fix generic/up builds in 2.6.12

Resending with a fixed subject & correct cc addr :)

Here is a resubmittal of Jesse Barnes' fix for generic UP builds, ported to 2.6.12. I've boot tested it on an rx2600.

Can someone test on an sn2 system?

diff -urN linux-2.6.12.orig/arch/ia64/kernel/Makefile linux-2.6.12/arch/ia64/kernel/Makefile
@@ -17,6 +17,7 @@
 obj-$(CONFIG_IOSAPIC)		+= iosapic.o
 obj-$(CONFIG_MODULES)		+= module.o
 obj-$(CONFIG_SMP)		+= smp.o smpboot.o domain.o
+obj-$(CONFIG_NUMA)		+= numa.o
 obj-$(CONFIG_PERFMON)		+= perfmon_default_smpl.o
 obj-$(CONFIG_IA64_CYCLONE)	+= cyclone.o
 obj-$(CONFIG_IA64_MCA_RECOVERY)	+= mca_recovery.o

diff -urN linux-2.6.12.orig/arch/ia64/kernel/acpi.c linux-2.6.12/arch/ia64/kernel/acpi.c
@@ -642,9 +642,11 @@
 			if (smp_boot_data.cpu_phys_id[cpu] != hard_smp_processor_id())
 				node_cpuid[i++].phys_id = smp_boot_data.cpu_phys_id[cpu];
 	}
-	build_cpu_to_node_map();
 # endif
 #endif
+#ifdef CONFIG_ACPI_NUMA
+	build_cpu_to_node_map();
+#endif
 	/* Make boot-up look pretty */
 	printk(KERN_INFO "%d CPUs available, %d CPUs total\n", available_cpus, total_cpus);
 	return 0;

diff -urN linux-2.6.12.orig/include/asm-ia64/sn/arch.h linux-2.6.12/include/asm-ia64/sn/arch.h
@@ -11,6 +11,7 @@
 #ifndef _ASM_IA64_SN_ARCH_H
 #define _ASM_IA64_SN_ARCH_H
 
+#include <linux/numa.h>
 #include <asm/types.h>
 #include <asm/percpu.h>
 #include <asm/sn/types.h>

diff -urN linux-2.6.12.orig/include/asm-ia64/sn/sn_cpuid.h linux-2.6.12/include/asm-ia64/sn/sn_cpuid.h
@@ -81,11 +81,6 @@
  *
  */
 
-#ifndef CONFIG_SMP
-#define cpu_physical_id(cpuid)			((ia64_getreg(_IA64_REG_CR_LID) >> 16) & 0xffff)
-#endif
-
-
 #define get_node_number(addr)			NASID_GET(addr)
 
 /*

dann frazier <[email protected]>

On Wed, Jun 29, 2005 at 01:33:32PM -0600, dann frazier wrote:

| Resending with a fixed subject & correct cc addr :)
| 
| Here is a resubmittal of Jesse Barnes' fix for generic UP builds, ported
| to 2.6.12.  I've boot tested it on an rx2600.
| 
| Can someone test on an sn2 system?

Dann,

Yep, works fine on various sn2 configs except for one particular hardware/prom configuation, which I believe is unrelated to this patch (we're still looking at it).

It would probably be worthwhile to incorporate Jesse's comments to the last post of this:

http://marc.theaimsgroup.com/?l=linux-ia64&m1077582813691&w=2

Greg

There was a conflict in discontig.c with Bob Picco's memory-less nodes patch ... which I resolved, it would be helpful if someone looked at it to see that I got it right. Patch is in my GIT tree (or will be when kernel.org mirroring makes it visible).

Generic kernels don't boot on Tiger ... they oops in tiocx_init(). Turning off CONFIG_SGI_TIOCX works around this for now, but this needs to be fixed.

Oh ... apologies to Dann ... in the commit commentary I said that Greg had dug this back up out of obscurity.

Yes, it would be nice to apply some more cleanup ... but I've let this patch get too stale before, so I put it in as-is for now.

-Tony