On Oct 13 06:01, Eric Blake wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> According to Christian Franke on 10/11/2009 2:45 PM:
> > 2009-10-11 Christian Franke <[email protected]>
> > Corinna Vinschen <[email protected]>
> >
> > * include/sys/cygwin.h: Add new cygwin_getinfo_type
> > CW_SET_EXTERNAL_TOKEN.
> > Add new enum CW_TOKEN_IMPERSONATION, CW_TOKEN_RESTRICTED.
>
> Shouldn't we also bump version.h when adding new CW_ flags?
You're right. Done in CVS.
> > + case CW_SET_EXTERNAL_TOKEN:
> > + {
> > + HANDLE token = va_arg (arg, HANDLE);
> > + int type = va_arg (arg, int);
> > + set_imp_token (token, type);
> > + return 0;
> > + }
>
> Not the first time this is done in this function. But generally,
> shouldn't we follow the good practice of using va_end any time we used
> va_arg, in case cygwin is ever ported to a system where va_end is more
> than a no-op? [At least, I'm assuming that __builtin_va_end() is a no-op
> for x86?]
That's probably a good idea, given that POSIX requires the usage of
va_end. PTC?
Corinna
--
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Project Co-Leader cygwin AT cygwin DOT com
Red Hat
