> It used to be ~1 BogoMIP per (x86 ISA) processor Mhz. Now, it's ~2.
2 cycles/loop for P6, 1 for most AMD CPUs, Pentium MMX, > 2 for original
Pentiums (broken branch prediction). It basically was a completel mess.
> The calculation seems much more stable, due to the TSC being used. I like
> this.
It is more stable because it doesn't depend on how much work is being done
but on how much time passes. In fact, it should be stable enough now to
skip the calibration and just directly set loops_per_sec to the detected
TSC frequency.
> Surely though, being 2x out would not be good for lots of user-land apps
> using this as a {rough, simply} processor power measurement.
I don't think any applications are actually that broken.
Philipp Rumpf
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 21:00:09 EST