Re: [PATCH] stop page_state stack waste

[Andrew Morton]

Hugh Dickins <hugh@xxxxxxxxxxx> wrote:
>
> Replace get_page_state (which memset most of full page_state to 0) by
> get_main_page_state, which just sets the small structure needed.  This
> helps 4k stacks not to overflow: cuts 224 bytes off try_to_free_pages
> and wakeup_bdflush (and sync_inodes_sb) stack usages: wakeup_bdflush
> doesn't do much, but is called by try_to_free_pages and mempool_alloc.

Yeah, I was looking at that.  I simply did:



The ____cacheline_aligned in there is a leftover from before the existence of
the percpu infrastructure.  It bloats struct page_state and structures which
contain it enormously, and we use these things on the stack deep in page
reclaim.

Signed-off-by: Andrew Morton <akpm@xxxxxxxx>
---

 25-akpm/include/linux/page-flags.h |    2 +-
 1 files changed, 1 insertion(+), 1 deletion(-)

diff -puN include/linux/page-flags.h~unalign-page_state include/linux/page-flags.h
--- 25/include/linux/page-flags.h~unalign-page_state	2004-06-08 03:08:02.766976760 -0700
+++ 25-akpm/include/linux/page-flags.h	2004-06-08 03:08:02.769976304 -0700
@@ -133,7 +133,7 @@ struct page_state {
 	unsigned long allocstall;	/* direct reclaim calls */
 
 	unsigned long pgrotated;	/* pages rotated to tail of the LRU */
-} ____cacheline_aligned;
+};
 
 DECLARE_PER_CPU(struct page_state, page_states);
 
_

-
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/