On Tue, 27 Sep 2005 15:23:39 EDT, Steve Grubb said:
On Tuesday 27 September 2005 01:57, Valdis.Kletnieks(a)vt.edu wrote:
> > 1500 - 1599 kernel LSPP events
> > 1700 - 1799 kernel crypto events
> > 1800 - 1999 future kernel use (maybe integrity labels and related events)
>
> < and so on..>
>
> Am I the only one who thinks "100 entries will be enough" sounds
> suspiciously like "640K should be enough for anybody"?
I'm thinking it should be enough unless vendors want to clip their programs
into the audit system and start inventing their own numbers.
That's basically what I was worried about.
Well, we could easily continue the same kind of messages in another
block. So
far, we've only consumed < 20 message types on any block. I really can't see
a 100 different kinds of LSPP kernel message types.
That's just waiting for a bug report when somebody's software doesn't play
nice
with "2100..2199,2700..2799" type ranges. Agreed the kernel probably won't
need
a lot more, but we might want to make sure we have bigger ranges available for
userspace. Maybe make each userspace range 1K wide rather than 100? (Yes,
I know that's just putting the Y2K problem off to 2038 ;)