summaryrefslogtreecommitdiff
path: root/src/mesa
AgeCommit message (Collapse)Author
2007-05-22fog: fix issues with negative fog coords (may fix #10529)Roland Scheidegger
Rework tnl fog a bit. Make sure we always use ABS(eyez) when fog coord source is depth, OTOH it does not seem to be necessary to use it (as was done before in some cases) if fog coord source is fogcoord (just to save some work). This fixes tests/fog (the first 2 cases) with i915/i915tex.
2007-05-22make sure optimized fog params get updatedRoland Scheidegger
2007-05-22i915/i915tex: minor cleanup (remove unneeded function callRoland Scheidegger
2007-05-22unbreak 3d textures (typo when setting tex layout)Roland Scheidegger
2007-05-21if light position is local, treat it as a homogeneous coord and divide by W ↵Brian
(see bug 11009)
2007-05-20Overhaul/simplify SWvertex and SWspan attribute handling.Brian
Instead of separate fog/specular/texcoord/varying code, just treat all of them as generic attributes. Simplifies the point/line/triangle functions.
2007-05-20add missing right-parenBrian
2007-05-20r300: Added TODO comment regarding immediate mode implementation.Oliver McFadden
2007-05-20r300: Just use "inline" rather than "__inline__".Oliver McFadden
2007-05-19need to copy new 1D/2D array texture objects in _mesa_PushAttrib()Brian
2007-05-19Implement GLX_EXT_texture_from_pixmap.Brian
Could be done more efficiently... but works.
2007-05-19fix miptree layout (i915) for small compressed mipmapsRoland Scheidegger
This seems to work now. Also fix i945 set_level_info, need to match i915 behaviour for storing mip height, as it's assumed to be the mip width (in texels) elsewhere and the division by 4 is done later (untested).
2007-05-19fix copy & paste bug of previous commit, breaking dxt5 formatsRoland Scheidegger
2007-05-19fix miptree comparison with compressed texturesRoland Scheidegger
TexelBytes is always 0 with compressed textures, thus would never match mt->cpp. This caused constant blitting around of textures, and it fixes at least the horrible performance of Q3 if compressed textures are enabled.
2007-05-19fix small s3tc mipmaps (#10968)Roland Scheidegger
make sure that always whole blocks are uploaded. (May still not work correctly if the top mip map is not at least a full block, that is 4 pixels wide - not sure, but probably doesn't happen in real world)
2007-05-18fix STATE_HALF_VECTOR value (bug 10987)Brian
2007-05-17change max anisotropy testChristoff Brill
2007-05-17use R200_DEBUG for debug outputChristoff Brill
2007-05-17remove CVS/XFree86 keywordsChristoff Brill
2007-05-17Add missing _EXT suffix to 2D_ARRAY target enums.Ian Romanick
2007-05-17Add array texture targets to list that can use compressed formats.Ian Romanick
2007-05-17Refactor determining whether a texture target can use compressed formatIan Romanick
2007-05-17make srcPtr param to make_1d_stack_mipmap() constBrian
2007-05-17r300: Removed the radeon_vertex_buffer structure.Tommy Schultz Lassen
2007-05-16Initial implementation of MESA_texture_arrayIan Romanick
Shadow sampling from texture arrays is still not implemented. Everything else should be there, though.
2007-05-15Bring framebuffer_texture's error checking more in-line with the spec.Ian Romanick
2007-05-14master/trunk is now for Mesa 7.1 devel: bump versionsBrian
2007-05-14comment about fixing uniform structsBrian
2007-05-13r300: Enable the non-user-buffers code to compile.Oliver McFadden
2007-05-13r300: Use e32(0x0) rather than e32(0); makes greping for hex values easier.Oliver McFadden
2007-05-13r300: Use the defined values when writing to R300_RS_ROUTE_0.Oliver McFadden
2007-05-13r300: Use the defined values when writing to R300_SE_VTE_CNTL.Oliver McFadden
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));
2007-05-13r300: Added a TODO comment for the R300_VAP_INPUT_ROUTE_[01]_0 values.Oliver McFadden
2007-05-13r300: Changed some more functions to static functions.Oliver McFadden
This required reordering some of the functions which is why the diff is a little larger.
2007-05-13r300: Changed a couple of functions in r300_state.c to static functions.Oliver McFadden
2007-05-13r300: Removed the unused r300_aos_rec (replaced by r300_dma_region) structure.Oliver McFadden
2007-05-13r300: Use #if 0 for disabled code, rather than commenting it out.Oliver McFadden
2007-05-13r300: Removed a couple of obsolete/commented out includes.Oliver McFadden
2007-05-13r300: A few very minor indenting corrections.Oliver McFadden
2007-05-13r300: Use the defined values for the interpolater (interp_magic) magic values.Oliver McFadden
2007-05-13r300: Use the defined values when writing to R300_VAP_INPUT_CNTL_[01].Oliver McFadden
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));
2007-05-12Revert "r300: Perform the locking closer to the DRM texture upload call."Oliver McFadden
This reverts commit 93881edb46fc95e1cfb4ded4e8a5db92612d3e4d.
2007-05-12bring over recent radeonMakeCurrent fixes for r300 to radeon/r200Roland Scheidegger
2007-05-12r300: Removed the name argument to ALLOC_STATE; do the name via the macro.Oliver McFadden
2007-05-12r300: Renamed r300DoEmitState to r300EmitAtoms.Oliver McFadden
2007-05-12r300: Perform the locking closer to the DRM texture upload call.Oliver McFadden
This might help if with attaching GDB if the driver gets into a -EAGAIN loop.
2007-05-11r300: Removed unused FORCE_32BITS_ELTS define.Oliver McFadden
2007-05-11r300: Removed the unused CPT macro.Oliver McFadden
2007-05-11r300: Moved r300PackFloat24 near r300PackFloat32.Oliver McFadden
2007-05-11r300: Added insert_at_tail to the ALLOC_STATE macro.Oliver McFadden
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-09 12:41:45 +0000