On Thu, 20 Apr 2000, Tigran Aivazian wrote:
> yes, he is probably right but do you know enough of discrete 82489 APIC
> internals to write the send IPI function that handles the subtleties he
> is talking about and also have such hardware to test it? I don't.
See how APIC_INTEGRATED macro is used when the NMI watchdog is being set
up in io_apic.c and take a similar approach. Apart from treating NMIs as
level-triggered, the 82489DX isn't much different in this matter. See how
INIT assert and deassert IPIs are being sent in do_boot_cpu(), too --
beside delays which are specifically needed for INIT IPIs, sending of NMIs
should be done in the same way for the 82489DX.
Testing changes is a different matter, though. I know of four 82489DX
systems out there, of which two boot Linux MP as of -pre6 (the other two
still have problems with the PIC compatibility mode, but that's unrelated
here). I don't know whether their users want to volunteer testing --
you'd need to ask at linux-smp, probably. The NMI watchdog works for
them.
-- + Maciej W. Rozycki, Technical University of Gdansk, Poland + +--------------------------------------------------------------+ + e-mail: macro@ds2.pg.gda.pl, PGP key available +- 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 : Sun Apr 23 2000 - 21:00:17 EST