| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | [DAGCombine] Avoid INSERT_SUBVECTOR reinsertions (PR28678) | Simon Pilgrim | 2016-08-10 | 1 | -1/+10 |
| * | [DAGCombiner] Better support for shifting large value type by constants | Simon Pilgrim | 2016-08-09 | 1 | -17/+42 |
| * | [X86] Heuristic to selectively build Newton-Raphson SQRT estimation | Nikolai Bozhenov | 2016-08-04 | 1 | -2/+6 |
| * | Typo fix in comment. NFC | Diana Picus | 2016-08-04 | 1 | -1/+1 |
| * | [DAGCombine] Make sext(setcc) combine respect getBooleanContents | Michael Kuperstein | 2016-08-01 | 1 | -9/+23 |
| * | MachineFunction: Return reference for getFrameInfo(); NFC | Matthias Braun | 2016-07-28 | 1 | -3/+3 |
| * | [DAGCombiner] Use APInt directly to detect out of range shift constants | Simon Pilgrim | 2016-07-27 | 1 | -3/+3 |
| * | [SelectionDAG] Get rid of bool parameters in SelectionDAG::getLoad, getStore,... | Justin Lebar | 2016-07-15 | 1 | -106/+83 |
| * | [DAG] make isConstantSplatVector() available to the rest of lowering | Sanjay Patel | 2016-07-10 | 1 | -32/+12 |
| * | reformat, fix comments/names; NFCI | Sanjay Patel | 2016-07-10 | 1 | -27/+22 |
| * | Reapply r274829 with fix for FP vectors | Matt Arsenault | 2016-07-08 | 1 | -2/+4 |
| * | Revert r274829, it caused PR28472. | Nico Weber | 2016-07-08 | 1 | -1/+1 |
| * | Do not expand SDIV when compiling for minimum code size | Sjoerd Meijer | 2016-07-08 | 1 | -0/+5 |
| * | Addressing post-commit comments regarding not expanding UDIV; | Sjoerd Meijer | 2016-07-08 | 1 | -2/+2 |
| * | Code size optimisation: don't expand a div to a mul and and a shift sequence. | Sjoerd Meijer | 2016-07-08 | 1 | -0/+5 |
| * | Bug 28444: Fix assertion when extract_vector_elt has mismatched type | Matt Arsenault | 2016-07-08 | 1 | -1/+1 |
| * | [DAGCombiner] Fix visitSTORE to continue processing current SDNode, if findBe... | Tim Shen | 2016-07-06 | 1 | -6/+12 |
| * | Revert r259387: "AArch64: Implement missed conditional compare sequences." | Balaram Makam | 2016-07-05 | 1 | -2/+2 |
| * | DAGCombiner: Fold away vector extract of insert with the same index | Matt Arsenault | 2016-07-05 | 1 | -0/+8 |
| * | [CodeGen] Teach OR combine of shuffles involving zero vectors to better handl... | Craig Topper | 2016-07-03 | 1 | -5/+10 |
| * | [CodeGen,Target] Remove the version of DAG.getVectorShuffle that takes a poin... | Craig Topper | 2016-07-01 | 1 | -11/+9 |
| * | [DAGCombine] Teach DAG combine to handle ORs of shuffles involving zero vecto... | Craig Topper | 2016-06-29 | 1 | -45/+52 |
| * | DAGCombiner: Don't narrow volatile vector loads + extract | Matt Arsenault | 2016-06-27 | 1 | -3/+8 |
| * | [SelectionDAG] Use DAG.getCommutedVectorShuffle instead of reimplementing it. | Craig Topper | 2016-06-26 | 1 | -15/+2 |
| * | Preserve DebugInfo when replacing values in DAGCombiner | Nirav Dave | 2016-06-23 | 1 | -2/+0 |
| * | Revert r273456, "Preserve DebugInfo when replacing values in DAGCombiner" as ... | Peter Collingbourne | 2016-06-23 | 1 | -0/+2 |
| * | Preserve DebugInfo when replacing values in DAGCombiner | Nirav Dave | 2016-06-22 | 1 | -2/+0 |
| * | Replace silly uses of 'signed' with 'int' | David Majnemer | 2016-06-21 | 1 | -1/+1 |
| * | [DAG] Remove redundant FMUL in Newton-Raphson SQRT code | Sanjay Patel | 2016-06-16 | 1 | -53/+86 |
| * | Revert "Preserve DebugInfo when replacing values in DAGCombiner" | Nirav Dave | 2016-06-15 | 1 | -0/+2 |
| * | Preserve DebugInfo when replacing values in DAGCombiner | Nirav Dave | 2016-06-15 | 1 | -2/+0 |
| * | Pass DebugLoc and SDLoc by const ref. | Benjamin Kramer | 2016-06-12 | 1 | -35/+33 |
| * | Avoid copies of std::strings and APInt/APFloats where we only read from it | Benjamin Kramer | 2016-06-08 | 1 | -3/+3 |
| * | transform obscured FP sign bit ops into a fabs/fneg using TLI hook | Sanjay Patel | 2016-06-02 | 1 | -0/+46 |
| * | [DAG] use getBitcast() to reduce code | Sanjay Patel | 2016-06-02 | 1 | -33/+25 |
| * | DAGCombiner: Fix broken size check in isAlias | Matt Arsenault | 2016-06-01 | 1 | -1/+1 |
| * | [X86] Detect SAD patterns and emit psadbw instructions. | Michael Kuperstein | 2016-05-27 | 1 | -7/+11 |
| * | Simplify std::all_of predicate (to one line) by using llvm::all_of. NFCI. | Simon Pilgrim | 2016-05-25 | 1 | -3/+1 |
| * | reduce indentation; NFCI | Sanjay Patel | 2016-05-19 | 1 | -9/+7 |
| * | Fix unused variable warning. | Simon Pilgrim | 2016-05-07 | 1 | -1/+0 |
| * | [SelectionDAG] Added bitreverse(bitreverse(v)) --> v | Simon Pilgrim | 2016-05-07 | 1 | -0/+12 |
| * | DAGCombiner: Reduce truncated shl width | Matt Arsenault | 2016-04-29 | 1 | -0/+19 |
| * | [DAGCombiner] Follow coding convention for function name (NFC) | Gerolf Hoflehner | 2016-04-27 | 1 | -2/+2 |
| * | Revert r267649, it caused PR27539. | Nico Weber | 2016-04-27 | 1 | -11/+7 |
| * | Detects the SAD pattern on X86 so that much better code will be emitted once ... | Cong Hou | 2016-04-27 | 1 | -7/+11 |
| * | [CodeGen] Add getBuildVector and getSplatBuildVector helpers. NFCI. | Ahmed Bougacha | 2016-04-26 | 1 | -14/+13 |
| * | [PR27390] [CodeGen] Reject indexed loads in CombinerDAG. | Marcin Koscielnicki | 2016-04-25 | 1 | -3/+8 |
| * | [MachineCombiner] Support for floating-point FMA on ARM64 (re-commit r267098) | Gerolf Hoflehner | 2016-04-24 | 1 | -1/+10 |
| * | [CodeGen] Teach DAG combine to fold select_cc seteq X, 0, sizeof(X), ctlz_zer... | Craig Topper | 2016-04-24 | 1 | -0/+35 |
| * | DAGCombiner: Relax alignment restriction when changing store type | Matt Arsenault | 2016-04-22 | 1 | -10/+14 |
