| Commit message (Expand) | Author | Age | Files | Lines |
* | BBVectorize: Enable vectorization of the fmuladd intrinsic | Hal Finkel | 2012-12-25 | 1 | -0/+28 |
* | Loosen scheduling restrictions on the PPC dcbt intrinsic | Hal Finkel | 2012-12-25 | 1 | -0/+22 |
* | Expand PPC64 atomic load and store | Hal Finkel | 2012-12-25 | 1 | -0/+20 |
* | [msan] Fix handling of vectors of pointers. | Evgeniy Stepanov | 2012-12-25 | 1 | -0/+15 |
* | [msan] Fix handling of select with vector condition. | Evgeniy Stepanov | 2012-12-25 | 1 | -0/+17 |
* | Harden test so it's not affected by changes to compare lowering. | Benjamin Kramer | 2012-12-25 | 1 | -1/+1 |
* | X86: Shave off one shuffle from the pcmpeqq sequence for SSE2 by making use o... | Benjamin Kramer | 2012-12-25 | 1 | -4/+2 |
* | X86: Custom lower <2 x i64> eq and ne when SSE41 is not available. | Benjamin Kramer | 2012-12-25 | 1 | -0/+26 |
* | Fix typo "Makre" -> "Make". | Nick Lewycky | 2012-12-24 | 1 | -6/+4 |
* | llvm/test/CodeGen/X86/fold-vex.ll: Add explicit triple. | NAKAMURA Takumi | 2012-12-24 | 1 | -1/+1 |
* | Some x86 instructions can load/store one of the operands to memory. On SSE, t... | Nadav Rotem | 2012-12-24 | 1 | -0/+16 |
* | LoopVectorizer: When checking for vectorizable types, also check | Nadav Rotem | 2012-12-24 | 1 | -0/+29 |
* | LoopVectorizer: Fix an endless loop in the code that looks for reductions. | Nadav Rotem | 2012-12-24 | 1 | -0/+44 |
* | CostModel: Change the default target-independent implementation for finding | Nadav Rotem | 2012-12-23 | 1 | -3/+3 |
* | We are not ready to estimate the cost of integer expansions based on the numb... | Nadav Rotem | 2012-12-23 | 1 | -2/+0 |
* | Loop Vectorizer: Update the cost model of scatter/gather operations and make | Nadav Rotem | 2012-12-23 | 1 | -1/+4 |
* | X86: Turn mul of <4 x i32> into pmuludq when no SSE4.1 is available. | Benjamin Kramer | 2012-12-22 | 1 | -0/+14 |
* | X86: Emit vector sext as shuffle + sra if vpmovsx is not available. | Benjamin Kramer | 2012-12-22 | 1 | -23/+96 |
* | In some cases, due to scheduling constraints we copy the EFLAGS. | Nadav Rotem | 2012-12-21 | 1 | -0/+37 |
* | [mips] Fix encoding of BAL instruction. Also, fix assembler test case which | Akira Hatanaka | 2012-12-21 | 1 | -1/+1 |
* | try to unbreak ppc buildbots. | Benjamin Kramer | 2012-12-21 | 1 | -4/+4 |
* | X86: Match pmin/pmax as a target specific dag combine. This occurs during vec... | Benjamin Kramer | 2012-12-21 | 2 | -3/+2790 |
* | R600: Expand vec4 INT <-> FP conversions | Tom Stellard | 2012-12-21 | 1 | -0/+52 |
* | [msan] Remove unreachable blocks before instrumenting a function. | Evgeniy Stepanov | 2012-12-21 | 1 | -0/+23 |
* | Improve the X86 cost model for loads and stores. | Nadav Rotem | 2012-12-21 | 2 | -2/+67 |
* | Add test case for r170674 | Reed Kotler | 2012-12-21 | 1 | -0/+29 |
* | Fix a bug in the code that checks if we can vectorize loops while using dynamic | Nadav Rotem | 2012-12-21 | 2 | -48/+110 |
* | Move these files over to the debug info directory. | Eric Christopher | 2012-12-21 | 2 | -0/+0 |
* | Revert "Adding support for llvm.arm.neon.vaddl[su].* and" | Bob Wilson | 2012-12-20 | 2 | -128/+0 |
* | LoopVectorize: Fix a bug in the scalarization of instructions. | Nadav Rotem | 2012-12-20 | 1 | -0/+48 |
* | On some ARM cpus, flags setting movs with shifter operand, i.e. lsl, lsr, asr, | Evan Cheng | 2012-12-20 | 1 | -13/+69 |
* | Change Lit error redirection to FileCheck to a more common syntax since it | Eli Bendersky | 2012-12-20 | 6 | -6/+6 |
* | Add a largish auto-generated test for the aligned bundling feature, along with | Eli Bendersky | 2012-12-20 | 1 | -0/+2674 |
* | Tests for the aligned bundling support added in r170718 | Eli Bendersky | 2012-12-20 | 13 | -0/+289 |
* | Simplify the testcase a bit. | Rafael Espindola | 2012-12-20 | 1 | -15/+4 |
* | Add a new attribute, 'noduplicate'. If a function contains a noduplicate call... | James Molloy | 2012-12-20 | 5 | -1/+168 |
* | Adding support for llvm.arm.neon.vaddl[su].* and | Renato Golin | 2012-12-20 | 2 | -0/+128 |
* | fix most of remaining issues with large frames. | Reed Kotler | 2012-12-20 | 1 | -2/+2 |
* | [mips] Use "or $r0, $r1, $zero" instead of "addu $r0, $zero, $r1" to copy | Akira Hatanaka | 2012-12-20 | 6 | -18/+18 |
* | Do not introduce vector operations in functions marked with noimplicitfloat. | Bob Wilson | 2012-12-20 | 1 | -0/+17 |
* | Split out abbreviations for the skeleton info from the rest of | Eric Christopher | 2012-12-19 | 1 | -0/+3 |
* | LLVM sdisel normalize bit extraction of the form: | Evan Cheng | 2012-12-19 | 1 | -0/+25 |
* | Remove edis - the enhanced disassembler. Fixes PR14654. | Roman Divacky | 2012-12-19 | 1 | -10/+0 |
* | Transform (x&C)>V into (x&C)!=0 where possible | Paul Redmond | 2012-12-19 | 1 | -0/+17 |
* | PowerPC: Expand VSELECT nodes. | Benjamin Kramer | 2012-12-19 | 1 | -0/+7 |
* | Make TargetLowering::getTypeConversion more resilient against odd illegal MVTs. | Benjamin Kramer | 2012-12-19 | 1 | -0/+22 |
* | [msan] Heuristically instrument unknown intrinsics. | Evgeniy Stepanov | 2012-12-19 | 1 | -2/+74 |
* | Optimized load + SIGN_EXTEND patterns in the X86 backend. | Elena Demikhovsky | 2012-12-19 | 3 | -3/+98 |
* | After reducing the size of an operation in the DAG we zero-extend the reduced | Nadav Rotem | 2012-12-19 | 1 | -0/+21 |
* | Teach SimplifySetCC that comparing AssertZext i1 against a constant 1 can be ... | Craig Topper | 2012-12-19 | 1 | -0/+15 |