Brian [Tue, 3 Jul 2007 23:19:30 +0000 (17:19 -0600)]
hook in state tracking for blend color
Brian [Tue, 3 Jul 2007 23:13:45 +0000 (17:13 -0600)]
added blend_color
Brian [Tue, 3 Jul 2007 23:13:04 +0000 (17:13 -0600)]
Blend MIN/MAX modes are special: the src/dest terms are always one.
Brian [Tue, 3 Jul 2007 23:12:11 +0000 (17:12 -0600)]
Rewrite blending in terms of SIMD operations.
Brian [Tue, 3 Jul 2007 16:58:03 +0000 (10:58 -0600)]
clarify that line_stipple_factor is in [0,255] corresponding to GL's range of [1,256].
Brian [Mon, 2 Jul 2007 21:32:12 +0000 (15:32 -0600)]
rename a few structs (use _state suffix consistantly), reorder/sort fields in some structs
Brian [Fri, 29 Jun 2007 19:10:47 +0000 (13:10 -0600)]
added buffer_data(), buffer_sub_data() to pipe_buffer
Brian [Fri, 29 Jun 2007 19:03:01 +0000 (13:03 -0600)]
more work on tex objects, surfaces, mappable buffers
Brian [Fri, 29 Jun 2007 18:59:58 +0000 (12:59 -0600)]
added texture types and mapping types
Brian [Fri, 29 Jun 2007 18:59:02 +0000 (12:59 -0600)]
texture object construction
Brian [Fri, 29 Jun 2007 18:16:37 +0000 (12:16 -0600)]
s/G_CONTEXT_H/SP_CONTEXT_H/
Brian [Tue, 26 Jun 2007 23:35:24 +0000 (17:35 -0600)]
consolidate point/line state into pipe_setup_state
Brian [Mon, 25 Jun 2007 15:46:04 +0000 (09:46 -0600)]
fix logic error in updating of quad's Z values. Basic depth testing works now
Brian [Mon, 25 Jun 2007 15:45:10 +0000 (09:45 -0600)]
clean-ups
Brian [Mon, 25 Jun 2007 15:44:28 +0000 (09:44 -0600)]
emit quad->outputs.depth[], hack need_z
Brian [Mon, 25 Jun 2007 15:43:02 +0000 (09:43 -0600)]
code for functional Z buffer surface
Brian [Mon, 25 Jun 2007 15:41:49 +0000 (09:41 -0600)]
better triangle Z coords
Keith Whitwell [Sun, 24 Jun 2007 09:32:48 +0000 (10:32 +0100)]
Restructure z test code slightly.
Make the logic slightly closer to an eventual SSE or SPE implementation.
Brian [Fri, 22 Jun 2007 19:37:47 +0000 (13:37 -0600)]
initial texture object, texture format code
Brian [Fri, 22 Jun 2007 19:08:49 +0000 (13:08 -0600)]
more texture sampler work
Brian [Fri, 22 Jun 2007 18:49:23 +0000 (12:49 -0600)]
added blend comments
Brian [Fri, 22 Jun 2007 18:47:04 +0000 (12:47 -0600)]
Assorted token renaming/removal, minor state changes, etc.
Brian [Fri, 22 Jun 2007 02:08:10 +0000 (20:08 -0600)]
alpha test and misc changes
Brian [Wed, 20 Jun 2007 23:20:02 +0000 (17:20 -0600)]
checkpoint: implement z/depth testing
Brian [Wed, 20 Jun 2007 23:19:13 +0000 (17:19 -0600)]
triangles w/ depth test
Keith Whitwell [Wed, 20 Jun 2007 22:05:23 +0000 (23:05 +0100)]
Only build i915tex
Keith Whitwell [Wed, 20 Jun 2007 22:04:56 +0000 (23:04 +0100)]
Add state tracker create/destroy calls to i915 driver.
Keith Whitwell [Wed, 20 Jun 2007 22:04:00 +0000 (23:04 +0100)]
Rename static variable in temporary tnl code
Keith Whitwell [Wed, 20 Jun 2007 22:03:26 +0000 (23:03 +0100)]
Remove dependency on vf.h in public headers
Brian [Wed, 20 Jun 2007 22:15:09 +0000 (16:15 -0600)]
update includes (s/tile/quad/)
Brian [Wed, 20 Jun 2007 22:14:55 +0000 (16:14 -0600)]
s/tile/quad/
Brian [Wed, 20 Jun 2007 22:12:26 +0000 (16:12 -0600)]
Rename "tile" files to "quad".
Brian [Wed, 20 Jun 2007 21:22:39 +0000 (15:22 -0600)]
only blend active pixels in the quad
Brian [Wed, 20 Jun 2007 21:14:47 +0000 (15:14 -0600)]
quad blending works now, but many blend terms need to be added in blend_quad().
Brian [Wed, 20 Jun 2007 20:29:14 +0000 (14:29 -0600)]
Initial implementation of a software pipeline for quad rasterization (fragment ops).
This is very much like the clipper/setup pipeline for primitives.
Brian [Wed, 20 Jun 2007 20:27:42 +0000 (14:27 -0600)]
better comment
Brian [Wed, 20 Jun 2007 19:10:48 +0000 (13:10 -0600)]
actually use new glClear code
Brian [Wed, 20 Jun 2007 18:50:16 +0000 (12:50 -0600)]
fix comment
Brian [Wed, 20 Jun 2007 18:23:47 +0000 (12:23 -0600)]
plug in write_mono_row_ub
Brian [Wed, 20 Jun 2007 15:41:41 +0000 (09:41 -0600)]
implement softpipe clearing (untested)
Brian [Wed, 20 Jun 2007 14:47:09 +0000 (08:47 -0600)]
Initial work for glClear(), clear color state.
Brian [Wed, 20 Jun 2007 01:52:25 +0000 (19:52 -0600)]
hook up point state
Brian [Wed, 20 Jun 2007 01:19:31 +0000 (19:19 -0600)]
Re-org of surface/framebuffer state.
We should be able to render to any depth/format of X window now.
Brian [Wed, 20 Jun 2007 01:17:48 +0000 (19:17 -0600)]
don't request GLUT_ALPHA
Brian [Tue, 19 Jun 2007 14:45:34 +0000 (08:45 -0600)]
just use regular malloc(), free(), memcpy()
Brian [Tue, 19 Jun 2007 14:43:05 +0000 (08:43 -0600)]
add point/line/polygon state
Brian [Mon, 18 Jun 2007 23:54:38 +0000 (17:54 -0600)]
Added alpha test state.
Brian [Mon, 18 Jun 2007 23:53:09 +0000 (17:53 -0600)]
Define just one set of PIPE_FUNC_LESS, GREATER, EQUAL, etc tokens.
Used for depth test, alpha test, stencil test, shadow test, etc.
Brian [Mon, 18 Jun 2007 12:57:36 +0000 (06:57 -0600)]
change clear color
Brian [Fri, 15 Jun 2007 16:57:44 +0000 (10:57 -0600)]
s/PRIM_H/SP_STATE_H/
Brian [Fri, 15 Jun 2007 16:55:12 +0000 (10:55 -0600)]
Added st_update_framebuffer struct/object.
Brian [Fri, 15 Jun 2007 16:54:08 +0000 (10:54 -0600)]
framebuffer state
Brian [Fri, 15 Jun 2007 16:50:48 +0000 (10:50 -0600)]
In softpipe_set_viewport() use ordinary (struct) assignment to update softpipe->viewport.
The previous memcpy() was incorrect since it only copied 4 bytes instead of 32.
With struct assignment we avoid data size errors.
Brian [Fri, 15 Jun 2007 15:51:53 +0000 (09:51 -0600)]
initial framebuffer state
Brian [Fri, 15 Jun 2007 15:51:27 +0000 (09:51 -0600)]
framebuffer state
Keith Whitwell [Thu, 14 Jun 2007 17:25:10 +0000 (18:25 +0100)]
Add GLUT_ALPHA for softpipe
Keith Whitwell [Thu, 14 Jun 2007 17:23:43 +0000 (18:23 +0100)]
Rename directories again?!
Some git wierdness going on.
Keith Whitwell [Thu, 14 Jun 2007 17:11:48 +0000 (18:11 +0100)]
Renamed softpipe directories and files to something less confusing.
softpipe/state_tracker --> state_tracker/
softpipe/ --> pipe/
softpipe/generic --> pipe/softpipe/
I don't think pipe is a great name, but I disliked all the others too.
Luckily it's fairly easy to rename with git, so this can be revisited
later.
Keith Whitwell [Thu, 14 Jun 2007 14:18:27 +0000 (15:18 +0100)]
Small optimization for line drawing.
And some comments on point sprites.
Keith Whitwell [Wed, 13 Jun 2007 14:37:52 +0000 (15:37 +0100)]
Remove some dead code.
Brian [Tue, 12 Jun 2007 14:47:49 +0000 (08:47 -0600)]
simple line drawing code
Brian [Wed, 30 May 2007 22:27:17 +0000 (16:27 -0600)]
added st_atom_depth.c
Brian [Wed, 30 May 2007 22:26:55 +0000 (16:26 -0600)]
sketch out texture sampler state
Brian [Wed, 30 May 2007 22:10:31 +0000 (16:10 -0600)]
comments, clean-up, alphabetize members
Brian [Wed, 30 May 2007 22:07:39 +0000 (16:07 -0600)]
added depth(Z) state support
Brian [Wed, 30 May 2007 22:07:27 +0000 (16:07 -0600)]
depth(Z) state
Brian [Mon, 28 May 2007 22:13:17 +0000 (16:13 -0600)]
implement point rendering
Brian [Fri, 25 May 2007 21:53:42 +0000 (15:53 -0600)]
INLINE goes before the function's return type
Brian [Fri, 25 May 2007 21:45:21 +0000 (15:45 -0600)]
Add some comments to explain things, code clarifications.
Also, add quad.facing field for front/back facing. See comments for details.
Brian [Fri, 25 May 2007 04:16:26 +0000 (22:16 -0600)]
use GLubyte for stencil ref/mask/clear
Brian [Thu, 24 May 2007 23:37:36 +0000 (17:37 -0600)]
Stencil state.
Brian [Thu, 24 May 2007 22:50:45 +0000 (16:50 -0600)]
added st_atom_blend.c, st_atom_scissor.c
Brian [Thu, 24 May 2007 22:50:22 +0000 (16:50 -0600)]
Implement blend state
Brian [Thu, 24 May 2007 22:49:27 +0000 (16:49 -0600)]
#ifndef SP_DEFINES_H protection, copyright
Brian [Thu, 24 May 2007 22:48:38 +0000 (16:48 -0600)]
scissor depends on glScissor() and framebuffer size
Brian [Thu, 24 May 2007 20:37:00 +0000 (14:37 -0600)]
add scissor state
Brian [Thu, 24 May 2007 20:36:48 +0000 (14:36 -0600)]
scissor state
Brian [Thu, 24 May 2007 20:36:02 +0000 (14:36 -0600)]
define, use SP_MAX_CLIP_PLANES
Keith Whitwell [Thu, 24 May 2007 10:37:08 +0000 (11:37 +0100)]
Add the vf module.
This is a cleaned up version of the code in tnl/t_vertex*.
Keith Whitwell [Thu, 24 May 2007 09:44:53 +0000 (10:44 +0100)]
Use the x11 driver as a test harness for the softpipe/state_tracker code.
This has some limitations as we currently require a mapped framebuffer,
so it only really works with double-buffered ximage rgba8888 windows.
Keith Whitwell [Thu, 24 May 2007 09:41:34 +0000 (10:41 +0100)]
First version of the softpipe rasterizer.
This will flesh out to hopefully acheive 3 things:
- New software rasterizer for mesa
- New driver model for mesa & dri with much smaller drivers
- The basis for the cell driver.
It's got a long way to go yet, but will at least run trivial/tri.c.
Keith Whitwell [Thu, 24 May 2007 09:38:57 +0000 (10:38 +0100)]
Modifications to run with softpipe test harness
RGB->RGBA
double buffer by default.
Brian [Mon, 14 May 2007 20:30:36 +0000 (14:30 -0600)]
master/trunk is now for Mesa 7.1 devel: bump versions
Bruce Merry [Mon, 14 May 2007 15:37:48 +0000 (09:37 -0600)]
in __glXGetArrayType() return type, not enabled (bug 10938)
Brian [Fri, 11 May 2007 21:50:33 +0000 (15:50 -0600)]
comment about fixing uniform structs
Oliver McFadden [Sun, 13 May 2007 13:38:38 +0000 (13:38 +0000)]
r300: Enable the non-user-buffers code to compile.
Oliver McFadden [Sun, 13 May 2007 12:59:41 +0000 (12:59 +0000)]
r300: Use e32(0x0) rather than e32(0); makes greping for hex values easier.
Oliver McFadden [Sun, 13 May 2007 12:56:59 +0000 (12:56 +0000)]
r300: Use the defined values when writing to R300_RS_ROUTE_0.
Oliver McFadden [Sun, 13 May 2007 12:43:40 +0000 (12:43 +0000)]
r300: Use the defined values when writing to R300_SE_VTE_CNTL.
You can verify this is correct with the following code.
assert (0x043f ==
(R300_VTX_W0_FMT | R300_VPORT_X_SCALE_ENA | R300_VPORT_X_OFFSET_ENA
| R300_VPORT_Y_SCALE_ENA | R300_VPORT_Y_OFFSET_ENA |
R300_VPORT_Z_SCALE_ENA | R300_VPORT_Z_OFFSET_ENA));
Oliver McFadden [Sun, 13 May 2007 12:36:27 +0000 (12:36 +0000)]
r300: Added a TODO comment for the R300_VAP_INPUT_ROUTE_[01]_0 values.
Oliver McFadden [Sun, 13 May 2007 09:09:30 +0000 (09:09 +0000)]
r300: Changed some more functions to static functions.
This required reordering some of the functions which is why the diff is a little
larger.
Oliver McFadden [Sun, 13 May 2007 08:59:54 +0000 (08:59 +0000)]
r300: Changed a couple of functions in r300_state.c to static functions.
Oliver McFadden [Sun, 13 May 2007 08:43:30 +0000 (08:43 +0000)]
r300: Removed the unused r300_aos_rec (replaced by r300_dma_region) structure.
Oliver McFadden [Sun, 13 May 2007 08:28:51 +0000 (08:28 +0000)]
r300: Use #if 0 for disabled code, rather than commenting it out.
Oliver McFadden [Sun, 13 May 2007 08:21:50 +0000 (08:21 +0000)]
r300: Removed a couple of obsolete/commented out includes.
Oliver McFadden [Sun, 13 May 2007 07:51:41 +0000 (07:51 +0000)]
r300: A few very minor indenting corrections.
Oliver McFadden [Sun, 13 May 2007 07:44:20 +0000 (07:44 +0000)]
r300: Use the defined values for the interpolater (interp_magic) magic values.
Oliver McFadden [Sun, 13 May 2007 06:39:24 +0000 (06:39 +0000)]
r300: Use the defined values when writing to R300_VAP_INPUT_CNTL_[01].
You can verify this is correct with the following code.
assert (0x00000001 == R300_INPUT_CNTL_0_COLOR);
assert (0x00000405 ==
(R300_INPUT_CNTL_POS | R300_INPUT_CNTL_COLOR |
R300_INPUT_CNTL_TC0));
Oliver McFadden [Sat, 12 May 2007 17:16:56 +0000 (17:16 +0000)]
Revert "r300: Perform the locking closer to the DRM texture upload call."
This reverts commit
93881edb46fc95e1cfb4ded4e8a5db92612d3e4d.
Roland Scheidegger [Sat, 12 May 2007 09:20:24 +0000 (11:20 +0200)]
bring over recent radeonMakeCurrent fixes for r300 to radeon/r200