Re: [PATCH] Add nid sanity on alloc_pages_node

From: Nish Aravamudan
Date: Sat Jul 14 2007 - 13:40:35 EST


On 7/13/07, Joe Jin <joe.jin@xxxxxxxxxx> wrote:
>
> Patch gone too ;) I deleted it. I was hoping that you'd send me the final
> finished product (please).
>

Ha.., the patch against 2.6.22, at your patch have use htlb_alloc_mask, but I
cannot found it at 2.6.22 kernel tree, I think you must use difference kernel
tree :)

I believe this patch will be unnecessary if my "Fix hugetlb pool
allocation with empty nodes" patch gets pulled into -mm.
alloc_fresh_huge_page() now takes a mempolicy which is created by the
two callers, rather than reinventing interleaving itself. I believe
this will avoid the oops you saw. I am still waiting on some test
results (annoying -mm config changes) before I repost them (and they
depend on Christoph's fixes for memoryless nodes).

Thanks,
Nish

--- linux-2.6.22/mm/hugetlb.c.orig 2007-07-12 15:02:19.000000000 +0800
+++ linux-2.6.22/mm/hugetlb.c 2007-07-13 17:33:45.000000000 +0800
@@ -101,13 +101,20 @@ static void free_huge_page(struct page *

static int alloc_fresh_huge_page(void)
{
- static int nid = 0;
+ static int prev_nid;
struct page *page;
- page = alloc_pages_node(nid, GFP_HIGHUSER|__GFP_COMP|__GFP_NOWARN,
- HUGETLB_PAGE_ORDER);
- nid = next_node(nid, node_online_map);
+ static DEFINE_SPINLOCK(nid_lock);
+ int nid;
+
+ spin_lock(&nid_lock);
+ nid = next_node(prev_nid, node_online_map);
if (nid == MAX_NUMNODES)
nid = first_node(node_online_map);
+ prev_nid = nid;
+ spin_unlock(&nid_lock);
+
+ page = alloc_pages_node(nid, GFP_HIGHUSER|__GFP_COMP|__GFP_NOWARN,
+ HUGETLB_PAGE_ORDER);
if (page) {
set_compound_page_dtor(page, free_huge_page);
spin_lock(&hugetlb_lock);
-
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/

-
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/