Re: page migration: Only migrate pages if allocation in the highest zone is possible

[Hugh Dickins <hugh@veritas.com>]

On Sun, 29 Apr 2007, Christoph Lameter wrote:

> Address spaces contain an allocation flag that specifies restriction on
> the zone for pages placed in the mapping. I.e. some device may require pages
> to be allocated from a DMA zone. Block devices may not be able to use pages
> from HIGHMEM.
> 
> Memory policies and the common use of page migration works only on the
> highest zone. If the address space does not allow allocation from the
> highest zone then the pages in the address space are not migratable simply
> because we can only allocate memory for a specified node if we allow
> allocation for the highest zone on each node.
> 
> Cc: Hugh Dickins <hugh@veritas.com>
> Signed-off-by: Christoph Lameter <clameter@sgi.com>

That looks about right to me: from the previous discussion it appeared
that a test in vma_migratable() is currently more appropriate than
refining the page allocation routines (and yours is nicer than mine).
Though that will change when page migration is used beyond CONFIG_NUMA.

There is still another argument against doing it in vma_migratable():
this way a private mapping of a blockdev filled with anon COWed
pages, perfectly migratable in itself, is ruled out as if all
its pages were shared.  But IIRC all the mempolicy stuff is really
geared towards shared mappings anyway, and I expect blockdev
mappings are usually shared too: I doubt it's a serious issue.

I would feel more comfortable if you #included linux/pagemap.h
and used the mapping_gfp_mask macro, just for documentation of
what's going on - that's how that field is accessed elsewhere;
though I agree that it looks like gfp_zone() will happen to
work on the raw flags field.

Conversely (why do I suggest safety on one line and unsafety
on another? incorribly inconsistent or what?) I contend that
testing vma->vm_file->f_mapping there is unnecessary, and I'd
rather we don't propagate unnecessary such tests (each one
sows another seed of doubt for the next): sys_fsync looks
like one syscall which would crash if f_mapping ever NULL.

Hugh

> 
> ---
>  include/linux/migrate.h |    9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> Index: linux-2.6.21-rc7-mm2/include/linux/migrate.h
> ===================================================================
> --- linux-2.6.21-rc7-mm2.orig/include/linux/migrate.h	2007-04-15 16:50:57.000000000 -0700
> +++ linux-2.6.21-rc7-mm2/include/linux/migrate.h	2007-04-27 23:14:49.000000000 -0700
> @@ -2,6 +2,7 @@
>  #define _LINUX_MIGRATE_H
>  
>  #include <linux/mm.h>
> +#include <linux/mempolicy.h>
>  
>  typedef struct page *new_page_t(struct page *, unsigned long private, int **);
>  
> @@ -10,6 +11,14 @@ static inline int vma_migratable(struct 
>  {
>  	if (vma->vm_flags & (VM_IO|VM_HUGETLB|VM_PFNMAP|VM_RESERVED))
>  		return 0;
> +	/*
> +	 * Migration allocates pages in the highest zone. If we cannot
> +	 * do so then migration (at least from node to node) is not
> +	 * possible.
> +	 */
> +	if (vma->vm_file && vma->vm_file->f_mapping &&
> +		gfp_zone(vma->vm_file->f_mapping->flags) < policy_zone)
> +			return 0;
>  	return 1;
>  }
>  

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>