Re: Linux 2.1.125 doesn't dump core on SIGSEGV

Alan Cox (alan@lxorguk.ukuu.org.uk)
Wed, 21 Oct 1998 00:42:47 +0100 (BST)

> On Tue, Oct 20, 1998 at 02:38:55PM +0100, David Woodhouse wrote:
> > How about providing a new flag to open(2) which tells it not to
> > follow links?
>
> Because the VFS doesn't support this - and it still doesn't help in
> the hard link instance.
>
> I've yet to here why stat then a sanity check won't work.

A stat is non atomic. Even an open and lstat is non atomic on NFS. Doing
local locking helps but you don't want to let lusers single thread the VFS
at while when we are targetting 64 processors...

To wear a linus impersonator hat
O_NOFOLLOW is the generic solution to that and a pile of other link handling
problems

Alan

-
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/