@@ -4708,6 +4708,30 @@ void fs_reclaim_release(gfp_t gfp_mask)
EXPORT_SYMBOL_GPL(fs_reclaim_release);
#endif
+/*
+ * Zonelists may change due to hotplug during allocation. Detect when zonelists
+ * have been rebuilt so allocation retries. Reader side does not lock and
+ * retries the allocation if zonelist changes. Writer side is protected by the
+ * embedded spin_lock.
+ */
+DEFINE_SEQLOCK(zonelist_update_seq);
+
+static unsigned int zonelist_iter_begin(void)
+{
+ if (IS_ENABLED(CONFIG_MEMORY_HOTREMOVE))
+ return 0;
+
+ return read_seqbegin(&zonelist_update_seq);
+}
+
+static unsigned int check_retry_zonelist(unsigned int seq)
+{
+ if (IS_ENABLED(CONFIG_MEMORY_HOTREMOVE))
+ return seq;
+
+ return read_seqretry(&zonelist_update_seq, seq);
+}
+
/* Perform direct synchronous page reclaim */
static unsigned long
__perform_reclaim(gfp_t gfp_mask, unsigned int order,
@@ -5001,6 +5025,7 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
int compaction_retries;
int no_progress_loops;
unsigned int cpuset_mems_cookie;
+ unsigned int zonelist_iter_cookie;
int reserve_flags;
/*
@@ -5011,11 +5036,12 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
(__GFP_ATOMIC|__GFP_DIRECT_RECLAIM)))
gfp_mask &= ~__GFP_ATOMIC;
-retry_cpuset:
+restart:
compaction_retries = 0;
no_progress_loops = 0;
compact_priority = DEF_COMPACT_PRIORITY;
cpuset_mems_cookie = read_mems_allowed_begin();
+ zonelist_iter_cookie = zonelist_iter_begin();
/*
* The fast path uses conservative alloc_flags to succeed only until
@@ -5187,9 +5213,13 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
goto retry;
- /* Deal with possible cpuset update races before we start OOM killing */
- if (check_retry_cpuset(cpuset_mems_cookie, ac))
- goto retry_cpuset;
+ /*
+ * Deal with possible cpuset update races or zonelist updates to avoid
+ * a unnecessary OOM kill.
+ */
+ if (check_retry_cpuset(cpuset_mems_cookie, ac) ||
+ check_retry_zonelist(zonelist_iter_cookie))
+ goto restart;
/* Reclaim has failed us, start killing things */
page = __alloc_pages_may_oom(gfp_mask, order, ac, &did_some_progress);
@@ -5209,9 +5239,13 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order,
}
nopage:
- /* Deal with possible cpuset update races before we fail */
- if (check_retry_cpuset(cpuset_mems_cookie, ac))
- goto retry_cpuset;
+ /*
+ * Deal with possible cpuset update races or zonelist updates to avoid
+ * a unnecessary OOM kill.
+ */
+ if (check_retry_cpuset(cpuset_mems_cookie, ac) ||
+ check_retry_zonelist(zonelist_iter_cookie))
+ goto restart;
/*
* Make sure that __GFP_NOFAIL request doesn't leak out and make sure
@@ -6514,9 +6548,8 @@ static void __build_all_zonelists(void *data)
int nid;
int __maybe_unused cpu;
pg_data_t *self = data;
- static DEFINE_SPINLOCK(lock);
- spin_lock(&lock);
+ write_seqlock(&zonelist_update_seq);
#ifdef CONFIG_NUMA
memset(node_load, 0, sizeof(node_load));
@@ -6553,7 +6586,7 @@ static void __build_all_zonelists(void *data)
#endif
}
- spin_unlock(&lock);
+ write_sequnlock(&zonelist_update_seq);
}
static noinline void __init
Patrick Daly reported the following problem; NODE_DATA(nid)->node_zonelists[ZONELIST_FALLBACK] - before offline operation [0] - ZONE_MOVABLE [1] - ZONE_NORMAL [2] - NULL For a GFP_KERNEL allocation, alloc_pages_slowpath() will save the offset of ZONE_NORMAL in ac->preferred_zoneref. If a concurrent memory_offline operation removes the last page from ZONE_MOVABLE, build_all_zonelists() & build_zonerefs_node() will update node_zonelists as shown below. Only populated zones are added. NODE_DATA(nid)->node_zonelists[ZONELIST_FALLBACK] - after offline operation [0] - ZONE_NORMAL [1] - NULL [2] - NULL The race is simple -- page allocation could be in progress when a memory hot-remove operation triggers a zonelist rebuild that removes zones. The allocation request will still have a valid ac->preferred_zoneref that is now pointing to NULL and triggers an OOM kill. This problem probably always existed but may be slightly easier to trigger due to 6aa303defb74 ("mm, vmscan: only allocate and reclaim from zones with pages managed by the buddy allocator") which distinguishes between zones that are completely unpopulated versus zones that have valid pages not managed by the buddy allocator (e.g. reserved, memblock, ballooning etc). Memory hotplug had multiple stages with timing considerations around managed/present page updates, the zonelist rebuild and the zone span updates. As David Hildenbrand puts it memory offlining adjusts managed+present pages of the zone essentially in one go. If after the adjustments, the zone is no longer populated (present==0), we rebuild the zone lists. Once that's done, we try shrinking the zone (start+spanned pages) -- which results in zone_start_pfn == 0 if there are no more pages. That happens *after* rebuilding the zonelists via remove_pfn_range_from_zone(). The only requirement to fix the race is that a page allocation request identifies when a zonelist rebuild has happened since the allocation request started and no page has yet been allocated. Use a seqlock_t to track zonelist updates with a lockless read-side of the zonelist and protecting the rebuild and update of the counter with a spinlock. Fixes: 6aa303defb74 ("mm, vmscan: only allocate and reclaim from zones with pages managed by the buddy allocator") Reported-by: Patrick Daly <quic_pdaly@quicinc.com> Signed-off-by: Mel Gorman <mgorman@techsingularity.net> Acked-by: Michal Hocko <mhocko@suse.com> Reviewed-by: David Hildenbrand <david@redhat.com> Cc: <stable@vger.kernel.org> # v4.9+ --- mm/page_alloc.c | 53 +++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 43 insertions(+), 10 deletions(-)