mesa.git
16 years agofix Xlib libGL.so build problem.
Brian Paul [Mon, 9 Jun 2008 22:06:44 +0000 (16:06 -0600)]
fix Xlib libGL.so build problem.

Also, build driverfuncs.c into libmesa.a since it's always needed.

16 years agos/SOLO/MESA/
Brian Paul [Mon, 9 Jun 2008 20:09:09 +0000 (14:09 -0600)]
s/SOLO/MESA/

16 years agoMakefile clean-ups
Brian Paul [Mon, 9 Jun 2008 19:50:44 +0000 (13:50 -0600)]
Makefile clean-ups

16 years agomore old stuff, rename SOLO->MESA, etc
Brian Paul [Mon, 9 Jun 2008 19:50:31 +0000 (13:50 -0600)]
more old stuff, rename SOLO->MESA, etc

16 years agoMakefile clean-ups
Brian Paul [Mon, 9 Jun 2008 19:50:11 +0000 (13:50 -0600)]
Makefile clean-ups

16 years agodocument makefile changes
Brian Paul [Fri, 6 Jun 2008 18:38:24 +0000 (12:38 -0600)]
document makefile changes

16 years agoCheck for NULL pointer
Alan Hourihane [Mon, 9 Jun 2008 15:24:15 +0000 (16:24 +0100)]
Check for NULL pointer

16 years agoglx: use goto's vs. nested if's ala xserver
George Sapountzis [Mon, 9 Jun 2008 14:47:49 +0000 (17:47 +0300)]
glx: use goto's vs. nested if's  ala xserver

compile tested only

16 years agoglx: add dri_common.c ala xserver
George Sapountzis [Mon, 9 Jun 2008 14:47:09 +0000 (17:47 +0300)]
glx: add dri_common.c ala xserver

also drop driFilterModes which is unused

in preparation of loading swrast_dri.so

16 years agodri/swrast: minor cosmetic
George Sapountzis [Mon, 9 Jun 2008 14:46:39 +0000 (17:46 +0300)]
dri/swrast: minor cosmetic

16 years agor300: Workaround hardware readcache problem
Nicolai Haehnle [Sun, 8 Jun 2008 20:36:20 +0000 (22:36 +0200)]
r300: Workaround hardware readcache problem

This workaround is similar to the one found in r200_span.c.
It seems like some part of the read hardware doesn't realize that
VRAM has changed. By reading from an arbitrary position, this is fixed.

The piglit test bugs/r300-readcache is a regression test for this bug.

16 years agoremove realglx.[ch] from build
Brian [Sun, 8 Jun 2008 17:25:30 +0000 (11:25 -0600)]
remove realglx.[ch] from build

16 years agoi965: fix OPCODE_TEX when additional ops are needed
Roland Scheidegger [Sun, 8 Jun 2008 12:04:39 +0000 (14:04 +0200)]
i965: fix OPCODE_TEX when additional ops are needed

16 years agor300: Allow driconf to set a default anisotropy.
Corbin Simpson [Sat, 7 Jun 2008 19:37:10 +0000 (12:37 -0700)]
r300: Allow driconf to set a default anisotropy.
If an app does not use GL_EXT_tex_filter_aniso, this lets driconf set it instead.

16 years agor300: Some cleanups in depth and stencil state handling
Nicolai Haehnle [Sat, 7 Jun 2008 19:07:28 +0000 (21:07 +0200)]
r300: Some cleanups in depth and stencil state handling

This also fixes a bug with Zfunc set to GL_NEVER in glean/paths.

16 years agor300: Apply writemask when clearing stencil buffer
Nicolai Haehnle [Sat, 7 Jun 2008 18:33:28 +0000 (20:33 +0200)]
r300: Apply writemask when clearing stencil buffer

This fixes the glean/maskedClear test.

16 years agor300: Fix stencil buffer clear
Nicolai Haehnle [Sat, 7 Jun 2008 18:25:47 +0000 (20:25 +0200)]
r300: Fix stencil buffer clear

The glean/readPixSanity test passes now.

16 years agor300: Further anisotropic filtering fixes
Nicolai Haehnle [Sat, 7 Jun 2008 16:54:35 +0000 (18:54 +0200)]
r300: Further anisotropic filtering fixes

Thanks to Corbin for the initial cut today. Fixed some minor stuff (in
particular, make sure we still use a MIP_LINEAR filtering mode; anisotropy
without MIP_LINEAR filtering is not the truly pleasing anisotropy).

16 years agor300: Update LOD_BIAS register constants
Nicolai Haehnle [Sat, 7 Jun 2008 12:17:36 +0000 (14:17 +0200)]
r300: Update LOD_BIAS register constants

16 years agor5xx: Unbreak anisotropic filtering?
Corbin Simpson [Sat, 7 Jun 2008 08:37:35 +0000 (01:37 -0700)]
r5xx: Unbreak anisotropic filtering?
Not quite finished, maybe? Not sure.

16 years agor5xx: Fix speedy LIT once and for all.
Corbin Simpson [Sat, 7 Jun 2008 04:53:05 +0000 (21:53 -0700)]
r5xx: Fix speedy LIT once and for all.
To do: Add a slightly more accurate LIT. Will do later.

16 years agoFix CRLF line endings.
Jose Fonseca [Sat, 7 Jun 2008 03:34:45 +0000 (12:34 +0900)]
Fix CRLF line endings.

16 years agoFix CRLF line endings.
Jose Fonseca [Sat, 7 Jun 2008 03:33:47 +0000 (12:33 +0900)]
Fix CRLF line endings.

16 years agor300: Improve texture layout calculations
Nicolai Haehnle [Fri, 6 Jun 2008 21:47:46 +0000 (23:47 +0200)]
r300: Improve texture layout calculations

The texture layout calculations for mipmapped cubemaps used to be completely
wrong, since the GPU expects images to be grouped by miplevel instead of by
face number.

This has been fixed now, though the memory layout is still slightly incorrect
for the smaller miplevels. Unfortunately, the docs are lacking in that area.

16 years agor300: Cleanup TX_MIN_FILTER defines
Nicolai Haehnle [Fri, 6 Jun 2008 19:28:08 +0000 (21:28 +0200)]
r300: Cleanup TX_MIN_FILTER defines

This commit should not affect the functionality at all, just cleanup
some of the original texture filter guesswork using information from
the register documentation.

16 years agor300: Remove unused variable r300_tex_obj::dirty_state and related defines
Nicolai Haehnle [Fri, 6 Jun 2008 19:13:12 +0000 (21:13 +0200)]
r300: Remove unused variable r300_tex_obj::dirty_state and related defines

This variable was only ever written to, but nobody used its value.

16 years agos/subdirs/asm_subdirs/
Brian Paul [Fri, 6 Jun 2008 18:27:55 +0000 (12:27 -0600)]
s/subdirs/asm_subdirs/

16 years agomove the install rules
Brian Paul [Fri, 6 Jun 2008 18:27:14 +0000 (12:27 -0600)]
move the install rules

16 years agoremove stray semicolon
Brian Paul [Fri, 6 Jun 2008 18:25:24 +0000 (12:25 -0600)]
remove stray semicolon

16 years agomove a beos-ism down into drivers/beos/Makefile
Brian Paul [Fri, 6 Jun 2008 18:24:44 +0000 (12:24 -0600)]
move a beos-ism down into drivers/beos/Makefile

16 years agoRemove the old, complicated default build rule; use the new driver_subdirs rule
Brian Paul [Fri, 6 Jun 2008 18:18:40 +0000 (12:18 -0600)]
Remove the old, complicated default build rule; use the new driver_subdirs rule

Also, all the old driver-specific rules are now gone.

16 years agoremove more old stuff, insert new driver_subdir build rules (but not enabled yet)
Brian Paul [Fri, 6 Jun 2008 18:12:59 +0000 (12:12 -0600)]
remove more old stuff, insert new driver_subdir build rules (but not enabled yet)

16 years agonew drivers/Makefile with simple default/clean/install rules
Brian Paul [Fri, 6 Jun 2008 18:09:55 +0000 (12:09 -0600)]
new drivers/Makefile with simple default/clean/install rules

16 years agomove directfb-libgl into drivers/directfb/Makefile
Brian Paul [Fri, 6 Jun 2008 18:09:01 +0000 (12:09 -0600)]
move directfb-libgl into drivers/directfb/Makefile

16 years agoMove fbdev driver construction into src/mesa/drivers/fbdev/Makefile
Brian Paul [Thu, 5 Jun 2008 22:44:11 +0000 (16:44 -0600)]
Move fbdev driver construction into src/mesa/drivers/fbdev/Makefile

16 years agoreorder some lines, more make clean cmds
Brian Paul [Thu, 5 Jun 2008 22:28:00 +0000 (16:28 -0600)]
reorder some lines, more make clean cmds

16 years agoremove X11, Glide, OSMesa sources
Brian Paul [Thu, 5 Jun 2008 22:27:43 +0000 (16:27 -0600)]
remove X11, Glide, OSMesa sources

16 years agoadded dependencies
Brian Paul [Thu, 5 Jun 2008 22:26:15 +0000 (16:26 -0600)]
added dependencies

16 years agoMove building of stand-alone Mesa into drivers/x11/Makefile
Brian Paul [Thu, 5 Jun 2008 22:08:42 +0000 (16:08 -0600)]
Move building of stand-alone Mesa into drivers/x11/Makefile

Also, some re-org, renaming, and general clean-up.
We're just a few steps away from removing the all the special case rules
for building drivers.

16 years agonew Makefile for building "stand-alone", Xlib-based libGL
Brian Paul [Thu, 5 Jun 2008 22:06:17 +0000 (16:06 -0600)]
new Makefile for building "stand-alone", Xlib-based libGL

16 years agoprefix some includes with mesa/ or glapi/
Brian Paul [Thu, 5 Jun 2008 22:05:02 +0000 (16:05 -0600)]
prefix some includes with mesa/ or glapi/

16 years agominor changes
Brian Paul [Thu, 5 Jun 2008 21:37:17 +0000 (15:37 -0600)]
minor changes

16 years agoMove construction of libOSMesa.so into src/mesa/drivers/osmesa/Makefile
Brian Paul [Thu, 5 Jun 2008 21:36:07 +0000 (15:36 -0600)]
Move construction of libOSMesa.so into src/mesa/drivers/osmesa/Makefile

This removes some cruft from src/mesa/Makefile.
Something similar could be done for stand-alone / Xlib-Mesa libGL...

16 years agoinsert mesa/ before include files
Brian Paul [Thu, 5 Jun 2008 21:23:46 +0000 (15:23 -0600)]
insert mesa/ before include files

16 years agoSilence warning
Alan Hourihane [Thu, 5 Jun 2008 10:47:20 +0000 (11:47 +0100)]
Silence warning

16 years agofix event handling issues (such as when resizing when not animating)
Brian Paul [Wed, 4 Jun 2008 22:11:24 +0000 (16:11 -0600)]
fix event handling issues (such as when resizing when not animating)

16 years agopress 'a' to toggle animation
Brian Paul [Wed, 4 Jun 2008 22:04:04 +0000 (16:04 -0600)]
press 'a' to toggle animation

16 years agocode refactoring (draw_frame(0, handle_event())
Brian Paul [Wed, 4 Jun 2008 22:01:47 +0000 (16:01 -0600)]
code refactoring (draw_frame(0, handle_event())

16 years agoremove references to incorrect -winwidth/-winheight, replace with -geometry
Brian Paul [Wed, 4 Jun 2008 21:56:30 +0000 (15:56 -0600)]
remove references to incorrect -winwidth/-winheight, replace with -geometry

16 years agoSet the attribute as used.
Brian Paul [Wed, 4 Jun 2008 20:41:33 +0000 (14:41 -0600)]
Set the attribute as used.

cherry-picked from gallium-0.1

16 years agoautoconf: Collect arch/platform settings in one location
Dan Nicholson [Wed, 4 Jun 2008 20:00:35 +0000 (13:00 -0700)]
autoconf: Collect arch/platform settings in one location

The architecture- and platorm-specific settings were scattered
throughout the configure script. This moves them near the beginning
before any of the driver-specific settings.

16 years agoremove old/unused foomesa.h example header
Brian Paul [Wed, 4 Jun 2008 15:33:02 +0000 (09:33 -0600)]
remove old/unused foomesa.h example header

16 years agocall glutDestroyWindow() before exit
Brian Paul [Wed, 4 Jun 2008 15:30:04 +0000 (09:30 -0600)]
call glutDestroyWindow() before exit

16 years agoi915: Fix GL_DEPTH_TEXTURE_MODE issue. (bug #16221)
Xiang, Haihao [Wed, 4 Jun 2008 07:44:38 +0000 (15:44 +0800)]
i915: Fix GL_DEPTH_TEXTURE_MODE issue. (bug #16221)

16 years agomesa/drm/ttm: allow build against non-TTM aware libdrm
Dave Airlie [Wed, 28 May 2008 05:55:44 +0000 (15:55 +1000)]
mesa/drm/ttm: allow build against non-TTM aware libdrm

I'll release a libdrm 2.3.1 without TTM apis included from a special
drm branch that should allow mesa 7.1 to build against it.

I've had to turn off DRI2 stuff.

16 years agobug #16503: use INSTALL macro instead of hard-coding to bin/minstall
Ian Romanick [Mon, 2 Jun 2008 18:19:35 +0000 (11:19 -0700)]
bug #16503: use INSTALL macro instead of hard-coding to bin/minstall

This is only part of the fix.  Someone with more autoconf-fu will have
to fix the rest

16 years ago[t_vp_build] Fix refcounting-related memory leak
Nicolai Haehnle [Sat, 31 May 2008 22:10:45 +0000 (00:10 +0200)]
[t_vp_build] Fix refcounting-related memory leak

This memory leak is identical to the earlier one in texenvprogram.
Fixed by not creating an additional reference to new programs in cache_item.

On top of that, remove some duplicated state setting.

16 years ago[texenvprogram] Fix refcounting-related memory leak.
Nicolai Haehnle [Sat, 31 May 2008 22:01:46 +0000 (00:01 +0200)]
[texenvprogram] Fix refcounting-related memory leak.

All newly created programs have RefCount == 1, but the fragment program cache
added an additional reference in cache_item, with the result being that none
of the programs were ever freed.

Solve the problem by not creating the additional reference in cache_item.

16 years agor500: add missing brackets around depth testing
Dave Airlie [Sun, 1 Jun 2008 21:18:21 +0000 (07:18 +1000)]
r500: add missing brackets around depth testing

16 years agor5xx: Enable depth write emission.
Corbin Simpson [Sun, 1 Jun 2008 18:45:30 +0000 (11:45 -0700)]
r5xx: Enable depth write emission.
Thanks to nh for the r3xx version.

16 years agor300: Writing to result.depth in fragment programs (R3xx; only stub for R5xx)
Nicolai Haehnle [Sun, 1 Jun 2008 17:53:52 +0000 (19:53 +0200)]
r300: Writing to result.depth in fragment programs (R3xx; only stub for R5xx)

Setup fg_depth_src for depth writing programs and change early Z (ztop)
semantics.

Piglit's version of glean/fragprog test passes now (unlike Glean, its
dependency on EXT_fog_coord, which we don't support, is optional).

R3xx only at the moment, but should be straightforward to adapt to R5xx
(I don't own an R5xx, and I don't want to break anything.)

16 years ago[r300] Fix KIL instruction and swizzling of texture coordinates
Nicolai Haehnle [Sun, 1 Jun 2008 11:13:32 +0000 (13:13 +0200)]
[r300] Fix KIL instruction and swizzling of texture coordinates

The KIL instruction only works if at least one texture unit is enabled
in hardware.

Texture instructions do not support swizzles, negations etc. natively,
so we now emit an explicit swizzling etc. operation when the texture coordinate
requires it.

This fixes the Piglit fp-kil test.

16 years agor5xx: Add OPCODE_FLR.
Corbin Simpson [Sun, 1 Jun 2008 10:23:00 +0000 (03:23 -0700)]
r5xx: Add OPCODE_FLR.
Why didn't anybody tell me we were missing this one? I thought I committed it ages ago!

16 years agor5xx: Last emit_sop() for now.
Corbin Simpson [Sun, 1 Jun 2008 05:41:54 +0000 (22:41 -0700)]
r5xx: Last emit_sop() for now.
This should also clean up LIT later on.

16 years agor5xx: More emit_sop(), stage 2.
Corbin Simpson [Sun, 1 Jun 2008 05:23:11 +0000 (22:23 -0700)]
r5xx: More emit_sop(), stage 2.
SIN/COS.

16 years agor5xx: Cleanup SOP with emit_sop().
Corbin Simpson [Sun, 1 Jun 2008 05:09:38 +0000 (22:09 -0700)]
r5xx: Cleanup SOP with emit_sop().
I wish I had a snarky commit message, but I'll save that for after I get the
trig working with this.

16 years agor5xx: Dump shader constants when dumping program assembly.
Corbin Simpson [Sun, 1 Jun 2008 03:48:15 +0000 (20:48 -0700)]
r5xx: Dump shader constants when dumping program assembly.

16 years agor300: disable the lowimpact fallbacks by default.
Dave Airlie [Sat, 31 May 2008 23:32:55 +0000 (09:32 +1000)]
r300: disable the lowimpact fallbacks by default.

because really we should be able to just fix the driver.

16 years agor500: Use 8bit fog alpha compare value.
Markus Amsler [Sat, 31 May 2008 01:10:21 +0000 (03:10 +0200)]
r500: Use 8bit fog alpha compare value.

This fixes transparency issues on r500 by enabling
backwards compatibility with the r300 alpha compare value.

16 years agoFix segfault on exit in unichrome driver (bug 16150)
Luc Verhaegen [Thu, 29 May 2008 17:00:06 +0000 (11:00 -0600)]
Fix segfault on exit in unichrome driver (bug 16150)

16 years agoundo prev change to this VC7 project file
Brian Paul [Thu, 29 May 2008 16:46:39 +0000 (10:46 -0600)]
undo prev change to this VC7 project file

Meant to update the VC8 project file.

16 years agoadded prog_uniform.[ch]
Karl Schultz [Thu, 29 May 2008 16:43:48 +0000 (10:43 -0600)]
added prog_uniform.[ch]

16 years agoAssorted Windows fixes (Karl Schultz)
Brian Paul [Thu, 29 May 2008 16:41:31 +0000 (10:41 -0600)]
Assorted Windows fixes (Karl Schultz)

16 years agoUse the GLuint64EXT type defined by GL_EXT_timer_query
Brian Paul [Wed, 28 May 2008 15:10:48 +0000 (09:10 -0600)]
Use the GLuint64EXT type defined by GL_EXT_timer_query

Removes some compiler #ifdef stuff.

16 years agoR1/2/3/4/5xx: fixed calculation of cliprects in CopyBuffer.
Dennis Kasprzyk [Thu, 29 May 2008 09:25:21 +0000 (11:25 +0200)]
R1/2/3/4/5xx: fixed calculation of cliprects in CopyBuffer.

16 years agoReport correct damage rectangle in CopySubBuffer.
Dennis Kasprzyk [Thu, 29 May 2008 09:24:16 +0000 (11:24 +0200)]
Report correct damage rectangle in CopySubBuffer.

16 years agor300: fix R300_FG_ALPHA_FUNC_VAL_MASK
Markus Amsler [Thu, 29 May 2008 00:11:59 +0000 (02:11 +0200)]
r300: fix R300_FG_ALPHA_FUNC_VAL_MASK

16 years agoupdated function list
Brian Paul [Wed, 28 May 2008 15:06:28 +0000 (09:06 -0600)]
updated function list

16 years agoreplace make with $(MAKE) (bug 16133)
Brian Paul [Wed, 28 May 2008 14:53:32 +0000 (08:53 -0600)]
replace make with $(MAKE) (bug 16133)

16 years agor300: disable cmdbuf debug again
Dave Airlie [Wed, 28 May 2008 00:23:25 +0000 (10:23 +1000)]
r300: disable cmdbuf debug again

16 years agoMerge branch 'r500-support'
Dave Airlie [Wed, 28 May 2008 00:20:14 +0000 (10:20 +1000)]
Merge branch 'r500-support'

Bring in all the r500 3D code into master should be as stable as the r300
code it is landing on top off.

Conflicts:

src/mesa/drivers/dri/r300/r300_reg.h
src/mesa/drivers/dri/r300/r300_state.c
src/mesa/drivers/dri/radeon/radeon_screen.c

16 years agor500: remove warnings and only start on newer drms.
Dave Airlie [Wed, 28 May 2008 00:07:30 +0000 (10:07 +1000)]
r500: remove warnings and only start on newer drms.

This removes lots of warnings to the user, and only allows the driver
to run on > .29 drms for r500 cards.

16 years agoR3/4/5xx: update to use drm get_param for num gb pipes
Alex Deucher [Wed, 28 May 2008 00:11:14 +0000 (10:11 +1000)]
R3/4/5xx: update to use drm get_param for num gb pipes

16 years agor500: cleanup warnings and include files
Dave Airlie [Wed, 28 May 2008 00:03:10 +0000 (10:03 +1000)]
r500: cleanup warnings and include files

16 years agoupdate, re-org build/install instructions
Brian Paul [Tue, 27 May 2008 19:27:57 +0000 (13:27 -0600)]
update, re-org build/install instructions

16 years agoset version string to 7.1 rc1
Brian Paul [Tue, 27 May 2008 18:51:37 +0000 (12:51 -0600)]
set version string to 7.1 rc1

16 years agoupdates for 7.1 release
Brian Paul [Tue, 27 May 2008 18:51:10 +0000 (12:51 -0600)]
updates for 7.1 release

16 years agoprep for 7.1 RC1
Brian Paul [Tue, 27 May 2008 18:50:59 +0000 (12:50 -0600)]
prep for 7.1 RC1

16 years agoallow GLX_SAMPLES_ARB==0 (bug 16073)
Brian Paul [Tue, 27 May 2008 15:49:18 +0000 (09:49 -0600)]
allow GLX_SAMPLES_ARB==0 (bug 16073)

16 years agoupdated for latest sources
Jouk Jansen [Tue, 27 May 2008 14:58:39 +0000 (08:58 -0600)]
updated for latest sources

16 years agodependency fixes (bug 13544)
Karl Schultz [Tue, 27 May 2008 14:48:25 +0000 (08:48 -0600)]
dependency fixes (bug 13544)

16 years agor5xx: Add OPCODE_XPD. In working condition, I might add.
Corbin Simpson [Tue, 27 May 2008 09:12:10 +0000 (02:12 -0700)]
r5xx: Add OPCODE_XPD. In working condition, I might add.
And we're officially finished with the ARB_fragment_program
instruction set. It's worth noting that LIT is still not reliable.
SIN and COS were fixed a few commits ago. We're finished with stage 1!
Whoohoo!

16 years agor5xx: Just a few small LIT fixes.
Corbin Simpson [Tue, 27 May 2008 08:01:46 +0000 (01:01 -0700)]
r5xx: Just a few small LIT fixes.
Still broken; will fix tomorrow.

16 years agor5xx: Fix emit_mov() regression.
Corbin Simpson [Tue, 27 May 2008 05:34:32 +0000 (22:34 -0700)]
r5xx: Fix emit_mov() regression.
Specifically, fix improper swizzling.

16 years agor500: the cs is shared per context - doh so don't use for program upload.
Dave Airlie [Tue, 27 May 2008 08:15:14 +0000 (18:15 +1000)]
r500: the cs is shared per context - doh so don't use for program upload.

Also remove some unused bits of the r500 fragprog struct

16 years agor500: need to re-setup inst offset/end for translated programs
Dave Airlie [Tue, 27 May 2008 07:39:35 +0000 (17:39 +1000)]
r500: need to re-setup inst offset/end for translated programs

this fixes texenv

16 years agor500: reset fp/fp_const counts
Dave Airlie [Tue, 27 May 2008 07:11:04 +0000 (17:11 +1000)]
r500: reset fp/fp_const counts

16 years agor300/r500: emit flush inside vap_cntl state atom
Dave Airlie [Tue, 27 May 2008 07:08:03 +0000 (17:08 +1000)]
r300/r500: emit flush inside vap_cntl state atom

Not sure if this is a good or bad plan, it certainly doesn't make things worse here.

16 years agor500: hopefully fix 4096 texture harder
Dave Airlie [Tue, 27 May 2008 05:29:39 +0000 (15:29 +1000)]
r500: hopefully fix 4096 texture harder

16 years agor5xx: Fix FP temp counting.
Corbin Simpson [Tue, 27 May 2008 05:12:24 +0000 (22:12 -0700)]
r5xx: Fix FP temp counting.
One of the ref counters wasn't being added to the temp counter.
Yet another product of late-night coding...