[Prev][Next][Index]

Re: SgFunctionSymb::numberOfParameters returns bogus value



This is not a previously known bug. But it looks like a bug  
nevertheless.


>Running dumpdep is unhelpful; the output looks fine with a parameter
>list of the correct length.


Could you run dumpdep again and compare symbol table entries for
 Processor_Main
before and after the instrumentation?
In particular, check out the field called in_l
it should be 

in_l=   ----
in both cases, but I suspect that it becomes something else after  
instrumentation. If this is indeed the case, then this would explain  
why numberOfParameters does not work correctly (and I can try to fix  
Sage so it won't mess up this field when a symbol is added inside a  
function). If not, then we have to look somewhere else for  
explanation.   --Beata