summaryrefslogtreecommitdiff
AgeCommit message (Expand)AuthorFilesLines
2016-05-04HACK test SIMD16 interference workaroundtest-fp64-compressed-waConnor Abbott1-1/+16
2016-04-29i965/fs: fix MOV_INDIRECT exec_size for doublesSamuel Iglesias Gonsálvez1-1/+7
2016-04-29i965/fs: take into account doubles when calculating read_size for MOV_INDIRECTSamuel Iglesias Gonsálvez1-2/+2
2016-04-29i965/fs: demote_pull_constants() did not take into account double typesSamuel Iglesias Gonsálvez1-1/+8
2016-04-29i965/fs: align access to double-based uniforms in push constant bufferSamuel Iglesias Gonsálvez5-9/+64
2016-04-29i965/fs: recognize writes with a subreg_offset > 0 as partialIago Toral Quiroga1-1/+2
2016-04-29i965/fs/lower_simd_width: Fix registers written for split instructionsIago Toral Quiroga1-2/+2
2016-04-29i965/fs: rename our lower_d2f pass to lower_d2xIago Toral Quiroga4-4/+4
2016-04-29i965/fs: implement i2d and u2dIago Toral Quiroga1-0/+2
2016-04-29i965/fs: implement d2i and d2uIago Toral Quiroga2-2/+6
2016-04-29i965/fs: implement d2bIago Toral Quiroga1-0/+13
2016-04-29i965/fs: implement fsign() for doublesIago Toral Quiroga1-17/+81
2016-04-29i965/fs: add null_reg_dfIago Toral Quiroga1-0/+7
2016-04-29i965/fs: We only support 32-bit integer ALU operations for nowIago Toral Quiroga1-3/+29
2016-04-29i965/fs: handle fp64 opcodes in brw_do_channel_expressionsIago Toral Quiroga1-9/+14
2016-04-29i965/fs: add support for f2d and d2fConnor Abbott1-0/+2
2016-04-29i965/fs: add a pass for legalizing d2fConnor Abbott4-0/+80
2016-04-29i965/fs: fix dst width calculation in CSEConnor Abbott1-1/+1
2016-04-29i965/fs: fix regs_written in LOAD_PAYLOAD for doublesConnor Abbott1-2/+6
2016-04-29i965/fs: fix is_copy_payload() for doublesConnor Abbott1-2/+2
2016-04-29i965/fs: fix regs_read() for uniformsConnor Abbott1-1/+1
2016-04-29i965/fs: fix compares for doublesConnor Abbott1-3/+31
2016-04-29i965/fs: extend exec_size halving in the generatorConnor Abbott1-6/+10
2016-04-29i965/fs: generalize SIMD16 interference workaroundConnor Abbott1-19/+48
2016-04-29i965/fs: fix assign_constant_locations() for doublesConnor Abbott1-2/+4
2016-04-29i965/fs: use byte_offset() in offset() for uniformsConnor Abbott1-3/+1
2016-04-29i965/fs: handle uniforms in byte_offset()Connor Abbott1-1/+5
2016-04-29i965/fs: fix type_size() for doublesConnor Abbott1-1/+2
2016-04-29i965/fs: optimize unpack doubleIago Toral Quiroga1-4/+24
2016-04-29i965/fs: optimize pack doubleIago Toral Quiroga1-0/+23
2016-04-29i965/fs/nir: translate double pack/unpackConnor Abbott1-0/+12
2016-04-29i965/fs: add a pass for lowering PACK opcodesConnor Abbott4-0/+66
2016-04-29i965/fs: add PACK opcodeConnor Abbott5-1/+15
2016-04-29i965/fs: add a stride helperConnor Abbott1-0/+8
2016-04-29i965/fs: always pass the bitsize to brw_type_for_nir_type()Connor Abbott1-3/+9
2016-04-29i965/fs: add support for printing double immediatesConnor Abbott1-0/+3
2016-04-29i965/fs: don't propagate 64-bit immediatesConnor Abbott1-0/+2
2016-04-29i965/fs: use the NIR bit size when creating registersConnor Abbott1-8/+25
2016-04-29i965/fs: print writemask_all when it's enabledConnor Abbott1-0/+3
2016-04-29i965: fixup uniform setup for doublesConnor Abbott1-1/+6
2016-04-29i965: two-argument instructions can only use 32-bit immediatesIago Toral Quiroga1-0/+2
2016-04-29i965: fix brw_abs_immediate() for doublesIago Toral Quiroga1-2/+4
2016-04-29i965: fix brw_saturate_immediate() for doublesIago Toral Quiroga1-6/+22
2016-04-29i965: fix is_zero(), is_one() and is_negative_one() for doublesConnor Abbott1-4/+24
2016-04-29i965: fix brw_negate_immediate() for doublesConnor Abbott1-2/+4
2016-04-29i965/eu: add support for DF immediatesConnor Abbott1-7/+21
2016-04-29i965: add support for disassembling DF immediatesConnor Abbott1-1/+1
2016-04-29i965: add support for getting/setting DF immediatesConnor Abbott1-0/+25
2016-04-29i965: add brw_imm_dfConnor Abbott2-0/+10
2016-04-29i965/eu: allow doubles in math instructionsConnor Abbott1-2/+4
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-21 17:23:10 +0000