This series came out of fixing dEQP failures on vc4's GLES2 context.
Mesa was allowing RGB565 textures, which is only valid with
GL_OES_required_internalformat. Rather than disable RGB565, I decided
the extension was easy enough to support.
I've sent one piglit test for renderbuffer sizing, and
This keeps us from promoting them up to , at the cost of not being
color-renderable.
---
src/gallium/drivers/vc4/vc4_formats.c | 5 ++---
src/gallium/drivers/vc4/vc4_uniforms.c | 1 +
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/gallium/drivers/vc4/vc4_formats.c
This extension is effectively a backport of GLES3's internalformat
handling to GLES 1/2. It guarantees that sized internalformats specified
for textures and renderbuffers have at least the specified size stored.
That's a pretty minimal requirement, so I think it can be dummy_true and
exposed as a
---
Do I have the swizzles right here? It's a bit complicated because I
have a reswizzle in vc4, so I may have just massaged things to work
out in my case. I tried a lot of combinations trying to specify BE
swizzles in a way that wouldn't assertion fail in the python script,
with no luck.
Previously, we were downconverting to automatically if the hardware
didn't suport it. However, with the advent of
GL_OES_required_internalformat, we have to actually store the
internalformats we advertise support for. And, it seems rather
disingenuous to advertise the extension if we don't
For supporting RGB5 in hardware with A in the low bit (vc4), we need this
format as well.
---
src/mesa/main/formats.c | 2 ++
src/mesa/main/formats.csv| 1 +
src/mesa/main/formats.h | 1 +
src/mesa/main/texformat.c| 1 +
src/mesa/swrast/s_texfetch.c | 1 +
5 files changed, 6
Samuel Iglesias Gonsálvez writes:
> El Viernes, 28 de abril de 2017 16:08:35 Francisco Jerez escribió:
>> Samuel Iglesias Gonsálvez writes:
>> > It was setting XYWZ swizzle and writemask to all uniforms, no matter if
>> > they were a vector or scalar,
On 05/02/2017 01:26 PM, Grazvydas Ignotas wrote:
On Tue, May 2, 2017 at 10:21 PM, Brian Paul > wrote:
Note, this is just some debug code.
---
src/gallium/auxiliary/draw/draw_pipe.c | 2 +-
1 file changed, 1 insertion(+), 1
Reviewed-by: Marek Olšák
Marek
On Tue, May 2, 2017 at 9:21 PM, Brian Paul wrote:
> ---
> src/mesa/main/api_validate.c | 2 +-
> src/mesa/main/api_validate.h | 2 +-
> 2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git
On Tue, May 2, 2017 at 10:21 PM, Brian Paul wrote:
> Note, this is just some debug code.
> ---
> src/gallium/auxiliary/draw/draw_pipe.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/gallium/auxiliary/draw/draw_pipe.c
>
---
src/mesa/main/api_validate.c | 2 +-
src/mesa/main/api_validate.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/mesa/main/api_validate.c b/src/mesa/main/api_validate.c
index bddde29..4694c36 100644
--- a/src/mesa/main/api_validate.c
+++
Note, this is just some debug code.
---
src/gallium/auxiliary/draw/draw_pipe.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/auxiliary/draw/draw_pipe.c
b/src/gallium/auxiliary/draw/draw_pipe.c
index 27a3762..3581698 100644
---
2017-04-28 13:11 GMT+02:00 Philipp Zabel :
> To restart interrupted system calls, use drmIoctl.
>
> Suggested-by: Emil Velikov
> Signed-off-by: Philipp Zabel
Reviewed-by: Christian Gmeiner
2017-04-28 12:19 GMT+02:00 Emil Velikov :
> On 27 April 2017 at 17:44, Philipp Zabel wrote:
>> The renderonly_scanout holds a reference on its prime pipe resource,
>> which should be released when it is destroyed. If it was created by
>>
2017-04-28 12:19 GMT+02:00 Emil Velikov :
> On 27 April 2017 at 17:44, Philipp Zabel wrote:
>> prime_fd is only used to transfer the scanout buffer to the GPU inside
>> renderonly_create_kms_dumb_buffer_for_resource. It should be closed
>>
Anuj Phogat writes:
> On Mon, Apr 24, 2017 at 9:15 PM, Ben Widawsky wrote:
>
>> On 17-04-18 18:18:39, Francisco Jerez wrote:
>>
>> Most, if not all of the unrelated changes that snuck in were due to rebase.
>> Anuj, would you mind fixing those? I tried
When we were assembling the unsigned 64-bit query return from its
two signed 32-bit component parts, the lower half was getting
sign-extended into the top half. Be more explicit about what we want to
do.
Fixes gbm_bo_get_modifier() returning ((1 << 64) - 1) rather than
((1 << 56) - 1), i.e.
On 2 May 2017 at 19:01, Rob Herring wrote:
> On Sun, Apr 30, 2017 at 4:52 AM, Emil Velikov
> wrote:
>> Hi Rob,
>>
>> On 27 April 2017 at 20:43, Rob Herring wrote:
>>
>>> --- a/src/gallium/targets/dri/Android.mk
>>> +++
On Tue, May 2, 2017 at 7:24 AM, Johnson Lin wrote:
> The matrix used for YCbCr to RGB is listed in Wiki https://en.wikipedia.org/
> wiki/YCbCr; There is minor error in the matrix constant: 0.0625=16/256 should
> be 16.0/255,and 0.5=128.0/256 should be 128.0/255.
> Note
On 2 May 2017 at 17:06, Rob Herring wrote:
> On Tue, May 2, 2017 at 10:14 AM, Emil Velikov
> wrote:
>> On 2 May 2017 at 15:41, Rob Herring wrote:
>>> On Mon, May 1, 2017 at 2:18 PM, Emil Velikov
>>> wrote:
"Kristian H. Kristensen" writes:
> Johnson Lin writes:
>
>> The matrix used for YCbCr to RGB is listed in Wiki
>> https://en.wikipedia.org/wiki/YCbCr;
>> There is minor error in the matrix constant: 0.0625=16/256 should be
>> 16.0/255,
>> and
Hi Eric,
On 2 May 2017 at 16:05, Eric Engestrom wrote:
> On Tuesday, 2017-05-02 11:52:06 +0100, Daniel Stone wrote:
>> --- a/common.c
>> +++ b/common.c
>> @@ -1,5 +1,6 @@
>> /*
>> * Copyright (c) 2017 Rob Clark
>> + * Copyright © 2013 Intel
Just like other type hash tables are destroyed in
_mesa_glsl_release_types(), also destroy the ones for function and
subroutine types.
Signed-off-by: Grazvydas Ignotas
---
src/compiler/glsl_types.cpp | 10 ++
1 file changed, 10 insertions(+)
diff --git
Since the enum is in the same header now, we can use it as the type of
the field.
Signed-off-by: Rafael Antognolli
---
PS: We can merge this with the previous patch too if that's better.
src/intel/compiler/brw_compiler.h | 2 +-
src/intel/compiler/brw_fs.cpp |
On Sun, Apr 30, 2017 at 4:52 AM, Emil Velikov wrote:
> Hi Rob,
>
> On 27 April 2017 at 20:43, Rob Herring wrote:
>
>> --- a/src/gallium/targets/dri/Android.mk
>> +++ b/src/gallium/targets/dri/Android.mk
>> @@ -105,10 +105,9 @@
This follows the same pattern as in _mesa_initialize_context(), but uses
pthread_once instead of a mutex. The goal is to avoid problems with
multiple instances.
Signed-off-by: Grazvydas Ignotas
---
src/intel/vulkan/anv_device.c | 19 ---
1 file changed, 16
Add a spirv_to_nir_fini() helper to clean up glsl_type allocations
resulting from spirv_to_nir() calls and call it on exit. This resolves
many reports from valgrind and should be helpful for application
developers that use valgrind.
Signed-off-by: Grazvydas Ignotas
---
Before we were doing RGBA4 on GLES3 only, but as of GLES2 2.0.22 it should
be RGBA4 as well. Fixes DEQP
functional.state_query.rbo.renderbuffer_internal_format.
---
src/mesa/main/renderbuffer.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/mesa/main/renderbuffer.c
Fixes DEQP's scoping.invalid.redeclare_function_fragment/vertex.
---
src/compiler/glsl/ast_to_hir.cpp | 10 ++
1 file changed, 10 insertions(+)
diff --git a/src/compiler/glsl/ast_to_hir.cpp b/src/compiler/glsl/ast_to_hir.cpp
index 9e02529dffb9..58088cec0d3a 100644
---
Here's a little set of changes for dEQP fixes for GLES2 contexts. I
haven't done a full run to confirm no regressions, as full runs on
hardware take a day or so. I'm hoping the Intel CI system might be
able to test these for me.
Eric Anholt (5):
glsl: Restrict functions to not return arrays
Fixes deqp_gles2 undefine_invalid_object_* failures.
---
src/compiler/glsl/glcpp/glcpp-parse.y | 7 ++-
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/src/compiler/glsl/glcpp/glcpp-parse.y
b/src/compiler/glsl/glcpp/glcpp-parse.y
index e113253061f6..5cb2a380605b 100644
---
The spec text cited above says you can't, but only the GLSL 3.00 (redefine
or overload) case was implemented.
Fixes dEQP scoping.invalid.redefine_builtin_fragment/vertex.
---
src/compiler/glsl/ast_to_hir.cpp | 15 +--
1 file changed, 13 insertions(+), 2 deletions(-)
diff --git
From the spec,
Arrays are allowed as arguments, but not as the return type. [...] The
return type can also be a structure if the structure does not contain
an array.
Fixes DEQP shaders.functions.invalid.return_array_in_struct_fragment.
---
src/compiler/glsl/ast_to_hir.cpp | 12
While we're at it, we could make the fields in
brw_wm_prog_data::computed_depth_mode use the actual enum type.
On Tue, May 2, 2017 at 10:26 AM, Jason Ekstrand
wrote:
> Reviewed-by: Jason Ekstrand
>
> On Tue, May 2, 2017 at 9:29 AM, Rafael Antognolli
Reviewed-by: Jason Ekstrand
On Tue, May 2, 2017 at 9:29 AM, Rafael Antognolli <
rafael.antogno...@intel.com> wrote:
> These enums live inside struct brw_wm_prog_data, so it makes sense to
> keep them in the same header. It also allows to use them without
> including
Samuel Iglesias Gonsálvez writes:
> On Mon, 2017-05-01 at 14:55 +0200, Samuel Iglesias Gonsálvez wrote:
>> El Viernes, 28 de abril de 2017 16:27:56 Francisco Jerez escribió:
>> > Samuel Iglesias Gonsálvez writes:
>> > > On gen7, the swizzles used in
On 05/02/2017 10:46 AM, Marek Olšák wrote:
switch (shader->Stage) {
case MESA_SHADER_VERTEX:
stvp = (struct st_vertex_program *)prog;
stvp->glsl_to_tgsi = v;
break;
case MESA_SHADER_FRAGMENT:
stfp = (struct st_fragment_program *)prog;
After successful drmGetDevices2() call, drmFreeDevices() needs to be
called.
Fixes: 743315f2 "radv: do not open random render node(s)"
Signed-off-by: Grazvydas Ignotas
---
src/amd/vulkan/radv_device.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git
drmGetDevices2 takes count and not size. Probably hasn't caused problems
yet in practice and was missed as setups with more than 8 DRM devices
are not very common.
Fixes: 743315f2 "radv: do not open random render node(s)"
Signed-off-by: Grazvydas Ignotas
---
After successful drmGetDevices2() call, drmFreeDevices() needs to be
called.
Fixes: b1fb6e8d "anv: do not open random render node(s)"
Signed-off-by: Grazvydas Ignotas
---
src/intel/vulkan/anv_device.c | 1 +
1 file changed, 1 insertion(+)
diff --git
drmGetDevices2 takes count and not size. Probably hasn't caused problems
yet in practice and was missed as setups with more than 8 DRM devices
are not very common.
Fixes: b1fb6e8d "anv: do not open random render node(s)"
Signed-off-by: Grazvydas Ignotas
---
On Tuesday, 2017-05-02 15:04:08 +0100, Daniel Stone wrote:
> The initial pattern of:
> initial_modeset();
> while (1) {
> pageflip();
> }
>
> was relying on getting a buffer despite not having rendered anything. It
> would also show undefined content. Replace this with just
On Tue, May 2, 2017 at 12:44 PM, Emil Velikov wrote:
> Hi Marek,
>
> Nice work.
>
> Small question - have you considered subclassing st_{vp,fp,cp}_program
> around st_common_program?
> It should allow us to remove even more code :-)
Yes, that would certainly be
>> switch (shader->Stage) {
>> case MESA_SHADER_VERTEX:
>> stvp = (struct st_vertex_program *)prog;
>> stvp->glsl_to_tgsi = v;
>> break;
>> case MESA_SHADER_FRAGMENT:
>> stfp = (struct st_fragment_program *)prog;
>> stfp->glsl_to_tgsi = v;
>>
These enums live inside struct brw_wm_prog_data, so it makes sense to
keep them in the same header. It also allows to use them without
including brw_eu_defines.h.
Signed-off-by: Rafael Antognolli
---
src/intel/compiler/brw_compiler.h | 21 +
On 29/04/17 14:51, Emil Velikov wrote:
From: Emil Velikov
Drop the misleading "will not match the one returned by thread_create"
hunk and provide more clarity as to what/why GetCurrentThread() isn't
the solution we're looking for.
Cc: José Fonseca
On Mon, Apr 24, 2017 at 9:15 PM, Ben Widawsky wrote:
> On 17-04-18 18:18:39, Francisco Jerez wrote:
>
> Most, if not all of the unrelated changes that snuck in were due to rebase.
> Anuj, would you mind fixing those? I tried my best to address the rest,
> but I'm
> admittedly
Thanks. I'm amending this:
diff --git a/src/gallium/state_trackers/nine/device9.c
b/src/gallium/state_trackers/nine/device9.c
index 5dc24d6..f26552c 100644
--- a/src/gallium/state_trackers/nine/device9.c
+++ b/src/gallium/state_trackers/nine/device9.c
@@ -2913,6 +2913,7 @@
On Tue, May 2, 2017 at 10:14 AM, Emil Velikov wrote:
> On 2 May 2017 at 15:41, Rob Herring wrote:
>> On Mon, May 1, 2017 at 2:18 PM, Emil Velikov
>> wrote:
>>> Hi Rob,
>>>
>>> On 27 April 2017 at 20:43, Rob Herring
Please prefix with "nir/lower_tex:" How about
nir/lower_tex: Fix minor error in YUV color conversion matrix
On Tue, May 2, 2017 at 7:24 AM, Johnson Lin wrote:
> The matrix used for YCbCr to RGB is listed in Wiki
> https://en.wikipedia.org/
> wiki/YCbCr; There is minor
On Tue, May 2, 2017 at 8:55 AM, Rafael Antognolli <
rafael.antogno...@intel.com> wrote:
> On Tue, May 02, 2017 at 08:44:05AM -0700, Jason Ekstrand wrote:
> > On Tue, May 2, 2017 at 8:28 AM, Rafael Antognolli <
> rafael.antogno...@intel.com>
> > wrote:
> >
> > On Tue, May 02, 2017 at
On Tue, May 02, 2017 at 08:44:05AM -0700, Jason Ekstrand wrote:
> On Tue, May 2, 2017 at 8:28 AM, Rafael Antognolli
>
> wrote:
>
> On Tue, May 02, 2017 at 07:26:53AM -0700, Jason Ekstrand wrote:
> > On Mon, May 1, 2017 at 6:43 PM, Rafael Antognolli <
>
On 29/04/17 14:24, Emil Velikov wrote:
Hi Ben,
On 27 April 2017 at 21:31, Ben Boeckel wrote:
LLVMDemangle, LLVMGlobalISel, and LLVMDebugInfoMSF are new.
Also update the comment to add irreader to the list of components.
CC:
On Tue, May 2, 2017 at 8:28 AM, Rafael Antognolli <
rafael.antogno...@intel.com> wrote:
> On Tue, May 02, 2017 at 07:26:53AM -0700, Jason Ekstrand wrote:
> > On Mon, May 1, 2017 at 6:43 PM, Rafael Antognolli <
> rafael.antogno...@intel.com>
> > wrote:
> >
> > We need to use some enums inside
On 2017-05-02 15:32, Emil Velikov wrote:
Carlos, care to throw a comment why it's using nmem and not something
else - say is_dmabuf_mem?
Hm, I do see a potential problem now. The code that maps/unmaps the
buffer and calls buf_to_fd() should also set nmems to 1. Then the
close() loop is
On Tue, May 02, 2017 at 07:26:53AM -0700, Jason Ekstrand wrote:
> On Mon, May 1, 2017 at 6:43 PM, Rafael Antognolli
>
> wrote:
>
> We need to use some enums inside genX_state_upload.c, but including the
> whole header will cause several conflicts between
On 2 May 2017 at 15:41, Rob Herring wrote:
> On Mon, May 1, 2017 at 2:18 PM, Emil Velikov wrote:
>> Hi Rob,
>>
>> On 27 April 2017 at 20:43, Rob Herring wrote:
>
> [...]
>
>>> diff --git a/src/gallium/Android.mk
On Tuesday, 2017-05-02 11:52:06 +0100, Daniel Stone wrote:
> eglGetProcAddress is allowed to return any old garbage for symbols it
> doesn't know about. To avoid any mishaps, check for the appropriate
> extension presence (split into EGL client extension, EGL display
> extension, and GL extension,
On 05/02/2017 12:12 AM, Timothy Arceri wrote:
On 01/05/17 19:57, Samuel Pitoiset wrote:
Thanks for fixing this. I guess "glsl: set vector_elements to 1 for
samplers" can now be applied without breaking the world? :-)
Yes you should be able to push it again (although wouldn't hurt to give
The matrix used for YCbCr to RGB is listed in Wiki https://en.wikipedia.org/
wiki/YCbCr; There is minor error in the matrix constant: 0.0625=16/256 should
be 16.0/255,and 0.5=128.0/256 should be 128.0/255.
Note that conversion from a 0-255 byte number to 0-1.0 float is to divide by
255 instead of
On 2 May 2017 at 15:01, Daniel Stone wrote:
> Hey Emil,
>
> On 2 May 2017 at 14:24, Emil Velikov wrote:
>> Seems to be the correct place to handle here, despite that both are
>> already handled.
>
> Honestly, I wouldn't have written this patch if I
On 2 May 2017 at 15:39, Lucas Stach wrote:
> With the modifier support it is possible to force a modifier which isn't
> compatible with any of the GPU targets. In that case the EGL platform will
> be unable to allocate the back buffers.
>
> Provide an error message in that
The TGSI DCE pass doesn't eliminate dead assignments like
MOV TEMP[0], TEMP[1] in presence of loops because it assumes
that the visitor doesn't emit dead code. This assumption is
actually wrong and this situation happens.
However, it appears that the merge_registers() pass accidentally
takes care
On Mon, May 1, 2017 at 2:18 PM, Emil Velikov wrote:
> Hi Rob,
>
> On 27 April 2017 at 20:43, Rob Herring wrote:
[...]
>> diff --git a/src/gallium/Android.mk b/src/gallium/Android.mk
>> index 7c6bda68d59f..d591aaf62e6a 100644
>> ---
With the modifier support it is possible to force a modifier which isn't
compatible with any of the GPU targets. In that case the EGL platform will
be unable to allocate the back buffers.
Provide an error message in that case, instead of crashing.
Signed-off-by: Lucas Stach
Mostly to pull in the Vivante tiling format modifiers, but some other
little changes included.
Signed-off-by: Lucas Stach
---
include/drm/drm_fourcc.h | 81
1 file changed, 81 insertions(+)
diff --git
On 2 May 2017 at 14:58, Rob Clark wrote:
> On Tue, May 2, 2017 at 9:32 AM, Emil Velikov wrote:
>> From: Emil Velikov
>>
>> Makes the code shorter and more explicit.
>>
>> Cc: Carlos Rafael Giani
On 05/02/2017 05:19 AM, Bartosz Tomczyk wrote:
We always use only single element.
v2: Change signle element arrays to variables
"single"
Looks good.
Reviewed-by: Brian Paul
---
src/mesa/vbo/vbo_exec_array.c | 74 +--
1 file
On Mon, May 1, 2017 at 6:43 PM, Rafael Antognolli <
rafael.antogno...@intel.com> wrote:
> We need to use some enums inside genX_state_upload.c, but including the
> whole header will cause several conflicts between things defined in this
> header and the genxml auto-generated headers.
>
> So
From: Marek Olšák
Cc: 17.1
---
src/gallium/drivers/radeonsi/si_shader.c| 43 +
src/gallium/drivers/radeonsi/si_state_shaders.c | 13 ++--
2 files changed, 48 insertions(+), 8 deletions(-)
diff --git
From: Marek Olšák
same as Vulkan
---
src/gallium/drivers/radeonsi/si_state.c | 10 +++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/src/gallium/drivers/radeonsi/si_state.c
b/src/gallium/drivers/radeonsi/si_state.c
index f50855d..00f1958 100644
---
On Tue, May 02, 2017 at 09:38:53AM +0100, Emil Velikov wrote:
> On 2 May 2017 at 09:32, Emil Velikov wrote:
> > Hi Rafael,
> >
> > On 2 May 2017 at 02:43, Rafael Antognolli
> > wrote:
> >> We need to use some enums inside
The initial pattern of:
initial_modeset();
while (1) {
pageflip();
}
was relying on getting a buffer despite not having rendered anything. It
would also show undefined content. Replace this with just using the
pageflip loop for the initial modeset as well.
Signed-off-by:
Hey Emil,
On 2 May 2017 at 14:24, Emil Velikov wrote:
> Seems to be the correct place to handle here, despite that both are
> already handled.
Honestly, I wouldn't have written this patch if I knew that; Lucas's
fix came in after the last time I'd pulled, and I didn't
On Tue, May 2, 2017 at 9:32 AM, Emil Velikov wrote:
> From: Emil Velikov
>
> Makes the code shorter and more explicit.
>
> Cc: Carlos Rafael Giani
> Signed-off-by: Emil Velikov
> ---
>
Hi,
On 2 May 2017 at 14:49, Rob Clark wrote:
> I think this is papering over an i965 bug.. in fact doing the
> DupNativeFenceFD could trigger a flush which you don't actually want
> on a tiler.
Well, when it's immediately before SwapBuffers, I don't think it
matters that
On Tue, May 2, 2017 at 6:52 AM, Daniel Stone wrote:
> Duplicating the fence before we flush means that the implementation can
> know that the fence needs to be CPU-visible. After flush it may be too
> late.
I think this is papering over an i965 bug.. in fact doing the
Juan,
Were you planning to review the rest of the series? Just wondering.
--Jason
On April 26, 2017 7:35:29 AM Jason Ekstrand wrote:
This absurdly long series does something fairly simple: It pulls the
block_pool into the state_pool and makes the state pool capable
Iago,
I'm still missing reviews on one or two of the patches here. Mind giving
them a quick look?
--Jason
On April 27, 2017 9:31:55 AM Jason Ekstrand wrote:
This is mostly a re-send of my earlier patches but there are a few changes.
I think, at this point, that I'm
From: Emil Velikov
Makes the code shorter and more explicit.
Cc: Carlos Rafael Giani
Signed-off-by: Emil Velikov
---
Worth moving the plane[i] bits as well?
The loop guard for the close() looks a big funky.
From: Emil Velikov
Assert in the very unlikely case when GST_VIDEO_INFO_N_PLANES returns
more than MAX_NUM_PLANES. Otherwise we'll get random memory corruption.
Signed-off-by: Emil Velikov
---
gst-decoder.c | 2 ++
1 file changed, 2
On 2 May 2017 at 11:52, Daniel Stone wrote:
> ...
>
> Signed-off-by: Daniel Stone
> ---
> drm-atomic.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drm-atomic.c b/drm-atomic.c
> index 25e04b1..6fb40fd 100644
> --- a/drm-atomic.c
> +++
I made a comment in patch 2 that I think needs to be fixed. With that
fixed patches 2, 3 and 8 are:
Reviewed-by: Iago Toral Quiroga
I also dropped a comment in patch 9 mostly to double-check that there
are no shenanigans with the annotation counts, in which case you can
add
On Mon, 2017-05-01 at 13:54 -0700, Matt Turner wrote:
> Which will allow us to print validation errors found in shader
> assembly
> in GPU hang error states.
> ---
> src/intel/tools/disasm.c | 71 +-
> --
> 1 file changed, 43 insertions(+), 28
2017-05-01 14:52 GMT+02:00 Marek Olšák :
> From: Marek Olšák
SNIP
>
> /**
> * Derived from Mesa gl_program:
> */
> -struct st_geometry_program
> +struct st_common_program
> {
> struct gl_program Base; /**< The Mesa geometry program */
I think
On Mon, 2017-05-01 at 13:54 -0700, Matt Turner wrote:
> This will allow us to more easily run brw_validate_instructions() on
> shader programs we find in GPU hang error states.
> ---
> src/intel/compiler/brw_eu.h | 3 ++-
> src/intel/compiler/brw_eu_validate.c | 10 --
From: Emil Velikov
Thus we can use the value to explicitly size arrays, instead of
__DRI_ATTRIB_FRAMEBUFFER_SRGB_CAPABLE + 1.
The latter seems magical and is error prone, as we add more dri
attributes.
v2: Fix off by one error (Tomasz)
Cc: Tomasz Figa
On Tue, May 02, 2017 at 12:34:44PM +0100, Daniel Stone wrote:
> Hi,
>
> On 2 May 2017 at 12:31, Chris Wilson wrote:
> > On Tue, May 02, 2017 at 11:52:07AM +0100, Daniel Stone wrote:
> >> + /* Wait on the CPU side for the _previous_ commit to
> >> +
On 1 May 2017 at 15:55, Tomasz Figa wrote:
> On Mon, May 1, 2017 at 11:17 PM, Mauro Rossi wrote:
>> Hi all,
>>
>> another try to merge android swrast patches in mesa 17.1 or mesa-dev
>> if they are somehow considered useful for android.
>>
>> Mauro
>>
Hi,
On 2 May 2017 at 12:31, Chris Wilson wrote:
> On Tue, May 02, 2017 at 11:52:07AM +0100, Daniel Stone wrote:
>> + /* Wait on the CPU side for the _previous_ commit to
>> + * complete before we post the flip through KMS, as
>>
On Tue, May 02, 2017 at 11:52:07AM +0100, Daniel Stone wrote:
> Atomic does not let us have multiple requests in flight at once; if we
> don't synchronise with the request completion on the CPU side, we can
> get -EBUSY from the atomic commit.
>
> We already have everything required to do this,
We always use only single element.
v2: Change signle element arrays to variables
---
src/mesa/vbo/vbo_exec_array.c | 74 +--
1 file changed, 37 insertions(+), 37 deletions(-)
diff --git a/src/mesa/vbo/vbo_exec_array.c b/src/mesa/vbo/vbo_exec_array.c
index
Sure, I will send updated patch soon.
On 02.05.2017 13:03, Nicolai Hähnle wrote:
On 02.05.2017 12:37, Bartosz Tomczyk wrote:
We always use only single element.
Can you just change prim to not be an array at all in that case?
Thanks,
Nicolai
---
src/mesa/vbo/vbo_exec_array.c | 4 ++--
1
On 02.05.2017 12:37, Bartosz Tomczyk wrote:
We always use only single element.
Can you just change prim to not be an array at all in that case?
Thanks,
Nicolai
---
src/mesa/vbo/vbo_exec_array.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git
On 02.05.2017 08:18, Dave Airlie wrote:
On 1 May 2017 at 22:43, Marek Olšák wrote:
From: Marek Olšák
Only very few shaders have them (from 48486 shaders):
For the series:
Reviewed-by: Dave Airlie
(you might want to stop radeonsi
Hi,
On 2 May 2017 at 11:52, Daniel Stone wrote:
> These four patches get kmscube -A working for me on Skylake. Probably
> the kicker is #4, though #2 does seem to be necessary. I'm not sure if
> #3 is perfectly correct for all drivers; I think Intel will still work
>
Hi,
These four patches get kmscube -A working for me on Skylake. Probably
the kicker is #4, though #2 does seem to be necessary. I'm not sure if
#3 is perfectly correct for all drivers; I think Intel will still work
duplicating after the fence, at the cost of creating an extra batch to
fence on.
Atomic does not let us have multiple requests in flight at once; if we
don't synchronise with the request completion on the CPU side, we can
get -EBUSY from the atomic commit.
We already have everything required to do this, namely the out-fence
from the last commit. Block on that with a CPU-side
...
Signed-off-by: Daniel Stone
---
drm-atomic.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drm-atomic.c b/drm-atomic.c
index 25e04b1..6fb40fd 100644
--- a/drm-atomic.c
+++ b/drm-atomic.c
@@ -368,6 +368,9 @@ const struct drm * init_drm_atomic(const char
Duplicating the fence before we flush means that the implementation can
know that the fence needs to be CPU-visible. After flush it may be too
late.
Signed-off-by: Daniel Stone
---
drm-atomic.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git
101 - 200 of 215 matches
Mail list logo