Re: [PATCH 06/15] mm: remove ppc highpte

[Hugh Dickins]

On Thu, 10 Nov 2005, Paul Mackerras wrote:
> Hugh Dickins writes:
> 
> > ppc's HIGHPTE config option was removed in 2.5.28, and nobody seems to
> > have wanted it enough to restore it: so remove its traces from pgtable.h
> > and pte_alloc_one.  Or supply an alternative patch to config it back?
> 
> I'm staggered.  We do want to be able to have pte pages in highmem.
> I would rather just have it always enabled if CONFIG_HIGHMEM=y, rather
> than putting the config option back.  I think that should just involve
> adding __GFP_HIGHMEM to the flags for alloc_pages in pte_alloc_one
> unconditionally, no?

I'm glad the patch has struck a spark!  Yes, there doesn't appear to
be anything else that you'd need to do.  Except, why was it disabled
in the first place?  I think you'll need to reassure yourselves that
whatever the reason was has gone away, before reenabling.

Andrew, please drop 06/15, I don't think anything after depends on it
(but it came about when I was looking at uninlining pte_offset_map_lock).

Hugh
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/