search for: vtable_access

...and are expressed neutrally in a simple reachability language" or "do we add language-specific kinds of nodes to the grammar, and have reachability rules that are fairly language specific". IE do you add, say, discriminated_union nodes to our current representation for ada, or "vtable_access" nodes to our current representation for C++ vtable accesses Or do you instead generate a metadata that has a unidirectional edge reachability (IE up only), or whatever it takes to do vtables generically. Both are completely and totally viable paths, and it's all about which way you want...

...I don't want language-specific kinds of nodes in the grammar, and I > believe that it's not necessary under some reasonable variant of this > scheme. Maybe I'm wrong. > > > IE do you add, say, discriminated_union nodes to our current > representation for ada, or "vtable_access" nodes to our current > representation for C++ vtable accesses > Or do you instead generate a metadata that has a unidirectional edge > reachability (IE up only), or whatever it takes to do vtables generically. > > Both are completely and totally viable paths, and it's all a...

On Thu, Aug 17, 2017 at 4:49 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > On 08/17/2017 04:49 PM, Daniel Berlin via llvm-dev wrote: > > >> >> For two given access sequences we can determine if the accessed >> objects are allowed to overlap by the rules of the input >> language. > > > Sadly, this is where this becomes "unlikely to want to