Re: [PATCH] Fix DEBUG_PAGEALLOC on 603/e300

[Xianghua Xiao <xiaoxianghua@gmail.com>]

On Wed, Apr 28, 2010 at 2:22 AM, Benjamin Herrenschmidt
<benh@kernel.crashing.org> wrote:
> So we tried to speed things up a bit using flush_hash_pages() directly
> but that falls over on 603 of course meaning we fail to flush the TLB
> properly and we may even end up having it corrupt memory randomly by
> accessing a hash table that doesn't exist.
>
> This removes the "optimization" by always going through flush_tlb_page()
> for now at least.
>
> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> ---
>
> Somebody with a 603 or e300 core based FSL SoC to try this out for me ?
>
> It's obviously completely untested :-)
>
> Cheers,
> Ben.
>
> diff --git a/arch/powerpc/mm/pgtable_32.c b/arch/powerpc/mm/pgtable_32.c
> index b9243e7..95774b4 100644
> --- a/arch/powerpc/mm/pgtable_32.c
> +++ b/arch/powerpc/mm/pgtable_32.c
> @@ -385,11 +385,7 @@ static int __change_page_attr(struct page *page, pgp=
rot_t prot)
> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0return -EINVAL;
> =C2=A0 =C2=A0 =C2=A0 =C2=A0__set_pte_at(&init_mm, address, kpte, mk_pte(p=
age, prot), 0);
> =C2=A0 =C2=A0 =C2=A0 =C2=A0wmb();
> -#ifdef CONFIG_PPC_STD_MMU
> - =C2=A0 =C2=A0 =C2=A0 flush_hash_pages(0, address, pmd_val(*kpmd), 1);
> -#else
> =C2=A0 =C2=A0 =C2=A0 =C2=A0flush_tlb_page(NULL, address);
> -#endif
> =C2=A0 =C2=A0 =C2=A0 =C2=A0pte_unmap(kpte);
>
> =C2=A0 =C2=A0 =C2=A0 =C2=A0return 0;
>
>
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev
>
This change works me on a 834x(e300) platform, tested with lmbench and
a production-ready application with 2.6.33.3.

Xianghua