X-Inside posts Linux bug, has it been reported anywhere else?

John Henders (jhenders@bogon.com)
16 Nov 1996 17:39:43 -0800

I saw this posted on comp.windows.x.i386unix. Has it been reported
anywhere else? Is it fixed in a 2.0.* kernel?

>From comp.windows.x.i386unix Thu Nov 14 14:06:51 1996
Path: bogon!van-bc!news.rmii.com!usenet
From: roell@xinside.com (Thomas Roell)
Newsgroups: comp.windows.x.i386unix
Subject: Re: I want to kill my computer (or, help me install X please)
Date: 12 Nov 1996 09:27:42 -0700
Organization: X Inside Inc.
Lines: 33
Sender: roell@xeno.xinside.com
Message-ID: <87enhzuvlt.fsf@xeno.xinside.com>
References: <568ftj$n5k@news.doit.wisc.edu>
NNTP-Posting-Host: fog.xinside.com
In-reply-to: orf@scrap.ssec.wisc.edu's message of 12 Nov 1996 00:21:07 GMT

In article <568ftj$n5k@news.doit.wisc.edu> orf@scrap.ssec.wisc.edu (Leigh Orf) writes:

> has really been the problem all along, it seems. So, initially, we
> install Slackware with whatever solid version of XFree86 was out there
> a couple months ago, and buy the acceleratedX server (v2.1) which is
> made by X Inside and which supports the Matrox Millenium. Everything is
> peachy... except that windows will suddenly just "go away" with either a
> KillClient message or return a signal of 255. So I get in touch with the
> X Inside people, and they say "Oh yeah, we're aware of that and we're
> working on it" and tell me to subscribe to their patches mailing list.

We found now the problem. Basically it's a LINUX kernel bug. It seems
to appear at least on kernels between 2.0.0 and 2.0.22. Older kernels
do work fine. In case I don't bore anybody, here what's really
happening. If you do a read() system call in a non-blocking device,
and it get's interrupted in the system call (which kind of should not
happen at all), then a posix compilant system (and LINUX for that
matter) should return an error of EAGAIN. However the kernels
mentioned above randomly return either EINTR (which is for blocking
devices) or ERESTARTSYS, which the docs say "Should never be seen by
user programs".

So you might want to go back to an older kernel. We are here trying to
figure out whether there is an easy patch the kernel to correct this
behaviour, since it breaks other programs, like 'telnet' as well.

- Thomas

Denver Office               THOMAS ROELL        /\      Das Reh springt hoch,
+1(303)298-7478             X INSIDE INC       /  \/\   das Reh springt weit,
1801 Broadway, Suite 1710                     /    \ \/\     was soll es tun,
Denver, CO 80202          roell@xinside.com  / Oelch! \ \     es hat ja Zeit.

      Artificial Intelligence stands no chance against Natural Stupidity.
                GAT d- -p+(--) c++++ l++ u++ t- m--- W--- !v
                     b+++ e* s-/+ n-(?) h++ f+g+ w+++ y*