netvm: propagate page->pfmemalloc to skb
authorMel Gorman <mgorman@suse.de>
Tue, 31 Jul 2012 23:44:23 +0000 (16:44 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 1 Aug 2012 01:42:46 +0000 (18:42 -0700)
The skb->pfmemalloc flag gets set to true iff during the slab allocation
of data in __alloc_skb that the the PFMEMALLOC reserves were used.  If the
packet is fragmented, it is possible that pages will be allocated from the
PFMEMALLOC reserve without propagating this information to the skb.  This
patch propagates page->pfmemalloc from pages allocated for fragments to
the skb.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
include/linux/skbuff.h

index 0336f02..b814bb8 100644 (file)
@@ -1247,6 +1247,17 @@ static inline void __skb_fill_page_desc(struct sk_buff *skb, int i,
 {
        skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
 
+       /*
+        * Propagate page->pfmemalloc to the skb if we can. The problem is
+        * that not all callers have unique ownership of the page. If
+        * pfmemalloc is set, we check the mapping as a mapping implies
+        * page->index is set (index and pfmemalloc share space).
+        * If it's a valid mapping, we cannot use page->pfmemalloc but we
+        * do not lose pfmemalloc information as the pages would not be
+        * allocated using __GFP_MEMALLOC.
+        */
+       if (page->pfmemalloc && !page->mapping)
+               skb->pfmemalloc = true;
        frag->page.p              = page;
        frag->page_offset         = off;
        skb_frag_size_set(frag, size);