oota-llvm.git
2009-11-02 Chris LattnerChris used to use '...' instead of proper grammar.
2009-11-02 Chris Lattnerremove some extraneous llvmcontext stuff.
2009-11-02 Chris Lattnerchange LatticeVal to use PointerIntPair to save some...
2009-11-02 Chris Lattnerfix instcombine to only do store sinking when the align...
2009-11-02 Chris Lattnermerge a test into store.ll
2009-11-02 Chris Lattnerconvert to filecheck
2009-11-02 Bill WendlingAdd missing end-tag.
2009-11-02 Bill WendlingSome formatting changes.
2009-11-02 Anton KorobeynikovHandle splats of undefs properly. This includes the...
2009-11-02 Anton KorobeynikovDo not infer the target type for COPY_TO_REGCLASS from...
2009-11-02 Anton Korobeynikov64-bit FP loads & stores operate on both NEON and VFP...
2009-11-02 Anton KorobeynikovUse NEON reg-reg moves, where profitable. This reduces...
2009-11-01 Evan ChengAdd PseudoSourceValue::mayAlias. It returns true if...
2009-11-01 Nick LewyckyLine this up as well.
2009-11-01 Nick LewyckyFix whitespace.
2009-11-01 Evan ChengFix a couple more places where we are creating ld ...
2009-11-01 Evan ChengMake use of imm12 version of Thumb2 ldr / str instructi...
2009-11-01 Chris Lattnerfix two strange things in the default passmgr:
2009-11-01 Chris Lattnermerge phi-merge.ll into phi.ll
2009-11-01 Chris Lattnerwhen merging two loads, make sure to take the min of...
2009-11-01 Chris Lattnersplit load sinking out to its own function, like gep...
2009-11-01 Chris Lattnerfix a bug noticed by inspection: when instcombine sinks...
2009-11-01 Chris LattnerIPSCCP apparently is not a superset of IPCP, this is...
2009-11-01 Chris Lattnerconvert to filecheck.
2009-11-01 Duncan SandsImprove the other instance of the comment.
2009-11-01 Duncan SandsAdd a missing closing parenthesis, and tweak to fit...
2009-11-01 Chris Lattneronly run GlobalDCE at -O3 and run it late instead of...
2009-11-01 Chris Lattnercleanups, switch GlobalDCE to SmallPtrSet instead of...
2009-11-01 Chris LattnerWe currently only run ipsccp at LTO time, which is...
2009-11-01 Chris Lattnerremove a bunch of locking from LLVMContextImpl. Since...
2009-11-01 Chris Lattnerimprove comment.
2009-11-01 Chris Lattneradd a comment about why we don't allow inlining indbr.
2009-11-01 Evan ChengFix tests.
2009-11-01 Chris Lattnerthe verifier shouldn't modify the IR.
2009-11-01 Douglas GregorReverting 85714, 85715, 85716, which are breaking the...
2009-11-01 Dan GohmanAdd a function to Passes.h to allow clients to create...
2009-11-01 Dan GohmanDon't #include Pass.h from CallGraph.h.
2009-11-01 Dan GohmanRemove the #include of Pass.h from PassManager.h. This...
2009-11-01 Chris Lattnerteach ipsccp and ipconstprop that a blockaddress doesn...
2009-11-01 Chris Lattnerchange llvm::MergeBlockIntoPredecessor to not merge...
2009-11-01 Chris Lattnerstrengthen an assumption: RevectorBlockTo knows that...
2009-11-01 Chris Lattnerfix an issue where the verifier would reject a function...
2009-11-01 Chris Lattnerif CostMetrics says to never duplicate some code, don...
2009-11-01 Chris Lattnerconstant fold indirectbr(blockaddress(%bb)) -> br label...
2009-11-01 Chris Lattnerimprove x86 codegen support for blockaddress. We now...
2009-11-01 Chris Lattnerpull check for return inst out of loop, never inline...
2009-11-01 Chris LattnerFix BlockAddress::replaceUsesOfWithOnConstant to correctly
2009-11-01 Chris Lattnerimplement linker support for BlockAddress.
2009-11-01 Chris LattnerRevert 85678/85680. The decision is to stay with the...
2009-10-31 Evan ChengUse cbz and cbnz instructions.
2009-10-31 Jim Grosbachvml[as].f32 cause stalls in following advanced SIMD...
2009-10-31 Jim GrosbachConsolidate test files
2009-10-31 Jim GrosbachChange to use FileCheck
2009-10-31 Jim GrosbachMake tests more explicit about which instructions are...
2009-10-31 Jim GrosbachGrammar tweak to comments
2009-10-31 Chris LattnerMake sure PRE doesn't split crit edges from indirectbr.
2009-10-31 Jim GrosbachUpdate test to be more explicit about what instruction...
2009-10-31 Chris Lattnerllvm::SplitEdge should refuse to split an edge from...
2009-10-31 Jim GrosbachUpdate test to be more explicit about what instruction...
2009-10-31 Chris Lattnerupdate the comment above llvm::SplitCriticalEdge, and...
2009-10-31 Jim GrosbachExpand 64-bit logical shift right inline
2009-10-31 Jim GrosbachExpand 64-bit arithmetic shift right inline
2009-10-31 Dan GohmanFix a missing newline in the dwarf output code.
2009-10-31 Dan GohmanMake -print-machineinstrs more readable.
2009-10-31 Dan GohmanFactor out more code into addCommonCodeGenPasses. The...
2009-10-31 Chris Lattneradjust a couple xforms to work with null bb's in BlockA...
2009-10-31 Chris LattnerMake blockaddress(@func, null) be valid, and make ...
2009-10-31 Benjamin KramerForce triple; darwin's ASM syntax differs from linux's.
2009-10-31 Jim GrosbachExpand 64 bit left shift inline rather than using the...
2009-10-31 Benjamin KramerAdd missing colons for FileCheck.
2009-10-31 Jim GrosbachConvert to FileCheck
2009-10-31 Jim GrosbachThe universal SDKROOT should only be assigned when...
2009-10-31 Chris Lattneradd a comment.
2009-10-31 Dan GohmanRevert r85667. LoopUnroll currently can't call utility...
2009-10-31 Dan GohmanRemove redundant code.
2009-10-31 Dan GohmanMerge the enhancements from LoopUnroll's FoldBlockIntoP...
2009-10-31 Dan GohmanRename forgetLoopBackedgeTakenCount to forgetLoop,...
2009-10-31 Dan GohmanReplace LoopUnrollPass.cpp's custom code-size estimatio...
2009-10-31 Dan GohmanSimplify this code.
2009-10-31 Dan GohmanRemove an unnecessary #include.
2009-10-31 Dan GohmanUpdate CMakeLists for recent renames.
2009-10-31 Dan GohmanRename UnrollLoop.cpp to LoopUnroll.cpp, and LoopUnroll...
2009-10-31 Dan GohmanRemove CodeGenLICM. It's largely obsoleted by MachineLI...
2009-10-31 Dan GohmanMake ScalarEvolutionAliasAnalysis slightly more aggress...
2009-10-31 Dan GohmanReapply r85634, with the bug fixed.
2009-10-31 Dan GohmanWhen discarding SrcValue information, discard all of...
2009-10-31 Dan GohmanFix 80-column violation.
2009-10-31 Eric ChristopherFix warning with gcc-4.0 and signed/unsigned.
2009-10-31 Evan ChengIt's safe to remat t2LDRpci; Add PseudoSourceValue...
2009-10-31 Evan ChengRevert 85634. It's breaking consumer-typeset (and others).
2009-10-31 Dan GohmanAdd a target triple so that this test behaves consisten...
2009-10-30 Dan GohmanAdd assertion checks here to turn silent miscompiles...
2009-10-30 Dan GohmanDon't mark registers dead here when processing nodes...
2009-10-30 Dan GohmanFix the -mattr line for this test so that it passes...
2009-10-30 Dan GohmanAdd a testcase for the recent duplicate PHI elimination...
2009-10-30 Dan GohmanAdd a comment about a missed opportunity.
2009-10-30 Dan GohmanOptimize around the fact that pred_iterator is slow...
2009-10-30 Kevin EnderbyUpdates to the ARM target assembler for llvm-mc per...
2009-10-30 Devang PatelIf string field is empty then return NULL.
2009-10-30 Chris Lattnerif basic blocks are destroyed while there are *just...
next