Re: [PATCH] mm: do not call bad_page on PG_reserved check

[Hugh Dickins <hugh@veritas.com>]

On Sat, 18 Nov 2006, David Rientjes wrote:

> The return value of free_pages_check() indicates if PG_reserved was set.
> If so, the calling functions return immediately and no pages are freed so
> there is no need to call bad_page().
> 
> Cc: Andi Kleen <ak@suse.de>
> Cc: Nick Piggin <npiggin@suse.de>
> Signed-off-by: David Rientjes <rientjes@cs.washington.edu>

NAK.  You're missing the point.  If an attempt is made to free a
reserved page, it implies that the page reference counting has
gone wrong: we want to hear about that (so call bad_page),
and we dare not reuse the page (so skip freeing it).

What might be a good change, is to avoid freeing a page which meets
_any_ of the criteria for calling bad_page: I often wonder whether
to do that, alongside abandoning that hopeless page_mapcount BUG in
page_remove_rmap, which has almost(?) never helped lead us to any fix.

Hugh

> ---
>  mm/page_alloc.c |    1 -
>  1 files changed, 0 insertions(+), 1 deletions(-)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index bf2f6cf..99bc29d 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -439,7 +439,6 @@ static inline int free_pages_check(struc
>  			1 << PG_slab	|
>  			1 << PG_swapcache |
>  			1 << PG_writeback |
> -			1 << PG_reserved |
>  			1 << PG_buddy ))))
>  		bad_page(page);
>  	if (PageDirty(page))
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/