The previous patch moved mcheck_cpu_init() out of identify_cpu() and put
it as the first CPU hotplug callback which is invoked on the target CPU
during bring up. It enables MCE and starts the MCE timer. If a CPU goes
down then those two things have to be reverted and this happens
currently in CPU_DOWN_PREPARE. This is not symmetrical because
CPU_DOWN_PREPARE is the counterpart of CPU_ONLINE.
Usually CPU_DOWN_FAILED and CPU_ONLINE can do the same thing but not in
this case since here (in CPU_DOWN_FAILED) it tries to revert what was
done in CPU_DOWN_PREPARE.

To make this simpler make CPU_DOWN_PREPARE the counterpart of
mcheck_cpu_starting() and just disable MCE and stop the timer. With this
change the callback is symmetrical again and we don't need CPU_DOWN_FAILED
including mce_reenable_cpu(). smp_call_function_single() can be dropped
because it is already invoked on the proper CPU and interrupts are disabled
at this point.

Cc: Tony Luck <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Sebastian Andrzej Siewior <[email protected]>
Signed-off-by: Thomas Gleixner <[email protected]>
---
 arch/x86/kernel/cpu/mcheck/mce.c | 44 +++++++++++-----------------------------
 1 file changed, 12 insertions(+), 32 deletions(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index 72af9db8526d..596a7128a46b 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -2469,43 +2469,22 @@ static void mce_device_remove(unsigned int cpu)
 }
 
 /* Make sure there are no machine checks on offlined CPUs. */
-static void mce_disable_cpu(void *h)
+static void mce_disable_cpu(void)
 {
-       unsigned long action = *(unsigned long *)h;
-
        if (!mce_available(raw_cpu_ptr(&cpu_info)))
                return;
 
-       if (!(action & CPU_TASKS_FROZEN))
+       if (!cpuhp_tasks_frozen)
                cmci_clear();
 
        vendor_disable_error_reporting();
 }
 
-static void mce_reenable_cpu(void *h)
-{
-       unsigned long action = *(unsigned long *)h;
-       int i;
-
-       if (!mce_available(raw_cpu_ptr(&cpu_info)))
-               return;
-
-       if (!(action & CPU_TASKS_FROZEN))
-               cmci_reenable();
-       for (i = 0; i < mca_cfg.banks; i++) {
-               struct mce_bank *b = &mce_banks[i];
-
-               if (b->init)
-                       wrmsrl(msr_ops.ctl(i), b->ctl);
-       }
-}
-
 /* Get notified when a cpu comes on/off. Be hotplug friendly. */
 static int
 mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
 {
        unsigned int cpu = (unsigned long)hcpu;
-       struct timer_list *t = &per_cpu(mce_timer, cpu);
 
        switch (action & ~CPU_TASKS_FROZEN) {
        case CPU_ONLINE:
@@ -2530,19 +2509,20 @@ mce_cpu_callback(struct notifier_block *nfb, unsigned 
long action, void *hcpu)
                if (!(action & CPU_TASKS_FROZEN))
                        cmci_rediscover();
                break;
-       case CPU_DOWN_PREPARE:
-               smp_call_function_single(cpu, mce_disable_cpu, &action, 1);
-               del_timer_sync(t);
-               break;
-       case CPU_DOWN_FAILED:
-               smp_call_function_single(cpu, mce_reenable_cpu, &action, 1);
-               mce_start_timer(cpu, t);
-               break;
        }
 
        return NOTIFY_OK;
 }
 
+static int mce_cpu_down_dying(unsigned int cpu)
+{
+       struct timer_list *t = this_cpu_ptr(&mce_timer);
+
+       mce_disable_cpu();
+       del_timer_sync(t);
+       return 0;
+}
+
 static struct notifier_block mce_cpu_notifier = {
        .notifier_call = mce_cpu_callback,
 };
@@ -2597,7 +2577,7 @@ static __init int mcheck_init_device(void)
                goto err_init_pool;
 
        err = cpuhp_setup_state(CPUHP_AP_X86_MCE_STARTING, "x86/mce:starting",
-                               mcheck_cpu_starting, NULL);
+                               mcheck_cpu_starting, mce_cpu_down_dying);
        if (err)
                goto err_init_pool;
 
-- 
2.10.2

Reply via email to