Ah ...
On Tue, 20 Feb 2001, Geoff Thorpe wrote:
> > @@ -67,8 +67,8 @@
> > {
> > MemCheck_off();
> > name_funcs = OPENSSL_malloc(sizeof(NAME_FUNCS));
> > - name_funcs->hash_func = lh_strhash;
> > - name_funcs->cmp_func = strcmp;
> > + name_funcs->hash_func = (LHASH_HASH_FN_TYPE)lh_strhash;
> > + name_funcs->cmp_func = (LHASH_COMP_FN_TYPE)strcmp;
> > name_funcs->free_func = 0; /* NULL is often declared to
> > * ((void *)0), which according
> > * to Compaq C is not really
>
> the cmp_func and hash_func entries of the NAME_FUNCS structure are the same
> prototypes as strcmp and lh_strhash. So these casts are currently unecessary and
> have in fact introduce an "incompatible" pointer warning. However, what's
> actually getting used in the lhash code *should* be of the
> LHASH_[HASH|COMP]_FN_TYPE types so I'll have to figure out where that's at.
> Working on it ...
I should have waited a couple of minutes before sending that ... those callbacks
*aren't* used in lhash callbacks, that's why the prototypes are fine as they
were. I understand now how easy the mistake was to make :-)
Cheers,
Geoff
______________________________________________________________________
OpenSSL Project http://www.openssl.org
Development Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
