Re: patch: new sysctl to remove hardcoded ELOOP limit

Andi Kleen (ak@muc.de)
09 Apr 1998 11:37:20 +0200


alan@lxorguk.ukuu.org.uk (Alan Cox) writes:

> > > Currently 5 symlinks will be followed before ELOOP is returned. Today
> > > I had a need for this to be 6 on my system. Rather than just hacking the
> > > value and recompiling it I thought it made more sense to have a sysctl so
> > > this is adjustable.
> >
> > Lovely little patch!!
> >
> > One comment: isn't there a /proc/sys/fs in 2.1.x? And if so wouldn't the
> > value be better suited there?
>
> This reminds me. At some point we need an internal or externally visibile
> O_DONTFOLLOWSYMLINKS for core dumping. FreeBSD has an externally visiible
> version. Im not sure if we need that or if its useful to users.

Another useful bit from FreeBSD: a O_NOFOLLOW option for open(). This
is a nice idea will prevent lots of /tmp races.

-Andi

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu