Re: Intel Pentium Bug: BSDI Releases a patch

Alan Cox (alan@lxorguk.ukuu.org.uk)
Thu, 13 Nov 1997 13:47:01 +0000 (GMT)


> > > instruction, all others starting at interrupt 7 which is the quite critical
> > > (FPU context switches) should not experience any performance degradation.
> > Yep
>
> So am I correct in understanding that intel will probably escape a
> multibillion dollar replacement program because the illegal
> instruction interupt number is lower than the FPU one?

Well that would assume there arent other alternative and maybe better fixes.
I suspect even if the FPU performance had dropped a bit Intel wouldn't be
replacing chips - they've have got a passable work around for a product that
isnt warranted to be correct.

At the moment its not clear if the page table trick fixes all cases, and there
may be other pentium bugs. Robert Collins has stated he knows an even more
serious bug that allows any task to become privileged, and done so on usenet
given his apparent nonfriendly relationship with intel the fact this has
neither been refuted or him sued makes me suspect he is telling the truth.

Alan