after about another 14 hours of hacking, several reboots, compiles,
and significant frustration I'd like to announce the second snapshot
of the PPSkit-0.6, also known as `nanokernel', because it brings
nanoseconds into the kernel.
Relative to the first snapshot I've made more renamings (the kernel
lords will curse me, I guess), and basically the thing is ready for
true nanoseconds resolution.
I only have one problem left: calibrate_tsc() is currently disabled,
because it will casue a divide overflow when trying to calibrate for
nanoseconds. Maybe I was to tired when I tried to fix it, but my
solution didn't work. Maybe someone can open my eyes...
I did two short basic tests on my Pentium 100: Run xntpd3-5.93
(quickly hacked and recompiled) with a "PPS" source (Meinberg
GPS167), and it looked good. Then I let the quartz run free and
measured the offset against the PPS pulse. It also gave a straight
line. So I guess time is still working...
Locations are master.kernel.org/pub/linux/daemons/ntp/PPS and
ftp://pcphy4.physik.uni-regensburg.de/pub/wiu09524/PPS, the file is
PPSkit-0.6-i386-pre2.diff.gz.
The kernel will call itself "2.2.1-NANO"...
Feedback is welcome. I hope to be able to finish the nanokernel soon
to return to other projects (like my tax declaration 8-().
Gurus for the non-i386 architecture can preview the stuff, but maybe
not yet implement similar routines as long as the interfaces haven't
settled down.
Especially we have every time routine doubled now, once for
microseconds, once for nanoseconds. I'd also like to minimize the
architecture-specific time.c. I'll have to see what I can do...
Regards,
Ulrich Windl
-
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/