aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/AMDGPU/AMDGPULibCalls.cpp
AgeCommit message (Expand)AuthorFilesLines
2025-12-15Revert "[APFloat] Add exp function for APFloat::IEEESsingle using expf implem...Mehdi Amini1-1/+1
2025-12-15[APFloat] Add exp function for APFloat::IEEESsingle using expf implementation...lntue1-1/+1
2025-11-14[AMDGPU] Make use of getFunction and getMF. NFC. (#167872)Jay Foad1-1/+1
2025-07-31[AMDGPU] Remove `UnsafeFPMath` uses (#151079)paperchalice1-11/+2
2025-06-03[ValueTracking] Make Depth last default arg (NFC) (#142384)Ramkumar Ramachandra1-3/+3
2025-05-24[AMDGPU] Remove unused includes (NFC) (#141376)Kazu Hirata1-1/+0
2025-04-07AMDGPU: Disable sincos fold for constant inputs (#134579)Matt Arsenault1-0/+5
2024-12-16AMDGPU: Verify function type matches when matching libcalls (#119043)Matt Arsenault1-1/+1
2024-11-13[AMDGPU] Remove unused includes (NFC) (#116154)Kazu Hirata1-3/+0
2024-10-17[SimplifyLibCall][Attribute] Fix bug where we may keep `range` attr with inco...goldsteinn1-1/+2
2024-10-11[NFC] Rename `Intrinsic::getDeclaration` to `getOrInsertDeclaration` (#111752)Rahul Joshi1-6/+7
2024-07-17[AMDGPU] Use range-based for loops. NFC. (#99047)Jay Foad1-6/+4
2024-07-16[AMDGPU] Fix and add namespace closing comments. NFC.Jay Foad1-2/+2
2024-07-16[AMDGPU] Use using instead of typedef. NFC.Jay Foad1-1/+1
2024-07-16[AMDGPU] clang-tidy: define trivial constructors with = default. NFC.Jay Foad1-1/+1
2024-07-01Revert "AMDGPU: Use real copysign in fast pow (#97152)"Matt Arsenault1-4/+3
2024-07-01AMDGPU: Use real copysign in fast pow (#97152)Matt Arsenault1-3/+4
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
2024-06-18AMDGPU: Fix vector handling in pown libcall simplification (#95832)Matt Arsenault1-7/+3
2024-05-21AMDGPU: Emit 1/llvm.sqrt(x) instead of rsqrt calls in libcall handling (#92863)Matt Arsenault1-9/+29
2024-05-21AMDGPU: Directly emit sqrt intrinsic when folding rootn(x, 2) (#92598)Matt Arsenault1-11/+25
2024-05-20AMDGPU: Don't fold rootn(x, 1) to input for strictfp functions (#92595)Matt Arsenault1-3/+8
2024-05-20AMDGPU: Relax vector restriction for rootn libcall folds (#92594)Matt Arsenault1-7/+3
2024-05-17AMDGPU: Handle undef correctly in isKnownIntegral (#92566)Matt Arsenault1-6/+14
2024-04-18[IR][PatternMatch] Only accept poison in getSplatValue() (#89159)Nikita Popov1-3/+3
2024-03-27[FPEnv][AMDGPU] Correct AMDGPUSimplifyLibCalls handling of strictfp attribute...Kevin P. Neal1-0/+2
2024-03-19[NFC][RemoveDIs] Use iterators for insertion at various call-sites (#84736)Jeremy Morse1-4/+6
2024-02-06[ValueTracking][NFC] Pass `SimplifyQuery` to `computeKnownFPClass` family (#8...Yingwei Zheng1-4/+5
2024-01-09AMDGPU: Replace sqrt OpenCL libcalls with llvm.sqrt (#74197)Matt Arsenault1-29/+3
2023-12-18[AMDGPU][NFC] Simplify needcopysign logic (#75176)Jakub Chlanda1-2/+1
2023-12-15[llvm] Remove no-op ptr-to-ptr casts (NFC)Youngsuk Kim1-2/+1
2023-09-01AMDGPU: Handle rounding intrinsic exponents in isKnownIntegralMatt Arsenault1-0/+18
2023-09-01AMDGPU: Use pown instead of pow if known integralMatt Arsenault1-0/+31
2023-09-01AMDGPU: Use exp2 and log2 intrinsics directly for f16/f32Matt Arsenault1-8/+21
2023-09-01AMDGPU: Handle sitofp and uitofp exponents in fast pow expansionMatt Arsenault1-2/+21
2023-09-01AMDGPU: Enable assumptions in AMDGPULibCallsMatt Arsenault1-8/+19
2023-09-01AMDGPU: Turn pow libcalls into powrMatt Arsenault1-1/+16
2023-09-01AMDGPU: Simplify handling of constant vectors in libcallsMatt Arsenault1-24/+14
2023-09-01AMDGPU: Don't require all flags to expand fast powrMatt Arsenault1-5/+7
2023-08-30AMDGPU: Cleanup check for integral exponents in pow foldsMatt Arsenault1-19/+25
2023-08-28AMDGPU: Fix assertion on half typed pow with constant exponentsMatt Arsenault1-6/+2
2023-08-25AMDGPU: Fix asserting on fast f16 pownMatt Arsenault1-2/+2
2023-08-25AMDGPU: Trim dead includesMatt Arsenault1-2/+0
2023-08-16AMDGPU: Fix verifier error on splatted opencl fmin/fmax and ldexp callsMatt Arsenault1-29/+52
2023-08-15AMDGPU: Replace log libcalls with log intrinsicsMatt Arsenault1-1/+16
2023-08-14AMDGPU: Remove special case constant folding of divideMatt Arsenault1-8/+0
2023-08-14AMDGPU: Remove special case folding of sqrtMatt Arsenault1-4/+0
2023-08-14AMDGPU: Remove special case folding of fma/madMatt Arsenault1-32/+8
2023-08-14AMDGPU: Replace ldexp libcalls with intrinsicMatt Arsenault1-1/+9
generated by cgit v1.2.3 (git 2.39.1) at 2026-05-19 20:02:33 +0000