Re: [PATCH] sched/numa: Fix NUMA_DIRECT topology identification

Re: [PATCH] sched/numa: Fix NUMA_DIRECT topology identification

[Rik van Riel]

On 08/10/2015 09:20 PM, Aravind Gopalakrishnan wrote:
> Systems which have all nodes at a distance of at most
> 1 hop should be identified as 'NUMA_DIRECT'.
> 
> However, the scheduler incorrectly identifies it as
> 'NUMA_BACKPLANE'. This is because 'n' is assigned to
> sched_max_numa_distance but the code (mis)interprets it
> to mean 'number of hops'.
> 
> Rik had actually used sched_domains_numa_levels for detecting
> a 'NUMA_DIRECT' topology-
> http://marc.info/?l=linux-kernel&m=141279712429834&w=2
> 
> But that was changed when he removed the hops table in the
> subsequent version-
> http://marc.info/?l=linux-kernel&m=141353106106771&w=2

Good catch! Thank you!

> Fixing the issue here.
> 
> Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>

Reviewed-by: Rik van Riel <riel@redhat.com>

-- 
All rights reversed

[PATCH] sched/numa: Fix NUMA_DIRECT topology identification

[Aravind Gopalakrishnan]

Systems which have all nodes at a distance of at most
1 hop should be identified as 'NUMA_DIRECT'.

However, the scheduler incorrectly identifies it as
'NUMA_BACKPLANE'. This is because 'n' is assigned to
sched_max_numa_distance but the code (mis)interprets it
to mean 'number of hops'.

Rik had actually used sched_domains_numa_levels for detecting
a 'NUMA_DIRECT' topology-
http://marc.info/?l=linux-kernel&m=141279712429834&w=2

But that was changed when he removed the hops table in the
subsequent version-
http://marc.info/?l=linux-kernel&m=141353106106771&w=2

Fixing the issue here.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
---
 kernel/sched/core.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 655557d..e51739e 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6442,8 +6442,10 @@ static void init_numa_topology_type(void)
 
 	n = sched_max_numa_distance;
 
-	if (n <= 1)
+	if (sched_domains_numa_levels <= 1) {
 		sched_numa_topology_type = NUMA_DIRECT;
+		return;
+	}
 
 	for_each_online_node(a) {
 		for_each_online_node(b) {
-- 
2.4.0

[tip:sched/core] sched/numa: Fix NUMA_DIRECT topology identification

[tip-bot for Aravind Gopalakrishnan]

Commit-ID:  e237882b8f83dd1a0eece1608bcb689d4f4b221b
Gitweb:     http://git.kernel.org/tip/e237882b8f83dd1a0eece1608bcb689d4f4b221b
Author:     Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
AuthorDate: Mon, 10 Aug 2015 20:20:48 -0500
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Wed, 12 Aug 2015 12:06:08 +0200

sched/numa: Fix NUMA_DIRECT topology identification

Systems which have all nodes at a distance of at most 1 hop should be
identified as 'NUMA_DIRECT'.

However, the scheduler incorrectly identifies it as 'NUMA_BACKPLANE'.
This is because 'n' is assigned to sched_max_numa_distance but the
code (mis)interprets it to mean 'number of hops'.

Rik had actually used sched_domains_numa_levels for detecting a
'NUMA_DIRECT' topology:

  http://marc.info/?l=linux-kernel&m=141279712429834&w=2

But that was changed when he removed the hops table in the
subsequent version:

  http://marc.info/?l=linux-kernel&m=141353106106771&w=2

Fixing the issue here.

Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1439256048-3748-1-git-send-email-Aravind.Gopalakrishnan@amd.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 kernel/sched/core.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index b11f624..ea6d743 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6444,8 +6444,10 @@ static void init_numa_topology_type(void)
 
 	n = sched_max_numa_distance;
 
-	if (n <= 1)
+	if (sched_domains_numa_levels <= 1) {
 		sched_numa_topology_type = NUMA_DIRECT;
+		return;
+	}
 
 	for_each_online_node(a) {
 		for_each_online_node(b) {