aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
AgeCommit message (Expand)AuthorFilesLines
2016-08-15AMDGPU: Don't fold subregister extracts into tied operandsMatt Arsenault1-3/+15
2016-06-30CodeGen: Use MachineInstr& in TargetInstrInfo, NFCDuncan P. N. Exon Smith1-4/+4
2016-06-24AMDGPU: Cleanup subtarget handling.Matt Arsenault1-4/+3
2016-04-25Add optimization bisect opt-in calls for AMDGPU passesAndrew Kaylor1-0/+3
2016-02-11AMDGPU: Fix passes depending on dominator tree for no reasonMatt Arsenault1-8/+2
2016-01-13AMDGPU/SI: Fix a bug in SIFoldOperandsMarek Olsak1-0/+11
2016-01-07AMDGPU/SI: Fold operands with sub-registersNicolai Haehnle1-4/+1
2015-10-21AMDGPU: Fix verifier error in SIFoldOperandsMatt Arsenault1-1/+4
2015-09-28Improved the interface of methods commuting operands, improved X86-FMA3 mem-f...Andrew Kaylor1-3/+12
2015-09-25AMDGPU: Fix recomputing dominator tree unnecessarilyMatt Arsenault1-0/+1
2015-09-10AMDGPU/SI: Fix creating v_mov_b32s without exec usesMatt Arsenault1-2/+14
2015-09-09AMDGPU/SI: Fold operands through REG_SEQUENCE instructionsTom Stellard1-0/+21
2015-08-29AMDGPU/SI: Fix some invaild assumptions when folding 64-bit immediatesTom Stellard1-1/+5
2015-08-28AMDGPU/SI: Factor operand folding code into its own functionTom Stellard1-67/+79
2015-07-13AMDGPU/SI: Select mad patterns to v_mac_f32Tom Stellard1-0/+31
2015-06-13R600 -> AMDGPU renameTom Stellard1-0/+288
generated by cgit v1.1 at 2025-10-15 19:13:19 +0000