Re: memory handling in pre5/pre6

From: Rik van Riel (riel@conectiva.com.br)
Date: Mon Apr 17 2000 - 17:23:30 EST

On 17 Apr 2000, Ulrich Drepper wrote:
> Rik van Riel <riel@conectiva.com.br> writes:
>
> > Question is, have any of you actually noticed a performance
> > hit because of this or are you just seeing different numbers
> > than what you're used to and automatically assuming that it
> > must be wrong?
>
> The machine is swapping when I go from one window to the others (from
> xemacs to netscape, mozilla, etc). And yes, this makes the system slower.

OK, I've changed mm/filemap.c a bit to:
- be closer to 2.2.15-pre<new> page eviction behaviour in
  shrink_mmap(), by preserving the accessed bit on still
  mapped pages
- remove a possible race condition in truncate_inode_pages(),
  I'm not sure if this fix will work, but it should ... ;)

I'd be interested in how much this patch helps interactive
performance, if at all... (I really need to setup X on my
test box;))

regards,

Rik 

--
The Internet is not a network of computers. It is a network
of people. That is its real strength.

Wanna talk about the kernel?  irc.openprojects.net / #kernelnewbies
http://www.conectiva.com/		http://www.surriel.com/

--- linux-2.3.99-pre6-3/mm/filemap.c.orig	Mon Apr 17 12:21:46 2000
+++ linux-2.3.99-pre6-3/mm/filemap.c	Mon Apr 17 19:20:54 2000
@@ -149,11 +149,13 @@
 
 		/* page wholly truncated - free it */
 		if (offset >= start) {
+			if (TryLockPage(page)) {
+				spin_unlock(&pagecache_lock);
+				goto repeat;
+			}
 			get_page(page);
 			spin_unlock(&pagecache_lock);
 
-			lock_page(page);
-
 			if (!page->buffers || block_flushpage(page, 0))
 				lru_cache_del(page);
 
@@ -191,11 +193,13 @@
 			continue;
 
 		/* partial truncate, clear end of page */
+		if (TryLockPage(page)) {
+			spin_unlock(&pagecache_lock);
+			goto repeat;
+		}
 		get_page(page);
 		spin_unlock(&pagecache_lock);
 
-		lock_page(page);
-
 		memclear_highpage_flush(page, partial, PAGE_CACHE_SIZE-partial);
 		if (page->buffers)
 			block_flushpage(page, partial);
@@ -234,6 +238,16 @@
 		list_del(page_lru);
 
 		dispose = &zone->lru_cache;
+		/* don't account passes over not DMA pages */
+		if (zone && (!memclass(page->zone, zone)))
+			goto dispose_continue;
+
+		count--;
+		dispose = &old;
+		/* the page is in use, we can't free it now */
+		if (!page->buffers && page_count(page) > 1)
+			goto dispose_continue;
+
 		if (test_and_clear_bit(PG_referenced, &page->flags))
 			/* Roll the page at the top of the lru list,
 			 * we could also be more aggressive putting
@@ -242,18 +256,7 @@
 			 */
 			goto dispose_continue;
 
-		dispose = &old;
-		/* don't account passes over not DMA pages */
-		if (zone && (!memclass(page->zone, zone)))
-			goto dispose_continue;
-
-		count--;
-
 		dispose = &young;
-
-		/* avoid unscalable SMP locking */
-		if (!page->buffers && page_count(page) > 1)
-			goto dispose_continue;
 
 		if (TryLockPage(page))
 			goto dispose_continue;

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

This archive was generated by hypermail 2b29 : Sun Apr 23 2000 - 21:00:12 EST