[RFC] buddy allocator without bitmap [4/4]

[RFC] buddy allocator without bitmap [4/4]

[Hiroyuki KAMEZAWA]

This patch 5th inserts prefetch().
I think These prefetch are reasonable and helpful.

-- Kame
====================================================================



---

  linux-2.6.8.1-mm4-kame-kamezawa/mm/page_alloc.c |    2 ++
  1 files changed, 2 insertions(+)

diff -puN mm/page_alloc.c~eliminate-bitmap-prefetch mm/page_alloc.c
--- linux-2.6.8.1-mm4-kame/mm/page_alloc.c~eliminate-bitmap-prefetch	2004-08-26 19:32:01.598461736 +0900
+++ linux-2.6.8.1-mm4-kame-kamezawa/mm/page_alloc.c	2004-08-26 19:32:01.602461128 +0900
@@ -257,6 +257,7 @@ static inline void __free_pages_bulk (st
  		order++;
  		mask <<= 1;
  		page_idx &= mask;
+		prefetch(base + (page_idx ^ (1 << order)));
  		list_del(&buddy->lru);
  		/* for propriety of PG_private bit, we clear it */
  		buddy->flags &= ~(1 << PG_private);
@@ -360,6 +361,7 @@ expand(struct zone *zone, struct page *p
  		area--;
  		high--;
  		size >>= 1;
+		prefetch(&page[size >> 1]);
  		BUG_ON(bad_range(zone, &page[size]));
  		list_add(&page[size].lru, &area->free_list);
  		page[size].flags |= (1 << PG_private);

_

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>

Re: [RFC] buddy allocator without bitmap [4/4]

[Dave Hansen]

On Thu, 2004-08-26 at 05:15, Hiroyuki KAMEZAWA wrote:
> This patch 5th inserts prefetch().
> I think These prefetch are reasonable and helpful.

Do you have any benchmark numbers to show it?

-- Dave

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>