On 24.07.2015 13:01, Dave Airlie wrote:
> From: Dave Airlie
>
> This adds support for fine derivatives and enables
> ARB_derivative_control on radeonsi.
>
> (just fell out of my working out interpolation)
>
> v2: cleanup some bits, and try and write an
> explaination.
Spelling: "explanation"
Jason Ekstrand writes:
> Ok, I've looked through this again and convinced myself that it's
> *mostly* correct. I am a bit skeptical of the address swizzling for
> Y-major tiling.
>
> I've also included some comments that I'd like to see added (assuming
> they're correct). Sometimes it's easier
On 07/23/2015 05:39 PM, Jason Ekstrand wrote:
> On Thu, Jul 23, 2015 at 1:01 AM, Eduardo Lima Mitev wrote:
>> On 07/23/2015 05:20 AM, Jason Ekstrand wrote:
>>> On Wed, Jul 22, 2015 at 4:37 AM, Eduardo Lima Mitev
>>> wrote:
On 07/13/2015 01:57 PM, Jason Ekstrand wrote:
> On Wed, Jul 8, 2
On 07/23/2015 05:39 PM, Jason Ekstrand wrote:
> On Thu, Jul 23, 2015 at 1:01 AM, Eduardo Lima Mitev wrote:
>> On 07/23/2015 05:20 AM, Jason Ekstrand wrote:
>>> On Wed, Jul 22, 2015 at 4:37 AM, Eduardo Lima Mitev
>>> wrote:
On 07/13/2015 01:57 PM, Jason Ekstrand wrote:
> On Wed, Jul 8, 2
Kenneth Graunke writes:
> On Saturday, July 18, 2015 05:34:47 PM Francisco Jerez wrote:
>> Each logical variant is largely equivalent to the original opcode but
>> instead of taking a single payload source it expects the arguments
>> separately as individual sources, like:
>>
>> tex_logical dst
Hi,
I have been looking a bit into the vec4 spilling code and this series
implements a few improvements. The main changes are in patches 1 and 4,
that add small optimizations. The remaining patches are all minor changes.
Also, I noticed that enabling spilling of everything (which is what I used
When we have code such as this:
mov vgrf1.0.x:F, vgrf2.:F
mov vgrf3.0.x:F, vgrf1.:F
...
mov vgrf3.0.x:F, vgrf1.:F
And vgrf1 is chosen for spilling, we can emit this:
mov vgrf1.0.x:F, vgrf2.:F
gen4_scratch_write hw_reg0:F, vgrf1.:D, 22D
mov vgrf3.0.x:F, vgrf1.:F
...
gen4_s
---
src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_vec4.cpp| 2 +-
src/mesa/drivers/dri/i965/intel_debug.c | 3 ++-
src/mesa/drivers/dri/i965/intel_debug.h | 5 +++--
4 files changed, 7 insertions(+), 5 deletions(-)
diff --g
In theory, GRF array access should have been moved to scratch by the time
we got here, so this should never happen. A full piglit run forcing
spilling of all registers seems to confirm this. The FS backend
does not seem to check for this either.
---
src/mesa/drivers/dri/i965/brw_vec4_reg_allocate.
Larger registers should have been moved to scratch (like GRF array access)
or split to size 1 by the split_virtual_grfs pass.
---
src/mesa/drivers/dri/i965/brw_vec4_reg_allocate.cpp | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/brw_vec4_reg_allocat
---
src/mesa/drivers/dri/i965/brw_vec4_reg_allocate.cpp | 14 +-
1 file changed, 9 insertions(+), 5 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_vec4_reg_allocate.cpp
b/src/mesa/drivers/dri/i965/brw_vec4_reg_allocate.cpp
index cd89edd..a9bf0d8 100644
--- a/src/mesa/drivers
Francisco Jerez writes:
> Kenneth Graunke writes:
>
>> On Saturday, July 18, 2015 05:34:47 PM Francisco Jerez wrote:
>>> Each logical variant is largely equivalent to the original opcode but
>>> instead of taking a single payload source it expects the arguments
>>> separately as individual sourc
Iago Toral Quiroga writes:
> When we have code such as this:
>
> mov vgrf1.0.x:F, vgrf2.:F
> mov vgrf3.0.x:F, vgrf1.:F
> ...
> mov vgrf3.0.x:F, vgrf1.:F
>
> And vgrf1 is chosen for spilling, we can emit this:
>
> mov vgrf1.0.x:F, vgrf2.:F
> gen4_scratch_write hw_reg0:F, vgrf1.
Iago Toral Quiroga writes:
> Larger registers should have been moved to scratch (like GRF array access)
> or split to size 1 by the split_virtual_grfs pass.
Not necessarily. split_virtual_grfs() won't be able to split stuff
which is read or written at once by the same instruction -- E.g. by
sen
This hasn't been overallocating space for the header for a long time.
It still leaves the header uninitialized though until the generator
fixes it.
---
src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/br
R-B me
On Jul 24, 2015 6:52 AM, "Francisco Jerez" wrote:
> This hasn't been overallocating space for the header for a long time.
> It still leaves the header uninitialized though until the generator
> fixes it.
> ---
> src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 5 +++--
> 1 file changed, 3 i
The commit subject line doesn't seem to match the code.
The code looks good though.
Reviewed-by: Brian Paul
On 07/23/2015 06:40 PM, Ilia Mirkin wrote:
Commit 17f714836 (mesa: rearrange texture error checking order) moved
the width/height/depth == 0 allowance before checking if the image was
On Jul 24, 2015 4:00 AM, "Francisco Jerez" wrote:
>
> Jason Ekstrand writes:
>
> > Ok, I've looked through this again and convinced myself that it's
> > *mostly* correct. I am a bit skeptical of the address swizzling for
> > Y-major tiling.
> >
> > I've also included some comments that I'd like
This adds a new vfunc to _EGLDriver, QueryDeviceName, which should
return a const string of the device name (usually in the format
'/dev/dri/cardN').
The EGLDevice could perhaps be cached in the EGLDisplay but there
usually aren't loads of devices and this lookup isn't particularly
costly so leave
Signed-off-by: Jonny Lamb
---
src/egl/main/eglapi.c | 47 +++
src/egl/main/eglglobals.c | 5 +
2 files changed, 52 insertions(+)
diff --git a/src/egl/main/eglapi.c b/src/egl/main/eglapi.c
index 824e51e..6c1e6c7 100644
--- a/src/egl/main/eglapi
This code is copied nearly verbatim from src/loader/loader.c. It
should be put somewhere so both files can reference the same code
without copy & pasting.
Signed-off-by: Jonny Lamb
---
src/egl/main/egldevice.c | 54
1 file changed, 54 insertions(+
Right now the _EGLDevice struct has nothing of interest apart from a
next pointer but this is what will be opaquely returned to clients in
eglQueryDevicesEXT.
The _EGLDeviceInfo struct is held in _eglGlobal and thanks to atexit()
it is cleaned up appropriately too.
Signed-off-by: Jonny Lamb
---
Enumerate udev devices by using:
SUBSYSTEM=="drm"
DEVTYPE=="drm_minor"
TAG=="master-of-seat"
Signed-off-by: Jonny Lamb
---
src/egl/main/eglapi.c| 7 ++-
src/egl/main/egldevice.c | 161 ++-
src/egl/main/egldevice.h | 7 ++-
3 files change
EGL_EXT_platform_query only defines one value for at the
moment, so just implement that.
Signed-off-by: Jonny Lamb
---
src/egl/main/eglapi.c| 9 -
src/egl/main/egldevice.c | 30 ++
src/egl/main/egldevice.h | 4
3 files changed, 42 insertions(+), 1
Here is an initial patchset implementing EGL_EXT_device_base (in
practice nothing to do), device_enumeration (for listing devices), and
device_query (for querying information about a device).
I worked on this because I want to then move onto adding an attribute
for eglGetPlatformDisplay so one can
EGL_EXT_device_query itself doesn't define any valid values for
so leave this function looking simple for values in the
future.
Signed-off-by: Jonny Lamb
---
src/egl/main/eglapi.c| 9 -
src/egl/main/egldevice.c | 17 +
src/egl/main/egldevice.h | 5 +
3 files ch
This is useful to ensure EGLDevices given to functions (such as
eglQueryDeviceAttrib or eglQueryDeviceString) are valid.
Signed-off-by: Jonny Lamb
---
src/egl/main/eglapi.c| 6 ++
src/egl/main/egldevice.c | 26 ++
src/egl/main/egldevice.h | 4
3 files chang
Reviewed-by: Marek Olšák
Marek
On Fri, Jul 24, 2015 at 6:43 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> This adds the frontend support, however the llvm
> backend produces the wrong pattern, however
> we can conditionalise enabling ARB_gpu_shader5
> on whatever version of llvm we fix this i
Jason Ekstrand writes:
> On Jul 24, 2015 4:00 AM, "Francisco Jerez" wrote:
>>
>> Jason Ekstrand writes:
>>
>> > Ok, I've looked through this again and convinced myself that it's
>> > *mostly* correct. I am a bit skeptical of the address swizzling for
>> > Y-major tiling.
>> >
>> > I've also in
On Fri, Jul 24, 2015 at 7:39 AM, Francisco Jerez wrote:
> Jason Ekstrand writes:
>
>> On Jul 24, 2015 4:00 AM, "Francisco Jerez" wrote:
>>>
>>> Jason Ekstrand writes:
>>>
>>> > Ok, I've looked through this again and convinced myself that it's
>>> > *mostly* correct. I am a bit skeptical of the
Jason Ekstrand writes:
> On Fri, Jul 24, 2015 at 7:39 AM, Francisco Jerez
> wrote:
>> Jason Ekstrand writes:
>>
>>> On Jul 24, 2015 4:00 AM, "Francisco Jerez" wrote:
Jason Ekstrand writes:
> Ok, I've looked through this again and convinced myself that it's
> *mostly*
On Jul 24, 2015 8:02 AM, "Francisco Jerez" wrote:
>
> Jason Ekstrand writes:
>
> > On Fri, Jul 24, 2015 at 7:39 AM, Francisco Jerez
wrote:
> >> Jason Ekstrand writes:
> >>
> >>> On Jul 24, 2015 4:00 AM, "Francisco Jerez"
wrote:
>
> Jason Ekstrand writes:
>
> > Ok, I've loo
On Wed, Jul 22, 2015 at 12:51 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> This is part of ARB_gpu_shader5, and this passes
> all the piglit tests currently available.
>
> Signed-off-by: Dave Airlie
> ---
> docs/GL3.txt | 2 +-
> src/gallium/drivers/radeonsi/si_s
Reviewed-by: Marek Olšák
Marek
On Fri, Jul 24, 2015 at 6:43 AM, Dave Airlie wrote:
> From: Dave Airlie
>
> This is required as part of ARB_gpu_shader5.
>
> no backend changes are required for this, or if
> any are, it's the same ones as for samplers.
>
> v2: use get_indirect_index (Marek)
>
>
Jason Ekstrand writes:
> On Jul 24, 2015 8:02 AM, "Francisco Jerez" wrote:
>>
>> Jason Ekstrand writes:
>>
>> > On Fri, Jul 24, 2015 at 7:39 AM, Francisco Jerez
> wrote:
>> >> Jason Ekstrand writes:
>> >>
>> >>> On Jul 24, 2015 4:00 AM, "Francisco Jerez"
> wrote:
>>
>> Jason Ekstra
On Fri, Jul 24, 2015 at 9:55 AM, Brian Paul wrote:
> The commit subject line doesn't seem to match the code.
It matches the code if you read the whole function... I think. Right
now if there's no image, it'll succeed, whereas after this change,
it'll return INVALID_OPERATION.
Happy to use a diff
Define a function to calculate the memory address of the image
location given by a vector of coordinates. This is required in cases
where we need to fall back to untyped surface access, which take a raw
memory offset and know nothing about surface coordinates, type
conversion or memory tiling and
From: Marek Olšák
Fixes piglit:
spec@glsl-1.30@execution@fs-texture-sampler2dshadow-10
spec@glsl-1.30@execution@fs-texture-sampler2dshadow-11
v2: use st_shader_stage_to_ptarget
---
src/mesa/state_tracker/st_atom_texture.c | 74 +-
src/mesa/state_tracker/st_
From: Marek Olšák
Fixes piglit:
spec@arb_texture_buffer_range@ranges-2
Cc: mesa-sta...@lists.freedesktop.org
---
src/mesa/state_tracker/st_atom_texture.c | 10 +-
src/mesa/state_tracker/st_context.c | 1 +
src/mesa/state_tracker/st_context.h | 1 +
3 files changed, 7 inser
On 14/05/15 20:38, Ian Romanick wrote:
Hi,
today I was able to go back to work on this patch. Sorry for the delay.
I changed the subject to the problem, as previous fix was wrong. I have
some questions (see below).
>>
>>
>> --- a/src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp
>> +++ b/src/mesa/d
Much better.
Reviewed-by: Jason Ekstrand
On Fri, Jul 24, 2015 at 9:33 AM, Francisco Jerez wrote:
> Define a function to calculate the memory address of the image
> location given by a vector of coordinates. This is required in cases
> where we need to fall back to untyped surface access, which
Reviewed-by: Brian Paul
On 07/24/2015 11:48 AM, Marek Olšák wrote:
From: Marek Olšák
Fixes piglit:
spec@arb_texture_buffer_range@ranges-2
Cc: mesa-sta...@lists.freedesktop.org
---
src/mesa/state_tracker/st_atom_texture.c | 10 +-
src/mesa/state_tracker/st_context.c | 1 +
On Thu, Jul 23, 2015 at 3:16 AM, Eduardo Lima Mitev wrote:
> These methods are essential for the implementation of the NIR->vec4 pass. They
> work similar to their fs_nir counter-parts.
>
> When processing instructions, these methods are invoked to resolve the
> brw registers (source or destinatio
Signed-off-by: Ilia Mirkin
---
There seems to be a huge overlap between what initialize_context and
initialize_context_to_defaults do, wasn't sure where to stick what.
src/glsl/main.cpp | 26 +-
1 file changed, 25 insertions(+), 1 deletion(-)
diff --git a/src/glsl/main.
https://bugs.freedesktop.org/show_bug.cgi?id=91456
Bug ID: 91456
Summary: Mesa won't compile with llvm 3.8
Product: Mesa
Version: git
Hardware: x86-64 (AMD64)
OS: Linux (All)
Status: NEW
Severity: major
https://bugs.freedesktop.org/show_bug.cgi?id=85712
--- Comment #4 from Steven Stewart-Gallus ---
A similar problem can happen where eglSwapBuffer returns EGL_FAILURE but then
sets the result of eglGetError to EGL_SUCCESS when the window is concurrently
closed.
--
You are receiving this mail bec
On 07/24/2015 10:05 AM, Ilia Mirkin wrote:
On Fri, Jul 24, 2015 at 9:55 AM, Brian Paul wrote:
The commit subject line doesn't seem to match the code.
It matches the code if you read the whole function... I think. Right
now if there's no image, it'll succeed, whereas after this change,
it'll r
The 420pack extension enables various GLSL rules that need to be applied
to any GLSL 4.20+ shader even if the extension is not explicitly
enabled.
Signed-off-by: Ilia Mirkin
---
src/glsl/glsl_parser.yy | 18 +-
src/glsl/glsl_parser_extras.h | 5 +
2 files changed, 14 i
Reviewed-by: Chris Forbes
On Sat, Jul 25, 2015 at 9:07 AM, Ilia Mirkin wrote:
> The 420pack extension enables various GLSL rules that need to be applied
> to any GLSL 4.20+ shader even if the extension is not explicitly
> enabled.
>
> Signed-off-by: Ilia Mirkin
> ---
> src/glsl/glsl_parser.yy
On Fri, Jul 24, 2015 at 12:19 PM, Jason Ekstrand wrote:
> On Thu, Jul 23, 2015 at 3:16 AM, Eduardo Lima Mitev wrote:
>> These methods are essential for the implementation of the NIR->vec4 pass.
>> They
>> work similar to their fs_nir counter-parts.
>>
>> When processing instructions, these metho
I think we already agreed to just copy data.location into
data.driver_location and we don't need this special-casing.
Just making a note of it as I review.
--Jason
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> Non-scalar backends like i965's NIR-vec4 need the orginal variable's v
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> The destination register from the instruction is stored in the output_reg
> variable at its original varying value. From there, vec4_visitor's
> emit_urb_slot() will pick it up and continue the URB setup code, so that
> part is shared.
>
Add in missed version checks in the GLSL parser
Signed-off-by: Ilia Mirkin
---
src/glsl/glsl_parser.yy | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index 4cce5b8..9ddf5ed 100644
--- a/src/glsl/glsl_parser.yy
+++ b/
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> From: Alejandro Piñeiro
>
> Uses the nir structure to get all the info needed (sources,
> dest reg, etc), and then it uses the common
> vec4_visitor::emit_texture to emit the final code.
> ---
> src/mesa/drivers/dri/i965/brw_vec4_nir.c
This patch needs to go *before* patch 71 so things continue to build.
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> From: Samuel Iglesias Gonsalvez
>
> So the implementation is independent of GLSL IR and the visit methods of the
> gen6 GS visitor. This way we will be able to reuse
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> From: Iago Toral Quiroga
>
> For the indirect case we need to take the index delivered by
> NIR and compute the parent uniform that we are accessing (the one
> that we uploaded to a surface) and the constant offset into that
> surface.
On Thu, Jul 23, 2015 at 3:17 AM, Eduardo Lima Mitev wrote:
> From: Alejandro Piñeiro
>
> Splitted in two. The emission is moved to a new vec4_visitor
> method, vec4_visitor::emit_texture, ir order to be reused
> on the nir path.
> ---
> src/mesa/drivers/dri/i965/brw_vec4.h | 15 +
> s
From: Dave Airlie
This adds support for fine derivatives and enables
ARB_derivative_control on radeonsi.
(just fell out of my working out interpolation)
v2: cleanup some bits, write a comment
v2.1: take Michel's comment from the mailing list
Signed-off-by: Dave Airlie
---
docs/GL3.txt
On 25 July 2015 at 01:15, Marek Olšák wrote:
> On Wed, Jul 22, 2015 at 12:51 AM, Dave Airlie wrote:
>> From: Dave Airlie
>>
>> This is part of ARB_gpu_shader5, and this passes
>> all the piglit tests currently available.
>>
>> Signed-off-by: Dave Airlie
>> ---
>> docs/GL3.txt
Alright, I got through it again...
I asked for a few trivial changes on a few of the patches. With those
fixed, everything except patch 65 and 66 are
Reviewed-by: Jason Ekstrand
While the requested changes on the texturing patches are not
complicated, I would like to see the updated version of
On Fri, 2015-07-24 at 18:49 -0400, Ilia Mirkin wrote:
> Add in missed version checks in the GLSL parser
>
> Signed-off-by: Ilia Mirkin
> ---
> src/glsl/glsl_parser.yy | 9 +
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser
Since commit c0cd5b var->data.binding was being used as a replacement
for atomic buffer index, but they don't have to be the same value they
just happen to end up the same when binding is 0.
Now we store atomic buffer index in the unused var->data.location
to avoid the extra memory of putting back
On Fri, Jul 24, 2015 at 10:25 PM, Timothy Arceri wrote:
> On Fri, 2015-07-24 at 18:49 -0400, Ilia Mirkin wrote:
>> Add in missed version checks in the GLSL parser
>>
>> Signed-off-by: Ilia Mirkin
>> ---
>> src/glsl/glsl_parser.yy | 9 +
>> 1 file changed, 5 insertions(+), 4 deletions(-)
On Fri, 2015-07-24 at 19:37 +0200, Alejandro Piñeiro wrote:
> On 14/05/15 20:38, Ian Romanick wrote:
>
> Hi,
>
> today I was able to go back to work on this patch. Sorry for the delay.
> I changed the subject to the problem, as previous fix was wrong. I have
> some questions (see below).
>
> > >
Add in missed version checks in the GLSL parser
Signed-off-by: Ilia Mirkin
---
v1 -> v2: drop AoA hunks to avoid conflicting with Timothy's changes
src/glsl/glsl_parser.yy | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
On Sat, 2015-07-25 at 01:06 -0400, Ilia Mirkin wrote:
> Add in missed version checks in the GLSL parser
>
> Signed-off-by: Ilia Mirkin
> ---
>
> v1 -> v2: drop AoA hunks to avoid conflicting with Timothy's changes
>
> src/glsl/glsl_parser.yy | 5 +++--
> 1 file changed, 3 insertions(+), 2 dele
---
ARB_dsa / GL 4.5 add indexed GL_TEXTURE_BINDING_* and GL_SAMPLER_BINDING
queries, as well as a GL_TEXTURE_TARGET query on texture objects.
The implementation for the GL_TEXTURE_BINDING_* and GL_SAMPLER_BINDING
queries is based on their non-indexed variants. To map the binding enum
to a textur
67 matches
Mail list logo
