| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| ... | |
| * | [APInt] Use lshrInPlace to replace lshr where possible | Craig Topper | 2017-04-18 | 1 | -2/+3 |
| * | [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization | Michael Zuckerman | 2017-04-16 | 1 | -0/+31 |
| * | [IR] Make paramHasAttr to use arg indices instead of attr indices | Reid Kleckner | 2017-04-14 | 1 | -1/+1 |
| * | [IR] Make getParamAttributes take argument numbers, not ArgNo+1 | Reid Kleckner | 2017-04-13 | 1 | -16/+19 |
| * | [InstCombine] Fix !prof metadata preservation for invokes | Reid Kleckner | 2017-04-13 | 1 | -18/+16 |
| * | [InstCombine] Simplify attribute code with new AttributeList::get NFC | Reid Kleckner | 2017-04-13 | 1 | -31/+20 |
| * | [IR] Take func, ret, and arg attrs separately in AttributeList::get | Reid Kleckner | 2017-04-13 | 1 | -11/+7 |
| * | [IR] Add AttributeSet to hide AttributeSetNode* again, NFC | Reid Kleckner | 2017-04-12 | 1 | -4/+3 |
| * | Reland "[IR] Make AttributeSetNode public, avoid temporary AttributeList copies" | Reid Kleckner | 2017-04-10 | 1 | -30/+19 |
| * | Revert "[IR] Make AttributeSetNode public, avoid temporary AttributeList copies" | Reid Kleckner | 2017-04-10 | 1 | -19/+30 |
| * | [IR] Make AttributeSetNode public, avoid temporary AttributeList copies | Reid Kleckner | 2017-04-10 | 1 | -30/+19 |
| * | Do not translate rint into nearbyint, but truncate it like nearbyint. | Joerg Sonnenberger | 2017-03-31 | 1 | -0/+1 |
| * | Fix the InstCombine to reserve the VP metadata and sets correct call count. | Dehao Chen | 2017-03-31 | 1 | -0/+6 |
| * | Spelling mistakes in comments. NFCI. | Simon Pilgrim | 2017-03-30 | 1 | -1/+1 |
| * | AMDGPU: Fold rcp/rsq of undef to undef | Matt Arsenault | 2017-03-24 | 1 | -2/+15 |
| * | Rename AttributeSet to AttributeList | Reid Kleckner | 2017-03-21 | 1 | -35/+35 |
| * | AMDGPU: Fold icmp/fcmp into icmp intrinsic | Matt Arsenault | 2017-03-13 | 1 | -0/+87 |
| * | Remove sometimes faulty rewrite of memcpy in instcombine. | Mikael Holmen | 2017-03-01 | 1 | -55/+12 |
| * | AMDGPU: Basic folds for fmed3 intrinsic | Matt Arsenault | 2017-02-27 | 1 | -0/+76 |
| * | AMDGPU: Replace disabled exp inputs with undef | Matt Arsenault | 2017-02-23 | 1 | -0/+28 |
| * | AMDGPU: Add replacement bfe intrinsics | Matt Arsenault | 2017-02-22 | 1 | -0/+73 |
| * | AMDGPU: Add cvt.pkrtz intrinsic | Matt Arsenault | 2017-02-22 | 1 | -0/+25 |
| * | InstCombine: Canonicalize fast fmuladd to fmul + fadd | Matt Arsenault | 2017-02-16 | 1 | -1/+14 |
| * | [AVX-512][InstCombine] Teach InstCombine to optimize 512-bit packss/packus in... | Craig Topper | 2017-02-16 | 1 | -2/+4 |
| * | [InstComobineCalls] Fix buildbot failures after r294453. | Igor Laevsky | 2017-02-08 | 1 | -1/+1 |
| * | [InstCombineCalls] Unfold element atomic memcpy instruction | Igor Laevsky | 2017-02-08 | 1 | -0/+81 |
| * | [InstCombineCalls] Remove zero length atomic memcpy intrinsics | Igor Laevsky | 2017-02-08 | 1 | -0/+6 |
| * | [InstCombine] Allow InstCombine to merge adjacent guards | Sanjoy Das | 2017-02-01 | 1 | -6/+14 |
| * | [Instcombine] Combine consecutive identical fences | Davide Italiano | 2017-01-31 | 1 | -0/+9 |
| * | [NVPTX] [InstCombine] Add llvm_unreachable to appease MSVC. | Justin Lebar | 2017-01-27 | 1 | -0/+1 |
| * | [NVPTX] Fix use-after-stack-free bug in InstCombineCalls. | Justin Lebar | 2017-01-27 | 1 | -1/+1 |
| * | [NVPTX] Upgrade NVVM intrinsics in InstCombineCalls. | Justin Lebar | 2017-01-27 | 1 | -0/+250 |
| * | Revert a couple of InstCombine/Guard checkins | Sanjoy Das | 2017-01-26 | 1 | -29/+0 |
| * | [X86] Add demanded elts support for the inputs to pclmul intrinsic | Craig Topper | 2017-01-26 | 1 | -0/+38 |
| * | [InstCombine] Canonicalize guards for NOT OR condition | Artur Pilipenko | 2017-01-25 | 1 | -0/+12 |
| * | [InstCombine][SSE] Add support for PACKSS/PACKUS constant folding | Simon Pilgrim | 2017-01-25 | 1 | -0/+94 |
| * | [InstCombine] Canonicalize guards for AND condition | Artur Pilipenko | 2017-01-25 | 1 | -0/+17 |
| * | [InstCombine] Allow InstrCombine to remove one of adjacent guards if they are... | Artur Pilipenko | 2017-01-25 | 1 | -0/+10 |
| * | [InstCombine][X86] MULDQ/MULUDQ undef -> zero | Simon Pilgrim | 2017-01-24 | 1 | -1/+1 |
| * | SimplifyLibCalls: Replace more unary libcalls with intrinsics | Matt Arsenault | 2017-01-23 | 1 | -1/+8 |
| * | [InstCombine][X86] Add MULDQ/MULUDQ constant folding support | Simon Pilgrim | 2017-01-23 | 1 | -3/+40 |
| * | [InstCombine][X86] MULDQ/MULUDQ undef -> zero | Simon Pilgrim | 2017-01-23 | 1 | -2/+2 |
| * | [InstCombine][X86] Add MULDQ/MULUDQ undef handling | Simon Pilgrim | 2017-01-20 | 1 | -0/+15 |
| * | [InstCombine] Remove unnecessary intrinsics demanded elts handling | Simon Pilgrim | 2017-01-18 | 1 | -22/+2 |
| * | [InstCombine][X86][AVX] Add DemandedElts support for VPERMILPD/VPERMILPS inst... | Simon Pilgrim | 2017-01-17 | 1 | -1/+11 |
| * | SimplifyLibCalls: Replace fabs libcalls with intrinsics | Matt Arsenault | 2017-01-17 | 1 | -0/+12 |
| * | [InstCombine][SSE] Add DemandedElts support for PSHUFB instructions | Simon Pilgrim | 2017-01-16 | 1 | -1/+11 |
| * | Make processing @llvm.assume more efficient - Add affected values to the assu... | Hal Finkel | 2017-01-11 | 1 | -0/+3 |
| * | InstCombine: Set operands instead of creating new call | Matt Arsenault | 2017-01-10 | 1 | -10/+6 |
| * | InstCombine: Fold cos(-x) -> cos(x) | Matt Arsenault | 2017-01-04 | 1 | -0/+14 |
