Re: Cyrix 6x86MX and Centaur C6 CPUs in 2.1.102

Michael L. Galbraith (mikeg@weiden.de)
Sat, 23 May 1998 08:53:57 +0200 (MET DST)


On Thu, 21 May 1998, [ISO-8859-1] André Derrick Balsa wrote:

> Hi,
>
> Vojtech Pavlik wrote:
> >
> > On Thu, May 21, 1998 at 03:44:27PM +0200, Martin Mares wrote:
> >
> > > > BTW TSC calibration in the setup code would allow showing MHz rating in
> > > > /proc/cpuinfo, so people would stop wondering why they get bogomips such
> > > > and such everytime a new x86 processor comes out.
> > >
> > > This would be interesting, but remember that CPU clock speed need not
> > > to be equal TSC frequency.
> >
> > Yes, but for all the chips produced up to now it is equal, and for later
> > chips we may decide based on the chip revision / manufacturer and if needed
> > apply a TSC per sec / CPU clocks per sec ratio for the respective chip to
> > get the correct CPU frequency.
> >
> > Vojtech
>
> Well, in fact the TSC *is* supposed to *exactly* count CPU clock cycles,
> so I don't see what Martin means by "CPU clock speed need not be equal
> to TSC frequency".

Not according to Intel :-/ We assume that realtime delta information
is present in TSC stamps while Intel clearly states that this is not
safe to assume.

Pentium Processor Family Developer's Manual Volume 1, Section 33.4.2:

"Portable software should not expect that the
counter reports absolute time or clock counts."

I read this to mean that only time sequence information is guaranteed
to be present in TSC stamps.. Cyrix breaks that.

-Mike

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu