[PATCH 1/5] UML - Fix ZONE_HIGHMEM compilation error

[PATCH 1/5] UML - Fix ZONE_HIGHMEM compilation error

[Jeff Dike]

References to ZONE_HIGHMEM need to depend on CONFIG_HIGHMEM.

Signed-off-by: Jeff Dike <jdike@addtoit.com>

Index: linux-2.6.17/arch/um/kernel/mem.c
===================================================================
--- linux-2.6.17.orig/arch/um/kernel/mem.c	2006-07-12 11:29:02.000000000 -0400
+++ linux-2.6.17/arch/um/kernel/mem.c	2006-07-12 11:29:11.000000000 -0400
@@ -226,7 +226,9 @@ void paging_init(void)
 	for(i=0;i<sizeof(zones_size)/sizeof(zones_size[0]);i++) 
 		zones_size[i] = 0;
 	zones_size[ZONE_DMA] = (end_iomem >> PAGE_SHIFT) - (uml_physmem >> PAGE_SHIFT);
+#ifdef CONFIG_HIGHMEM
 	zones_size[ZONE_HIGHMEM] = highmem >> PAGE_SHIFT;
+#endif
 	free_area_init(zones_size);
 
 	/*

Re: [PATCH 1/5] UML - Fix ZONE_HIGHMEM compilation error

[Andrew Morton]

On Wed, 12 Jul 2006 12:39:43 -0400
Jeff Dike <jdike@addtoit.com> wrote:

> References to ZONE_HIGHMEM need to depend on CONFIG_HIGHMEM.
> 

There are several such references in mm/page_alloc.c

> 
> Index: linux-2.6.17/arch/um/kernel/mem.c
> ===================================================================
> --- linux-2.6.17.orig/arch/um/kernel/mem.c	2006-07-12 11:29:02.000000000 -0400
> +++ linux-2.6.17/arch/um/kernel/mem.c	2006-07-12 11:29:11.000000000 -0400
> @@ -226,7 +226,9 @@ void paging_init(void)
>  	for(i=0;i<sizeof(zones_size)/sizeof(zones_size[0]);i++) 

I spy an ARRAY_SIZE().

>  		zones_size[i] = 0;
>  	zones_size[ZONE_DMA] = (end_iomem >> PAGE_SHIFT) - (uml_physmem >> PAGE_SHIFT);
> +#ifdef CONFIG_HIGHMEM
>  	zones_size[ZONE_HIGHMEM] = highmem >> PAGE_SHIFT;
> +#endif
>  	free_area_init(zones_size);
>  

Maybe this is an rc1-mm1 fix?  Did Christoph's patches break UML, perhaps??

Re: [PATCH 1/5] UML - Fix ZONE_HIGHMEM compilation error

[Jeff Dike]

On Thu, Jul 13, 2006 at 01:24:21AM -0700, Andrew Morton wrote:
> >  	for(i=0;i<sizeof(zones_size)/sizeof(zones_size[0]);i++) 
> 
> I spy an ARRAY_SIZE().

Yup, I did an ARRAY_SIZE pass a while ago, but I missed that somehow,
and some grepping shows there are a bunch more.

> Maybe this is an rc1-mm1 fix?  Did Christoph's patches break UML, perhaps??

Yes, it's this bit in mmzone.h:
    #ifdef CONFIG_HIGHMEM
	/*
	 * A memory area that is only addressable by the kernel through
	 * mapping portions into its own address space. This is for example
	 * used by i386 to allow the kernel to address the memory beyond
	 * 900MB. The kernel will set up special mappings (page
	 * table entries on i386) for each page that the kernel needs to
	 * access.
	 */
	ZONE_HIGHMEM,
    #endif

				Jeff