Hi Neil,
On Tuesday, October 28, 2014 18:22:33 Neil Roberts wrote:
> On inspection it looks like this would potentially break
> _mesa_meta_Clear when it is using GLSL because that does not save the
> MESA_META_TRANSFORM state.
>
> I wonder if MESA_META_TRANSFORM is not the right state flag for t
A pattern in certain shaders is:
uniform vec4 colors[NUM_LIGHTS];
for (int i = 0; i < NUM_LIGHTS; i++) {
...use colors[i]...
}
In this case, the application author expects the shader compiler to
unroll the loop. By doing so, it replaces variable indexing of the
array with constan
On Wed, Oct 29, 2014 at 5:16 PM, Kenneth Graunke wrote:
> diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources
> index 0c55327..6aed52d 100644
> --- a/src/glsl/Makefile.sources
> +++ b/src/glsl/Makefile.sources
> @@ -58,6 +58,7 @@ LIBGLSL_FILES = \
> $(GLSL_SRCDIR)/loop_analy
On Wed, Oct 29, 2014 at 6:11 PM, Thomas Helland
wrote:
> This will be used later on in opt_minmax
>
> Signed-off-by: Thomas Helland
> ---
> src/glsl/ir_constant_util.h | 103
>
> src/glsl/opt_algebraic.cpp | 95 ++---
Add functions for is_greater_than_one
and is_less_than_zero
Signed-off-by: Thomas Helland
---
src/glsl/ir_constant_util.h | 31 +++
1 file changed, 31 insertions(+)
diff --git a/src/glsl/ir_constant_util.h b/src/glsl/ir_constant_util.h
index b3b9a19..9dae974 100644
-
This will make expansion easier and less cluttered.
Signed-off-by: Thomas Helland
---
src/glsl/opt_minmax.cpp | 15 ++-
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/glsl/opt_minmax.cpp b/src/glsl/opt_minmax.cpp
index e4141bc..b21daca 100644
--- a/src/glsl/opt_mi
This gets rid of extra instructions in some
shaders I purposefully wrote to test this.
Works for shaders similar to the following:
vec3 c = {8, 8, 8};
gl_FragColor.rgb = max(sin(d), c);
Signed-off-by: Thomas Helland
---
src/glsl/opt_minmax.cpp | 17 +
1 file changed, 17 insertio
All of these are guaranteed to be larger than 0
Signed-off-by: Thomas Helland
---
src/glsl/opt_minmax.cpp | 8
1 file changed, 8 insertions(+)
diff --git a/src/glsl/opt_minmax.cpp b/src/glsl/opt_minmax.cpp
index 4088c80..e768857 100644
--- a/src/glsl/opt_minmax.cpp
+++ b/src/glsl/opt_m
Also, if the operand has bounds between 0.0 and 1.0
then copy that range up.
Signed-off-by: Thomas Helland
---
src/glsl/opt_minmax.cpp | 14 ++
1 file changed, 14 insertions(+)
diff --git a/src/glsl/opt_minmax.cpp b/src/glsl/opt_minmax.cpp
index 0b9ddc2..4088c80 100644
--- a/src/gls
This will be used later on in opt_minmax
Signed-off-by: Thomas Helland
---
src/glsl/ir_constant_util.h | 103
src/glsl/opt_algebraic.cpp | 95 ++--
src/glsl/opt_minmax.cpp | 19 ++--
3 files changed, 109
This series does some initial work to make expansion of
the get_range function a lot cleaner.
It also adds a couple simple initial ranges.
These patches are by no means perfect, but I hope
they will provide some feedback and ideas.
I'm hoping to expand this to do the following:
-Add get_range for
On Tue, Oct 28, 2014 at 5:30 PM, Matt Turner wrote:
> On Tue, Oct 28, 2014 at 3:59 PM, Matt Turner wrote:
- assert(stage == MESA_SHADER_FRAGMENT);
>>>
>>> I like removing these asserts from the function bodies, but I'm
>>> confused why you're doing it. The VS isn't going to call
>>> fire_f
---
src/mesa/main/bitset.h | 1 +
src/mesa/main/macros.h | 27 ---
src/util/macros.h | 27 +++
3 files changed, 28 insertions(+), 27 deletions(-)
diff --git a/src/mesa/main/bitset.h b/src/mesa/main/bitset.h
index f50b14f..dbf1af9 100644
--- a/
---
configure.ac | 1 +
scons/gallium.py | 2 +
src/gallium/auxiliary/tgsi/tgsi_exec.c | 1 +
src/gallium/auxiliary/tgsi/tgsi_scan.c | 2 +-
src/gallium/auxiliary/util/u_helpers.c |
---
.../drivers/dri/i965/brw_fs_copy_propagation.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 2 +-
.../drivers/dri/i965/brw_performance_monitor.c | 2 +-
.../drivers/dri/i965/brw_vec4_live_variables.h | 2 +-
src/mesa/drivers/dri/nouveau/nouveau_context.h
In the compiler, we'd like to generate implicit uniforms for internal
use. These should not be visible via the GL uniform introspection API.
To support that, we add a new ir_variable::how_declared value of
ir_var_hidden, and plumb that through to gl_uniform_storage.
v2 (idr): Fix some memory man
Consider GLSL code such as:
const ivec2 offsets[] =
ivec2[](ivec2(-1, -1), ivec2(-1, 0), ivec2(-1, 1),
ivec2(0, -1), ivec2(0, 0), ivec2(0, 1),
ivec2(1, -1), ivec2(1, 0), ivec2(1, 1));
ivec2 offset = offsets[];
Both i965 and nv50 currently handle this v
On Wed, Oct 29, 2014 at 3:51 PM, Matt Turner wrote:
> On Wed, Oct 29, 2014 at 3:42 PM, Jason Ekstrand
> wrote:
> > diff --git a/configure.ac b/configure.ac
> > index 03f1bca..e2258eb 100644
> > --- a/configure.ac
> > +++ b/configure.ac
> > @@ -131,6 +131,7 @@ dnl Check for compiler builtins
> >
On Tue, Oct 28, 2014 at 5:48 PM, Kenneth Graunke wrote:
> On Tuesday, October 28, 2014 04:25:05 PM Matt Turner wrote:
>> On Tue, Oct 28, 2014 at 3:17 PM, Kristian Høgsberg
> wrote:
>> > This flag signals that we have a SIMD8 VS shader so we can set up the
>> > corresponding state accordingly. Th
On Thursday, October 23, 2014 04:19:19 PM Matt Turner wrote:
> Helps a small number of vertex shaders in the games Dungeon Defenders
> and Shank, as well as an internal benchmark.
>
> instructions in affected programs: 2801 -> 2719 (-2.93%)
> ---
> src/glsl/opt_algebraic.cpp | 25
https://bugs.freedesktop.org/show_bug.cgi?id=57702
--- Comment #3 from fjhenigman ---
If I'm understanding correctly:
- this bug asks that, for example, /usr/lib64/dri/i965_dri.so pull in libglapi
- at the moment it's deliberately not pulled because some use case didn't want
that
- that other use
On Wednesday, October 29, 2014 02:09:55 PM Matt Turner wrote:
> Multiplication is commutative.
>
> instructions in affected programs: 48314 -> 47954 (-0.75%)
> ---
> src/mesa/drivers/dri/i965/brw_fs_cse.cpp | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/src/mes
On Wed, Oct 29, 2014 at 2:10 PM, Matt Turner wrote:
> ---
> .../drivers/dri/i965/brw_fs_live_variables.cpp | 35
> ++
> src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 5
> 2 files changed, 40 insertions(+)
>
> diff --git a/src/mesa/drivers/dri/i965/brw_fs_live
On Wed, Oct 29, 2014 at 3:42 PM, Jason Ekstrand wrote:
> diff --git a/configure.ac b/configure.ac
> index 03f1bca..e2258eb 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -131,6 +131,7 @@ dnl Check for compiler builtins
> AX_GCC_BUILTIN([__builtin_bswap32])
> AX_GCC_BUILTIN([__builtin_bswa
---
configure.ac | 1 +
src/gallium/auxiliary/tgsi/tgsi_exec.c | 1 +
src/gallium/auxiliary/tgsi/tgsi_scan.c | 2 +-
src/gallium/auxiliary/util/u_helpers.c | 1 +
src/gallium/auxiliary/util/u_math.h|
On Wed, Oct 29, 2014 at 1:45 PM, Roland Scheidegger
wrote:
> I like the idea of the series, however gallium still uses its own
> definitions (by the looks of it supporting more compilers for native
> definitions but in some cases with worse code for the fallback)
> sometimes with different names
On Wednesday, October 29, 2014 02:31:53 PM Matt Turner wrote:
> ... or on Gen >= 7 at all. We use load_payload to gather results for the
> FB write(s) now, so we never write to MRFs directly. It's still called
> after lower_load_payload() since that will generate MOVs to MRFs on
> platforms with MR
https://bugs.freedesktop.org/show_bug.cgi?id=84186
--- Comment #6 from Mathias Brodala ---
Created attachment 108659
--> https://bugs.freedesktop.org/attachment.cgi?id=108659&action=edit
X startup with EGL debug info
I am experiencing basically the same issue with my HD4670 when trying to enab
https://bugs.freedesktop.org/show_bug.cgi?id=57702
fjhenigman changed:
What|Removed |Added
CC||fjhmesa...@gmail.com
--
You are receiving
https://bugs.freedesktop.org/show_bug.cgi?id=84186
Mathias Brodala changed:
What|Removed |Added
CC||i...@noctus.net
--
You are receiving
... or on Gen >= 7 at all. We use load_payload to gather results for the
FB write(s) now, so we never write to MRFs directly. It's still called
after lower_load_payload() since that will generate MOVs to MRFs on
platforms with MRFs.
No differences in shader-db on Haswell (Gen 7.5).
---
src/mesa/d
On Tue, Oct 28, 2014 at 4:50 PM, Matt Turner wrote:
> On Tue, Oct 28, 2014 at 3:17 PM, Kristian Høgsberg wrote:
>> This patch uses the previous refactoring to add a new run_vs() method
>> that generates vertex shader code using the scalar visitor and
>> optimizer.
>>
>> Signed-off-by: Kristian Hø
Also while we're touching var_from_reg, just make it an inline function.
---
src/mesa/drivers/dri/i965/brw_fs_dead_code_eliminate.cpp | 8
src/mesa/drivers/dri/i965/brw_fs_live_variables.cpp | 14 --
src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 11 ++
The next patch will be simplified because of this, and makes reading the
code a lot easier.
---
.../dri/i965/brw_fs_dead_code_eliminate.cpp| 2 +-
.../drivers/dri/i965/brw_fs_live_variables.cpp | 54 --
src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 6 +--
.
Like in fs_visitor.
---
src/mesa/drivers/dri/i965/brw_vec4.h | 5 +++--
src/mesa/drivers/dri/i965/brw_vec4_live_variables.cpp | 15 +++
src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp| 2 +-
3 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/src/
---
.../drivers/dri/i965/brw_fs_live_variables.cpp | 35 ++
src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 5
2 files changed, 40 insertions(+)
diff --git a/src/mesa/drivers/dri/i965/brw_fs_live_variables.cpp
b/src/mesa/drivers/dri/i965/brw_fs_live_variables.c
Most prominently helps Natural Selection 2, which has a surprising
number shaders that do very complicated things before drawing black.
instructions in affected programs: 23824 -> 19570 (-17.86%)
---
.../dri/i965/brw_fs_dead_code_eliminate.cpp| 23 +++---
1 file change
---
src/mesa/drivers/dri/i965/brw_vec4_cse.cpp | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/brw_vec4_cse.cpp
b/src/mesa/drivers/dri/i965/brw_vec4_cse.cpp
index 28c69ca..630d335 100644
--- a/src/mesa/drivers/dri/i965/brw_vec4_cse.cpp
+++ b/src/
Multiplication is commutative.
instructions in affected programs: 48314 -> 47954 (-0.75%)
---
src/mesa/drivers/dri/i965/brw_fs_cse.cpp | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/brw_fs_cse.cpp
b/src/mesa/drivers/dri/i965/brw_fs_cse.cpp
I like the idea of the series, however gallium still uses its own
definitions (by the looks of it supporting more compilers for native
definitions but in some cases with worse code for the fallback)
sometimes with different names (fls/util_last_bit) and sometimes with
the same even (ffs and util_bi
Kenneth Graunke writes:
> v2: Use the UST value provided in the PRESENT_COMPLETE_NOTIFY event
> rather than gettimeofday(), which gives us the presentation time
> instead of the time when SwapBuffers was called. Suggested by
> Keith Packard. This relies on the fact that the X DRI3/P
On Wednesday, October 29, 2014 11:27:56 AM Jason Ekstrand wrote:
> ---
> src/gallium/state_trackers/glx/xlib/glx_api.c | 6 +-
> src/gallium/state_trackers/glx/xlib/xm_api.c | 10 +-
> src/mesa/drivers/common/meta.c| 3 +-
> src/mesa/drivers/dri/i965/brw_blorp_bli
On 29/10/14 10:43, Matthew Waters wrote:
> From: Matthew Waters
>
> As of version 15 of the EGL_KHR_create_context spec, debug contexts
> are allowed for ES contexts. We should allow creation instead of
> erroring.
>
By moving the check from the dri module to the loader we can end up with
combi
v2: Use the UST value provided in the PRESENT_COMPLETE_NOTIFY event
rather than gettimeofday(), which gives us the presentation time
instead of the time when SwapBuffers was called. Suggested by
Keith Packard. This relies on the fact that the X DRI3/Present
implementations use mic
On 21/10/14 16:19, Jan Vesely wrote:
> Signed-off-by: Jan Vesely
> ---
> configure.ac | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/configure.ac b/configure.ac
> index 93b25a2..a588d55 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -1970,7 +1970,7 @@ if te
https://bugs.freedesktop.org/show_bug.cgi?id=70410
Emil Velikov changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
On 27/10/14 21:03, Jan Vesely wrote:
> On Mon, 2014-10-27 at 20:22 +, Emil Velikov wrote:
>> On 27/10/14 18:05, Jan Vesely wrote:
>>> On Mon, 2014-10-27 at 02:24 +, Emil Velikov wrote:
On 26/10/14 19:36, Jan Vesely wrote:
> On Fri, 2014-10-24 at 23:54 +, Emil Velikov wrote:
>>>
---
src/gallium/state_trackers/glx/xlib/glx_api.c | 6 +-
src/gallium/state_trackers/glx/xlib/xm_api.c | 10 +-
src/mesa/drivers/common/meta.c| 3 +-
src/mesa/drivers/dri/i965/brw_blorp_blit.cpp | 4 +-
src/mesa/drivers/dri/i965/brw_curbe.c |
On Tue, Oct 28, 2014 at 4:36 PM, Matt Turner wrote:
> On Tue, Oct 28, 2014 at 3:17 PM, Kristian Høgsberg wrote:
>> These last few operations all only apply when we've actually generated code,
>> optimized and allocated registers. The dummy and the repclear shaders don't
>> touch uncompressed_sta
---
.../drivers/dri/i965/brw_fs_copy_propagation.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_live_variables.h | 2 +-
.../drivers/dri/i965/brw_performance_monitor.c | 2 +-
.../drivers/dri/i965/brw_vec4_live_variables.h | 2 +-
src/mesa/drivers/dri/nouveau/nouveau_context.h
On Tue, Oct 28, 2014 at 4:33 PM, Matt Turner wrote:
> On Tue, Oct 28, 2014 at 3:17 PM, Kristian Høgsberg wrote:
>> The scalar vertex shader will use the ATTR register file for vertex
>> attributes. This patch adds support for the ATTR file to fs_visitor.
>>
>> Signed-off-by: Kristian Høgsberg
>
On Tue, Oct 28, 2014 at 4:17 PM, Matt Turner wrote:
> On Tue, Oct 28, 2014 at 3:17 PM, Kristian Høgsberg wrote:
>> This is all we need from the generator for SIMD8 vertex shaders. This
>> opcode is just the send instruction, all the hard work will happen
>> in the visitor using LOAD_PAYLOAD.
>>
https://bugs.freedesktop.org/show_bug.cgi?id=77449
Neil Roberts changed:
What|Removed |Added
Depends on||83908
--
You are receiving this mail bec
https://bugs.freedesktop.org/show_bug.cgi?id=84566
--- Comment #46 from Jason Ekstrand ---
(In reply to Iago Toral from comment #43)
> Jason, we are running into some issues when attempting to use
> _mesa_format_convert for glReadPixels and glGetTexImage.
>
> Generally, one thing that is differe
Reviewed-by: Matt Turner
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
https://bugs.freedesktop.org/show_bug.cgi?id=84566
--- Comment #45 from Jason Ekstrand ---
(In reply to Iago Toral from comment #44)
> (In reply to Iago Toral from comment #43)
> (...)
> > 3) Luminance formats have special requirements. A conversion to Luminance
> > from RGBA requires to do L=R+G
On Mon, Oct 27, 2014 at 6:34 AM, Alexandre Courbot wrote:
> GK20A does not have dedicated VRAM, therefore allocating in VRAM can be
> sub-optimal and sometimes even harmful. Set its VRAM domain to
> NOUVEAU_BO_GART so all objects are allocated in system memory.
>
> Signed-off-by: Alexandre Courbot
Kenneth Graunke writes:
> v2: Use the UST value provided in the PRESENT_COMPLETE_NOTIFY event
> rather than gettimeofday(), which gives us the presentation time
> instead of the time when SwapBuffers was called. Suggested by
> Keith Packard. This relies on the fact that the X Presen
On Tue, 2014-10-28 at 19:27 -0700, Kenneth Graunke wrote:
> Hello,
>
> I'd like to rename some files in i965:
>
> - brw_gs.c -> brw_ff_gs.c
> - brw_gs.h -> brw_ff_gs.h
> - brw_gs_emit.c -> brw_ff_gs_emit.c
> - brw_vec4_gs.c -> brw_gs.c
> - brw_vec4_gs.h -> brw_gs.h
>
> The current "brw
On Wed, 2014-10-29 at 16:58 +1100, Timothy Arceri wrote:
> On Tue, 2014-10-28 at 22:14 +, Bruno Jimenez wrote:
> > Hi,
> >
> > I haven't had time to play yet with OpenMP, but I have seen the assembly
> > it produces in my computer. If I enable SSE2 it can use it, and if I
> > enable SSE4.1 it
Makes use of SSE to speed up compute of min and max elements
Callgrind cpu usage results from pts benchmarks:
Openarena 0.8.8: 3.67% -> 1.03%
UrbanTerror: 2.36% -> 0.81%
V5:
- actually make use of the optimisation in android (Emil Velikov)
- set a better array size limit for using SSE and added
- rebase and resend.
v3:
- fix up the EGL_CONTEXT_OPENGL_ROBUST_ACCESS_BIT_KHR check
v2:
- replace &= with |=
- remove offset="assign" from the api xml
Matthew Waters (3):
egl: rework handling EGL_CONTEXT_FLAGS for ES debug contexts
glapi: add function pointers for KHR_debug for gles
ma
From: Matthew Waters
Although GL_CONTEXT_FLAGS is not explicitly added by KHR_debug,
it contains,
"It is implementation defined how much debug output is generated if
the context was created without the CONTEXT_DEBUG_BIT set. This is a new
query bit added to the existing GL_CONTEXT_FL
From: Matthew Waters
As of version 15 of the EGL_KHR_create_context spec, debug contexts
are allowed for ES contexts. We should allow creation instead of
erroring.
Signed-off-by: Matthew Waters
---
src/egl/main/eglcontext.c | 51 ++
src/mesa/driver
From: Matthew Waters
Signed-off-by: Matthew Waters
---
src/mapi/glapi/gen/KHR_debug.xml| 73 +
src/mesa/main/extensions.c | 2 +-
src/mesa/main/tests/dispatch_sanity.cpp | 25 +++
3 files changed, 99 insertions(+), 1 deletion(-)
dif
Reviewed-by: Marek Olšák
Marek
On Wed, Oct 29, 2014 at 8:58 AM, Michel Dänzer wrote:
> From: Michel Dänzer
>
> This prevents us from silently overflowing the stack arrays, and allows
> arbitrary stack depths.
>
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85454
>
> Reported-and-Tes
The CSE pass now prints out why it thinks a value is not a candidate for
adding to the AE set.
Signed-off-by: Kenneth Graunke
---
src/glsl/opt_cse.cpp | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/src/glsl/opt_cse.cpp b/src/glsl/opt_cse.cpp
index 9c96835..b0b67f4 10
v2: Use the UST value provided in the PRESENT_COMPLETE_NOTIFY event
rather than gettimeofday(), which gives us the presentation time
instead of the time when SwapBuffers was called. Suggested by
Keith Packard. This relies on the fact that the X Present
implementation uses microsec
From: Michel Dänzer
This prevents us from silently overflowing the stack arrays, and allows
arbitrary stack depths.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85454
Reported-and-Tested-by: Nick Sarnie
Signed-off-by: Michel Dänzer
---
src/gallium/drivers/radeon/radeon_llvm.h
https://bugs.freedesktop.org/show_bug.cgi?id=84570
--- Comment #31 from Michel Dänzer ---
(In reply to Kai from comment #30)
> Michel, is there any chance attachment 107544 [details] [review] will be
> part of 3.18?
No, but it's in Alex's queue for 3.19.
--
You are receiving this mail because:
https://bugs.freedesktop.org/show_bug.cgi?id=85586
--- Comment #1 from Michel Dänzer ---
Using current LLVM 3.6 Git snapshot.
--
You are receiving this mail because:
You are the assignee for the bug.
___
mesa-dev mailing list
mesa-dev@lists.freedeskto
https://bugs.freedesktop.org/show_bug.cgi?id=85586
Bug ID: 85586
Summary: Draw module crashes in LLVM generated code since
commit 60ec95fa1e0c42bd42358185970b20c9b81591fa
Product: Mesa
Version: git
Hardware: Other
72 matches
Mail list logo
