]> git.kernelconcepts.de Git - karo-tx-linux.git/commitdiff
genirq/affinity: Fix calculating vectors to assign
authorKeith Busch <keith.busch@intel.com>
Wed, 19 Apr 2017 23:51:10 +0000 (19:51 -0400)
committerThomas Gleixner <tglx@linutronix.de>
Thu, 20 Apr 2017 14:03:09 +0000 (16:03 +0200)
The vectors_per_node is calculated from the remaining available vectors.
The current vector starts after pre_vectors, so we need to subtract that
from the current to properly account for the number of remaining vectors
to assign.

Fixes: 3412386b531 ("irq/affinity: Fix extra vecs calculation")
Reported-by: Andrei Vagin <avagin@virtuozzo.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Link: http://lkml.kernel.org/r/1492645870-13019-1-git-send-email-keith.busch@intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
kernel/irq/affinity.c

index d052947fe78507ac80bbe58a09cc612bca7bab73..e2d356dd75812df8e42dfac3feb132edd5612e33 100644 (file)
@@ -98,7 +98,7 @@ irq_create_affinity_masks(int nvecs, const struct irq_affinity *affd)
                int ncpus, v, vecs_to_assign, vecs_per_node;
 
                /* Spread the vectors per node */
-               vecs_per_node = (affv - curvec) / nodes;
+               vecs_per_node = (affv - (curvec - affd->pre_vectors)) / nodes;
 
                /* Get the cpus on this node which are in the mask */
                cpumask_and(nmsk, cpu_online_mask, cpumask_of_node(n));