2005-12-13 |
Evan Cheng | Add inc + dec patterns which fold load + stores. |
commit | commitdiff | tree | snapshot |
2005-12-13 |
Evan Cheng | Add neg and not patterns which fold load + stores. |
commit | commitdiff | tree | snapshot |
2005-12-13 |
Evan Cheng | Missed a couple redundant explicit type casts. |
commit | commitdiff | tree | snapshot |
2005-12-13 |
Evan Cheng | Fix some bad choice of names: i16SExt8 ->i16immSExt8... |
commit | commitdiff | tree | snapshot |
2005-12-13 |
Evan Cheng | * Split immSExt8 to i16SExt8 and i32SExt8 for i16 and... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | Add some integer mul patterns. |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | Bug fix: CodeGenMap[N] = ... -> CodeGenMap[N.getValue... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | At top of generated isel SelectCode() is this: |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | Accept and ignore prefetches for now |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | Fix CodeGen/Generic/2005-12-12-ExpandSextInreg.ll |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | Testcase for a problem that reid ran into |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | Add some sub patterns. |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | When SelectLEAAddr() fails, it shouldn't cause the... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | For ISD::RET, if # of operands >= 2, try selection... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Andrew Lenharth | fix FP selects |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | remove some never-completed and now-obsolete code. |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | Add a few more add / store patterns. e.g. ADD32mi8. |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Evan Cheng | Bug fix: finding the correct incoming chain for pattern... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Andrew Lenharth | restore a more restricted select |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | Implement a little hack for parity with GCC on crafty... |
commit | commitdiff | tree | snapshot |
2005-12-12 |
Chris Lattner | new testcase: sra should be able to eliminate all of... |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Send an indicator to llvm-testresults if the build... |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Fix typo :( |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Minor tweak to get isel opt |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | add selectcc |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Remove type casts that are no longer needed |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Realize the constant pool & global addrs must always... |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Chris Lattner | Fix the JIT failures from last night. |
commit | commitdiff | tree | snapshot |
2005-12-11 |
Andrew Lenharth | FP select improvements (and likely breakage), oh and... |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Nate Begeman | Add support for TargetConstantPool nodes to the dag... |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Evan Cheng | Use SDTCisPtrTy type property for store address. |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Evan Cheng | Stop emitting a redudant type check for complex pattern... |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Evan Cheng | * Added X86 store patterns. |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Evan Cheng | Added new getNode and getTargetNode variants for X86... |
commit | commitdiff | tree | snapshot |
2005-12-10 |
Evan Cheng | For instructions which produce no result, e.g. store... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Nate Begeman | Add support patterns to many load and store instruction... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Chris Lattner | Add SDTCisPtrTy and use it for loads, to indicate that... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Chris Lattner | Add a new SDTCisPtrTy constraint, which indicates that... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Evan Cheng | Added patterns for ADD8rm, etc. These fold load operand... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Evan Cheng | * Do not allow nodes which produce chain results (e... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Chris Lattner | Avoid emitting two tabs when switching to a named section |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Chris Lattner | Teach the PPC backend about the ctor and dtor list... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Chris Lattner | Teach legalize how to promote sext_inreg to fix a probl... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Evan Cheng | Prevent folding of instructions which produce chains... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Evan Cheng | * Make sure complex pattern operands are selected first... |
commit | commitdiff | tree | snapshot |
2005-12-09 |
Andrew Lenharth | it helps if your conditionals are not reversed |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Chris Lattner | improve code insertion in two ways: |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Chris Lattner | Add another important case we miss |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Evan Cheng | Added support for ComplexPattern. |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Evan Cheng | Added explicit type field to ComplexPattern. |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Evan Cheng | * Added an explicit type field to ComplexPattern. |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Evan Cheng | * Added intelligence to X86 LEA addressing mode matchin... |
commit | commitdiff | tree | snapshot |
2005-12-08 |
Evan Cheng | Added support for ComplexPattern. These are patterns... |
commit | commitdiff | tree | snapshot |
2005-12-07 |
Nate Begeman | Fix a crash where ConstantVec nodes were being generate... |
commit | commitdiff | tree | snapshot |
2005-12-07 |
Chris Lattner | Only transform (sext (truncate x)) -> (sextinreg x... |
commit | commitdiff | tree | snapshot |
2005-12-07 |
Chris Lattner | X86 doesn't support sextinreg for 8-bit things either. |
commit | commitdiff | tree | snapshot |
2005-12-07 |
Chris Lattner | Teach the dag combiner to turn a truncate/sign_extend... |
commit | commitdiff | tree | snapshot |
2005-12-07 |
Chris Lattner | Remove a now-dead map, patch by Saem Ghani, thanks! |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Andrew Lenharth | fix divide and remainder |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Chris Lattner | Silence another annoying GCC warning |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Andrew Lenharth | This solves the problem of the CBE renaming symbols... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Andrew Lenharth | more decent branches for FP. I might have to make... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Andrew Lenharth | OK, this does wonders for broken stuff |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Sumant Kowshik | Added comment for removing assert |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Sumant Kowshik | Collapsing node if variable length struct with final... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Sumant Kowshik | Handling of zero length last fields in struct used... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Chris Lattner | Fix test/Regression/Linker/2005-12-06-AppendingZeroLeng... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Chris Lattner | new testcase for PR662 |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Chris Lattner | Implement external storage for post-order iteration... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Nate Begeman | Teach the SelectionDAG ISel how to turn ConstantPacked... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Chris Lattner | Use new PPC-specific nodes to represent shifts which... |
commit | commitdiff | tree | snapshot |
2005-12-06 |
Andrew Lenharth | added instructions with inverted immediates |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Andrew Lenharth | yea, it helps to have your path set right when testing |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Andrew Lenharth | These never trigger, but whatever |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Evan Cheng | Remove unnecessary let hasCtrlDep=1 now it can be inferred. |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Evan Cheng | * Infer instruction property hasCtrlDep from pattern... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Andrew Lenharth | move this over to the dag |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | getRawValue zero extens for unsigned values, use getsex... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Andrew Lenharth | fix constant pool loads |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Fix the #1 code quality problem that I have seen on... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Add a flag to Module::getGlobalVariable to allow it... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | attribute((used)) is now supported |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Wrap a long line, never internalize llvm.used. |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | New testcase for PR660 |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | testcase for PR664 |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Several things: |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | On some targets (e.g. X86), shift amounts are not the... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Implement PR673: for explicit register references,... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Add some methods |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Add some explicit type casts so that tblgen knows the... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Add some explicit type casts so that tblgen knows the... |
commit | commitdiff | tree | snapshot |
2005-12-05 |
Chris Lattner | Generate code to silence bogus GCC warnings. |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Chris Lattner | The basic fneg cases are already autogen'd |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Chris Lattner | Autogen matching code for ADJCALLSTACK[UP|DOWN], thanks... |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Chris Lattner | Finish moving uncond br over to .td file, remove from... |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Chris Lattner | Define BR in the .td file now that Evan made tblgen... |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Evan Cheng | Added isel patterns for RET, JMP, and WRITEPORT. |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Evan Cheng | * Commit the fix (by Chris) for a tblgen type inferenci... |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Evan Cheng | * Added instruction property hasCtrlDep for those which... |
commit | commitdiff | tree | snapshot |
2005-12-04 |
Chris Lattner | Fix PR672 another way which should be more robust |
commit | commitdiff | tree | snapshot |
next |