aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
AgeCommit message (Expand)AuthorFilesLines
2016-09-09AMDGPU: Run LoadStoreVectorizer pass by defaultMatt Arsenault1-0/+3
2016-07-28MachineFunction: Return reference for getFrameInfo(); NFCMatthias Braun1-2/+2
2016-07-28AMDGPU: Remove analyzeImmediateMatt Arsenault1-5/+12
2016-07-12AMDGPU: Unify MOVRELSOffset and MOVRELDOffsetNicolai Haehnle1-30/+6
2016-07-09AMDGPU: Improve offset folding for register indexingMatt Arsenault1-0/+49
2016-07-05AMDGPU/SI: Remove address space query functions from AMDGPUDAGToDAGISelTom Stellard1-56/+3
2016-07-05AMDGPU/R600: Add PatFrags for selecting the correct vtx id for loadsTom Stellard1-5/+0
2016-07-04AMDGPU/SI: Remove hack for selecting < 32-bit loads to MUBUF instructionsTom Stellard1-4/+0
2016-06-24AMDGPU: Cleanup subtarget handling.Matt Arsenault1-1/+1
2016-06-22AMDGPU: Fix gcc warningsMatt Arsenault1-90/+0
2016-06-21Delete more dead code.Rafael Espindola1-32/+0
2016-06-21Delete some dead code.Rafael Espindola1-5/+0
2016-06-20Reformat blank lines.NAKAMURA Takumi1-1/+0
2016-06-20Untabify.NAKAMURA Takumi1-5/+3
2016-06-15AMDGPU: Fix MUBUF offset bugs affecting llvm.amdgcn.buffer.* intrinsicsNicolai Haehnle1-13/+30
2016-06-12Pass DebugLoc and SDLoc by const ref.Benjamin Kramer1-3/+4
2016-06-10AMDGPU/SI: Make sure to emit TargetConstant nodes when matching ds_*permuteTom Stellard1-2/+2
2016-06-09AMDGPU: Fix flat atomicsMatt Arsenault1-0/+17
2016-06-09AMDGPU: Fix i64 global cmpxchgMatt Arsenault1-6/+75
2016-05-13AMDGPU/R600: Implement memory loads from constant ASJan Vesely1-3/+10
2016-05-12SDAG: Implement Select instead of SelectImpl in AMDGPUDAGToDAGISelJustin Bogner1-49/+67
2016-05-09Fixed unused but set variable warningSimon Pilgrim1-3/+0
2016-05-05SDAG: Rename Select->SelectImpl and repurpose Select as returning voidJustin Bogner1-2/+2
2016-05-02AMDGPU: Make i64 loads/stores promote to v2i32Matt Arsenault1-55/+0
2016-04-29AMDGPU/SI: Add offset field to ds_permute/ds_bpermute instructionsTom Stellard1-2/+2
2016-04-25AMDGPU: Implement addrspacecastMatt Arsenault1-66/+0
2016-04-22AMDGPU: sext_inreg (srl x, K), vt -> bfe x, K, vt.SizeMatt Arsenault1-0/+16
2016-04-14[StructurizeCFG] Annotate branches that were treated as uniformNicolai Haehnle1-1/+3
2016-04-12AMDGPU: Add atomic_inc + atomic_dec intrinsicsMatt Arsenault1-1/+2
2016-04-07AMDGPU/SI: Implement atomic load/store for i32 and i64Jan Vesely1-12/+33
2016-03-24Fix sequence point warning. NFC.Vasileios Kalintiris1-1/+1
2016-03-23AMDGPU: Insert moves of frame index to value operandsMatt Arsenault1-0/+56
2016-03-21AMDGPU: Remove SignBitIsZero for mubuf scratch offsetsMatt Arsenault1-1/+1
2016-03-18AMDGPU: use ComplexPattern for offsets in llvm.amdgcn.buffer.load/store.formatNicolai Haehnle1-0/+79
2016-03-02AMDGPU: Simplify boolean conditional return statementsMatt Arsenault1-10/+7
2016-02-24AMDGPU: Check cheaper condition before SignBitIsZeroMatt Arsenault1-7/+6
2016-02-13AMDGPU: Cleanup includes and random macrosMatt Arsenault1-11/+4
2016-02-12AMDGPU/SI: Detect uniform branches and emit s_cbranch instructionsTom Stellard1-0/+55
2016-02-02Refactor backend diagnostics for unsupported featuresOliver Stannard1-3/+3
2016-01-28Revert r259035, it introduces a cyclic library dependencyOliver Stannard1-2/+2
2016-01-28Add backend dignostic printer for unsupported featuresOliver Stannard1-2/+2
2016-01-28Revert r258951 (and r258950), "Refactor backend diagnostics for unsupported f...NAKAMURA Takumi1-2/+2
2016-01-27Refactor backend diagnostics for unsupported featuresOliver Stannard1-2/+2
2016-01-20AMDGPU: Fix old comments that mention AMDILMatt Arsenault1-1/+1
2015-12-22 AMDGPU/SI: Use flat for global load/store when targeting HSAChangpeng Fang1-9/+17
2015-12-22Revert "AMDGPU/SI: Use flat for global load/store when targeting HSA"Rafael Espindola1-17/+9
2015-12-22AMDGPU/SI: Use flat for global load/store when targeting HSAChangpeng Fang1-9/+17
2015-12-01AMDGPU: Error on addrspacecasts that aren't actually implementedMatt Arsenault1-4/+7
2015-12-01AMDGPU/SI: Remove REGISTER_STORE/REGISTER_LOAD code which is now deadTom Stellard1-35/+0
2015-11-30AMDGPU: Rework how private buffer passed for HSAMatt Arsenault1-5/+1
generated by cgit v1.1 at 2025-10-17 20:32:12 +0000