[PATCH] fix Summit srat.h includes

[PATCH] fix Summit srat.h includes

[Dave Hansen]

[-- Attachment #1: Type: text/plain, Size: 823 bytes --]

I was compiling for my plain 'ol PC, and was getting unresolved symbols
for get_memcfg_from_srat() and get_zholes_size().  The CONFIG_NUMA
definition right now allows it to be turned on for plain old X86_PC. 
Does anyone know why this is?  

depends on SMP && HIGHMEM64G && 
          (X86_PC || X86_NUMAQ || X86_GENERICARCH || 
                  (X86_SUMMIT && ACPI && !ACPI_HT_ONLY))

In any case, the summit code incorrectly assumes in at least 2 places
that NUMA && !NUMAQ means summit.  Someone was evidently trying to cover
the generic subarch case, but that's already taken care of by the lovely
config system and CONFIG_ACPI_SRAT.  This patch fixes those assumptions
and adds a nice little warning for people that try to #include srat.h
without having srat support turned on.  

-- 
Dave Hansen
haveblue@us.ibm.com

[-- Attachment #2: srat-include-2.6.0-test4-0.patch --]
[-- Type: text/x-patch, Size: 1508 bytes --]

diff -ru linux-2.6.0-test4/include/asm-i386/mmzone.h linux-2.6.0-test4-srat/include/asm-i386/mmzone.h
--- linux-2.6.0-test4/include/asm-i386/mmzone.h	2003-08-22 16:56:14.000000000 -0700
+++ linux-2.6.0-test4-srat/include/asm-i386/mmzone.h	2003-09-15 08:22:32.000000000 -0700
@@ -119,7 +119,7 @@
 
 #ifdef CONFIG_X86_NUMAQ
 #include <asm/numaq.h>
-#elif CONFIG_NUMA	/* summit or generic arch */
+#elif CONFIG_ACPI_SRAT
 #include <asm/srat.h>
 #elif CONFIG_X86_PC
 #define get_memcfg_numa get_memcfg_numa_flat
diff -ru linux-2.6.0-test4/include/asm-i386/numnodes.h linux-2.6.0-test4-srat/include/asm-i386/numnodes.h
--- linux-2.6.0-test4/include/asm-i386/numnodes.h	2003-08-22 16:52:57.000000000 -0700
+++ linux-2.6.0-test4-srat/include/asm-i386/numnodes.h	2003-09-15 08:22:50.000000000 -0700
@@ -5,7 +5,7 @@
 
 #ifdef CONFIG_X86_NUMAQ
 #include <asm/numaq.h>
-#elif CONFIG_NUMA
+#elif CONFIG_ACPI_SRAT
 #include <asm/srat.h>
 #else
 #define MAX_NUMNODES	1
diff -ru linux-2.6.0-test4/include/asm-i386/srat.h linux-2.6.0-test4-srat/include/asm-i386/srat.h
--- linux-2.6.0-test4/include/asm-i386/srat.h	2003-08-22 16:53:43.000000000 -0700
+++ linux-2.6.0-test4-srat/include/asm-i386/srat.h	2003-09-15 08:20:49.000000000 -0700
@@ -27,6 +27,10 @@
 #ifndef _ASM_SRAT_H_
 #define _ASM_SRAT_H_
 
+#ifndef CONFIG_ACPI_SRAT
+#error CONFIG_ACPI_SRAT not defined, and srat.h header has been included
+#endif
+
 #define MAX_NUMNODES		8
 extern void get_memcfg_from_srat(void);
 extern unsigned long *get_zholes_size(int);

Re: [PATCH] fix Summit srat.h includes

[Martin J. Bligh]

> I was compiling for my plain 'ol PC, and was getting unresolved symbols
> for get_memcfg_from_srat() and get_zholes_size().  The CONFIG_NUMA
> definition right now allows it to be turned on for plain old X86_PC. 
> Does anyone know why this is?  

Yeah, distros want it, so they can ship one kernel for SMP that covers
NUMA and non-NUMA, switching dynamically.
 
> depends on SMP && HIGHMEM64G && 
>           (X86_PC || X86_NUMAQ || X86_GENERICARCH || 
>                   (X86_SUMMIT && ACPI && !ACPI_HT_ONLY))
> 
> In any case, the summit code incorrectly assumes in at least 2 places
> that NUMA && !NUMAQ means summit.  Someone was evidently trying to cover
> the generic subarch case, but that's already taken care of by the lovely
> config system and CONFIG_ACPI_SRAT.  This patch fixes those assumptions
> and adds a nice little warning for people that try to #include srat.h
> without having srat support turned on.  

Nice, thanks. Yeah, the generic NUMA on generic PC code will need a bit
of tidy up, but it should more or less work ;-)

Thanks,

M.