2011-01-03 |
Duncan Sands | Speed up instsimplify by about 10-15% by not bothering... |
tree | commitdiff |
2011-01-03 |
Cameron Zwarich | Switch a worklist in CodeGenPrepare to SmallVector... |
tree | commitdiff |
2011-01-03 |
Chris Lattner | earlycse can do trivial with-a-block dead store |
tree | commitdiff |
2011-01-03 |
Chris Lattner | switch the load table to use a recycling bump pointer... |
tree | commitdiff |
2011-01-03 |
Chris Lattner | now that loads are in their own table, we can implement |
tree | commitdiff |
2011-01-03 |
Chris Lattner | split loads and calls into separate tables. Loads... |
tree | commitdiff |
2011-01-03 |
Chris Lattner | various cleanups, no functionality change. |
tree | commitdiff |
2011-01-03 |
Chris Lattner | Teach EarlyCSE to do trivial CSE of loads and read... |
tree | commitdiff |
2011-01-03 |
Chris Lattner | rename InstValue to SimpleValue, add some comments. |
tree | commitdiff |
2011-01-03 |
Michael J. Spencer | CMake: Add missing source file. |
tree | commitdiff |
2011-01-03 |
Chris Lattner | Allocate nodes for the scoped hash table from a recylin... |
tree | commitdiff |
2011-01-03 |
Chris Lattner | reduce redundancy in the hashing code and other misc... |
tree | commitdiff |
2011-01-03 |
Cameron Zwarich | Add a new loop-instsimplify pass, with the intention... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | fix some pastos |
tree | commitdiff |
2011-01-02 |
Chris Lattner | add DEBUG and -stats output to earlycse. |
tree | commitdiff |
2011-01-02 |
Chris Lattner | Enhance earlycse to do CSE of casts, instsimplify and... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | split dom frontier handling stuff out to its own Domina... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | sketch out a new early cse pass. No functionality... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | fix a miscompilation of tramp3d-v4: when forming a... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | If a loop iterates exactly once (has backedge count... |
tree | commitdiff |
2011-01-02 |
Nick Lewycky | Also remove functions that use complex constant express... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | enhance loop idiom recognition to scan *all* unconditio... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | make inSubLoop much more efficient. |
tree | commitdiff |
2011-01-02 |
Chris Lattner | rip out isExitBlockDominatedByBlockInLoop, calling... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | add a list of opportunities for future improvement. |
tree | commitdiff |
2011-01-02 |
Duncan Sands | Fix PR8702 by not having LoopSimplify claim to preserve... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | Allow loop-idiom to run on multiple BB loops, but still... |
tree | commitdiff |
2011-01-02 |
Chris Lattner | remove debugging code. |
tree | commitdiff |
2011-01-02 |
Chris Lattner | add some -stats output. |
tree | commitdiff |
2011-01-02 |
Chris Lattner | improve loop rotation to use CodeMetrics to analyze the |
tree | commitdiff |
2011-01-02 |
Chris Lattner | teach loop idiom recognition to form memcpy's from... |
tree | commitdiff |
2011-01-02 |
Nick Lewycky | Remove functions from the FnSet when one of their calle... |
tree | commitdiff |
2011-01-01 |
Chris Lattner | fix a globalopt crash on two Adobe-C++ testcases that... |
tree | commitdiff |
2011-01-01 |
Chris Lattner | add a validity check that was missed, fixing a crash... |
tree | commitdiff |
2011-01-01 |
Chris Lattner | improve validity check to handle constant-trip-count... |
tree | commitdiff |
2011-01-01 |
Chris Lattner | implement the "no aliasing accesses in loop" safety... |
tree | commitdiff |
2010-12-31 |
Duncan Sands | Simplify this pass by using a depth-first iterator... |
tree | commitdiff |
2010-12-31 |
Duncan Sands | Zap dead instructions harder. |
tree | commitdiff |
2010-12-30 |
Benjamin Kramer | Make a bunch of symbols internal. |
tree | commitdiff |
2010-12-28 |
Chris Lattner | simplify this, isBytewiseValue handles the extra check... |
tree | commitdiff |
2010-12-28 |
Duncan Sands | Silence gcc warning about an unused variable when doing... |
tree | commitdiff |
2010-12-27 |
Chris Lattner | fix some issues Frits noticed, add AliasAnalysis as... |
tree | commitdiff |
2010-12-27 |
Benjamin Kramer | BuildLibCalls: Nuke EmitMemCpy, EmitMemMove and EmitMem... |
tree | commitdiff |
2010-12-27 |
Benjamin Kramer | SimplifyLibCalls: Use IRBuilder to simplify code. |
tree | commitdiff |
2010-12-27 |
Chris Lattner | have loop-idiom nuke instructions that feed stores... |
tree | commitdiff |
2010-12-26 |
Chris Lattner | implement enough of the memset inference algorithm... |
tree | commitdiff |
2010-12-26 |
Chris Lattner | start using irbuilder to make mem intrinsics in a few... |
tree | commitdiff |
2010-12-26 |
Chris Lattner | sketch more of this out. |
tree | commitdiff |
2010-12-26 |
Chris Lattner | move isBytewiseValue out to ValueTracking.h/cpp |
tree | commitdiff |
2010-12-26 |
Chris Lattner | actually add the file... |
tree | commitdiff |
2010-12-26 |
Chris Lattner | Start of a pass for recognizing memset and memcpy idioms. |
tree | commitdiff |
2010-12-26 |
Benjamin Kramer | Simplify code. |
tree | commitdiff |
2010-12-25 |
Chris Lattner | don't lose TD info |
tree | commitdiff |
2010-12-25 |
Chris Lattner | switch the inliner alignment enforcement stuff to use the |
tree | commitdiff |
2010-12-25 |
Chris Lattner | Move getOrEnforceKnownAlignment out of instcombine... |
tree | commitdiff |
2010-12-24 |
Benjamin Kramer | Fix a thinko pointed out by Frits van Bommel: looking... |
tree | commitdiff |
2010-12-24 |
Benjamin Kramer | MemCpyOpt: Turn memcpys from a constant into a memset... |
tree | commitdiff |
2010-12-23 |
Owen Anderson | When determining if we can fold (x >> C1) << C2, the... |
tree | commitdiff |
2010-12-23 |
Owen Anderson | It is possible for SimplifyCFG to cause PHI nodes to... |
tree | commitdiff |
2010-12-23 |
Mon P Wang | Preserve the address space when generating bitcasts... |
tree | commitdiff |
2010-12-23 |
Jeffrey Yasskin | Change all self assignments X=X to (void)X, so that... |
tree | commitdiff |
2010-12-22 |
Benjamin Kramer | InstCombine: creating selects from -1 and 0 is fine... |
tree | commitdiff |
2010-12-22 |
Duncan Sands | Add a generic expansion transform: A op (B op' C) ... |
tree | commitdiff |
2010-12-22 |
Duncan Sands | Add some statistics, good for understanding how much... |
tree | commitdiff |
2010-12-21 |
Owen Anderson | Give GVN back the ability to perform simple conditional... |
tree | commitdiff |
2010-12-21 |
Owen Anderson | Remove dead code. |
tree | commitdiff |
2010-12-21 |
Benjamin Kramer | GVN's Expression is not POD-like (it contains a SmallVe... |
tree | commitdiff |
2010-12-21 |
Duncan Sands | Visit instructions deterministically. Use a FIFO so... |
tree | commitdiff |
2010-12-21 |
Duncan Sands | If an instruction simplifies, try again to simplify... |
tree | commitdiff |
2010-12-21 |
Duncan Sands | Pull a few more simplifications out of instcombine... |
tree | commitdiff |
2010-12-20 |
Duncan Sands | Oops, forgot to add the pass itself! |
tree | commitdiff |
2010-12-20 |
Duncan Sands | Add a new convenience pass for testing InstructionSimpl... |
tree | commitdiff |
2010-12-20 |
Benjamin Kramer | Add a check missing from my last commit and avoid a... |
tree | commitdiff |
2010-12-20 |
Benjamin Kramer | Reduce indentation. |
tree | commitdiff |
2010-12-20 |
Benjamin Kramer | Teach InstCombine to merge (icmp ult (X + CA), C1)... |
tree | commitdiff |
2010-12-20 |
Chris Lattner | fix PR8807 by making transformConstExprCastCall aware... |
tree | commitdiff |
2010-12-20 |
Chris Lattner | various cleanups for transformConstExprCastCall |
tree | commitdiff |
2010-12-20 |
Chris Lattner | when eliding a byval copy due to inlining a readonly... |
tree | commitdiff |
2010-12-20 |
Chris Lattner | pull byval processing out to its own helper function. |
tree | commitdiff |
2010-12-20 |
Chris Lattner | fix PR8769, a miscompilation by inliner when inlining... |
tree | commitdiff |
2010-12-20 |
Mon P Wang | Avoid dropping the address space when InstCombine optim... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | fix an oversight caught by Frits! |
tree | commitdiff |
2010-12-19 |
Chris Lattner | tidy up |
tree | commitdiff |
2010-12-19 |
Chris Lattner | move a transformation to a more logical place, simplify... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | recognize an unsigned add with overflow idiom into... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | optimize uadd(x, cst) into a comparison when the normal |
tree | commitdiff |
2010-12-19 |
Chris Lattner | use IC.ReplaceInstUsesWith instead of a raw RAUW so... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | generalize the sadd creation code to not require that the |
tree | commitdiff |
2010-12-19 |
Chris Lattner | fix another miscompile in the llvm.sadd formation logic... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | fix a bug (possibly 8816) in the sadd forming xform... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | rework the code added in r122072 to pull it out to... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | Enhance LICM to promote alias sets whose pointers thems... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | fix PR8602, a bug in an assertion: a volatile store... |
tree | commitdiff |
2010-12-19 |
Chris Lattner | revert r122164, I'm going to go with a different approach. |
tree | commitdiff |
2010-12-19 |
Chris Lattner | first step to fixing PR8642: don't fold away empty... |
tree | commitdiff |
2010-12-18 |
Chris Lattner | simplify this a bit. |
tree | commitdiff |
2010-12-17 |
Bill Wendling | Whitespace fixes. No functionality change. |
tree | commitdiff |
2010-12-17 |
Nate Begeman | Add vector versions of some existing scalar transforms... |
tree | commitdiff |
2010-12-17 |
Owen Anderson | Reapply r121905 (automatic synthesis of @llvm.sadd... |
tree | commitdiff |
2010-12-17 |
Benjamin Kramer | SimplifyCFG: Ranges can be larger than 64 bits. Fixes... |
tree | commitdiff |
next |