| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | Fix spelling error | Chris Lattner | 2002-11-08 | 1 | -1/+1 | |
| | | | | | llvm-svn: 4645 | |||||
| * | Instcombine this away: | Chris Lattner | 2002-11-04 | 1 | -1/+7 | |
| | | | | | | | | %A = getelementptr [4 x int]* %M, long 0, long 0 %B = getelementptr int* %A, long 2 llvm-svn: 4536 | |||||
| * | Add a transformation to turn: | Chris Lattner | 2002-11-04 | 1 | -0/+36 | |
| | | | | | | | | | malloc Ty, C int malloc [C x Ty], 1 llvm-svn: 4534 | |||||
| * | Make sure to _delete_ memory allocated by worklists | Chris Lattner | 2002-11-04 | 1 | -0/+2 | |
| | | | | | llvm-svn: 4523 | |||||
| * | Fixes to the reassociate pass to make it respect dominance properties | Chris Lattner | 2002-10-31 | 1 | -57/+54 | |
| | | | | | | | | Huge thanks go to Casey Carter for writing this fix, reassociate is now reoperational! llvm-svn: 4471 | |||||
| * | BreakCriticalEdges should update dominance frontier information as well as | Chris Lattner | 2002-10-31 | 1 | -2/+14 | |
| | | | | | | | other dominance stuff. Patch contributed by Casey Carter llvm-svn: 4457 | |||||
| * | Fix spelling of `propagate'. | Misha Brukman | 2002-10-29 | 3 | -48/+48 | |
| | | | | | llvm-svn: 4423 | |||||
| * | - Rename AnalysisUsage::preservesAll to getPreservesAll & preservesCFG to | Chris Lattner | 2002-10-21 | 9 | -10/+10 | |
| | | | | | | | setPreservesCFG to be less confusing. llvm-svn: 4255 | |||||
| * | - Fix bug: cee/2002-10-07-NoImmediateDominator.ll | Chris Lattner | 2002-10-08 | 1 | -1/+2 | |
| | | | | | llvm-svn: 4081 | |||||
| * | - Checkin LARGE number of Changes to CEE pass that will make it much more | Chris Lattner | 2002-10-08 | 1 | -53/+394 | |
| | | | | | | | | | | | | powerful, but that are largely disabled. The basic idea here is that it is trying to forward branches across basic blocks that have PHI nodes in it, which are crucial to be able to handle cases like whet.ll. Unfortunately we are not updating SSA correctly, causing sim.c to die, and I don't have time to fix the regression now, so I must disable the functionality. llvm-svn: 4077 | |||||
| * | Expose isCriticalEdge & SplitCriticalEdge methods from crit-edges pass | Chris Lattner | 2002-10-08 | 1 | -9/+10 | |
| | | | | | llvm-svn: 4075 | |||||
| * | It is illegal for PHI nodes to have zero values, delete the code to handle them | Chris Lattner | 2002-10-08 | 1 | -2/+0 | |
| | | | | | llvm-svn: 4071 | |||||
| * | Fold ashr -1, X into -1 | Chris Lattner | 2002-10-08 | 1 | -0/+6 | |
| | | | | | llvm-svn: 4070 | |||||
| * | Updates to work with recent Statistic's changes: | Chris Lattner | 2002-10-01 | 16 | -58/+54 | |
| | | | | | | | | | | | | | * Renamed StatisticReporter.h/cpp to Statistic.h/cpp * Broke constructor to take two const char * arguments instead of one, so that indendation can be taken care of automatically. * Sort the list by pass name when printing * Make sure to print all statistics as a group, instead of randomly when the statistics dtors are called. * Updated ProgrammersManual with new semantics. llvm-svn: 4002 | |||||
| * | Fix bug in LICM that caused the previous big win. :( | Chris Lattner | 2002-09-29 | 1 | -13/+16 | |
| | | | | | llvm-svn: 3980 | |||||
| * | Hoist the contents of Loops in depth first order in the dominator tree, | Chris Lattner | 2002-09-29 | 1 | -6/+36 | |
| | | | | | | | | | | rather than in random order. This causes LICM to be DRAMATICALLY more effective. For example, on bzip2.c, it is able to hoist 302 loads and 2380 total instructions, as opposed to 44/338 before. This obviously makes other transformations much more powerful as well! llvm-svn: 3978 | |||||
| * | Fix major bugs in dominator set & tree information updating | Chris Lattner | 2002-09-29 | 1 | -6/+9 | |
| | | | | | llvm-svn: 3975 | |||||
| * | - Further cleanups of LICM pass, remove extra work from previous implementation | Chris Lattner | 2002-09-26 | 1 | -59/+25 | |
| | | | | | | | - Do not clone instructions then insert clone outside of loop. Just move them. llvm-svn: 3951 | |||||
| * | Improve comments, doxygenize more | Chris Lattner | 2002-09-26 | 1 | -20/+36 | |
| | | | | | llvm-svn: 3950 | |||||
| * | Clean up LICM significantly now that it is guaranteed to have loop preheaders | Chris Lattner | 2002-09-26 | 1 | -99/+14 | |
| | | | | | llvm-svn: 3947 | |||||
| * | Change pass name to something sane | Chris Lattner | 2002-09-26 | 1 | -2/+1 | |
| | | | | | llvm-svn: 3946 | |||||
| * | Loop invariant code motion now depends on the LoopPreheader pass. Dead code | Chris Lattner | 2002-09-26 | 1 | -5/+10 | |
| | | | | | | | has not yet been removed. llvm-svn: 3945 | |||||
| * | - Cleanup break-crit-edges pass by making SplitCriticalEdge a member method. | Chris Lattner | 2002-09-26 | 1 | -29/+30 | |
| | | | | | | | - break-crit-edges pass does not invalidate loop-preheader pass. llvm-svn: 3944 | |||||
| * | Checkin new loop-preheader insertion pass. | Chris Lattner | 2002-09-26 | 1 | -0/+189 | |
| | | | | | llvm-svn: 3942 | |||||
| * | Change LowerAllocations pass to 'require' TargetData instead of it being | Chris Lattner | 2002-09-25 | 1 | -28/+28 | |
| | | | | | | | passed in. llvm-svn: 3929 | |||||
| * | Eliminate extraneous #include | Chris Lattner | 2002-09-24 | 1 | -1/+0 | |
| | | | | | llvm-svn: 3916 | |||||
| * | - Do not expose Critical Edge breaking mechanics outside the BCE pass, thus | Chris Lattner | 2002-09-24 | 1 | -9/+94 | |
| | | | | | | | removing it from Transforms/Local.h and from Transforms/Utils/* llvm-svn: 3910 | |||||
| * | Correlated Exprs pass now requires BCE pass instead of doing it manually | Chris Lattner | 2002-09-24 | 1 | -20/+1 | |
| | | | | | llvm-svn: 3908 | |||||
| * | - Expose passinfo from BreakCriticalEdges pass so that it may be "Required" | Chris Lattner | 2002-09-24 | 1 | -14/+17 | |
| | | | | | | | by other passes. Make BCE pass be in anonymous namespace now. llvm-svn: 3907 | |||||
| * | Add new BreakCriticalEdges pass | Chris Lattner | 2002-09-24 | 1 | -0/+52 | |
| | | | | | llvm-svn: 3903 | |||||
| * | Optimize away cases like: | Chris Lattner | 2002-09-23 | 1 | -0/+6 | |
| | | | | | | | | | %cast109 = cast uint %cast212 to short ; <short> [#uses=1] %cast214 = cast short %cast109 to uint ; <uint> [#uses=1] %cast215 = cast uint %cast214 to short ; <short> [#uses=1] llvm-svn: 3897 | |||||
| * | * Fix bug: CorrelatedExprs/2002-09-23-PHIUpdateBug.ll | Chris Lattner | 2002-09-23 | 1 | -5/+20 | |
| | | | | | | | * Make sure "Changed" is updated correctly llvm-svn: 3891 | |||||
| * | Fix bug: test/Regression/Transforms/InstCombine/2002-09-17-GetElementPtrCrash.ll | Chris Lattner | 2002-09-17 | 1 | -1/+2 | |
| | | | | | | | Thanks to Casey for finding it! llvm-svn: 3783 | |||||
| * | Extract most of the transformation into an externally accessible | Vikram S. Adve | 2002-09-16 | 1 | -27/+31 | |
| | | | | | | | | function -- DecomposeArrayRef(GetElementPtrInst* GEP) -- that can be invoked on a single instruction at a time. llvm-svn: 3755 | |||||
| * | Function.h is unnecessary when Module.h is included. | Misha Brukman | 2002-09-14 | 1 | -1/+0 | |
| | | | | | llvm-svn: 3716 | |||||
| * | - Change getelementptr instruction to use long indexes instead of uint | Chris Lattner | 2002-09-11 | 2 | -4/+4 | |
| | | | | | | | indexes for sequential types. llvm-svn: 3682 | |||||
| * | Fix the last of the silly bugs I just introduced. :( | Chris Lattner | 2002-09-10 | 1 | -3/+3 | |
| | | | | | llvm-svn: 3674 | |||||
| * | Add cannonicalization of shl X, 1 -> add X, X | Chris Lattner | 2002-09-10 | 1 | -4/+12 | |
| | | | | | llvm-svn: 3671 | |||||
| * | Clean up code due to auto-insert constructors | Chris Lattner | 2002-09-10 | 4 | -32/+15 | |
| | | | | | llvm-svn: 3665 | |||||
| * | Simplify code (somtimes dramatically), by using the new "auto-insert" feature | Chris Lattner | 2002-09-10 | 3 | -73/+38 | |
| | | | | | | | of instruction constructors. llvm-svn: 3656 | |||||
| * | * Clean up code a little bit | Chris Lattner | 2002-09-10 | 1 | -19/+27 | |
| | | | | | | | * Fix bug: test/Regression/Transforms/IndVarsSimplify/2002-09-09-PointerIndVar.ll llvm-svn: 3644 | |||||
| * | Hack unneccesary now that shifts of pointers are no longer legal! | Chris Lattner | 2002-09-10 | 1 | -1/+0 | |
| | | | | | llvm-svn: 3640 | |||||
| * | Fix Bug: test/Regression/Transforms/InstCombine/2002-09-08-PointerShiftBug.ll | Chris Lattner | 2002-09-08 | 1 | -0/+1 | |
| | | | | | llvm-svn: 3626 | |||||
| * | Minor simplification | Chris Lattner | 2002-09-08 | 1 | -6/+5 | |
| | | | | | llvm-svn: 3619 | |||||
| * | Initial checkin of Correlated Expression Elimination Pass | Chris Lattner | 2002-09-06 | 1 | -0/+974 | |
| | | | | | llvm-svn: 3599 | |||||
| * | - Renamed Type::isIntegral() to Type::isInteger() | Chris Lattner | 2002-09-03 | 1 | -11/+4 | |
| | | | | | | | | - Added new method Type::isIntegral() that is the same as isInteger, but also accepts bool. llvm-svn: 3574 | |||||
| * | Add constant prop & DIE to InstCombine, so it cleans up after itself | Chris Lattner | 2002-09-02 | 1 | -9/+45 | |
| | | | | | llvm-svn: 3568 | |||||
| * | - Implement SCCP of getelementptr instructions | Chris Lattner | 2002-08-30 | 1 | -5/+31 | |
| | | | | | | | | - Implement SCCP of globals into ConstantExprs hopefully opening new opportunities for more SCCP. llvm-svn: 3555 | |||||
| * | Really minor cleanups | Chris Lattner | 2002-08-30 | 1 | -9/+4 | |
| | | | | | llvm-svn: 3549 | |||||
| * | - GCSE now no longer counts instructions not removed (due to no common | Chris Lattner | 2002-08-30 | 1 | -257/+108 | |
| | | | | | | | | | dominator as being removed) - GCSE now uses new Value #'ing interface, instead of dealing with AA itself - GCSE worklist implementation much simpler, class cleaned up. llvm-svn: 3533 | |||||

