2008-05-23 |
Dan Gohman | Use PMULDQ for v2i64 multiplies when SSE4.1 is availabl... |
blob | commitdiff | raw |
2008-05-14 |
Dan Gohman | Change target-specific classes to use more precise... |
blob | commitdiff | raw | diff to current |
2008-05-12 |
Bill Wendling | Constify the machine instruction passed into the |
blob | commitdiff | raw | diff to current |
2008-05-02 |
Evan Cheng | Undo r50574. We are already ensuring the folded load... |
blob | commitdiff | raw | diff to current |
2008-05-02 |
Evan Cheng | Not safe folding a load + FsXORPSrr into FsXORPSrm... |
blob | commitdiff | raw | diff to current |
2008-04-20 |
Nicolas Geoffray | Don't forget to update the current operand when getting... |
blob | commitdiff | raw | diff to current |
2008-04-16 |
Evan Cheng | Don't forget about sub-register indices when rematting... |
blob | commitdiff | raw | diff to current |
2008-04-16 |
Nicolas Geoffray | Infrastructure for getting the machine code size of... |
blob | commitdiff | raw | diff to current |
2008-04-07 |
Dan Gohman | Rename MemOperand to MachineMemOperand. This was sugges... |
blob | commitdiff | raw | diff to current |
2008-04-01 |
Evan Cheng | ReMat of load from stub in pic mode extends the life... |
blob | commitdiff | raw | diff to current |
2008-04-01 |
Evan Cheng | Remove unnecessary and non-deterministic checking code... |
blob | commitdiff | raw | diff to current |
2008-04-01 |
Evan Cheng | Disabling remat of load from gv stub (temporarily)... |
blob | commitdiff | raw | diff to current |
2008-03-31 |
Evan Cheng | It's not safe to fold a load from GV stub or constantpo... |
blob | commitdiff | raw | diff to current |
2008-03-31 |
Evan Cheng | Move reMaterialize() from TargetRegisterInfo to TargetI... |
blob | commitdiff | raw | diff to current |
2008-03-31 |
Evan Cheng | Re-apply 48911. |
blob | commitdiff | raw | diff to current |
2008-03-28 |
Evan Cheng | Backing out 48911 for now. It's breaking stuff. |
blob | commitdiff | raw | diff to current |
2008-03-28 |
Evan Cheng | Load from stub is already re-materializable. |
blob | commitdiff | raw | diff to current |
2008-03-27 |
Evan Cheng | Code clean up. |
blob | commitdiff | raw | diff to current |
2008-03-27 |
Evan Cheng | Allow certain lea instructions to be rematerialized. |
blob | commitdiff | raw | diff to current |
2008-03-27 |
Evan Cheng | Remove an unused command line option. |
blob | commitdiff | raw | diff to current |
2008-03-25 |
Dan Gohman | Add CMP32mr and friends to the load-unfolding table... |
blob | commitdiff | raw | diff to current |
2008-03-21 |
Chris Lattner | Add support for calls that return two FP values in |
blob | commitdiff | raw | diff to current |
2008-03-16 |
Christopher Lamb | Make insert_subreg a two-address instruction, vastly... |
blob | commitdiff | raw | diff to current |
2008-03-13 |
Christopher Lamb | Get rid of a pseudo instruction and replace it with... |
blob | commitdiff | raw | diff to current |
2008-03-11 |
Chris Lattner | coalesce away 80-bit floating point copies. |
blob | commitdiff | raw | diff to current |
2008-03-11 |
Chris Lattner | convert a massive if statement to a switch. |
blob | commitdiff | raw | diff to current |
2008-03-11 |
Christopher Lamb | Missed part of recommit. |
blob | commitdiff | raw | diff to current |
2008-03-10 |
Chris Lattner | abort with an assert instead of a cerr to get line# |
blob | commitdiff | raw | diff to current |
2008-03-10 |
Evan Cheng | Revert 48125, 48126, and 48130 for now to unbreak some... |
blob | commitdiff | raw | diff to current |
2008-03-10 |
Christopher Lamb | Allow insert_subreg into implicit, target-specific... |
blob | commitdiff | raw | diff to current |
2008-03-09 |
Chris Lattner | teach X86InstrInfo::copyRegToReg how to copy into ST... |
blob | commitdiff | raw | diff to current |
2008-03-09 |
Chris Lattner | add some code to support cross-register class copying... |
blob | commitdiff | raw | diff to current |
2008-03-09 |
Chris Lattner | rearrange some code, no functionality change. |
blob | commitdiff | raw | diff to current |
2008-02-23 |
Evan Cheng | Turning on remat of pic loads. |
blob | commitdiff | raw | diff to current |
2008-02-23 |
Evan Cheng | No need recognize load from a fixed argument slot as... |
blob | commitdiff | raw | diff to current |
2008-02-22 |
Evan Cheng | Allow re-materialization of pic load (controlled by... |
blob | commitdiff | raw | diff to current |
2008-02-13 |
Evan Cheng | commuteInstr() can now commute non-ssa machine instrs. |
blob | commitdiff | raw | diff to current |
2008-02-08 |
Evan Cheng | It's not always safe to fold movsd into xorpd, etc... |
blob | commitdiff | raw | diff to current |
2008-02-08 |
Evan Cheng | Added missing entries in X86 load / store folding tables. |
blob | commitdiff | raw | diff to current |
2008-02-07 |
Evan Cheng | In some cases, e.g. ADD32ri, no transformation is made... |
blob | commitdiff | raw | diff to current |
2008-01-12 |
Chris Lattner | fix a wordo that gordon noticed :) |
blob | commitdiff | raw | diff to current |
2008-01-12 |
Chris Lattner | Any x86 instruction that reads from an invariant locati... |
blob | commitdiff | raw | diff to current |
2008-01-11 |
Chris Lattner | rename MachineInstr::setInstrDescriptor -> setDesc |
blob | commitdiff | raw | diff to current |
2008-01-11 |
Chris Lattner | remove xchg and shift-reg-by-1 instructions, which... |
blob | commitdiff | raw | diff to current |
2008-01-10 |
Chris Lattner | Simplify the side effect stuff a bit more and make... |
blob | commitdiff | raw | diff to current |
2008-01-10 |
Chris Lattner | verify that the frame index is immutable before remat... |
blob | commitdiff | raw | diff to current |
2008-01-09 |
Chris Lattner | add a testcase |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Bill Wendling | Operand 1 should be a register. We don't care if it... |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Chris Lattner | rename TargetInstrDescriptor -> TargetInstrDesc. |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Chris Lattner | Move a bunch more accessors from TargetInstrInfo to... |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Chris Lattner | remove MachineOpCode typedef. |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Chris Lattner | Add predicates methods to TargetOperandInfo, and switch... |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Chris Lattner | Rename MachineInstr::getInstrDescriptor -> getDesc... |
blob | commitdiff | raw | diff to current |
2008-01-07 |
Owen Anderson | Move even more functionality from MRegisterInfo into... |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Bill Wendling | Fix comment. |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Bill Wendling | Chris and Evan noticed that this check was compleatly... |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Chris Lattner | enable sinking and licm of loads from the argument... |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Chris Lattner | simplify some code by using shorter accessors. |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Chris Lattner | revert my previous patch. |
blob | commitdiff | raw | diff to current |
2008-01-05 |
Chris Lattner | factor some code better to avoid redundancy between |
blob | commitdiff | raw | diff to current |
2008-01-04 |
Owen Anderson | Move some more functionality from MRegisterInfo to... |
blob | commitdiff | raw | diff to current |
2008-01-02 |
Bill Wendling | Machine LICM will check that operands are defined outsi... |
blob | commitdiff | raw | diff to current |
2008-01-01 |
Owen Anderson | Move some more instruction creation methods from Regist... |
blob | commitdiff | raw | diff to current |
2008-01-01 |
Chris Lattner | Fix a bug in my previous patch: refer to the impl not... |
blob | commitdiff | raw | diff to current |
2008-01-01 |
Chris Lattner | Fix a problem where lib/Target/TargetInstrInfo.h would... |
blob | commitdiff | raw | diff to current |
2007-12-31 |
Owen Anderson | Move copyRegToReg from MRegisterInfo to TargetInstrInfo... |
blob | commitdiff | raw | diff to current |
2007-12-31 |
Chris Lattner | Rename SSARegMap -> MachineRegisterInfo in keeping... |
blob | commitdiff | raw | diff to current |
2007-12-30 |
Chris Lattner | Add new shorter predicates for testing machine operands... |
blob | commitdiff | raw | diff to current |
2007-12-30 |
Chris Lattner | Use MachineOperand::getImm instead of MachineOperand... |
blob | commitdiff | raw | diff to current |
2007-12-30 |
Bill Wendling | If we have a load of a global address that's not modifi... |
blob | commitdiff | raw | diff to current |
2007-12-29 |
Chris Lattner | Remove attribution from file headers, per discussion... |
blob | commitdiff | raw | diff to current |
2007-12-17 |
Bill Wendling | Add "mayHaveSideEffects" and "neverHasSideEffects"... |
blob | commitdiff | raw | diff to current |
2007-12-08 |
Bill Wendling | Reverting 44702. It wasn't correct to rename them. |
blob | commitdiff | raw | diff to current |
2007-12-08 |
Bill Wendling | Renaming: |
blob | commitdiff | raw | diff to current |
2007-10-11 |
Arnold Schwaighofer | Added tail call optimization to the x86 back end. It... |
blob | commitdiff | raw | diff to current |
2007-10-09 |
Evan Cheng | Under 64-bit mode use LEA64_32r instead of LEA64r to... |
blob | commitdiff | raw | diff to current |
2007-10-08 |
Evan Cheng | Allow x86 compare to be commutable by default. |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | Commute x86 cmove instructions by swapping the operands... |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | Enable convertToThreeAddress for X86 by default. |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | INC64_32r -> LEA64_32r is better than INC64_32r ->... |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | In 64-bit mode, avoid using leal with 32-bit 32-bit... |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | Add support to convert more 64-bit instructions to... |
blob | commitdiff | raw | diff to current |
2007-10-05 |
Evan Cheng | Testing convertToThreeeAddress as X86 llcbeta. |
blob | commitdiff | raw | diff to current |
2007-09-29 |
Evan Cheng | Enabling new condition code modeling scheme. |
blob | commitdiff | raw | diff to current |
2007-09-25 |
Evan Cheng | Added support for new condition code modeling scheme... |
blob | commitdiff | raw | diff to current |
2007-09-17 |
Dan Gohman | Add 64-bit jmp instructions to the list of instructions... |
blob | commitdiff | raw | diff to current |
2007-09-14 |
Dan Gohman | Add patterns for SHLD64* and SHRD64*. |
blob | commitdiff | raw | diff to current |
2007-09-14 |
Evan Cheng | Add implicit def of EFLAGS on those instructions that... |
blob | commitdiff | raw | diff to current |
2007-09-07 |
Owen Anderson | Add lengthof and endof templates that hide a lot of... |
blob | commitdiff | raw | diff to current |
2007-09-06 |
Evan Cheng | Fix a bug in X86InstrInfo::convertToThreeAddress that... |
blob | commitdiff | raw | diff to current |
2007-08-10 |
Christopher Lamb | Add 2-addr to 3-addr promotion code that allows 32... |
blob | commitdiff | raw | diff to current |
2007-07-26 |
Evan Cheng | Don't pollute the meaning of isUnpredicatedTerminator. |
blob | commitdiff | raw | diff to current |
2007-07-06 |
Evan Cheng | isUnpredicatedTerminator should treat conditional branc... |
blob | commitdiff | raw | diff to current |
2007-07-04 |
Dale Johannesen | Refactor X87 instructions. As a side effect, all |
blob | commitdiff | raw | diff to current |
2007-07-03 |
Dale Johannesen | Fix for PR 1505 (and 1489). Rewrite X87 register |
blob | commitdiff | raw | diff to current |
2007-06-26 |
Dan Gohman | Revert the earlier change that removed the M_REMATERIAL... |
blob | commitdiff | raw | diff to current |
2007-06-19 |
Dan Gohman | Replace M_REMATERIALIZIBLE and the newly-added isOtherR... |
blob | commitdiff | raw | diff to current |
2007-06-14 |
Dale Johannesen | Do not treat FP_REG_KILL as terminator in branch analys... |
blob | commitdiff | raw | diff to current |
2007-06-14 |
Dan Gohman | Add a target hook to allow loads from constant pools... |
blob | commitdiff | raw | diff to current |
2007-06-13 |
Dale Johannesen | Handle blocks with 2 unconditional branches in AnalyzeB... |
blob | commitdiff | raw | diff to current |
next |