[cfe-commits] [PATCH][Review request]Support for c++0x nullptr in static analyzer

[Ted Kremenek]

On Apr 20, 2011, at 10:56 PM, Jim Goodnow II wrote:

> Ok, so the key difference in the static analyzer is run-time actions as opposed to compile-time choices.

Yes, that's one way to look at it.  SVals represent the abstract "semantics" of an expression.  Essentially think of the static analyzer as a symbolic simulator for C.

> It still seems like there could be a run-time situation where you cared about the difference between a zero-valued pointer and a 'null' pointer. But, in reality, I'm guessing that the 'nullptr' type would still be implemented as a zero-valued pointer, although it could be anything including 0xDEADBEEF!

As Argiris pointed out in his email, I don't think that is ever the case.

> And cases where a zero valued pointer is an intended use, such as the offsetof() macro,

In this case, I think nullptr would likely be casted first to the pointer of the appropriate type.

> would be unlikely enough to be passed a 'nullptr' that it's not worth modeling?

If we care about nullptr, we can locally walk the AST to see if a null value arose from a nullptr.  Modeling the semantics specially seems to have no real value.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20110422/1e5fc489/attachment.html>