Re: [tip:irq/numa] x86, apic: Fix dummy apic read operation together with broken MP handling

[Cyrill Gorcunov <gorcunov@gmail.com>]

[Cyrill Gorcunov - Mon, Jun 08, 2009 at 07:16:46PM +0400]
...
| | should be other case:
| | when MADT is right, but disablelapic is used.
| | will get cpu_has_apic == 0, and we are not using dummy apic read/write.
| | 
| | so don't need to check
| |         /*
| |          * If we couldn't find a local APIC, then get out of here now!
| |          */
| |         if (APIC_INTEGRATED(apic_version[boot_cpu_physical_apicid]) &&
| |             !cpu_has_apic) {
| |                 if (!disable_apic) {
| |                         pr_err("BIOS bug, local APIC #%d not detected!...\n",
| |                                 boot_cpu_physical_apicid);
| |                         pr_err("... forcing use of dummy APIC emulation."
| |                                 "(tell your hw vendor)\n");
| |                 }
| |                 smpboot_clear_io_apic();
| |                 arch_disable_smp_support();
| |                 return -1;
| |         }
|
...

I've just emulated the situation where this if() triggered via
disable_apic boot option and without if (!disable_apic) we
have just a wrong message about BIOS bug (which is not since
we've disabled apic by hands). So no need to remove this
"if (!disable_apic)" snippet. In most cases it will be
always-true condition but for rare cases it'll be usefull
as well.

Now I'm trying to fake apic integrated case so we could
need additional check here...
 
	-- Cyrill