summaryrefslogtreecommitdiff
AgeCommit message (Expand)AuthorFilesLines
2014-11-14i965: Use df uniform loading for hsw alsofp64_lower_df_hswTopi Pohjolainen3-7/+0
2014-11-14i965/hsw: Support for df register regionsTopi Pohjolainen1-8/+24
2014-11-13Revert "denorm wip"Topi Pohjolainen4-7/+1
2014-11-13denorm wipTopi Pohjolainen4-1/+7
2014-11-12Revert "wip"Topi Pohjolainen1-4/+0
2014-11-12wipTopi Pohjolainen1-0/+4
2014-11-11dbg/fs: Dump instructions before double loweringfp64_lower_dfTopi Pohjolainen1-0/+4
2014-11-11i965/eu: Use simd8 for treating hi/lo 32-bits of 16-width doublesTopi Pohjolainen1-0/+2
2014-11-11i965/eu: Allow 3-src float ops with doublesTopi Pohjolainen1-6/+14
2014-11-11i965/vec4: Make sure math isn't tried with doublesTopi Pohjolainen1-0/+5
2014-11-11i965/vec4: Treat double precision floats as any other basic typeTopi Pohjolainen1-1/+1
2014-11-11i965/fs: Double precision to single conversion supportTopi Pohjolainen2-2/+2
2014-11-11lower dfTopi Pohjolainen1-0/+40
2014-11-11i965: Add lowering pass splitting df-ops into twoTopi Pohjolainen4-0/+194
2014-11-11i965/fs: Generator support for packing d2f resultsTopi Pohjolainen1-0/+27
2014-11-11i965: Add new opcode for packing d2f resultsTopi Pohjolainen2-0/+14
2014-11-11i965/fs: Add support for ir_unop_unpack_double_2x32Topi Pohjolainen2-1/+13
2014-11-11i965/fs: Add unpack_double_2x32 virtual opcodeTopi Pohjolainen4-0/+93
2014-11-11i965/fs: Lower double precision scalars into vectorsTopi Pohjolainen1-4/+39
2014-11-11i965: Add helper telling if uniform is double and requires special loadTopi Pohjolainen2-0/+38
2014-11-11i965/fs/gen7: Add generator support for loading double precision uniformsTopi Pohjolainen4-0/+39
2014-11-11i965/gen7: Add support for loading double float scalars in 16-widthTopi Pohjolainen2-0/+28
2014-11-11i965/gen7: Add support for double precision constant operandsTopi Pohjolainen1-0/+5
2014-11-11i965/gen8: Add support for double precision constant operandsTopi Pohjolainen5-0/+26
2014-11-11i965/fs: Add support for ir_unop_pack_double_2x32Topi Pohjolainen3-0/+25
2014-11-11i965/fs: Add pack_double_2x32 virtual opcodeTopi Pohjolainen4-0/+85
2014-11-11i965: Add helper telling if a register is scalarTopi Pohjolainen1-0/+8
2014-11-11i965/fs: Add support for double precision uniform loadingTopi Pohjolainen1-7/+9
2014-11-11i965/fs: Prepare live interval analysis for double precisionTopi Pohjolainen1-1/+6
2014-11-11i965/fs: Take double float into account in register offsetsTopi Pohjolainen2-3/+9
2014-11-11i965/fs: Prepare virtual registers for double precision floatsTopi Pohjolainen1-1/+4
2014-11-11i965: Tell backend register about double precision typeTopi Pohjolainen1-1/+2
2014-11-11i965: Determine size of double precision float registerTopi Pohjolainen1-0/+2
2014-11-11i965: Lower DFSQRTTopi Pohjolainen1-1/+2
2014-11-11glsl: Add lowering pass approximating sqrt(double) with sqrt(float)Topi Pohjolainen2-0/+213
2014-11-11i965: Lower DFRACEXP/DLDEXPTopi Pohjolainen1-1/+2
2014-11-11glsl: add a lowering pass for frexp/ldexp with double argumentsIlia Mirkin2-2/+280
2014-11-11glsl: handle d2f/f2d in mat to vec visitorIlia Mirkin1-0/+2
2014-11-11st_glsl_to_tgsi: handle fp64 constants betterDave Airlie1-32/+51
2014-11-11tgsi: add support for flt64 constantsDave Airlie8-6/+85
2014-11-11st/glsl: add st fp64 support (v2)Dave Airlie2-139/+539
2014-11-11tgsi: expose doubles for softpipe.Dave Airlie1-1/+1
2014-11-11gallium: add double opcodes and TGSI execution (v2)Dave Airlie3-14/+779
2014-11-11i965: just avoid warnings with fp64Dave Airlie4-0/+38
2014-11-11glsl: lower double optional passesDave Airlie2-0/+211
2014-11-11glsl: implement double builtin functionsDave Airlie1-259/+492
2014-11-11glsl/lower_instructions: add double lowering passesDave Airlie1-0/+83
2014-11-11glsl: enable/disable certain lowering passes for doublesDave Airlie1-3/+3
2014-11-11SQUASH: support double when constructing matrix from scalarTapani Pälli2-9/+21
2014-11-11SQUASH: set correct size for double array uniformsTapani Pälli2-2/+4
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-24 00:21:27 +0000