OSDN Git Service
Brian Paul [Wed, 6 May 2009 18:37:10 +0000 (12:37 -0600)]
mesa: code consolidation in glDraw[Range]Elements() validation
Brian Paul [Wed, 6 May 2009 18:27:38 +0000 (12:27 -0600)]
mesa: new comments, minor reformatting
Brian Paul [Wed, 6 May 2009 18:23:28 +0000 (12:23 -0600)]
mesa: use elementBuf local var instead of ctx->Array.ElementArrayBufferObj
Makes no real difference, but more consistant.
Eric Anholt [Wed, 6 May 2009 17:51:58 +0000 (10:51 -0700)]
i965: Remove the forced lack of caching for renderbuffer surface state.
This snuck in with the multi-draw-buffers commit, and is a major penalty
to performance. It doesn't appear to be required, as the only dependency
the surface BO has is on the state key (and if there's some other dependency,
it should just be in the key).
This brings openarena performance up to almost 2% faster than Mesa 7.4.
Eric Anholt [Wed, 6 May 2009 14:58:40 +0000 (07:58 -0700)]
i965: Remove _NEW_PROGRAM from brw_wm_surfaces setup dependencies.
This was a leftover from the brw_wm_constant_buffer change.
Brian Paul [Wed, 6 May 2009 14:45:55 +0000 (08:45 -0600)]
mesa: be smarter about allocating vert/frag program constants
Try to re-use constants/literals more often to make best use of the
constant buffer space.
See bug 21354.
Eric Anholt [Wed, 6 May 2009 00:16:15 +0000 (17:16 -0700)]
i965: Split WM constant buffer update from other WM surfaces.
This can avoid re-uploading constant data when it isn't necessary, and is
a step towards not updating other surfaces just because constants change.
It also brings the upload of the constant buffer next to the creation.
This brings openarena performance up another 4%, to 91% of the Mesa 7.4 branch.
Eric Anholt [Tue, 5 May 2009 21:05:54 +0000 (14:05 -0700)]
i965: Disentangle VS constant surface state from WM surface state.
Also, only create VS surface state if there's a VS constant buffer to be
uploaded, and set the contents of the buffer at the same time as creation.
Eric Anholt [Tue, 5 May 2009 20:07:33 +0000 (13:07 -0700)]
i965: Don't create constant buffers if they won't be used.
Really, the creation and upload of constants should be in the same place,
since they should only happen together, and a state flag should be
triggered by them so that we don't thrash state around so much for just
updating constants. But this still recovers openarena performance by
another 19%, leaving us 16% behind Mesa 7.4 branch.
Brian Paul [Tue, 5 May 2009 22:29:04 +0000 (16:29 -0600)]
mesa: only use fallback texture when using shaders, not fixed-function (take two)
The semantics are a little different for shaders vs. fixed-function when
trying to use an incomplete texture. The fallback texture returning
(0,0,0,1) should only be used with shaders. For fixed function, the texture
unit is truly disabled/ignored.
Fixes glean fbo test regression.
Brian Paul [Tue, 5 May 2009 22:30:30 +0000 (16:30 -0600)]
Revert "mesa: only use fallback texture when using shaders, not fixed-function"
This reverts commit
a0edbfb28fb2e670c657d52190a7e8b1ccf4f46e.
This patch didn't completely fix the problem. The next patch will.
Brian Paul [Tue, 5 May 2009 22:29:04 +0000 (16:29 -0600)]
mesa: only use fallback texture when using shaders, not fixed-function
The semantics are a little different for shaders vs. fixed-function when
trying to use an incomplete texture. The fallback texture returning
(0,0,0,1) should only be used with shaders.
Fixes glean fbo test regression.
Brian Paul [Tue, 5 May 2009 22:21:20 +0000 (16:21 -0600)]
mesa: only use fallback texture when using shaders, not fixed-function
The semantics are a little different for shaders vs. fixed-function when
trying to use an incomplete texture. The fallback texture returning
(0,0,0,1) should only be used with shaders.
Fixes glean fbo test regression.
Brian Paul [Tue, 5 May 2009 22:17:55 +0000 (16:17 -0600)]
mesa: minor simplification in enable_texture(), updated comments
Brian Paul [Tue, 5 May 2009 16:13:13 +0000 (10:13 -0600)]
Merge branch 'mesa_7_5_branch'
Brian Paul [Tue, 5 May 2009 15:39:39 +0000 (09:39 -0600)]
gallium: remove redundant cube face #defines
Brian Paul [Tue, 5 May 2009 16:11:51 +0000 (10:11 -0600)]
gallium: fix texcoords for cubemap mipmap generation
We need to use (s,t,r) cubemap coords when sampling from the cubemap texture
so the rendered quad gets texture samples from the correct cube face.
Keith Whitwell [Tue, 5 May 2009 12:00:44 +0000 (13:00 +0100)]
progs/trivial: add test for vertex program invarient transform
Keith Whitwell [Tue, 5 May 2009 11:12:28 +0000 (12:12 +0100)]
mesa: more complete fix for transform_invarient glitches
Add a new flag mvp_with_dp4 in the context, and use that to switch
both ffvertex.c and programopt.c vertex transformation code to
either DP4 or MUL/MAD implementations.
Keith Whitwell [Fri, 1 May 2009 17:20:42 +0000 (18:20 +0100)]
mesa/main: set PREFER_DP4 to match position_invarient code
This is a quick fix for z fighting in quake4 caused by the mismatch
between vertex transformation here and in the position_invarient code.
Full fix would be to make this driver-tunable and adjust both
position_invarient and ffvertex_prog.c code to respect driver
preferences.
José Fonseca [Fri, 1 May 2009 15:30:08 +0000 (16:30 +0100)]
wgl: Implemente SwapLayerBuffers.
José Fonseca [Fri, 1 May 2009 15:12:17 +0000 (16:12 +0100)]
scons: Don't use deprecated Options.
José Fonseca [Thu, 30 Apr 2009 12:10:58 +0000 (13:10 +0100)]
util: Limit the stack walk to avoid referencing undefined memory.
Keith Whitwell [Thu, 30 Apr 2009 12:09:34 +0000 (13:09 +0100)]
gallium/tgsi: hack around linker/archiver breakage
Add a dummy function which exists only so that tgsi_text_translate()
doesn't get magic-ed out of the libtgsi.a archive by the build system.
Don't remove unless you know this has been fixed - check on
mingw/scons builds as well.
Keith Whitwell [Thu, 30 Apr 2009 11:35:59 +0000 (12:35 +0100)]
progs/trivial: add vbo-noninterleaved test
José Fonseca [Thu, 30 Apr 2009 11:24:08 +0000 (12:24 +0100)]
wgl: Include alpha bits in pixel format's cColorBits field.
Keith Whitwell [Thu, 30 Apr 2009 09:59:19 +0000 (10:59 +0100)]
gallium/draw: cope with unused vertex_elements
Keith Whitwell [Wed, 29 Apr 2009 13:21:41 +0000 (14:21 +0100)]
trivial: add line-flat.c
José Fonseca [Tue, 28 Apr 2009 18:46:56 +0000 (19:46 +0100)]
wgl: UINT_PTR null value is an integral type, so return 0 instead of NULL.
José Fonseca [Tue, 28 Apr 2009 17:53:52 +0000 (18:53 +0100)]
pb: Save the stack backtrace when creating/mapping a debug buffer.
José Fonseca [Tue, 28 Apr 2009 16:07:01 +0000 (17:07 +0100)]
pb: Dump the fenced buffer sizes.
Stephane Marchesin [Tue, 5 May 2009 10:03:38 +0000 (12:03 +0200)]
nv04: small fix again
Stephane Marchesin [Tue, 5 May 2009 09:29:58 +0000 (11:29 +0200)]
Fix an nv04 bug (thanks Mhopf).
Brian Paul [Mon, 4 May 2009 17:14:35 +0000 (11:14 -0600)]
mesa: also print program params/constants when dumping shaders to disk
Brian Paul [Mon, 4 May 2009 17:13:35 +0000 (11:13 -0600)]
mesa: remove some unfinished/devel code
Brian Paul [Mon, 4 May 2009 13:52:55 +0000 (07:52 -0600)]
glx: replace Xmalloc() calls with Xcalloc()
Fixes a bug where psp->WaitX was uninitialized. Reported by Chris Clayton.
Brian Paul [Fri, 1 May 2009 23:52:23 +0000 (17:52 -0600)]
draw: added some inf/nan debug code (disabled)
Brian Paul [Thu, 23 Apr 2009 16:24:19 +0000 (10:24 -0600)]
gallium/glx: say we're direct rendering
Brian Paul [Sat, 2 May 2009 00:31:04 +0000 (18:31 -0600)]
mesa: in glReadBufer() set _NEW_BUFFERS, not _NEW_PIXEL
Since GL_READ_BUFFER is historically part of the gl_pixel_attrib group
it made sense to signal changes with _NEW_PIXEL. But now with FBOs it's
also part of the framebuffer state.
Now _NEW_PIXEL strictly indicates pixels transfer state changes.
This change avoids framebuffer state validation when any random bit of
pixel-transfer state is set.
DRI drivers updated too: don't check _NEW_COLOR when updating framebuffer
state. I think that was just copied from the Xlib driver because we care
about dither enable/disable state there.
Brian Paul [Sat, 2 May 2009 00:21:02 +0000 (18:21 -0600)]
Merge branch 'mesa_7_5_branch'
Brian Paul [Sat, 2 May 2009 00:17:34 +0000 (18:17 -0600)]
st: create renderbuffer's pipe_surface in st_render_texture()
Previously we created the pipe_surface during framebuffer validation.
But if we did a glCopyTex[Sub]Image() before anything else we wouldn't yet
have the surface. This fixes that.
Brian Paul [Sat, 2 May 2009 00:04:07 +0000 (18:04 -0600)]
mesa: fix state validation bug for glCopyTex[Sub]Image()
We need to make sure the framebuffer state is up to date to make sure we
read pixels from the right buffer when doing a texture image copy.
Brian Paul [Fri, 1 May 2009 22:44:04 +0000 (16:44 -0600)]
st: if st_swapbuffers() is called for single-buffered visual don't crash
Furthermore, return pointer(s) to the front color buffer(s).
Brian Paul [Fri, 1 May 2009 18:28:57 +0000 (12:28 -0600)]
xlib winsys: use new st_swapbuffers() function
The front/back buffer pointers are truly swapped (when there is an actual
front buffer).
This fixes some issues seen with apps/tests that draw to both the front
and back color buffers. The true swap allows us to avoid the (potentially)
slow surface_copy() call in update_framebuffer_state() and is cleaner
overall.
Brian Paul [Fri, 1 May 2009 18:48:46 +0000 (12:48 -0600)]
st: when creating an on-demand front color buffer, init to back buffer image
When we create a new front color buffer (user called glDrawBuffer(GL_FRONT))
initialize it to the contents of the back buffer. Any previous call to
SwapBuffers() would have done that in effect, so make it reality.
Brian Paul [Fri, 1 May 2009 18:25:42 +0000 (12:25 -0600)]
st: added st_renderbuffer::defined flag
Indicates whether there's defined image contents, or garbage/don't care.
This is set when we draw into a renderbuffer and cleared when we resize/
reallocate a renderbuffer or do a buffer swap (back buffer becomes undefined).
We use this to determine whether the front color buffer has been drawn to,
and whether to display its contents upon glFlush/Finish(), when the new
st_swapbuffers() function is used.
Brian Paul [Fri, 1 May 2009 18:12:43 +0000 (12:12 -0600)]
mesa: fix comment
Brian Paul [Fri, 1 May 2009 17:57:00 +0000 (11:57 -0600)]
Merge branch 'mesa_7_5_branch'
Tom Fogal [Wed, 29 Apr 2009 04:03:12 +0000 (21:03 -0700)]
Fix symbol list for mangled Mesa on Darwin.
When building mangled Mesa on Darwin, the exported symbols are
named `_mgluWhatever' instead of simply `_gluWhatever'. When
using a list of exported symbols via the system ld's
`-exported_symbols_list' command line option (as done by mklib),
this resulted in error messages about exporting symbols which do
not exist.
Fortunately the file format accepts simple wildcards. This throws
a wildcard so that the symbol list will match both the mangled and
non-mangled names, preventing the warning and actually exporting
the correct symbols in one shot.
Brian Paul [Fri, 1 May 2009 16:50:04 +0000 (10:50 -0600)]
set: new st_swapbuffers() which does a true front/back buffer swap
The pointers to the front/back renderbuffers are exchanged.
This new function isn't actually used yet...
Zack Rusin [Fri, 1 May 2009 16:50:38 +0000 (12:50 -0400)]
actually add debug to the debug config
José Fonseca [Tue, 24 Mar 2009 19:18:06 +0000 (19:18 +0000)]
egl_xlib: Pass RTLD_LAZY to dlopen.
dlopen manpage mandates that either RTLD_LAZY or RTLD_NOW flags must be
passed. Not doing so was causing a NULL return on debian unstable x86-64.
Ian Romanick [Tue, 28 Apr 2009 17:08:57 +0000 (10:08 -0700)]
Test either GL_FRONT_LEFT or GL_FRONT for front-buffer rendering
For non-stereo visuals, which is all we support, we treat
GL_FRONT_LEFT as GL_FRONT. However, they are technically different,
and they have different enum values. Test for either one to determine
if we're in front-buffer rendering mode.
This fix was suggested by Pierre Willenbrock.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Brian Paul [Fri, 1 May 2009 16:15:21 +0000 (10:15 -0600)]
st: when double buffered, only create front color buffer on demand
Before we always created the front color buffer, even if was never used.
This can save some memory.
Brian Paul [Fri, 1 May 2009 15:37:14 +0000 (09:37 -0600)]
Merge branch 'const-buffer-changes'
Conflicts:
src/mesa/drivers/dri/i965/brw_curbe.c
src/mesa/drivers/dri/i965/brw_vs_emit.c
src/mesa/drivers/dri/i965/brw_wm_glsl.c
Brian Paul [Fri, 1 May 2009 15:30:32 +0000 (09:30 -0600)]
mesa: create/use a fallback texture when bound texture is incomplete
When a GLSL sampler reads from an incomplete texture it should
return (0,0,0,1). Instead of jumping through hoops in all the drivers
to make this happen, just create/install a fallback texture with those
texel values.
Fixes piglit/fp-incomplete-tex on i965 and more importantly, fixes some
GPU lockups when trying to sample from missing surfaces. If a binding
table entry is NULL, it seems that sampling sometimes works, but not
always (lockup).
Todo: create a fallback texture for each type of texture target?
Brian Paul [Fri, 1 May 2009 15:24:19 +0000 (09:24 -0600)]
docs: initial 7.6 release notes page
Brian Paul [Fri, 1 May 2009 15:22:20 +0000 (09:22 -0600)]
mesa: bump version to 7.6 (devel)
Brian Paul [Fri, 1 May 2009 15:19:59 +0000 (09:19 -0600)]
mesa: remove -devel suffix from version
Brian Paul [Fri, 1 May 2009 15:15:14 +0000 (09:15 -0600)]
configs: bump MESA_MINOR to 5
Brian Paul [Thu, 30 Apr 2009 23:27:51 +0000 (17:27 -0600)]
docs: fix typo
Corbin Simpson [Fri, 1 May 2009 14:29:14 +0000 (07:29 -0700)]
radeon: Don't even bother with things too big to fit into our card.
Corbin Simpson [Fri, 1 May 2009 13:01:52 +0000 (06:01 -0700)]
r300-gallium, radeon-winsys: Hide radeon_cs from r300 pipe.
Corbin Simpson [Fri, 1 May 2009 12:54:53 +0000 (05:54 -0700)]
r300-gallium, radeon-winsys: Space accounting.
It is no longer optional in current libdrm, so it was time to actually
start counting our BOs.
Corbin Simpson [Fri, 1 May 2009 12:03:56 +0000 (05:03 -0700)]
r300-gallium, radeon-winsys: Reorganize r300_winsys header, break ABI.
Make things more consistent, prepare for more function hooks.
Corbin Simpson [Fri, 1 May 2009 11:47:50 +0000 (04:47 -0700)]
r300-gallium: Don't bother with conditional double define.
We'll just forever leave it in r300_winsys.h since it's needed for
whichever winsys is hosting the pipe.
Corbin Simpson [Tue, 28 Apr 2009 11:37:56 +0000 (04:37 -0700)]
radeon: Fix cast and comment.
Brian Paul [Thu, 30 Apr 2009 23:17:31 +0000 (17:17 -0600)]
docs: asst. updates for 7.5 release
Brian Paul [Thu, 30 Apr 2009 23:13:22 +0000 (17:13 -0600)]
docs: bring in 7.4 doc updates from mesa_7_4_branch
Brian Paul [Thu, 30 Apr 2009 23:11:20 +0000 (17:11 -0600)]
docs: notes about new version number meaning in release notes
Brian Paul [Thu, 30 Apr 2009 23:04:01 +0000 (17:04 -0600)]
mesa: update linux-static, linux-x86-static configs
Brian Paul [Thu, 30 Apr 2009 23:03:54 +0000 (17:03 -0600)]
demos: silence warning
Tormod Volden [Thu, 30 Apr 2009 22:55:54 +0000 (16:55 -0600)]
mklib: replace if/expr with case
Saves forking an expr for every object.
Tormod Volden [Thu, 30 Apr 2009 22:52:56 +0000 (16:52 -0600)]
mesa: Prepend "-Wl," to linking options
Let mklib ignore -Wl options inside the object list when building
static libraries
Brian Paul [Thu, 30 Apr 2009 22:52:02 +0000 (16:52 -0600)]
i915simple: remove duplicated i915_context.c in C_SOURCES
Michel Dänzer [Thu, 30 Apr 2009 11:21:08 +0000 (13:21 +0200)]
r300: Increase reference count of texture objects referenced by current state.
Fixes a use-after-free reported in
http://bugs.freedesktop.org/show_bug.cgi?id=20539, so this possibly fixes that
bug. It has been confirmed to fix
http://bugs.freedesktop.org/show_bug.cgi?id=17895 .
Michel Dänzer [Thu, 30 Apr 2009 07:39:29 +0000 (09:39 +0200)]
gallium: Add SCons build support for the DRI state tracker.
Dan Nicholson [Thu, 30 Apr 2009 03:33:50 +0000 (20:33 -0700)]
egl: Don't install demodriver
I don't think anyone besides a developer would ever want to use the demo
egl driver. Furthermore, egl would only ever load demodriver if it was
set via EGL_DRIVER in the environment. In that case, I think you can
point it to your mesa source directory.
Jakob Bornecrantz [Wed, 29 Apr 2009 19:44:03 +0000 (20:44 +0100)]
progs/tests: Update ignores
Jakob Bornecrantz [Wed, 29 Apr 2009 19:17:21 +0000 (20:17 +0100)]
progs/tests: Add mipmap_comp for mipmap testing with compressed textures
Tom Fogal [Wed, 29 Apr 2009 16:32:46 +0000 (10:32 -0600)]
Use variable library name in pkg-config output.
Previously the pkg-config output files would contain e.g. `-lGL'
and `-lGLU', even if the user modified their configuration to
build libraries with different names. This modifies the
pkg-config inputs, and corresponding makery, so that modifying the
output library name will cause the appropriate updated name to
appear in the pkg-config `-l' option.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Dan Nicholson [Wed, 29 Apr 2009 19:11:43 +0000 (12:11 -0700)]
autoconf: Add switch for optional EGL
EGL doesn't build on all platforms, so allow people to opt out.
Dan Nicholson [Wed, 29 Apr 2009 13:49:27 +0000 (06:49 -0700)]
autoconf: Clean up some m4 usage
m4_fatal is equivalent to m4_errprint + m4_exit.
Brian Paul [Wed, 29 Apr 2009 17:52:06 +0000 (11:52 -0600)]
mesa: added _mesa_check_soa_dependencies() function
This function will check an instruction to see if there's data dependencies
between the dst and src registers if executed in an SOA manner.
Brian Paul [Wed, 29 Apr 2009 17:56:57 +0000 (11:56 -0600)]
tgsi: added tgsi_check_soa_dependencies() and related debug code (disabled)
The TGSI interpeter operates in SOA style. We need to check for data
dependencies in instructions which read from and write to the same register.
For now just adding some debug code to detect that condition. Actual fixes
to follow.
Brian Paul [Wed, 29 Apr 2009 17:52:06 +0000 (11:52 -0600)]
mesa: added _mesa_check_soa_dependencies() function
This function will check an instruction to see if there's data dependencies
between the dst and src registers if executed in an SOA manner.
Brian Paul [Tue, 28 Apr 2009 20:29:27 +0000 (14:29 -0600)]
softpipe: return PIPE_UNREFERENCED in softpipe_is_buffer/texture_referenced()
This allows the engine demo to run again (avoid crash in VBO code).
This stuff still needs to be revisited someday though...
Keith Whitwell [Tue, 28 Apr 2009 16:50:19 +0000 (17:50 +0100)]
mesa/st: protect internal flushes with FLUSH_CURRENT
Already doing this for driver.flush()
Keith Whitwell [Tue, 28 Apr 2009 16:49:50 +0000 (17:49 +0100)]
mesa/main: protect driver.finish with FLUSH_CURRENT
Already doing this for driver.flush()
Keith Whitwell [Tue, 28 Apr 2009 13:51:11 +0000 (14:51 +0100)]
mesa/st: remove duplicate offset calculation
Keith Whitwell [Tue, 28 Apr 2009 13:50:05 +0000 (14:50 +0100)]
gallium/draw: add ability to print out active pipeline stages
Keith Whitwell [Mon, 27 Apr 2009 14:33:44 +0000 (15:33 +0100)]
mesa/st: translate VERT_ATTRIB_GENERIC8..15 in st_translate_vertex_program
It seems quake4 can hit these attributes sometimes.
Keith Whitwell [Tue, 28 Apr 2009 10:49:55 +0000 (11:49 +0100)]
progs: add fflushes for cygwin
Keith Whitwell [Mon, 27 Apr 2009 18:36:42 +0000 (19:36 +0100)]
util/indices: remove debug prints
Keith Whitwell [Mon, 27 Apr 2009 17:56:26 +0000 (18:56 +0100)]
mesa/st: workaround for crashes in st_copy_texsubimage
Proper fix for this hasn't been identified, but avoid crashing.
José Fonseca [Mon, 27 Apr 2009 19:24:55 +0000 (20:24 +0100)]
wgl: Store current HDC/HGLRC in stw_context.
Less TLS lookups.
Alex Deucher [Tue, 28 Apr 2009 16:37:29 +0000 (12:37 -0400)]
R300: add quadpipe overrides
RV410 SE chips only have 1 quadpipe.
Also, handle other R300 chip with quadpipe override
Brian Paul [Tue, 28 Apr 2009 15:58:44 +0000 (09:58 -0600)]
i965: avoid segfault in intel_update_renderbuffers() if using DRI1
Brian Paul [Tue, 28 Apr 2009 15:05:19 +0000 (09:05 -0600)]
swrast: add missing break in clamp_rect_coord_linear()
See bug 21461.
Brian Paul [Mon, 27 Apr 2009 23:01:59 +0000 (17:01 -0600)]
demos: asst. updates, clean-ups