[Chicken-users] Re: questionable http-server behaviour under heavyload: broken pipe

[Johannes Groedem]

* Felix Winkelmann <address@hidden>:

> which internally does a signal(SIGPIPE, SIG_IGN).  So I assumed the
> signal will be disabled. Apparently I was wrong. Or not?

It should be possible to ignore it, and if you do, you'll get an
EPIPE-error from write(2), if I remember correctly. (The manpage seems
to agree with me.)

-- 
Johannes Groedem <OpenPGP: 5055654C>