On Thursday, July 10, 2014 06:11:44 PM Viresh Kumar wrote:
> This is only relevant to implementations with multiple clusters, where
> clusters
> have separate clock lines but all CPUs within a cluster share it.
>
> Consider a dual cluster platform with 2 cores per cluster. During suspend we
> start offlining CPUs from 1 to 3. When CPU2 is remove, policy->kobj would be
> moved to CPU3 and when CPU3 goes down we wouldn't free policy or its kobj.
>
> Now on resume, we will get CPU2 before CPU3 and will call __cpufreq_add_dev().
> We will recover the old policy and update policy->cpu from 3 to 2 from
> update_policy_cpu().
>
> But the kobj is still tied to CPU3 and wasn't moved to CPU2. We wouldn't
> create
> a link for CPU2, but would try that while bringing CPU3 online. Which will
> report errors as CPU3 already has kobj assigned to it.
>
> This bug got introduced with commit 42f921a, which overlooked this scenario.
>
> To fix this, lets move kobj to the new policy->cpu while bringing first CPU
> of a
> cluster back. We already have update_policy_cpu() routine which can be updated
> with this change. That would get rid of the cpufreq_nominate_new_policy_cpu()
> as
> update_policy_cpu() is also called on CPU removal.
>
> To achieve that we add another parameter to update_policy_cpu() as cpu_dev is
> present with both the callers.
>
> Fixes: ("42f921a cpufreq: remove sysfs files for CPUs which failed to come
> back after resume")
> Cc: Stable <stable@vger.kernel.org> # 3.13+
> Reported-by: Bu Yitian <y...@qti.qualcomm.com>
> Reported-by: Saravana Kannan <skan...@codeaurora.org>
> Signed-off-by: Viresh Kumar <viresh.ku...@linaro.org>
> ---
> V1->V2: Move kobject_move() call to update_policy_cpu(), which makes
> cpufreq_nominate_new_policy_cpu() almost empty. So we remove it completely.
>
> @Yitian: Sorry, but you need to test this again as there were enough
> modifications in V2.
>
> drivers/cpufreq/cpufreq.c | 73
> +++++++++++++++++++++++------------------------
> 1 file changed, 36 insertions(+), 37 deletions(-)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 62259d2..c81d9ec6 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1076,10 +1076,20 @@ static void cpufreq_policy_free(struct cpufreq_policy
> *policy)
> kfree(policy);
> }
>
> -static void update_policy_cpu(struct cpufreq_policy *policy, unsigned int
> cpu)
> +static int update_policy_cpu(struct cpufreq_policy *policy, unsigned int cpu,
> + struct device *cpu_dev)
> {
> + int ret;
> +
> if (WARN_ON(cpu == policy->cpu))
> - return;
> + return 0;
> +
> + /* Move kobject to the new policy->cpu */
> + ret = kobject_move(&policy->kobj, &cpu_dev->kobj);
> + if (ret) {
> + pr_err("%s: Failed to move kobj: %d\n", __func__, ret);
> + return ret;
Previously, we returned -EINVAL in the kobject_move() failure case. Why are
we changing that now?
> + }
>
> down_write(&policy->rwsem);
>
> @@ -1090,6 +1100,8 @@ static void update_policy_cpu(struct cpufreq_policy
> *policy, unsigned int cpu)
>
> blocking_notifier_call_chain(&cpufreq_policy_notifier_list,
> CPUFREQ_UPDATE_POLICY_CPU, policy);
> +
> + return 0;
> }
>
> static int __cpufreq_add_dev(struct device *dev, struct subsys_interface
> *sif)
> @@ -1154,7 +1166,7 @@ static int __cpufreq_add_dev(struct device *dev, struct
> subsys_interface *sif)
> * by invoking update_policy_cpu().
> */
> if (recover_policy && cpu != policy->cpu)
> - update_policy_cpu(policy, cpu);
> + WARN_ON(update_policy_cpu(policy, cpu, dev));
This is an arbitrary difference in the handling of update_policy_cpu() return
value. Why do we want the WARN_ON() here and not in the other place?
Don't we want to recover from kobject_move() failures here as well?
> else
> policy->cpu = cpu;
>
> @@ -1307,38 +1319,11 @@ static int cpufreq_add_dev(struct device *dev, struct
> subsys_interface *sif)
> return __cpufreq_add_dev(dev, sif);
> }
>
> -static int cpufreq_nominate_new_policy_cpu(struct cpufreq_policy *policy,
> - unsigned int old_cpu)
> -{
> - struct device *cpu_dev;
> - int ret;
> -
> - /* first sibling now owns the new sysfs dir */
> - cpu_dev = get_cpu_device(cpumask_any_but(policy->cpus, old_cpu));
> -
> - sysfs_remove_link(&cpu_dev->kobj, "cpufreq");
> - ret = kobject_move(&policy->kobj, &cpu_dev->kobj);
> - if (ret) {
> - pr_err("%s: Failed to move kobj: %d\n", __func__, ret);
> -
> - down_write(&policy->rwsem);
> - cpumask_set_cpu(old_cpu, policy->cpus);
> - up_write(&policy->rwsem);
Why don't we need the above three lines in the new code?
> -
> - ret = sysfs_create_link(&cpu_dev->kobj, &policy->kobj,
> - "cpufreq");
> -
> - return -EINVAL;
> - }
> -
> - return cpu_dev->id;
> -}
> -
> static int __cpufreq_remove_dev_prepare(struct device *dev,
> struct subsys_interface *sif)
> {
> unsigned int cpu = dev->id, cpus;
> - int new_cpu, ret;
> + int ret;
> unsigned long flags;
> struct cpufreq_policy *policy;
>
> @@ -1378,14 +1363,28 @@ static int __cpufreq_remove_dev_prepare(struct device
> *dev,
> if (cpu != policy->cpu) {
> sysfs_remove_link(&dev->kobj, "cpufreq");
> } else if (cpus > 1) {
> - new_cpu = cpufreq_nominate_new_policy_cpu(policy, cpu);
> - if (new_cpu >= 0) {
> - update_policy_cpu(policy, new_cpu);
> + /* Nominate new CPU */
> + int new_cpu = cpumask_any_but(policy->cpus, cpu);
> + struct device *cpu_dev = get_cpu_device(new_cpu);
>
> - if (!cpufreq_suspended)
> - pr_debug("%s: policy Kobject moved to cpu: %d
> from: %d\n",
> - __func__, new_cpu, cpu);
> + sysfs_remove_link(&cpu_dev->kobj, "cpufreq");
> + ret = update_policy_cpu(policy, new_cpu, cpu_dev);
> + if (ret) {
> + /*
> + * To supress compilation warning about return value of
> + * sysfs_create_link().
> + */
> + int temp;
> +
> + /* Create link again if we failed. */
> + temp = sysfs_create_link(&cpu_dev->kobj, &policy->kobj,
> + "cpufreq");
And this is *ugly*.
> + return ret;
> }
> +
> + if (!cpufreq_suspended)
> + pr_debug("%s: policy Kobject moved to cpu: %d from:
> %d\n",
> + __func__, new_cpu, cpu);
> } else if (cpufreq_driver->stop_cpu && cpufreq_driver->setpolicy) {
> cpufreq_driver->stop_cpu(policy);
> }
>
--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.
--
To unsubscribe from this list: send the line "unsubscribe stable" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
