On Mon, Dec 15, 2014 at 10:12 PM, Jason Ekstrand wrote:
> This pass uses the previously built algebraic transformations framework and
> should act as an example for anyone else wanting to make an algebraic
> transformation pass for NIR.
> ---
> src/glsl/Makefile.am | 10 -
On 06.01.2015 05:20, Marek Olšák wrote:
> Hi,
>
> The motivation for this patch series is to reduce the number of TC L2
> cache flushes.
>
> The result is that TC L2 is only flushed at IB boundary and when the
> framebuffer is changed. This is achieved by switching all clients to
> use the cache,
Patches 100, 102-104, 106-109, 114-117, 121-146 are:
Reviewed-by: Connor Abbott
Whew! I think I've looked through everything.
On Tue, Dec 16, 2014 at 1:04 AM, Jason Ekstrand
wrote:
> NIR (pronounced "ner") is a new IR (internal representation) for the Mesa
> shader compiler that will sit betw
Can you make the fields lowercase to match the algebraic properties
enum? Or did you make that uppercase? Lowercase feels better to me
since it's an enum, but I don't really care too much.
On Fri, Dec 19, 2014 at 8:02 PM, Jason Ekstrand wrote:
> This should be much better for debugging as GDB wil
On Wed, Dec 17, 2014 at 8:04 PM, Jason Ekstrand wrote:
> Previously, we were doing a lazy creation of the parallel copy
> instructions. This is confusing, hard to get right, and involves some
> extra state tracking of the copies. This commit adds an extra walk over
> the basic blocks to add the
The commit message should probably say something like "nir: Make
nir_ssa_undef_instr_create initialize the destination". That we now
need to pass in the number of components is just a side effect.
On Tue, Dec 16, 2014 at 1:13 AM, Jason Ekstrand wrote:
> ---
> src/glsl/nir/nir.c |
Again, not my area of expertise here.
On Tue, Dec 16, 2014 at 1:13 AM, Jason Ekstrand wrote:
> ---
> src/mesa/drivers/dri/i965/brw_fs_nir.cpp | 24 ++--
> 1 file changed, 22 insertions(+), 2 deletions(-)
>
> diff --git a/src/mesa/drivers/dri/i965/brw_fs_nir.cpp
> b/src/mesa/
I don't really feel qualified to review this, since I didn't write any
of the original code (I just copied-n-pasted it from what i965 was
doing...). I'm not sure who would be able to, maybe Chris Forbes since
he did the indirect sampler work?
On Tue, Dec 16, 2014 at 1:13 AM, Jason Ekstrand wrote:
I created nir_tex_src_sampler_index for exactly this purpose, which
fits in with the "stick all the sources in an array so we can easily
iterate over them" philosophy. If you decide to keep with this
solution, though, at least remove that.
On Tue, Dec 16, 2014 at 1:13 AM, Jason Ekstrand wrote:
>
>>
>> I was mislead by the "leaf" name the first time I reviewed this; having a
>> comment explaining what it does helps, but I still think that it's a pretty
>> misleading name. "Leaf," at least to me, implies that it's a leaf of the
>> dereference tree, which in this case isn't true unless I'm mi
On Tue, Dec 16, 2014 at 1:12 AM, Jason Ekstrand wrote:
> ---
> src/glsl/Makefile.sources| 1 +
> src/glsl/nir/nir.h | 1 +
> src/glsl/nir/nir_opt_constant_folding.c | 283
> +++
> src/mesa/drivers/dri/i965/brw_fs_nir.cpp |
On Monday, January 05, 2015 09:27:36 PM Marek Olšák wrote:
> Radeon hardware doesn't support primitive restart for DrawArrays either.
>
> Marek
Current Intel hardware doesn't support primitive restart for DrawArrays
either. I'm betting the DrawArraysIndirect case is broken on i965 currently,
and
On Tue, Dec 16, 2014 at 1:12 AM, Jason Ekstrand wrote:
> This pass uses the previously built algebraic transformations framework and
> should act as an example for anyone else wanting to make an algebraic
> transformation pass for NIR.
> ---
> src/glsl/Makefile.am | 10 -
>
On Sun, Jan 4, 2015 at 8:56 PM, Connor Abbott wrote:
> Reviewed-by: Connor Abbott
>
> Nice to see that this idea worked out well!
>
> On Tue, Dec 16, 2014 at 1:11 AM, Jason Ekstrand
> wrote:
>
>> We used to have the number of components built into the intrinsic. This
>> meant that all of our l
On Sun, Jan 4, 2015 at 9:15 PM, Connor Abbott wrote:
> This is a general question for the interpolation support:
>
> Why are we using the variable-based intrinsics directly, instead of
> lowering it to something index-based in the lower_io pass just like we do
> for normal inputs?
>
I knew you w
Wow, this commit is a mess. Yeah, I should clean that up.
On Sun, Jan 4, 2015 at 9:19 PM, Connor Abbott wrote:
>
>
> On Tue, Dec 16, 2014 at 1:12 AM, Jason Ekstrand
> wrote:
>
>> ---
>> src/glsl/nir/nir_intrinsics.h | 32 +++-
>> src/glsl/nir/nir_lower_io.c | 16
On Tue, Dec 16, 2014 at 1:12 AM, Jason Ekstrand wrote:
> This commit builds on the nir_search.h infastructure by adds a bit of
adding
> python code that makes it stupid easy to write an algebraic transformation
> pass. The nir_algebraic.py file contains four python classes that
> correspond dir
On Mon, Jan 5, 2015 at 9:12 PM, Connor Abbott wrote:
> Hi,
>
> Was it your intention to not support non-per-component things like dot
> product at all? I've made a few inline comments about how to do it,
> and it doesn't seem like it's that hard.
>
No, It was just never tested on them. All your
Hi,
Was it your intention to not support non-per-component things like dot
product at all? I've made a few inline comments about how to do it,
and it doesn't seem like it's that hard.
On Tue, Dec 16, 2014 at 1:12 AM, Jason Ekstrand wrote:
>
> This framework provides a simple way to do simple sea
On Mon, Jan 05, 2015 at 03:18:54PM -0800, Vinson Lee wrote:
> Merge the following upstream autoconf-archive patches.
>
> ax_prog_flex: change grep syntax to accept e.g. "flex.real" in case a wrapper
> or symlink is used.
> AX_PROG_FLEX: avoid use of grep empty string escape extension (fix for
>
Hi,
Please, could you remove completly all my bugzilla account and all the
informations attached to it ?
I don't have found this feature in the website to do that...
Thank you.
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freed
https://bugs.freedesktop.org/show_bug.cgi?id=88079
Mike Mason changed:
What|Removed |Added
CC||chad.vers...@intel.com,
|
https://bugs.freedesktop.org/show_bug.cgi?id=88079
Mike Mason changed:
What|Removed |Added
See Also||https://bugs.freedesktop.or
https://bugs.freedesktop.org/show_bug.cgi?id=88079
Bug ID: 88079
Summary: dEQP-GLES3.functional.fbo.completeness.renderable.rend
erbuffer.color0 tests fail due to enabling of GL_RGB
and GL_RGBA
Product: Mesa
Versi
This has been fairly extensively updated, per your comments:
http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=9a3a8c4fcc95cc6e8834cd7af6b5d8c262942d9a
Thanks.
Laura
On Tue, Dec 16, 2014 at 7:46 AM, Brian Paul wrote:
> On 12/15/2014 06:22 PM, Laura Ekstrand wrote:
>
>> ---
>>
These comments have been addressed.
Thanks.
Laura
On Tue, Dec 30, 2014 at 5:54 PM, Anuj Phogat wrote:
> On Tue, Dec 16, 2014 at 11:54 PM, Laura Ekstrand
> wrote:
> > No. I have a 78 column line highlight in my editor.
> >
> It does fit when you remove the whitespaces in function braces.
> >
I've removed extra spaces in my function headers in this patch and those
previous. As I continue to address comments on patches, I will try to
remove any other extra spaces I see.
Thanks.
Laura
On Fri, Jan 2, 2015 at 7:50 AM, Brian Paul wrote:
> On 12/30/2014 06:45 PM, Anuj Phogat wrote:
>
>>
On Mon, Jan 5, 2015 at 2:38 PM, Chad Versace wrote:
> On 01/03/2015 11:54 AM, Jason Ekstrand wrote:
> > From: Sisinty Sasmita Patra
> >
> > This commit refactors the tiled_memcpy code in intel_tex_subimage.c and
> > moves it into its own file intel_tiled_memcpy files. Also, xtile_copy
> and
> >
Hi,
About this patch:
1. It is not tested. I'll test it after 12th.
2. It implements atomic buffers as a surface which can be reused for
ARB_shader_image_load_store
3. You can ignore the first patch.
My questions:
1. What does R_028AC0_ALU_ATOM_CACHE_GS_0 represent?
2. What determines the values
---
src/gallium/include/pipe/p_context.h | 5 +
src/gallium/include/pipe/p_defines.h | 7 ++-
src/gallium/include/pipe/p_state.h | 10 ++
3 files changed, 21 insertions(+), 1 deletion(-)
diff --git a/src/gallium/include/pipe/p_context.h
b/src/gallium/include/pipe/p_context.h
Fixes piglit vs-*-main-return.
---
src/glsl/lower_packed_varyings.cpp | 32 +++-
1 file changed, 31 insertions(+), 1 deletion(-)
diff --git a/src/glsl/lower_packed_varyings.cpp
b/src/glsl/lower_packed_varyings.cpp
index 5e844c7..b8c3228 100644
--- a/src/glsl/lower_pac
On Mon, Jan 5, 2015 at 3:47 PM, Matt Turner wrote:
> On Mon, Jan 5, 2015 at 3:27 PM, Jason Ekstrand
> wrote:
> > From: Sisinty Sasmita Patra
> >
> > Added intel_readpixels_tiled_mempcpy and
> intel_gettexsubimage_tiled_mempcpy
> > functions. These are the fast paths for glReadPixels and glGetTe
On Mon, Jan 5, 2015 at 3:27 PM, Jason Ekstrand wrote:
> From: Sisinty Sasmita Patra
>
> Added intel_readpixels_tiled_mempcpy and intel_gettexsubimage_tiled_mempcpy
> functions. These are the fast paths for glReadPixels and glGetTexImage.
>
> On chrome, using the RoboHornet 2D Canvas toDataURL tes
On Mon, Jan 5, 2015 at 3:18 PM, Vinson Lee wrote:
> Merge the following upstream autoconf-archive patches.
>
> ax_prog_flex: change grep syntax to accept e.g. "flex.real" in case a wrapper
> or symlink is used.
> AX_PROG_FLEX: avoid use of grep empty string escape extension (fix for
> OpenBSD)
>
From: Sisinty Sasmita Patra
Added intel_readpixels_tiled_mempcpy and intel_gettexsubimage_tiled_mempcpy
functions. These are the fast paths for glReadPixels and glGetTexImage.
On chrome, using the RoboHornet 2D Canvas toDataURL test, this patch cuts
amount of time spent in glReadPixels by more t
This comment has been addressed.
On Tue, Dec 30, 2014 at 4:09 PM, Anuj Phogat wrote:
> On Tue, Dec 16, 2014 at 6:52 AM, Laura Ekstrand
> wrote:
> > This is a convenience function for *Texture*Parameter functions.
> > ---
> > src/mesa/main/texparam.c | 38 ++
Merge the following upstream autoconf-archive patches.
ax_prog_flex: change grep syntax to accept e.g. "flex.real" in case a wrapper
or symlink is used.
AX_PROG_FLEX: avoid use of grep empty string escape extension (fix for OpenBSD)
AX_PROG_FLEX: Also accept gflex.
Signed-off-by: Vinson Lee
---
These comments have been addressed:
http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=fe7abbcd2356a25758791bf1a66654936a7a1837
On Tue, Dec 30, 2014 at 4:03 PM, Anuj Phogat wrote:
> On Tue, Dec 16, 2014 at 6:52 AM, Laura Ekstrand
> wrote:
> > Beginning in the OpenGL 4.3 core spe
These comments have been addressed.
On Tue, Dec 30, 2014 at 3:46 PM, Anuj Phogat wrote:
> On Tue, Dec 16, 2014 at 7:46 AM, Brian Paul wrote:
> > On 12/15/2014 06:22 PM, Laura Ekstrand wrote:
> >>
> >> The following preparations were made in texstate.c and texstate.h to
> >> better facilitate th
On Sun, Jan 4, 2015 at 1:07 PM, Ben Widawsky wrote:
> I just did a very cursory review. I assume someone smarter than me will do
> a
> real review, but if not, feel free to ping me.
>
> I think all the comments apply to both functions.
>
> On Sat, Jan 03, 2015 at 11:54:15AM -0800, Jason Ekstrand
On 01/03/2015 11:54 AM, Jason Ekstrand wrote:
> From: Sisinty Sasmita Patra
>
> This commit addes tiled copy functions for coping from tiled memory to
> linear memory. These are very similar to the existing linear-to-tiled
> paths.
>
> v2: Jason Ekstrand
>- New commit message
>- Variou
On 01/03/2015 11:54 AM, Jason Ekstrand wrote:
> From: Sisinty Sasmita Patra
>
> This commit refactors the tiled_memcpy code in intel_tex_subimage.c and
> moves it into its own file intel_tiled_memcpy files. Also, xtile_copy and
> ytile_copy are renamed to linear_to_xtiled and linear_to_ytiled
>
On 01/03/2015 11:54 AM, Jason Ekstrand wrote:
> There's no reason why we should be doing this for 2D textures and not
> rectangles. Just a matter of adding another hunk to the condition.
>
> Signed-off-by: Jason Ekstrand
> ---
> src/mesa/drivers/dri/i965/intel_tex_subimage.c | 3 ++-
> 1 file c
I've addressed Brian Paul's comment about dsa ? "ture" : "" in this patch.
On Tue, Dec 30, 2014 at 1:25 PM, Laura Ekstrand
wrote:
>
>
> On Mon, Dec 29, 2014 at 3:02 PM, Anuj Phogat
> wrote:
>
>> On Tue, Dec 16, 2014 at 6:52 AM, Laura Ekstrand
>> wrote:
>> > ---
>> > src/mapi/glapi/gen/ARB_dir
These comments have been addressed:
http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=a2f20c936f6db31986220938db06cf28885e7ee6
.
Thanks.
Laura
On Wed, Dec 31, 2014 at 6:03 PM, Anuj Phogat wrote:
> On Tue, Dec 16, 2014 at 10:59 AM, Laura Ekstrand
> wrote:
> > This happens almo
On 01/05/2015 10:34 AM, Laura Ekstrand wrote:
> I added a clearer explanation based on our irc discussion:
>
> http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=db418437915bba959c0b6c8babe40f675bbdf31c
Thanks. LGTM.
signature.asc
Description: OpenPGP digital signature
These problems have all been corrected except one:
In glGenTextures it is create_textures(ctx, 0 ... rather than GL_NONE
because the original glGenTextures had the [now removed] line:
- GLenum target = 0;
So I passed in 0 to preserve the original functionality.
On Wed, Dec 17, 2014 at 6
Eric Anholt writes:
> If you had a conditional assignment of an array or struct (say, from the
> if-lowering pass), we'd try doing swizzle_for_size() on the aggregate
> type, and it would assertion fail due to vector_elements==0. Instead,
> extend emit_block_mov() to handle emitting the conditio
These are all:
Reviewed-by: Eric Anholt
signature.asc
Description: PGP signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Radeon hardware doesn't support primitive restart for DrawArrays either.
Marek
On Mon, Jan 5, 2015 at 5:48 PM, Roland Scheidegger wrote:
> I'm wondering how this is supposed to work with draw (and draw_info in
> general). Looks like draw will not do prim restart for non-indexed calls
> - d3d10 a
From: Marek Olšák
---
src/gallium/drivers/radeonsi/si_descriptors.c | 4
src/gallium/drivers/radeonsi/si_state.c | 6 ++
2 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_descriptors.c
b/src/gallium/drivers/radeonsi/si_descriptors.c
in
From: Marek Olšák
This allows not flushing TC L2 on CIK later.
---
src/gallium/drivers/radeonsi/si_descriptors.c | 13 -
src/gallium/drivers/radeonsi/sid.h| 2 ++
2 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_descriptors.
From: Marek Olšák
This fixes a case where a transform feedback buffer is fed back as an index
buffer, because SURFACE_SYNC must be after VS_PARTIAL_FLUSH.
---
src/gallium/drivers/radeonsi/si_state_draw.c | 58 +---
1 file changed, 35 insertions(+), 23 deletions(-)
diff -
From: Marek Olšák
---
src/gallium/drivers/radeonsi/si_descriptors.c | 20
src/gallium/drivers/radeonsi/si_hw_context.c | 5 +
src/gallium/drivers/radeonsi/si_pipe.h| 5 +
3 files changed, 10 insertions(+), 20 deletions(-)
diff --git a/src/gallium/drivers/
From: Marek Olšák
So that TC L2 doesn't need to be flushed.
The only problem is with index buffers, which don't use TC.
A simple solution is added that flushes TC L2 before a draw call (TC_L2_dirty).
---
src/gallium/drivers/radeon/r600_pipe_common.h | 12 ++
src/gallium/drivers/radeonsi
From: Marek Olšák
I will rename them for radeonsi.
---
src/gallium/drivers/r600/r600_pipe.h | 15 +++
src/gallium/drivers/radeon/r600_pipe_common.h | 24 ++--
src/gallium/drivers/radeonsi/si_descriptors.c | 4 ++--
src/gallium/drivers/radeonsi/si_pipe.h
From: Marek Olšák
---
src/gallium/drivers/r600/r600_hw_context.c | 9 ++---
src/gallium/drivers/radeon/r600_streamout.c | 7 +--
2 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/src/gallium/drivers/r600/r600_hw_context.c
b/src/gallium/drivers/r600/r600_hw_context.c
index
From: Marek Olšák
This is easier to read and will work better with shader image stores.
---
src/gallium/drivers/radeonsi/si_state.c | 18 +++---
1 file changed, 7 insertions(+), 11 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_state.c
b/src/gallium/drivers/radeonsi/si_s
From: Marek Olšák
That's either framebuffer caches or caches for shader resources.
The motivation is that framebuffer caches need to be flushed very rarely
here.
---
src/gallium/drivers/r600/r600_blit.c | 3 +-
src/gallium/drivers/radeon/r600_pipe_common.c | 5 +--
src/gallium/drivers
From: Marek Olšák
---
src/gallium/drivers/radeonsi/si_state_draw.c | 26 +-
1 file changed, 17 insertions(+), 9 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_state_draw.c
b/src/gallium/drivers/radeonsi/si_state_draw.c
index bcb85b3..ccc44d5 100644
--- a/src/
From: Marek Olšák
- we don't usually need to flush TC L2
- we should flush KCACHE
(not really an issue now since we always flush KCACHE when updating
descriptors, but it could be a problem if we used CE, which doesn't
require flushing KCACHE)
- add an explicit VS_PARTIAL_FLUSH flag
---
s
From: Marek Olšák
It's causing problems, because we mix uncached CP DMA with cached WRITE_DATA
when updating the same memory.
The solution for SI is to use uncached access here, because CP DMA doesn't
support cached access.
CIK will be handled in the next patch.
---
src/gallium/drivers/radeons
From: Marek Olšák
---
src/gallium/drivers/radeonsi/si_compute.c | 22 +
src/gallium/drivers/radeonsi/si_descriptors.c | 64 ++---
src/gallium/drivers/radeonsi/si_hw_context.c| 20
src/gallium/drivers/radeonsi/si_pipe.h | 42 +-
Hi,
The motivation for this patch series is to reduce the number of TC L2 cache
flushes.
The result is that TC L2 is only flushed at IB boundary and when the
framebuffer is changed. This is achieved by switching all clients to use the
cache, so that all data is coherent between clients. In thi
I just ran
git grep -e "_err(" --and -e "lookup",
followed by
git grep -e "_error(" --and -e "lookup".
It looks like there is precedence for the naming convention
_mesa_lookup_[object_name]_err instead of error.
On Wed, Dec 17, 2014 at 5:51 PM, Laura Ekstrand
wrote:
> That makes sense. It's now
Same new explanation as patch 2:
http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=e43d8e347eddc85ff7ecadff1a5c44d646b8e9dc
On Thu, Jan 1, 2015 at 11:16 AM, Chad Versace
wrote:
> On 12/31/2014 05:26 PM, Laura Ekstrand wrote:
> > This is part of a potential solution to Khronos B
I added a clearer explanation based on our irc discussion:
http://cgit.freedesktop.org/~ldeks/mesa/commit/?h=adsa-textures&id=db418437915bba959c0b6c8babe40f675bbdf31c
On Thu, Jan 1, 2015 at 11:15 AM, Chad Versace
wrote:
> On 12/31/2014 05:26 PM, Laura Ekstrand wrote:
> > This is part of a poten
https://bugs.freedesktop.org/show_bug.cgi?id=86837
--- Comment #22 from bgunte...@gmail.com ---
(In reply to Aaron Watry from comment #21)
> (In reply to bgunteriv from comment #17)
> > (In reply to Andy Furniss from comment #13)
> > > (In reply to Emil Velikov from comment #12)
> > > > Seems like
Chad and Anuj, thanks for the catch. I've fixed it.
Laura
On Thu, Jan 1, 2015 at 11:08 AM, Chad Versace
wrote:
> On 12/31/2014 05:26 PM, Laura Ekstrand wrote:
>
> > +/**
> > + * Check if the given cube map texture is "cube complete" as defined in
> > + * the OpenGL specification.
> > + */
> >
The old way made it impossible for the optimizer to reason about what
was going on. The new way is the same number of instructions (the neg
gets folded into the cvt) but enables the optimizer to be cleverer if
comparing to a constant (most common case). [The optimizer is presently
not sufficiently
This comment is vague. Do you have a specific recommendation for the code
here?
Thanks.
Laura
On Sat, Jan 3, 2015 at 7:17 AM, Emil Velikov
wrote:
> On 30/12/14 22:20, Laura Ekstrand wrote:
> > To run this partial implementation,
> >
> > export MESA_EXTENSION_OVERRIDE=+GL_ARB_direct_state_acce
https://bugs.freedesktop.org/show_bug.cgi?id=87926
--- Comment #1 from Roland Scheidegger ---
Hmm I guess we'd need to drop the draw call somewhere (since results are
unspecified). Not sure where the optimal place to do it is. Or invoke
out-of-bounds behavior somewhere else. I think llvmpipe (whi
https://bugs.freedesktop.org/show_bug.cgi?id=87913
Roland Scheidegger changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://bugs.freedesktop.org/show_bug.cgi?id=87658
Roland Scheidegger changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
From: Rob Clark
NOTE IN[] and OUT[] don't need (have?) ArrayID's.. and TEMP[] can
optionally have them. So we implicitly assume that ArrayID==0 always
exists for each file. This is why array_max[file] is never less than
zero.
You can tell from indirect_files(_read/written) if the legacy array-
I'm wondering how this is supposed to work with draw (and draw_info in
general). Looks like draw will not do prim restart for non-indexed calls
- d3d10 apparently works like the FIXED_INDEX case of gl (so, arrays
won't trigger restarts). I wasn't even aware that GL triggers restarts
for DrawArray c
Reviewed-by: Roland Scheidegger
Am 04.01.2015 um 14:21 schrieb Rob Clark:
> From: Rob Clark
>
> At least temporarily, I need to fallback to old compiler still for
> relative dest (for freedreno), but I can do relative src temp. Only
> a temporary situation, but seems easy/reasonable for tgsi-s
https://bugs.freedesktop.org/show_bug.cgi?id=86837
--- Comment #21 from Aaron Watry ---
(In reply to bgunteriv from comment #17)
> (In reply to Andy Furniss from comment #13)
> > (In reply to Emil Velikov from comment #12)
> > > Seems like Christian dropped the link with the tentative fix.
> > >
I will drop this patch. There is up to -15% change in performance for
legacy apps. Not worth it.
Marek
On Mon, Jan 5, 2015 at 12:18 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> This can be done using the SPI mapping only. If two_side is disabled,
> VS COLOR is loaded to both PS COLOR and PS B
https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #12 from Eero Tamminen ---
(In reply to Stéphane Travostino from comment #11)
> After spending half a day bisecting, I don't think there's any real
> difference between Mesa 10.3.2 and master: the fps drops happen in both
> releases,
https://bugs.freedesktop.org/show_bug.cgi?id=86837
--- Comment #20 from Andy Furniss ---
(In reply to Andy Furniss from comment #19)
> (In reply to bgunteriv from comment #17)
> > @Andy Furniss, what is your command line for building mesa?
Forgot to put I also have
--enable-texture-float
Not
https://bugs.freedesktop.org/show_bug.cgi?id=86837
--- Comment #19 from Andy Furniss ---
(In reply to bgunteriv from comment #17)
> (In reply to Andy Furniss from comment #13)
> > (In reply to Emil Velikov from comment #12)
> > > Seems like Christian dropped the link with the tentative fix.
> > >
On 12/16/2014 09:15 AM, Eduardo Lima Mitev wrote:
> '3.8.2 Sampler Objects' section of the GL-ES 3.0 specification states:
>
> "An INVALID_OPERATION error is generated if sampler is not the name
> of a sampler object previously returned from a call to GenSamplers."
>
> In desktop GL, an G
On 01/05/2015 05:41 AM, Kenneth Graunke wrote:
> On Sunday, January 04, 2015 04:04:20 PM Abdiel Janulgue wrote:
>> Signed-off-by: Abdiel Janulgue
>> ---
>> src/mesa/drivers/dri/i965/gen7_blorp.cpp | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/src/mesa/drivers/dri/i965/gen7_blorp
On Mon, Dec 22, 2014 at 05:50:33PM -0800, Ben Widawsky wrote:
> On Mon, Dec 22, 2014 at 01:28:32AM -0800, Kenneth Graunke wrote:
> > Our ability to perform register writes depends on the hardware and
> > kernel version. It shouldn't ever change on a per-context basis,
> > so we only need to check
85 matches
Mail list logo
