> On Thu, Mar 14, 2002 at 10:53:01AM -0500, Bill Davidsen wrote:
> It's not a matter of codesize, it's a correctness issue in the source.
> #ifndef CONFIG_M686 is wrong. It assumes a P6 is the only CPU family
> in existence without the bug, despite the fact there are probably close
> to a dozen others.
No problem with that, it obviously should be done for all chips which
could have the bug, both Intel and faithful clones :-(
> > that many people would set it off not knowing was it was much less whether
> > they needed it. This is not like a missing FPU where you can do a graceful
> > reject of the instructions, if you have the bug and not the fix you are
> > vulnerable to sudden total failures, correct?
>
> No. You at worse vulnerable to a malicious user running hand-crafted code
> (no compiler generates this code-sequence) bringing down the machine.
Having the machine lock suddenly when user mode code is executed fits MY
definition of sudden total failure! Malicious user meaning anyone who d/l
something and runs it. Doesn't take malicious, just dumb. Somone wanting
to try the new screen saver or some such.
> The proposal however was not to remove anything that we currently have.
> Every kernel that is possible to be run on an affected box (i386/i486/i586)
> would still have the workaround present. We just won't generate it in
> Cyrix III, Athlon, Pentium 4, etc kernels..
As long as it's not something the user can easily set wrong. If they
build a kernel for K6 and they have a P-II "stupidity, like virtue, is its
own reward."
-- bill davidsen <davidsen@tmr.com> CTO, TMR Associates, Inc Doing interesting things with little computers since 1979.- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/