sched/topology: improve topology_span_sane speed

JIRA: https://issues.redhat.com/browse/RHEL-110301

commit f55dac1
Author: Steve Wahl <[email protected]>
Date:   Tue Mar 4 10:08:43 2025 -0600

    sched/topology: improve topology_span_sane speed

    Use a different approach to topology_span_sane(), that checks for the
    same constraint of no partial overlaps for any two CPU sets for
    non-NUMA topology levels, but does so in a way that is O(N) rather
    than O(N^2).

    Instead of comparing with all other masks to detect collisions, keep
    one mask that includes all CPUs seen so far and detect collisions with
    a single cpumask_intersects test.

    If the current mask has no collisions with previously seen masks, it
    should be a new mask, which can be uniquely identified by the lowest
    bit set in this mask.  Keep a pointer to this mask for future
    reference (in an array indexed by the lowest bit set), and add the
    CPUs in this mask to the list of those seen.

    If the current mask does collide with previously seen masks, it should
    be exactly equal to a mask seen before, looked up in the same array
    indexed by the lowest bit set in the mask, a single comparison.

    Move the topology_span_sane() check out of the existing topology level
    loop, let it use its own loop so that the array allocation can be done
    only once, shared across levels.

    On a system with 1920 processors (16 sockets, 60 cores, 2 threads),
    the average time to take one processor offline is reduced from 2.18
    seconds to 1.01 seconds.  (Off-lining 959 of 1920 processors took
    34m49.765s without this change, 16m10.038s with this change in place.)

    Signed-off-by: Steve Wahl <[email protected]>
    Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
    Reviewed-by: Valentin Schneider <[email protected]>
    Reviewed-by: Madadi Vineeth Reddy <[email protected]>
    Tested-by: K Prateek Nayak <[email protected]>
    Tested-by: Valentin Schneider <[email protected]>
    Tested-by: Madadi Vineeth Reddy <[email protected]>
    Link: https://lore.kernel.org/r/[email protected]

Signed-off-by: Phil Auld <[email protected]>

Author: auldp

kernel/sched/topology.c

@@ -2395,36 +2395,69 @@ static struct sched_domain *build_sched_domain(struct sched_domain_topology_leve
 
 /*
  * Ensure topology masks are sane, i.e. there are no conflicts (overlaps) for
- * any two given CPUs at this (non-NUMA) topology level.
+ * any two given CPUs on non-NUMA topology levels.
  */
-static bool topology_span_sane(struct sched_domain_topology_level *tl,
-			      const struct cpumask *cpu_map, int cpu)
+static bool topology_span_sane(const struct cpumask *cpu_map)
 {
-	int i = cpu + 1;
+	struct sched_domain_topology_level *tl;
+	const struct cpumask **masks;
+	struct cpumask *covered;
+	int cpu, id;
+	bool ret = false;
 
-	/* NUMA levels are allowed to overlap */
-	if (tl->flags & SDTL_OVERLAP)
-		return true;
+	lockdep_assert_held(&sched_domains_mutex);
+	covered = sched_domains_tmpmask;
+
+	masks = kmalloc_array(nr_cpu_ids, sizeof(struct cpumask *), GFP_KERNEL);
+	if (!masks)
+		return ret;
+
+	for_each_sd_topology(tl) {
+
+		/* NUMA levels are allowed to overlap */
+		if (tl->flags & SDTL_OVERLAP)
+			continue;
+
+		cpumask_clear(covered);
+		memset(masks, 0, nr_cpu_ids * sizeof(struct cpumask *));
 
-	/*
-	 * Non-NUMA levels cannot partially overlap - they must be either
-	 * completely equal or completely disjoint. Otherwise we can end up
-	 * breaking the sched_group lists - i.e. a later get_group() pass
-	 * breaks the linking done for an earlier span.
-	 */
-	for_each_cpu_from(i, cpu_map) {
 		/*
-		 * We should 'and' all those masks with 'cpu_map' to exactly
-		 * match the topology we're about to build, but that can only
-		 * remove CPUs, which only lessens our ability to detect
-		 * overlaps
+		 * Non-NUMA levels cannot partially overlap - they must be either
+		 * completely equal or completely disjoint. Otherwise we can end up
+		 * breaking the sched_group lists - i.e. a later get_group() pass
+		 * breaks the linking done for an earlier span.
 		 */
-		if (!cpumask_equal(tl->mask(cpu), tl->mask(i)) &&
-		    cpumask_intersects(tl->mask(cpu), tl->mask(i)))
-			return false;
+		for_each_cpu(cpu, cpu_map) {
+			/* lowest bit set in this mask is used as a unique id */
+			id = cpumask_first(tl->mask(cpu));
+
+			/* zeroed masks cannot possibly collide */
+			if (id >= nr_cpu_ids)
+				continue;
+
+			/* if this mask doesn't collide with what we've already seen */
+			if (!cpumask_intersects(tl->mask(cpu), covered)) {
+				/* this failing would be an error in this algorithm */
+				if (WARN_ON(masks[id]))
+					goto notsane;
+
+				/* record the mask we saw for this id */
+				masks[id] = tl->mask(cpu);
+				cpumask_or(covered, tl->mask(cpu), covered);
+			} else if ((!masks[id]) || !cpumask_equal(masks[id], tl->mask(cpu))) {
+				/*
+				 * a collision with covered should have exactly matched
+				 * a previously seen mask with the same id
+				 */
+				goto notsane;
+			}
+		}
 	}
+	ret = true;
 
-	return true;
+ notsane:
+	kfree(masks);
+	return ret;
 }
 
 /*
@@ -2456,9 +2489,6 @@ build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_attr *att
 		sd = NULL;
 		for_each_sd_topology(tl) {
 
-			if (WARN_ON(!topology_span_sane(tl, cpu_map, i)))
-				goto error;
-
 			sd = build_sched_domain(tl, cpu_map, attr, sd, i);
 
 			has_asym |= sd->flags & SD_ASYM_CPUCAPACITY;
@@ -2472,6 +2502,9 @@ build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_attr *att
 		}
 	}
 
+	if (WARN_ON(!topology_span_sane(cpu_map)))
+		goto error;
+
 	/* Build the groups for the domains */
 	for_each_cpu(i, cpu_map) {
 		for (sd = *per_cpu_ptr(d.sd, i); sd; sd = sd->parent) {