[RFC PATCH 9/9] workqueue: Affine unbound workqueues to housekeeping cpumask

[Frederic Weisbecker]

Although the unbound workqueue cpumask can be overriden through sysfs,
the housekeeping cpumask which drives the CPU isolation provides a
relevant default value.

Signed-off-by: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Cc: Chris Metcalf <cmetcalf@xxxxxxxxxxxx>
Cc: Rik van Riel <riel@xxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Mike Galbraith <efault@xxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Christoph Lameter <cl@xxxxxxxxx>
Cc: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
Cc: Wanpeng Li <kernellwp@xxxxxxxxx>
Cc: Luiz Capitulino <lcapitulino@xxxxxxxxxx>
---
 kernel/workqueue.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index a86688f..4303c06 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -48,6 +48,7 @@
 #include <linux/nodemask.h>
 #include <linux/moduleparam.h>
 #include <linux/uaccess.h>
+#include <linux/housekeeping.h>
 
 #include "workqueue_internal.h"
 
@@ -5524,7 +5525,7 @@ int __init workqueue_init_early(void)
 	WARN_ON(__alignof__(struct pool_workqueue) < __alignof__(long long));
 
 	BUG_ON(!alloc_cpumask_var(&wq_unbound_cpumask, GFP_KERNEL));
-	cpumask_copy(wq_unbound_cpumask, cpu_possible_mask);
+	cpumask_copy(wq_unbound_cpumask, housekeeping_cpumask());
 
 	pwq_cache = KMEM_CACHE(pool_workqueue, SLAB_PANIC);
 
-- 
2.7.4