Shane Kerr wrote:
>
> On 2001-04-16 00:49:55 +0000, Karim Yaghmour wrote:
> >
> > May I point out that "should be" is different from "can be" and I'm
> > afraid you'll find it hard to reach the 0.1% mark unless you resort to
> > self-modifying code.
>
> Huh? On modern CPU's, self-modifying code is a huge performance killer.
> Which is good, because it's also a nightmare to maintain. I think this,
> along with the bitmask idea, can be safely sent to /dev/null. Certainly
> nobody is going to stop anyone who wants to play around with performance
> tweaks like this, but I hate to see anyone waste a lot of time on
> techniques that won't work. :(
Thanks, but I didn't mean eternally self-modifying code. What I
meant is that you insert a couple of nops in the normal flow
of code and when you want to hook onto an event, you simply
overwrite the nops with a call to the function you want. This
isn't something that should happen more than once in normal
system behavior. Hence, it costs a lot the first time, but
after that, it's business as usual.
Cheers,
Karim
===================================================
Karim Yaghmour
karymat_private
Embedded and Real-Time Linux Expert
===================================================
_______________________________________________
linux-security-module mailing list
linux-security-moduleat_private
http://mail.wirex.com/mailman/listinfo/linux-security-module
This archive was generated by hypermail 2b30 : Mon Apr 16 2001 - 03:13:01 PDT