Re: [TOMOYO #15 0/8] TOMOYO Linux

[Pavel Machek <pavel@ucw.cz>]

Hi!

> > Yes. maybe ioctl() is worse, but I don't think c-like language parser
> > in kernel is acceptable.
> 
> for just clarification to me.
> 
> IIUC, many developers said UNNECESSARY parser is BAD (yes, I also think so),
> but nobody said any parser is bad.
> 
> Therefore, I think point is that the patch have enough reasonable reason or not.
> and, I thought "pavel, good job. you're right" at you oppositing time because
> tomoyo did't explain any reason at that time.
> 
> However, they changed. the patch description of the "[TOMOYO #15 3/8] Common functions for TOMOYO Linux."
> explain the reason.
> for me, I feel it's reasonable reason. then I didn't oppose current tomoyo posting.
> 
> So, I don't understand which you oppose
>    (1) ANY parser is bad.
>    (2) current description still don't explain enough reason.
> 
> May I ask you?

I'm not sure if I've seen all the TOMOYO patches... But from what I've
seen of TOMOYO design, putting the parser into kernel was "just
because"; it did not have any good reason. I hate to say that, but
AppArmor was better designed there.
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html