In smp_prepare_cpus(), after max_cpus are booted, instead of breaking
from the loop, the 'continue' statement is used which results in
unnecessarily wasting time by looping NR_CPUS times!
Many things around this could be pulled into generic code in the future, but
for now, fix this particular piece of code locally (because I am unable to
convince myself to ignore it even temporarily, given that it is such a gem!).
And also rewrite the 'if' statement in a more natural way.
Cc: David Howells <dhowells@redhat.com>
Cc: Koichi Yasutake <redacted>
Cc: Thomas Gleixner <redacted>
Cc: Yong Zhang <redacted>
Cc: <redacted>
Signed-off-by: Srivatsa S. Bhat <redacted>
---
arch/mn10300/kernel/smp.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/mn10300/kernel/smp.c b/arch/mn10300/kernel/smp.c
index e62c223..b19e75d2 100644
--- a/arch/mn10300/kernel/smp.c
+++ b/arch/mn10300/kernel/smp.c
@@ -695,9 +695,9 @@ void __init smp_prepare_cpus(unsigned int max_cpus)
/* Boot secondary CPUs (for which phy_id > 0) */
for (phy_id = 0; phy_id < NR_CPUS; phy_id++) {
+ if (cpucount + 1 >= max_cpus)
+ break;
/* Don't boot primary CPU */
- if (max_cpus <= cpucount + 1)
- continue;
if (phy_id != 0)
do_boot_cpu(phy_id);
set_cpu_possible(phy_id, true);