| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| ... | |
| * | [DAGCombiner] Just call isConstOrConstSplat directly. NFCI. | Simon Pilgrim | 2016-10-19 | 1 | -8/+4 |
| * | [DAGCombine] Generalize distributeTruncateThroughAnd to work with any non-opa... | Simon Pilgrim | 2016-10-19 | 1 | -13/+9 |
| * | revert r284495: [Target] remove TargetRecip class | Sanjay Patel | 2016-10-18 | 1 | -32/+9 |
| * | [Target] remove TargetRecip class; move reciprocal estimate isel functionalit... | Sanjay Patel | 2016-10-18 | 1 | -9/+32 |
| * | [DAGCombiner] Add splatted vector support to (udiv x, (shl pow2, y)) -> x >>u... | Simon Pilgrim | 2016-10-18 | 1 | -2/+3 |
| * | Strip trailing whitespace (NFCI) | Simon Pilgrim | 2016-10-18 | 1 | -1/+1 |
| * | [DAG] make isConstOrConstSplat and isConstOrConstSplatFP more accessible; NFC | Sanjay Patel | 2016-10-17 | 1 | -38/+0 |
| * | [DAG] optimize away an arithmetic-right-shift of a 0 or -1 value | Sanjay Patel | 2016-10-17 | 1 | -0/+4 |
| * | [DAG] avoid creating illegal node when transforming negated shifted sign bit | Sanjay Patel | 2016-10-14 | 1 | -2/+3 |
| * | [DAG] add folds for negated shifted sign bit | Sanjay Patel | 2016-10-14 | 1 | -0/+13 |
| * | Fix use-after-frees | Nicolai Haehnle | 2016-10-14 | 1 | -2/+2 |
| * | [DAGCombiner] Teach createBuildVecShuffle to handle cases where input vectors... | Craig Topper | 2016-10-14 | 1 | -5/+9 |
| * | [DAG] hoist DL(N) and fix formatting; NFC | Sanjay Patel | 2016-10-13 | 1 | -24/+31 |
| * | Revert "In visitSTORE, always use FindBetterChain, rather than only when UseA... | Nirav Dave | 2016-10-13 | 1 | -120/+271 |
| * | In visitSTORE, always use FindBetterChain, rather than only when UseAA is ena... | Nirav Dave | 2016-10-13 | 1 | -271/+120 |
| * | [DAGCombiner] Add vector support to (mul (shl X, Y), Z) -> (shl (mul X, Z), Y... | Simon Pilgrim | 2016-10-13 | 1 | -7/+6 |
| * | [DAGCombiner] Add vector support to C2-(A+C1) -> (C2-C1)-A folding | Simon Pilgrim | 2016-10-13 | 1 | -5/+5 |
| * | [DAGCombiner] Add vector support to (sub -1, x) -> (xor x, -1) canonicalization | Simon Pilgrim | 2016-10-13 | 1 | -1/+12 |
| * | [DAGCombiner] Update most ADD combines to support general vector combines | Simon Pilgrim | 2016-10-12 | 1 | -12/+54 |
| * | [DAGCombiner] Do not remove the load of stored values when optimizations are ... | Konstantin Zhuravlyov | 2016-10-12 | 1 | -1/+2 |
| * | [DAG] Fix crash in build_vector -> vector_shuffle combine | Michael Kuperstein | 2016-10-11 | 1 | -0/+5 |
| * | [DAG] add fold for masked negated sign-extended bool | Sanjay Patel | 2016-10-11 | 1 | -5/+11 |
| * | [DAG] add fold for masked negated extended bool | Sanjay Patel | 2016-10-11 | 1 | -2/+15 |
| * | [DAG] simplify logic; NFC | Sanjay Patel | 2016-10-11 | 1 | -8/+6 |
| * | [DAG] hoist DL(N) and fix formatting; NFC | Sanjay Patel | 2016-10-11 | 1 | -25/+32 |
| * | [DAG] fix formatting; NFC | Sanjay Patel | 2016-10-11 | 1 | -72/+68 |
| * | [DAG] clean up foldSelectOfConstants(); NFCI | Sanjay Patel | 2016-10-07 | 1 | -19/+14 |
| * | [DAG] move fold (select C, 0, 1 -> xor C, 1) to a helper function; NFC | Sanjay Patel | 2016-10-07 | 1 | -16/+31 |
| * | Delete some dead code in SelectionDAG (NFC) | Vedant Kumar | 2016-10-06 | 1 | -4/+0 |
| * | [DAG] Generalize build_vector -> vector_shuffle combine for more than 2 inputs | Michael Kuperstein | 2016-10-06 | 1 | -135/+191 |
| * | fix formatting; NFC | Sanjay Patel | 2016-10-03 | 1 | -8/+5 |
| * | Revert "In visitSTORE, always use FindBetterChain, rather than only when UseA... | Nirav Dave | 2016-09-28 | 1 | -120/+271 |
| * | In visitSTORE, always use FindBetterChain, rather than only when UseAA is ena... | Nirav Dave | 2016-09-28 | 1 | -271/+120 |
| * | [DAG] Remove isVectorClearMaskLegal() check from vector_build dagcombine | Michael Kuperstein | 2016-09-28 | 1 | -7/+0 |
| * | Change the order of the splitted store from high - low to low - high. | Wei Mi | 2016-09-18 | 1 | -2/+2 |
| * | getValueType().getSizeInBits() -> getValueSizeInBits() ; NFCI | Sanjay Patel | 2016-09-14 | 1 | -9/+8 |
| * | getValueType().getScalarSizeInBits() -> getScalarValueSizeInBits() ; NFCI | Sanjay Patel | 2016-09-14 | 1 | -32/+24 |
| * | getScalarType().getSizeInBits() -> getScalarSizeInBits() ; NFCI | Sanjay Patel | 2016-09-14 | 1 | -32/+32 |
| * | [DAG] Allow build-to-shuffle combine to combine builds from two wide vectors. | Michael Kuperstein | 2016-09-13 | 1 | -27/+53 |
| * | [DAGCombiner] Use APInt directly in (shl (zext (srl x, C)), C) combine range ... | Simon Pilgrim | 2016-09-13 | 1 | -2/+2 |
| * | [DAGCombiner] Use APInt directly in (shl (ext (shl x, c1)), c2) combine | Simon Pilgrim | 2016-09-13 | 1 | -11/+15 |
| * | Remove MVT:i1 xor instruction before SELECT. (Performance improvement). | Ayman Musa | 2016-09-13 | 1 | -0/+16 |
| * | [DAG] Refactor BUILD_VECTOR combine to make it easier to extend. NFCI. | Michael Kuperstein | 2016-09-13 | 1 | -123/+156 |
| * | [CodeGen] Split out the notions of MI invariance and MI dereferenceability. | Justin Lebar | 2016-09-11 | 1 | -0/+2 |
| * | [SelectionDAG] Ensure DAG::getZeroExtendInReg is called with a scalar type | Simon Pilgrim | 2016-09-09 | 1 | -1/+1 |
| * | [DAGCombiner] Enable AND combines of splatted constant vectors | Simon Pilgrim | 2016-09-08 | 1 | -7/+7 |
| * | [DAGCombine] More fixups to SETCC legality checking (visitANDLike/visitORLike) | Hal Finkel | 2016-09-06 | 1 | -28/+46 |
| * | Split the store of a wide value merged from an int-fp pair into multiple stores. | Wei Mi | 2016-09-02 | 1 | -0/+103 |
| * | [DAGcombiner] Fix incorrect sinking of a truncate into the operand of a shift. | Andrea Di Biagio | 2016-09-02 | 1 | -3/+3 |
| * | [DAGCombine] Don't fold a trunc if it feeds an anyext | Michael Kuperstein | 2016-09-01 | 1 | -0/+4 |
