V2: Don't remove ShaderSupported() test. It sets up some function pointers for
the CompileShader framework.
Signed-off-by: Fabian Bieler fabianbie...@fastmail.fm
---
src/glsl/gsraytrace.cpp | 24 +---
1 file changed, 9 insertions(+), 15 deletions(-)
diff --git
On Thu, Feb 6, 2014 at 2:40 AM, Roland Scheidegger srol...@vmware.com wrote:
I don't think that would work. The reason for this stuff to exist is
because new hw makes that possible on the hw level directly.
I don't think this has anything to do with new hardware. This stuff
has always been
On 05/02/14 05:02, Ilia Mirkin wrote:
Apparently some players are ill-prepared for us claiming that a decoder
exists only to have creating it fail, and express this poor preparation
with crashes (e.g. flash). Check that firmware is there to increase the
chances of there being a high
On 02/05/2014 06:19 PM, Marek Olšák wrote:
From: Marek Olšák marek.ol...@amd.com
This binds a NULL sampler view in that case.
Cc: 10.1 10.0 mesa-sta...@lists.freedesktop.org
---
src/mesa/state_tracker/st_cb_texture.c | 6 ++
1 file changed, 6 insertions(+)
diff --git
Am 06.02.2014 12:42, schrieb Marek Olšák:
On Thu, Feb 6, 2014 at 2:40 AM, Roland Scheidegger srol...@vmware.com wrote:
I don't think that would work. The reason for this stuff to exist is
because new hw makes that possible on the hw level directly.
I don't think this has anything to do with
In the specification text of NV_vertex_program1_1, the upper
limit of the RCC instruction is written as 1.884467e+19 in
scientific notation, but as 0x5F80 in binary. But the binary
version translates to 1.84467e+19 rather than 1.884467e+19 in
scientific notation.
Since the lower-limit equals
On 02/06/2014 09:09 AM, Erik Faye-Lund wrote:
In the specification text of NV_vertex_program1_1, the upper
limit of the RCC instruction is written as 1.884467e+19 in
scientific notation, but as 0x5F80 in binary. But the binary
version translates to 1.84467e+19 rather than 1.884467e+19 in
From: Tom Stellard thomas.stell...@amd.com
DS instructions that access local memory can only uses addresses that
are less than or equal to the value of M0. When M0 is uninitialized,
then we experience undefined behavior.
---
lib/Target/R600/SIInstrInfo.cpp| 12
On Sun, Jan 19, 2014 at 11:18 PM, Eric Anholt e...@anholt.net wrote:
---
src/glsl/opt_algebraic.cpp | 12
1 file changed, 12 insertions(+)
diff --git a/src/glsl/opt_algebraic.cpp b/src/glsl/opt_algebraic.cpp
index 332f0b7..6b0d992 100644
--- a/src/glsl/opt_algebraic.cpp
+++
On 4 February 2014 15:07, Jordan Justen jordan.l.jus...@intel.com wrote:
v2:
* Make gl_InstanceID a system value
v3:
* Properly shift from R0.1 into DST.4 by adding
GS_OPCODE_GET_INSTANCE_ID
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
Sir Anthony anth...@adsorbtion.org writes:
This is tough question, I thought about it some time and concluded
that some function class must include body range, otherwise there will
be no easy way to get function end
Good. That's all I wanted to hear, that you had intentionally thought
about
Matt Turner matts...@gmail.com writes:
On Sun, Jan 19, 2014 at 11:18 PM, Eric Anholt e...@anholt.net wrote:
---
src/glsl/opt_algebraic.cpp | 12
1 file changed, 12 insertions(+)
diff --git a/src/glsl/opt_algebraic.cpp b/src/glsl/opt_algebraic.cpp
index 332f0b7..6b0d992 100644
Marek Olšák mar...@gmail.com writes:
From: Marek Olšák marek.ol...@amd.com
OpenGL allows a buffer to be mapped only once, but we also map buffers
internally, e.g. in the software primitive restart fallback, for PBOs,
vbo_get_minmax_index, etc. This has always been a problem, but it will
be
On 02/05/2014 03:06 PM, Marek Olšák wrote:
From: Marek Olšák marek.ol...@amd.com
OpenGL allows a buffer to be mapped only once, but we also map buffers
internally, e.g. in the software primitive restart fallback, for PBOs,
vbo_get_minmax_index, etc. This has always been a problem, but it will
On 02/05/2014 02:06 PM, Marek Olšák wrote:
From: Marek Olšák marek.ol...@amd.com
OpenGL allows a buffer to be mapped only once, but we also map buffers
internally, e.g. in the software primitive restart fallback, for PBOs,
vbo_get_minmax_index, etc. This has always been a problem, but it
On Thursday 06 February 2014, Eric Anholt wrote:
Marek Olšák mar...@gmail.com writes:
From: Marek Olšák marek.ol...@amd.com
OpenGL allows a buffer to be mapped only once, but we also map buffers
internally, e.g. in the software primitive restart fallback, for PBOs,
On 02/04/2014 09:08 PM, Kenneth Graunke wrote:
According to the latest documentation, any PIPE_CONTROL with the
Command Streamer Stall bit set must also have another bit set,
with five different options:
- Render Target Cache Flush
- Depth Cache Flush
- Stall at Pixel Scoreboard
On 02/04/2014 09:07 PM, Kenneth Graunke wrote:
This makes it work on Broadwell, too.
Eric convinced me (on IRC) that the extra BEGIN_BATCH / ADVANCE_BATCH
calls won't cause problems. Series is
Reviewed-by: Ian Romanick ian.d.roman...@intel.com
Signed-off-by: Kenneth Graunke
On 02/04/2014 02:50 PM, Ilia Mirkin wrote:
On Tue, Jan 21, 2014 at 1:09 AM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
this just ties the mesa code to the pre-existing gallium interface,
I'm not sure what to do with the CSO stuff yet.
0.2: fix min/max bounds
On 02/04/2014 01:29 AM, Rogovin, Kevin wrote:
I don't believe our hardware can support GL_ARB_shader_stencil_export.
The render target write message can take RGBA, depth, and sample masks,
but not stencil. Without that, it's not at all obvious how to implement it.
There is a terrible
On 02/05/2014 09:42 PM, Kenneth Graunke wrote:
Mesa fails to retain the precision qualifier when parsing:
#version 300 es
centroid in mediump vec2 v;
Consider how the parser's type_qualifier production is applied.
First, the precision_qualifier rule creates a new ast_type_qualifier:
https://bugs.freedesktop.org/show_bug.cgi?id=69101
--- Comment #18 from Alexander Mezin mezin.alexan...@gmail.com ---
On Arch Linux:
Xorg 1.15, Mesa 10.0.3, Linux 3.13.2, KWin 4.11.6 with OpenGL 3.1 compositing.
libdrm 2.4.52
xf86-video-ati 7.3.0
xf86-video-intel 2.99.909 (sna)
No more black
On 02/03/2014 11:15 AM, Kenneth Graunke wrote:
Since commit 9cee3ff562f3e4b51bfd30338fd1ba7716ac5737, INTEL_DEBUG=vs
has caused a NULL pointer dereference for fixed-function/ARB programs.
In the vec4 generators, prog is a gl_program, and shader_prog is the
gl_shader_program. This is
https://bugs.freedesktop.org/show_bug.cgi?id=69101
Alexander Mezin mezin.alexan...@gmail.com changed:
What|Removed |Added
Status|NEW |RESOLVED
On 01/29/2014 02:36 PM, Kenneth Graunke wrote:
I'd neglected to port these to Broadwell. Most of this code is copy
and pasted from Gen7, but instead of using F32TO16/F16TO32, we just
use MOV with HF register types.
Fixes fs-packHalf2x16 and fs-unpackHalf2x16 tests (both the ARB
extension
On 01/29/2014 02:36 PM, Kenneth Graunke wrote:
Signed-off-by: Kenneth Graunke kenn...@whitecape.org
This patch is
Reviewed-by: Ian Romanick ian.d.roman...@intel.com
---
src/mesa/drivers/dri/i965/gen8_disasm.c | 1 +
1 file changed, 1 insertion(+)
diff --git
Forgot to cc the list.
-- Forwarded message --
From: Chris Forbes chr...@ijw.co.nz
Date: Fri, Feb 7, 2014 at 1:24 PM
Subject: Re: [Mesa-dev] [PATCH 2/2] i965: Use the new
brw_load_register_mem helper for draw indirect.
To: Kenneth Graunke kenn...@whitecape.org
+ } else {
On 01/29/2014 11:08 PM, Kenneth Graunke wrote:
On 01/29/2014 02:36 PM, Kenneth Graunke wrote:
Broadwell removed the F32TO16 and F16TO32 instructions. However, it has
actual support for HF values, so they're actually just MOV.
Fixes vs-packHalf2x16 and vs-unpackHalf2x16 tests (both the ARB
the folder:
http://www.x.org/docs/AMD/
has an outdated version of Southern Islands Series Instruction Set
Architecture. The latest version can be found here:
http://developer.amd.com/wordpress/media/2012/12/AMD
_Southern_Islands_Instruction_Set_Architecture.pdf
Who has permission to update this
---
src/glsl/ast.h | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/glsl/ast.h b/src/glsl/ast.h
index 8e54ddc..50b95bf 100644
--- a/src/glsl/ast.h
+++ b/src/glsl/ast.h
@@ -116,10 +116,10 @@ public:
*/
struct {
unsigned source; /** GLSL
This is tough question, I thought about it some time and concluded that some
function class must include body range, otherwise there will be no easy way to
get function end (it contents may be iterated (I did it in my code before I
realized I need to change locations placing itself, it was
From: Kristian Høgsberg k...@bitplanet.net
Making this its own function cleans up loader_get_driver_for_fd() a bit
and simplifies the control flow.
Signed-off-by: Kristian Høgsberg k...@bitplanet.net
v2:
- change patch order and rebase (Emil)
---
src/loader/loader.c | 48
Having one function that tries to accomodate the different decision trees
for the three different #ifdef cases makes the logic hard to follow. Now
that most of the complexity in loader_get_driver_for_fd() has been split
out in helper functions, we can just move loader_get_driver_for_fd() into
the
The udev hwdb is a mechanism for applying udev properties to devices at
hotplug time. The hwdb text files are compiled into a binary database
that lets udev efficiently look up and apply properties to devices that
match a given modalias.
This patch exports the mesa PCI ID tables as hwdb files
This lets us match any device on any bus, including platform devices.
Signed-off-by: Kristian Høgsberg k...@bitplanet.net
v2: (Emil)
- correctly handle chip_id = 0 devices under android
- fnmatch.h is only available under linux
- cleanup a few typos s/driver-map/modalias-map/
XXX:
- scons
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
src/loader/loader.c | 17 -
1 file changed, 8 insertions(+), 9 deletions(-)
diff --git a/src/loader/loader.c b/src/loader/loader.c
index cbb501d..d75380b 100644
--- a/src/loader/loader.c
+++ b/src/loader/loader.c
@@ -264,13
From: Kristian Høgsberg k...@bitplanet.net
Signed-off-by: Kristian Høgsberg k...@bitplanet.net
v2:
- change patch order and rebase (Emil)
---
src/loader/loader.c | 44 +---
1 file changed, 25 insertions(+), 19 deletions(-)
diff --git
Helo Kristian,
Rather than nitpicking like an old bat, I took the liberty of respinning
your patchset. Most patches are left as is, minus a reshuffle
in their order + rebase.
Available in the hwdb-loader-v4 branch at https://github.com/evelikov/Mesa/
Highlights comparing to v3:
- Dropped
Trying to figure out where a driver name comes from by looking at
whether or not chip_id is -1 isn't very roboust.
Signed-off-by: Kristian Høgsberg k...@bitplanet.net
v2:
- change patch order and rebase (Emil)
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
src/loader/loader.c | 25
Make sure that we unref the udev, if we fail to retrieve the device.
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
src/loader/loader.c | 9 -
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/src/loader/loader.c b/src/loader/loader.c
index d75380b..422079a 100644
ping ?
On 29/01/14 20:10, Emil Velikov wrote:
v2:
Cleanup driver name if dri2_load_driver() fails. Spotted by Chad
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
Hmm... I swear I had that hunk in mind when writing the patch.
Or maybe I was day dreaming.
Thanks Chad
Add the explicit note about the required version during configure.
Require the same version (151) of udev when building the pipe-loader.
Mention the udev version requirement in GBM Requires.private.
v2: Resolve a couple of silly typos. Spotted by Ilia
Signed-off-by: Emil Velikov
ping ?
On 29/01/14 20:23, Emil Velikov wrote:
Hello list,
Now that everything libudev related is moved to src/loader and we
handle the library by dlopen-ing it, we can cleanup a few includes,
CFLAGS and LIBS that were behind.
The final commit adds explicit version check for udev, where
On 01/02/14 09:27, Lauri Kasanen wrote:
On Fri, 31 Jan 2014 10:58:48 -0800
Kenneth Graunke kenn...@whitecape.org wrote:
On 01/31/2014 03:14 AM, Lauri Kasanen wrote:
It's important not to break existing uses. There is no need to require
udev for 3d. This is an alternative to Stéphane
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Reviewed-by: Paul Berry stereotype...@gmail.com
---
src/mesa/main/mtypes.h | 1 +
src/mesa/main/shaderapi.c | 1 +
src/mesa/program/program.c | 1 +
On Don, 2014-02-06 at 09:40 -0800, Tom Stellard wrote:
From: Tom Stellard thomas.stell...@amd.com
DS instructions that access local memory can only uses addresses that
are less than or equal to the value of M0. When M0 is uninitialized,
then we experience undefined behavior.
[...]
@@
From: Dave Airlie airl...@redhat.com
This respects the CAP and adds glsl-tgsi support for the TG4 opcode.
Signed-off-by: Dave Airlie airl...@redhat.com
---
src/mesa/state_tracker/st_extensions.c | 4
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 11 ---
2 files changed, 12
On Mon, 2014-02-03 at 18:07 +0100, Marek Olšák wrote:
Do you mean something like:
write_domains = VRAM
read_domains = VRAM | GTT
I don't think it's perfect.
That's what I meant, and of course it's not perfect. :)
We should treat rendering by an app and blitting by u_blitter (=
texture
From: Dave Airlie airl...@redhat.com
This just adds the TG4 opcode, and a CAP for the
max texture gather components.
v2: add component selection src and a cap to say
hw can do it. (st can use to help control
GL_ARB_gpu_shader5/GLSL 4.00). Add docs.
Signed-off-by: Dave Airlie airl...@redhat.com
From: Dave Airlie airl...@redhat.com
This respects the CAP and adds glsl-tgsi support for the TG4 opcode.
v2: propogate the component selection to gallium
Signed-off-by: Dave Airlie airl...@redhat.com
---
src/mesa/state_tracker/st_extensions.c | 4
We introduce a new merge_in_qualifier ast_type_qualifier
which allows specialized handling of merging input layout
qualifiers.
By merging layout qualifiers into state-in_qualifier, we
allow multiple input qualifiers. For example, the primitive
type can be specified specified separately from the
On Fri, Feb 7, 2014 at 12:36 PM, Roland Scheidegger srol...@vmware.com wrote:
Am 07.02.2014 02:56, schrieb Ilia Mirkin:
On Thu, Feb 6, 2014 at 8:52 PM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
This just adds the TG4 opcode, and a CAP for the
max texture
From: Dave Airlie airl...@redhat.com
I'll merge this with the previous patches.
---
src/gallium/docs/source/screen.rst | 5 +
1 file changed, 5 insertions(+)
diff --git a/src/gallium/docs/source/screen.rst
b/src/gallium/docs/source/screen.rst
index 90e8278..0677e93 100644
---
https://bugs.freedesktop.org/show_bug.cgi?id=72895
--- Comment #13 from Barto mister.free...@laposte.net ---
here is the apitrace of the bug :
http://demo.ovh.eu/download/4460655c06da77b352f00024805c88ac/fgfs.trace.tar.gz
the bug starts when the engine starts, the trees will disapear
--
You
This just adds the gallium opcode/cap + the st/mesa pieces
to enable it, I have r600g code but I want to go over it again
before sending it. I also have a couple of other extensions that
have trivial gallium bits I'd like to line up.
(GL_ARB_texture_query_lod, and GL_ARB_texture_query_levels).
From: Dave Airlie airl...@redhat.com
This just adds the TG4 opcode, and a CAP for the
max texture gather components.
Signed-off-by: Dave Airlie airl...@redhat.com
---
src/gallium/auxiliary/tgsi/tgsi_info.c | 1 +
src/gallium/include/pipe/p_defines.h | 3 ++-
https://bugs.freedesktop.org/show_bug.cgi?id=72895
--- Comment #14 from Ilia Mirkin imir...@alum.mit.edu ---
In case it's of any interest, I also see the bug with nouveau/nv50 (nv98 card)
by replaying the apitrace -- so yeah, probably a core bug. Or mesa/st.
--
You are receiving this mail
v4:
* Merge with recent compute shader parsing of
input layout qualifiers
v3:
* Fix major brokenness of dual instance mode operation
using Paul's suggestions
* Update parsing to allow separate primitive and
invocation declarations. Fixes piglit test:
On Thu, Feb 6, 2014 at 8:52 PM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
This just adds the TG4 opcode, and a CAP for the
max texture gather components.
Is this different from GATHER4? If it is, should probably be
documented in docs/source/tgsi.rst...
v3:
* Properly prevent dual object mode execution when
the invocation count 1
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Reviewed-by: Paul Berry stereotype...@gmail.com
---
src/mesa/drivers/dri/i965/brw_context.h | 2 ++
src/mesa/drivers/dri/i965/brw_defines.h
Fixes various piglit tests:
spec/glsl-1.50/compiler/incorrect-in-layout-qualifier-*.geom
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/glsl/ast_type.cpp | 13 +
1 file changed, 13 insertions(+)
diff --git a/src/glsl/ast_type.cpp b/src/glsl/ast_type.cpp
index
Grab the parsed invocation count, check for consistency
during linking, and finally save the result in
gl_shader_program Geom.Invocations.
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Reviewed-by: Paul Berry stereotype...@gmail.com
---
src/glsl/glsl_parser_extras.cpp | 4
Am 07.02.2014 02:56, schrieb Ilia Mirkin:
On Thu, Feb 6, 2014 at 8:52 PM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
This just adds the TG4 opcode, and a CAP for the
max texture gather components.
Is this different from GATHER4? If it is, should probably be
On Friday 07 February 2014 00:58:47 Emil Velikov wrote:
Add the explicit note about the required version during configure.
Require the same version (151) of udev when building the pipe-loader.
Mention the udev version requirement in GBM Requires.private.
v2: Resolve a couple of silly typos.
v2:
* Make gl_InvocationID a system value
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Reviewed-by: Paul Berry stereotype...@gmail.com
---
src/glsl/builtin_variables.cpp | 2 ++
src/mesa/main/mtypes.h | 1 +
2 files changed, 3 insertions(+)
diff --git
v2:
* Make gl_InvocationID a system value
v3:
* Properly shift from R0.1 into DST.4 by adding
GS_OPCODE_GET_INSTANCE_ID
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Acked-by: Paul Berry stereotype...@gmail.com
---
src/mesa/drivers/dri/i965/brw_defines.h | 12
On Thu, Feb 6, 2014 at 11:08 PM, Eric Anholt e...@anholt.net wrote:
Marek Olšák mar...@gmail.com writes:
From: Marek Olšák marek.ol...@amd.com
OpenGL allows a buffer to be mapped only once, but we also map buffers
internally, e.g. in the software primitive restart fallback, for PBOs,
v3:
* Add check for ARB_gpu_shader5
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
Reviewed-by: Paul Berry stereotype...@gmail.com
---
src/mesa/main/shaderapi.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/mesa/main/shaderapi.c b/src/mesa/main/shaderapi.c
index
_mesa_glsl_parse_state in_qualifier-invocations will store the
invocations count.
v3:
* Use in_qualifier to allow the primitive to be specied
separately from the invocations count (merge_qualifiers)
Signed-off-by: Jordan Justen jordan.l.jus...@intel.com
---
src/glsl/ast.h | 10
https://bugs.freedesktop.org/show_bug.cgi?id=72895
--- Comment #12 from Barto mister.free...@laposte.net ---
I tried the 3 patches from Fredrik Höglund ( bug 73504 ) with the source code
of mesa 10.0.3 ( the official source code, not the last git version )
but the bug is still here, the trees
Yes, I will be very grateful if someone will do it.
Thanks.
On Thu, 06 Feb 2014 12:40:21 -0800
Carl Worth cwo...@cworth.org wrote:
Sir Anthony anth...@adsorbtion.org writes:
This is tough question, I thought about it some time and concluded
that some function class must include body range,
Hi,
So I believe that we can all agree that the tree-based representation
that GLSL IR currently uses for shaders needs to go. For the benefit
of those that didn't watch Ian Romanick's talk at FOSDEM, I'll
reiterate some of the problems with it as of now:
- All the ir_dereference chains blow up
72 matches
Mail list logo
