Coda File System

Re: still some problems to solve in 5.3.17

From: Jan Harkes <jaharkes_at_cs.cmu.edu>
Date: Mon, 7 Jan 2002 11:26:14 -0500
On Sun, Jan 06, 2002 at 12:25:10AM +0100, Ivan Popov wrote:
> I'm here to bother you with the same things again and again :)

That's ok, a reminder once in a while saves me from keeping a TODO list
around ;)

> Sorry for that but I assume the developers want to know -
> symbolic link traversal during exec() is unreliable.

I haven't gotten a clue on this one yet, it's either a bug in Venus that
noone has noticed until you did, or some bug/race in the generic symlink
traversal code in recent Linux kernels that is triggered by the slowness
of Coda upcalls and in real life doesn't affect 'typical' filesystems.

> PATH=/coda/A
> /coda/A/sed -> /coda/B/sed
> /coda/B/sed:
>    #!/bin/sh
>    exec /coda/C/sed "$@"
> 
> Well, *sometimes* I get a message:
> can not execute /coda/A/sed: no such file or directory

Hmm, tried this, but even with a 'while true; do sed ... ; done' I
couldn't get it to fail.

> Another thing - if I let compilations or similar massive updates on /coda
> run without "cfs strong; cfs wf -on", venus dies rather soon
> after several "reintegration: Success" messages, with Signal 11.

I'll look into this as well. Sig 11's are typically null pointer
dereferences and definitely bugs.

Jan
Received on 2002-01-07 11:27:01