| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | avoid redundant lookups in BBExecutable, and make it a SmallPtrSet. | Chris Lattner | 2009-11-02 | 1 | -14/+16 | |
| | | | | | llvm-svn: 85790 | |||||
| * | Use the libanalysis 'ConstantFoldLoadFromConstPtr' function | Chris Lattner | 2009-11-02 | 1 | -18/+11 | |
| | | | | | | | | instead of reinventing SCCP-specific logic. This gives us new powers. llvm-svn: 85789 | |||||
| * | switch the main 'ValueState' map from being an std::map to being | Chris Lattner | 2009-11-02 | 1 | -311/+281 | |
| | | | | | | | | | | | | a DenseMap. Doing this required being aware of subtle iterator invalidation issues, but it provides a big speedup. In a release-asserts build, this sped up optimizing 403.gcc from 1.34s -> 0.79s (IPSCCP) and 1.11s -> 0.44s (SCCP). This commit also conflates in a bunch of general cleanups, sorry. llvm-svn: 85788 | |||||
| * | fix a bug exposed by moving SRoA earlier which caused a crash building kc++ | Chris Lattner | 2009-11-02 | 1 | -1/+6 | |
| | | | | | llvm-svn: 85786 | |||||
| * | only IPSCCP incoming arguments if the function is executable, this fixes | Chris Lattner | 2009-11-02 | 1 | -15/+17 | |
| | | | | | | | an assertion on the buildbot. llvm-svn: 85784 | |||||
| * | add a new ValueState::getConstantInt() helper, use it to | Chris Lattner | 2009-11-02 | 1 | -52/+53 | |
| | | | | | | | simplify some code. llvm-svn: 85783 | |||||
| * | tidy up some more: remove some extraneous inline specifiers, return harder. | Chris Lattner | 2009-11-02 | 1 | -126/+95 | |
| | | | | | llvm-svn: 85780 | |||||
| * | eliminate the SCCPSolver::getValueMapping method. | Chris Lattner | 2009-11-02 | 1 | -25/+23 | |
| | | | | | llvm-svn: 85778 | |||||
| * | fix failures introduced in r85774 | Chris Lattner | 2009-11-02 | 1 | -1/+2 | |
| | | | | | llvm-svn: 85777 | |||||
| * | factor duplicated code into a new DeleteInstructionInBlock | Chris Lattner | 2009-11-02 | 1 | -92/+80 | |
| | | | | | | | function, eliminate temporary (and pointless) smallvector. llvm-svn: 85776 | |||||
| * | Chris used to use '...' instead of proper grammar. | Chris Lattner | 2009-11-02 | 1 | -30/+32 | |
| | | | | | llvm-svn: 85775 | |||||
| * | remove some extraneous llvmcontext stuff. | Chris Lattner | 2009-11-02 | 1 | -31/+26 | |
| | | | | | llvm-svn: 85774 | |||||
| * | change LatticeVal to use PointerIntPair to save some space. | Chris Lattner | 2009-11-02 | 1 | -49/+52 | |
| | | | | | llvm-svn: 85773 | |||||
| * | fix instcombine to only do store sinking when the alignments | Chris Lattner | 2009-11-02 | 1 | -4/+8 | |
| | | | | | | | of the two loads agree. Propagate that onto the new store. llvm-svn: 85772 | |||||
| * | when merging two loads, make sure to take the min of their alignment, | Chris Lattner | 2009-11-01 | 1 | -1/+1 | |
| | | | | | | | | not the max. This didn't matter until the previous patch because instcombine would refuse to sink loads with differenting alignments. llvm-svn: 85738 | |||||
| * | split load sinking out to its own function, like gep sinking. | Chris Lattner | 2009-11-01 | 1 | -66/+101 | |
| | | | | | llvm-svn: 85737 | |||||
| * | fix a bug noticed by inspection: when instcombine sinks loads through | Chris Lattner | 2009-11-01 | 1 | -4/+24 | |
| | | | | | | | | | phis, it didn't preserve the alignment of the load. This is a missed optimization of the alignment is high and a miscompilation when the alignment is low. llvm-svn: 85736 | |||||
| * | cleanups, switch GlobalDCE to SmallPtrSet instead of std::set | Chris Lattner | 2009-11-01 | 2 | -16/+10 | |
| | | | | | llvm-svn: 85730 | |||||
| * | remove a bunch of locking from LLVMContextImpl. Since only one thread | Chris Lattner | 2009-11-01 | 1 | -0/+3 | |
| | | | | | | | | can be banging on a context at a time, this isn't needed. Owen, please review. llvm-svn: 85728 | |||||
| * | improve comment. | Chris Lattner | 2009-11-01 | 1 | -1/+1 | |
| | | | | | llvm-svn: 85725 | |||||
| * | Reverting 85714, 85715, 85716, which are breaking the build | Douglas Gregor | 2009-11-01 | 1 | -1/+0 | |
| | | | | | llvm-svn: 85717 | |||||
| * | Remove the #include of Pass.h from PassManager.h. This breaks a significant | Dan Gohman | 2009-11-01 | 1 | -0/+1 | |
| | | | | | | | #include dependency, as frontends commonly pull in PassManager.h. llvm-svn: 85714 | |||||
| * | teach ipsccp and ipconstprop that a blockaddress doesn't 'take the address' ↵ | Chris Lattner | 2009-11-01 | 2 | -2/+7 | |
| | | | | | | | | | | of a function in a way that should prevent ip constprop. This allows clang/test/CodeGen/indirect-goto.c to pass with the new indirect goto lowering. llvm-svn: 85709 | |||||
| * | change llvm::MergeBlockIntoPredecessor to not merge two blocks BB1->BB2 | Chris Lattner | 2009-11-01 | 1 | -3/+7 | |
| | | | | | | | | | | when BB2 has its address taken. Since it ends up doing BB2->rauw(BB1), this can cause the address of the entry block to be taken. Since it is generally undesirable to nuke blocks whose address is taken, even when we can, just unconditionally stop this xform. llvm-svn: 85708 | |||||
| * | strengthen an assumption: RevectorBlockTo knows that PredBB | Chris Lattner | 2009-11-01 | 1 | -12/+14 | |
| | | | | | | | | | ended in an uncond branch because the pass requires BreakCriticalEdges. However, BCE doesn't eliminate critical adges from indbrs. llvm-svn: 85707 | |||||
| * | if CostMetrics says to never duplicate some code, don't unswitch a loop. | Chris Lattner | 2009-11-01 | 1 | -1/+2 | |
| | | | | | | | This prevents unswitching from duplicating indbr's. llvm-svn: 85705 | |||||
| * | constant fold indirectbr(blockaddress(%bb)) -> br label %bb. | Chris Lattner | 2009-11-01 | 1 | -9/+46 | |
| | | | | | llvm-svn: 85704 | |||||
| * | Revert 85678/85680. The decision is to stay with the current form of | Chris Lattner | 2009-11-01 | 2 | -8/+4 | |
| | | | | | | | | indirectbr, thus we don't need "blockaddr(@func, null)". Eliminate it for simplicity. llvm-svn: 85699 | |||||
| * | Make sure PRE doesn't split crit edges from indirectbr. | Chris Lattner | 2009-10-31 | 1 | -1/+5 | |
| | | | | | llvm-svn: 85692 | |||||
| * | llvm::SplitEdge should refuse to split an edge from an indirectbr. | Chris Lattner | 2009-10-31 | 2 | -1/+4 | |
| | | | | | | | Fix CodeGenPrepare to not try to split edges from indirectbr. llvm-svn: 85690 | |||||
| * | update the comment above llvm::SplitCriticalEdge, and make | Chris Lattner | 2009-10-31 | 1 | -6/+21 | |
| | | | | | | | it abort on IndirectBrInst as describe in the comment. llvm-svn: 85688 | |||||
| * | adjust a couple xforms to work with null bb's in BlockAddress. | Chris Lattner | 2009-10-31 | 2 | -4/+8 | |
| | | | | | llvm-svn: 85680 | |||||
| * | add a comment. | Chris Lattner | 2009-10-31 | 1 | -0/+4 | |
| | | | | | llvm-svn: 85671 | |||||
| * | Revert r85667. LoopUnroll currently can't call utility functions which | Dan Gohman | 2009-10-31 | 2 | -24/+94 | |
| | | | | | | | | auto-update the DominatorTree because it doesn't keep the DominatorTree current while it works. llvm-svn: 85670 | |||||
| * | Remove redundant code. | Dan Gohman | 2009-10-31 | 1 | -3/+0 | |
| | | | | | llvm-svn: 85668 | |||||
| * | Merge the enhancements from LoopUnroll's FoldBlockIntoPredecessor into | Dan Gohman | 2009-10-31 | 2 | -94/+24 | |
| | | | | | | | | MergeBlockIntoPredecessor. This makes SimplifyCFG slightly more aggressive, and makes it unnecessary for LoopUnroll to have its own copy of this code. llvm-svn: 85667 | |||||
| * | Rename forgetLoopBackedgeTakenCount to forgetLoop, because it | Dan Gohman | 2009-10-31 | 3 | -3/+3 | |
| | | | | | | | clears out more information than just the stored backedge taken count. llvm-svn: 85664 | |||||
| * | Replace LoopUnrollPass.cpp's custom code-size estimation code using | Dan Gohman | 2009-10-31 | 1 | -30/+5 | |
| | | | | | | | the new common CodeMetrics code. llvm-svn: 85663 | |||||
| * | Simplify this code. | Dan Gohman | 2009-10-31 | 1 | -2/+2 | |
| | | | | | llvm-svn: 85662 | |||||
| * | Remove an unnecessary #include. | Dan Gohman | 2009-10-31 | 1 | -1/+0 | |
| | | | | | llvm-svn: 85661 | |||||
| * | Update CMakeLists for recent renames. | Dan Gohman | 2009-10-31 | 2 | -2/+2 | |
| | | | | | llvm-svn: 85660 | |||||
| * | Rename UnrollLoop.cpp to LoopUnroll.cpp, and LoopUnroll.cpp to | Dan Gohman | 2009-10-31 | 2 | -0/+0 | |
| | | | | | | | | LoopUnrollPass.cpp, for consistency with other passes which are similarly split. llvm-svn: 85659 | |||||
| * | Remove CodeGenLICM. It's largely obsoleted by MachineLICM's new ability | Dan Gohman | 2009-10-31 | 2 | -113/+0 | |
| | | | | | | | to unfold loop-invariant loads. llvm-svn: 85657 | |||||
| * | Reapply r85634, with the bug fixed. | Dan Gohman | 2009-10-31 | 1 | -15/+17 | |
| | | | | | llvm-svn: 85655 | |||||
| * | Revert 85634. It's breaking consumer-typeset (and others). | Evan Cheng | 2009-10-31 | 1 | -17/+15 | |
| | | | | | llvm-svn: 85641 | |||||
| * | Add a comment about a missed opportunity. | Dan Gohman | 2009-10-30 | 1 | -0/+5 | |
| | | | | | llvm-svn: 85635 | |||||
| * | Optimize around the fact that pred_iterator is slow: instead of sorting | Dan Gohman | 2009-10-30 | 1 | -15/+17 | |
| | | | | | | | | PHI operands by the predecessor order, sort them by the order used by the first PHI in the block. This is still suffucient to expose duplicates. llvm-svn: 85634 | |||||
| * | Teach SimplifyCFG how to eliminate duplicate PHI nodes within a block. | Dan Gohman | 2009-10-30 | 1 | -0/+61 | |
| | | | | | | | | This reduces codesize on a variety of codes by 1-2% on x86-64. It also helps clean up after SSAUpdater. llvm-svn: 85626 | |||||
| * | Sort the incoming values in PHI nodes to match the predecessor order. | Dan Gohman | 2009-10-30 | 1 | -0/+19 | |
| | | | | | | | | This helps expose duplicate PHIs, which will make it easier for them to be eliminated. llvm-svn: 85623 | |||||
| * | Add option to createGVNPass to disable PRE. | Evan Cheng | 2009-10-30 | 1 | -2/+3 | |
| | | | | | llvm-svn: 85609 | |||||

