2011-12-07 |
Eli Friedman | Make sure we correctly set LiveRegGens when a call... |
blob | commitdiff | raw |
2011-12-07 |
Eli Friedman | Fix an assertion in the scheduler. PR11386. No testca... |
blob | commitdiff | raw | diff to current |
2011-12-07 |
Nick Lewycky | These global variables aren't thread-safe, STATISTIC... |
blob | commitdiff | raw | diff to current |
2011-11-16 |
Owen Anderson | Rename MVT::untyped to MVT::Untyped to match similar... |
blob | commitdiff | raw | diff to current |
2011-11-15 |
Pete Cooper | Added custom lowering for load->dec->store sequence... |
blob | commitdiff | raw | diff to current |
2011-11-10 |
Evan Cheng | Use a bigger hammer to fix PR11314 by disabling the... |
blob | commitdiff | raw | diff to current |
2011-11-09 |
Duncan Sands | Speculatively revert commit 144124 (djg) in the hope... |
blob | commitdiff | raw | diff to current |
2011-11-08 |
Dan Gohman | Add a hack to the scheduler to disable pseudo-two-addre... |
blob | commitdiff | raw | diff to current |
2011-11-03 |
Dan Gohman | Reapply r143206, with fixes. Disallow physical register... |
blob | commitdiff | raw | diff to current |
2011-10-29 |
Dan Gohman | Revert r143206, as there are still some failing tests. |
blob | commitdiff | raw | diff to current |
2011-10-28 |
Dan Gohman | Reapply r143177 and r143179 (reverting r143188), with... |
blob | commitdiff | raw | diff to current |
2011-10-28 |
Duncan Sands | Speculatively disable Dan's commits 143177 and 143179... |
blob | commitdiff | raw | diff to current |
2011-10-28 |
Dan Gohman | Eliminate LegalizeOps' LegalizedNodes map and have... |
blob | commitdiff | raw | diff to current |
2011-10-24 |
Dan Gohman | Change this overloaded use of Sched::Latency to be... |
blob | commitdiff | raw | diff to current |
2011-10-21 |
Chandler Carruth | Remove a now dead function, fixing -Wunused-function... |
blob | commitdiff | raw | diff to current |
2011-10-20 |
Dan Gohman | Delete the list-tdrr scheduler. Top-down schedulers... |
blob | commitdiff | raw | diff to current |
2011-09-01 |
Andrew Trick | PreRA scheduler should avoid cloning compares. |
blob | commitdiff | raw | diff to current |
2011-06-28 |
Evan Cheng | - Rename TargetInstrDesc, TargetOperandInfo to MCInstrD... |
blob | commitdiff | raw | diff to current |
2011-06-27 |
Evan Cheng | More refactoring. Move getRegClass from TargetOperandIn... |
blob | commitdiff | raw | diff to current |
2011-06-27 |
Andrew Trick | pre-RA-sched: Cleanup register pressure tracking. |
blob | commitdiff | raw | diff to current |
2011-06-27 |
Jakob Stoklund Olesen | Distinguish early clobber output operands from clobbere... |
blob | commitdiff | raw | diff to current |
2011-06-21 |
Owen Anderson | Fix some trailing issues from my introduction of MVT... |
blob | commitdiff | raw | diff to current |
2011-06-18 |
Benjamin Kramer | Remove unused but set variables. |
blob | commitdiff | raw | diff to current |
2011-06-15 |
Owen Anderson | Add a new MVT::untyped. This will be used in future... |
blob | commitdiff | raw | diff to current |
2011-06-15 |
Andrew Trick | Added -stress-sched flag in the Asserts build. |
blob | commitdiff | raw | diff to current |
2011-06-08 |
Andrew Trick | Remove a temporary test case probe in CheckForLiveRegDef. |
blob | commitdiff | raw | diff to current |
2011-06-07 |
Andrew Trick | Fix a merge bug in preRAsched for handling physreg... |
blob | commitdiff | raw | diff to current |
2011-04-26 |
Evan Cheng | Be careful about scheduling nodes above previous calls... |
blob | commitdiff | raw | diff to current |
2011-04-26 |
Evan Cheng | Fix typo |
blob | commitdiff | raw | diff to current |
2011-04-14 |
Andrew Trick | In the pre-RA scheduler, maintain cmp+br proximity. |
blob | commitdiff | raw | diff to current |
2011-04-13 |
Andrew Trick | Recommit r129383. PreRA scheduler heuristic fixes:... |
blob | commitdiff | raw | diff to current |
2011-04-12 |
Andrew Trick | Revert 129383. It causes some targets to hit a schedule... |
blob | commitdiff | raw | diff to current |
2011-04-12 |
Andrew Trick | PreRA scheduler heuristic fixes: VRegCycle, TokenFactor... |
blob | commitdiff | raw | diff to current |
2011-04-07 |
Andrew Trick | Added a check in the preRA scheduler for potential... |
blob | commitdiff | raw | diff to current |
2011-03-25 |
Andrew Trick | Fix for -pre-RA-sched=source. |
blob | commitdiff | raw | diff to current |
2011-03-23 |
Andrew Trick | Ensure that def-side physreg copies are scheduled above... |
blob | commitdiff | raw | diff to current |
2011-03-23 |
Andrew Trick | whitespace |
blob | commitdiff | raw | diff to current |
2011-03-21 |
Eric Christopher | Grammar-o. |
blob | commitdiff | raw | diff to current |
2011-03-10 |
Evan Cheng | Re-commit 127368 and 127371. They are exonerated. |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Evan Cheng | Revert 127368 and 127371 for now. |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Evan Cheng | Change the definition of TargetRegisterInfo::getCrossCo... |
blob | commitdiff | raw | diff to current |
2011-03-09 |
Benjamin Kramer | Fix typo, make helper static. |
blob | commitdiff | raw | diff to current |
2011-03-08 |
Eric Christopher | Fix some latent bugs if the nodes are unschedulable... |
blob | commitdiff | raw | diff to current |
2011-03-08 |
Andrew Trick | Further improvements to pre-RA-sched=list-ilp. |
blob | commitdiff | raw | diff to current |
2011-03-07 |
Cameron Zwarich | Move getRegPressureLimit() from TargetLoweringInfo... |
blob | commitdiff | raw | diff to current |
2011-03-06 |
Eric Christopher | Typo. |
blob | commitdiff | raw | diff to current |
2011-03-06 |
Andrew Trick | Disable a couple of experimental heuristics to get... |
blob | commitdiff | raw | diff to current |
2011-03-05 |
Andrew Trick | Be explicit with abs(). Visual Studio workaround. |
blob | commitdiff | raw | diff to current |
2011-03-05 |
Andrew Trick | Missing comment. |
blob | commitdiff | raw | diff to current |
2011-03-05 |
Andrew Trick | Increased the register pressure limit on x86_64 from... |
blob | commitdiff | raw | diff to current |
2011-03-04 |
Andrew Trick | Minor pre-RA-sched fixes and cleanup. |
blob | commitdiff | raw | diff to current |
2011-02-04 |
Andrew Trick | Introducing a new method of tracking register pressure... |
blob | commitdiff | raw | diff to current |
2011-01-27 |
Andrew Trick | Remove a temporary workaround for a lencod miscompile... |
blob | commitdiff | raw | diff to current |
2011-01-24 |
Andrew Trick | Temporarily workaround JM/lencod miscompile (SIGSEGV). |
blob | commitdiff | raw | diff to current |
2011-01-21 |
Andrew Trick | Enable support for precise scheduling of the instructio... |
blob | commitdiff | raw | diff to current |
2011-01-21 |
Andrew Trick | Convert -enable-sched-cycles and -enable-sched-hazard... |
blob | commitdiff | raw | diff to current |
2011-01-20 |
Andrew Trick | Selection DAG scheduler register pressure heuristic... |
blob | commitdiff | raw | diff to current |
2011-01-14 |
Andrew Trick | Support for precise scheduling of the instruction selec... |
blob | commitdiff | raw | diff to current |
2010-12-24 |
Andrew Trick | Minor cleanup related to my latest scheduler changes. |
blob | commitdiff | raw | diff to current |
2010-12-24 |
Andrew Trick | Fix a few cases where the scheduler is not checking... |
blob | commitdiff | raw | diff to current |
2010-12-24 |
Andrew Trick | Various bits of framework needed for precise machine... |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Chris Lattner | flags -> glue for selectiondag |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Andrew Trick | Reorganize ListScheduleBottomUp in preparation for... |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Andrew Trick | Converted LiveRegCycles to LiveRegGens. It's easier... |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Andrew Trick | In CheckForLiveRegDef use TRI->getOverlaps. |
blob | commitdiff | raw | diff to current |
2010-12-23 |
Andrew Trick | Fixes PR8823: add-with-overflow-128.ll |
blob | commitdiff | raw | diff to current |
2010-12-21 |
Andrew Trick | In DelayForLiveRegsBottomUp, handle instructions that... |
blob | commitdiff | raw | diff to current |
2010-12-21 |
Andrew Trick | whitespace |
blob | commitdiff | raw | diff to current |
2010-12-21 |
Chris Lattner | rename MVT::Flag to MVT::Glue. "Flag" is a terrible... |
blob | commitdiff | raw | diff to current |
2010-12-20 |
Chris Lattner | Fix a bug in the scheduler's handling of "unspillable... |
blob | commitdiff | raw | diff to current |
2010-12-20 |
Chris Lattner | the result of CheckForLiveRegDef is dead, remove it. |
blob | commitdiff | raw | diff to current |
2010-11-03 |
Evan Cheng | Two sets of changes. Sorry they are intermingled. |
blob | commitdiff | raw | diff to current |
2010-10-29 |
Evan Cheng | Avoiding overly aggressive latency scheduling. If the... |
blob | commitdiff | raw | diff to current |
2010-07-26 |
Evan Cheng | The "excess register pressure" returned by HighRegPress... |
blob | commitdiff | raw | diff to current |
2010-07-26 |
Duncan Sands | Pacify gcc-4.5 which wrongly thinks that RExcess (passe... |
blob | commitdiff | raw | diff to current |
2010-07-25 |
Evan Cheng | Add comments. |
blob | commitdiff | raw | diff to current |
2010-07-25 |
Bob Wilson | Fix crashes when scheduling a CopyToReg node -- getMach... |
blob | commitdiff | raw | diff to current |
2010-07-24 |
Evan Cheng | Add an ILP scheduler. This is a register pressure aware... |
blob | commitdiff | raw | diff to current |
2010-07-23 |
Evan Cheng | - Allow target to specify when is register pressure... |
blob | commitdiff | raw | diff to current |
2010-07-22 |
Evan Cheng | Re-apply r109079 with fix. |
blob | commitdiff | raw | diff to current |
2010-07-22 |
Owen Anderson | Revert r109079, which broke a lot of CodeGen tests. |
blob | commitdiff | raw | diff to current |
2010-07-22 |
Evan Cheng | Initialize RegLimit only when register pressure is... |
blob | commitdiff | raw | diff to current |
2010-07-21 |
Evan Cheng | More register pressure aware scheduling work. |
blob | commitdiff | raw | diff to current |
2010-07-21 |
Evan Cheng | Teach bottom up pre-ra scheduler to track register... |
blob | commitdiff | raw | diff to current |
2010-06-29 |
Rafael Espindola | Add a VT argument to getMinimalPhysRegClass and replace... |
blob | commitdiff | raw | diff to current |
2010-05-30 |
Oscar Fuentes | Use `llvm::next' instead of `next' to make VC++ 2010... |
blob | commitdiff | raw | diff to current |
2010-05-28 |
Evan Cheng | Fix some latency computation bugs: if the use is not... |
blob | commitdiff | raw | diff to current |
2010-05-26 |
Dan Gohman | Eliminate the use of PriorityQueue and just use a std... |
blob | commitdiff | raw | diff to current |
2010-05-26 |
Dan Gohman | Delete an unused function. |
blob | commitdiff | raw | diff to current |
2010-05-26 |
Dan Gohman | Change push_all to a non-virtual function and implement... |
blob | commitdiff | raw | diff to current |
2010-05-21 |
Evan Cheng | Rename -pre-RA-sched=hybrid to -pre-RA-sched=list-hybrid. |
blob | commitdiff | raw | diff to current |
2010-05-20 |
Evan Cheng | Allow targets more controls on what nodes are scheduled... |
blob | commitdiff | raw | diff to current |
2010-05-20 |
Evan Cheng | Add a hybrid bottom up scheduler that reduce register... |
blob | commitdiff | raw | diff to current |
2010-04-07 |
Chris Lattner | Three changes: |
blob | commitdiff | raw | diff to current |
2010-02-09 |
Chris Lattner | move target-independent opcodes out of TargetInstrInfo |
blob | commitdiff | raw | diff to current |
2010-02-05 |
Evan Cheng | When the scheduler unfold a load folding instruction... |
blob | commitdiff | raw | diff to current |
2010-01-23 |
Bill Wendling | Remove the '-disable-scheduling' flag and replace it... |
blob | commitdiff | raw | diff to current |
2010-01-06 |
Bill Wendling | The previous code could potentially cause a cycle.... |
blob | commitdiff | raw | diff to current |
2010-01-06 |
Bill Wendling | Only check the ordering if there is an ordering for... |
blob | commitdiff | raw | diff to current |
2010-01-05 |
Bill Wendling | Add a semi-primitive form of scheduling via the "SDNode... |
blob | commitdiff | raw | diff to current |
next |