Home » Mailing list archives » linux-kernel » 2007 » September » 10

[PATCH 7/13] Drain per-cpu lists when high-order allocations fail

!MAILaRCHIVE_VOTE_RePLACE
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
[view in full thread]
From: Mel Gorman <mel@...>
To: <akpm@...>
Cc: Mel Gorman <mel@...>, <linux-kernel@...>, <linux-mm@...>
Subject: [PATCH 7/13] Drain per-cpu lists when high-order allocations fail
Date: Monday, September 10, 2007 - 7:22 am

Subject: Drain per-cpu lists when high-order allocations fail

Per-cpu pages can accidentally cause fragmentation because they are free, but
pinned pages in an otherwise contiguous block.  When this patch is applied,
the per-cpu caches are drained after the direct-reclaim is entered if the
requested order is greater than 0.  It simply reuses the code used by suspend
and hotplug.

Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 mm/page_alloc.c |   24 +++++++++++++++++++++++-
 1 file changed, 23 insertions(+), 1 deletion(-)

diff -rup -X /usr/src/patchset-0.6/bin//dontdiff linux-2.6.23-rc5-006-group-short-lived-and-reclaimable-kernel-allocations/mm/page_alloc.c linux-2.6.23-rc5-007-drain-per-cpu-lists-when-high-order-allocations-fail/mm/page_alloc.c
--- linux-2.6.23-rc5-006-group-short-lived-and-reclaimable-kernel-allocations/mm/page_alloc.c	2007-09-02 16:20:31.000000000 +0100
+++ linux-2.6.23-rc5-007-drain-per-cpu-lists-when-high-order-allocations-fail/mm/page_alloc.c	2007-09-02 16:20:48.000000000 +0100
@@ -852,6 +852,7 @@ void mark_free_pages(struct zone *zone)
 	}
 	spin_unlock_irqrestore(&zone->lock, flags);
 }
+#endif /* CONFIG_PM */
 
 /*
  * Spill all of this CPU's per-cpu pages back into the buddy allocator.
@@ -864,7 +865,25 @@ void drain_local_pages(void)
 	__drain_pages(smp_processor_id());
 	local_irq_restore(flags);	
 }
-#endif /* CONFIG_HIBERNATION */
+
+void smp_drain_local_pages(void *arg)
+{
+	drain_local_pages();
+}
+
+/*
+ * Spill all the per-cpu pages from all CPUs back into the buddy allocator
+ */
+void drain_all_local_pages(void)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+	__drain_pages(smp_processor_id());
+	local_irq_restore(flags);
+
+	smp_call_function(smp_drain_local_pages, NULL, 0, 1);
+}
 
 /*
  * Free a 0-order page
@@ -1452,6 +1471,9 @@ nofail_alloc:
 
 	cond_resched();
 
+	if (order != 0)
+		drain_all_local_pages();
+
 	if (likely(did_some_progress)) {
 		page = get_page_from_freelist(gfp_mask, order,
 						zonelist, alloc_flags);
-
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
Messages in current thread:
[PATCH 0/13] Reduce external fragmentation by grouping pages..., Mel Gorman, (Mon Sep 10, 7:20 am)
Re: [PATCH 0/13] Reduce external fragmentation by grouping p..., Andrew Morton, (Thu Sep 13, 9:01 pm)
Re: [PATCH 0/13] Reduce external fragmentation by grouping p..., Mel Gorman, (Fri Sep 14, 10:33 am)
Re: [PATCH 0/13] Reduce external fragmentation by grouping p..., Andrew Morton, (Sun Sep 16, 6:34 am)
[PATCH 13/13] Print out statistics in relation to fragmentat..., Mel Gorman, (Mon Sep 10, 7:24 am)
[PATCH 12/13] Be more agressive about stealing when MIGRATE_..., Mel Gorman, (Mon Sep 10, 7:24 am)
[PATCH 11/13] Bias the placement of kernel pages at lower pfns, Mel Gorman, (Mon Sep 10, 7:23 am)
[PATCH 10/13] Bias the location of pages freed for min_free_..., Mel Gorman, (Mon Sep 10, 7:23 am)
[PATCH 9/13] Do not group pages by mobility type on low memo..., Mel Gorman, (Mon Sep 10, 7:23 am)
[PATCH 8/13] Move free pages between lists on steal, Mel Gorman, (Mon Sep 10, 7:22 am)
[PATCH 7/13] Drain per-cpu lists when high-order allocations..., Mel Gorman, (Mon Sep 10, 7:22 am)
Re: [PATCH 7/13] Drain per-cpu lists when high-order allocat..., Nick Piggin, (Mon Sep 10, 11:05 am)
Re: [PATCH 7/13] Drain per-cpu lists when high-order allocat..., Mel Gorman, (Tue Sep 11, 5:34 am)
[PATCH 6/13] Group short-lived and reclaimable kernel alloca..., Mel Gorman, (Mon Sep 10, 7:22 am)
Re: Group short-lived and reclaimable kernel allocations, Paul Jackson, (Mon Sep 10, 3:44 pm)
Re: Group short-lived and reclaimable kernel allocations, Mel Gorman, (Mon Sep 10, 5:15 pm)
[PATCH 5/13] Choose pages from the per cpu list-based on mig..., Mel Gorman, (Mon Sep 10, 7:21 am)
[PATCH 4/13] Split the free lists for movable and unmovable ..., Mel Gorman, (Mon Sep 10, 7:21 am)
[PATCH 3/13] Fix corruption of memmap on ia64-sparsemem when..., Mel Gorman, (Mon Sep 10, 7:21 am)
[PATCH 2/13] Add a bitmap that is used to track flags affect..., Mel Gorman, (Mon Sep 10, 7:20 am)
[PATCH 1/13] ia64: parse kernel parameter hugepagesz= in ear..., Mel Gorman, (Mon Sep 10, 7:20 am)