[PATCH] x86: fix section mismatch in LAPIC initialization

[PATCH] x86: fix section mismatch in LAPIC initialization

[Jan Beulich]

Doing things conditionally upon spm_processor_id() being zero is
generally a bad idea, as this means CPU 0 cannot be offlined and
brought back online later again. While there may be other places where
this is done, I think adding more of those should be avoided so that
some day SMP really becomes "symmetrical".

Signed-off-by: Jan Beulich <jbeulich@novell.com>

---
 arch/x86/include/asm/apic.h |    1 +
 arch/x86/kernel/apic/apic.c |    9 +++++++--
 arch/x86/kernel/smpboot.c   |    4 ++--
 3 files changed, 10 insertions(+), 4 deletions(-)

--- 2.6.37-rc8/arch/x86/include/asm/apic.h
+++ 2.6.37-rc8-x86-end-lapic-setup/arch/x86/include/asm/apic.h
@@ -233,6 +233,7 @@ extern void sync_Arb_IDs(void);
 extern void init_bsp_APIC(void);
 extern void setup_local_APIC(void);
 extern void end_local_APIC_setup(void);
+extern void bsp_end_local_APIC_setup(void);
 extern void init_apic_mappings(void);
 extern void setup_boot_APIC_clock(void);
 extern void setup_secondary_APIC_clock(void);
--- 2.6.37-rc8/arch/x86/kernel/apic/apic.c
+++ 2.6.37-rc8-x86-end-lapic-setup/arch/x86/kernel/apic/apic.c
@@ -1389,12 +1389,17 @@ void __cpuinit end_local_APIC_setup(void
 
 	setup_apic_nmi_watchdog(NULL);
 	apic_pm_activate();
+}
+
+void __init bsp_end_local_APIC_setup(void)
+{
+	end_local_APIC_setup();
 
 	/*
 	 * Now that local APIC setup is completed for BP, configure the fault
 	 * handling for interrupt remapping.
 	 */
-	if (!smp_processor_id() && intr_remapping_enabled)
+	if (intr_remapping_enabled)
 		enable_drhd_fault_handling();
 
 }
@@ -1751,7 +1756,7 @@ int __init APIC_init_uniprocessor(void)
 		enable_IO_APIC();
 #endif
 
-	end_local_APIC_setup();
+	bsp_end_local_APIC_setup();
 
 #ifdef CONFIG_X86_IO_APIC
 	if (smp_found_config && !skip_ioapic_setup && nr_ioapics)
--- 2.6.37-rc8/arch/x86/kernel/smpboot.c
+++ 2.6.37-rc8-x86-end-lapic-setup/arch/x86/kernel/smpboot.c
@@ -1065,7 +1065,7 @@ static int __init smp_sanity_check(unsig
 
 		connect_bsp_APIC();
 		setup_local_APIC();
-		end_local_APIC_setup();
+		bsp_end_local_APIC_setup();
 		return -1;
 	}
 
@@ -1142,7 +1142,7 @@ void __init native_smp_prepare_cpus(unsi
 	if (!skip_ioapic_setup && nr_ioapics)
 		enable_IO_APIC();
 
-	end_local_APIC_setup();
+	bsp_end_local_APIC_setup();
 
 	map_cpu_to_logical_apicid();
 

Re: [PATCH] x86: fix section mismatch in LAPIC initialization

[H. Peter Anvin]

On 01/03/2011 07:03 AM, Jan Beulich wrote:
> Doing things conditionally upon spm_processor_id() being zero is
> generally a bad idea, as this means CPU 0 cannot be offlined and
> brought back online later again. While there may be other places where
> this is done, I think adding more of those should be avoided so that
> some day SMP really becomes "symmetrical".
> 
> Signed-off-by: Jan Beulich <jbeulich@novell.com>

The description doesn't really match the subject line...

	-hpa

Re: [PATCH] x86: fix section mismatch in LAPIC initialization

[Jan Beulich]

>>> On 03.01.11 at 19:29, "H. Peter Anvin" <hpa@zytor.com> wrote:
> On 01/03/2011 07:03 AM, Jan Beulich wrote:
>> Doing things conditionally upon spm_processor_id() being zero is
>> generally a bad idea, as this means CPU 0 cannot be offlined and
>> brought back online later again. While there may be other places where
>> this is done, I think adding more of those should be avoided so that
>> some day SMP really becomes "symmetrical".
>> 
>> Signed-off-by: Jan Beulich <jbeulich@novell.com>
> 
> The description doesn't really match the subject line...

The subject, imo, doesn't need further explanation. The description
was meant to explain why the function gets broken up rather than
just fiddling with other (called) functions' annotations. Are you
suggesting to drop the description altogether?

Jan