summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* 80-column and whitespace fixes.Eric Christopher2009-10-071-135/+156
| | | | llvm-svn: 83489
* Fixed MCSectionMachO::ParseSectionSpecifier to allow an attribute of "none" soKevin Enderby2009-10-072-43/+58
| | | | | | | that a symbol stub section with no attributes can be parsed as in: .section __TEXT,__picsymbolstub4,symbol_stubs,none,16 llvm-svn: 83488
* Convert test to FileCheck.Bob Wilson2009-10-071-9/+27
| | | | llvm-svn: 83487
* Add codegen support for NEON vst4 intrinsics with 128-bit vectors.Bob Wilson2009-10-074-9/+135
| | | | llvm-svn: 83486
* Fixes problem in finding visible convertion functions of a classFariborz Jahanian2009-10-073-30/+113
| | | | | | | | where matching conversion types in base classes were still visible. Plus refactoring and cleanup. Added a test case. llvm-svn: 83485
* Add codegen support for NEON vst3 intrinsics with 128-bit vectors.Bob Wilson2009-10-074-9/+129
| | | | llvm-svn: 83484
* grammarJim Grosbach2009-10-071-1/+1
| | | | llvm-svn: 83483
* Add codegen support for NEON vst2 intrinsics with 128-bit vectors.Bob Wilson2009-10-074-7/+84
| | | | llvm-svn: 83482
* add initializers for clarity. Add missing assignment of PrevLastUseOp.Jim Grosbach2009-10-071-2/+3
| | | | llvm-svn: 83481
* Remove LazyLiveness from the tree. It doesn't work right now, and I'm not ↵Owen Anderson2009-10-071-168/+0
| | | | | | | | going to have the time to finish it any time soon. If someone's interested it, they can resurrect it from SVN history. llvm-svn: 83480
* Add codegen support for NEON vld4 intrinsics with 128-bit vectors.Bob Wilson2009-10-074-7/+137
| | | | llvm-svn: 83479
* Add another bit of the ARM target assembler to llvm-mc to parse registersKevin Enderby2009-10-071-3/+17
| | | | | | | | with writeback, things like "sp!", etc. Also added some more stuff to the temporarily hacked methods ARMAsmParser::MatchRegisterName and ARMAsmParser::MatchInstruction to allow more parser testing. llvm-svn: 83477
* Replace some code for aggressive-remat with MachineInstr::isInvariantLoad, andDan Gohman2009-10-071-23/+28
| | | | | | teach it how to recognize invariant physical registers. llvm-svn: 83476
* Replace TargetInstrInfo::isInvariantLoad and its target-specificDan Gohman2009-10-0712-107/+56
| | | | | | | | | implementations with a new MachineInstr::isInvariantLoad, which uses MachineMemOperands and is target-independent. This brings MachineLICM and other functionality to targets which previously lacked an isInvariantLoad implementation. llvm-svn: 83475
* Add a few simple MachineVerifier checks for MachineMemOperands.Dan Gohman2009-10-071-0/+10
| | | | llvm-svn: 83474
* Diagnose explicit instantiations and specializations that occur in class scopeDouglas Gregor2009-10-073-4/+15
| | | | llvm-svn: 83473
* When building candidate set for built-ins; when looking forFariborz Jahanian2009-10-072-4/+2
| | | | | | | convesion functions, look in base classes to. (Removes a FIXME). llvm-svn: 83472
* Add codegen support for NEON vld3 intrinsics with 128-bit vectors.Bob Wilson2009-10-074-13/+146
| | | | llvm-svn: 83471
* Rearrange code for selecting vld2 intrinsics. No functionality change.Bob Wilson2009-10-071-9/+14
| | | | | | This is just to be more consistent with the forthcoming code for vld3/4. llvm-svn: 83470
* Class template partial specializations can be declared anywhere thatDouglas Gregor2009-10-073-4/+41
| | | | | | | | | | its definition may be defined, including in a class. Also, put in an assertion when trying to instantiate a class template partial specialization of a member template, which is not yet implemented. llvm-svn: 83469
* Add tests for vld2 of 128-bit vectors.Bob Wilson2009-10-071-0/+50
| | | | llvm-svn: 83468
* Add register-reuse to frame-index register scavenging. When a target usesJim Grosbach2009-10-0731-100/+262
| | | | | | | | | | | | | | | | | | | | a virtual register to eliminate a frame index, it can return that register and the constant stored there to PEI to track. When scavenging to allocate for those registers, PEI then tracks the last-used register and value, and if it is still available and matches the value for the next index, reuses the existing value rather and removes the re-materialization instructions. Fancier tracking and adjustment of scavenger allocations to keep more values live for longer is possible, but not yet implemented and would likely be better done via a different, less special-purpose, approach to the problem. eliminateFrameIndex() is modified so the target implementations can return the registers they wish to be tracked for reuse. ARM Thumb1 implements and utilizes the new mechanism. All other targets are simply modified to adjust for the changed eliminateFrameIndex() prototype. llvm-svn: 83467
* Removed couple of unnecessary canonicalizationFariborz Jahanian2009-10-071-2/+2
| | | | | | per Doug's obsevation. llvm-svn: 83464
* Do not assume that the module is set.Devang Patel2009-10-071-15/+18
| | | | llvm-svn: 83462
* Add a comment explaining how DenseMap::insert works, because it is notTorok Edwin2009-10-071-0/+3
| | | | | | | | | intuitive. It does NOT update the value if the key is already in the map, it also returns false if the key is already in the map, regardless if the value matched. llvm-svn: 83458
* Add PR to this FIXME, looks like I didn't commit this change after all.Torok Edwin2009-10-071-1/+2
| | | | llvm-svn: 83457
* Make getPointerTo return a const PointerType* rather thanDuncan Sands2009-10-072-2/+2
| | | | | | an unqualified PointerType* because it seems more correct. llvm-svn: 83454
* Add 'template' keyword to disambiguate the use of a member template within a ↵Ted Kremenek2009-10-071-1/+1
| | | | | | templated class. Hopefully this will please the buildbots. llvm-svn: 83452
* Revert r83443.Ted Kremenek2009-10-074-14/+7
| | | | llvm-svn: 83451
* INTRINSIC_W_CHAIN and INTRINSIC_VOID do not use MemSDNode. TheyDan Gohman2009-10-071-2/+0
| | | | | | may access memory, but they don't carry a MachineMemOperand. llvm-svn: 83449
* Mangle anonymous structs/unions correctly. Fixes PR5139.Anders Carlsson2009-10-073-3/+59
| | | | llvm-svn: 83448
* Fixup pathnames.Mike Stump2009-10-074-7/+14
| | | | llvm-svn: 83443
* Add a MangleContext and pass it to all mangle functions. It will be used for ↵Anders Carlsson2009-10-075-42/+58
| | | | | | keeping state, such as identifiers assigned to anonymous structs as well as scope encoding. llvm-svn: 83442
* Add FreeInst to the "is a call" check for Insts that are calls, butEric Christopher2009-10-071-2/+2
| | | | | | not intrinsics. llvm-svn: 83441
* More testing of explicit specializationsDouglas Gregor2009-10-071-15/+25
| | | | llvm-svn: 83440
* Change ExplodedNode to have its NodeGroups all BumpPtrAllocated, avoiding ↵Ted Kremenek2009-10-075-57/+253
| | | | | | | | malloc() traffic when adding successors/predecessors to a node. This was done by introducing BumpVector, which is essentially SmallVector with all memory being BumpPtrAllocated (this can certainly be cleaned up or moved into llvm/ADT). This change yields a 1.8% speed increase when running the analyzer (with -analyzer-store=region) on a small benchmark file. llvm-svn: 83439
* Force a specific target triple in c-index-api-test.m, making the behavior of ↵Ted Kremenek2009-10-071-8/+8
| | | | | | the test not be dependent on the host. llvm-svn: 83438
* Fix this comment. The loop header is the loop entry point.Dan Gohman2009-10-071-2/+2
| | | | llvm-svn: 83437
* Refactor checking of the scope of explicit template specializationDouglas Gregor2009-10-077-113/+297
| | | | | | | | | declarations and explicit template instantiations, improving diagnostics and making the code usable for function template specializations (as well as class template specializations and partial specializations). llvm-svn: 83436
* Add PseudoSourceValues for constpool stuff on ELF (Darwin should use ↵Anton Korobeynikov2009-10-072-12/+34
| | | | | | | | something similar) and register spills. llvm-svn: 83435
* While we still have a MallocInst treat it as a call like any otherEric Christopher2009-10-071-0/+4
| | | | | | | | | for inlining. When MallocInst goes away this code will be subsumed as part of calls and work just fine... llvm-svn: 83434
* Fix eof.Mike Stump2009-10-061-1/+1
| | | | llvm-svn: 83433
* Fix eof.Mike Stump2009-10-061-1/+1
| | | | llvm-svn: 83432
* Patch to implement C++ [over.built]p11 of overload resolution.Fariborz Jahanian2009-10-062-1/+58
| | | | | | | | Doug, please review. There is a FIXME in the test case with a question which is unrelated to this patch (that is, error is issued before set of builtins are added to the candidate list). llvm-svn: 83429
* Uncomment some commented out code.Anders Carlsson2009-10-062-5/+5
| | | | llvm-svn: 83428
* Add codegen for __builtin_abort. Convert to FileCheck.Mike Stump2009-10-062-44/+52
| | | | llvm-svn: 83427
* Change GetAddressCXXOfBaseClass to use CXXBasePaths for calculating base ↵Anders Carlsson2009-10-064-66/+144
| | | | | | class offsets. Fix the code to handle virtual bases as well. llvm-svn: 83426
* Added bits of the ARM target assembler to llvm-mc to parse some load instructionKevin Enderby2009-10-061-1/+404
| | | | | | | operands. Some parsing of arm memory operands for preindexing and postindexing forms including with register controled shifts. This is a work in progress. llvm-svn: 83424
* Add files to xcode project.Anders Carlsson2009-10-061-0/+6
| | | | llvm-svn: 83423
* Add codegen support for NEON vld2 operations on quad registers.Bob Wilson2009-10-064-1/+62
| | | | llvm-svn: 83422
OpenPOWER on IntegriCloud