linux-kernel - [tip: x86/cpu] x86/cpu: Take NUMA node into account when allocating per-CPU cpumasks

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Date: Wed, 10 Apr 2024 05:23:10 -0000
From: "tip-bot2 for Li RongQing" <tip-bot2@...utronix.de>
To: linux-tip-commits@...r.kernel.org
Cc: Li RongQing <lirongqing@...du.com>, Ingo Molnar <mingo@...nel.org>,
 x86@...nel.org, linux-kernel@...r.kernel.org
Subject: [tip: x86/cpu] x86/cpu: Take NUMA node into account when allocating
 per-CPU cpumasks

The following commit has been merged into the x86/cpu branch of tip:

Commit-ID:     e0a9ac192fd62322b932c6018db60217b3ad866d
Gitweb:        https://git.kernel.org/tip/e0a9ac192fd62322b932c6018db60217b3ad866d
Author:        Li RongQing <lirongqing@...du.com>
AuthorDate:    Wed, 10 Apr 2024 11:01:14 +08:00
Committer:     Ingo Molnar <mingo@...nel.org>
CommitterDate: Wed, 10 Apr 2024 06:55:31 +02:00

x86/cpu: Take NUMA node into account when allocating per-CPU cpumasks

per-CPU cpumasks are dominantly accessed from their own local CPUs,
so allocate them node-local to improve performance.

[ mingo: Rewrote the changelog. ]

Signed-off-by: Li RongQing <lirongqing@...du.com>
Signed-off-by: Ingo Molnar <mingo@...nel.org>
Link: https://lore.kernel.org/r/20240410030114.6201-1-lirongqing@baidu.com
---
 arch/x86/kernel/apic/x2apic_cluster.c |  3 ++-
 arch/x86/kernel/smpboot.c             | 13 +++++++------
 2 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c
index 567dbd2..afbb885 100644
--- a/arch/x86/kernel/apic/x2apic_cluster.c
+++ b/arch/x86/kernel/apic/x2apic_cluster.c
@@ -183,7 +183,8 @@ static int x2apic_prepare_cpu(unsigned int cpu)
 
 	if (alloc_clustermask(cpu, cluster, cpu_to_node(cpu)) < 0)
 		return -ENOMEM;
-	if (!zalloc_cpumask_var(&per_cpu(ipi_mask, cpu), GFP_KERNEL))
+	if (!zalloc_cpumask_var_node(&per_cpu(ipi_mask, cpu), GFP_KERNEL,
+				cpu_to_node(cpu)))
 		return -ENOMEM;
 	return 0;
 }
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 76bb650..536dad1 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -1033,7 +1033,7 @@ static __init void disable_smp(void)
 
 void __init smp_prepare_cpus_common(void)
 {
-	unsigned int i;
+	unsigned int i, n;
 
 	/* Mark all except the boot CPU as hotpluggable */
 	for_each_possible_cpu(i) {
@@ -1042,11 +1042,12 @@ void __init smp_prepare_cpus_common(void)
 	}
 
 	for_each_possible_cpu(i) {
-		zalloc_cpumask_var(&per_cpu(cpu_sibling_map, i), GFP_KERNEL);
-		zalloc_cpumask_var(&per_cpu(cpu_core_map, i), GFP_KERNEL);
-		zalloc_cpumask_var(&per_cpu(cpu_die_map, i), GFP_KERNEL);
-		zalloc_cpumask_var(&per_cpu(cpu_llc_shared_map, i), GFP_KERNEL);
-		zalloc_cpumask_var(&per_cpu(cpu_l2c_shared_map, i), GFP_KERNEL);
+		n = cpu_to_node(i);
+		zalloc_cpumask_var_node(&per_cpu(cpu_sibling_map, i), GFP_KERNEL, n);
+		zalloc_cpumask_var_node(&per_cpu(cpu_core_map, i), GFP_KERNEL, n);
+		zalloc_cpumask_var_node(&per_cpu(cpu_die_map, i), GFP_KERNEL, n);
+		zalloc_cpumask_var_node(&per_cpu(cpu_llc_shared_map, i), GFP_KERNEL, n);
+		zalloc_cpumask_var_node(&per_cpu(cpu_l2c_shared_map, i), GFP_KERNEL, n);
 	}
 
 	set_cpu_sibling_map(0);