>I am releasing now a new patch. It will have the fixes for the bug
>triggered in 2.2.6_andrea3 and will have the fixes for the
>performances-bugs I found previously in 2.2.6_andrea2. It will return to
I just have a patch against 2.2.6_andrea4.bz2.
Index: mm/filemap.c
===================================================================
RCS file: /var/cvs/linux/mm/filemap.c,v
retrieving revision 1.1.2.57
diff -u -r1.1.2.57 filemap.c
--- linux/mm/filemap.c 1999/04/27 18:52:52 1.1.2.57
+++ linux/mm/filemap.c 1999/04/28 12:53:01
@@ -172,7 +172,7 @@
LIST_HEAD(old);
LIST_HEAD(young);
- while (count-- > 0 && (tmp = lru_head[i]->prev) != lru_head[i])
+ while (count > 0 && (tmp = lru_head[i]->prev) != lru_head[i])
{
struct page * page = list_entry(tmp, struct page, lru);
@@ -183,6 +183,7 @@
list_add(tmp, &young);
continue;
}
+ count--;
if ((gfp_mask & __GFP_DMA) && !PageDMA(page))
{
It's not really a bugfix but I think it's better. The only meaning of the
the referenced bit now has to be to reorder the lru list, not at all to
drive swapout or not. I did such change because it looked better in low
memory to swapout more, but it was not the right thing to do.
If somebody will try out 2.2.6_andrea4.bz2 with and then w/o the patch
above let me know (let me know the details of your memory too ;).
Thanks.
Andrea Arcangeli
-
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/