Messages in this thread | | | From | wtenhave@sybase ... | Date | Wed, 15 Sep 1999 17:19:44 +0200 (CEST) | Subject | Re: mlockability |
| |
On 15 Sep, Juan Carlos Castro y Castro wrote: > un-swappable, one uses the mlock(2) call, which only root has the right > to use. Hence the suid.
The suid root is well in place here. What cripples is that the max size for the to phys locked memory segment which does not allow for more then 512 MByte.
> Couldn't there be a kernel option (maybe tunable by /proc/something) > through wich the kernel could grant a limited amount of mlock-able > memory for each non-root process? This could default to 0 but be > increased when sensitive software is being used.
Implementing something like SIGDANGER (available under AIX 4.x for example) is much nicer. I.e. reserve a signal for processes becoming a thread for the total available virtual memory pool. This to suite any virtual memory exhaustion.
-- Wim ten Have.
- 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/
| |