Messages in this thread | | | Date | Tue, 17 May 2005 17:57:56 +0100 | From | Al Viro <> | Subject | Re: [patch 2/7] BSD Secure Levels: move bd claim from inode to filp |
| |
On Tue, May 17, 2005 at 05:49:22PM +0100, Al Viro wrote: > a) have caused open() on that device node, even though caller did > not ask for that and actually had not planned to do anything with actual > device. > b) have caused all subsequent permission() for MAY_WRITE fail for > that sucker [*] until somebody opens and closes device in question (for > read, obviously).
Actually, that is correct only if that somebody gets the same task_struct as deceased caller of utime()/whatever had triggered call of permission(). Due to (c) below...
> c) seclvl_bd_release() expects, for some reason, to be called when > task that had called seclvl_bd_claim() to be still alive. Use of current > in setting/checking ->i_security is a bad joke. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |