summaryrefslogtreecommitdiff
path: root/lib/Analysis/ValueTracking.cpp
Commit message (Expand)AuthorAge
* teach ComputeNumSignBits to look through PHI nodes.Chris Lattner2010-01-07
* Move remaining stuff to the isInteger predicate.Benjamin Kramer2010-01-05
* Fix GetConstantStringInfo to not look into MDString (it works on Chris Lattner2009-12-15
* Fixed an assertion failure for tracking sext of a vector of integersMon P Wang2009-12-02
* implement a FIXME: limit the depth that DecomposeGEPExpression goes the sameChris Lattner2009-11-28
* limit the recursion depth of GetLinearExpression. ThisChris Lattner2009-11-27
* Implement PR1143 (at -m64) by making basicaa look through extensions. WeChris Lattner2009-11-26
* fix two transposed lines duncan caught and add an explanatory comment.Chris Lattner2009-11-26
* this todo is resolved.Chris Lattner2009-11-26
* move DecomposeGEPExpression out into ValueTracking.cppChris Lattner2009-11-26
* remove some redundant bracesChris Lattner2009-11-26
* Remove unused LLVMContext.Nick Lewycky2009-11-23
* Simplify ComputeMultiple so that it doesn't depend on TargetData.Dan Gohman2009-11-18
* Add ComputeMultiple() analysis function that recursively determines if a Valu...Victor Hernandez2009-11-10
* Remove AllocationInst. Since MallocInst went away, AllocaInst is the only su...Victor Hernandez2009-10-23
* Remove MallocInst from LLVM Instructions.Victor Hernandez2009-10-17
* Handle sqrt in CannotBeNegativeZero. absf and abslDale Johannesen2009-09-25
* Teach ValueTracking how to look through GlobalAliases. GlobalAliases areDan Gohman2009-09-15
* fix ComputeMaskedBits handling of zext/sext/trunc to work with vectors.Chris Lattner2009-09-08
* add some comments to describe the invariants.Chris Lattner2009-09-08
* Fix DbgStopPointInst->getFileName/getDirectory, broken by the MDNodification inTorok Edwin2009-09-02
* Handle TargetData with const.Dan Gohman2009-08-27
* Use hasDefinitiveInitializer() instead of testing the same thingDan Gohman2009-08-19
* Fix SimplifyLibcalls and ValueTracking to check mayBeOverriddenDan Gohman2009-08-19
* Push LLVMContexts through the IntegerType APIs.Owen Anderson2009-08-13
* Don't assume that external global variables are aligned at their preferredDan Gohman2009-08-11
* Fix a bunch of namespace pollution.Dan Gohman2009-08-07
* Move a few more APIs back to 2.5 forms. The only remaining ones left to chan...Owen Anderson2009-07-31
* Move more code back to 2.5 APIs.Owen Anderson2009-07-30
* Remove Value::getNameLenDaniel Dunbar2009-07-26
* Get rid of the Pass+Context magic.Owen Anderson2009-07-22
* Add a new Operator class, for handling Instructions and ConstantExprsDan Gohman2009-07-17
* "LLVMContext* " --> "LLVMContext *"Owen Anderson2009-07-06
* Finish LLVMContext-ing lib/Analysis. This required pushing LLVMContext's thr...Owen Anderson2009-07-06
* do not try to analyze bitcasts from i64 to <2 x i32> in ComputedMaskedBits. ...Chris Lattner2009-07-02
* Fix llvm::ComputeNumSignBits to handle pointer typesDan Gohman2009-06-22
* Support vector casts in more places, fixing a variety of assertionDan Gohman2009-06-15
* Split the Add, Sub, and Mul instruction opcodes into separateDan Gohman2009-06-04
* When the low bits of one operand of an add are zero, that numberDan Gohman2009-05-24
* Teach ValueTracking a new way to analyze PHI nodes, and and teachDan Gohman2009-05-21
* Rename PaddedSize to AllocSize, in the hope that thisDuncan Sands2009-05-09
* Oops...I committed too much.Bill Wendling2009-03-13
* Temporarily XFAIL this test.Bill Wendling2009-03-13
* Fix a ValueTracking rule: RHS means operand 1, not 0. Add a simpleDan Gohman2009-02-24
* Eliminate use of uint32_t to improve compatibility with cygwinChris Lattner2009-01-20
* Rename getABITypeSize to getTypePaddedSize, asDuncan Sands2009-01-12
* ValueTracker can't assume that an alloca with no specified alignment Chris Lattner2009-01-08
* Re-apply 55137 with fixes.David Greene2008-10-27
* Temporarily reverting r55137. This was causing the bootstrap to go into anBill Wendling2008-08-21
* Fix ComputeMaskedBits to handle phis correctly. We need to take theDavid Greene2008-08-21