| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | Revert r97245 which seems to be causing performance problems. | Bob Wilson | 2010-02-28 | 1 | -6/+5 |
| * | Move the EnableFullLoadPRE flag from a separate command-line option to an | Bob Wilson | 2010-02-26 | 1 | -5/+6 |
| * | Remove unused "NoPRE" parameter in GVN and createGVNPass(). | Bob Wilson | 2010-02-26 | 1 | -5/+4 |
| * | Reapply r97010, the speculative revert failed. | Daniel Dunbar | 2010-02-24 | 1 | -6/+1 |
| * | Speculatively revert r97010, "Add an argument to PHITranslateValue to specify | Daniel Dunbar | 2010-02-24 | 1 | -1/+6 |
| * | Add an argument to PHITranslateValue to specify the DominatorTree. If this | Bob Wilson | 2010-02-24 | 1 | -6/+1 |
| * | Update memdep when load PRE inserts a new load, and add some debug output. | Bob Wilson | 2010-02-23 | 1 | -0/+2 |
| * | Erase deleted instructions from GVN's ValueTable. This fixes assertion | Bob Wilson | 2010-02-22 | 1 | -0/+7 |
| * | Rename SuccessorNumber to GetSuccessorNumber. | Bob Wilson | 2010-02-16 | 1 | -2/+2 |
| * | Split critical edges as needed for load PRE. | Bob Wilson | 2010-02-16 | 1 | -8/+30 |
| * | Refactor to share code to find the position of a basic block successor in the | Bob Wilson | 2010-02-16 | 1 | -8/+1 |
| * | There are two ways of checking for a given type, for example isa<PointerType>(T) | Duncan Sands | 2010-02-16 | 1 | -24/+24 |
| * | Ignore dbg info intrinsics. | Devang Patel | 2010-02-11 | 1 | -0/+4 |
| * | Fix some comment typos. | Bob Wilson | 2010-02-03 | 1 | -3/+3 |
| * | Add an option to GVN to remove all partially redundant loads. This is currently | Bob Wilson | 2010-02-01 | 1 | -80/+107 |
| * | Check alignment of loads when deciding whether it is safe to execute them | Bob Wilson | 2010-01-30 | 1 | -1/+2 |
| * | Improve isSafeToLoadUnconditionally to recognize that GEPs with constant | Bob Wilson | 2010-01-29 | 1 | -1/+2 |
| * | Remove check for an impossible condition: the condition of the while loop has | Bob Wilson | 2010-01-25 | 1 | -2/+0 |
| * | Convert some of the dynamic opcode lookups into static ones. | Owen Anderson | 2010-01-17 | 1 | -59/+40 |
| * | Use do+while instead of while for loops which obviously have a | Dan Gohman | 2010-01-05 | 1 | -2/+2 |
| * | Change errs() to dbgs(). | David Greene | 2010-01-05 | 1 | -20/+20 |
| * | split code that doesn't need to be templated out of IRBuilder into a new | Chris Lattner | 2009-12-28 | 1 | -0/+1 |
| * | The phi translated pointer can be computed when returning a partially cached ... | Chris Lattner | 2009-12-22 | 1 | -1/+1 |
| * | Add a fastpath to Load GVN to special case when we have exactly one dominating | Chris Lattner | 2009-12-21 | 1 | -2/+10 |
| * | refactor some code out to a new helper method. | Chris Lattner | 2009-12-21 | 1 | -22/+27 |
| * | Eliminte unnecessary uses of <cstdio>. | Dan Gohman | 2009-12-18 | 1 | -4/+3 |
| * | Remove isPod() from DenseMapInfo, splitting it out to its own | Chris Lattner | 2009-12-15 | 1 | -1/+4 |
| * | Fix PR5744, a case where we were getting the pointer size instead of the | Chris Lattner | 2009-12-10 | 1 | -1/+1 |
| * | allow this to build when the #if 0's are enabled. No functionality change. | Chris Lattner | 2009-12-10 | 1 | -6/+2 |
| * | fix hte last remaining known (by me) phi translation bug. When we reanalyze | Chris Lattner | 2009-12-09 | 1 | -6/+9 |
| * | change GetStoreValueForLoad to use IRBuilder, which is cleaner and | Chris Lattner | 2009-12-09 | 1 | -7/+7 |
| * | change AnalyzeLoadFromClobberingMemInst/AnalyzeLoadFromClobberingStore | Chris Lattner | 2009-12-09 | 1 | -13/+22 |
| * | change AnalyzeLoadFromClobberingWrite and clients to pass in type | Chris Lattner | 2009-12-09 | 1 | -10/+12 |
| * | change NonLocalDepEntry from being a typedef for an std::pair to be its | Chris Lattner | 2009-12-09 | 1 | -10/+10 |
| * | add some aborts to #if 0's. | Chris Lattner | 2009-12-09 | 1 | -0/+2 |
| * | Switch GVN and memdep to use PHITransAddr, which correctly handles | Chris Lattner | 2009-12-09 | 1 | -17/+22 |
| * | constant fold loads from memcpy's from global constants. This is important | Chris Lattner | 2009-12-06 | 1 | -6/+46 |
| * | add support for forwarding mem intrinsic values to non-local loads. | Chris Lattner | 2009-12-06 | 1 | -33/+71 |
| * | Handle forwarding local memsets to loads. For example, we optimize this: | Chris Lattner | 2009-12-06 | 1 | -43/+136 |
| * | Fix this crasher, and add a FIXME for a missed optimization. | Owen Anderson | 2009-12-03 | 1 | -0/+10 |
| * | Cleanup/remove some parts of the lifetime region handling code in memdep and ... | Owen Anderson | 2009-12-02 | 1 | -8/+6 |
| * | factor some code better. | Chris Lattner | 2009-12-02 | 1 | -10/+10 |
| * | reenable load address insertion in load pre. This allows us to | Chris Lattner | 2009-11-28 | 1 | -11/+19 |
| * | Enhance InsertPHITranslatedPointer to be able to return a list of newly | Chris Lattner | 2009-11-28 | 1 | -1/+3 |
| * | disable value insertion for now, I need to figure out how | Chris Lattner | 2009-11-27 | 1 | -1/+7 |
| * | Rework InsertPHITranslatedPointer to handle the recursive case, this | Chris Lattner | 2009-11-27 | 1 | -24/+14 |
| * | teach GVN's load PRE to insert computations of the address in predecessors | Chris Lattner | 2009-11-27 | 1 | -18/+26 |
| * | Fix phi translation in load PRE to agree with the phi | Chris Lattner | 2009-11-27 | 1 | -2/+8 |
| * | Add an option for running GVN with redundant load processing disabled. | Dan Gohman | 2009-11-14 | 1 | -9/+23 |
| * | Fix DenseMap iterator constness. | Jeffrey Yasskin | 2009-11-10 | 1 | -4/+4 |
