The quilt patch titled
Subject: smpboot: use atomic_try_cmpxchg in cpu_wait_death and cpu_report_death
has been removed from the -mm tree. Its filename was
smpboot-use-atomic_try_cmpxchg-in-cpu_wait_death-and-cpu_report_death.patch
This patch was dropped because it was merged into the mm-nonmm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: Uros Bizjak <ubizjak@xxxxxxxxx>
Subject: smpboot: use atomic_try_cmpxchg in cpu_wait_death and cpu_report_death
Date: Thu, 25 Aug 2022 16:56:03 +0200
Use atomic_try_cmpxchg instead of atomic_cmpxchg (*ptr, old, new) == old
in cpu_wait_death and cpu_report_death. x86 CMPXCHG instruction returns
success in ZF flag, so this change saves a compare after cmpxchg (and
related move instruction in front of cmpxchg). Also, atomic_try_cmpxchg
implicitly assigns old *ptr value to "old" when cmpxchg fails, enabling
further code simplifications.
No functional change intended.
Link: https://lkml.kernel.org/r/20220825145603.5811-1-ubizjak@xxxxxxxxx
Signed-off-by: Uros Bizjak <ubizjak@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
kernel/smpboot.c | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
--- a/kernel/smpboot.c~smpboot-use-atomic_try_cmpxchg-in-cpu_wait_death-and-cpu_report_death
+++ a/kernel/smpboot.c
@@ -433,7 +433,7 @@ bool cpu_wait_death(unsigned int cpu, in
/* The outgoing CPU will normally get done quite quickly. */
if (atomic_read(&per_cpu(cpu_hotplug_state, cpu)) == CPU_DEAD)
- goto update_state;
+ goto update_state_early;
udelay(5);
/* But if the outgoing CPU dawdles, wait increasingly long times. */
@@ -444,16 +444,17 @@ bool cpu_wait_death(unsigned int cpu, in
break;
sleep_jf = DIV_ROUND_UP(sleep_jf * 11, 10);
}
-update_state:
+update_state_early:
oldstate = atomic_read(&per_cpu(cpu_hotplug_state, cpu));
+update_state:
if (oldstate == CPU_DEAD) {
/* Outgoing CPU died normally, update state. */
smp_mb(); /* atomic_read() before update. */
atomic_set(&per_cpu(cpu_hotplug_state, cpu), CPU_POST_DEAD);
} else {
/* Outgoing CPU still hasn't died, set state accordingly. */
- if (atomic_cmpxchg(&per_cpu(cpu_hotplug_state, cpu),
- oldstate, CPU_BROKEN) != oldstate)
+ if (!atomic_try_cmpxchg(&per_cpu(cpu_hotplug_state, cpu),
+ &oldstate, CPU_BROKEN))
goto update_state;
ret = false;
}
@@ -475,14 +476,14 @@ bool cpu_report_death(void)
int newstate;
int cpu = smp_processor_id();
+ oldstate = atomic_read(&per_cpu(cpu_hotplug_state, cpu));
do {
- oldstate = atomic_read(&per_cpu(cpu_hotplug_state, cpu));
if (oldstate != CPU_BROKEN)
newstate = CPU_DEAD;
else
newstate = CPU_DEAD_FROZEN;
- } while (atomic_cmpxchg(&per_cpu(cpu_hotplug_state, cpu),
- oldstate, newstate) != oldstate);
+ } while (!atomic_try_cmpxchg(&per_cpu(cpu_hotplug_state, cpu),
+ &oldstate, newstate));
return newstate == CPU_DEAD;
}
_
Patches currently in -mm which might be from ubizjak@xxxxxxxxx are
