net: use __GFP_NORETRY for high order allocations

Eric Dumazet · davem330 · commit ed98df3361f0 · 2014-02-06T21:28:06.000-08:00
sock_alloc_send_pskb() &amp; sk_page_frag_refill()
have a loop trying high order allocations to prepare
skb with low number of fragments as this increases performance.

Problem is that under memory pressure/fragmentation, this can
trigger OOM while the intent was only to try the high order
allocations, then fallback to order-0 allocations.

We had various reports from unexpected regressions.

According to David, setting __GFP_NORETRY should be fine,
as the asynchronous compaction is still enabled, and this
will prevent OOM from kicking as in :

CFSClientEventm invoked oom-killer: gfp_mask=0x42d0, order=3, oom_adj=0,
oom_score_adj=0, oom_score_badness=2 (enabled),memcg_scoring=disabled
CFSClientEventm

Call Trace:
 [&lt;ffffffff8043766c&gt;] dump_header+0xe1/0x23e
 [&lt;ffffffff80437a02&gt;] oom_kill_process+0x6a/0x323
 [&lt;ffffffff80438443&gt;] out_of_memory+0x4b3/0x50d
 [&lt;ffffffff8043a4a6&gt;] __alloc_pages_may_oom+0xa2/0xc7
 [&lt;ffffffff80236f42&gt;] __alloc_pages_nodemask+0x1002/0x17f0
 [&lt;ffffffff8024bd23&gt;] alloc_pages_current+0x103/0x2b0
 [&lt;ffffffff8028567f&gt;] sk_page_frag_refill+0x8f/0x160
 [&lt;ffffffff80295fa0&gt;] tcp_sendmsg+0x560/0xee0
 [&lt;ffffffff802a5037&gt;] inet_sendmsg+0x67/0x100
 [&lt;ffffffff80283c9c&gt;] __sock_sendmsg_nosec+0x6c/0x90
 [&lt;ffffffff80283e85&gt;] sock_sendmsg+0xc5/0xf0
 [&lt;ffffffff802847b6&gt;] __sys_sendmsg+0x136/0x430
 [&lt;ffffffff80284ec8&gt;] sys_sendmsg+0x88/0x110
 [&lt;ffffffff80711472&gt;] system_call_fastpath+0x16/0x1b
Out of Memory: Kill process 2856 (bash) score 9999 or sacrifice child

Signed-off-by: Eric Dumazet &lt;edumazet@google.com&gt;
Acked-by: David Rientjes &lt;rientjes@google.com&gt;
Acked-by: "Eric W. Biederman" &lt;ebiederm@xmission.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
diff --git a/net/core/sock.c b/net/core/sock.c
@@ -1775,7 +1775,9 @@ struct sk_buff *sock_alloc_send_pskb(struct sock *sk, unsigned long header_len,
 			while (order) {
 				if (npages >= 1 << order) {
 					page = alloc_pages(sk->sk_allocation |
-							   __GFP_COMP | __GFP_NOWARN,
+							   __GFP_COMP |
+							   __GFP_NOWARN |
+							   __GFP_NORETRY,
 							   order);
 					if (page)
 						goto fill_page;
@@ -1845,7 +1847,7 @@ bool skb_page_frag_refill(unsigned int sz, struct page_frag *pfrag, gfp_t prio)
 		gfp_t gfp = prio;
 
 		if (order)
-			gfp |= __GFP_COMP | __GFP_NOWARN;
+			gfp |= __GFP_COMP | __GFP_NOWARN | __GFP_NORETRY;
 		pfrag->page = alloc_pages(gfp, order);
 		if (likely(pfrag->page)) {
 			pfrag->offset = 0;
