Date: Sun, 23 Nov 1997 22:52:07 -0600 (CST) From: "Daniel M. Eischen" <deischen@iworks.InterWorks.org> To: julian@whistle.com, nash@mcs.com Cc: freebsd-hackers@freebsd.org Subject: Re: pthread_cond_timedwait returning wrong error? Message-ID: <199711240452.WAA24529@iworks.InterWorks.org>
next in thread | raw e-mail | index | archive | help
> don't forget 2.2.x
> (it has basically the same code)
>
> On Sun, 23 Nov 1997, Alex Nash wrote:
>
> > On 23 Nov, Daniel M. Eischen wrote:
> > >
> > > Is EAGAIN the correct error for a timeout from a
> > > pthread_cond_timedwait() call? I would think that
> > > ETIMEDOUT would be more appropriate.
> >
> > According to my 1996 edition of ANSI/IEEE std 1003.1, you are correct --
> > ETIMEDOUT should be returned. I've committed a fix for this into
> > -current.
> >
> > Thanks.
> >
> > Alex
While we're on the subject, can someone please add uthread_sigwait
to Makefile.inc in lib/libc_r/uthread? I've been wanting this for
a while; the GNAT port needs this to work. The code is there, it's
just that the function isn't included in the make.
*** Makefile.inc.orig Sun Nov 16 19:06:14 1997
--- Makefile.inc Sun Nov 23 23:42:23 1997
***************
*** 83,88 ****
--- 83,89 ----
uthread_sigprocmask.c \
uthread_sigsetmask.c \
uthread_sigsuspend.c \
+ uthread_sigwait.c \
uthread_socket.c \
uthread_socketpair.c \
uthread_spec.c \
One last question. The other GNAT targets (Linux, Open/VMS,
RTEMS, etc) all seem to want the errno to be _returned_ by
pthread_cond_timedwait (and a couple of others too). This
seems wrong, -1 should be returned, and the error code should
be retrieved from the thread-safe errno. What does the POSIX
spec say about the return value of pthread_cond_timedwait()?
Dan Eischen
deischen@iworks.InterWorks.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199711240452.WAA24529>