- 6679906 Avoid creating shift-by-zero SDNodes in the common case of by Dan Gohman · 17 years ago
- 3c8f36f Reapply r55191 and r55192. by Dan Gohman · 17 years ago
- 449416d Reverting r55190, r55191, and r55192. They broke the build with this error message: by Bill Wendling · 17 years ago
- e0182ec Support non-fallthrough unconditional branches in FastISel. by Dan Gohman · 17 years ago
- 83785c8 First cut, un-optimized (and untested) fast isel lowering of GetElementPtrInst. by Evan Cheng · 17 years ago
- bb46633 Simplify FastISel's constructor argument list, make the FastISel by Dan Gohman · 17 years ago
- 2385852 Dump the instruction that foiled ISel even when -debug is not used. by Dan Gohman · 17 years ago
- d420789 Kill off SimpleBBISel, it's replaced by FastISel. by Evan Cheng · 17 years ago
- a267651 Disable DAGCombine's alignment inference in "fast" codegen mode. by Dan Gohman · 17 years ago
- c53ec49 Use BitVector instead of std::vector<unsigned char>. by Dan Gohman · 17 years ago
- cc8430f Fix a leak in the FastISel code that Chris pointed out. by Dan Gohman · 17 years ago
- 3e697cf Add support for running SelectionDAG if FastISel fails. This is under by Dan Gohman · 17 years ago
- 78eca17 Add code to call FastISel, and a command-line option to enable it. by Dan Gohman · 17 years ago
- 5eca075 Rename some GC classes so that their roll will hopefully be clearer. by Gordon Henriksen · 17 years ago
- 5a29c9e Factor GC metadata table assembly generation out of Collector in preparation for splitting AsmPrinter into its own library. by Gordon Henriksen · 17 years ago
- 925a7e8 Rename SelectionDAGISel's FastISel to Fast, to begin to make by Dan Gohman · 17 years ago
- 381cb07 Add skeleton of simple basic block instruction selector. by Evan Cheng · 17 years ago
- 39fd6e8 Factor code that finalize PHI nodes, jump tables, etc. out of SelectBasicBlock. No functionality changes. by Evan Cheng · 17 years ago
- 3bc9c6f Remove the -disable-correct-folding option, which was ugly and is no longer needed. by Owen Anderson · 17 years ago
- 6ac8df7 This option doesn't need to be a target option. It can be in SDISel instead. by Owen Anderson · 17 years ago
- bd3ba46 - Fix SelectionDAG to generate correct CFGs. by Owen Anderson · 17 years ago
- 6f498b0 Fix SDISel lowering of PHI nodes to use ComputeValueVTs. by Dan Gohman · 17 years ago
- 1f565bc Fix SDISel lowering of zeroinitializer and undef to use ComputeValueVTs. by Dan Gohman · 17 years ago
- 7232464 Add a flag to disable jump table generation (all by Dale Johannesen · 17 years ago
- 90d33ee Don't look for leaf values to store when lowering stores of by Dan Gohman · 17 years ago
- 5bc1ea0 Add vector shifts to the IR, patch by Eli Friedman. by Nate Begeman · 17 years ago
- fed90b6 Fold the useful features of alist and alist_node into ilist, and by Dan Gohman · 17 years ago
- 475871a Rename SDOperand to SDValue. by Dan Gohman · 17 years ago
- 8b4588f Use AliasAnalysis::pointsToConstantMemory in SDISel to avoid unnecessary by Dan Gohman · 17 years ago
- fc74abf Enable first-class aggregates support. by Dan Gohman · 17 years ago
- 2dbc167 After early-lowering the FORMAL_ARGUMENTS node, delete it. by Dan Gohman · 17 years ago
- 462dc7f Add titles to the various SelectionDAG viewGraph calls by Dan Gohman · 17 years ago
- f00e74f Turn LegalizeTypes back off again for the moment: by Duncan Sands · 17 years ago
- 7f1d8ae Turn on LegalizeTypes by default. by Duncan Sands · 17 years ago
- 5e84368 Reapply 53476 and 53480, with a fix so that it properly updates by Dan Gohman · 17 years ago
- b5eec33 Back out 53476 and 53480 for now. Somehow they cause llc to miscompile 179.art. by Evan Cheng · 17 years ago
- dbe2a5a Add support for putting NamedRegionTimers in TimerGroups, and by Dan Gohman · 17 years ago
- 70587ea make legalize types be a command line option: -enable-legalize-types. by Chris Lattner · 17 years ago
- 0e5f130 Pool-allocation for SDNodes. The pool is allocated once for each function, by Dan Gohman · 17 years ago
- 126d907 Rather than having a different custom legalization by Duncan Sands · 17 years ago
- 10fff60 Revert my previous check-in that split up MachineModuleInfo. It turns out to by Bill Wendling · 17 years ago
- 4bdcb61 Add a new getMergeValues method that does not need by Duncan Sands · 17 years ago
- f1a792b Do run ComputeLiveOutVRegInfo with -fast. by Evan Cheng · 17 years ago
- 4576f6d Do not use computationally expensive scheduling heuristics with -fast. by Evan Cheng · 17 years ago
- ebffb66 Apply Chris' suggestion. by Evan Cheng · 17 years ago
- 4406604 Split ISD::LABEL into ISD::DBG_LABEL and ISD::EH_LABEL, eliminating by Dan Gohman · 17 years ago
- 99fe47b Use a simpler but equivalent form of RecordSource. by Dan Gohman · 17 years ago
- f667480 Add timing report for various sub-passes under SelectionDAGISel. by Evan Cheng · 17 years ago
- 7f46020 Rename ISD::LOCATION to ISD::DBG_STOPPOINT to better reflect its by Dan Gohman · 17 years ago
- db8d56b Split scheduling from instruction selection. by Evan Cheng · 17 years ago
- a44b674 Replace some std::vectors that showed up in heap profiling with by Dan Gohman · 17 years ago
- f951620 Revert the SelectionDAG optimization that makes by Duncan Sands · 17 years ago
- 305635a Refactor the DebugInfoDesc stuff out of the MachineModuleInfo file. Clean up by Bill Wendling · 17 years ago
- fd4418f Remove the OrigVT member from AtomicSDNode, as it is redundant with by Dan Gohman · 17 years ago
- 2887310 Added MemOperands to Atomic operations since Atomics touches memory. by Mon P Wang · 17 years ago
- ab8ec0a Teach ReturnInst lowering about aggregate return values. by Dan Gohman · 17 years ago
- dded0fd Fix the index calculations for the extractvalue lowering code. by Dan Gohman · 17 years ago
- 8f36f6d Simplify the ComputeLinearIndex logic and fix a few bugs. by Dan Gohman · 17 years ago
- ead0d88 add a new -enable-value-prop flag for llcbeta, that enables propagation by Chris Lattner · 17 years ago
- a47c6c3 Fix spelling. by Duncan Sands · 17 years ago
- 507a58a add missing atomic intrinsic from gcc by Andrew Lenharth · 17 years ago
- f5025cf CodeGen support for aggregate-value function arguments. by Dan Gohman · 17 years ago
- 8a6ccb5 Handle empty aggregate values. by Dan Gohman · 17 years ago
- 8e4eb09 Remove comparison methods for MVT. The main cause by Duncan Sands · 17 years ago
- 1d685a4 CodeGen support for insertvalue and extractvalue, and for loads and by Dan Gohman · 17 years ago
- 2d389e8 Connect successors before creating the DAG node for the branch. This has by Owen Anderson · 17 years ago
- 83ec4b6 Wrap MVT::ValueType in a struct to get type safety by Duncan Sands · 17 years ago
- 4fa2a3f Use isSingleValueType instead of isFirstClassType to by Dan Gohman · 17 years ago
- 041e2eb IR support for extractvalue and insertvalue instructions. Also, begin by Dan Gohman · 17 years ago
- 3521334 Really silence compiler warnings. by Evan Cheng · 17 years ago
- 5087124 Silence some compiler warnings. by Evan Cheng · 17 years ago
- 844731a Clean up the use of static and anonymous namespaces. This turned up by Dan Gohman · 17 years ago
- b43e9c1 Add support for vicmp/vfcmp codegen, more legalize support coming. by Nate Begeman · 17 years ago
- 0b12ecf Turn StripPointerCast() into a method by Anton Korobeynikov · 17 years ago
- b04addd Make StripPointerCast a common function (should we mak it method of Value instead?) by Anton Korobeynikov · 17 years ago
- 4a3f6c8 Make several variable declarations static. by Dan Gohman · 17 years ago
- 63307c3 Added addition atomic instrinsics and, or, xor, min, and max. by Mon P Wang · 17 years ago
- 1fa850b Use push_back(...) instead of resize(1, ...), per review feedback. by Dan Gohman · 17 years ago
- cba3b44 Fix uninitialized uses of the FPC variable. by Dan Gohman · 17 years ago
- 6bf30ab don't randomly miscompile seto/setuo just because we are in by Chris Lattner · 17 years ago
- 30e62c0 Tail call optimization improvements: by Arnold Schwaighofer · 17 years ago
- 3fb2968 make the vector conversion magic handle multiple results. by Chris Lattner · 17 years ago
- 41f6259 add support for multiple return values in inline asm. This is a step by Chris Lattner · 17 years ago
- 33bf38a Fix a bug in RegsForValue::getCopyToRegs() that causes cyclical scheduling units. If it's creating multiple CopyToReg nodes that are "flagged" together, it should not create a TokenFactor for it's chain outputs: by Evan Cheng · 17 years ago
- ebc3a3a Delete an unused constructor. by Dan Gohman · 17 years ago
- 10a6b7a Add a comment to CreateRegForValue that clarifies the handling of by Dan Gohman · 17 years ago
- 0fe0090 Rewrite the comments for RegsForValue and its members, and by Dan Gohman · 17 years ago
- b20d4f8 Don't call size() on each iteration of the loop. by Dan Gohman · 17 years ago
- 6833b06 Another collection of random cleanups. No functionality change. by Chris Lattner · 17 years ago
- b606dba Remove the SmallVector ctor that converts from a SmallVectorImpl. This by Chris Lattner · 17 years ago
- 8eaff04 switch RegsForValue::Regs to be a SmallVector to avoid by Chris Lattner · 17 years ago
- f899fce move static function out of anon namespace, no functionality change. by Chris Lattner · 17 years ago
- acf8b01 Another step to getting multiple result inline asm to work. by Chris Lattner · 17 years ago
- 5a09690 Implement a signficant optimization for inline asm: by Chris Lattner · 17 years ago
- 507ffd2 isa+cast -> dyn_cast by Chris Lattner · 17 years ago
- 4376fea Move a bunch of inline asm code out of line. by Chris Lattner · 17 years ago
- 23ce502 Remove the code from CodeGenPrepare that moved getresult instructions by Dan Gohman · 17 years ago
- 67780f1 Use isa instead of dyn_cast. by Dan Gohman · 17 years ago
- 3dc34f6 Add support to codegen for getresult instructions with undef operands. by Dan Gohman · 17 years ago
- 6ccbbd8 Change Divided flag to Split, as suggested by Evan by Nicolas Geoffray · 17 years ago