summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Cleanup up LegalizeTypes handling of loads andDuncan Sands2008-06-235-24/+22
| | | | | | stores. llvm-svn: 52620
* Make test work on non-x86 machines (like my G4 PPC).Bill Wendling2008-06-231-2/+2
| | | | llvm-svn: 52619
* At Chris' suggestion, move the liveness and worklist datastructures intoOwen Anderson2008-06-231-7/+11
| | | | | | | instance variables so they can be allocated just once, and reuse the worklist as the dead list as well. llvm-svn: 52618
* Improve LSR's dead-phi detection to handle use-def cyclesDan Gohman2008-06-222-20/+38
| | | | | | with more than two nodes. llvm-svn: 52617
* Use Loop::block_iterator.Dan Gohman2008-06-224-16/+21
| | | | llvm-svn: 52616
* Generalize createSCEV to be able to form SCEV expressions fromDan Gohman2008-06-221-110/+117
| | | | | | ConstantExprs. llvm-svn: 52615
* Use SCEVAddRecExpr::isAffine.Dan Gohman2008-06-222-3/+3
| | | | llvm-svn: 52614
* Remove unnecessary #includes.Dan Gohman2008-06-224-4/+0
| | | | llvm-svn: 52613
* Move a few more SCEVExpander methods out-of-line.Dan Gohman2008-06-222-17/+25
| | | | llvm-svn: 52612
* Fix PR2369 by making scalarrepl more careful about promoting Chris Lattner2008-06-222-5/+32
| | | | | | | | | structures. Its default threshold is to promote things that are smaller than 128 bytes, which is sane. However, it is not sane to do this for things that turn into 128 *registers*. Add a cap on the number of registers introduced, defaulting to 128/4=32. llvm-svn: 52611
* Make custom lowering of ADD work correctly. ThisDuncan Sands2008-06-221-3/+3
| | | | | | | | | | fixes PR2476; patch by Richard Osborne. The same problem exists for a bunch of other operators, but I'm ignoring this because they will be automagically fixed when the new LegalizeTypes infrastructure lands, since it already solves this problem centrally. llvm-svn: 52610
* add a new clang::APValue class at Eli's request. It is a discriminatedChris Lattner2008-06-222-5/+190
| | | | | | union between [potentially complex] APInt/APFloat. llvm-svn: 52609
* Fix for PR2479: correctly optimize expressions like (a > 13) & (a == Eli Friedman2008-06-212-2/+12
| | | | | | | | 15). See also PR1800, which is about the signed case. llvm-svn: 52608
* add a fixme backChris Lattner2008-06-211-0/+1
| | | | llvm-svn: 52607
* Use Instruction::eraseFromParent().Dan Gohman2008-06-217-13/+13
| | | | llvm-svn: 52606
* Use Function's arg_size() and size() methods.Dan Gohman2008-06-211-2/+2
| | | | llvm-svn: 52605
* Simplify some getNode calls.Dan Gohman2008-06-211-6/+3
| | | | llvm-svn: 52604
* canClobberPhysRegDefs shouldn't called without checking hasPhysRegDefs;Dan Gohman2008-06-211-3/+2
| | | | | | check this with an assert. llvm-svn: 52603
* Avoid creating a redundant zero APInt.Dan Gohman2008-06-211-1/+1
| | | | llvm-svn: 52602
* Use clear() to zero an existing APInt.Dan Gohman2008-06-211-1/+1
| | | | llvm-svn: 52601
* Use back() instead of [size()-1].Dan Gohman2008-06-212-3/+3
| | | | llvm-svn: 52600
* "this patch adds code generation hooks for Objective-C constant strings. It ↵Chris Lattner2008-06-216-19/+20
| | | | | | | | also modifies Sema so that Objective-C constant strings are treated as untyped objects if the interface for the constant string class can not be found. This is consistent with Apple GCC. I thought it was consistent with GNU GCC, since this was causing failures when trying to compile GNUstep with (GNU) GCC, but it appears that this is not the case when attempting to produce a simple test case to demonstrate it. Possibly there is a way of making the error go away, but I haven't found out what it is yet." Patch by David Chisnall! llvm-svn: 52599
* Implement printing for ObjCClassDecl's, patch by Jean-Daniel Dupas!Chris Lattner2008-06-211-2/+9
| | | | llvm-svn: 52598
* Don't add isa with @defs only to work around it in the code generator, patchChris Lattner2008-06-212-8/+0
| | | | | | by David Chisnall. llvm-svn: 52597
* This file is empty.Duncan Sands2008-06-211-0/+0
| | | | llvm-svn: 52596
* Turn off llvm-gcc warnings when running "make check".Duncan Sands2008-06-211-2/+2
| | | | llvm-svn: 52595
* Use MachineBasicBlock::transferSuccessors.Dan Gohman2008-06-213-21/+7
| | | | llvm-svn: 52594
* "This moves built-in Objective-C types up the scope chains to where they can ↵Chris Lattner2008-06-213-54/+42
| | | | | | | | | | be replaced by versions included from the runtime library's headers." This makes it ok to use @"foo" without a declaration for NSConstantString. Patch by David Chisnall! llvm-svn: 52593
* Use static_cast instead of reinterpret_cast for casting void*.Dan Gohman2008-06-212-5/+5
| | | | llvm-svn: 52592
* Fix warning when assertions disabled.Chris Lattner2008-06-211-1/+1
| | | | llvm-svn: 52590
* fix warning when assertions disabled.Chris Lattner2008-06-211-2/+1
| | | | llvm-svn: 52589
* fix warning when assertion disabled.Chris Lattner2008-06-211-1/+1
| | | | llvm-svn: 52588
* fix some warnings when assertions are disabled.Chris Lattner2008-06-213-5/+5
| | | | llvm-svn: 52587
* add parser and sema support for the funny ObjC '@defs' thing.Chris Lattner2008-06-216-13/+103
| | | | | | Patch by David Chisnall! llvm-svn: 52586
* Remove a redundant return.Dan Gohman2008-06-211-2/+0
| | | | llvm-svn: 52585
* Reorder some fields in SDNode. This avoids padding,Dan Gohman2008-06-211-5/+5
| | | | | | reducings its size from 80 to 72 on 64-bit hosts. llvm-svn: 52584
* Remove ScheduleDAG's SUnitMap altogether. Instead, use SDNode's NodeIdDan Gohman2008-06-215-53/+39
| | | | | | | field, which is otherwise unused after instruction selection, as an index into the SUnit array. llvm-svn: 52583
* Add a priority queue class, which is a wrapper around std::priority_queueDan Gohman2008-06-214-12/+85
| | | | | | | and provides fairly efficient removal of arbitrary elements. Switch ScheduleDAGRRList from std::set to this new priority queue. llvm-svn: 52582
* remove dead enums.Chris Lattner2008-06-212-4/+0
| | | | llvm-svn: 52581
* Switch 'super' from being a weird cast thing to being a predefined expr node.Chris Lattner2008-06-216-44/+25
| | | | | | Patch by David Chisnall with objc rewriter and stmtdumper updates from me. llvm-svn: 52580
* "This maps the -pthread option to -lpthread - this isn't 100% correct, Chris Lattner2008-06-211-0/+6
| | | | | | | | | | | | | | | since handling this correctly is quite complex, and on some platforms requires additional -D options and on some implies linking against a different libc, but this works better than just ignoring the option. The other change passes the -x option across to clang, which allows compiling .c files as Objective-C and so on. For some reason a lot of configure scripts seem to be under the misguided impression that this is a sensible thing to do." Patch by David Chisnall! llvm-svn: 52579
* Include stdint.h instead of stdio.h.Ted Kremenek2008-06-211-1/+1
| | | | llvm-svn: 52578
* Support for load/store of expanded float types. IDuncan Sands2008-06-214-1/+75
| | | | | | | don't know if a truncating store is possible here, but added support for it anyway. llvm-svn: 52577
* Change ScheduleDAG's SUnitMap from DenseMap<SDNode*, vector<SUnit*> >Dan Gohman2008-06-215-42/+54
| | | | | | | | to DenseMap<SDNode*, SUnit*>, and adjust the way cloned SUnit nodes are handled so that only the original node needs to be in the map. This speeds up llc on 447.dealII.llvm.bc by about 2%. llvm-svn: 52576
* K&R-style functions not allowed in C++.Argyrios Kyrtzidis2008-06-211-2/+3
| | | | llvm-svn: 52575
* Enable PRE.Evan Cheng2008-06-211-1/+1
| | | | llvm-svn: 52574
* Undo spill weight tweak. Need to investigate the performance regressions.Evan Cheng2008-06-216-19/+10
| | | | llvm-svn: 52572
* Simplify some template parameterization.Dan Gohman2008-06-211-19/+12
| | | | llvm-svn: 52571
* Back out Matthijs' DAE patches. It's miscompiling gcc driver.Evan Cheng2008-06-212-557/+367
| | | | llvm-svn: 52570
* Updated latest static analyzer builder to checker-43.Ted Kremenek2008-06-201-1/+1
| | | | llvm-svn: 52569
OpenPOWER on IntegriCloud