gallivm: make use of new iround code in lp_bld_conv.
[mesa.git] / src / gallium / auxiliary / gallivm /
2010-10-08 Roland Scheideggergallivm: make use of new iround code in lp_bld_conv.
2010-10-08 Roland Scheideggergallivm: optimize soa linear clamp to edge wrap mode...
2010-10-08 Roland Scheideggergallivm: avoid unnecessary URem in linear wrap repeat...
2010-10-08 Roland Scheideggergallivm: more linear tex wrap mode calculation simplifi...
2010-10-08 Roland Scheideggergallivm: optimize some tex wrap mode calculations a bit
2010-10-08 Roland Scheideggergallivm: replace sub/floor/ifloor combo with ifloor_fract
2010-10-08 Roland Scheideggergallivm: faster iround implementation for sse2
2010-10-08 Roland Scheideggergallivm: fix trunc/itrunc comment
2010-10-08 Vinson Leegallivm: Remove unnecessary header.
2010-10-08 José Fonsecagallivm: Help for combined extraction and broadcasting.
2010-10-08 José Fonsecallvmpipe: First minify the texture size, then broadcast.
2010-10-08 José Fonsecagallivm: Move into the as much of the second level...
2010-10-08 José Fonsecagallivm: Warn when doing inefficient integer comparisons.
2010-10-08 Keith Whitwellgallivm: round rather than truncate in new 4x4f->1x16ub...
2010-10-08 José Fonsecagallivm: Use the wrappers for SSE pack intrinsics.
2010-10-08 Keith Whitwellgallivm: special case conversion 4x4f to 1x16ub
2010-10-08 José Fonsecagallivm: Implement brilinear filtering.
2010-10-08 José Fonsecagallivm: Fix copy'n'paste typo in previous commit.
2010-10-08 José Fonsecagallivm: Clamp mipmap level and zero mip weight simulta...
2010-10-08 José Fonsecagallivm: Use lp_build_ifloor_fract for lod computation.
2010-10-08 José Fonsecagallivm: Don't compute the second mipmap level when...
2010-10-08 José Fonsecagallivm: Simplify lp_build_mipmap_level_sizes' interface.
2010-10-08 José Fonsecagallivm: Do not do mipfiltering when magnifying.
2010-10-07 José Fonsecagallivm: Vectorize the rho computation.
2010-10-06 José Fonsecagallivm: Compute lod as integer whenever possible.
2010-10-06 José Fonsecagallivm: Only apply min/max_lod when necessary.
2010-10-06 Keith Whitwellgallivm: don't apply zero lod_bias
2010-10-06 José Fonsecagallivm: Combined ifloor & fract helper.
2010-10-06 José Fonsecagallivm: Fast implementation of iround(log2(x))
2010-10-06 José Fonsecagallivm: Use a faster (and less accurate) log2 in lod...
2010-10-06 José Fonsecagallivm: Take the type signedness in consideration...
2010-09-30 Brian Paulgallivm: added some comments
2010-09-30 Brian Paulgallivm: check for level=0 case in lp_build_minify()
2010-09-30 José Fonsecagallivm: More comprehensive border usage logic.
2010-09-29 José Fonsecagallivm: Use SSE4.1's ROUNDSS/ROUNDSD for scalar rounding.
2010-09-27 Brian Paulllvmpipe: fix swizzling of texture border color
2010-09-25 Brian Paulgallivm: fix repeat() function for NPOT textures
2010-09-25 José Fonsecagallivm: Remove dead experimental code.
2010-09-25 José Fonsecagallivm: Fetch the lod from the dynamic state when...
2010-09-25 Roland Scheideggergallivm: optimize yuv decoding
2010-09-25 Roland Scheideggergallivm: don't use URem/UDiv when calculating offsets...
2010-09-25 Roland Scheideggergallivm: fix copy&paste bug
2010-09-24 Brian Paulllvmpipe: make texture border_color dynamic state
2010-09-24 Brian Paulllvmpipe: make min/max lod and lod bias dynamic state
2010-09-22 José Fonsecagallivm: Add unorm support to lp_build_lerp()
2010-09-21 Brian Paulgallivm: fix lp_build_sample_compare()
2010-09-21 Brian PaulMerge branch 'sprite-coord'
2010-09-20 Brian Paulgallivm: remove debug code
2010-09-17 Brian Paulgallivm: added missing case for PIPE_TEXTURE_RECT
2010-09-16 Vinson Leegallivm: Remove unnecessary header.
2010-09-16 Brian Paulgallivm: fix wrong return value in bitwise functions
2010-09-16 José Fonsecagallivm: Clamp indirect register indices to file_max.
2010-09-16 José Fonsecagallivm: Start collecting bitwise arithmetic helpers...
2010-09-16 José Fonsecagallivm: Fix address register swizzle.
2010-09-16 Keith Whitwellgallivm: make lp_build_sample_nop public
2010-09-16 Brian Paulgallivm: move i32_vec_type inside the #ifdef
2010-09-16 Brian Paulgallivm: fix incorrect vector shuffle datatype
2010-09-16 Vinson Leegallivm: Remove unnecessary headers.
2010-09-15 Brian Paulgallivm: expand AoS sampling to cover all filtering...
2010-09-15 Christoph BumillerMerge remote branch 'origin/nv50-compiler'
2010-09-14 Vinson Leegallivm: Remove unnecessary header.
2010-09-11 José Fonsecagallivm: nr_channels is only valid for formats with...
2010-09-11 José Fonsecagallivm: Use const keyword on swizzles.
2010-09-11 José Fonsecagallivm: Allow to TGSI AoS translation to happen in...
2010-09-11 José Fonsecagallivm: Add a new debug flag to warn about performance...
2010-09-11 José Fonsecagallivm: Helper functions for pointer indirection.
2010-09-11 José Fonsecagallivm: Cleanup the TGSI <-> sampler interface.
2010-09-11 José Fonsecagallivm: Add some utility functions to set/get array...
2010-09-11 José Fonsecagallivm: Basic AoS TGSI -> LLVM IR.
2010-09-11 José Fonsecagallivm: Move the texture modifiers to the header.
2010-09-11 José Fonsecagallivm: s/lp_build_broadcast_aos/lp_build_swizzle_scal...
2010-09-05 José Fonsecagallivm: Pass condition masks as an unsigned bitmask.
2010-09-05 José Fonsecagallivm: Cope with tgsi instruction reallocation failure.
2010-09-02 Christoph BumillerMerge remote branch 'origin/master' into nv50-compiler
2010-09-01 Brian Paulgallivm: fix bug in nested conditionals
2010-08-30 José Fonsecagallivm: Compute the 4 texel offsets for linear filteri...
2010-08-30 José Fonsecagallivm: Disable LLVM's pretty stack trace dumper.
2010-08-30 José Fonsecagallivm: Correct copy'n'pasted comments.
2010-08-30 José Fonsecagallivm: Fix lp_build_sum_vector.
2010-08-29 Marek Olšákutil: remove util_is_pot in favor of util_is_power_of_two
2010-08-25 Vinson Leegallivm: Include missing header in lp_bld_sample.h.
2010-08-25 Vinson Leegallivm: Include missing header in lp_bld_pack.h.
2010-08-21 José Fonsecagallivm: Emit DIVPS instead of RCPPS.
2010-08-20 Luca Barbierigallium: make all checks for PIPE_TEXTURE_2D check...
2010-08-18 Christoph BumillerMerge remote branch 'origin/master' into nv50-compiler
2010-08-18 michalgallivm: Use proper index to lookup predicate register...
2010-08-17 nobledgallivm: Fix and re-enable MMX-disabling code
2010-08-17 Ian RomanickMerge branch 'glsl2'
2010-08-14 Vinson Leegallivm: Remove unnecessary header.
2010-08-14 Luca Barbieriu_cpu_detect: remove arch and little_endian
2010-08-14 José Fonsecagallivm: Refactor the Newton-Rapshon steps, and disable...
2010-08-14 Ian RomanickMerge branch 'master' into glsl2
2010-08-11 José Fonsecagallivm: Fix and enable the extra Newton/Raphson step...
2010-08-11 José Fonsecagallivm: Use lp_build_div instead of lp_build_mul ...
2010-08-11 José Fonsecagallivm: Use unsigned shift in lp_build_minify.
2010-08-10 Brian Paulgallivm: fix non-SSE4.1 case in lp_build_pack2()
2010-08-10 nobledgallivm: Fix bitwise operations for floats, division...
2010-08-10 nobledgallivm: Even more type checking
2010-08-09 José Fonsecagallivm: More type checks.
2010-08-09 José Fonsecagallivm: Don't call LLVMBuildFNeg on llvm-2.6.
next