| Commit message (Expand) | Author | Age | Files | Lines |
* | Minor optimization: | Evan Cheng | 2009-03-01 | 1 | -29/+237 |
* | If two-address def is dead and the instruction does not define other register... | Evan Cheng | 2009-02-21 | 1 | -0/+30 |
* | Rename getAnalysisToUpdate to getAnalysisIfAvailable. | Duncan Sands | 2009-01-28 | 1 | -1/+1 |
* | Teach 2addr pass to be do more commuting. If both uses of a two-address instr... | Evan Cheng | 2009-01-25 | 1 | -6/+104 |
* | Refactor code. No functionality change. | Evan Cheng | 2009-01-23 | 1 | -20/+38 |
* | Tidy up #includes, deleting a bunch of unnecessary #includes. | Dan Gohman | 2009-01-05 | 1 | -1/+1 |
* | Do the LiveVariables update before printing the instruction in | Dan Gohman | 2008-11-12 | 1 | -1/+2 |
* | Add an option to enable StrongPHIElimination, for ease of testing. | Owen Anderson | 2008-10-07 | 1 | -1/+5 |
* | Switch the MachineOperand accessors back to the short names like | Dan Gohman | 2008-10-03 | 1 | -7/+7 |
* | Tidy up several unbeseeming casts from pointer to intptr_t. | Dan Gohman | 2008-09-04 | 1 | -1/+1 |
* | Move the check whether it's worth remating to caller. | Evan Cheng | 2008-08-27 | 1 | -0/+1 |
* | Refactor isSafeToReMat out of 2addr pass. | Evan Cheng | 2008-08-27 | 1 | -29/+1 |
* | - Remove calls to copyKillDeadInfo which is an N^2 function. Instead, propaga... | Evan Cheng | 2008-07-03 | 1 | -7/+4 |
* | - Add LiveVariables::replaceKillInstruction. This does a subset of instructio... | Evan Cheng | 2008-07-03 | 1 | -3/+2 |
* | Make LiveVariables even more optional, by making it optional in the call to T... | Owen Anderson | 2008-07-02 | 1 | -46/+1 |
* | TwoAddressInstructionPass doesn't really require LiveVariables, it just needs... | Owen Anderson | 2008-07-02 | 1 | -14/+67 |
* | Remove unneeded include. | Evan Cheng | 2008-06-30 | 1 | -1/+0 |
* | Enable two-address remat by default. | Evan Cheng | 2008-06-25 | 1 | -35/+33 |
* | Missed a check. | Evan Cheng | 2008-06-19 | 1 | -1/+1 |
* | Complete support for two-address pass rematerialization. Now *almost* always ... | Evan Cheng | 2008-06-18 | 1 | -49/+128 |
* | Implement "AsCheapAsAMove" for some obviously cheap instructions: xor and the | Bill Wendling | 2008-05-29 | 1 | -1/+4 |
* | Check the "isSafeToMove" predicate, which has a series of tests to make sure | Bill Wendling | 2008-05-28 | 1 | -1/+3 |
* | Incorporated feedback: Check that the implicitly defined operands aren't used | Bill Wendling | 2008-05-27 | 1 | -9/+8 |
* | The enabling of remat in 2-address conversion breaks this test: | Bill Wendling | 2008-05-26 | 1 | -23/+30 |
* | A problem that's exposed when machine LICM is enabled. Consider this code: | Bill Wendling | 2008-05-26 | 1 | -1/+40 |
* | Change class' public PassInfo variables to by initialized with the | Dan Gohman | 2008-05-13 | 1 | -1/+1 |
* | Clean up the use of static and anonymous namespaces. This turned up | Dan Gohman | 2008-05-13 | 1 | -4/+4 |
* | Cosmetic changes: | Bill Wendling | 2008-05-10 | 1 | -36/+48 |
* | Fix a memory bug: increment an iterator of a deleted machine instr. | Evan Cheng | 2008-03-27 | 1 | -2/+7 |
* | Typo. | Evan Cheng | 2008-03-13 | 1 | -1/+1 |
* | Don't try to sink 3-address instruction if convertToThreeAddress created more... | Evan Cheng | 2008-03-13 | 1 | -1/+6 |
* | Remove an unused command line option. | Evan Cheng | 2008-03-13 | 1 | -9/+0 |
* | TwoAddressInstructionPass enhancement. After it converts a two address instru... | Evan Cheng | 2008-03-13 | 1 | -13/+131 |
* | Refactor code. Remove duplicated functions that basically do the same thing as | Evan Cheng | 2008-03-05 | 1 | -2/+2 |
* | Rename MRegisterInfo to TargetRegisterInfo. | Dan Gohman | 2008-02-10 | 1 | -3/+3 |
* | Remove DefInst from LiveVariables::VarInfo. Use the facilities on MachineReg... | Owen Anderson | 2008-01-15 | 1 | -4/+0 |
* | rename TargetInstrDescriptor -> TargetInstrDesc. | Chris Lattner | 2008-01-07 | 1 | -7/+7 |
* | Rename all the M_* flags to be namespace qualified enums, and switch | Chris Lattner | 2008-01-07 | 1 | -1/+1 |
* | add more and significantly better comments to the rest of the machineinstr | Chris Lattner | 2008-01-07 | 1 | -1/+1 |
* | Move a bunch more accessors from TargetInstrInfo to TargetInstrDescriptor | Chris Lattner | 2008-01-07 | 1 | -2/+2 |
* | Rename MachineInstr::getInstrDescriptor -> getDesc(), which reflects | Chris Lattner | 2008-01-07 | 1 | -1/+1 |
* | Don't recalculate the loop info and loop dominators analyses if they're | Bill Wendling | 2008-01-04 | 1 | -0/+2 |
* | Move copyRegToReg from MRegisterInfo to TargetInstrInfo. This is part of the | Owen Anderson | 2007-12-31 | 1 | -2/+1 |
* | Rename SSARegMap -> MachineRegisterInfo in keeping with the idea | Chris Lattner | 2007-12-31 | 1 | -2/+2 |
* | Remove attribution from file headers, per discussion on llvmdev. | Chris Lattner | 2007-12-29 | 1 | -2/+2 |
* | As Chris and Evan pointed out, BreakCriticalMachineEdges doesn't really need | Owen Anderson | 2007-11-12 | 1 | -1/+0 |
* | This preserves critical edge breaking. | Owen Anderson | 2007-11-08 | 1 | -0/+1 |
* | Bring UsedBlocks back. StrongPHIElimination needs this information. | Owen Anderson | 2007-11-08 | 1 | -0/+4 |
* | It's possible to commute instrctions with more than 3 operands. | Evan Cheng | 2007-10-23 | 1 | -1/+1 |
* | Added missing curly braces which renders the if clause useless in debug build. | Evan Cheng | 2007-10-20 | 1 | -1/+2 |