This is the mail archive of the
mailing list for the glibc project.
Re: PI mutex support for pthread_cond_* now in nptl
On Tue, Feb 19, 2013 at 05:18:20PM +0000, Joseph S. Myers wrote:
> There are however two different groups of those headers, those where e.g.
> lll_futex_timed_wait ends with
> INTERNAL_SYSCALL_ERROR_P (__ret, __err) ? -__ret : __ret;
> (e.g. powerpc) and those (e.g. s390) where it just ends with a call to
> INTERNAL_SYSCALL and relies on the return value of INTERNAL_SYSCALL being
> the correct return from the macro. I'm not sure whether those differences
> actually serve a useful purpose or whether the return value of
> INTERNAL_SYSCALL differs in a relevant way between architectures.
>From the very superficial analysis I had done last year, it should be
possible to consolidate all that.