2008-02-19 |
Owen Anderson | Factor the profitability check for return slot optimiza... |
tree | commitdiff |
2008-02-19 |
Owen Anderson | An sret parameter is required to be the first parameter... |
tree | commitdiff |
2008-02-19 |
Owen Anderson | Cleanup some of my patches from yesterday. Refactor... |
tree | commitdiff |
2008-02-19 |
Owen Anderson | Fix Transforms/GVN/memcpy.ll, which Chris broke in... |
tree | commitdiff |
2008-02-18 |
Nick Lewycky | Correctly fold divide-by-constant, even when faced... |
tree | commitdiff |
2008-02-18 |
Chris Lattner | Transforming -A + -B --> -(A + B) isn't safe for... |
tree | commitdiff |
2008-02-18 |
Chris Lattner | minor code simplification, no functionality change. |
tree | commitdiff |
2008-02-18 |
Duncan Sands | Simplify caller updating using a CallSite, as |
tree | commitdiff |
2008-02-18 |
Owen Anderson | Add support to GVN for performing sret return slot... |
tree | commitdiff |
2008-02-18 |
Chris Lattner | switch simplifycfg from using vectors for most things... |
tree | commitdiff |
2008-02-18 |
Chris Lattner | optimize away stackrestore calls that have no interveni... |
tree | commitdiff |
2008-02-18 |
Chris Lattner | simplify some code, BreakUpSubtract always returns... |
tree | commitdiff |
2008-02-17 |
Chris Lattner | Fold (-x + -y) -> -(x+y) which promotes better associat... |
tree | commitdiff |
2008-02-17 |
Chris Lattner | fix pasto |
tree | commitdiff |
2008-02-17 |
Chris Lattner | Split up subtracts into add+negate if they have a reass... |
tree | commitdiff |
2008-02-17 |
Chris Lattner | make the logic for breaking up subtracts more explicit, no |
tree | commitdiff |
2008-02-16 |
Duncan Sands | Remove any 'nest' parameter attributes if the function |
tree | commitdiff |
2008-02-15 |
Devang Patel | Fix PR2028 |
tree | commitdiff |
2008-02-14 |
Devang Patel | If loop header is also loop exiting block then OrigPN... |
tree | commitdiff |
2008-02-14 |
Chris Lattner | Fix PR2029 |
tree | commitdiff |
2008-02-14 |
Chris Lattner | simplify code, no functionality change. |
tree | commitdiff |
2008-02-14 |
Nick Lewycky | Fix PR2032. Inform the alias analysis of changes to... |
tree | commitdiff |
2008-02-13 |
Devang Patel | A loop latch phi node may have uses inside loop, not... |
tree | commitdiff |
2008-02-13 |
Dan Gohman | Rename APInt's isPositive to isNonNegative, to reflect... |
tree | commitdiff |
2008-02-13 |
Devang Patel | While moving exit condition, do not drop loop latch... |
tree | commitdiff |
2008-02-13 |
Devang Patel | Keep track of exit value operand number when operands... |
tree | commitdiff |
2008-02-13 |
Chris Lattner | remove some dead code. |
tree | commitdiff |
2008-02-12 |
Owen Anderson | Re-apply the patch to improve the optimizations of... |
tree | commitdiff |
2008-02-12 |
Eli Friedman | Fix for bug 1996: optimize out loads of undef. This... |
tree | commitdiff |
2008-02-10 |
Chris Lattner | Fix scalarrepl to not 'miscompile' undefined code,... |
tree | commitdiff |
2008-02-08 |
Devang Patel | Fix PR 1995. |
tree | commitdiff |
2008-02-06 |
Bill Wendling | Temporarily reverting: |
tree | commitdiff |
2008-02-05 |
Chris Lattner | Make RenamePass faster by making the 'is this a new... |
tree | commitdiff |
2008-02-05 |
Chris Lattner | Fix a bug compiling PR1978 (perhaps not the only one... |
tree | commitdiff |
2008-02-04 |
Owen Anderson | Be more precise when eliminating pointers bue to memcpy... |
tree | commitdiff |
2008-02-04 |
Owen Anderson | Allow GVN to hack on memcpy's, making them open to... |
tree | commitdiff |
2008-02-03 |
Nick Lewycky | There are some cases where icmp(add) can be folded... |
tree | commitdiff |
2008-02-03 |
Nick Lewycky | Hack on vectors too. |
tree | commitdiff |
2008-02-03 |
Nick Lewycky | Fold away one multiply in instcombine. This would norma... |
tree | commitdiff |
2008-02-01 |
Duncan Sands | Don't drop function/call return attributes like 'nounwind'. |
tree | commitdiff |
2008-01-30 |
Nick Lewycky | Remove a couple more cases of "getNumUses() == 0".... |
tree | commitdiff |
2008-01-30 |
Nick Lewycky | Use empty() instead of comparing size() with zero. |
tree | commitdiff |
2008-01-30 |
Nick Lewycky | Don't DCE FreeInst's. We were using those! Patch from... |
tree | commitdiff |
2008-01-30 |
Owen Anderson | Make DSE much more aggressive by performing DCE earlier... |
tree | commitdiff |
2008-01-30 |
Chris Lattner | Fix a bug where scalarrepl would discard offset if... |
tree | commitdiff |
2008-01-29 |
Chris Lattner | Don't let globalopt hack on volatile loads or stores. |
tree | commitdiff |
2008-01-29 |
Dan Gohman | Use empty() instead of comparing size() with zero. |
tree | commitdiff |
2008-01-29 |
Chris Lattner | eliminate additions of 0.0 when they are obviously... |
tree | commitdiff |
2008-01-29 |
Owen Anderson | Add support for eliminating memcpy's at the end of... |
tree | commitdiff |
2008-01-29 |
Devang Patel | Filter loops that subtract induction variables. |
tree | commitdiff |
2008-01-28 |
Chris Lattner | Transform calls to memcpy into llvm.memcpy calls, patch... |
tree | commitdiff |
2008-01-28 |
Nick Lewycky | Handle some more combinations of extend and icmp. Fixes... |
tree | commitdiff |
2008-01-28 |
Chris Lattner | Fix PR1932 by disabling an xform invalid for fdiv. |
tree | commitdiff |
2008-01-28 |
Chris Lattner | Fix PR1938 by forcing the code that uses an undefined... |
tree | commitdiff |
2008-01-27 |
Nick Lewycky | Be more careful modifying the use_list while also itera... |
tree | commitdiff |
2008-01-27 |
Duncan Sands | Revert r46393: readonly/readnone functions are no |
tree | commitdiff |
2008-01-27 |
Bill Wendling | The CorrelatedExpressionElimination pass is known to... |
tree | commitdiff |
2008-01-27 |
Chris Lattner | Fold fptrunc(add (fpextend x), (fpextend y)) -> add... |
tree | commitdiff |
2008-01-26 |
Bill Wendling | If there are no machine instructions emitted for a... |
tree | commitdiff |
2008-01-26 |
Duncan Sands | Create an explicit copy for byval parameters even |
tree | commitdiff |
2008-01-26 |
Bill Wendling | If we have a function like this: |
tree | commitdiff |
2008-01-25 |
Duncan Sands | Do this more neatly. |
tree | commitdiff |
2008-01-25 |
Owen Anderson | DeadStoreElimination can treat byval parameters as... |
tree | commitdiff |
2008-01-22 |
Nick Lewycky | Enable the fix I just checked in, silly me. |
tree | commitdiff |
2008-01-22 |
Nick Lewycky | Multiply can be evaluated in a different type, so long... |
tree | commitdiff |
2008-01-20 |
Duncan Sands | Make sure the caller doesn't use freed memory. |
tree | commitdiff |
2008-01-20 |
Duncan Sands | Initializing an unsigned with ~0UL causes the compiler |
tree | commitdiff |
2008-01-17 |
Evan Cheng | DAE bug fix. Don't lose parameter attributes on vararg... |
tree | commitdiff |
2008-01-17 |
Chris Lattner | Fix arg promotion to propagate the correct attrs on... |
tree | commitdiff |
2008-01-16 |
Chris Lattner | Handle attribute(used) global variables that are i8. |
tree | commitdiff |
2008-01-16 |
Devang Patel | Do not strip llvm.used values. |
tree | commitdiff |
2008-01-14 |
Duncan Sands | I noticed that the trampoline straightening transformat... |
tree | commitdiff |
2008-01-14 |
Chris Lattner | Fix the miscompilation of MiBench/consumer-lame that... |
tree | commitdiff |
2008-01-14 |
Chris Lattner | The isNotSuitableForSRA property is now dead, don't... |
tree | commitdiff |
2008-01-14 |
Chris Lattner | Change SRAGlobal to not depend on isNotSuitableForSRA... |
tree | commitdiff |
2008-01-14 |
Chris Lattner | Make the 'shrink global to bool' optimization more... |
tree | commitdiff |
2008-01-14 |
Chris Lattner | Turn a memcpy from a double* into a load/store of doubl... |
tree | commitdiff |
2008-01-13 |
Chris Lattner | factor memcpy/memmove simplification out to its own... |
tree | commitdiff |
2008-01-13 |
Chris Lattner | simplify some code. If we can infer alignment for... |
tree | commitdiff |
2008-01-13 |
Chris Lattner | simplify some code by adding a InsertBitCastBefore... |
tree | commitdiff |
2008-01-13 |
Chris Lattner | Fix PR1907, a nasty miscompilation because instcombine... |
tree | commitdiff |
2008-01-13 |
Duncan Sands | Small simplification. |
tree | commitdiff |
2008-01-13 |
Duncan Sands | When turning a call to a bitcast function into a direct... |
tree | commitdiff |
2008-01-12 |
Chris Lattner | we don't have to make an explicit copy of a byval argum... |
tree | commitdiff |
2008-01-12 |
Chris Lattner | Allow clients to specify the inline threshold when... |
tree | commitdiff |
2008-01-11 |
Duncan Sands | When DAE drops the varargs part of a function, ensure any |
tree | commitdiff |
2008-01-11 |
Chris Lattner | Teach argpromote to ruthlessly hack small byval structs... |
tree | commitdiff |
2008-01-11 |
Chris Lattner | Use smallptrset instead of std::set for efficiency. |
tree | commitdiff |
2008-01-11 |
Chris Lattner | a byval argument is guaranteed to be valid to load. |
tree | commitdiff |
2008-01-11 |
Chris Lattner | Update this code to use eraseFromParent where possible... |
tree | commitdiff |
2008-01-11 |
Chris Lattner | replace a loop with a constant time check. |
tree | commitdiff |
2008-01-11 |
Chris Lattner | another minor datastructure tweak. |
tree | commitdiff |
2008-01-11 |
Chris Lattner | start using smallvector to avoid vector heap thrashing. |
tree | commitdiff |
2008-01-11 |
Chris Lattner | When inlining a functino with a byval argument, make... |
tree | commitdiff |
2008-01-08 |
Chris Lattner | Implement PR1795, an instcombine hack for forming GEPs... |
tree | commitdiff |
2008-01-07 |
Duncan Sands | Small cleanup for handling of type/parameter attribute |
tree | commitdiff |
2008-01-07 |
Gordon Henriksen | Deleting an empty file. Thanks, /usr/bin/patch! |
tree | commitdiff |
2008-01-07 |
Gordon Henriksen | With this patch, the LowerGC transformation becomes the |
tree | commitdiff |
2008-01-06 |
Duncan Sands | The transform that tries to turn calls to bitcast funct... |
tree | commitdiff |
2008-01-06 |
Duncan Sands | When transforming a call to a bitcast function into |
tree | commitdiff |
next |