BPF List
 help / color / mirror / Atom feed
From: Andrea Righi <arighi@nvidia.com>
To: Tejun Heo <tj@kernel.org>, David Vernet <void@manifault.com>,
	Changwoo Min <changwoo@igalia.com>
Cc: Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Juri Lelli <juri.lelli@redhat.com>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	Ben Segall <bsegall@google.com>, Mel Gorman <mgorman@suse.de>,
	Valentin Schneider <vschneid@redhat.com>,
	Ian May <ianm@nvidia.com>,
	bpf@vger.kernel.org, linux-kernel@vger.kernel.org,
	Yury Norov <yury.norov@gmail.com>
Subject: [PATCH 1/5] sched/topology: Introduce for_each_numa_node() iterator
Date: Thu,  6 Feb 2025 21:15:31 +0100	[thread overview]
Message-ID: <20250206202109.384179-2-arighi@nvidia.com> (raw)
In-Reply-To: <20250206202109.384179-1-arighi@nvidia.com>

Introduce for_each_numa_node() and sched_numa_node() helpers to iterate
over node IDs in order of increasing NUMA distance from a given starting
node.

These iterator functions are similar to for_each_numa_hop_mask() and
sched_numa_hop_mask(), but instead of providing a cpumask at each
iteration, they provide a node ID.

Example usage:

  nodemask_t visited = NODE_MASK_NONE;
  int start = cpu_to_node(smp_processor_id());

  for_each_numa_node(node, start, visited, N_ONLINE)
  	pr_info("node (%d, %d) -> %d\n",
  		 start, node, node_distance(start, node));

On a system with equidistant nodes:

 $ numactl -H
 ...
 node distances:
 node     0    1    2    3
    0:   10   20   20   20
    1:   20   10   20   20
    2:   20   20   10   20
    3:   20   20   20   10

Output of the example above (on node 0):

[    7.367022] node (0, 0) -> 10
[    7.367151] node (0, 1) -> 20
[    7.367186] node (0, 2) -> 20
[    7.367247] node (0, 3) -> 20

On a system with non-equidistant nodes (simulated using virtme-ng):

 $ numactl -H
 ...
 node distances:
 node     0    1    2    3
    0:   10   51   31   41
    1:   51   10   21   61
    2:   31   21   10   11
    3:   41   61   11   10

Output of the example above (on node 0):

 [    8.953644] node (0, 0) -> 10
 [    8.953712] node (0, 2) -> 31
 [    8.953764] node (0, 3) -> 41
 [    8.953817] node (0, 1) -> 51

Cc: Yury Norov <yury.norov@gmail.com>
Signed-off-by: Andrea Righi <arighi@nvidia.com>
---
 include/linux/topology.h | 31 ++++++++++++++++++++++++++++-
 kernel/sched/topology.c  | 42 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 72 insertions(+), 1 deletion(-)

diff --git a/include/linux/topology.h b/include/linux/topology.h
index 52f5850730b3e..0c82b913a8814 100644
--- a/include/linux/topology.h
+++ b/include/linux/topology.h
@@ -248,12 +248,18 @@ static inline const struct cpumask *cpu_cpu_mask(int cpu)
 #ifdef CONFIG_NUMA
 int sched_numa_find_nth_cpu(const struct cpumask *cpus, int cpu, int node);
 extern const struct cpumask *sched_numa_hop_mask(unsigned int node, unsigned int hops);
-#else
+extern int sched_numa_node(nodemask_t *visited, int start, unsigned int state);
+#else /* !CONFIG_NUMA */
 static __always_inline int sched_numa_find_nth_cpu(const struct cpumask *cpus, int cpu, int node)
 {
 	return cpumask_nth_and(cpu, cpus, cpu_online_mask);
 }
 
+static inline int sched_numa_node(nodemask_t *visited, int start, unsigned int state)
+{
+	return MAX_NUMNODES;
+}
+
 static inline const struct cpumask *
 sched_numa_hop_mask(unsigned int node, unsigned int hops)
 {
@@ -261,6 +267,29 @@ sched_numa_hop_mask(unsigned int node, unsigned int hops)
 }
 #endif	/* CONFIG_NUMA */
 
+/**
+ * for_each_numa_node - iterate over NUMA nodes at increasing hop distances
+ *                      from a given starting node.
+ * @node: the iteration variable, representing the current NUMA node.
+ * @start: the NUMA node to start the iteration from.
+ * @visited: a nodemask_t to track the visited nodes.
+ * @state: state of NUMA nodes to iterate.
+ *
+ * This macro iterates over NUMA nodes in increasing distance from
+ * @start_node and yields MAX_NUMNODES when all the nodes have been
+ * visited.
+ *
+ * The difference between for_each_node() and for_each_numa_node() is that
+ * the former allows to iterate over nodes in no particular order, whereas
+ * the latter iterates over nodes in increasing order of distance.
+ *
+ * Requires rcu_lock to be held.
+ */
+#define for_each_numa_node(node, start, visited, state)				\
+	for (node = start;							\
+	     node != MAX_NUMNODES;						\
+	     node = sched_numa_node(&(visited), start, state))
+
 /**
  * for_each_numa_hop_mask - iterate over cpumasks of increasing NUMA distance
  *                          from a given node.
diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index da33ec9e94ab2..e1d0a33415fb5 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -2183,6 +2183,48 @@ int sched_numa_find_nth_cpu(const struct cpumask *cpus, int cpu, int node)
 }
 EXPORT_SYMBOL_GPL(sched_numa_find_nth_cpu);
 
+/**
+ * sched_numa_node - Find the NUMA node at the closest distance from
+ *		     node @start.
+ *
+ * @visited: a pointer to a nodemask_t representing the visited nodes.
+ * @start: the node to start the search from.
+ * @state: the node state to filter nodes by.
+ *
+ * This function iterates over all nodes in the given state and calculates
+ * the distance to the starting node. It returns the node that is the
+ * closest in terms of distance that has not already been considered (not
+ * set in @visited and not the starting node). If the node is found, it is
+ * marked as visited in the @visited node mask.
+ *
+ * Returns the node ID closest in terms of hop distance from the @start
+ * node, or MAX_NUMNODES if no node is found (or all nodes have been
+ * visited).
+ */
+int sched_numa_node(nodemask_t *visited, int start, unsigned int state)
+{
+	int dist, n, min_node, min_dist;
+
+	min_node = MAX_NUMNODES;
+	min_dist = INT_MAX;
+
+	/* Find the nearest unvisted node */
+	for_each_node_state(n, state) {
+		if (n == start || node_isset(n, *visited))
+			continue;
+		dist = node_distance(start, n);
+		if (dist < min_dist) {
+			min_dist = dist;
+			min_node = n;
+		}
+	}
+	if (min_node != MAX_NUMNODES)
+		node_set(min_node, *visited);
+
+	return min_node;
+}
+EXPORT_SYMBOL_GPL(sched_numa_node);
+
 /**
  * sched_numa_hop_mask() - Get the cpumask of CPUs at most @hops hops away from
  *                         @node
-- 
2.48.1


  reply	other threads:[~2025-02-06 20:21 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-06 20:15 [PATCHSET v9 sched_ext/for-6.15] sched_ext: split global idle cpumask into per-NUMA cpumasks Andrea Righi
2025-02-06 20:15 ` Andrea Righi [this message]
2025-02-07  3:57   ` [PATCH 1/5] sched/topology: Introduce for_each_numa_node() iterator Yury Norov
2025-02-07 15:44     ` Andrea Righi
2025-02-07 17:07       ` Yury Norov
2025-02-06 20:15 ` [PATCH 2/5] sched_ext: idle: Introduce SCX_OPS_BUILTIN_IDLE_PER_NODE Andrea Righi
2025-02-06 20:15 ` [PATCH 3/5] sched_ext: idle: introduce SCX_PICK_IDLE_IN_NODE Andrea Righi
2025-02-06 20:15 ` [PATCH 4/5] sched_ext: idle: Per-node idle cpumasks Andrea Righi
2025-02-06 20:15 ` [PATCH 5/5] sched_ext: idle: Introduce node-aware idle cpu kfunc helpers Andrea Righi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250206202109.384179-2-arighi@nvidia.com \
    --to=arighi@nvidia.com \
    --cc=bpf@vger.kernel.org \
    --cc=bsegall@google.com \
    --cc=changwoo@igalia.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=ianm@nvidia.com \
    --cc=juri.lelli@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mgorman@suse.de \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=tj@kernel.org \
    --cc=vincent.guittot@linaro.org \
    --cc=void@manifault.com \
    --cc=vschneid@redhat.com \
    --cc=yury.norov@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox