Re: A fifo and signal bug

H.J. Lu (hjl@lucon.org)
Sat, 21 Nov 1998 10:38:33 -0800 (PST)

>
>
>
> On Sat, 21 Nov 1998, H.J. Lu wrote:
> >
> > No. It is a kernel bug. You can see it by just looking at fs/filo.c.
> > Basically, when a fifo reader is interrupted by a signal in fifo_open,
> > there is no way for a fifo writer to tell someone is waiting to read
> > but interrupted by a signal. But I am afraid a real fix is not trivial.
>
> So send me a program that shows this. The test program I got appears to
> pass, and as such I don't see your point at all.
>

It is very strange. My test program fails for everyone else. Could
you please try it on an UP machine running the stock 2.1.129 UP kernel?

BTW, I put 5 sec wait between sending SIGTSTP and trying to open
the fifo with O_WRONLY|O_NONBLOCK. It may not be enough or someone
else was reading/trying to open the fifo to read.

Thanks.

-- 
H.J. Lu (hjl@gnu.org)


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/