2010-07-21 |
Evan Cheng | Teach bottom up pre-ra scheduler to track register...
|
commit | commitdiff | tree |
2010-07-19 |
Evan Cheng | ARM has to provide its own TargetLowering::findRepresentativ...
|
commit | commitdiff | tree |
2010-07-19 |
Evan Cheng | Teach computeRegisterProperties() to compute "representative...
|
commit | commitdiff | tree |
2010-07-15 |
Evan Cheng | Split -enable-finite-only-fp-math to two options:
|
commit | commitdiff | tree |
2010-07-14 |
Evan Cheng | Fix for PR7193 was overly conservative. The only case...
|
commit | commitdiff | tree |
2010-07-14 |
Evan Cheng | Re-enable the test with fix.
|
commit | commitdiff | tree |
2010-07-14 |
Evan Cheng | Teach ProcessImplicitDefs to transform more COPY instruction...
|
commit | commitdiff | tree |
2010-07-13 |
Evan Cheng | Update comment.
|
commit | commitdiff | tree |
2010-07-13 |
Evan Cheng | Extend the r107852 optimization which turns some fp...
|
commit | commitdiff | tree |
2010-07-13 |
Evan Cheng | Add an ARM "feature". Cortex-a8 fp comparison is very...
|
commit | commitdiff | tree |
2010-07-13 |
Evan Cheng | -enable-unsafe-fp-math should not imply -enable-finite...
|
commit | commitdiff | tree |
2010-07-08 |
Evan Cheng | Check for FiniteOnlyFPMath as well.
|
commit | commitdiff | tree |
2010-07-08 |
Evan Cheng | r107852 is only safe with -enable-unsafe-fp-math to...
|
commit | commitdiff | tree |
2010-07-08 |
Evan Cheng | Optimize some vfp comparisons to integer ones. This...
|
commit | commitdiff | tree |
2010-07-07 |
Evan Cheng | Move getExtLoad() and (some) getLoad() DebugLoc argument...
|
commit | commitdiff | tree |
2010-07-04 |
Evan Cheng | Infer alignments of fixed frame objects when they are...
|
commit | commitdiff | tree |
2010-07-03 |
Evan Cheng | Remove isSS argument from CreateFixedObject. Fixed...
|
commit | commitdiff | tree |
2010-07-02 |
Evan Cheng | Remove early IT block formation. It's not used.
|
commit | commitdiff | tree |
2010-07-02 |
Evan Cheng | - Two-address pass should not assume unfolding is always...
|
commit | commitdiff | tree |
2010-06-29 |
Evan Cheng | PR7503: uxtb16 is not available for ARMv7-M. Patch...
|
commit | commitdiff | tree |
2010-06-29 |
Evan Cheng | Change if-cvt options to something that actually as...
|
commit | commitdiff | tree |
2010-06-26 |
Evan Cheng | It's now possible to run code placement pass for ARM.
|
commit | commitdiff | tree |
2010-06-25 |
Evan Cheng | Change if-conversion block size limit checks to add...
|
commit | commitdiff | tree |
2010-06-24 |
Evan Cheng | Oops. IT block formation pass needs to be run at any...
|
commit | commitdiff | tree |
2010-06-22 |
Evan Cheng | Tail merging pass shall not break up IT blocks. rdar...
|
commit | commitdiff | tree |
2010-06-21 |
Evan Cheng | Fix PR7421: bug in kill transferring logic. It was...
|
commit | commitdiff | tree |
2010-06-20 |
Evan Cheng | Fix a crash caused by dereference of MBB.end(). rdar...
|
commit | commitdiff | tree |
2010-06-19 |
Evan Cheng | Ignore dbg_value's.
|
commit | commitdiff | tree |
2010-06-19 |
Evan Cheng | Disable sibcall optimization for Thumb1 for now since...
|
commit | commitdiff | tree |
2010-06-19 |
Evan Cheng | Indentation and remove dead code.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Move ARM if-conversion before post-ra scheduling.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Update cmake list.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Thumb2 hazard recognizer.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Allow ARM if-converter to be run after post allocation...
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Fix an inverted condition.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Fix cross initialization compilation error.
|
commit | commitdiff | tree |
2010-06-18 |
Evan Cheng | Teach iff-converter to properly count # of dups. It...
|
commit | commitdiff | tree |
2010-06-16 |
Evan Cheng | Make post-ra scheduling, anti-dep breaking, and register...
|
commit | commitdiff | tree |
2010-06-14 |
Evan Cheng | - Do away with SimpleHazardRecognizer.h. It's not used...
|
commit | commitdiff | tree |
2010-06-14 |
Evan Cheng | Avoid uncessary array copying.
|
commit | commitdiff | tree |
2010-06-12 |
Evan Cheng | Allow target to provide its own hazard recognizer to...
|
commit | commitdiff | tree |
2010-06-12 |
Evan Cheng | Code formatting.
|
commit | commitdiff | tree |
2010-06-10 |
Evan Cheng | Code refactoring, no functionality changes.
|
commit | commitdiff | tree |
2010-06-10 |
Evan Cheng | Delete code that's not safe.
|
commit | commitdiff | tree |
2010-06-09 |
Evan Cheng | Allow target to place 2-address pass inserted copies...
|
commit | commitdiff | tree |
2010-06-09 |
Evan Cheng | Revert 105540, 105542, 105544, 105546, and 105548 to...
|
commit | commitdiff | tree |
2010-06-09 |
Evan Cheng | Typo.
|
commit | commitdiff | tree |
2010-06-09 |
Evan Cheng | Thumb2 IT blocks are fairly expensive. When there are...
|
commit | commitdiff | tree |
2010-06-04 |
Evan Cheng | Re-apply 105308 with fix.
|
commit | commitdiff | tree |
2010-06-02 |
Evan Cheng | Enable machine cse of instructions which define physical...
|
commit | commitdiff | tree |
2010-05-29 |
Evan Cheng | Remove schedule-livein-copies. It's not being used.
|
commit | commitdiff | tree |
2010-05-29 |
Evan Cheng | Fix PR7193: if sibling call address can take a register...
|
commit | commitdiff | tree |
2010-05-29 |
Evan Cheng | Doh. Machine LICM is re-initializing the CSE map over...
|
commit | commitdiff | tree |
2010-05-28 |
Evan Cheng | Fix some latency computation bugs: if the use is not...
|
commit | commitdiff | tree |
2010-05-28 |
Evan Cheng | Schedule high latency instructions for latency reduction...
|
commit | commitdiff | tree |
2010-05-27 |
Evan Cheng | Use report_fatal_error, not llvm_unreachable.
|
commit | commitdiff | tree |
2010-05-27 |
Evan Cheng | llvm can't correctly support 'H', 'Q' and 'R' modifiers...
|
commit | commitdiff | tree |
2010-05-24 |
Evan Cheng | Avoid adding duplicate function live-in's.
|
commit | commitdiff | tree |
2010-05-24 |
Evan Cheng | LR is in GPR, not tGPR even in Thumb1 mode.
|
commit | commitdiff | tree |
2010-05-22 |
Evan Cheng | Implement @llvm.returnaddress. rdar://8015977.
|
commit | commitdiff | tree |
2010-05-21 |
Evan Cheng | Allow machine cse to cse instructions which define...
|
commit | commitdiff | tree |
2010-05-21 |
Evan Cheng | - Change MachineInstr::findRegisterDefOperandIdx so...
|
commit | commitdiff | tree |
2010-05-21 |
Evan Cheng | Change ARM scheduling default to list-hybrid if the...
|
commit | commitdiff | tree |
2010-05-21 |
Evan Cheng | Rename -pre-RA-sched=hybrid to -pre-RA-sched=list-hybrid.
|
commit | commitdiff | tree |
2010-05-20 |
Evan Cheng | Allow targets more controls on what nodes are scheduled...
|
commit | commitdiff | tree |
2010-05-20 |
Evan Cheng | Add a hybrid bottom up scheduler that reduce register...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Code clean up.
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Code refactoring: pull SchedPreference enum from TargetLower...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | t2LEApcrel and tLEApcrel are re-materializable. This...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Use 'adr' for LEApcrel and LEApcrel. Mark LEApcrel...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Mark pattern-less mayLoad / mayStore instructions neverHasSi...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Target instruction selection should copy memoperands.
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Mark a few more pattern-less instructions with neverHasSideE...
|
commit | commitdiff | tree |
2010-05-19 |
Evan Cheng | Intrinsics which do a vector compare (results are all...
|
commit | commitdiff | tree |
2010-05-18 |
Evan Cheng | Sink dag combine's post index load / store code that...
|
commit | commitdiff | tree |
2010-05-18 |
Evan Cheng | Continuously refine the register class of REG_SEQUENCE...
|
commit | commitdiff | tree |
2010-05-18 |
Evan Cheng | Fix PR7162: Use source register classes and sub-indices...
|
commit | commitdiff | tree |
2010-05-18 |
Evan Cheng | FIX PR7158. SimplifyVBinOp was asserting when it fails...
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | Fix PR7175. Insert copies of a REG_SEQUENCE source...
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | Fix PR7156. If the sources of a REG_SEQUENCE are all...
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | vmov of immediates are trivially re-materializable.
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | Careful with reg_sequence coalescing to not to overwrite...
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | Turn on -neon-reg-sequence by default.
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | Yes, if the redef is a copy, update the old val# with...
|
commit | commitdiff | tree |
2010-05-17 |
Evan Cheng | No reason not to run the NEON domain croassing fix...
|
commit | commitdiff | tree |
2010-05-16 |
Evan Cheng | Model vst lane instructions with REG_SEQUENCE.
|
commit | commitdiff | tree |
2010-05-15 |
Evan Cheng | Model 128-bit vld lane with REG_SEQUENCE.
|
commit | commitdiff | tree |
2010-05-15 |
Evan Cheng | v4i64 and v8i64 are only synthesizable when NEON is...
|
commit | commitdiff | tree |
2010-05-15 |
Evan Cheng | Allow TargetLowering::getRegClassFor() to be called...
|
commit | commitdiff | tree |
2010-05-15 |
Evan Cheng | Model 64-bit lane vld with REG_SEQUENCE.
|
commit | commitdiff | tree |
2010-05-15 |
Evan Cheng | A partial re-def instruction may be a copy.
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Teach two-address pass to do some coalescing while...
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Model VST*_UPD and VST*oddUPD pair with REG_SEQUENCE.
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Model VLD*_UPD and VLD*odd_UPD pair with REG_SEQUENCE.
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Get rid of the bit twiddling to read / set OpActions...
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Added a QQQQ register file to model 4-consecutive Q...
|
commit | commitdiff | tree |
2010-05-14 |
Evan Cheng | Fix comments.
|
commit | commitdiff | tree |
2010-05-13 |
Evan Cheng | Adding a v8i64 512-bit vector type. This will be used...
|
commit | commitdiff | tree |
2010-05-13 |
Evan Cheng | Eliminate use of magic numbers to access OpActions...
|
commit | commitdiff | tree |
2010-05-13 |
Evan Cheng | Fix up LoadExtActions, TruncStoreActions, and IndexedModeAct...
|
commit | commitdiff | tree |
next |