[tip:core/urgent] stop_machine: Mark per cpu stopper enabled early

From: tip-bot for Thomas Gleixner
Date: Tue Feb 26 2013 - 16:39:45 EST

Commit-ID: 46c498c2cdee5efe44f617bcd4f388179be36115
Gitweb: http://git.kernel.org/tip/46c498c2cdee5efe44f617bcd4f388179be36115
Author: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
AuthorDate: Tue, 26 Feb 2013 18:44:33 +0100
Committer: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
CommitDate: Tue, 26 Feb 2013 22:25:17 +0100

stop_machine: Mark per cpu stopper enabled early

commit 14e568e78 (stop_machine: Use smpboot threads) introduced the
following regression:

Before this commit the stopper enabled bit was set in the online

cpu online
stopper(CPU1)->enabled = true;

The conversion to smpboot threads moved the enablement to the wakeup
path of the parked thread. The majority of users seem to have the
following working order:

cpu online
stopper thread runs
stopper(CPU1)->enabled = true;

But Konrad and Sander have observed:

cpu online
stopper thread runs
stopper(CPU1)->enabled = true;

Now the stop machinery kicks CPU0 into the stop loop, where it gets
stuck forever because the queue code saw stopper(CPU1)->enabled ==
false, so CPU0 waits for CPU1 to enter stomp_machine, but the CPU1
stopper work got discarded due to enabled == false.

Add a pre_unpark function to the smpboot thread descriptor and call it
before waking the thread.

This fixes the problem at hand, but the stop_machine code should be
more robust. The stopper->enabled flag smells fishy at best.

Thanks to Konrad for going through a loop of debug patches and
providing the information to decode this issue.

Reported-and-tested-by: Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>
Reported-and-tested-by: Sander Eikelenboom <linux@xxxxxxxxxxxxxx>
Cc: Srivatsa S. Bhat <srivatsa.bhat@xxxxxxxxxxxxxxxxxx>
Cc: Rusty Russell <rusty@xxxxxxxxxxxxxxx>
Link: alpine.LFD.2.02.1302261843240.22263@ionos">http://lkml.kernel.org/r/alpine.LFD.2.02.1302261843240.22263@ionos
Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
include/linux/smpboot.h | 4 ++++
kernel/smpboot.c | 2 ++
kernel/stop_machine.c | 2 +-
3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/include/linux/smpboot.h b/include/linux/smpboot.h
index c65dee0..13e9296 100644
--- a/include/linux/smpboot.h
+++ b/include/linux/smpboot.h
@@ -24,6 +24,9 @@ struct smpboot_thread_data;
* parked (cpu offline)
* @unpark: Optional unpark function, called when the thread is
* unparked (cpu online)
+ * @pre_unpark: Optional unpark function, called before the thread is
+ * unparked (cpu online). This is not guaranteed to be
+ * called on the target cpu of the thread. Careful!
* @selfparking: Thread is not parked by the park function.
* @thread_comm: The base name of the thread
@@ -37,6 +40,7 @@ struct smp_hotplug_thread {
void (*cleanup)(unsigned int cpu, bool online);
void (*park)(unsigned int cpu);
void (*unpark)(unsigned int cpu);
+ void (*pre_unpark)(unsigned int cpu);
bool selfparking;
const char *thread_comm;
diff --git a/kernel/smpboot.c b/kernel/smpboot.c
index d4abac2..8eaed9a 100644
--- a/kernel/smpboot.c
+++ b/kernel/smpboot.c
@@ -209,6 +209,8 @@ static void smpboot_unpark_thread(struct smp_hotplug_thread *ht, unsigned int cp
struct task_struct *tsk = *per_cpu_ptr(ht->store, cpu);

+ if (ht->pre_unpark)
+ ht->pre_unpark(cpu);

diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
index 95d178c..c09f295 100644
--- a/kernel/stop_machine.c
+++ b/kernel/stop_machine.c
@@ -336,7 +336,7 @@ static struct smp_hotplug_thread cpu_stop_threads = {
.create = cpu_stop_create,
.setup = cpu_stop_unpark,
.park = cpu_stop_park,
- .unpark = cpu_stop_unpark,
+ .pre_unpark = cpu_stop_unpark,
.selfparking = true,

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/