- 68675c6 misched interface: rename Begin/End to RegionBegin/RegionEnd since they are not private. by Andrew Trick · 13 years ago
- ed395c8 misched prep: Expose the ScheduleDAGInstrs interface so targets may by Andrew Trick · 13 years ago
- cf46b5a misched prep: rename InsertPos to End. by Andrew Trick · 13 years ago
- 953be89 misched preparation: rename core scheduler methods for consistency. by Andrew Trick · 13 years ago
- 47c1445 misched preparation: clarify ScheduleDAG and ScheduleDAGInstrs roles. by Andrew Trick · 13 years ago
- 84b454d misched preparation: modularize schedule emission. by Andrew Trick · 13 years ago
- 73ba69b misched preparation: modularize schedule printing. by Andrew Trick · 13 years ago
- 4c72720 misched preparation: modularize schedule verification. by Andrew Trick · 13 years ago
- 9ebfbf8 Convert more GenRegisterInfo tables from unsigned to uint16_t to reduce static data size. by Craig Topper · 13 years ago
- b6bd8cc BitVectorize loop. by Benjamin Kramer · 13 years ago
- 46252d8 post-ra-sched: Turn the KillIndices vector into a bitvector, it only stored two meaningful states. by Benjamin Kramer · 13 years ago
- 49b726c post-ra-sched: Replace a std::set of regs with a bitvector. by Benjamin Kramer · 13 years ago
- 976647d Make calls scheduling boundaries post-ra. by Jakob Stoklund Olesen · 13 years ago
- f19a592 Handle regmasks in FixupKills. by Jakob Stoklund Olesen · 13 years ago
- 44d2382 Make all pointers to TargetRegisterClass const since they are all pointers to static data that should not be modified. by Craig Topper · 13 years ago
- 1dd8c85 Codegen pass definition cleanup. No functionality. by Andrew Trick · 13 years ago
- c7d081b Move pass configuration out of pass constructors: PostRAScheduler. by Andrew Trick · 13 years ago
- 5e920d7 misched: Added ScheduleDAGInstrs::IsPostRA by Andrew Trick · 13 years ago
- ddfd137 - Add MachineInstrBundle.h and MachineInstrBundle.cpp. This includes a function by Evan Cheng · 13 years ago
- 5a96b3d Add bundle aware API for querying instruction properties and switch the code by Evan Cheng · 13 years ago
- a7b0cb7 Remove all remaining uses of Value::getNameStr(). by Benjamin Kramer · 13 years ago
- 5b1b4489 Rename TargetSubtarget to TargetSubtargetInfo for consistency. by Evan Cheng · 13 years ago
- fa796dd Teach antidependency breakers to use RegisterClassInfo. by Jakob Stoklund Olesen · 13 years ago
- e29e8e1 Update DBG_VALUEs while breaking anti dependencies. by Devang Patel · 13 years ago
- cf9aa28 Add an issue width check to the postRA scheduler. Patch by Max Kazakov! by Andrew Trick · 13 years ago
- 89fd437 Typo: Reviewed by Alistair. by Andrew Trick · 13 years ago
- 15ab359 Post-RA scheduler compile time fix. Quadratic computation of DAG node depth. by Andrew Trick · 13 years ago
- 2da8bc8 Various bits of framework needed for precise machine-level selection by Andrew Trick · 14 years ago
- 3ef1c87 Teach if-converter to be more careful with predicating instructions that would by Evan Cheng · 14 years ago
- 90c579d Reapply r110396, with fixes to appease the Linux buildbot gods. by Owen Anderson · 14 years ago
- 1f74590 Revert r110396 to fix buildbots. by Owen Anderson · 14 years ago
- 9ccaf53 Don't use PassInfo* as a type identifier for passes. Instead, use the address of the static by Owen Anderson · 14 years ago
- 24173da Use std::vector instead of TargetRegisterInfo::FirstVirtualRegister. by Bill Wendling · 14 years ago
- 86050dc Allow ARM if-converter to be run after post allocation scheduling. by Evan Cheng · 14 years ago
- 774bc88 - Do away with SimpleHazardRecognizer.h. It's not used and offers little value. by Evan Cheng · 14 years ago
- 729aab3 Allow target to provide its own hazard recognizer to post-ra scheduler. by Evan Cheng · 14 years ago
- 1015ba7 - Change MachineInstr::findRegisterDefOperandIdx so it can also look for defs by Evan Cheng · 14 years ago
- 76526f8 Remove dbg_value workaround and associated command line option by Jim Grosbach · 14 years ago
- de70b1f Enable preserving debug information through post-RA scheduling by Jim Grosbach · 14 years ago
- 9001303 80 column and trailing whitespace cleanup by Jim Grosbach · 14 years ago
- 5468e09 add cmd line option to leave dbgvalues in during post-RA sceduling. Useful by Jim Grosbach · 14 years ago
- af1d8ca Get rid of the EdgeMapping map. Instead, just check for BasicBlock by Dan Gohman · 14 years ago
- 8295d4c As a temporary workaround for post-RA not handling DebugValue instructions, by Bob Wilson · 14 years ago
- 8b3d668 Remove a #include. by Dan Gohman · 14 years ago
- b0812f1 Fix some more places where dbg_value affected codegen. by Dale Johannesen · 15 years ago
- e1b2129 Change errs() to dbgs(). by David Greene · 15 years ago
- 990d285 <rdar://problem/7453528>. Track only physical registers that are valid for the target. by David Goodwin · 15 years ago
- 15d75d9 Don't hang on to pointers or references after vector::push_back. by Jakob Stoklund Olesen · 15 years ago
- 557bbe6 Remove some old experimental code that is no longer needed. Remove additional, speculative scheduling pass as its cost did not translate into significant performance improvement. Minor tweaks. by David Goodwin · 15 years ago
- 87d21b9 Allow target to specify regclass for which antideps will only be broken along the critical path. by David Goodwin · 15 years ago
- 12dd99d Rename registers to break output dependencies in addition to anti-dependencies. by David Goodwin · 15 years ago
- c2e8a7e Fixed to address code review. No functional changes. by David Goodwin · 15 years ago
- 0855dee Allow targets to specify register classes whose member registers should not be renamed to break anti-dependencies. by David Goodwin · 15 years ago
- 5409783 Break anti-dependencies using free registers in a round-robin manner to avoid introducing new anti-dependencies. by David Goodwin · 15 years ago
- 4de099d Do a scheduling pass ignoring anti-dependencies to identify candidate registers that should be renamed. by David Goodwin · 15 years ago
- 0ba90f3 Make -print-machineinstrs more readable. by Dan Gohman · 15 years ago
- 82c7248 Make AntiDepReg.h internal. by David Goodwin · 15 years ago
- e10deca Allow the aggressive anti-dep breaker to process the same region multiple times. This is necessary because new anti-dependencies are exposed when "current" ones are broken. by David Goodwin · 15 years ago
- ada0ef8 Define virtual destructor in *.cpp file. by David Goodwin · 15 years ago
- 3487771 Add aggressive anti-dependence breaker. Currently it is not the default for any target. Enable with -break-anti-dependencies=all. by David Goodwin · 15 years ago
- 2e7be61 Break anti-dependence breaking out into its own class. by David Goodwin · 15 years ago
- f5a86f4 Remove includes of Support/Compiler.h that are no longer needed after the by Nick Lewycky · 15 years ago
- 6726b6d Remove VISIBILITY_HIDDEN from class/struct found inside anonymous namespaces. by Nick Lewycky · 15 years ago
- 4c3715c Allow the target to select the level of anti-dependence breaking that should be performed by the post-RA scheduler. The default is none. by David Goodwin · 15 years ago
- c1ae8c9 Revert r84658 and r84691. They were causing llvm-gcc bootstrap to fail. by Dan Gohman · 15 years ago
- 7441d14 Respect src register allocation requirements when breaking anti-dependencies. Remove some dead code. by David Goodwin · 15 years ago
- 480c529 Checkpoint more aggressive anti-dependency breaking for post-ra scheduler. by David Goodwin · 15 years ago
- fa16354 Change createPostRAScheduler so it can be turned off at llc -O1. by Evan Cheng · 15 years ago
- c83da2f9 If post-alloc scheduler is not enabled, it should return false, not true. by Evan Cheng · 15 years ago
- d452ea6 Add debugging output. by David Goodwin · 15 years ago
- 5bf7c2a Fix a missing initialization of PostRAScheduler's AA member. by Dan Gohman · 15 years ago
- a70dca1 Factor out LiveIntervalAnalysis' code to determine whether an instruction by Dan Gohman · 15 years ago
- 8bff4af Fix a use-after-free in post-ra-scheduling. by Benjamin Kramer · 15 years ago
- 63bcbb7 All callee-saved registers are live-out of a return block. by David Goodwin · 15 years ago
- 9843a93 Remove neonfp attribute and instead set default based on CPU string. Add -arm-use-neon-fp to override the default. by David Goodwin · 15 years ago
- 471850a Restore the -post-RA-scheduler flag as an override for the target specification. Remove -mattr for setting PostRAScheduler enable and instead use CPU string. by David Goodwin · 15 years ago
- c7951f8 Use MachineFrameInfo.getPristineRegs() to determine which callee-saved registers are available for anti-dependency breaking. Some cleanup. by David Goodwin · 15 years ago
- 714e8bc Observe hasExtraSrcRegAllocReq and hasExtraDefRegAllocReq. Do not change by Evan Cheng · 15 years ago
- c277ab0 Fix integer overflow in instruction scheduling. This can happen if we have by Reid Kleckner · 15 years ago
- 0dad89f Remove -post-RA-schedule flag and add a TargetSubtarget method to enable post-register-allocation scheduling. By default it is off. For ARM, enable/disable with -mattr=+/-postrasched. Enable by default for cortex-a8. by David Goodwin · 15 years ago
- 544df36 Use KILL instead of IMPLICIT_DEF in LowerSubregs pass. by Jakob Stoklund Olesen · 15 years ago
- 8f90934 Fix bug in kill flag updating for post-register-allocation scheduling. When the kill flag of a superreg needs to be cleared because there are one or more subregs live, we instead add implicit-defs of those subregs and leave the kill flag on the superreg. This allows us to end the live-range of the superreg without ending the live-ranges of the subregs. by David Goodwin · 15 years ago
- fb2e752 Enhance EmitInstrWithCustomInserter() so target can specify CFG changes that sdisel will use to properly complete phi nodes. by Evan Cheng · 15 years ago
- be441c0 It's a bool, so treat it like one. Fixes a MSVC warning. by Benjamin Kramer · 15 years ago
- 5e41178 Create our own block initializer for kill fixups as the scheduling one wasn't doing the right thing. by David Goodwin · 15 years ago
- 1f15228 Add hidden flags to allow binary search of post-RA scheduling errors. by David Goodwin · 15 years ago
- a3251db Don't mark a register live at an undef use. by David Goodwin · 15 years ago
- 7886cd8 Another stab at fixing up register kill flags after post-RA scheduling. by David Goodwin · 15 years ago
- 88a589c Fixup register kills after scheduling. by David Goodwin · 15 years ago
- 103289e convert LoopInfo.h and GraphWriter.h to use raw_ostream by Chris Lattner · 15 years ago
- 2ffb0ce Fix counting of Post-RA scheduling stalls. Improve debug output. by David Goodwin · 15 years ago
- da27757 This logic was accidentally inverted in r78767. by Dan Gohman · 15 years ago
- 26255ad Factor out the code for finding an available register for use by Dan Gohman · 15 years ago
- 7cd0118 Use DEBUG macro for debug output. by David Goodwin · 15 years ago
- c93d837 Add some debug output. by David Goodwin · 15 years ago
- 3a5f0d4 Replace DOUT. by David Goodwin · 15 years ago
- d94a4e5 Post RA scheduler changes. Introduce a hazard recognizer that uses the target schedule information to accurately model the pipeline. Update the scheduler to correctly handle multi-issue targets. by David Goodwin · 15 years ago
- 845012e Use setPreservesAll and setPreservesCFG in CodeGen passes. by Dan Gohman · 15 years ago
- 2a38688 inline the global 'getInstrOperandRegClass' function into its callers by Chris Lattner · 15 years ago
- c23197a llvm_unreachable->llvm_unreachable(0), LLVM_UNREACHABLE->llvm_unreachable. by Torok Edwin · 15 years ago