Re: [PATCH] Scheduling problems wrt SMP

[MOLNAR Ingo]

On Fri, 18 Dec 1998, David S. Miller wrote:

> We already do stuff like this, see sys_getppid().

unrelated, but i dont think sys_getppid() is conceptually safe. We could
theoretically have the parent switched twice under us, with
allocation-reuse, which makes us accept a pid taken from a structure that
was not allocated at the time of dereferencing ... This is extremely
unlikely, and we dont care much about the parent pid in this case, but
what if the invalid parent pid happens to be a pid of a security-relevant
process, and the getppid() is done as root. (even more unlikely but
theoretically possible ;)

just nitpicking,

-- mingo


-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]