| Commit message (Expand) | Author | Age | Files | Lines |
... | |
* | Added getArchNameForAssembler method to the Triple class for which returns OS... | Viktor Kutuzov | 2009-11-17 | 1 | -0/+27 |
* | Remove a special case for tail merging that seems to be both broken and | Bob Wilson | 2009-11-17 | 1 | -33/+0 |
* | Generalize OptimizeLoopTermCond to optimize more loop terminating icmp to use... | Evan Cheng | 2009-11-17 | 1 | -87/+99 |
* | Set MadeChange instead of MadeChangeThisIteration. | Dan Gohman | 2009-11-17 | 1 | -1/+1 |
* | Revert CPU detection code to return "generic" instead of an empty string in case | Benjamin Kramer | 2009-11-17 | 2 | -4/+4 |
* | Remove trailing whitespace | Jim Grosbach | 2009-11-17 | 1 | -74/+74 |
* | Update a comment, now that tail duplication happens after other branch | Bob Wilson | 2009-11-17 | 1 | -2/+2 |
* | Set Inst{15-12} (Rd/Rt) to 0b1111 (PC) for BR_JTadd, BR_JTr, and BR_JTm to | Johnny Chen | 2009-11-17 | 1 | -0/+3 |
* | Perform tail duplication only once, after tail merging is complete. | Bob Wilson | 2009-11-17 | 2 | -33/+68 |
* | Revert 89021. It's miscompiling llvm-gcc driver driver at -O0. | Evan Cheng | 2009-11-17 | 1 | -1/+1 |
* | Re-apply 89011. It's not to be blamed. | Evan Cheng | 2009-11-17 | 2 | -4/+7 |
* | Revert 89011. Buildbot thinks it might be breaking stuff. | Evan Cheng | 2009-11-17 | 2 | -7/+4 |
* | Remove VISIBILITY_HIDDEN from the classes in this directory. Fixes bug 5507. | Nick Lewycky | 2009-11-17 | 5 | -19/+18 |
* | Revert r88939. | Nick Lewycky | 2009-11-17 | 1 | -2/+2 |
* | Fail less mysteriously; inform the user that their LLVM was not built with | Nick Lewycky | 2009-11-17 | 1 | -0/+3 |
* | Fix a race condition in the Timer class. | Owen Anderson | 2009-11-17 | 1 | -38/+5 |
* | Refactor the code that creates the "dot-label" difference. This may be used in | Bill Wendling | 2009-11-17 | 2 | -18/+29 |
* | When moving a block for table jumps, make sure the prior block terminator | Jim Grosbach | 2009-11-17 | 1 | -7/+9 |
* | Enable -split-phi-edges by default | Jakob Stoklund Olesen | 2009-11-17 | 1 | -1/+1 |
* | MOV64rm should be marked isReMaterializable. | Evan Cheng | 2009-11-17 | 1 | -1/+1 |
* | Remove the optimizations that convert BRCOND and BR_CC into | Dan Gohman | 2009-11-17 | 1 | -17/+12 |
* | Remove debug info attached with an instruction. | Devang Patel | 2009-11-17 | 1 | -36/+18 |
* | In GlobalVariable::setInitializer, assert that the initializer has the | Jeffrey Yasskin | 2009-11-17 | 1 | -0/+15 |
* | A few more instructions that should be marked re-materializable. | Evan Cheng | 2009-11-17 | 2 | -4/+7 |
* | Set Rm bits of BX_RET to 0b1110 (R14); and set condition code bits of BRIND to | Johnny Chen | 2009-11-16 | 1 | -0/+2 |
* | Initialize the new AsmPrinterFlags field to 0, fixing uses of | Dan Gohman | 2009-11-16 | 1 | -6/+8 |
* | Make X86-64 in the Large model always emit 64-bit calls. | Jeffrey Yasskin | 2009-11-16 | 5 | -56/+30 |
* | - Check memoperand alignment instead of checking stack alignment. Most load /... | Evan Cheng | 2009-11-16 | 2 | -17/+14 |
* | Revert r88939. | Devang Patel | 2009-11-16 | 1 | -2/+2 |
* | Fix an expensive-checks error. | David Greene | 2009-11-16 | 1 | -20/+24 |
* | Make the pass class name more explicit. | Jim Grosbach | 2009-11-16 | 1 | -4/+6 |
* | make pass name a bit more clear | Jim Grosbach | 2009-11-16 | 1 | -1/+1 |
* | Sink a #include <map> to where it's actually needed. | Dan Gohman | 2009-11-16 | 1 | -0/+1 |
* | Fix a typo in a comment. | Dan Gohman | 2009-11-16 | 2 | -2/+2 |
* | Fix a comment. | Bob Wilson | 2009-11-16 | 1 | -1/+1 |
* | Add VISIBILITY_HIDDEN marker. | Devang Patel | 2009-11-16 | 1 | -2/+2 |
* | Simplify thumb2 jump table adjustments. Remove unnecessary calculation and | Jim Grosbach | 2009-11-16 | 1 | -58/+14 |
* | clarify comment | Jim Grosbach | 2009-11-16 | 1 | -2/+2 |
* | Fix some comments. | Bob Wilson | 2009-11-16 | 1 | -5/+4 |
* | Whitespace: be consistent with pointer syntax. | Bob Wilson | 2009-11-16 | 1 | -6/+6 |
* | Clean up whitespace. | Bob Wilson | 2009-11-16 | 1 | -6/+6 |
* | back off for a bit. tracking down weirdness | Jim Grosbach | 2009-11-16 | 1 | -1/+1 |
* | Analyze has to be before checking the condition, obviously. Properly construc... | Jim Grosbach | 2009-11-16 | 1 | -6/+8 |
* | Support spill comments. | David Greene | 2009-11-16 | 1 | -0/+52 |
* | BuildIntCast takes an additional parameter, isSigned. | Duncan Sands | 2009-11-16 | 1 | -2/+3 |
* | CreateIntCast takes an "isSigned" parameter. Pass "true" for it, rather than | Duncan Sands | 2009-11-16 | 1 | -1/+3 |
* | Special case FixedStackPseudoSourceValueVal as well. Do we really need to dif... | Evan Cheng | 2009-11-16 | 1 | -1/+2 |
* | Check if subreg index is zero. | Evan Cheng | 2009-11-16 | 1 | -1/+3 |
* | For some targets, a copy can use a register multiple times, e.g. ppc. | Evan Cheng | 2009-11-16 | 1 | -0/+5 |
* | Disable ldc1/sdc1 instructions for mips1 targets. | Bruno Cardoso Lopes | 2009-11-16 | 1 | -1/+2 |