On Tue, Apr 29, 2008 at 10:59 AM, Diego Novillo <[EMAIL PROTECTED]> wrote: > On 4/29/08 1:31 PM, Joe Buck wrote: > > > > Such a facility can have other uses, particularly for static analysis, > > by allowing simple preconditions and postconditions to be specified. > > For example: > > > > * a returned pointer is guaranteed to be non-null. > > * a supplied pointer is always dereferenced. > > * a supplied pointer must be dereferenceable on input, and that pointer > > is no longer dereferenceable after return, e.g. free(). > > > > Indeed.
We already have the second one, called nonnull The first one is recorded as PR 20318. Thanks, Andrew Pinski
