2011-04-20 |
Cameron Zwarich | The bitcast case here is actually handled uniformly... |
blob | commitdiff | raw | diff to current |
2011-04-20 |
Cameron Zwarich | Cleanup some code to better use an early return style... |
blob | commitdiff | raw | diff to current |
2011-04-14 |
Mon P Wang | Cleanup r129509 based on comments by Chris |
blob | commitdiff | raw | diff to current |
2011-04-14 |
Mon P Wang | Cleanup r129472 by using a utility routine as suggested... |
blob | commitdiff | raw | diff to current |
2011-04-13 |
Mon P Wang | Vectors with different number of elements of the same... |
blob | commitdiff | raw | diff to current |
2011-03-30 |
Jay Foad | Remove PHINode::reserveOperandSpace(). Instead, add... |
blob | commitdiff | raw | diff to current |
2011-03-30 |
Jay Foad | (Almost) always call reserveOperandSpace() on newly... |
blob | commitdiff | raw | diff to current |
2011-03-29 |
Cameron Zwarich | Do some simple copy propagation through integer loads... |
blob | commitdiff | raw | diff to current |
2011-03-26 |
Cameron Zwarich | Fix a typo and add a test. |
blob | commitdiff | raw | diff to current |
2011-03-23 |
Cameron Zwarich | Fix PR9464 by correcting some math that just happened... |
blob | commitdiff | raw | diff to current |
2011-03-16 |
Cameron Zwarich | Fix a comment. |
blob | commitdiff | raw | diff to current |
2011-03-16 |
Cameron Zwarich | Only convert allocas to scalars if it is profitable... |
blob | commitdiff | raw | diff to current |
2011-03-16 |
Cameron Zwarich | Better use initializer lists. |
blob | commitdiff | raw | diff to current |
2011-03-16 |
Cameron Zwarich | Add a clarifying comment. |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Cameron Zwarich | Fix a crasher introduced by r127317 that is seen on... |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Cameron Zwarich | Add support to scalar replacement for partial vector... |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Cameron Zwarich | Move vector type merging to a separate function in... |
blob | commitdiff | raw | diff to current |
2011-02-15 |
Chris Lattner | convert ConstantVector::get to use ArrayRef. |
blob | commitdiff | raw | diff to current |
2011-02-14 |
Chris Lattner | revert my ConstantVector patch, it seems to have made... |
blob | commitdiff | raw | diff to current |
2011-02-14 |
Chris Lattner | Switch ConstantVector::get to use ArrayRef instead... |
blob | commitdiff | raw | diff to current |
2011-01-24 |
Dan Gohman | Give GetUnderlyingObject a TargetData, to keep it in... |
blob | commitdiff | raw | diff to current |
2011-01-24 |
Chris Lattner | enhance SRoA to promote allocas that are used by PHI... |
blob | commitdiff | raw | diff to current |
2011-01-23 |
Chris Lattner | Enhance SRoA to promote allocas that are used by select... |
blob | commitdiff | raw | diff to current |
2011-01-23 |
Chris Lattner | Enhance SRoA to be more aggressive about scalarization... |
blob | commitdiff | raw | diff to current |
2011-01-23 |
Chris Lattner | have AllocaInfo store the alloca being inspected, simpl... |
blob | commitdiff | raw | diff to current |
2011-01-23 |
Chris Lattner | Rearrange some code a bit. Change MarkUnsafe to |
blob | commitdiff | raw | diff to current |
2011-01-23 |
Chris Lattner | remove an old hack that avoided creating MMX datatypes... |
blob | commitdiff | raw | diff to current |
2011-01-18 |
Cameron Zwarich | Remove outdated references to dominance frontiers. |
blob | commitdiff | raw | diff to current |
2011-01-17 |
Cameron Zwarich | Roll r123609 back in with two changes that fix test... |
blob | commitdiff | raw | diff to current |
2011-01-17 |
Cameron Zwarich | Roll out r123609 due to failures on the llvm-x86_64... |
blob | commitdiff | raw | diff to current |
2011-01-17 |
Cameron Zwarich | Eliminate the use of dominance frontiers in PromoteMemT... |
blob | commitdiff | raw | diff to current |
2011-01-16 |
Chris Lattner | tidy up a comment, as suggested by duncan |
blob | commitdiff | raw | diff to current |
2011-01-16 |
Chris Lattner | if an alloca is only ever accessed as a unit, and is... |
blob | commitdiff | raw | diff to current |
2011-01-16 |
Chris Lattner | Use an irbuilder to get some trivial constant folding... |
blob | commitdiff | raw | diff to current |
2011-01-16 |
Chris Lattner | enhance FoldOpIntoPhi in instcombine to try harder... |
blob | commitdiff | raw | diff to current |
2011-01-15 |
Chris Lattner | Generalize LoadAndStorePromoter a bit and switch LICM |
blob | commitdiff | raw | diff to current |
2011-01-14 |
Chris Lattner | switch SRoA to use LoadAndStorePromoter instead of... |
blob | commitdiff | raw | diff to current |
2011-01-14 |
Chris Lattner | split SROA into two passes: one that uses DomFrontiers... |
blob | commitdiff | raw | diff to current |
2011-01-14 |
Chris Lattner | Implement full support for promoting allocas to registe... |
blob | commitdiff | raw | diff to current |
2011-01-13 |
Bob Wilson | Fix whitespace. |
blob | commitdiff | raw | diff to current |
2011-01-13 |
Bob Wilson | Check for empty structs, and for consistency, zero... |
blob | commitdiff | raw | diff to current |
2011-01-13 |
Bob Wilson | Extend SROA to handle arrays accessed as homogeneous... |
blob | commitdiff | raw | diff to current |
2011-01-13 |
Bob Wilson | Make SROA more aggressive with allocas containing padding. |
blob | commitdiff | raw | diff to current |
2011-01-02 |
Chris Lattner | split dom frontier handling stuff out to its own Domina... |
blob | commitdiff | raw | diff to current |
2010-12-26 |
Chris Lattner | start using irbuilder to make mem intrinsics in a few... |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Mon P Wang | Preserve the address space when generating bitcasts... |
blob | commitdiff | raw | diff to current |
2010-12-15 |
Dan Gohman | Move Value::getUnderlyingObject to be a standalone |
blob | commitdiff | raw | diff to current |
2010-11-24 |
Nick Lewycky | Treat a call of function pointer like a load of the... |
blob | commitdiff | raw | diff to current |
2010-11-20 |
Benjamin Kramer | Simplify code. No change in functionality. |
blob | commitdiff | raw | diff to current |
2010-11-18 |
Chris Lattner | finish a thought. |
blob | commitdiff | raw | diff to current |
2010-11-18 |
Chris Lattner | allow eliminating an alloca that is just copied from... |
blob | commitdiff | raw | diff to current |
2010-11-18 |
Chris Lattner | enhance the "alloca is just a memcpy from constant... |
blob | commitdiff | raw | diff to current |
2010-11-18 |
Chris Lattner | fix a small oversight in the "eliminate memcpy from... |
blob | commitdiff | raw | diff to current |
2010-10-19 |
Owen Anderson | Get rid of static constructors for pass registration... |
blob | commitdiff | raw | diff to current |
2010-10-16 |
Benjamin Kramer | Eliminate some calls to Value::getNameStr. |
blob | commitdiff | raw | diff to current |
2010-10-12 |
Owen Anderson | Begin adding static dependence information to passes... |
blob | commitdiff | raw | diff to current |
2010-10-07 |
Owen Anderson | Now with fewer extraneous semicolons! |
blob | commitdiff | raw | diff to current |
2010-09-30 |
Dale Johannesen | Massive rewrite of MMX: |
blob | commitdiff | raw | diff to current |
2010-09-01 |
Chris Lattner | deepen my MMX/SRoA hack to avoid hurting non-x86 codegen. |
blob | commitdiff | raw | diff to current |
2010-09-01 |
Chris Lattner | add a gross hack to work around a problem that Argiris... |
blob | commitdiff | raw | diff to current |
2010-08-18 |
Chris Lattner | remove dead prototype. |
blob | commitdiff | raw | diff to current |
2010-08-06 |
Owen Anderson | Reapply r110396, with fixes to appease the Linux buildb... |
blob | commitdiff | raw | diff to current |
2010-08-06 |
Owen Anderson | Revert r110396 to fix buildbots. |
blob | commitdiff | raw | diff to current |
2010-08-05 |
Owen Anderson | Don't use PassInfo* as a type identifier for passes... |
blob | commitdiff | raw | diff to current |
2010-07-21 |
Owen Anderson | Fix batch of converting RegisterPass<> to INTIALIZE_PASS(). |
blob | commitdiff | raw | diff to current |
2010-07-16 |
Gabor Greif | eliminate CallInst::ArgOffset |
blob | commitdiff | raw | diff to current |
2010-07-08 |
Chris Lattner | Fix the second half of PR7437: scalarrepl wasn't preserving |
blob | commitdiff | raw | diff to current |
2010-06-30 |
Gabor Greif | use getArgOperand instead of getOperand |
blob | commitdiff | raw | diff to current |
2010-06-28 |
Gabor Greif | employ CallInst::ArgOffset (for now) |
blob | commitdiff | raw | diff to current |
2010-06-28 |
Gabor Greif | use cached value |
blob | commitdiff | raw | diff to current |
2010-06-27 |
Chris Lattner | minor cleanup to SROA: when lowering type unsafe access... |
blob | commitdiff | raw | diff to current |
2010-06-22 |
Dan Gohman | Use pre-increment instead of post-increment when the... |
blob | commitdiff | raw | diff to current |
2010-04-20 |
Gabor Greif | use abstract accessors to CallInst |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Eric Christopher | Revert 101465, it broke internal OpenGL testing. |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Gabor Greif | reapply r101434 |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Chris Lattner | fix comment noticed by Bob |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Gabor Greif | back out r101423 and r101397, they break llvm-gcc self... |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Chris Lattner | fix PR6832: we were using the alignment of a pointer... |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Chris Lattner | improve comments. |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Chris Lattner | pull all the ConvertToScalarInfo code together into one |
blob | commitdiff | raw | diff to current |
2010-04-16 |
Chris Lattner | more refactoring: suck some stuff out of SRoA into |
blob | commitdiff | raw | diff to current |
2010-04-15 |
Chris Lattner | introduce a new ConvertToScalarInfo struct to simplify |
blob | commitdiff | raw | diff to current |
2010-04-15 |
Chris Lattner | tidy interface to isOnlyCopiedFromConstantGlobal |
blob | commitdiff | raw | diff to current |
2010-04-15 |
Gabor Greif | reapply r101364, which has been backed out in r101368 |
blob | commitdiff | raw | diff to current |
2010-04-15 |
Gabor Greif | back out r101364, as it trips the linux nightlybot... |
blob | commitdiff | raw | diff to current |
2010-04-15 |
Gabor Greif | rotate CallInst operands, i.e. move callee to the back |
blob | commitdiff | raw | diff to current |
2010-04-06 |
Gabor Greif | performance: get rid of repeated dereferencing of use_i... |
blob | commitdiff | raw | diff to current |
2010-04-04 |
Mon P Wang | Reapply address space patch after fixing an issue in... |
blob | commitdiff | raw | diff to current |
2010-04-02 |
Mon P Wang | Revert r100191 since it breaks objc in clang |
blob | commitdiff | raw | diff to current |
2010-04-02 |
Mon P Wang | Reapply address space patch after fixing an issue in... |
blob | commitdiff | raw | diff to current |
2010-03-30 |
Bob Wilson | Revert Mon Ping's change 99928, since it broke all... |
blob | commitdiff | raw | diff to current |
2010-03-30 |
Mon P Wang | Added support for address spaces and added a isVolatile... |
blob | commitdiff | raw | diff to current |
2010-02-16 |
Duncan Sands | There are two ways of checking for a given type, for... |
blob | commitdiff | raw | diff to current |
2010-02-15 |
Duncan Sands | Uniformize the names of type predicates: rather than... |
blob | commitdiff | raw | diff to current |
2010-02-03 |
Bob Wilson | Adjust the heuristics used to decide when SROA is likel... |
blob | commitdiff | raw | diff to current |
2010-01-27 |
Benjamin Kramer | Use the less expensive getName function instead of... |
blob | commitdiff | raw | diff to current |
2010-01-25 |
Bob Wilson | Change Value::getUnderlyingObject to have the MaxLookup... |
blob | commitdiff | raw | diff to current |
2010-01-21 |
Victor Hernandez | DbgInfoIntrinsics no longer appear in an instruction... |
blob | commitdiff | raw | diff to current |
2010-01-19 |
Bob Wilson | Fix a crash in scalarrepl for memcpy/memmove where... |
blob | commitdiff | raw | diff to current |
2010-01-05 |
Dan Gohman | Use do+while instead of while for loops which obviously... |
blob | commitdiff | raw | diff to current |
next |