On Wed, Jul 13, 2011 at 2:48 AM, Phil Blundell <[email protected]> wrote: > On Tue, 2011-07-12 at 21:19 -0700, Khem Raj wrote: >> ++#ifndef sigmask >> ++/* Compute mask for signal SIG. */ >> ++#define sigmask(sig) __sigmask(sig) >> ++#endif /* sigmask */ > > That looks a bit dubious to me. If this is relying on some uclibc > implementation detail then I think it should either be applied for > uclibc only, or guarded with something like "#if defined(__UCLIBC__) > && !defined(sigmask)". Otherwise you might end up trying to call > __sigmask() on non-uclibc systems which could have unwanted effects.
in general for eglibc/glibc it wont matter since signal.h defines sigmask and is defined to point to __sigmask but for non glibc/eglibc libc yes there is a chance. > > p. > > > > _______________________________________________ > Openembedded-core mailing list > [email protected] > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core > _______________________________________________ Openembedded-core mailing list [email protected] http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
