Re: bug#32452: 26.1; gnutls_try_handshake maxes out cpu retrying when server is a bit busy

[Eli Zaretskii]

> From: Lars Ingebrigtsen <larsi@gnus.org>
> Cc: eggert@cs.ucla.edu,  bug-gnulib@gnu.org,  npostavs@gmail.com,
>   32452@debbugs.gnu.org
> Date: Mon, 28 Feb 2022 13:35:47 +0100
> 
> Lars Ingebrigtsen <larsi@gnus.org> writes:
> 
> > grep --color=auto -nH --null -e rpl_nanosleep `find . -type f`
> > ./lib/time.h647:#    define nanosleep rpl_nanosleep
> > ./lib/time.in.h140:#    define nanosleep rpl_nanosleep
> >
> > And that's it.
> 
> The header file continues:
> 
> # if 0
> #  if GNULIB_PORTCHECK
> #   if !(defined __cplusplus && defined GNULIB_NAMESPACE)
> #    define nanosleep rpl_nanosleep
> #   endif
> _GL_FUNCDECL_RPL (nanosleep, int,
>                   (struct timespec const *__rqtp, struct timespec *__rmtp)
>                   _GL_ARG_NONNULL ((1)));
> _GL_CXXALIAS_RPL (nanosleep, int,
>                   (struct timespec const *__rqtp, struct timespec *__rmtp));

My guess is that _GL_FUNCDECL_RPL should produce a rpl_nanosleep.

But I'm sure Gnulib folks will see the reason much faster than my
guesses.