[PATCH 4.4 21/28] sparc64: Fix find_node warning if numa node cannot be found

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

4.4-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Thomas Tai <thomas.tai@oracle.com>


[ Upstream commit 74a5ed5c4f692df2ff0a2313ea71e81243525519 ]

When booting up LDOM, find_node() warns that a physical address
doesn't match a NUMA node.

WARNING: CPU: 0 PID: 0 at arch/sparc/mm/init_64.c:835
find_node+0xf4/0x120 find_node: A physical address doesn't
match a NUMA node rule. Some physical memory will be
owned by node 0.Modules linked in:

CPU: 0 PID: 0 Comm: swapper Not tainted 4.9.0-rc3 #4
Call Trace:
 [0000000000468ba0] __warn+0xc0/0xe0
 [0000000000468c74] warn_slowpath_fmt+0x34/0x60
 [00000000004592f4] find_node+0xf4/0x120
 [0000000000dd0774] add_node_ranges+0x38/0xe4
 [0000000000dd0b1c] numa_parse_mdesc+0x268/0x2e4
 [0000000000dd0e9c] bootmem_init+0xb8/0x160
 [0000000000dd174c] paging_init+0x808/0x8fc
 [0000000000dcb0d0] setup_arch+0x2c8/0x2f0
 [0000000000dc68a0] start_kernel+0x48/0x424
 [0000000000dcb374] start_early_boot+0x27c/0x28c
 [0000000000a32c08] tlb_fixup_done+0x4c/0x64
 [0000000000027f08] 0x27f08

It is because linux use an internal structure node_masks[] to
keep the best memory latency node only. However, LDOM mdesc can
contain single latency-group with multiple memory latency nodes.

If the address doesn't match the best latency node within
node_masks[], it should check for an alternative via mdesc.
The warning message should only be printed if the address
doesn't match any node_masks[] nor within mdesc. To minimize
the impact of searching mdesc every time, the last matched
mask and index is stored in a variable.

Signed-off-by: Thomas Tai <thomas.tai@oracle.com>
Reviewed-by: Chris Hyser <chris.hyser@oracle.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 arch/sparc/mm/init_64.c |   65 +++++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 61 insertions(+), 4 deletions(-)

--- a/arch/sparc/mm/init_64.c
+++ b/arch/sparc/mm/init_64.c
@@ -800,6 +800,8 @@ struct mdesc_mblock {
 };
 static struct mdesc_mblock *mblocks;
 static int num_mblocks;
+static int find_numa_node_for_addr(unsigned long pa,
+				   struct node_mem_mask *pnode_mask);
 
 static unsigned long ra_to_pa(unsigned long addr)
 {
@@ -819,6 +821,9 @@ static unsigned long ra_to_pa(unsigned l
 
 static int find_node(unsigned long addr)
 {
+	static bool search_mdesc = true;
+	static struct node_mem_mask last_mem_mask = { ~0UL, ~0UL };
+	static int last_index;
 	int i;
 
 	addr = ra_to_pa(addr);
@@ -828,10 +833,27 @@ static int find_node(unsigned long addr)
 		if ((addr & p->mask) == p->val)
 			return i;
 	}
-	/* The following condition has been observed on LDOM guests.*/
-	WARN_ONCE(1, "find_node: A physical address doesn't match a NUMA node"
-		" rule. Some physical memory will be owned by node 0.");
-	return 0;
+	/* The following condition has been observed on LDOM guests because
+	 * node_masks only contains the best latency mask and value.
+	 * LDOM guest's mdesc can contain a single latency group to
+	 * cover multiple address range. Print warning message only if the
+	 * address cannot be found in node_masks nor mdesc.
+	 */
+	if ((search_mdesc) &&
+	    ((addr & last_mem_mask.mask) != last_mem_mask.val)) {
+		/* find the available node in the mdesc */
+		last_index = find_numa_node_for_addr(addr, &last_mem_mask);
+		numadbg("find_node: latency group for address 0x%lx is %d\n",
+			addr, last_index);
+		if ((last_index < 0) || (last_index >= num_node_masks)) {
+			/* WARN_ONCE() and use default group 0 */
+			WARN_ONCE(1, "find_node: A physical address doesn't match a NUMA node rule. Some physical memory will be owned by node 0.");
+			search_mdesc = false;
+			last_index = 0;
+		}
+	}
+
+	return last_index;
 }
 
 static u64 memblock_nid_range(u64 start, u64 end, int *nid)
@@ -1158,6 +1180,41 @@ int __node_distance(int from, int to)
 	return numa_latency[from][to];
 }
 
+static int find_numa_node_for_addr(unsigned long pa,
+				   struct node_mem_mask *pnode_mask)
+{
+	struct mdesc_handle *md = mdesc_grab();
+	u64 node, arc;
+	int i = 0;
+
+	node = mdesc_node_by_name(md, MDESC_NODE_NULL, "latency-groups");
+	if (node == MDESC_NODE_NULL)
+		goto out;
+
+	mdesc_for_each_node_by_name(md, node, "group") {
+		mdesc_for_each_arc(arc, md, node, MDESC_ARC_TYPE_FWD) {
+			u64 target = mdesc_arc_target(md, arc);
+			struct mdesc_mlgroup *m = find_mlgroup(target);
+
+			if (!m)
+				continue;
+			if ((pa & m->mask) == m->match) {
+				if (pnode_mask) {
+					pnode_mask->mask = m->mask;
+					pnode_mask->val = m->match;
+				}
+				mdesc_release(md);
+				return i;
+			}
+		}
+		i++;
+	}
+
+out:
+	mdesc_release(md);
+	return -1;
+}
+
 static int find_best_numa_node_for_mlgroup(struct mdesc_mlgroup *grp)
 {
 	int i;