Re: Switch APIC (+nmi, +oprofile) to driver model

From: Pavel Machek (pavel@ucw.cz)
Date: Mon Feb 10 2003 - 11:10:31 EST


Hi!

> >+ set_fixmap_nocache(FIX_APIC_BASE, apic_phys); /* FIXME: this is needed for S3 resume, but why? */
>
> This is horrible! The only reason this might be needed is if
> the page tables weren't restored properly at resume, and that
> indicates a bug somewhere else.

Pagetables should be restored okay, I do not know what is going on.

> >+static int __init init_apic_devicefs(void)
> > {
> >+ driver_register(&apic_driver);
> > if (apic_pm_state.active)
> >- pm_register(PM_SYS_DEV, 0, apic_pm_callback);
> >+ return sys_device_register(&device_apic);
> >+ return 0;
> ...
> >+device_initcall(init_apic_devicefs);
>
> init_apic_devicefs() registers apic_driver even if active is false.
> This probably breaks any machine where cpu_has_apic is false, since
> apic_suspend/resume will access non-existent APIC registers & MSRs.
>
> I don't like the idea of registering apic_driver when !cpu_has_apic,
> but it might be needed for the local-APIC NMI watchdog. A fix could
> be to guard apic_suspend/resume with "if(!cpu_has_apic)return;".

cpu_has_apic test seems wrong: AFAICS, there are CPUs that have apic
but linux does not know how to use it.

> > nmi_watchdog = nmi;
>
> You just killed NMI_LOCAL_APIC support on Intel.

Fixed.

> >+device_initcall(init_nmi_devicefs);
>
> Doesn't this require that init_apic_devicefs() has been done?
> Can you guarantee that?

Moved to late_initcall().

> >
> >+#define __pminit
> > #endif /* CONFIG_PM */
>
> __pminit is no longer defined when !CONFIG_PM, which will
> cause compile errors. Possibly the remaining uses of __pminit
> should be removed.

Fixed.
                                                                Pavel

-- 
Worst form of spam? Adding advertisment signatures ala sourceforge.net.
What goes next? Inserting advertisment *into* email?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sat Feb 15 2003 - 22:00:33 EST