OSDN Git Service

android-x86/external-mesa.git
14 years agonvfx: implement DP2 in vp and fp
Luca Barbieri [Sat, 21 Aug 2010 17:43:46 +0000 (19:43 +0200)]
nvfx: implement DP2 in vp and fp

14 years agonvfx: implement TRUNC in vp and fp
Luca Barbieri [Sat, 21 Aug 2010 17:35:06 +0000 (19:35 +0200)]
nvfx: implement TRUNC in vp and fp

14 years agonvfx: implement NOP
Luca Barbieri [Sat, 21 Aug 2010 17:45:06 +0000 (19:45 +0200)]
nvfx: implement NOP

14 years agonvfx: add vertex program control flow
Luca Barbieri [Sat, 21 Aug 2010 16:37:21 +0000 (18:37 +0200)]
nvfx: add vertex program control flow

14 years agonvfx: fix vertex shader headers
Luca Barbieri [Sat, 21 Aug 2010 16:37:01 +0000 (18:37 +0200)]
nvfx: fix vertex shader headers

14 years agonv40: add fragment program control flow
Luca Barbieri [Fri, 20 Aug 2010 19:16:49 +0000 (21:16 +0200)]
nv40: add fragment program control flow

14 years agonvfx: refactor shader assembler
Luca Barbieri [Sat, 21 Aug 2010 10:32:59 +0000 (12:32 +0200)]
nvfx: refactor shader assembler

14 years agonvfx: add option to dump shaders in TGSI and native code
Luca Barbieri [Sat, 21 Aug 2010 11:28:38 +0000 (13:28 +0200)]
nvfx: add option to dump shaders in TGSI and native code

14 years agonvfx: improve and correct nvfx_shader.h
Luca Barbieri [Thu, 25 Feb 2010 16:46:37 +0000 (17:46 +0100)]
nvfx: improve and correct nvfx_shader.h

14 years agonvfx: fix lodbias
Luca Barbieri [Thu, 19 Aug 2010 20:47:03 +0000 (22:47 +0200)]
nvfx: fix lodbias

14 years agonvfx: mostly fix inline corruption magically
Luca Barbieri [Thu, 19 Aug 2010 20:36:00 +0000 (22:36 +0200)]
nvfx: mostly fix inline corruption magically

Not sure why this mostly works.

14 years agonvfx: fix GPU hardlocks when depth buffer is absent
Luca Barbieri [Thu, 19 Aug 2010 10:58:14 +0000 (12:58 +0200)]
nvfx: fix GPU hardlocks when depth buffer is absent

14 years agonvfx: fire ring after transfers
Luca Barbieri [Mon, 16 Aug 2010 23:01:42 +0000 (01:01 +0200)]
nvfx: fire ring after transfers

Might reduce the risk of running out of memory

14 years agonv30: band-aid viewport issues
Luca Barbieri [Mon, 16 Aug 2010 14:55:00 +0000 (16:55 +0200)]
nv30: band-aid viewport issues

For some reason nv30 seems to like to reset the viewport, even though
attempts to isolate where exactly it does that have currently been
inconclusive.

14 years agonvfx: support flatshade_first
Luca Barbieri [Sun, 15 Aug 2010 08:15:40 +0000 (10:15 +0200)]
nvfx: support flatshade_first

14 years agonvfx: expose GLSL
Luca Barbieri [Sat, 13 Mar 2010 01:28:59 +0000 (02:28 +0100)]
nvfx: expose GLSL

Still no control flow support, but basic stuff works.

14 years agonvfx: support proper shader linkage - adds glsl support
Luca Barbieri [Tue, 10 Aug 2010 21:09:53 +0000 (23:09 +0200)]
nvfx: support proper shader linkage - adds glsl support

14 years agonvfx: rewrite draw code and buffer code
Luca Barbieri [Sat, 7 Aug 2010 03:39:18 +0000 (05:39 +0200)]
nvfx: rewrite draw code and buffer code

This is a full rewrite of the drawing and buffer management logic.

It offers a lot of improvements:
1. A copy of buffers is now always kept in system memory. This is
   necessary to allow software processing of them, which is necessary
   or improves performance in many cases.
2. Support for pushing vertices on the FIFO, with index lookup if necessary.
3. "Smart" draw code that tries to intelligently choose the cheapest
  way to draw something: whether to use inline vertices or hardware
  vertex buffer, and whether to use hardware index buffers
4. Support for all vertex formats supported by the hardware
5. Usage of translate to push vertices, supporting all formats that are
   sensible to use as vertex formats
6. Support for base vertex
7. Usage of Ben Skeggs' primitive splitter originally for nv50, allowing
   correct splitting of line loops, triangle fans, etc.
8. Support for instancing
9. Precomputation using the vertex elements CSO

Thanks to Ben Skeggs for his primitive splitter originally for nv50.

Thanks to Christoph Bumiller for his nv50 push code, that was the basis
of this work, even though I changed his code dramatically, in particular
to replace his ad-hoc vertex data emitter with translate.

The changes could also go into nv50 too, but there are substantial
differences due to the additional nv50 hardware features.

14 years agonvfx: refactor sampling code, add support for swizzles and depth tex
Luca Barbieri [Sat, 7 Aug 2010 01:47:25 +0000 (03:47 +0200)]
nvfx: refactor sampling code, add support for swizzles and depth tex

This is a significant refactoring of the sampling code that:
- Moves all generic functions in nvfx_fragtex.c
- Adds a driver-specific sampler view structure and uses it to
  precompute texture setup as it should be done
- Unifies a bit more of code between nv30 and nv40
- Adds support for sampler view swizzles
- Support for specifying as sampler view format different from the
  resource one (only trivially)
- Support for sampler view specification of first and last level
- Support for depth textures on nv30, both for reading depth and
  for compare
- Support for sRGB textures
- Unifies the format table between nv30 and nv40
- Expands the format table to include essentially all supportable formats
  except mixed sign and "autonormal" formats
- Fixes the "is format supported" logic, which was quite broken, and
  makes it use the format table

Only tested on nv30 currently.

14 years agonvfx: new 2D: unify textures and buffers
Luca Barbieri [Tue, 3 Aug 2010 20:49:19 +0000 (22:49 +0200)]
nvfx: new 2D: unify textures and buffers

Stop using the vtbl, and use real transfers for buffers too.

14 years agonvfx: new 2D: use a CPU copy for up to 4 pixels, up from 0
Luca Barbieri [Sun, 18 Apr 2010 14:43:19 +0000 (16:43 +0200)]
nvfx: new 2D: use a CPU copy for up to 4 pixels, up from 0

Seems a reasonable threshold for now.

Significantly speeds up Piglit's 1x1 glReadPixels (but, you know,
reading pixels in 1x1 blocks is NOT a good idea, especially if you
might be running on a less-than-perfect driver).

14 years agonvfx: new 2D: new render temporaries with resources
Luca Barbieri [Tue, 3 Aug 2010 03:47:41 +0000 (05:47 +0200)]
nvfx: new 2D: new render temporaries with resources

This patch adds support for creating temporary surfaces to allow
rendering to surfaces that cannot be rendered to.
It uses the _second_ version of the render temporary infrastructure.

This is necessary for swizzled 3D textures and small mipmaps of
swizzled 2D textures.

This version of the patch creates a resource to use as a temporary
instead of a raw BO, making the code simpler.

14 years agonv30: new 2D: support ARB_texture_rectangle
Luca Barbieri [Thu, 11 Mar 2010 17:06:28 +0000 (18:06 +0100)]
nv30: new 2D: support ARB_texture_rectangle

This uses nv30's _RECT formats.

14 years agonvfx: new 2D: optimize fragtex format lookup
Luca Barbieri [Tue, 23 Feb 2010 11:47:45 +0000 (12:47 +0100)]
nvfx: new 2D: optimize fragtex format lookup

Use an array indexed by the pipe format instead of doing a linear scan.

14 years agonvfx: new 2D: enable swizzling for all surfaces
Luca Barbieri [Mon, 18 Jan 2010 22:40:22 +0000 (23:40 +0100)]
nvfx: new 2D: enable swizzling for all surfaces

Now that the new 2D code is in place, swizzling can be safely enabled.

Render temporaries are needed in some cases, so this may degrade nv30
a bit until it gets render temporaries too.

14 years agonvfx: new 2D: use new 2D engine in Gallium
Luca Barbieri [Tue, 3 Aug 2010 04:24:22 +0000 (06:24 +0200)]
nvfx: new 2D: use new 2D engine in Gallium

This patch implements nv04_surface_copy/fill using the new 2D engine module.
It supports falling back to the 3D engine using the u_blitter module, which will be
added in a later patch.

Also adds support for using the 3D engine, reusing the u_blitter module
created for r300.
This is used for unswizzling and copies between swizzled surfaces.

14 years agonv04-nv40: new 2D: add new Gallium-independent 2D engine
Luca Barbieri [Tue, 19 Jan 2010 17:51:10 +0000 (18:51 +0100)]
nv04-nv40: new 2D: add new Gallium-independent 2D engine

This patch add a brand new nv04-nv40 2D engine module.
It should correctly implement all operations involving swizzled, and 3D-swizzled surfaces.

This code is independent from the Gallium framework and can thus be reused in the DDX and classic Mesa drivers (it's only likely to be useful in the latter, though).

Currently, surface_copy and surface_fill are broken for 3D textures, for swizzled source textures and possibly for some misaligned cases

The code is based around the new nv04_region structure, which encapsulates the information from pipe_surface needed for the 2D engine and CPU copies.
The use of nv04_region makes the code independent of the Gallium framework and allows to transform the nv04_region without clobbering the nv04_region.
The existing M2MF, blitter, and SWIZZLED_SURFACE paths have been improved and a new CPU path has been added.
There is also support to tell the caller to use the 3D engine.

The main feature of the copy/fill setup algorithm is linearization/contiguous-linearization of swizzled surfaces.
The idea of linearization is that some swizzled surfaces are laid out like linear ones (1xN, 2xN, Nx1) and can thus be used as such (e.g. useful for copying single pixels).
Also, some rectangles (e.g. the whole surface) are contiguous in memory. If both the source and destination rectangles are swizzled but contiguous, then they can be regarded as both linear: this is the idea of "contiguous linearization".
This, for instance, allows to use the 2D engine to duplicate the content of a swizzled surface to another swizzled surface, by pretending they are actually linear.
After linearization, the result may not be 64-byte aligned. Another transformation is done to enlarge the linear surface so that it becomes 64-byte aligned.
This is also used to 64-byte align swizzled texture mipmaps.

The inner loop of the CPU path is as optimized as possible without using SSE/SSE2.
Future improvements could include SSE/SSE2 support, and possibly a faster coordinate swizzling algorithm (which is however not used in the inner loop).
It may be a good idea to autogenerate swizzling code at least for all possible POT 2D texture dimensions  (less than 256), maybe for all 3D ones too (less than 4096).
Also, it woud be a very good idea to make a copy with the GPU first if the source surface is in uncached memory.

14 years agonvfx: new 2D: rewrite transfer code to use staging transfers
Luca Barbieri [Tue, 3 Aug 2010 19:32:42 +0000 (21:32 +0200)]
nvfx: new 2D: rewrite transfer code to use staging transfers

This greatly simplifies the code, and avoids ad-hoc copy code.

Also, these new transfers work for buffers too, even though they
are still used for miptrees only.

14 years agonvfx: new 2D: rewrite miptree code, adapt transfers
Luca Barbieri [Fri, 15 Jan 2010 09:11:11 +0000 (10:11 +0100)]
nvfx: new 2D: rewrite miptree code, adapt transfers

Changes:
- Disable swizzling on non-RGBA 2D textures, since the current 2D
  code is mostly broken in those cases. A later patch will fix this.
  Thanks to Andrew Randrianasulu who reported this.
- Fix compressed texture transfers and hack around the current 2D
  code inability to copy compressed textures by using direct access.
  Thanks to Andrew Randrianasulu who reported this.

This patch rewrites all the miptree layout and transfer code in the
nvfx driver.

The current code is broken in several ways:
1. 3D textures are laid out first by face, then by level, which is
incorrect
2. Cube maps should have 128-byte aligned faces
3. Swizzled textures have a strange alignment test that seems
unnecessary
4. We store the image_offsets for each face/slice but they can be
easily computed instead
5. "Swizzling" is not supported for compressed formats. They can be
"swizzled" but swizzling only means that there are no gaps (pitch is
level-dependant) and the layout is still linear
6. Swizzling is not supported for non-RGBA formats. All formats (except
possibly depth) can be swizzled according to my testing.

The miptree layout is rewritten based on my empirical testing, which I
posted in the "miptree findings" mail.
The image_offset array is removed, since it can be calculated with a
simple multiplication; the only array in the miptree structure is now
the one for mipmap level starts, which it seems cannot be easily
computed in constant time.

Also, we now directly store a nouveau_bo instead of a pipe_buffer in
the miptree structure, like nv50 does.

Support for render temporaries is removed, and will be readded in a
later patch.

Note that the current temporary code is broken, because it does not
copy the temporary back on render cache flushes.

14 years agonvfx: add nouveau_resource_on_gpu
Luca Barbieri [Fri, 6 Aug 2010 23:56:01 +0000 (01:56 +0200)]
nvfx: add nouveau_resource_on_gpu

Add a function to get whether a resource is likely on the GPU or not.

Currently always returns TRUE.

14 years agonvfx: add linear flag for buffers
Luca Barbieri [Tue, 3 Aug 2010 20:50:19 +0000 (22:50 +0200)]
nvfx: add linear flag for buffers

14 years agonvfx: properly unreference bound objects on context destruction
Luca Barbieri [Mon, 9 Aug 2010 03:05:12 +0000 (05:05 +0200)]
nvfx: properly unreference bound objects on context destruction

14 years agonvfx: reference count bound objects
Luca Barbieri [Fri, 6 Aug 2010 07:15:00 +0000 (09:15 +0200)]
nvfx: reference count bound objects

14 years agonvfx: fix format support code for compressed texture
Luca Barbieri [Sat, 7 Aug 2010 00:49:48 +0000 (02:49 +0200)]
nvfx: fix format support code for compressed texture

A source line was put in the wrong place.

14 years agogallium/auxiliary: add semantic linkage utility code
Luca Barbieri [Thu, 25 Feb 2010 12:08:35 +0000 (13:08 +0100)]
gallium/auxiliary: add semantic linkage utility code

14 years agou_debug_describe: use switch instead of if chain
Luca Barbieri [Sat, 21 Aug 2010 10:47:18 +0000 (12:47 +0200)]
u_debug_describe: use switch instead of if chain

14 years agou_debug_describe: add PIPE_TEXTURE_RECT
Luca Barbieri [Sat, 21 Aug 2010 10:45:39 +0000 (12:45 +0200)]
u_debug_describe: add PIPE_TEXTURE_RECT

14 years agoauxiliary: add copyright headers
Luca Barbieri [Sat, 21 Aug 2010 10:37:18 +0000 (12:37 +0200)]
auxiliary: add copyright headers

Thanks to Jose Fonseca for pointing out they were missing.

14 years agoutil: Match printf format to silence warning.
José Fonseca [Sat, 21 Aug 2010 09:38:22 +0000 (10:38 +0100)]
util: Match printf format to silence warning.

14 years agomesa: Remove unsused local variable.
José Fonseca [Sat, 21 Aug 2010 09:34:57 +0000 (10:34 +0100)]
mesa: Remove unsused local variable.

14 years agoutil: Make the reference debuggin code more C++ friendly.
José Fonseca [Sat, 21 Aug 2010 09:34:42 +0000 (10:34 +0100)]
util: Make the reference debuggin code more C++ friendly.

C++ doesn't accept function <-> void* conversions without a putting a
fight.

14 years agoutil: Remove the x86 exception handlers.
José Fonseca [Sat, 21 Aug 2010 09:07:12 +0000 (10:07 +0100)]
util: Remove the x86 exception handlers.

Unused now that check_os_katmai_support was removed.

14 years agotrace: Don't immediately destroy the pipe's sampler view in the trace driver.
Alex Corscadden [Tue, 17 Aug 2010 20:45:31 +0000 (13:45 -0700)]
trace: Don't immediately destroy the pipe's sampler view in the trace driver.

The trace driver's implementation of sampler_view_destroy was calling
directly into the underlying pipe's sampler_view_destroy implementation.
This causes problems for pipes that keep references to sampler views
even after the state tracker has released them.  Instead, we'll simply
drop the trace driver's reference to the pipe's sampler view.

Signed-off-by: José Fonseca <jfonseca@vmware.com>
14 years agotrace: Trace the correct version of the resource when setting the index buffer.
Alex Corscadden [Tue, 17 Aug 2010 18:35:29 +0000 (11:35 -0700)]
trace: Trace the correct version of the resource when setting the index buffer.

The trace driver was tracing the unwrapped version of the index buffer
when setting the index buffer.  This caused an assert validating that
a resource belonged to the trace driver to fail.  Instead, we'll log
the unmodified index buffer structure when setting the index buffer.

Signed-off-by: José Fonseca <jfonseca@vmware.com>
14 years agogenerate_builtins.py: Remove unused import sys.
Vinson Lee [Sat, 21 Aug 2010 08:13:16 +0000 (01:13 -0700)]
generate_builtins.py: Remove unused import sys.

14 years agomapi: Use MAPI_EXPORT to export public functions.
Chia-I Wu [Sat, 21 Aug 2010 05:03:50 +0000 (13:03 +0800)]
mapi: Use MAPI_EXPORT to export public functions.

mapi.h is included by vgapi and st/vega.  On win32, the macro expands to
dllexport and dllimport respectively.

14 years agomapi: Prefix functions in u_current.h by u_current.
Chia-I Wu [Sat, 21 Aug 2010 04:10:02 +0000 (12:10 +0800)]
mapi: Prefix functions in u_current.h by u_current.

That is, replace the old _glapi_* names by new names that start with
u_current_.  When MAPI_GLAPI_CURRENT is defined, u_current.h defines
rename macros to restore the old names.  That is done for ABI
compatibility.

14 years agoglapi: Move public function/variable declarations to glapi.h.
Chia-I Wu [Sat, 21 Aug 2010 03:50:22 +0000 (11:50 +0800)]
glapi: Move public function/variable declarations to glapi.h.

glapi defines an interface that is used by DRI drivers.  It must not be
changed in an ABI incompatible way.  This commit moves all
functions/variables belong to the interface to glapi.h.  Instead of
including u_current.h from glapi.h, u_current.h now includes glapi.h.

14 years agoevergreen : initial support driver code.
richard [Fri, 20 Aug 2010 23:09:25 +0000 (19:09 -0400)]
evergreen : initial support driver code.

14 years agoi965: Rename nr_depth_regs to nr_payload_regs.
Eric Anholt [Tue, 17 Aug 2010 01:35:04 +0000 (18:35 -0700)]
i965: Rename nr_depth_regs to nr_payload_regs.

Only 8 out of the up to 13 regs are for source/dest depth, so the name
wasn't particularly appropriate.  Note that this doesn't count the
constant or URB payload regs.  Also, don't pre-divide by 2, so it's
actually a number of registers.

14 years agoi965: Also use the SIMD8 FB writes for SIMD8 mode on non-SNB.
Eric Anholt [Fri, 20 Aug 2010 22:32:17 +0000 (15:32 -0700)]
i965: Also use the SIMD8 FB writes for SIMD8 mode on non-SNB.

14 years agoi965: Add support for FB writes on Sandybridge.
Zhenyu Wang [Fri, 20 Aug 2010 22:02:19 +0000 (15:02 -0700)]
i965: Add support for FB writes on Sandybridge.

14 years agoi965: Set the destination horiz stride even for da16, as SNB seems to need it.
Zhenyu Wang [Fri, 20 Aug 2010 22:01:11 +0000 (15:01 -0700)]
i965: Set the destination horiz stride even for da16, as SNB seems to need it.

14 years agoi965: Set the maximum number of threads on Sandybridge.
Zhenyu Wang [Fri, 20 Aug 2010 21:38:56 +0000 (14:38 -0700)]
i965: Set the maximum number of threads on Sandybridge.

14 years agoi965: Add AccWrCtl support on Sandybridge.
Zhenyu Wang [Fri, 20 Aug 2010 21:37:19 +0000 (14:37 -0700)]
i965: Add AccWrCtl support on Sandybridge.

Whenever the accumulator results are needed, this bit must be set.

14 years agoi965: Mention the mlen and rlen for URB reads.
Zhenyu Wang [Fri, 20 Aug 2010 21:36:46 +0000 (14:36 -0700)]
i965: Mention the mlen and rlen for URB reads.

14 years agoi965: Sandybridge doesn't have Compr4 mode, since it's not needed any more.
Zhenyu Wang [Fri, 20 Aug 2010 21:32:44 +0000 (14:32 -0700)]
i965: Sandybridge doesn't have Compr4 mode, since it's not needed any more.

14 years agoi965: Adjust disasm of subreg numbers to be in units of the register type.
Zhenyu Wang [Fri, 20 Aug 2010 21:28:39 +0000 (14:28 -0700)]
i965: Adjust disasm of subreg numbers to be in units of the register type.

This makes reading the code easier when matching up to the specs,
which also use this format.

14 years agoi965: Fix DP write channel ordering on Sandybridge.
Eric Anholt [Thu, 8 Jul 2010 20:16:40 +0000 (13:16 -0700)]
i965: Fix DP write channel ordering on Sandybridge.

The SIMD16 message no longer has the goofy interleaved format that
made Compr4 compression necessary before.

14 years agoos_stream: fix bugs in allocation path
Luca Barbieri [Fri, 20 Aug 2010 22:39:48 +0000 (00:39 +0200)]
os_stream: fix bugs in allocation path

14 years agop_compiler: add replacement va_copy
Luca Barbieri [Fri, 20 Aug 2010 22:39:31 +0000 (00:39 +0200)]
p_compiler: add replacement va_copy

This might technically not always be correct, because va_copy might
be a function, or a system might not have va_copy, and not work with
assignment.

Hopefully this is never the case.
Without configure tests, it doesn't seem possible to do better.

14 years agoDelete more vestiges of the old shader compiler.
Kenneth Graunke [Fri, 20 Aug 2010 20:06:02 +0000 (13:06 -0700)]
Delete more vestiges of the old shader compiler.

14 years agoglsl: Remove bogus "ambient" field from vec4 gl_TextureEnvColor.
Kenneth Graunke [Fri, 20 Aug 2010 19:57:21 +0000 (12:57 -0700)]
glsl: Remove bogus "ambient" field from vec4 gl_TextureEnvColor.

14 years agoglsl: add missing ambient field to gl_LightModel
Luca Barbieri [Wed, 18 Aug 2010 10:52:57 +0000 (12:52 +0200)]
glsl: add missing ambient field to gl_LightModel

Again, this is a one-element struct that was incorrectly missing the
field.

14 years agoglsl: don't crash if a field is specified for a non-struct uniform
Luca Barbieri [Wed, 18 Aug 2010 10:52:56 +0000 (12:52 +0200)]
glsl: don't crash if a field is specified for a non-struct uniform

This was triggered by the previous bug, but is a separate problem
in the general sense.

14 years agoglsl: add missing sceneColor field to gl_{Front, Back}LightModelProduct
Luca Barbieri [Wed, 18 Aug 2010 10:52:55 +0000 (12:52 +0200)]
glsl: add missing sceneColor field to gl_{Front, Back}LightModelProduct

According to both GLSL 1.20 and 4.0, these are a struct with one field
called "sceneColor".

Fixes a crash on loading in FlightGear.

14 years agointel: Don't try to do work for BufferSubData with a size of 0.
Eric Anholt [Fri, 20 Aug 2010 19:36:34 +0000 (12:36 -0700)]
intel: Don't try to do work for BufferSubData with a size of 0.

If we hit the linear blit path, we'd come up with a pitch of 0, then
divide by zero.

Fixes vbo-subdata-zero, made for bug #28931 (warsow).

14 years agomesa: Fix GetUniformLocation while compiling display lists.
Nick Bowler [Thu, 19 Aug 2010 16:19:55 +0000 (12:19 -0400)]
mesa: Fix GetUniformLocation while compiling display lists.

This function was apparently missing from the display list dispatch
table, causing the generic no-op function to be called instead.  To make
matters worse, the no-op function is indistinguishable from a successful
call to GetUniformLocation.  GL specifies that GetUniformLocation is
executed immediately when compiling display lists.

Fixes fdo bug 29622.

Signed-off-by: Nick Bowler <nbowler@draconx.ca>
14 years agoRemove remnants of the old glsl compiler.
Eric Anholt [Fri, 20 Aug 2010 17:52:14 +0000 (10:52 -0700)]
Remove remnants of the old glsl compiler.

14 years agor600g: add POW instruction
Jerome Glisse [Fri, 20 Aug 2010 16:53:29 +0000 (18:53 +0200)]
r600g: add POW instruction

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
14 years agor600g: cleanup definition, fix segfault when no valid pixel shader
Jerome Glisse [Fri, 20 Aug 2010 15:57:38 +0000 (17:57 +0200)]
r600g: cleanup definition, fix segfault when no valid pixel shader

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
14 years agoutil: Fix build for C++ compilers.
Michal Krol [Fri, 20 Aug 2010 16:51:22 +0000 (18:51 +0200)]
util: Fix build for C++ compilers.

14 years agogallium: hook up reference count debugging code
Luca Barbieri [Tue, 17 Aug 2010 22:41:10 +0000 (00:41 +0200)]
gallium: hook up reference count debugging code

This commit adds the ability to produce a log file containing all
reference count changes, and object creation/destruction, on Gallium
objects.

The data allows to answer these crucial questions:
1. This app is exhausting all my memory due to a resource leak: where
   is the bug?
2. Which resources is this app using at a given moment? Which parts of
   the code created them?
3. What kinds of resources does this app use?
4. How fast does this app create and destroy resources? Which parts of
   the code create resources fast?

The output is compatible with the one produced by the similar facility
in Mozilla Firefox, allowing to use Mozilla's tools to analyze the data.

To get the log file:
export GALLIUM_REFCNT_LOG=<file>

To get function names and source lines in the log file:
tools/addr2line.sh <file>

To process the log file, see:
http://www.mozilla.org/performance/refcnt-balancer.html

14 years agoauxiliary: add reference count debugging code
Luca Barbieri [Tue, 17 Aug 2010 22:40:33 +0000 (00:40 +0200)]
auxiliary: add reference count debugging code

14 years agoos_stream: add printf facility
Luca Barbieri [Fri, 20 Aug 2010 09:31:24 +0000 (11:31 +0200)]
os_stream: add printf facility

14 years agou_debug_symbol: add symbol name caching
Luca Barbieri [Wed, 18 Aug 2010 14:38:17 +0000 (16:38 +0200)]
u_debug_symbol: add symbol name caching

Without this, any form of logging that dumps stack traces continuously
will spend a lot of time resolving symbol names.

14 years agou_debug_symbol: add support for getting symbol names from glibc
Luca Barbieri [Tue, 17 Aug 2010 22:39:49 +0000 (00:39 +0200)]
u_debug_symbol: add support for getting symbol names from glibc

14 years agou_debug_symbol: support getting a string without output
Luca Barbieri [Tue, 17 Aug 2010 22:38:19 +0000 (00:38 +0200)]
u_debug_symbol: support getting a string without output

14 years agoauxiliary: add functions to describe gallium objects
Luca Barbieri [Tue, 17 Aug 2010 22:36:14 +0000 (00:36 +0200)]
auxiliary: add functions to describe gallium objects

14 years agor600g: add occlusion query support
Dave Airlie [Tue, 17 Aug 2010 06:07:23 +0000 (16:07 +1000)]
r600g: add occlusion query support

Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
14 years agou_staging: improve interface
Luca Barbieri [Fri, 20 Aug 2010 14:35:34 +0000 (16:35 +0200)]
u_staging: improve interface

14 years agogallium/docs: improve documentation for resources
Luca Barbieri [Fri, 20 Aug 2010 11:11:58 +0000 (13:11 +0200)]
gallium/docs: improve documentation for resources

14 years agomesa: Add missing ES sources to SConscript.
Chia-I Wu [Thu, 19 Aug 2010 12:42:10 +0000 (20:42 +0800)]
mesa: Add missing ES sources to SConscript.

These sources compile to nothing when FEATURE_ES is not defined and thus
were overlooked.  Note that api_exec_es[12].c are still missing on the
list.  They should be added when they can be generated on the fly.

14 years agomesa: Decorate functions with GL_APIENTRY in es_generator.py.
Chia-I Wu [Thu, 19 Aug 2010 13:20:01 +0000 (21:20 +0800)]
mesa: Decorate functions with GL_APIENTRY in es_generator.py.

Note that GLES headers use GL_APIENTRY, not GLAPIENTRY.

14 years agomesa: Include compiler.h for ASSERT.
Chia-I Wu [Fri, 20 Aug 2010 02:27:26 +0000 (10:27 +0800)]
mesa: Include compiler.h for ASSERT.

mfeatures.h defines ASSERT_NO_FEATURE to ASSERT, which is defined in
compiler.h.  Header files using the macro should include compiler.h.

14 years agotargets/egl-gdi: Implement guess_gl_api.
Chia-I Wu [Fri, 20 Aug 2010 04:58:52 +0000 (12:58 +0800)]
targets/egl-gdi: Implement guess_gl_api.

It is needed to support calling eglGetProcAddress before eglInitialize.

14 years agoegl: Fix context API check and be verbose.
Chia-I Wu [Fri, 20 Aug 2010 05:19:10 +0000 (13:19 +0800)]
egl: Fix context API check and be verbose.

The API of the context was not checked against EGL_RENDERABLE_TYPE when there
was no attribute list.  Move the check to _eglInitContext, and be verbose about
common mistakes (EGL_RENDERABLE_TYPE not set, EGL_CONTEXT_CLIENT_VERSION not
set, or eglBindAPI not called).

14 years agoegl: Add egl.def for win32 build.
Chia-I Wu [Fri, 20 Aug 2010 04:41:46 +0000 (12:41 +0800)]
egl: Add egl.def for win32 build.

Without the .def file, function names are decorated and cannot be
queried by GetProcAddress easily.

14 years agoegl: Allow core functions to be queried.
Chia-I Wu [Fri, 20 Aug 2010 04:35:02 +0000 (12:35 +0800)]
egl: Allow core functions to be queried.

When _EGL_GET_CORE_ADDRESSES is defined, eglGetProcAddress can be used to query
core functions.  This is non-standard, but some apps expect it.

14 years agost/mesa: fix code/declaration mixing
Keith Whitwell [Fri, 20 Aug 2010 10:52:01 +0000 (11:52 +0100)]
st/mesa: fix code/declaration mixing

14 years agogalahad: remove incorrect comment just added
Luca Barbieri [Fri, 20 Aug 2010 10:06:44 +0000 (12:06 +0200)]
galahad: remove incorrect comment just added

14 years agonv50: use NV50TIC_0_2_TARGET_RECT
Luca Barbieri [Wed, 18 Aug 2010 15:29:59 +0000 (17:29 +0200)]
nv50: use NV50TIC_0_2_TARGET_RECT

14 years agogalahad: check resource_create template
Luca Barbieri [Wed, 18 Aug 2010 23:38:33 +0000 (01:38 +0200)]
galahad: check resource_create template

14 years agomesa/st: use PIPE_TEXTURE_RECT for GL_TEXTURE_RECTANGLE
Luca Barbieri [Thu, 15 Apr 2010 07:04:20 +0000 (09:04 +0200)]
mesa/st: use PIPE_TEXTURE_RECT for GL_TEXTURE_RECTANGLE

14 years agowinsys/sw: use PIPE_TEXTURE_RECT if appropriate
Luca Barbieri [Wed, 18 Aug 2010 23:44:02 +0000 (01:44 +0200)]
winsys/sw: use PIPE_TEXTURE_RECT if appropriate

14 years agost/glx: use PIPE_TEXTURE_RECT if appropriate
Luca Barbieri [Wed, 18 Aug 2010 22:55:13 +0000 (00:55 +0200)]
st/glx: use PIPE_TEXTURE_RECT if appropriate

14 years agost/dri: use PIPE_TEXTURE_RECT if appropriate
Luca Barbieri [Wed, 18 Aug 2010 20:57:22 +0000 (22:57 +0200)]
st/dri: use PIPE_TEXTURE_RECT if appropriate

14 years agoauxiliary: support using PIPE_TEXTURE_RECT internally
Luca Barbieri [Fri, 6 Aug 2010 05:39:21 +0000 (07:39 +0200)]
auxiliary: support using PIPE_TEXTURE_RECT internally

Currently Gallium internals always use PIPE_TEXTURE_2D and normalized
coordinates to access textures.

However, PIPE_TEXTURE_2D is not always supported for NPOT textures,
and PIPE_TEXTURE_RECT requires unnormalized coordinates.

Hence, this change adds support for both kinds of normalization.

14 years agomesa/st: support using PIPE_TEXTURE_RECT internally
Luca Barbieri [Fri, 6 Aug 2010 05:39:21 +0000 (07:39 +0200)]
mesa/st: support using PIPE_TEXTURE_RECT internally

Currently Gallium internals always use PIPE_TEXTURE_2D and normalized
coordinates to access textures.

However, PIPE_TEXTURE_2D is not always supported for NPOT textures,
and PIPE_TEXTURE_RECT requires unnormalized coordinates.

Hence, this change adds support for both kinds of normalization.

14 years agou_staging: use PIPE_TEXTURE_RECT
Luca Barbieri [Wed, 18 Aug 2010 20:57:02 +0000 (22:57 +0200)]
u_staging: use PIPE_TEXTURE_RECT