search for: isptrorptrvectorty

...ns, the compiler will get a lot larger and a lot slower) On Mon, Aug 29, 2016 at 9:10 AM, Daniel Berlin <dberlin at dberlin.org> wrote: > it would also, for that matter, say the same about an array of pointers. > > It's not clear to me what will break if you change this to > isPtrOrPtrVectorTy. > > In fact, i know it doesn't fix this bug. > It's a pretty deep rabbit hole of things not quite prepared to understand > vectors of pointers. > > (we prepare memorylocations of them, but memory locations expect to be one > thing, not a group of things, etc). > &gt...

this is definitely a bug in AA. 225 for (auto I = CS2.arg_begin(), E = CS2.arg_end(); I != E; ++I) { 226 const Value *Arg = *I; 227 if (!Arg->getType()->isPointerTy()) -> 228 continue; 229 unsigned CS2ArgIdx = std::distance(CS2.arg_begin(), I); 230 auto CS2ArgLoc = MemoryLocation::getForArgument(CS2, CS2ArgIdx, TLI);

...and > a lot slower) > > On Mon, Aug 29, 2016 at 9:10 AM, Daniel Berlin <dberlin at dberlin.org> > wrote: > >> it would also, for that matter, say the same about an array of pointers. >> >> It's not clear to me what will break if you change this to >> isPtrOrPtrVectorTy. >> >> In fact, i know it doesn't fix this bug. >> It's a pretty deep rabbit hole of things not quite prepared to understand >> vectors of pointers. >> >> (we prepare memorylocations of them, but memory locations expect to be >> one thing, not a gro...

...gt; > it would also, for that matter, say the same about an array > > > > > of > > > > > pointers. > > > > > > > > > > > > > > > It's not clear to me what will break if you change this to > > > > > isPtrOrPtrVectorTy. > > > > > > > > > > > > > > > In fact, i know it doesn't fix this bug. > > > > > > > > > > > > > > > It's a pretty deep rabbit hole of things not quite prepared > > > > > to &g...

...t dberlin.org> >>>> wrote: >>>> >>>>> it would also, for that matter, say the same about an array of >>>>> pointers. >>>>> >>>>> It's not clear to me what will break if you change this to >>>>> isPtrOrPtrVectorTy. >>>>> >>>>> In fact, i know it doesn't fix this bug. >>>>> It's a pretty deep rabbit hole of things not quite prepared to >>>>> understand vectors of pointers. >>>>> >>>>> (we prepare memorylocations...

...t;>> wrote: >>>>> >>>>>> it would also, for that matter, say the same about an array of >>>>>> pointers. >>>>>> >>>>>> It's not clear to me what will break if you change this to >>>>>> isPtrOrPtrVectorTy. >>>>>> >>>>>> In fact, i know it doesn't fix this bug. >>>>>> It's a pretty deep rabbit hole of things not quite prepared to >>>>>> understand vectors of pointers. >>>>>> >>>>>> (we...