| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
because single-pipe cards have bigger CMASK RAM
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
[ Squashed port of the following r600g commits: - Michel Dänzer ]
commit 428e37c2da420f7dc14a2ea265f2387270f9bee1
Author: Marek Olšák <[email protected]>
Date: Tue Oct 2 22:02:54 2012 +0200
r600g: add in-place DB decompression and texturing with DB tiling
The decompression is done in-place and only the compressed tiles are
decompressed. Note: R6xx-R7xx can do that only with Z16 and Z32F.
The texture unit is programmed to use non-displayable tiling and depth
ordering of samples, so that it can fetch the texture in the native DB format.
The latest version of the libdrm surface allocator is required for stencil
texturing to work. The old one didn't create the mipmap tree correctly.
We need a separate mipmap tree for stencil, because the stencil mipmap
offsets are not really depth offsets/4.
There are still some known bugs, but this should save some memory and it also
improves performance a little bit in Lightsmark (especially with low
resolutions; tested with Radeon HD 5000).
The DB->CB copy is still used for transfers.
commit e2f623f1d6da9bc987582ff68d0471061ae44030
Author: Marek Olšák <[email protected]>
Date: Sat Jul 28 13:55:59 2012 +0200
r600g: don't decompress depth or stencil if there isn't any
commit 43e226b6efb77db2247741cc2057d9625a2cfa05
Author: Marek Olšák <[email protected]>
Date: Wed Jul 18 00:32:50 2012 +0200
r600g: optimize uploading depth textures
Make it only copy the portion of a depth texture being uploaded and
not the whole 2D layer.
There is also a little code cleanup.
commit b242adbe5cfa165b252064a1ea36f802d8251ef1
Author: Marek Olšák <[email protected]>
Date: Wed Jul 18 00:17:46 2012 +0200
r600g: remove needless wrapper r600_texture_depth_flush
commit 611dd529425281d73f1f0ad2000362d4a5525a25
Author: Marek Olšák <[email protected]>
Date: Wed Jul 18 00:05:14 2012 +0200
r600g: init_flushed_depth_texture should be able to report errors
commit 80755ff56317446a8c89e611edc1fdf320d6779b
Author: Marek Olšák <[email protected]>
Date: Sat Jul 14 17:06:27 2012 +0200
r600g: properly track which textures are depth
This fixes the issue with have_depth_texture never being set to false.
commit fe1fd675565231b49d3ac53d0b4bec39d8bc6781
Author: Marek Olšák <[email protected]>
Date: Sun Jul 8 03:10:37 2012 +0200
r600g: don't flush depth textures set as colorbuffers
The only case a depth buffer can be set as a color buffer is when flushing.
That wasn't always the case, but now this code isn't required anymore.
commit 5a17d8318ec2c20bf86275044dc8f715105a88e7
Author: Marek Olšák <[email protected]>
Date: Sun Jul 8 02:14:18 2012 +0200
r600g: flush depth textures bound to vertex shaders
This was missing/broken. There are also minor code cleanups.
commit dee58f94af833906863b0ff2955b20f3ab407e63
Author: Marek Olšák <[email protected]>
Date: Sun Jul 8 01:54:24 2012 +0200
r600g: do fine-grained depth texture flushing
- maintain a mask of which mipmap levels are dirty (instead of one big flag)
- only flush what was requested at a given point and not the whole resource
(most often only one level and one layer has to be flushed)
Signed-off-by: Michel Dänzer <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use r600_resource_texture::flished_depth_texture for GPU access, and
allocate it in the VRAM. For transfers we'll allocate texture in the GTT
and store it in the r600_transfer::staging.
Improves performance when flushed depth texture is frequently used by the
GPU, e.g. in Lightsmark
[ Ported from r600g commit 37708479608af877986b76302a9c92611d1e23d0 ]
Signed-off-by: Michel Dänzer <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
[ Squashed port of the following r600g commits: - Michel Dänzer ]
commit c1e8c845ea9c6f843cc5bba5974668c007799bbc
Author: Marek Olšák <[email protected]>
Date: Sat Jul 7 19:10:00 2012 +0200
r600g: inline r600_hw_copy_region
commit 4891c5dc64ccd8cf2bf8a8550ae23e1a61806a7d
Author: Marek Olšák <[email protected]>
Date: Mon Jun 25 22:53:21 2012 +0200
r600g: inline r600_blit_push_depth and use resource_copy_region
We are going to have a separate resource for depth texturing and transfers
and this is just a transfer thing.
commit da98bb6fc105e1a2f688a1713ca9e50f0ac8fbed
Author: Marek Olšák <[email protected]>
Date: Mon Jun 25 12:45:32 2012 +0200
r600g: split flushed depth texture creation and flushing
Signed-off-by: Michel Dänzer <[email protected]>
|
|
|
|
|
|
|
| |
No piglit regressions anymore thanks to fixes in libdrm_radeon and here.
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Michel Dänzer <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Michel Dänzer <[email protected]>
|
|
|
|
|
|
|
| |
The static Makefiles used it.
Reviewed-by: Kenneth Graunke <[email protected]>
Reviewed-by: Andreas Boll <[email protected]>
|
|
|
|
| |
Reviewed-by: Andreas Boll <[email protected]>
|
|
|
|
| |
Reviewed-by: Andreas Boll <[email protected]>
|
|
|
|
|
|
|
| |
Fixes missing options.h when doing 'make check' in dri/common before
'make' has been run.
Reviewed-by: Andreas Boll <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
|
|
|
| |
Allow additional format/type combinations based on the
color render buffer to fix failures with gles3-gtf.
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
|
|
|
|
| |
For GLES2/3 allow reading of pixels with format/type based on:
* GL_IMPLEMENTATION_COLOR_READ_FORMAT
* GL_IMPLEMENTATION_COLOR_READ_TYPE
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
|
|
|
|
| |
[mattst88] v2: Enable only for ES3 per spec.
[mattst88] v3: Use _mesa_is_gles3 since EXT_color_buffer_float is
ES3-only.
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
|
|
| |
Changes based on GTF/gles3 conformance test suite.
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
Signed-off-by: Jordan Justen <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
For now I'm just enabling this on the same subset of hardware that has
OpenGL 3.0 enabled. This same functionality is part of OpenGL 3.0, and
there is no matching desktop extension.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
| |
They're part of GL_OES_depth_texture_cube_map, and we'll always enable
that extension in ES3 contexts.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
| |
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
| |
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The vs part hasn't been wired up since tgsi_sse2 was disabled in:
commit 4eb3225b38ce12cb34ab3d90804c9683bd7b4ed3
Author: José Fonseca <[email protected]>
Date: Tue Nov 8 00:10:47 2011 +0000
Remove tgsi_sse2.
And it would certainly not work correctly in its current state:
draw/draw_vs_ppc.c: In function ‘draw_create_vs_ppc’:
draw/draw_vs_ppc.c:190:24: warning: assignment from incompatible pointer
type [enabled by default]
As with the sse2 backend, this should be done in llvm anyway.
Reviewed-by: Brian Paul <[email protected]>
Signed-off-by: Adam Jackson <[email protected]>
|
| |
|
|
|
|
| |
I really did build test, my compiler just doesn't seem to care.
|
|
|
|
| |
NOTE: This is a candidate for the 9.0 branch.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is required by OpenGL ES 3.0 and desktop OpenGL 4.2. Previous
version were ambiguous. This also matches the behavior of NVIDIA's
closed-source driver (version 304.64).
Fixed gles3conformance test uniform_buffer_object_getactiveuniformsiv
and uniform_buffer_object_structure_and_array_element_names (on my
in-progress branch that fixes a bunch of other stuff...YMMV).
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is required by OpenGL ES 3.0 and desktop OpenGL 4.2. Previous
version were ambiguous. This also matches the behavior of NVIDIA's
closed-source driver (version 304.64).
Fixed gles3conformance test uniform_buffer_object_getactiveuniform.
Several piglit tests expect glGetActiveUniform to *not* include the [0]
on the end. These tests were already failing on NVIDIA, and this change
regresses them on Mesa. Patches have been sent to the piglit mailing
list to fix the tests.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
| |
We currently have a bug in this code, and I don't want to fix it in two
places.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
| |
It always returns true, so there's no point in having a return value.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
| |
This looks like a copy-and-paste left over.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The GLSL 1.40 spec says:
"Uniform block names and variable names declared within uniform
blocks are scoped at the program level."
Track the block name in the symbol table and emit errors when conflicts
exist.
Fixes es3conform's uniform_buffer_object_block_name_conflict test, and
fixes the piglit block-name-clashes-with-{variable,function,struct}.vert
tests.
NOTE: This is a candidate for the 9.0 branch.
v2: Fix bad constructor initialization. Noticed by Topi Pohjolainen.
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
About both row_major and column_major layout qualifiers, the GLSL spec
says:
"It only affects the layout of matrices."
However, the OpenGL ES 3.0 conformance tests have taken this to mean it
is an error use it elsewhere. This seems logical given that
'layout(row_major) vec4 foo' is probably not what the programmer meant.
The only catch is dealing with structures that contain matrices. Layout
qualifiers cannot be applied directly to fields of structures, so the
only way to affect the layout of the fields is to apply a qualifier to
the structure declaration itself. There is ongoing debate about this
within Khronos, and it seems to be settling in favor of allowing the
qualifiers on structures. I light of this, I have chosen to allow the
qualifiers on structures but emit a warning since the usage may not be
portable.
Fixes gles3conform test
uniform_buffer_object_layouts_not_for_matrix_type and causes no
regressions.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
| |
Between the previous commit and this one, improves GLBenchmark 2.1
offscreen performance by 0.48% +/- 0.24% (n=22, throttling outliers
removed).
Reviewed-by: Kenneth Graunke <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
| |
Reviewed-by: Kenneth Graunke <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
| |
It's never going to have texture fetches.
Reviewed-by: Kenneth Graunke <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
| |
Fixes a crash in GLB2.1 offscreen on the glthread branch.
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
| |
Tested with piglit ARB_texture_buffer_object/formats.
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
| |
This is for GL_ARB_texture_buffer_object_rgb32 support, but it also
causes the format to get used for float32 rgb textures as well on
Ironlake and later. Since that came with some surprises, separate
the change from the enable commit.
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
We almost never want a stride in pixels -- if you're doing anything with
a stride, you're specifying an offset or incrementing a pointer, and in
both cases you had to multiply by cpp to get the bytes value you wanted.
But worse, on the way to creating a region from a new tiled BO, we
divided by cpp to get pitch in pixels, and for an RGB32 buffer (an
upcoming change) the pitch wouldn't divide exactly, and we'd end up with
a wrong stride in our region.
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
|
|
| |
As we gain support for NPOT cpp, a pitch may not divide by cpp cleanly.
Reviewed-by: Kenneth Graunke <[email protected]>
|
|
|
|
| |
Reviewed-by: Tom Stellard <[email protected]>
|
|
|
|
| |
Reviewed-by: Tom Stellard <[email protected]>
|
|
|
|
| |
Reviewed-by: Tom Stellard <[email protected]>
|