aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
AgeCommit message (Expand)AuthorFilesLines
2025-08-22[AMDGPU][NFC] Only include CodeGenPassBuilder.h where needed. (#154769)Ivan Kosarev1-0/+2
2025-07-31[AMDGPU] Remove `UnsafeFPMath` uses (#151079)paperchalice1-15/+8
2025-07-16[AMDGPU] Remove widen-16-bit-ops from CGP (#145483)Pierre van Houtryve1-294/+0
2025-06-03[ValueTracking] Make Depth last default arg (NFC) (#142384)Ramkumar Ramachandra1-12/+10
2025-06-02AMDGPUCodeGenPrepare.cpp - fix MSVC operator precedence warning. NFC.Simon Pilgrim1-2/+2
2025-05-29AMDGPU: Handle other fmin flavors in fract combine (#141987)Matt Arsenault1-5/+14
2025-05-29[AMDGPU] Handle CreateBinOp not returning BinaryOperator (#137791)anjenner1-1/+4
2025-05-24[AMDGPU] Remove unused includes (NFC) (#141376)Kazu Hirata1-1/+0
2025-05-16[AMDGPU] Do not promote uniform i16 operations to i32 in CGP (#140208)Pierre van Houtryve1-4/+4
2025-05-02[IRBuilder] Add versions of createInsertVector/createExtractVector that take ...Craig Topper1-2/+1
2025-05-02[AMDGPU] Check for nonnull loads feeding addrspacecast (#138184)Jay Foad1-0/+5
2025-04-24[AMDGPU] Use variadic isa<>. NFC. (#137016)Jay Foad1-1/+1
2025-04-17Re apply 130577 narrow math for and operand (#133896)Shoreshen1-0/+77
2025-04-07[NFC][LLVM][AMDGPU] Cleanup pass initialization for AMDGPU (#134410)Rahul Joshi1-3/+1
2025-04-01Revert "[AMDGPU][CodeGenPrepare] Narrow 64 bit math to 32 bit if profitable" ...Shoreshen1-84/+0
2025-04-01[AMDGPU][CodeGenPrepare] Narrow 64 bit math to 32 bit if profitable (#130577)Shoreshen1-0/+84
2025-03-31[IRBuilder] Add new overload for CreateIntrinsic (#131942)Rahul Joshi1-1/+1
2025-03-28[Analysis][NFC] Extract KnownFPClass (#133457)Tim Gymnich1-0/+1
2025-03-03[AMDGPU] Simplify conditional expressions. NFC. (#129228)Jay Foad1-2/+2
2025-01-09[AMDGPU] Rework getDivNumBits API (#119768)choikwa1-24/+24
2025-01-07[AMDGPU] Calculate getDivNumBits' AtLeast using bitwidth (#121758)choikwa1-1/+2
2024-12-13PatternMatch: migrate to CmpPredicate (#118534)Ramkumar Ramachandra1-1/+1
2024-12-12 Reapply [AMDGPU] prevent shrinking udiv/urem if either operand exceeds signe...choikwa1-10/+29
2024-12-09Revert "Reapply "[AMDGPU] prevent shrinking udiv/urem if either operand is in...Joseph Huber1-22/+6
2024-12-06Reapply "[AMDGPU] prevent shrinking udiv/urem if either operand is in… (#11...choikwa1-6/+22
2024-12-03[AMDGPU] Refine AMDGPUCodeGenPrepareImpl class. NFC. (#118461)Jay Foad1-107/+91
2024-11-28[AMDGPU] Preserve all analyses if nothing changed (#117994)Jay Foad1-1/+3
2024-11-20Revert "[AMDGPU] prevent shrinking udiv/urem if either operand is in (SignedM...Joseph Huber1-29/+13
2024-11-20[AMDGPU] prevent shrinking udiv/urem if either operand is in (SignedMax,Unsig...choikwa1-13/+29
2024-10-17[LLVM] Make more use of IRBuilder::CreateIntrinsic. NFC. (#112706)Jay Foad1-9/+5
2024-10-11[NFC] Rename `Intrinsic::getDeclaration` to `getOrInsertDeclaration` (#111752)Rahul Joshi1-7/+8
2024-07-02AMDGPU: Fix assert from wrong address space size assumption (#97267)Matt Arsenault1-1/+8
2024-06-24Revert "[IR][NFC] Update IRBuilder to use InsertPosition (#96497)"Stephen Tozer1-1/+1
2024-06-24[IR][NFC] Update IRBuilder to use InsertPosition (#96497)Stephen Tozer1-1/+1
2024-03-27[AMDGPU] Fix missing `IsExact` flag when expanding vector binary operator (#8...Shilei Tian1-0/+3
2024-03-19[AMDGCN] Use ZExt when handling indices in insertment element (#85718)Peter Rong1-2/+2
2024-03-18[RemoveDIs] Use getFirstNonPHIIt to fix crash #85472 (#85618)Orlando Cazalet-Hyams1-1/+1
2024-03-01[AMDGPU] Improve detection of non-null addrspacecast operands (#82311)Pierre van Houtryve1-0/+73
2024-02-06[AMDGPU] Use correct number of bits needed for div/rem shrinking (#80622)choikwa1-9/+12
2024-02-06[ValueTracking][NFC] Pass `SimplifyQuery` to `computeKnownFPClass` family (#8...Yingwei Zheng1-1/+2
2023-12-13[AMDGPU] Update IEEE and DX10_CLAMP for GFX12 (#75030)Piotr Sobczak1-2/+2
2023-11-30[AMDGPU] Don't create mulhi_24 in CGP (#72983)Pierre van Houtryve1-47/+13
2023-09-13AMDGPU: Avoid creating vector extracts if we aren't going to do anythingMatt Arsenault1-5/+4
2023-09-12AMDGPU: Correctly lower llvm.sqrt.f32Matt Arsenault1-9/+119
2023-08-30AMDGPU: Fix sqrt fast math flags spreading to fdiv fast math flagsMatt Arsenault1-4/+3
2023-08-23AMDGPU: Permit more rsq formation in AMDGPUCodeGenPrepareMatt Arsenault1-17/+17
2023-08-11[AMDGPU] Clear BreakPhiNodesCache in-between functionspvanhout1-0/+1
2023-08-03[AMDGPU] Break Large PHIs: Take whole PHI chains into accountpvanhout1-45/+77
2023-07-21[AMDGPU] Fix an unused variable warningKazu Hirata1-2/+1
2023-07-21AMDGPU: Fix variables only used in assertsMatt Arsenault1-4/+2
generated by cgit v1.1 at 2025-09-29 19:51:14 +0000