Re: exceptions & dataflow

Sergey Solyanik (sergey@solyanik.com) wrote:
>: In Java, hardware exceptions can ony happen in very controlled manner
>: (e. g. to test for null). If your language contains pointers, things
>: are much worse and you'd have to break basic block with essentially
>: any pointer dereference when you cannot prove that location exists.


mcdirmid@beaver.cs.washington.edu (Sean McDirmid) writes:
> I would be very interested to see how it was done by Sun or other
> compilers with exception support (modula 2, C++, etc..).


In C++, dereferencing a null or dangling pointer results in undefined
behaviour. Typically on Unix systems the program will get a signal,
not an exception. I suppose the user might try to throw an exception
from the signal handler, but that would be undefined behaviour too.
--
Fergus Henderson <fjh@cs.mu.oz.au>
WWW: <http://www.cs.mu.oz.au/~fjh>
PGP: finger fjh@128.250.37.3
--

Return to the comp.compilers page.
Search the comp.compilers archives again.