search for: valuetracking_8cpp_source

...ne Instruction Scheduling" behaves correctly. Is this the right thing to do? Would other callers to llvm::GetUnderlyingObject not want the additional behavior? Thanks, Matthew Curtis. [1] http://llvm.org/docs/doxygen/html/StackColoring_8cpp_source.html [2] http://llvm.org/docs/doxygen/html/ValueTracking_8cpp_source.html#l01780 [3] http://llvm.org/docs/doxygen/html/ScheduleDAGInstrs_8cpp_source.html#l00087 -- Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation

...orrectly. > > Is this the right thing to do? Would other callers to llvm::GetUnderlyingObject > not want the additional behavior? > > Thanks, > Matthew Curtis. > > [1] http://llvm.org/docs/doxygen/html/StackColoring_8cpp_source.html > [2] http://llvm.org/docs/doxygen/html/ValueTracking_8cpp_source.html#l01780 > [3] http://llvm.org/docs/doxygen/html/ScheduleDAGInstrs_8cpp_source.html#l00087 >

...o account and will report > that %idx must be all zeros. However, %idx might turn into undef if > %x has the bottom bit set. In fact, it doesn't appear to be > conservative at all in the face of flags. If anything, it takes > advantage of them: > http://llvm.org/docs/doxygen/html/ValueTracking_8cpp_source.html#l00275 > > > This tells me that we have three options if we want the instruction > to stay speculatable: > 1. Make a variant of ComputeKnownBits (or something along those > lines) which is explicitly pessimistic in the face of flags. > > 2. Kick the can to the backen...

----- Original Message ----- > From: "Pete Cooper" <peter_cooper at apple.com> > To: "Hal Finkel" <hfinkel at anl.gov> > Cc: "LLVMdev" <llvmdev at cs.uiuc.edu>, "Paweł Bylica" <chfast at gmail.com> > Sent: Wednesday, July 1, 2015 6:42:41 PM > Subject: Re: [LLVMdev] extractelement causes memory access violation - what to

Michael makes a great point about aliasing here and different indexing that accesses the same element! Another note: x = A[0][2] is fundamentally different depending on the type of `A`. If e.g. A was declared: int A[10][20], there's only _one_ load. A is a (and is treated as) a linear buffer, and GEPs only pinpoint the specific position of A[0][2] in this buffer (i.e. 0*10 + 2). But if A was