Remove definition from egl_dri2.
Defining this is egl_dri2.h breaks as soon as
a new dri2 buffer token is added like with commit
4501a5d6e8d00fd0d87625352ed5ba1a8861f72e.
---
include/GL/internal/dri_interface.h |1 +
src/egl/drivers/dri2/egl_dri2.h |2 --
2 files changed, 1
so, export EGL_PLATFORM=drm, is still working, right?
On Mon, Aug 15, 2011 at 3:47 PM, Benjamin Franzke
b...@kemper.freedesktop.org wrote:
Module: Mesa
Branch: master
Commit: 85fe9484945cb57ffd49df248b0e5057eba6af04
URL:
On Mon, 2011-08-15 at 09:57 +0200, Benjamin Franzke wrote:
Remove definition from egl_dri2.
Defining this is egl_dri2.h breaks as soon as
a new dri2 buffer token is added like with commit
4501a5d6e8d00fd0d87625352ed5ba1a8861f72e.
---
include/GL/internal/dri_interface.h |1 +
2011/8/15 Cooper Yuan coopery...@gmail.com:
so, export EGL_PLATFORM=drm, is still working, right?
Yes, and has a higher priority than the autodetection.
On Mon, Aug 15, 2011 at 3:47 PM, Benjamin Franzke
b...@kemper.freedesktop.org wrote:
Module: Mesa
Branch: master
Commit:
Hi All, please review this changelist that's going to enable eglWaitClient()
which has the same effect with glFinish according to specification.
before, egl applications using pixmap or pbuffer flicker always because of
no flush.
Cooper
---
src/gallium/state_trackers/dri/drm/dri2.c |4
- Original Message -
2011/8/12 Christian König deathsim...@vodafone.de:
Am Freitag, den 12.08.2011, 10:49 -0400 schrieb Younes Manton:
Sorry, by incompatible I didn't mean that you couldn't use them
together, but that one is more restrictive than the other. Like
the
discussion
On 13 August 2011 09:58, Kenneth Graunke kenn...@whitecape.org wrote:
On 08/12/2011 10:38 AM, Paul Berry wrote:
This patch fixes a bug when lowering an integer division:
x/y
to a multiplication by a reciprocal:
int(float(x)*reciprocal(float(y)))
If x was a a plain int and y was an
- Original Message -
On 13 August 2011 09:58, Kenneth Graunke kenn...@whitecape.org
wrote:
On 08/12/2011 10:38 AM, Paul Berry wrote:
This patch fixes a bug when lowering an integer division:
x/y
to a multiplication by a reciprocal:
From: Vincent Lejeune v...@ovi.com
This optimisation pass will look for and pack together float,
vec2, vec3 varyings in fragment shaders and transform the vertex
shader accordingly. It might improve performance depending on
the hardware.
---
src/glsl/Makefile |1 +
On 15 August 2011 08:50, Jose Fonseca jfons...@vmware.com wrote:
In places you don't have native int division support, you could use one
Newton-Raphson iteration step for almost accurate results, assuggested
accuracy of SSE2's RCPPS instructions. See for reference the following
llvmpipe
On 15 August 2011 10:11, Dan McCabe zen3d.li...@gmail.com wrote:
You might also want to consider implementing
quotient = int((float(x) + 0.5 * float(y)) * reciprocal(float(y)));
This rounds the result to the nearest integer rather then flooring the
result and is arguably faster (assuming
Have you had a chance to think about any of these issues, Ian? I'd
like to commit this patch series to Mesa.
On 5 August 2011 17:16, Paul Berry stereotype...@gmail.com wrote:
On 2 August 2011 18:27, Ian Romanick i...@freedesktop.org wrote:
+ *
+ * Also, perform implicit conversion
Here's my mti-tested branch of the map-texture-image-v4 work. It's
not regressing on intel or softpipe, and in fact fixes a couple of
tests now on those. It is not as complete a rework as
map-texture-image-v4, but it's quite a bit of the change and it it
should be bisectable (I've had to several
From: Brian Paul bri...@vmware.com
Several drivers have these fields in their subclasses of gl_texture_image.
They'll be useful for core Mesa too...
---
src/mesa/main/mtypes.h |2 ++
src/mesa/main/teximage.c |2 ++
2 files changed, 4 insertions(+), 0 deletions(-)
diff --git
This was produced by sed, except for one hunk in driverfuncs.c where
trailing whitespace was dropped.
---
src/mesa/drivers/common/driverfuncs.c |2 +-
src/mesa/drivers/dri/intel/intel_tex.c |2 +-
src/mesa/drivers/dri/intel/intel_tex_image.c |2 +-
From: Brian Paul bri...@vmware.com
---
src/mesa/state_tracker/st_cb_texture.c | 30 +++---
src/mesa/state_tracker/st_gen_mipmap.c |1 -
src/mesa/state_tracker/st_texture.c|4 ++--
src/mesa/state_tracker/st_texture.h|5 -
4 files changed, 17
From: Brian Paul bri...@vmware.com
ctx-Driver.MapTextureImage() / UnmapTextureImage() will be called by
the glTex[Sub]Image(), glGetTexImage() functions, etc. when we're
accessing texture data, and also for software rendering when accessing
texture data.
---
src/mesa/main/dd.h | 19
From: Brian Paul bri...@vmware.com
---
src/mesa/state_tracker/st_cb_texture.c | 45
1 files changed, 45 insertions(+), 0 deletions(-)
diff --git a/src/mesa/state_tracker/st_cb_texture.c
b/src/mesa/state_tracker/st_cb_texture.c
index 098a4d6..e4be7fb 100644
From: Brian Paul bri...@vmware.com
---
src/mesa/drivers/common/meta.c | 259 ++--
1 files changed, 167 insertions(+), 92 deletions(-)
diff --git a/src/mesa/drivers/common/meta.c b/src/mesa/drivers/common/meta.c
index f9b4755..19eae42 100644
---
All driver implementations of FreeTextureImageBuffer already check
that Data != NULL and free it. However, this means that we will also
free driver storage if the driver storage wasn't in the form of a Data
pointer.
This was produced by the following semantic patch:
@@
expression C;
expression
From: Brian Paul bri...@vmware.com
Decompress a texture by drawing quad with the texture.
---
src/mesa/drivers/common/meta.c | 200
src/mesa/drivers/common/meta.h |6 +
2 files changed, 206 insertions(+), 0 deletions(-)
diff --git
---
src/mesa/drivers/dri/intel/intel_tex.c | 84 ++
src/mesa/drivers/dri/intel/intel_tex.h |6 ++
src/mesa/drivers/dri/intel/intel_tex_image.c |4 +-
3 files changed, 92 insertions(+), 2 deletions(-)
diff --git
---
src/mesa/drivers/dri/intel/intel_tex.c |5 +++--
src/mesa/drivers/dri/unichrome/via_tex.c |6 +++---
src/mesa/state_tracker/st_cb_texture.c |4 ++--
3 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/src/mesa/drivers/dri/intel/intel_tex.c
From: Brian Paul bri...@vmware.com
No driver used the eject function, or set the validate hook that made
that function do anything.
---
src/mesa/swrast/s_context.c | 79 ---
1 files changed, 0 insertions(+), 79 deletions(-)
diff --git
This is untested, but should be close to working since it's basically
a copy of nouveau_teximage_map().
---
src/mesa/drivers/dri/nouveau/nouveau_texture.c | 81
1 files changed, 81 insertions(+), 0 deletions(-)
diff --git
Before, if we ended up here without a BO for our image, but did choose
a miptree that had active rendering in the command buffer, our
teximage data would jump ahead of the rendering using the old texture
contents.
This showed up as breakage in gen-teximage and friends in the
following commit.
---
From: Brian Paul bri...@vmware.com
---
src/mesa/main/debug.c | 94 ++--
src/mesa/main/debug.h |2 +-
2 files changed, 52 insertions(+), 44 deletions(-)
diff --git a/src/mesa/main/debug.c b/src/mesa/main/debug.c
index e7f6be9..4e05a45 100644
---
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c| 103 ++--
src/mesa/state_tracker/st_cb_texture.c | 38
2 files changed, 57 insertions(+), 84 deletions(-)
diff --git a/src/mesa/main/texgetimage.c
From: Brian Paul bri...@vmware.com
---
src/mesa/drivers/common/driverfuncs.c |2 +
src/mesa/sources.mak |1 +
src/mesa/swrast/s_texture.c | 109 +
src/mesa/swrast/swrast.h | 14
4 files changed, 126
It uses MapTextureImage() now, so we don't need our own mapping.
---
src/mesa/drivers/dri/intel/intel_tex_subimage.c | 149 ++-
1 files changed, 65 insertions(+), 84 deletions(-)
diff --git a/src/mesa/drivers/dri/intel/intel_tex_subimage.c
---
src/mesa/drivers/dri/radeon/radeon_texture.c | 77 ++
1 files changed, 77 insertions(+), 0 deletions(-)
diff --git a/src/mesa/drivers/dri/radeon/radeon_texture.c
b/src/mesa/drivers/dri/radeon/radeon_texture.c
index d9bc1c0..9964663 100644
---
---
src/mesa/drivers/dri/r200/r200_context.h |2 -
src/mesa/drivers/dri/r200/r200_tex.c | 41 ++
src/mesa/drivers/dri/r300/r300_tex.c | 28 ++
src/mesa/drivers/dri/r600/evergreen_tex.c| 28 ++
From: Brian Paul bri...@vmware.com
This continues to allocate texImage-Data as before, so
drivers calling these functions need to use that when present.
---
src/mesa/main/texstore.c | 336 +++---
1 files changed, 225 insertions(+), 111 deletions(-)
diff
It's handled by MapTextureImage() now.
---
src/mesa/drivers/dri/r200/Makefile|1 -
src/mesa/drivers/dri/r300/Makefile|1 -
src/mesa/drivers/dri/r600/Makefile|1 -
src/mesa/drivers/dri/radeon/Makefile |1 -
Now that Mesa core knows how to map teximages, we no longer needed the
compressed paths here.
---
src/mesa/drivers/dri/intel/intel_tex_subimage.c | 116 +++
1 files changed, 33 insertions(+), 83 deletions(-)
diff --git a/src/mesa/drivers/dri/intel/intel_tex_subimage.c
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c | 24 +++-
1 files changed, 15 insertions(+), 9 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index b98d708..a4e466c 100644
--- a/src/mesa/main/texgetimage.c
+++
From: Brian Paul bri...@vmware.com
This code is about to stop needing -Data and using MapTextureImage().
---
src/mesa/main/texgetimage.c |7 +++
1 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index 4c8a412..da07800
---
src/mesa/state_tracker/st_context.c |4
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/src/mesa/state_tracker/st_context.c
b/src/mesa/state_tracker/st_context.c
index 6d4bc54..e994e2e 100644
--- a/src/mesa/state_tracker/st_context.c
+++
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c | 11 ++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index 9f64b26..f5278e8 100644
--- a/src/mesa/main/texgetimage.c
+++
From: Brian Paul bri...@vmware.com
Our other unpacking code tends to use a datatype/format combo, while
these operate on data with a MESA_FORMAT type. Also note that
handling any BaseFormat overrides for channels is up to the caller.
---
src/mesa/SConscript |1 +
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c | 21 +++--
1 files changed, 15 insertions(+), 6 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index da07800..b98d708 100644
--- a/src/mesa/main/texgetimage.c
+++
It's handled by MapTextureImage() now.
---
src/mesa/drivers/dri/nouveau/nouveau_texture.c | 14 --
1 files changed, 0 insertions(+), 14 deletions(-)
diff --git a/src/mesa/drivers/dri/nouveau/nouveau_texture.c
b/src/mesa/drivers/dri/nouveau/nouveau_texture.c
index 38b89de..85eac83
The mesa core code uses MapTextureImage() like we need now.
---
src/mesa/drivers/dri/intel/intel_tex_image.c | 61 --
1 files changed, 0 insertions(+), 61 deletions(-)
diff --git a/src/mesa/drivers/dri/intel/intel_tex_image.c
Fixes regressions in compressed texture mipmap generation, which
relies on GetTexImage internally.
---
src/mesa/state_tracker/st_cb_texture.c | 33 +--
1 files changed, 2 insertions(+), 31 deletions(-)
diff --git a/src/mesa/state_tracker/st_cb_texture.c
From: Brian Paul bri...@vmware.com
Now that we can zero-copy generate the mipmaps into brand new
glTexImage()-generated storage using MapTextureImage(), we no longer
need to allocate image-Data in mipmap generate. This requires
deleting the drivers' old overrides of the miptree tracking after
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c| 22 +++---
src/mesa/state_tracker/st_cb_texture.c |3 ++-
2 files changed, 17 insertions(+), 8 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index
From: Brian Paul bri...@vmware.com
---
src/mesa/main/texgetimage.c | 18 +-
1 files changed, 13 insertions(+), 5 deletions(-)
diff --git a/src/mesa/main/texgetimage.c b/src/mesa/main/texgetimage.c
index a4e466c..aeff0da 100644
--- a/src/mesa/main/texgetimage.c
+++
From: Ian Romanick ian.d.roman...@intel.com
Since our logic values are 0.0 (false) and 1.0 (true), 1.0 - x
accurately implements logical not.
---
src/mesa/program/ir_to_mesa.cpp |8 +++-
1 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/src/mesa/program/ir_to_mesa.cpp
From: Ian Romanick ian.d.roman...@intel.com
Logical-or is implemented using addition (followed by clampling to
[0,1]) on values of 0.0 and 1.0. Replacing the logical-or operators
with addition gives a + b which has a result on the range [0, 2].
Previously a SNE instruction was used to clamp the
From: Ian Romanick ian.d.roman...@intel.com
---
src/mesa/program/ir_to_mesa.cpp | 14 +++---
1 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/src/mesa/program/ir_to_mesa.cpp b/src/mesa/program/ir_to_mesa.cpp
index 5442976..3d5ba5b 100644
---
From: Ian Romanick ian.d.roman...@intel.com
This is just like the ir_binop_logic_or case. The operation
ir_unop_any is (a.x || a.y || a.z || a.w). Logical-or is implemented
using addition (followed by clampling to [0,1]) on values of 0.0 and
1.0. Replacing the logical-or operators with
From: Ian Romanick ian.d.roman...@intel.com
The operation ir_binop_all_equal is !(a.x != b.x || a.y != b.y || a.z
!= b.z || a.w != b.w). Logical-or is implemented using addition
(followed by clampling to [0,1]) on values of 0.0 and 1.0. Replacing
the logical-or operators with addition gives
From: Ian Romanick ian.d.roman...@intel.com
!a b occurs frequently when nexted if-statements have been
flattened. It should also be possible use a MAD for (a b) || c,
though that would require a MAD_SAT.
---
src/mesa/program/ir_to_mesa.cpp | 52 +++
1
From: Ian Romanick ian.d.roman...@intel.com
This cleans up some code generated by the IR-to-Mesa pass for i915.
In particular, some shaders involving arrays of constant matrices
result in really bad code.
---
src/mesa/SConscript |1 +
From: Ian Romanick ian.d.roman...@intel.com
Shader Model 3.0[1] requires that shaders be able to execute at least
65536 instructions. Bump Mesa maxExec to that limit. This allows
several vertex shaders in the OpenGL ES 2.0 conformance test suite to
run to completion.
1:
I implemented functions for horizontal/vertical alignment units separately
because I find it easier to read that way...especially with all the
corner-cases.
Regresses piglit test fbo-generatemipmap-formats GL_ARB_depth_texture
on Sandybridge: the 1x1 quad is 0.0 rather than 0.5. This is caused
On Mon, 15 Aug 2011 12:21:06 -0700, Kenneth Graunke kenn...@whitecape.org
wrote:
I implemented functions for horizontal/vertical alignment units separately
because I find it easier to read that way...especially with all the
corner-cases.
Regresses piglit test fbo-generatemipmap-formats
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/05/2011 05:16 PM, Paul Berry wrote:
On 2 August 2011 18:27, Ian Romanick i...@freedesktop.org wrote:
+ *
+ * Also, perform implicit conversion of arguments. Note: to
+ * implicitly convert out parameters, we need to place
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/12/2011 10:38 AM, Paul Berry wrote:
This patch fixes a bug when lowering an integer division:
x/y
to a multiplication by a reciprocal:
int(float(x)*reciprocal(float(y)))
If x was a a plain int and y was an ivecN, the lowering
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/11/2011 05:38 PM, Kenneth Graunke wrote:
Core Mesa now loads float/integer/boolean uniform data for drivers that
claim to support GLSL 1.30, thanks to Bryan Cain's recent work. When we
run in 1.30 mode, we need to avoid double conversions.
On 08/08/2011 08:04 AM, Brian Paul wrote:
On 08/05/2011 06:21 PM, Kenneth Graunke wrote:
On 08/05/2011 06:45 AM, Brian Paul wrote:
On Thu, Aug 4, 2011 at 4:26 PM, Kenneth
Graunkekenn...@whitecape.org wrote:
Hey,
I'd like to remove libGLw from the main Mesa repository. It never
changes,
This optimisation pass will look for and pack together float,
vec2, vec3 varyings in fragment shaders and transform the vertex
shader accordingly. It might improve performance depending on
the hardware.
---
src/glsl/Makefile |1 +
src/glsl/SConscript|1 +
On Mon, 15 Aug 2011 12:02:42 -0700, Ian Romanick i...@freedesktop.org wrote:
From: Ian Romanick ian.d.roman...@intel.com
This cleans up some code generated by the IR-to-Mesa pass for i915.
In particular, some shaders involving arrays of constant matrices
result in really bad code.
I'm
On Mon, 15 Aug 2011 12:02:35 -0700, Ian Romanick i...@freedesktop.org wrote:
From: Ian Romanick ian.d.roman...@intel.com
Since our logic values are 0.0 (false) and 1.0 (true), 1.0 - x
accurately implements logical not.
This series is:
Reviewed-by: Eric Anholt e...@anholt.net
except for the
On 08/15/2011 01:41 PM, Kenneth Graunke wrote:
The new repo is up now (thanks Eric!):
git://git.freedesktop.org/git/mesa/glw
Apparently what I meant to say was:
git://anongit.freedesktop.org/mesa/glw
___
mesa-dev mailing list
On Mon, 15 Aug 2011 11:26:51 -0700, Chad Versace c...@chad-versace.us wrote:
Remove the inclusion of fpu_control.h from compiler.h. Since compiler.h
uses no symbol from fpu_control.h, the removal is safe.
Since Bionic lacks fpu_control.h, its inclusion breaks the Android build.
I think this
On 08/15/2011 02:54 PM, Eric Anholt wrote:
On Mon, 15 Aug 2011 11:26:51 -0700, Chad Versacec...@chad-versace.us wrote:
Remove the inclusion of fpu_control.h from compiler.h. Since compiler.h
uses no symbol from fpu_control.h, the removal is safe.
Since Bionic lacks fpu_control.h, its
On Thu, 11 Aug 2011 17:38:37 -0700, Kenneth Graunke kenn...@whitecape.org
wrote:
Prior to this patch, POW was the only two-operand math operation, so a
few cases needed to be updated.
Signed-off-by: Kenneth Graunke kenn...@whitecape.org
diff --git
On Thu, 11 Aug 2011 17:38:38 -0700, Kenneth Graunke kenn...@whitecape.org
wrote:
The MATH instruction unfortunately only operates on GRF registers with
an HStride of 1, so we often need to create a temporary GRF. For
integer division and modulus, we need to use an integer register.
On 08/15/2011 10:45 AM, Paul Berry wrote:
On 15 August 2011 10:11, Dan McCabezen3d.li...@gmail.com wrote:
You might also want to consider implementing
quotient = int((float(x) + 0.5 * float(y)) * reciprocal(float(y)));
This rounds the result to the nearest integer rather then flooring the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
This series is huge, so I'm going to review it in batches. I've
reviewed 01/37 through 10/37, inclusive. I'm sending replies with
comments to 01/37 and 05/37. The rest look good and get my R-b.
On 08/15/2011 11:53 AM, Eric Anholt wrote:
From:
Previously, native integer support was based on whether the driver
advertised GLSL 1.30 or not. However, drivers that natively support
integers may wish to do so for older GLSL versions as well. Adding this
new opt-in flag allows them to do so.
Currently disabled by default on all drivers,
On Mon, 15 Aug 2011 15:02:31 -0700, Kenneth Graunke kenn...@whitecape.org
wrote:
Previously, native integer support was based on whether the driver
advertised GLSL 1.30 or not. However, drivers that natively support
integers may wish to do so for older GLSL versions as well. Adding this
new
On Mon, 15 Aug 2011 15:02:32 -0700, Kenneth Graunke kenn...@whitecape.org
wrote:
When ctx-Const.NativeIntegers is set, Core Mesa loads integer/boolean
uniforms directly, rather than loading the floating point equivalent.
So, when that's set, we don't need to perform any conversions.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/15/2011 01:44 PM, Eric Anholt wrote:
On Mon, 15 Aug 2011 12:02:42 -0700, Ian Romanick i...@freedesktop.org
wrote:
From: Ian Romanick ian.d.roman...@intel.com
This cleans up some code generated by the IR-to-Mesa pass for i915.
In
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/15/2011 01:58 PM, Brian Paul wrote:
On 08/15/2011 02:54 PM, Eric Anholt wrote:
On Mon, 15 Aug 2011 11:26:51 -0700, Chad Versacec...@chad-versace.us
wrote:
Remove the inclusion of fpu_control.h from compiler.h. Since compiler.h
uses no
On Mon, 15 Aug 2011 16:17:20 -0700, Ian Romanick i...@freedesktop.org wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/15/2011 01:44 PM, Eric Anholt wrote:
On Mon, 15 Aug 2011 12:02:42 -0700, Ian Romanick i...@freedesktop.org
wrote:
From: Ian Romanick ian.d.roman...@intel.com
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/15/2011 11:53 AM, Eric Anholt wrote:
From: Brian Paul bri...@vmware.com
This continues to allocate texImage-Data as before, so
drivers calling these functions need to use that when present.
---
src/mesa/main/texstore.c | 336
On Mon, 15 Aug 2011 16:55:45 -0700, Chad Versace c...@chad-versace.us wrote:
Remove the inclusion of fpu_control.h from compiler.h. Since Bionic lacks
fpu_control.h, this fixes the Android build.
Also remove the sole use of the fpu_control bits, which was in debug.c.
Those were brianp's
Hi,
I tried to give some thought to how this interface would work for
decoders that can't or would prefer not to support multiple decode
buffers, and most of the scenarios I came up with seem to work out, so
overall I'm not opposed to it. Even though I still think this can be
done by each driver
80 matches
Mail list logo