Patchwork powerpc: is_hugepage_only_range() must account for both 4kB and 64kB slices

login
register
mail settings
Submitter Dave Kleikamp
Date Jan. 14, 2009, 7:09 p.m.
Message ID <1231960174.31244.8.camel@norville.austin.ibm.com>
Download mbox | patch
Permalink /patch/18519/
State Accepted, archived
Commit 9ba0fdbfaed2e74005d87fab948c5522b86ff733
Delegated to: Benjamin Herrenschmidt
Headers show

Comments

Dave Kleikamp - Jan. 14, 2009, 7:09 p.m.
powerpc: is_hugepage_only_range() must account for both 4kB and 64kB slices

The subpage_prot syscall fails on second and subsequent calls for a given
region, because is_hugepage_only_range() is mis-identifying the 4 kB
slices when the process has a 64 kB page size.

Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt - Jan. 15, 2009, 1:54 a.m.
On Wed, 2009-01-14 at 13:09 -0600, Dave Kleikamp wrote:
> powerpc: is_hugepage_only_range() must account for both 4kB and 64kB slices
> 
> The subpage_prot syscall fails on second and subsequent calls for a given
> region, because is_hugepage_only_range() is mis-identifying the 4 kB
> slices when the process has a 64 kB page size.
> 
> Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Looks good. I'll put that in for .29

Cheers,
Ben.

> diff --git a/arch/powerpc/mm/slice.c b/arch/powerpc/mm/slice.c
> index db44e02..ba51948 100644
> --- a/arch/powerpc/mm/slice.c
> +++ b/arch/powerpc/mm/slice.c
> @@ -710,9 +710,18 @@ int is_hugepage_only_range(struct mm_struct *mm, unsigned long addr,
>  			   unsigned long len)
>  {
>  	struct slice_mask mask, available;
> +	unsigned int psize = mm->context.user_psize;
>  
>  	mask = slice_range_to_mask(addr, len);
> -	available = slice_mask_for_size(mm, mm->context.user_psize);
> +	available = slice_mask_for_size(mm, psize);
> +#ifdef CONFIG_PPC_64K_PAGES
> +	/* We need to account for 4k slices too */
> +	if (psize == MMU_PAGE_64K) {
> +		struct slice_mask compat_mask;
> +		compat_mask = slice_mask_for_size(mm, MMU_PAGE_4K);
> +		or_mask(available, compat_mask);
> +	}
> +#endif
>  
>  #if 0 /* too verbose */
>  	slice_dbg("is_hugepage_only_range(mm=%p, addr=%lx, len=%lx)\n",
>

Patch

diff --git a/arch/powerpc/mm/slice.c b/arch/powerpc/mm/slice.c
index db44e02..ba51948 100644
--- a/arch/powerpc/mm/slice.c
+++ b/arch/powerpc/mm/slice.c
@@ -710,9 +710,18 @@  int is_hugepage_only_range(struct mm_struct *mm, unsigned long addr,
 			   unsigned long len)
 {
 	struct slice_mask mask, available;
+	unsigned int psize = mm->context.user_psize;
 
 	mask = slice_range_to_mask(addr, len);
-	available = slice_mask_for_size(mm, mm->context.user_psize);
+	available = slice_mask_for_size(mm, psize);
+#ifdef CONFIG_PPC_64K_PAGES
+	/* We need to account for 4k slices too */
+	if (psize == MMU_PAGE_64K) {
+		struct slice_mask compat_mask;
+		compat_mask = slice_mask_for_size(mm, MMU_PAGE_4K);
+		or_mask(available, compat_mask);
+	}
+#endif
 
 #if 0 /* too verbose */
 	slice_dbg("is_hugepage_only_range(mm=%p, addr=%lx, len=%lx)\n",