summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Provide static methods in BinaryOperator to determine if an opcode is an ↵Ted Kremenek2008-07-171-3/+10
| | | | | | equality opcode, a relational opcode, or a logical opcode. llvm-svn: 53744
* Move GRTransferFunc* into ValueStateManager, and move the assumption logic ↵Ted Kremenek2008-07-175-230/+245
| | | | | | there as well. llvm-svn: 53743
* Added headersTed Kremenek2008-07-171-4/+8
| | | | llvm-svn: 53742
* Remove redundant logic.Ted Kremenek2008-07-172-18/+3
| | | | llvm-svn: 53740
* Begin major changes to EvalXXX methods in GRTransferFuncs. Currently some ↵Ted Kremenek2008-07-175-5/+93
| | | | | | of the methods only return an RVal; we want them to be able to create an arbitrary number of states. llvm-svn: 53739
* When printing MemOperand nodes, only use print() forDan Gohman2008-07-171-3/+10
| | | | | | | | PseudoSourceValue values, which never have names. Use getName() for all other values, because we want to print just a short summary of the value, not the entire instruction. llvm-svn: 53738
* Fix for codegen crash on multibit bool bitfield initializationDaniel Dunbar2008-07-171-0/+7
| | | | | | <rdar://problem/6078606> llvm-svn: 53737
* Use MergeBlockIntoPredecessor to simplify some code.Owen Anderson2008-07-171-52/+9
| | | | llvm-svn: 53735
* Subreg live interval valno may not have a corresponding def machineinstr ↵Evan Cheng2008-07-172-1/+35
| | | | | | since it's less precise. llvm-svn: 53734
* Add nounwind.Evan Cheng2008-07-171-1/+1
| | | | llvm-svn: 53733
* test commit accessDaniel Dunbar2008-07-171-0/+2
| | | | llvm-svn: 53732
* Make MergeBlockIntoPredecessor more aggressive when the same successor appearsOwen Anderson2008-07-172-6/+25
| | | | | | more than once. llvm-svn: 53731
* Enable PRE. My last batch of changes fixed the miscompile.Owen Anderson2008-07-171-1/+1
| | | | llvm-svn: 53730
* Use a legal type for elements of the vector_shuffleDuncan Sands2008-07-171-3/+4
| | | | | | | | | | | | | | | | | | mask. These are just indices into the shuffled vector so their type is unrelated to the type of the shuffled elements (which is what was being used before). This fixes vec_shuffle-11.ll when using LegalizeTypes. What seems to have happened is that Dan's recent change r53687, which corrected the result type of the shuffle, somehow caused LegalizeTypes to notice that the mask operand was a BUILD_VECTOR with a legal type but elements of an illegal type (i64). LegalizeTypes legalized this by introducing a new BUILD_VECTOR of i32 and bitcasting it to the old type. But the mask operand is not supposed to be a bitcast but a straight BUILD_VECTOR of constants, causing a crash. llvm-svn: 53729
* Add a new function, ReplaceAllUsesOfValuesWith, which handles bulkDan Gohman2008-07-1721-310/+520
| | | | | | | | | | | | | | | | | | | | | | | | | | | replacement of multiple values. This is slightly more efficient than doing multiple ReplaceAllUsesOfValueWith calls, and theoretically could be optimized even further. However, an important property of this new function is that it handles the case where the source value set and destination value set overlap. This makes it feasible for isel to use SelectNodeTo in many very common cases, which is advantageous because SelectNodeTo avoids a temporary node and it doesn't require CSEMap updates for users of values that don't change position. Revamp MorphNodeTo, which is what does all the work of SelectNodeTo, to handle operand lists more efficiently, and to correctly handle a number of corner cases to which its new wider use exposes it. This commit also includes a change to the encoding of post-isel opcodes in SDNodes; now instead of being sandwiched between the target-independent pre-isel opcodes and the target-dependent pre-isel opcodes, post-isel opcodes are now represented as negative values. This makes it possible to test if an opcode is pre-isel or post-isel without having to know the size of the current target's post-isel instruction set. These changes speed up llc overall by 3% and reduce memory usage by 10% on the InstructionCombining.cpp testcase with -fast and -regalloc=local. llvm-svn: 53728
* Indentation.Evan Cheng2008-07-171-3/+3
| | | | llvm-svn: 53727
* Moved RemoveDeadBindings logic for the contents of 'Store' to a virtual ↵Ted Kremenek2008-07-174-90/+122
| | | | | | RemoveDeadBindings method in StoreManager. llvm-svn: 53726
* We now build universal binaries of the checker.Ted Kremenek2008-07-171-1/+1
| | | | llvm-svn: 53725
* Unify ctx_iterator/ctx_begin()/ctx_end() and iterator/begin()/end() so that ↵Argyrios Kyrtzidis2008-07-173-159/+102
| | | | | | a single iterator type is used for both traversing decls of the same declaration context *and* of the parent declaration contexts, depending on the value of the bool parameter 'LookInParentCtx' that is passed to IdentifierResolver::begin(). llvm-svn: 53724
* LegalizeTypes support for what seems to be theDuncan Sands2008-07-173-0/+36
| | | | | | | only missing ppc long double operations: FNEG and FP_EXTEND. llvm-svn: 53723
* Remove unnecessary readme entryNate Begeman2008-07-171-9/+0
| | | | llvm-svn: 53722
* Turn LegalizeTypes back off again for the moment:Duncan Sands2008-07-176-6/+10
| | | | | | | it is breaking Darwin bootstrap due to missing functionality. llvm-svn: 53721
* Fix a typo in last commitNate Begeman2008-07-171-1/+1
| | | | llvm-svn: 53720
* SSE codegen for vsetcc nodesNate Begeman2008-07-175-43/+209
| | | | llvm-svn: 53719
* Make GlobalOpt preserve address spaces when scalar replacing aggregate globals.Matthijs Kooijman2008-07-172-2/+32
| | | | llvm-svn: 53716
* Fix PR2553Chris Lattner2008-07-172-1/+10
| | | | llvm-svn: 53715
* When lowering certain atomics, we need to copy the memoperand from the oldMon P Wang2008-07-171-1/+5
| | | | | | atomic operation to the new one. llvm-svn: 53714
* Factorize some code for determining which libcall to use.Duncan Sands2008-07-175-563/+231
| | | | llvm-svn: 53713
* Inliner tweak. Function calls should cost more than one instruction!Evan Cheng2008-07-171-0/+5
| | | | llvm-svn: 53712
* Factor MergeBlockIntoPredecessor out into BasicBlockUtils.Owen Anderson2008-07-173-51/+60
| | | | llvm-svn: 53705
* RewriteObjC::SynthMessageExpr(): Fix super rewrite (a fairly recent regression).Steve Naroff2008-07-161-1/+4
| | | | | | Fixes <rdar://problem/6046663> clang ObjC rewriter: Regression in handling of 'super' in latest build. llvm-svn: 53703
* When in C++, make EnumConstant names hide tag names in the same scope, ↵Argyrios Kyrtzidis2008-07-162-1/+8
| | | | | | instead of colliding with them. llvm-svn: 53702
* Name string length is end position marker - begin position marker.Devang Patel2008-07-161-1/+1
| | | | llvm-svn: 53697
* Remove a FIXME from RewriteObjC::RewriteObjCSynchronizedStmt().Steve Naroff2008-07-161-2/+5
| | | | llvm-svn: 53696
* RewriteObjC::Initialize(): add function decls used by @synchronized.Steve Naroff2008-07-161-0/+3
| | | | | | This fixes <rdar://problem/6040143> clang ObjC rewriter: @synchronized keyword inserts undeclared objc functions. llvm-svn: 53695
* Two fixes:Steve Naroff2008-07-164-14/+21
| | | | | | | - Make sure ObjCIvarDecl propagates the bitfield width. - RewriteObjC::SynthesizeIvarOffsetComputation(): Avoid using the __OFFSETOF__ mumbo jumbo for bitfields (since it isn't legal C). This fixes <rdar://problem/5986079> clang ObjC rewriter: bitfields and ivar access don't mix. llvm-svn: 53694
* Do not forget global definitions from inline asm code block.Devang Patel2008-07-162-0/+40
| | | | llvm-svn: 53693
* Mark function used by asm block as used, otherwise optimizer may not see the ↵Devang Patel2008-07-162-1/+8
| | | | | | use and may delete the function. llvm-svn: 53692
* There's no need to iterate block merging and PRE. In fact, iterating the latterOwen Anderson2008-07-161-14/+12
| | | | | | could cause problems for memdep when it breaks critical edges. llvm-svn: 53691
* Somehow, custom lowering of i64 multiplications got dropped along the way.Scott Michel2008-07-161-104/+105
| | | | llvm-svn: 53689
* Fix the result type of X86's truncate to i8.Dan Gohman2008-07-161-5/+3
| | | | llvm-svn: 53688
* Fix the result type of a VECTOR_SHUFFLE+BIT_CONVERT dagcombine. ThisDan Gohman2008-07-161-3/+2
| | | | | | | was turned up by some new SelectionDAG assertion checks that I'm working on. llvm-svn: 53687
* Fix the name of BreakCriticalMachineEdge.h's include guardDan Gohman2008-07-161-2/+2
| | | | llvm-svn: 53686
* Correct a top-level comment.Dan Gohman2008-07-161-1/+1
| | | | llvm-svn: 53685
* Add support for promoting and expanding AssertZextDuncan Sands2008-07-164-0/+3116
| | | | | | | and AssertSext. Needed when passing huge integer parameters with the zeroext or signext attributes. llvm-svn: 53684
* Clarify the comments here, to make slightly more clear theDan Gohman2008-07-161-2/+2
| | | | | | | difference in purpose of TargetInstrInfo and TargetInstrDesc, which isn't immediately obvious from the name. llvm-svn: 53683
* Tidy a doxygen comment.Dan Gohman2008-07-161-1/+1
| | | | llvm-svn: 53682
* Fix a comment to say nonnegative instead of positive.Dan Gohman2008-07-161-1/+1
| | | | llvm-svn: 53681
* Add an assert to check for empty flags for MachineMemOperand.Dan Gohman2008-07-161-0/+1
| | | | llvm-svn: 53680
* RewriteObjC::RewriteObjCTryStmt():Don't synthesize a catch begin if there ↵Steve Naroff2008-07-162-20/+24
| | | | | | | | are 0 catch clauses. This fixes <rdar://problem/5987211> clang ObjC rewriter: @try / @finally block produces unbalanced output. llvm-svn: 53679
OpenPOWER on IntegriCloud