| Age | Commit message (Expand) | Author |
|---|
| 2009-10-01 | If location info is attached with an instruction then keep track of alloca sl... | Devang Patel |
| 2009-09-30 | Use MDNode * directly as an RecordSourceLine() argument. | Devang Patel |
| 2009-09-30 | Silence comparison always false warning in -Asserts mode. | Reid Kleckner |
| 2009-09-30 | Fix integer overflow in instruction scheduling. This can happen if we have | Reid Kleckner |
| 2009-09-29 | Remove unnecessary cast. | Devang Patel |
| 2009-09-28 | s/class Metadata/class MetadataContext/g | Devang Patel |
| 2009-09-28 | Do not use global typedef for MDKindID. | Devang Patel |
| 2009-09-28 | Use VerifySchedule instead of doing the work manually. | Dan Gohman |
| 2009-09-26 | Convert comparisons like (x == infinity) to (x >= infinity) on targets | Dan Gohman |
| 2009-09-25 | Improve MachineMemOperand handling. | Dan Gohman |
| 2009-09-25 | Rename getTargetNode to getMachineNode, for consistency with the | Dan Gohman |
| 2009-09-25 | Make sure sin, cos, sqrt calls are marked readonly | Dale Johannesen |
| 2009-09-25 | Generate FSQRT from calls to the sqrt function, which | Dale Johannesen |
| 2009-09-25 | Fix combiner-aa issue with bases which are different, but can alias. | Nate Begeman |
| 2009-09-25 | Add a version of dumpr() that has a SelectionDAG* argument. | Dan Gohman |
| 2009-09-23 | Use getStoreSize() instead of getStoreSizeInBits()/8. | Dan Gohman |
| 2009-09-23 | Rename several variables from EVT to more descriptive names, now that EVT | Dan Gohman |
| 2009-09-23 | Give MachineMemOperand an operator<<, factoring out code from | Dan Gohman |
| 2009-09-21 | Change MachineMemOperand's alignment value to be the alignment of | Dan Gohman |
| 2009-09-20 | tidy up | Chris Lattner |
| 2009-09-20 | Tabs -> spaces, and remove trailing whitespace. | Daniel Dunbar |
| 2009-09-19 | Fix PR4926. When target hook EmitInstrWithCustomInserter() insert new basic b... | Evan Cheng |
| 2009-09-18 | Enhance EmitInstrWithCustomInserter() so target can specify CFG changes that ... | Evan Cheng |
| 2009-09-18 | duncan points out the EH selector values are signed. | Chris Lattner |
| 2009-09-18 | Revert r82214. It broke 403.gcc on x86_64 / Darwin. | Evan Cheng |
| 2009-09-18 | Fix a bug in sdisel switch lowering code. When it updates the phi nodes in sw... | Evan Cheng |
| 2009-09-17 | tolerate llvm.eh.selector.i64 on 32-bit systems and llvm.eh.selector.i32 on | Chris Lattner |
| 2009-09-16 | Fix typo. | Devang Patel |
| 2009-09-16 | At iSel time, update DebugLoc based on debug info attached with an instruction. | Devang Patel |
| 2009-09-15 | Do not add the SVOffset to the Node CSE ID. The same pointer argument cannot... | Nate Begeman |
| 2009-09-15 | Better solution for tracking both the original alignment of the access, and t... | Nate Begeman |
| 2009-09-15 | Remove incorrect CSE code from r81813. | Nate Begeman |
| 2009-09-15 | Substantially speed up combiner-aa in the following ways: | Nate Begeman |
| 2009-09-15 | Teach the legalizer to propagate the original alignment of loads and store when | Nate Begeman |
| 2009-09-15 | Add an "original alignment" field to load and store nodes. This enables the | Nate Begeman |
| 2009-09-13 | kill off the last use of TRI::AsmName. | Chris Lattner |
| 2009-09-11 | Make fast-isel try ISD::FNEG before resorting to bitcasts and xors. | Dan Gohman |
| 2009-09-11 | Reapply r81171 with a fix: don't try to use i64 when it | Dan Gohman |
| 2009-09-10 | Don't swap the operands of a subtraction when trying to create a | Bob Wilson |
| 2009-09-10 | Revert r81171 which was causing pr4927. | Bob Wilson |
| 2009-09-09 | When widening a vector load, use the correct chain. This fixes PR4891. | Dan Gohman |
| 2009-09-08 | change selectiondag to add the sign extended versions of immediate operands | Chris Lattner |
| 2009-09-08 | Fix an abort on a store of an empty struct member. getValue returns | Dan Gohman |
| 2009-09-07 | Fix a thinko: When lowering fneg with xor, bitcast the operands | Dan Gohman |
| 2009-09-06 | Simplify. Testing shows that this is not equivalent to BBI = CR.CaseBB + 1. | Duncan Sands |
| 2009-09-06 | Remove some not-really-used variables, as warned | Duncan Sands |
| 2009-09-06 | Remove some unused variables and methods warned about by | Duncan Sands |
| 2009-09-05 | Detect VLAs. | Devang Patel |
| 2009-09-03 | LLVM currently represents floating-point negation as -0.0 - x. Fix | Dan Gohman |
| 2009-09-03 | Recognize more opportunities to use SSE min and max instructions, | Dan Gohman |
