Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Valentin Schneider <valentin.schneider@arm.com>
To:     linux-kernel@vger.kernel.org
Cc:     mingo@kernel.org, peterz@infradead.org, vincent.guittot@linaro.org,
        dietmar.eggemann@arm.com, morten.rasmussen@arm.com,
        mgorman@suse.de, song.bao.hua@hisilicon.com
Subject: [PATCH 1/1] sched/topology: Make sched_init_numa() use a set for the deduplicating sort
Date:   Fri, 22 Jan 2021 12:39:43 +0000
Message-Id: <20210122123943.1217-2-valentin.schneider@arm.com>
In-Reply-To: <20210122123943.1217-1-valentin.schneider@arm.com>
References: <20210122123943.1217-1-valentin.schneider@arm.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

The deduplicating sort in sched_init_numa() assumes that the first line in
the distance table contains all unique values in the entire table. I've
been trying to pen what this exactly means for the topology, but it's not
straightforward. For instance, topology.c uses this example:

  node   0   1   2   3
    0:  10  20  20  30
    1:  20  10  20  20
    2:  20  20  10  20
    3:  30  20  20  10

  0 ----- 1
  |     / |
  |   /   |
  | /     |
  2 ----- 3

Which works out just fine. However, if we swap nodes 0 and 1:

  1 ----- 0
  |     / |
  |   /   |
  | /     |
  2 ----- 3

we get this distance table:

  node   0  1  2  3
    0:  10 20 20 20
    1:  20 10 20 30
    2:  20 20 10 20
    3:  20 30 20 10

Which breaks the deduplicating sort (non-representative first line). In
this case this would just be a renumbering exercise, but it so happens that
we can have a deduplicating sort that goes through the whole table in O(n²)
at the extra cost of a temporary memory allocation (i.e. any form of set).

The ACPI spec (SLIT) mentions distances are encoded on 8 bits. Following
this, implement the set as a 256-bits bitmap. Should this not be
satisfactory (i.e. we want to support 32-bit values), then we'll have to go
for some other sparse set implementation.

This has the added benefit of letting us allocate just the right amount of
memory for sched_domains_numa_distance[], rather than an arbitrary
(nr_node_ids + 1).

Note: DT binding equivalent (distance-map) decodes distances as 32-bit
values.

Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
---
 include/linux/topology.h |  1 +
 kernel/sched/topology.c  | 99 +++++++++++++++++++---------------------
 2 files changed, 49 insertions(+), 51 deletions(-)

diff --git a/include/linux/topology.h b/include/linux/topology.h
index ad03df1cc266..7634cd737061 100644
--- a/include/linux/topology.h
+++ b/include/linux/topology.h
@@ -48,6 +48,7 @@ int arch_update_cpu_topology(void);
 /* Conform to ACPI 2.0 SLIT distance definitions */
 #define LOCAL_DISTANCE		10
 #define REMOTE_DISTANCE		20
+#define DISTANCE_BITS           8
 #ifndef node_distance
 #define node_distance(from,to)	((from) == (to) ? LOCAL_DISTANCE : REMOTE_DISTANCE)
 #endif
diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index 5d3675c7a76b..bf5c9bd10bfe 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -1596,66 +1596,58 @@ static void init_numa_topology_type(void)
 	}
 }
 
+
+#define NR_DISTANCE_VALUES (1 << DISTANCE_BITS)
+
 void sched_init_numa(void)
 {
-	int next_distance, curr_distance = node_distance(0, 0);
 	struct sched_domain_topology_level *tl;
-	int level = 0;
-	int i, j, k;
-
-	sched_domains_numa_distance = kzalloc(sizeof(int) * (nr_node_ids + 1), GFP_KERNEL);
-	if (!sched_domains_numa_distance)
-		return;
-
-	/* Includes NUMA identity node at level 0. */
-	sched_domains_numa_distance[level++] = curr_distance;
-	sched_domains_numa_levels = level;
+	unsigned long *distance_map;
+	int nr_levels = 0;
+	int i, j;
 
 	/*
 	 * O(nr_nodes^2) deduplicating selection sort -- in order to find the
 	 * unique distances in the node_distance() table.
-	 *
-	 * Assumes node_distance(0,j) includes all distances in
-	 * node_distance(i,j) in order to avoid cubic time.
 	 */
-	next_distance = curr_distance;
+	distance_map = bitmap_alloc(NR_DISTANCE_VALUES, GFP_KERNEL);
+	if (!distance_map)
+		return;
+
+	bitmap_zero(distance_map, NR_DISTANCE_VALUES);
 	for (i = 0; i < nr_node_ids; i++) {
 		for (j = 0; j < nr_node_ids; j++) {
-			for (k = 0; k < nr_node_ids; k++) {
-				int distance = node_distance(i, k);
-
-				if (distance > curr_distance &&
-				    (distance < next_distance ||
-				     next_distance == curr_distance))
-					next_distance = distance;
-
-				/*
-				 * While not a strong assumption it would be nice to know
-				 * about cases where if node A is connected to B, B is not
-				 * equally connected to A.
-				 */
-				if (sched_debug() && node_distance(k, i) != distance)
-					sched_numa_warn("Node-distance not symmetric");
+			int distance = node_distance(i, j);
 
-				if (sched_debug() && i && !find_numa_distance(distance))
-					sched_numa_warn("Node-0 not representative");
+			if (distance < LOCAL_DISTANCE || distance >= NR_DISTANCE_VALUES) {
+				sched_numa_warn("Invalid distance value range");
+				return;
 			}
-			if (next_distance != curr_distance) {
-				sched_domains_numa_distance[level++] = next_distance;
-				sched_domains_numa_levels = level;
-				curr_distance = next_distance;
-			} else break;
+
+			bitmap_set(distance_map, distance, 1);
 		}
+	}
+	/*
+	 * We can now figure out how many unique distance values there are and
+	 * allocate memory accordingly.
+	 */
+	nr_levels = bitmap_weight(distance_map, NR_DISTANCE_VALUES);
 
-		/*
-		 * In case of sched_debug() we verify the above assumption.
-		 */
-		if (!sched_debug())
-			break;
+	sched_domains_numa_distance = kcalloc(nr_levels, sizeof(int), GFP_KERNEL);
+	if (!sched_domains_numa_distance) {
+		bitmap_free(distance_map);
+		return;
+	}
+
+	for (i = 0, j = 0; i < nr_levels; i++, j++) {
+		j = find_next_bit(distance_map, NR_DISTANCE_VALUES, j);
+		sched_domains_numa_distance[i] = j;
 	}
 
+	bitmap_free(distance_map);
+
 	/*
-	 * 'level' contains the number of unique distances
+	 * 'nr_levels' contains the number of unique distances
 	 *
 	 * The sched_domains_numa_distance[] array includes the actual distance
 	 * numbers.
@@ -1664,15 +1656,15 @@ void sched_init_numa(void)
 	/*
 	 * Here, we should temporarily reset sched_domains_numa_levels to 0.
 	 * If it fails to allocate memory for array sched_domains_numa_masks[][],
-	 * the array will contain less then 'level' members. This could be
+	 * the array will contain less then 'nr_levels' members. This could be
 	 * dangerous when we use it to iterate array sched_domains_numa_masks[][]
 	 * in other functions.
 	 *
-	 * We reset it to 'level' at the end of this function.
+	 * We reset it to 'nr_levels' at the end of this function.
 	 */
 	sched_domains_numa_levels = 0;
 
-	sched_domains_numa_masks = kzalloc(sizeof(void *) * level, GFP_KERNEL);
+	sched_domains_numa_masks = kzalloc(sizeof(void *) * nr_levels, GFP_KERNEL);
 	if (!sched_domains_numa_masks)
 		return;
 
@@ -1680,7 +1672,7 @@ void sched_init_numa(void)
 	 * Now for each level, construct a mask per node which contains all
 	 * CPUs of nodes that are that many hops away from us.
 	 */
-	for (i = 0; i < level; i++) {
+	for (i = 0; i < nr_levels; i++) {
 		sched_domains_numa_masks[i] =
 			kzalloc(nr_node_ids * sizeof(void *), GFP_KERNEL);
 		if (!sched_domains_numa_masks[i])
@@ -1688,12 +1680,17 @@ void sched_init_numa(void)
 
 		for (j = 0; j < nr_node_ids; j++) {
 			struct cpumask *mask = kzalloc(cpumask_size(), GFP_KERNEL);
+			int k;
+
 			if (!mask)
 				return;
 
 			sched_domains_numa_masks[i][j] = mask;
 
 			for_each_node(k) {
+				if (sched_debug() && (node_distance(j, k) != node_distance(k, j)))
+					sched_numa_warn("Node-distance not symmetric");
+
 				if (node_distance(j, k) > sched_domains_numa_distance[i])
 					continue;
 
@@ -1705,7 +1702,7 @@ void sched_init_numa(void)
 	/* Compute default topology size */
 	for (i = 0; sched_domain_topology[i].mask; i++);
 
-	tl = kzalloc((i + level + 1) *
+	tl = kzalloc((i + nr_levels) *
 			sizeof(struct sched_domain_topology_level), GFP_KERNEL);
 	if (!tl)
 		return;
@@ -1728,7 +1725,7 @@ void sched_init_numa(void)
 	/*
 	 * .. and append 'j' levels of NUMA goodness.
 	 */
-	for (j = 1; j < level; i++, j++) {
+	for (j = 1; j < nr_levels; i++, j++) {
 		tl[i] = (struct sched_domain_topology_level){
 			.mask = sd_numa_mask,
 			.sd_flags = cpu_numa_flags,
@@ -1740,8 +1737,8 @@ void sched_init_numa(void)
 
 	sched_domain_topology = tl;
 
-	sched_domains_numa_levels = level;
-	sched_max_numa_distance = sched_domains_numa_distance[level - 1];
+	sched_domains_numa_levels = nr_levels;
+	sched_max_numa_distance = sched_domains_numa_distance[nr_levels - 1];
 
 	init_numa_topology_type();
 }
-- 
2.27.0