On Mon, Jan 02, 2017 at 06:37:21PM -0800, Ben Widawsky wrote:
> On Gen9 hardware, the display engine is able to scanout a compressed
> framebuffer by providing an offset to auxiliary compression information.
> Unfortunately, the hardware is incapable of doing the same thing for the
> fast clear
On Wed, Jan 04, 2017 at 10:26:50AM +0200, Pohjolainen, Topi wrote:
> On Mon, Jan 02, 2017 at 06:37:15PM -0800, Ben Widawsky wrote:
> > Allows us to continue utilizing common miptree creation using __DRIimage
> > without creating a new DRIimage (for the intel_process_dri2_buffer()
> > case).
> >
>
On Tue, 2017-01-03 at 08:59 -0800, Jason Ekstrand wrote:
> Dave recently added a spirv-specific structure for this sort of
> feature enabling. I think it would be better to use that rather than
> nir_options.
>
Great! This change also avoids me to modify nir_options so I can remove
some
Changes from v1:
- Use Chris's patch to check if we can do pipelined register writes instead
of trying to reuse the infrastructure from intel_batchbuffer
- Add a kernel_features bitfield to the intel screen that we can check to
see if specific features are available. Drop
We can use this to track various features that may or may not be supported
by the hw / kernel. Currently, we usually do this by checking the generation
and supported command parser versions in various places thoughtout the driver
code. With this patch, we centralize all these checks in just once
---
src/mesa/drivers/dri/i965/intel_extensions.c | 2 ++
src/mesa/drivers/dri/i965/intel_screen.c | 3 ++-
2 files changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/intel_extensions.c
b/src/mesa/drivers/dri/i965/intel_extensions.c
index 8e67b57..aa89380 100644
From: Chris Wilson
Moving the test to the screen places it alongside the other global HW
feature tests that want to be shared between contexts.
Also, we need to know if we support pipelined register writes at
screen creation time so that we can tell if we can expose
Instead, check the screen field directly.
---
src/mesa/drivers/dri/i965/brw_context.c | 2 --
src/mesa/drivers/dri/i965/brw_context.h | 5 -
src/mesa/drivers/dri/i965/gen7_l3_state.c| 5 +++--
src/mesa/drivers/dri/i965/intel_extensions.c | 2 +-
Here's my theory about the Witcher 2 issue:
The problematic shader contains KILL. Reloading inputs after KILL is
unsafe, because KILL breaks the WQM mode, thus the inputs are not
loaded for the whole quad. Control flow statements have a similar
issue.
These are the cases when inputs can be
Am 04.01.2017 um 11:47 schrieb Marek Olšák:
From: Marek Olšák
The context may be used by texture_get_handle.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99158
Ah, yes of course. Did you added the context to resource_get_handle
recently? If not CC stable
On Tue, 2017-01-03 at 14:41 -0800, Jason Ekstrand wrote:
> I made a few pretty trivial comments. With those addressed,
>
> Reviewed-by: Jason Ekstrand
>
> On Dec 16, 2016 8:55 AM, "Juan A. Suarez Romero" > wrote:
> So far, input_reads was a bitmap
This series is,
Acked-by: Edward O'Callaghan
On 01/04/2017 06:17 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> ---
> src/gallium/drivers/radeonsi/si_compute.c | 2 +-
> src/gallium/drivers/radeonsi/si_debug.c | 2 +-
>
Ah, right. Thanks for the explanation!
On Jan 4, 2017 14:02, "Ilia Mirkin" wrote:
> On Wed, Jan 4, 2017 at 4:24 AM, Erik Faye-Lund
> wrote:
> > On Jan 2, 2017 06:03, "Ilia Mirkin" wrote:
> >
> > Signed-off-by: Ilia Mirkin
On Jan 4, 2017 4:54 AM, "Erik Faye-Lund" wrote:
On Jan 3, 2017 16:34, "Jason Ekstrand" wrote:
On Fri, Dec 16, 2016 at 6:49 AM, Juan A. Suarez Romero
wrote:
> From: Samuel Iglesias Gonsálvez
>
>
On Wed, Jan 4, 2017 at 4:24 AM, Erik Faye-Lund wrote:
> On Jan 2, 2017 06:03, "Ilia Mirkin" wrote:
>
> Signed-off-by: Ilia Mirkin
> ---
> src/gallium/auxiliary/tgsi/tgsi_info.c | 2 +-
> src/gallium/docs/source/tgsi.rst
On Jan 4, 2017 5:46 AM, "Juan A. Suarez Romero" wrote:
On Tue, 2017-01-03 at 14:41 -0800, Jason Ekstrand wrote:
I made a few pretty trivial comments. With those addressed,
Reviewed-by: Jason Ekstrand
On Dec 16, 2016 8:55 AM, "Juan A. Suarez Romero"
Hi,
On 04.01.2017 13:07, Kenneth Graunke wrote:
This series reworks i965's handling of gl_TessLevelInner/Outer[] arrays.
Instead of using lower_tess_levels to turn them into vec4/vec2s, we pass
them through to NIR and make them compact arrays (where array indexing
translates to enhanced layouts
Hello
> Last time I looked at this extension I think there was at least one of
> the mainstream emulators using it. Can't recall which one ... pcsx2
> maybe.
I tried to use it on PCSX2. Unfortunately the X / GLX / (EGL?) / Mesa
implementation of context creation doesn't allow extra (AKA
On Jan 4, 2017 4:39 AM, "Eero Tamminen" wrote:
Hi,
Tested-by: Eero Tamminen
May also impact other programs having float indexed loops:
- Invisible Inc
- Talos Principle
I think I saw most Talos loops get unrolled. There was one or two
We need to move this to the shared layer.
Signed-off-by: Kenneth Graunke
---
src/mesa/drivers/dri/i965/brw_vec4.cpp | 5 +
src/mesa/drivers/dri/i965/brw_vs.c | 3 ---
2 files changed, 5 insertions(+), 3 deletions(-)
diff --git
On Wed, 2017-01-04 at 06:50 -0800, Jason Ekstrand wrote:
> On Jan 4, 2017 4:39 AM, "Eero Tamminen"
> wrote:
> Hi,
>
> Tested-by: Eero Tamminen
>
> May also impact other programs having float indexed loops:
> - Invisible Inc
> - Talos
On Mon, 2016-12-12 at 19:39 -0800, Jason Ekstrand wrote:
> ---
> src/compiler/nir/nir_remove_dead_variables.c | 66
>
> 1 file changed, 58 insertions(+), 8 deletions(-)
>
> diff --git a/src/compiler/nir/nir_remove_dead_variables.c
>
On 17-01-04 12:33:56, Chad Versace wrote:
Fixes crash in piglit
`egl_khr_gl_renderbuffer_image-clear-shared-image GL_DEPTH_COMPONENT24`
on Skylake.
The crash happened because blorp attempted to execute a pending hiz
clear after the hiz buffer was deleted. Deleting the pending hiz ops
when the
https://bugs.freedesktop.org/show_bug.cgi?id=98604
--- Comment #11 from Michel Dänzer ---
Can you get a backtrace of a crash with gdb?
--
You are receiving this mail because:
You are on the CC list for the bug.___
mesa-dev mailing
On 17-01-04 10:34:17, Topi Pohjolainen Topi Pohjolainen wrote:
On Wed, Jan 04, 2017 at 09:40:51AM +0200, Pohjolainen, Topi wrote:
On Mon, Jan 02, 2017 at 06:37:10PM -0800, Ben Widawsky wrote:
> This code will disable actually creating these buffers for the scanout,
> but it puts the allocation
On 17-01-04 09:51:20, Topi Pohjolainen Topi Pohjolainen wrote:
On Mon, Jan 02, 2017 at 06:37:13PM -0800, Ben Widawsky wrote:
v2: Leave "image+mod" (Topi)
Signed-off-by: Ben Widawsky
Acked-by: Daniel Stone
---
src/mesa/drivers/dri/i965/intel_screen.c
Looks fine.
Reviewed-by: Timothy Arceri
On Mon, 2016-12-12 at 19:39 -0800, Jason Ekstrand wrote:
> ---
> src/compiler/nir/nir_builder.h | 16
> 1 file changed, 16 insertions(+)
>
> diff --git a/src/compiler/nir/nir_builder.h
>
From: Roland Scheidegger
Doing these operations with blend format means that we have to convert
the destination into blend format, which is entirely pointless if we don't
do blending. For instance, we'd convert half floats to floats, or 10/10/10/2
to unorm16, just to apply
On 17-01-04 10:00:59, Topi Pohjolainen Topi Pohjolainen wrote:
On Mon, Jan 02, 2017 at 06:37:18PM -0800, Ben Widawsky wrote:
In the foreseeable future it doesn't seem to make sense to have multiple
resolve flags. What does make sense is to have the caller give an
indication to the lower layers
On 17-01-04 10:57:40, Topi Pohjolainen Topi Pohjolainen wrote:
On Wed, Jan 04, 2017 at 10:26:50AM +0200, Pohjolainen, Topi wrote:
On Mon, Jan 02, 2017 at 06:37:15PM -0800, Ben Widawsky wrote:
> Allows us to continue utilizing common miptree creation using __DRIimage
> without creating a new
On 17-01-04 10:29:45, Topi Pohjolainen Topi Pohjolainen wrote:
On Mon, Jan 02, 2017 at 06:37:21PM -0800, Ben Widawsky wrote:
On Gen9 hardware, the display engine is able to scanout a compressed
framebuffer by providing an offset to auxiliary compression information.
Unfortunately, the hardware
Reviewed-by: Timothy Arceri
On Mon, 2016-12-12 at 19:39 -0800, Jason Ekstrand wrote:
> ---
> src/compiler/nir/nir_builder.h | 7 +++
> 1 file changed, 7 insertions(+)
>
> diff --git a/src/compiler/nir/nir_builder.h
> b/src/compiler/nir/nir_builder.h
> index
Ok, given what you said below, I think everything is sane.
Reviewed-by: Jason Ekstrand
Note that I didn't actually verify that all of the code you deleted is
going to work once we switch over to NIR handling it with varyings.
On Wed, Jan 4, 2017 at 2:12 PM, Kenneth
Eero Tamminen writes:
> Hi,
>
> On 04.01.2017 13:07, Kenneth Graunke wrote:
>> This series reworks i965's handling of gl_TessLevelInner/Outer[] arrays.
>> Instead of using lower_tess_levels to turn them into vec4/vec2s, we pass
>> them through to NIR and make them
On 17-01-04 12:21:36, Chad Versace wrote:
Pre-patch, if the user created an EGLImage from an intel_mipmap_tree
before the miptree's first use, then intel_miptree_make_shareable()
failed to prevent later creation of the miptree's auxilliary surface.
This results in the original, exported miptree
On 17-01-04 10:41:58, Topi Pohjolainen Topi Pohjolainen wrote:
On Mon, Jan 02, 2017 at 06:37:22PM -0800, Ben Widawsky wrote:
v2: Try to keep the assert as recommended by Topi. This requires
modifying the num_samples check to be <= 1 because internally created
buffers set num_samples = 0.
Cc:
On Wednesday, January 4, 2017 3:16:41 PM PST Eero Tamminen wrote:
> Hi,
>
> On 04.01.2017 13:07, Kenneth Graunke wrote:
> > This series reworks i965's handling of gl_TessLevelInner/Outer[] arrays.
> > Instead of using lower_tess_levels to turn them into vec4/vec2s, we pass
> > them through to NIR
I resubmitted the kernel patch to lkml and cc'ed Daniel.
On Mon, Jan 2, 2017 at 10:32 AM, Daniel Vetter wrote:
> On Thu, Dec 29, 2016 at 04:43:34PM -0800, Ben Widawsky wrote:
> > On 16-12-16 21:27:51, Rainer Hochecker wrote:
> > > From: Rainer Hochecker
On Mon, Dec 19, 2016 at 10:19 PM, Thierry Reding
wrote:
> On Mon, Dec 19, 2016 at 04:04:34PM +, Emil Velikov wrote:
>> On Monday, 19 December 2016, Thierry Reding
>> wrote:
>>
>> > On Wed, Nov 30, 2016 at 02:44:36PM +0100, Christian Gmeiner
On 12/22/2016 05:07 PM, Nicolai Hähnle wrote:
> On 19.12.2016 23:26, Christian Inci wrote:
>> Hashcat needs MAX_GLOBAL_BUFFERS to be 21 or even 22 for some modes. It'll
>> crash otherwise.
>> I'm adding an assert to see if programs need it to be even higher.
>>
>> Signed-off-by: Christian Inci
v2: Make the error return be -1 instead of 0 because I think 0 is
actually valid.
Cc: Daniel Stone
Signed-off-by: Ben Widawsky
---
src/gbm/backends/dri/gbm_dri.c | 28
src/gbm/gbm-symbols-check | 1 +
If the VUE map has slots at the end which the shader does not write,
then we'd "flush" (constructing an URB write) on the last output it
actually wrote. Then, we'd construct another SEND with EOT, but with
no actual payload data. That's not legal.
For example, SSO programs have clip distance
On Wed, 2017-01-04 at 16:17 +0100, Erik Faye-Lund wrote:
>
>
> On Jan 4, 2017 14:45, "Jason Ekstrand" wrote:
> On Jan 4, 2017 4:54 AM, "Erik Faye-Lund" wrote:
> On Jan 3, 2017 16:34, "Jason Ekstrand" wrote:
> On Fri, Dec 16,
On Wed, 2017-01-04 at 15:35 -0800, Kenneth Graunke wrote:
> On Wednesday, January 4, 2017 1:26:23 PM PST Iago Toral Quiroga
> wrote:
> >
> > We can use this to track various features that may or may not be
> > supported
> > by the hw / kernel. Currently, we usually do this by checking the
> >
On Wed, 2017-01-04 at 15:34 -0800, Kenneth Graunke wrote:
> On Wednesday, January 4, 2017 1:26:20 PM PST Iago Toral Quiroga
> wrote:
> >
> > From: Chris Wilson
> [snip]
> >
> > + /* And afterwards clear the register */
> > + if (reset) {
> > + *batch++ =
---
docs/relnotes/13.1.0.html | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/relnotes/13.1.0.html b/docs/relnotes/13.1.0.html
index 4dce843..124da93 100644
--- a/docs/relnotes/13.1.0.html
+++ b/docs/relnotes/13.1.0.html
@@ -47,6 +47,8 @@ Note: some of the new features are only
There was a bit to take in here but it seems ok to me. I've made a
bunch of trivial suggestions/comments below otherwise:
Reviewed-by: Timothy Arceri
On Mon, 2016-12-12 at 19:39 -0800, Jason Ekstrand wrote:
> ---
> src/compiler/Makefile.sources | 1
On 20/12/16 07:37 PM, Timothy Arceri wrote:
> Reviewed-by: Nicolai Hähnle
> ---
> src/mesa/state_tracker/st_atom_texture.c | 5 +
> 1 file changed, 1 insertion(+), 4 deletions(-)
>
> diff --git a/src/mesa/state_tracker/st_atom_texture.c
>
On Jan 4, 2017 05:08, "Kenneth Graunke" wrote:
It's only used in one place, it ignores the offset parameter currently,
and I want to add more parameters...at which point, passing in a bunch
of integers seems less obvious than writing it out.
Signed-off-by: Kenneth Graunke
On 21/12/16 04:01, srol...@vmware.com wrote:
From: Roland Scheidegger
Using bit replication. This path now resembles something which might make
sense. (The logic was mostly copied from llvmpipe fs backend.)
I am not convinced though it is actually faster than SoA sampling
On 21/12/16 04:01, srol...@vmware.com wrote:
From: Roland Scheidegger
simd instruction sets usually have comparisons for equal, not unequal.
So use a different comparison against the mask itself - which also means
we don't need a all-zero as well as a all-one (for the pxor)
On Jan 4, 2017 14:45, "Jason Ekstrand" wrote:
On Jan 4, 2017 4:54 AM, "Erik Faye-Lund" wrote:
On Jan 3, 2017 16:34, "Jason Ekstrand" wrote:
On Fri, Dec 16, 2016 at 6:49 AM, Juan A. Suarez Romero
wrote:
>
On 21/12/16 04:01, srol...@vmware.com wrote:
From: Roland Scheidegger
This code uses a vector shift which has to be emulated on x86 unless
there's AVX2. Luckily in some cases we can actually avoid the shift
altogether, so do that.
Also make sure we hit the fast
On 21/12/16 04:01, srol...@vmware.com wrote:
From: Roland Scheidegger
If we only feed one source vector at a time, we cannot use pack intrinsics
(as we only have a 64bit destination dst vector). lp_bld_conv_auto is
specifically designed to alter the length and number of
On 22/12/16 03:39, srol...@vmware.com wrote:
From: Roland Scheidegger
Generally we should do tranpose after conversion, if the format has less than
32 bits per channel (if it has 32 bits, conversion is going to be a no-op
anyway...). This is obviously because there's less
On Mon, Jan 02, 2017 at 06:37:18PM -0800, Ben Widawsky wrote:
> In the foreseeable future it doesn't seem to make sense to have multiple
> resolve flags. What does make sense is to have the caller give an
> indication to the lower layers what it things should be done for
> resolve. The enum change
On Wednesday, January 4, 2017 1:26:23 PM PST Iago Toral Quiroga wrote:
> We can use this to track various features that may or may not be supported
> by the hw / kernel. Currently, we usually do this by checking the generation
> and supported command parser versions in various places thoughtout
On Wednesday, January 4, 2017 1:26:20 PM PST Iago Toral Quiroga wrote:
> From: Chris Wilson
[snip]
> + /* And afterwards clear the register */
> + if (reset) {
> + *batch++ = MI_LOAD_REGISTER_IMM | (3 - 2);
> + *batch++ = reg;
> + *batch++ = 0;
> + }
On 17-01-04 22:55:17, Varad Gautam wrote:
Hi Ben,
On Wed, Jan 4, 2017 at 3:25 AM, Ben Widawsky wrote:
On 16-11-24 20:50:37, Varad Gautam wrote:
This is the second revision to the EGL_EXT_image_dma_buf_import_modifiers
[1]
series at [2], addressing the comments received.
On Wed, Jan 04, 2017 at 09:40:51AM +0200, Pohjolainen, Topi wrote:
> On Mon, Jan 02, 2017 at 06:37:10PM -0800, Ben Widawsky wrote:
> > This code will disable actually creating these buffers for the scanout,
> > but it puts the allocation in place.
> >
> > Primarily this patch is split out for
On Mon, Jan 02, 2017 at 06:37:22PM -0800, Ben Widawsky wrote:
> v2: Try to keep the assert as recommended by Topi. This requires
> modifying the num_samples check to be <= 1 because internally created
> buffers set num_samples = 0.
>
> Cc: Topi Pohjolainen
>
On Tue, 2017-01-03 at 14:41 -0800, Jason Ekstrand wrote:
> I made a few pretty trivial comments. With those addressed,
>
> Reviewed-by: Jason Ekstrand
>
> On Dec 16, 2016 8:55 AM, "Juan A. Suarez Romero" > wrote:
> So far, input_reads was a bitmap
On Mon, Jan 02, 2017 at 06:37:15PM -0800, Ben Widawsky wrote:
> Allows us to continue utilizing common miptree creation using __DRIimage
> without creating a new DRIimage (for the intel_process_dri2_buffer()
> case).
>
> This is a bit ugly, but I think it's the best one can do.
>
>
On Fri, Dec 23, 2016 at 11:04:51PM +0100, Christian Gmeiner wrote:
[...]
> +struct pipe_screen *imx_drm_screen_create(int fd)
> +{
> + struct renderonly ro = {
> + .create_for_resource = renderonly_create_kms_dumb_buffer_for_resource,
> + .kms_fd = fd,
> + .gpu_fd =
On Mon, Jan 02, 2017 at 06:37:20PM -0800, Ben Widawsky wrote:
> Upper layers of the code will have the need to specify full or partial
> resolves (more on this in the next patch). This code simply adds the new
> enums and plumbs it in as minimally as necessary.
>
> Signed-off-by: Ben Widawsky
On Jan 2, 2017 06:03, "Ilia Mirkin" wrote:
Signed-off-by: Ilia Mirkin
---
src/gallium/auxiliary/tgsi/tgsi_info.c | 2 +-
src/gallium/docs/source/tgsi.rst | 11 +++
src/gallium/include/pipe/p_shader_tokens.h | 2 +-
3 files
On Jan 4, 2017 12:46, "Lionel Landwerlin"
wrote:
On 04/01/17 18:16, Jason Ekstrand wrote:
On Jan 4, 2017 12:02, "Lionel Landwerlin"
wrote:
v2: Move relative push constant relative offset computation down to
I've got an updated patch out.
Note that the Python scripts were largely unmodified from the ones in
the libglvnd tree. I don't know if divergence between the two is
something anyone is likely to care about, though, and I'd expect the
EGL-specific changes to libglvnd's version to be pretty
---
configure.ac | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 799f5ebda029..d1ffb57f57e3 100644
--- a/configure.ac
+++ b/configure.ac
@@ -2382,7 +2382,7 @@ dnl
dnl Gallium helper functions
dnl
gallium_require_llvm() {
-if test
On 04/01/17 18:16, Jason Ekstrand wrote:
On Jan 4, 2017 12:02, "Lionel Landwerlin"
>
wrote:
v2: Move relative push constant relative offset computation down to
_vtn_load_store_tail() (Jason)
Hm... I may not
Reviewed-by: Matt Turner
signature.asc
Description: Digital signature
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
I need to read 8 for real but the rest seems sensible. 1-7 are
Reviewed-by: Jason Ekstrand
On Jan 4, 2017 10:49, "Jason Ekstrand" wrote:
> On Jan 4, 2017 05:08, "Kenneth Graunke" wrote:
>
> It's only used in one place, it
this avoids an extra copy which occurs in case of dri2
v1.1: fallback to dri2 if dri3 fails to initialize
Suggested-by: Christian König
Signed-off-by: Nayan Deshmukh
---
src/gallium/state_trackers/vdpau/presentation.c | 58
dri3 allows us to send handle of a texture directly to X
so this patch allows a state tracker to directly send its
texture to X to be used as back buffer and avoids extra
copying
v2: use clip width/height to display a portion of the surface
v3: remove redundant variables, fix wrapping, rename
the hack was introduced to avoid an extra copying
but now with dri3 we don't need it anymore
v1.1: rebasing
Signed-off-by: Nayan Deshmukh
---
src/gallium/state_trackers/vdpau/bitmap.c| 2 -
src/gallium/state_trackers/vdpau/device.c| 50 -
Ping?
On Dec 12, 2016 21:39, "Jason Ekstrand" wrote:
> ---
> src/compiler/Makefile.sources | 1 +
> src/compiler/nir/nir.h| 2 +
> src/compiler/nir/nir_opt_copy_prop_vars.c | 799
> ++
> 3 files changed, 802
On Wed, Jan 4, 2017 at 6:40 PM, Alex Deucher wrote:
> On Wed, Jan 4, 2017 at 5:47 AM, Marek Olšák wrote:
>> From: Marek Olšák
>>
>> The context may be used by texture_get_handle.
>
> The the omx state tracker need this as well?
The
The header file should, instead, have an extern "C" in it. Header
files should be safe to include in C++ code, most have extern "C"
guards, but some get missed.
Cheers,
-ilia
On Wed, Jan 4, 2017 at 12:41 PM, George Kyriazis
wrote:
> Explicitly declare
On Wed, Jan 4, 2017 at 5:47 AM, Marek Olšák wrote:
> From: Marek Olšák
>
> The context may be used by texture_get_handle.
The the omx state tracker need this as well?
Alex
> ---
> src/gallium/state_trackers/va/buffer.c | 7 +--
> 1 file changed, 5
On Jan 4, 2017 12:02, "Lionel Landwerlin"
wrote:
v2: Move relative push constant relative offset computation down to
_vtn_load_store_tail() (Jason)
Hm... I may not have been particularly clear. I meant, more or less, to
have get_io_offset (is that what its
Marek Olšák writes:
> From: Marek Olšák
>
> GLSL compilation now takes 24% less time with the Gallium noop driver.
> I used my shader-db for the measurement. The difference for the whole
> radeonsi driver can be ~10%.
>
> The generated TGSI is mostly the
Thanks for doing this!
On Dec 17, 2016 17:44, "Lionel Landwerlin"
wrote:
Signed-off-by: Lionel Landwerlin
---
src/compiler/spirv/vtn_variables.c | 85
+++--
Hi Ben,
On Wed, Jan 4, 2017 at 3:25 AM, Ben Widawsky wrote:
> On 16-11-24 20:50:37, Varad Gautam wrote:
>>
>> This is the second revision to the EGL_EXT_image_dma_buf_import_modifiers
>> [1]
>> series at [2], addressing the comments received. This diverges from v1 due
>> to
>>
On Wed, Jan 4, 2017 at 6:36 AM, Samuel Iglesias Gonsálvez
wrote:
> On Tue, 2017-01-03 at 12:14 -0500, Matt Turner wrote:
>> On Tue, Jan 3, 2017 at 7:27 AM, Samuel Iglesias Gonsálvez
>> wrote:
>> > Signed-off-by: Samuel Iglesias Gonsálvez
v2: Move relative push constant relative offset computation down to
_vtn_load_store_tail() (Jason)
Signed-off-by: Lionel Landwerlin
---
src/compiler/spirv/vtn_variables.c | 78 -
Explicitly declare lp_native_vector_width inside an extern "C",
since we cannot include the correct header file inside extern "C".
---
src/gallium/drivers/swr/swr_screen.cpp | 16 +---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git
Hi,
Tested-by: Eero Tamminen
May also impact other programs having float indexed loops:
- Invisible Inc
- Talos Principle
- Steam Big Picture
- Serious Sam 3
- Eero
On 04.01.2017 02:59, Timothy Arceri wrote:
Fixes performance regression in SynMark PSPom
Hello,
This series reworks i965's handling of gl_TessLevelInner/Outer[] arrays.
Instead of using lower_tess_levels to turn them into vec4/vec2s, we pass
them through to NIR and make them compact arrays (where array indexing
translates to enhanced layouts components).
This has some nice benefits.
On Tue, 2017-01-03 at 08:57 -0800, Jason Ekstrand wrote:
>
>
> On Dec 16, 2016 8:55 AM, "Juan A. Suarez Romero" > wrote:
> From: Samuel Iglesias Gonsálvez
>
> We use *64*_PASSTHRU formats to upload vertex attributes of 64 bits
> to avoid
https://bugs.freedesktop.org/show_bug.cgi?id=98833
--- Comment #10 from Eero Tamminen ---
Ping... Were you able to reproduce the issue with patched weston-simple-egl?
--
You are receiving this mail because:
You are the QA Contact for the
On Wed, 2017-01-04 at 12:36 +0100, Samuel Iglesias Gonsálvez wrote:
> On Tue, 2017-01-03 at 12:14 -0500, Matt Turner wrote:
> > On Tue, Jan 3, 2017 at 7:27 AM, Samuel Iglesias Gonsálvez
> > wrote:
> > > Signed-off-by: Samuel Iglesias Gonsálvez
> > >
From: Marek Olšák
The context may be used by texture_get_handle.
---
src/gallium/state_trackers/va/buffer.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/src/gallium/state_trackers/va/buffer.c
b/src/gallium/state_trackers/va/buffer.c
index
From: Marek Olšák
The context may be used by texture_get_handle.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99158
---
src/gallium/state_trackers/vdpau/output.c | 7 +--
src/gallium/state_trackers/vdpau/surface.c | 12
2 files changed, 13
Signed-off-by: Kenneth Graunke
---
src/compiler/glsl/glsl_to_nir.cpp | 13 -
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/src/compiler/glsl/glsl_to_nir.cpp
b/src/compiler/glsl/glsl_to_nir.cpp
index 4e0a33a9e74..8b1aec1b3b4 100644
---
It's only used in one place, it ignores the offset parameter currently,
and I want to add more parameters...at which point, passing in a bunch
of integers seems less obvious than writing it out.
Signed-off-by: Kenneth Graunke
---
This wraps glsl_type::count_attribute_slots(), but will soon contain a
couple of overrides for a couple of GLSL built-ins variables.
Signed-off-by: Kenneth Graunke
---
src/compiler/glsl/ir.cpp| 7 +++
src/compiler/glsl/ir.h | 2
Treating everything as scalar arrays allows us to drop a bunch of
special case input/output munging all throughout the backend.
Instead, we just need to remap the TessLevel components to the
appropriate patch URB header locations in remap_patch_urb_offsets().
We also switch to treating the TES
Upcoming reworks in i965 are going to make it easy to handle this
like any other input. Having it as a system value will just require
additional code for no benefit.
Signed-off-by: Kenneth Graunke
---
src/compiler/glsl/builtin_variables.cpp | 15 +++
Right now, this shouldn't have any effect, as all drivers use
LowerClipDist and LowerTessFactors to turn the float[] arrays into
vectors.
However, it should help make it possible for drivers to avoid that
lowering.
Signed-off-by: Kenneth Graunke
---
There's no point in trying to mark partial array access for
gl_ClipDistance, gl_TessLevelOuter, or gl_TessLevelInner - they're
special built-in variables that control fixed function hardware,
and will likely be used in an all-or-nothing fashion.
Since these arrays only occupy 1-2 varying slots,
1 - 100 of 116 matches
Mail list logo