Xiang, Haihao [Mon, 30 Jul 2007 16:11:22 +0000 (00:11 +0800)]
Fix an error in _save_OBE_DrawElements
In the case that a buffer object is bound to
ELEMENT_ARRARY_BUFFER, it is invalid to directly
dereference indices passed to glDrawElements.
Xiang, Haihao [Mon, 30 Jul 2007 15:50:52 +0000 (23:50 +0800)]
Fix an error in _mesa_validate_DrawElements
The field 'Data' in an obj is NULL in some dri drivers(such
as i965). Use mapped pointer instead.
Brian [Mon, 30 Jul 2007 15:01:02 +0000 (09:01 -0600)]
New sunos5-v9-cc-gcc config (Roland Egger)
Brian [Mon, 30 Jul 2007 14:52:57 +0000 (08:52 -0600)]
misc comments, clean-ups
Brian [Mon, 30 Jul 2007 14:13:40 +0000 (08:13 -0600)]
remove old comment, minor formatting fixes
Xiang, Haihao [Mon, 30 Jul 2007 08:30:32 +0000 (16:30 +0800)]
handle LSB_FIRST in _mesa_pack_bitmap by the way used in
_mesa_unpack_bitmap
Xiang, Haihao [Mon, 30 Jul 2007 08:17:40 +0000 (16:17 +0800)]
Fix an error related to glPolygonStipple.
As glPixelStore(GL_UNPACK) affect the bits into a stipple pattern,
hence 128 bytes used to store the pattern in a display list aren't
enough sometimes.
Zou Nan hai [Mon, 30 Jul 2007 02:18:11 +0000 (10:18 +0800)]
ARB sprite point support on i965
Brian [Mon, 30 Jul 2007 00:28:46 +0000 (18:28 -0600)]
added a few more functions
Jan Dvorak [Mon, 30 Jul 2007 00:26:22 +0000 (18:26 -0600)]
fix glPointParameteriv bug 11754
Brian [Mon, 30 Jul 2007 00:04:23 +0000 (18:04 -0600)]
Added shader points and shader bitmap demos
Roland Scheidegger [Sun, 29 Jul 2007 17:44:55 +0000 (19:44 +0200)]
fix cos/sin range reduction for i915 driver too
Roland Scheidegger [Sun, 29 Jul 2007 17:40:50 +0000 (19:40 +0200)]
fix range reduction for sin/cos in i915tex (#11609)
Roland Scheidegger [Sun, 29 Jul 2007 16:31:36 +0000 (18:31 +0200)]
fix fallback crashes when driver can't handle frag prog for i915 driver too (untested)
Roland Scheidegger [Sun, 29 Jul 2007 16:04:28 +0000 (18:04 +0200)]
Fix crashes when the frag prog can't be handled in hardware (#11131)
Must not change to/from swrast after Render.Start or bad things will happen.
(Driver will still somewhat incorrectly report an implementation error,
and apps can't really figure out if a prog is natively supported as validation
is later - could try doing it earlier to give some hint at least, even though
native status may still change later due to fog etc.)
Zhang [Fri, 27 Jul 2007 17:19:35 +0000 (11:19 -0600)]
more Mingw32 fixes
Brian [Thu, 26 Jul 2007 22:42:05 +0000 (16:42 -0600)]
fix-up inlined/non-inlined function inconsistencies
Brian [Thu, 26 Jul 2007 21:32:02 +0000 (15:32 -0600)]
Fix function call bug 11731. Also, fix up IR_CALL/IR_FUNC confusion.
Brian [Thu, 26 Jul 2007 17:39:11 +0000 (11:39 -0600)]
generate error upon writing to varying var in fragment program (bug 11733)
Brian [Thu, 26 Jul 2007 14:22:47 +0000 (08:22 -0600)]
clamp float colors
Brian [Thu, 26 Jul 2007 14:22:28 +0000 (08:22 -0600)]
fix color interpolation for CHAN_BITS==32
Brian [Thu, 26 Jul 2007 14:22:09 +0000 (08:22 -0600)]
don't use rgba_line() if CHAN_BITS==32
Brian [Tue, 24 Jul 2007 23:45:14 +0000 (17:45 -0600)]
remove unused MAX_3D_TEXTURE_SIZE, reformattting
Brian [Tue, 24 Jul 2007 15:56:44 +0000 (09:56 -0600)]
call ctx->Driver.NewProgram() instead of _mesa_new_program()
Brian [Tue, 24 Jul 2007 15:19:22 +0000 (09:19 -0600)]
fix logic error, typos
Brian [Mon, 23 Jul 2007 22:51:57 +0000 (16:51 -0600)]
fix GLX_STEREO handling (bug 11705)
Zhang [Sat, 21 Jul 2007 17:28:06 +0000 (11:28 -0600)]
Fix a number of MINGW32 issues
Brian [Sat, 21 Jul 2007 16:06:18 +0000 (10:06 -0600)]
Remove ctx->Point._Size and ctx->Line._Width.
The clamping for these values depends on whether we're drawing AA or non-AA
points, lines. Defer clamping until drawing time. Drivers could compute and
keep clamped AA and clamped non-AA values if desired.
Brian [Sat, 21 Jul 2007 16:04:14 +0000 (10:04 -0600)]
remove VC6/VC7 project files until updated
Roland Scheidegger [Fri, 20 Jul 2007 09:34:26 +0000 (11:34 +0200)]
fix bogus recently introduced function param comments
Roland Scheidegger [Wed, 18 Jul 2007 18:17:14 +0000 (20:17 +0200)]
fix mesa's handling of fbo's / window fb (again)
Make sure the relevant fields in window fbs get updated at appropriate time
(those are NOT the same as fbos!!!), and fix up related code accordingly.
This is a bit ugly, but there's a reason the issues section in EXT_fbo is
a couple hundred pages long...
Hopefully correct now.
Roland Scheidegger [Wed, 18 Jul 2007 16:07:12 +0000 (18:07 +0200)]
more fixes for mesa's fbo handling (fixes tests/fbotest1/2)
Roland Scheidegger [Thu, 19 Jul 2007 15:58:21 +0000 (17:58 +0200)]
fix up mesa's probably bogus framebuffer updates with different read/write fbs
Conflicts:
src/mesa/main/framebuffer.c
Roland Scheidegger [Tue, 17 Jul 2007 15:29:55 +0000 (17:29 +0200)]
fix mesa fb binding
Make sure that we bind the right buffer (draw or read) when rebinding
the window framebuffer (the api doesn't allow binding different draw and
read buffers at the same time, but the default window framebuffer is basically
2 fb objects, one for read, one for write, which can be different). Pass both
of these two down the driver api (no driver uses this right now).
Roland Scheidegger [Tue, 17 Jul 2007 10:56:30 +0000 (12:56 +0200)]
increase MAX_RELOCS so never run out before batch buffer is full (fixes xdemos/shape)
Oliver McFadden [Wed, 18 Jul 2007 02:59:08 +0000 (02:59 +0000)]
Oliver McFadden [Wed, 18 Jul 2007 02:49:26 +0000 (02:49 +0000)]
r300: Cleaned up vertprog construction.
Construct the vertprog instruction in the 4 DWORD parts...
DWORD 0: Opcode and Output.
DWORD 1: First Argument.
DWORD 2: Second Argument.
DWORD 3: Third Argument.
Allow the opcode translation functions to generate more than one instruction;
useful for when an instruction must be emulated. FLR, XPD, etc.
Tommy Schultz Lassen [Wed, 18 Jul 2007 00:37:43 +0000 (00:37 +0000)]
r300: Corrected texcoord start when BFC1 is enabled.
Oliver McFadden [Mon, 16 Jul 2007 21:43:39 +0000 (21:43 +0000)]
r300: Corrected some progs/fp/* regressions from the BFC patch.
I'm not completely sure this is correct; it restores the old behaviour.
Oliver McFadden [Mon, 16 Jul 2007 20:24:41 +0000 (20:24 +0000)]
r300: Use _mesa_num_inst_src_regs for number of arguments.
Oliver McFadden [Mon, 16 Jul 2007 20:07:30 +0000 (20:07 +0000)]
r300: Corrected vertprog FLR and XPD instruction regression.
Oliver McFadden [Mon, 16 Jul 2007 19:20:32 +0000 (19:20 +0000)]
r300: Added code for vertprog opcode ARL.
Oliver McFadden [Mon, 16 Jul 2007 19:13:46 +0000 (19:13 +0000)]
r300: Don't need vertprog scalar flag anymore; it's handled explicitly...
Oliver McFadden [Mon, 16 Jul 2007 19:10:13 +0000 (19:10 +0000)]
r300: Removed broken RCC vertprog opcode.
Oliver McFadden [Mon, 16 Jul 2007 19:09:29 +0000 (19:09 +0000)]
r300: More vertprog rework; give each opcode it's own function.
Oliver McFadden [Mon, 16 Jul 2007 16:21:21 +0000 (16:21 +0000)]
Revert "r300: Gracefully exit after GART memory is exhausted."
This reverts commit
9457bf62bbba3b9226ebbbea5dc7798ca22485f6.
Causes the X server to die with Compiz and Beryl.
Oliver McFadden [Mon, 16 Jul 2007 16:17:50 +0000 (16:17 +0000)]
r300: Reorder the vertprog code to the ARB specification.
Roland Scheidegger [Mon, 16 Jul 2007 16:21:36 +0000 (18:21 +0200)]
fix bogus fb/drawable information
the framebuffer objects attached to drawables can have invalidate state
associated with them, since for the window framebuffer this is per-context
state and not per-fbo state. Since drivers rely on that information
(otherwise would need to check if currently the window-framebuffer is
bound in a lot of places) fix it up in _mesa_make_current (ugly).
(Brought over from i915tex_privbuffers, where it fixes xdemos/wincopy
when switching to front buffer rendering.)
Roland Scheidegger [Mon, 16 Jul 2007 16:09:30 +0000 (18:09 +0200)]
fix segfault with i915 drivers in swrast drawpixels path when resizing windows
Oliver McFadden [Mon, 16 Jul 2007 15:42:16 +0000 (15:42 +0000)]
r300: Use the R300_PVS_UPLOAD_* defines.
Oliver McFadden [Mon, 16 Jul 2007 15:39:12 +0000 (15:39 +0000)]
r300: Added the clip plane upload defines.
Oliver McFadden [Mon, 16 Jul 2007 12:21:38 +0000 (12:21 +0000)]
r300: Enable the vertprog point size again.
Oliver McFadden [Mon, 16 Jul 2007 11:34:40 +0000 (11:34 +0000)]
r300: Corrected position bug with position invariant option. Bug #11594.
Oliver McFadden [Mon, 16 Jul 2007 07:17:22 +0000 (07:17 +0000)]
r300: Removed Vim modeline I left in the file by mistake. :-)
Oliver McFadden [Mon, 16 Jul 2007 07:12:29 +0000 (07:12 +0000)]
r300: Make sure the modified value of u_temp_i is respected.
Oliver McFadden [Mon, 16 Jul 2007 06:38:12 +0000 (06:38 +0000)]
r300: Added a function for vertex program default opcode translation.
Oliver McFadden [Mon, 16 Jul 2007 06:33:53 +0000 (06:33 +0000)]
r300: Clean up the vertex program maximum length check.
Oliver McFadden [Mon, 16 Jul 2007 06:27:29 +0000 (06:27 +0000)]
r300: Corrected indenting in r300_vertprog.c.
Oliver McFadden [Mon, 16 Jul 2007 06:16:35 +0000 (06:16 +0000)]
r300: Major vertex program code clean up and rework.
Oliver McFadden [Mon, 16 Jul 2007 05:04:10 +0000 (05:04 +0000)]
r300: Replaced the ugly VERTEX_SHADER_INSTRUCTION typedef.
Oliver McFadden [Sun, 15 Jul 2007 01:17:54 +0000 (01:17 +0000)]
r300: Gracefully exit after GART memory is exhausted.
Oliver McFadden [Sat, 14 Jul 2007 18:31:29 +0000 (18:31 +0000)]
r300: Disable vertex program point size; it's almost certainly wrong.
See the conversation between myself and Tommy Schultz Lassen on mesa3d-dev.
Oliver McFadden [Sat, 14 Jul 2007 18:11:45 +0000 (18:11 +0000)]
r300: Clean up the vertex program output assignment.
Patrice Mandin [Sat, 14 Jul 2007 16:54:22 +0000 (18:54 +0200)]
nouveau: nv10 and nv11,15 are different
Jan Dvorak [Fri, 13 Jul 2007 22:36:00 +0000 (16:36 -0600)]
fix shader/info string length queries (bug 11588)
Ben Skeggs [Fri, 13 Jul 2007 13:39:24 +0000 (23:39 +1000)]
Play "nuke the typedef"
Oliver McFadden [Fri, 13 Jul 2007 09:39:23 +0000 (09:39 +0000)]
Revert "r300: Corrected off-by-one error in r300_vertprog.c."
This reverts commit
07ac2386f5c0ab9c2432d4b5e3490b1e13d033fc.
Oliver McFadden [Fri, 13 Jul 2007 04:58:32 +0000 (04:58 +0000)]
r300: Corrected off-by-one error in r300_vertprog.c.
Oliver McFadden [Fri, 13 Jul 2007 02:11:58 +0000 (02:11 +0000)]
r300: Moved some hardware initialization into appropriate functions.
Oliver McFadden [Fri, 13 Jul 2007 01:45:05 +0000 (01:45 +0000)]
r300: Use a switch statement in r300UpdateCulling.
Oliver McFadden [Thu, 12 Jul 2007 23:52:06 +0000 (23:52 +0000)]
r300: Setup the back facing color bits correctly.
Apparently for back facing color to work you must set all 3 color bits; I guess
the hardware cannot handle them separately.
Brian [Fri, 13 Jul 2007 03:32:35 +0000 (21:32 -0600)]
fix width/depth mix-up (bug 11577)
Brian [Thu, 12 Jul 2007 20:50:53 +0000 (14:50 -0600)]
fix swizzle-related bug 11534
Patrice Mandin [Thu, 12 Jul 2007 16:36:41 +0000 (18:36 +0200)]
nv20 also don't support viewport_origin, like nv10
Ben Skeggs [Thu, 12 Jul 2007 01:08:17 +0000 (11:08 +1000)]
nouveau: match drm 0.0.9 interface
Brian [Wed, 11 Jul 2007 22:33:15 +0000 (16:33 -0600)]
clean-up of stencil code
Brian [Wed, 11 Jul 2007 22:19:51 +0000 (16:19 -0600)]
Minor clean-up of polygon offset logic. Properly compute _MRD field.
Brian [Wed, 11 Jul 2007 19:11:47 +0000 (13:11 -0600)]
Improve this demo in a few ways.
1. Use more reasonable hither/yon clip planes to make better use of shallow
(16-bit) z buffers.
2. Use different colors on cube faces to help detect Z fighting if it occurs.
3. Report GL_DEPTH_BITS on start-up.
Tommy Schultz Lassen [Wed, 11 Jul 2007 17:37:12 +0000 (17:37 +0000)]
r300: Added support for Back Facing Color.
Tommy Schultz Lassen [Wed, 11 Jul 2007 17:35:35 +0000 (17:35 +0000)]
r300: Corrected r300UpdateCulling; would only set font face if culling was set.
Patrice Mandin [Wed, 11 Jul 2007 17:25:32 +0000 (19:25 +0200)]
nouveau: nv20 does not support hw scissors
Arthur Huillet [Wed, 11 Jul 2007 00:48:40 +0000 (02:48 +0200)]
bumped nouveau DRM interface version number
Michel Dänzer [Tue, 10 Jul 2007 09:02:18 +0000 (11:02 +0200)]
i915tex: Better procedure for dropping batchbuffer on virtual resolution change.
The previous procedure would often result in a GPU lockup.
Michel Dänzer [Tue, 10 Jul 2007 08:54:51 +0000 (10:54 +0200)]
i915tex: Only wait for vblank when really necessary.
This avoids superfluous waits for vblank timing out under some circumstances.
Michel Dänzer [Tue, 10 Jul 2007 08:49:28 +0000 (10:49 +0200)]
Clear pointers to freed cliprects.
Not doing this could lead to double frees under rare circumstances.
Brian [Sun, 8 Jul 2007 15:58:18 +0000 (09:58 -0600)]
Fix size test bug in _mesa_test_proxy_teximage().
width/height/depth == 0 is a legal texture size (no error generated).
Later, the texture will be considered incomplete, however, and texturing
will effectively be disabled.
See bug 11309.
Brian [Sun, 8 Jul 2007 15:20:42 +0000 (09:20 -0600)]
Check if 'indices' parameter is NULL (bug 11314)
Brian [Sun, 8 Jul 2007 15:02:36 +0000 (09:02 -0600)]
Change float depthScale param to _mesa_unpack_depth_span() to GLuint depthMax.
Brian [Sun, 8 Jul 2007 14:53:19 +0000 (08:53 -0600)]
check depthScale value for optimized ushort->uint case (fixes bug 11474)
Brian [Fri, 6 Jul 2007 22:51:19 +0000 (16:51 -0600)]
In _swsetup_Translate(), update dest->attrib[FRAG_ATTRIB_COL0].
Also, check if we're in RGB vs. CI mode. This fixes a problem with
incorrect rendering color seen with the redbook/polys demo.
Brian [Fri, 6 Jul 2007 22:50:13 +0000 (16:50 -0600)]
Add case for GL_QUADS in i915_reduced_primitive_state().
The t_dd_tritemp.h code can emit GL_QUADS primitives. We need to catch
that case to determine if polygon stipple should be enabled.
Fixes bug reported by Carlos Diógenes on 4 July 2007.
Patrice Mandin [Fri, 6 Jul 2007 22:08:35 +0000 (00:08 +0200)]
Merge branch 'master' of git+ssh://pmandin@git.freedesktop.org/git/mesa/mesa
Patrice Mandin [Fri, 6 Jul 2007 21:59:29 +0000 (23:59 +0200)]
NV_17 is different from NV_10
Patrice Mandin [Fri, 6 Jul 2007 21:55:51 +0000 (23:55 +0200)]
NV1X don't support VIEWPORT_ORIGIN in hardware
Xavier Bachelot [Fri, 6 Jul 2007 18:56:21 +0000 (12:56 -0600)]
call glutInit(), bug 11486
Brian [Thu, 5 Jul 2007 15:37:46 +0000 (09:37 -0600)]
stencil pixel map didn't work in _mesa_unpack_stencil_span(), bug 11475
Brian [Wed, 4 Jul 2007 19:15:20 +0000 (13:15 -0600)]
Be more consistant with paths in #includes. Eventually, eliminate a bunch of -I flags.
Brian [Wed, 4 Jul 2007 15:23:12 +0000 (09:23 -0600)]
support more modes, added -f cmd line option
Brian [Wed, 4 Jul 2007 15:22:15 +0000 (09:22 -0600)]
assorted clean-ups
Eric Anholt [Wed, 4 Jul 2007 13:37:14 +0000 (07:37 -0600)]
fix LogicOp/bitmap problem, bug 11133