform (that actually runs these tests on
our driver); it looks like there may be some bugs when using client
arrays. All applicable non-compatibility tests
Provide the information about indirectly addressable arrays (ranges of temps) in
the shader to the drivers. TGSI representation itself isn't modified, array
information is passed as an additional data in the pipe_shader_state, so the
drivers can use it as a hint for optimization.
---
It's far from
Hi,
I'm not very familiar with the code, but I should mention that this should
really help with the lima compiler backend. In fact, I was thinking of this
very problem recently when figuring out how to translate TGSI into our
backend. Fragment shaders on the Mali 200/400 have 6 vector registers,
w
https://bugs.freedesktop.org/show_bug.cgi?id=31598
LoneVVolf changed:
What|Removed |Added
Status|RESOLVED|REOPENED
Resolution|FIXED
Chad Versace writes:
> ir->remove();
> diff --git a/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
> b/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
> index 324e665..0ff296c 100644
> --- a/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
> +++ b/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
> @@ -923,6 +923,34 @
From: Christian König
Signed-off-by: Christian König
---
lib/Target/R600/AMDGPUStructurizeCFG.cpp | 71 +++---
1 file changed, 36 insertions(+), 35 deletions(-)
diff --git a/lib/Target/R600/AMDGPUStructurizeCFG.cpp
b/lib/Target/R600/AMDGPUStructurizeCFG.cpp
index 5be
v2: don't mess up other loops
Signed-off-by: Christian König
---
lib/Target/R600/AMDGPUStructurizeCFG.cpp | 11 ++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/lib/Target/R600/AMDGPUStructurizeCFG.cpp
b/lib/Target/R600/AMDGPUStructurizeCFG.cpp
index 22338b5..5be40de
Chad Versace writes:
> diff --git a/src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp
> b/src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp
> index ebf8990..b5f1aae 100644
> --- a/src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp
> +++ b/src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp
> @@ -348,6 +348,143
Kenneth Graunke writes:
> ---
> src/mesa/drivers/dri/i965/brw_state.h | 3 ++-
> src/mesa/drivers/dri/i965/brw_wm.c| 2 +-
> src/mesa/drivers/dri/i965/brw_wm_surface_state.c | 8 ++--
> src/mesa/drivers/dri/i965/gen7_wm_surface_state.c | 4 ++--
> 4 files changed,
On Mon, Jan 21, 2013 at 04:40:47PM +0100, Michel Dänzer wrote:
> From: Michel Dänzer
>
>
I don't think we emit the int_AMDGPU_mul intrinsic anymore, but it probably
doesn't hurt to keep it around until we sort out all of the legacy vs non-legacy
instruction issues.
Reviewed-by: Tom Stellard
On Mon, Jan 21, 2013 at 07:15:21PM +0100, Vincent Lejeune wrote:
> ---
> lib/Target/R600/AMDILISelDAGToDAG.cpp | 3 +++
> lib/Target/R600/R600ISelLowering.cpp | 6 ++
> 2 files changed, 9 insertions(+)
>
> diff --git a/lib/Target/R600/AMDILISelDAGToDAG.cpp
> b/lib/Target/R600/AMDILISelDAGTo
On Mon, Jan 21, 2013 at 07:15:22PM +0100, Vincent Lejeune wrote:
> ---
> lib/Target/R600/AMDILISelDAGToDAG.cpp | 51
> ---
> 1 file changed, 48 insertions(+), 3 deletions(-)
>
Reviewed-by: Tom Stellard
> diff --git a/lib/Target/R600/AMDILISelDAGToDAG.cpp
> b/l
Ian Romanick writes:
> On 01/18/2013 02:30 PM, Eric Anholt wrote:
>> This will let us copy from the Exec dispatch to deal with our commands that
>> don't get compiled into display lists.
>> ---
>> src/mesa/main/context.c |5 -
>> src/mesa/main/dlist.c | 12 +++-
>> src/mes
This patch series adds support for varying structs, which are a
required part of GLSL ES 3.00 and GLSL 1.50.
I can see two principal ways to implement this feature: a "flattening"
approach, and a "packing" approach. In the "flattening" approach, the
linker replaces each varying struct with a set
This patch replaces the three ir_variable_mode enums:
- ir_var_in
- ir_var_out
- ir_var_inout
with the following five:
- ir_var_shader_in
- ir_var_shader_out
- ir_var_function_in
- ir_var_function_out
- ir_var_function_inout
This eliminates a frustrating ambiguity: it used to be impossible to
t
Previously, it didn't matter whether structure splitting tried to
split shader ins/outs, because structs were prohibited from being used
for shader ins/outs. However, GLSL 3.00 ES supports varying structs.
In order for varying structs to work, we need to make sure that
structure splitting doesn't
This patch paves the way for allowing varying structs by generalizing
varying_matches::compute_packing_order to handle any type of varying.
Previously, we packed in the order (vec4, vec2, float, vec3), with
matrices being packed according to the size of their columns. Now, we
pack everything accor
This patch adds code to lower_packed_varyings to handle varyings of
type struct. Varying structs are currently packed in the most naive
possible way (in declaration order, with no gaps), so there is a
potential loss of runtime efficiency. In a later patch it would be
nice to replace this with a "
It is not clear from the GLSL ES 3.00 spec how transform feedback is
supposed to apply to varying structs:
- There is no specification for how the structure is to be packed when
it is recorded into the transform feedback buffer.
- There is no reasonable value for GetTransformFeedbackVarying to
---
src/glsl/ast_to_hir.cpp | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
index 3c2ce15..33e20a8 100644
--- a/src/glsl/ast_to_hir.cpp
+++ b/src/glsl/ast_to_hir.cpp
@@ -2031,10 +2031,8 @@ apply_type_qualifier_to_variabl
On 01/21/2013 01:38 PM, Eric Anholt wrote:
Kenneth Graunke writes:
---
src/mesa/drivers/dri/i965/brw_state.h | 3 ++-
src/mesa/drivers/dri/i965/brw_wm.c| 2 +-
src/mesa/drivers/dri/i965/brw_wm_surface_state.c | 8 ++--
src/mesa/drivers/dri/i965/gen7_wm_surf
https://bugs.freedesktop.org/show_bug.cgi?id=59094
Vinson Lee changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
On 01/21/2013 01:18 PM, Eric Anholt wrote:
> Chad Versace writes:
>> ir->remove();
>> diff --git a/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
>> b/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
>> index 324e665..0ff296c 100644
>> --- a/src/mesa/drivers/dri/i965/brw_fs_emit.cpp
>> +++ b/src/mesa/dri
Pedantic whitespace comments:
On Mon, Jan 21, 2013 at 12:49 AM, Chad Versace
wrote:
> That is, evaluate constant expressions of the following functions:
> packSnorm2x16 unpackSnorm2x16
> packUnorm2x16 unpackUnorm2x16
> packHalf2x16 unpackHalf2x16
>
> v2: Reuse _mesa_pack_float_to_half a
https://bugs.freedesktop.org/show_bug.cgi?id=59688
Priority: medium
Bug ID: 59688
Keywords: regression
CC: i...@freedesktop.org
Assignee: mesa-dev@lists.freedesktop.org
Summary: piglit glsl-uniform-out-of-bounds-2 regression
https://bugs.freedesktop.org/show_bug.cgi?id=45920
Vinson Lee changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
1 - 100 of 108 matches
Mail list logo
