Subject: [PATCH] ACPI/sbshc: Fix rare oops when removing modules.

There was a small race when removing the sbshc module where
smbus_alarm() had queued acpi_smbus_callback() for deferred execution
but it hadn't been run yet, so that when it did run hc had been freed
and the module unloaded, resulting in an invalid paging request.

A similar race existed when removing the sbs module with regards to
acpi_sbs_callback() (which is called from acpi_smbus_callback()).

We therefore need to ensure no callbacks are pending or executing before
the cleanups are done and the modules are removed.

Signed-off-by: Ronald Tschal?r <[email protected]>
---
drivers/acpi/osl.c | 1 +
drivers/acpi/sbshc.c | 2 ++
2 files changed, 3 insertions(+)

diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
index 8df9abfa947b..9d139727f164 100644
--- a/drivers/acpi/osl.c
+++ b/drivers/acpi/osl.c
@@ -1129,6 +1129,7 @@ void acpi_os_wait_events_complete(void)
flush_workqueue(kacpid_wq);
flush_workqueue(kacpi_notify_wq);
}
+EXPORT_SYMBOL(acpi_os_wait_events_complete);

struct acpi_hp_work {
struct work_struct work;
diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c
index 7a3431018e0a..5008ead4609a 100644
--- a/drivers/acpi/sbshc.c
+++ b/drivers/acpi/sbshc.c
@@ -196,6 +196,7 @@ int acpi_smbus_unregister_callback(struct acpi_smb_hc *hc)
hc->callback = NULL;
hc->context = NULL;
mutex_unlock(&hc->lock);
+ acpi_os_wait_events_complete();
return 0;
}

@@ -292,6 +293,7 @@ static int acpi_smbus_hc_remove(struct acpi_device *device)

hc = acpi_driver_data(device);
acpi_ec_remove_query_handler(hc->ec, hc->query_bit);
+ acpi_os_wait_events_complete();
kfree(hc);
device->driver_data = NULL;
return 0;
--
2.17.1



2018-10-03 09:30:41

by Rafael J. Wysocki

[permalink] [raw]
Subject: Re: [PATCH] ACPI/sbshc: Fix rare oops when removing modules.

On Monday, October 1, 2018 4:53:13 AM CEST Ronald Tschal?r wrote:
> There was a small race when removing the sbshc module where
> smbus_alarm() had queued acpi_smbus_callback() for deferred execution
> but it hadn't been run yet, so that when it did run hc had been freed
> and the module unloaded, resulting in an invalid paging request.
>
> A similar race existed when removing the sbs module with regards to
> acpi_sbs_callback() (which is called from acpi_smbus_callback()).
>
> We therefore need to ensure no callbacks are pending or executing before
> the cleanups are done and the modules are removed.
>
> Signed-off-by: Ronald Tschal?r <[email protected]>
> ---
> drivers/acpi/osl.c | 1 +
> drivers/acpi/sbshc.c | 2 ++
> 2 files changed, 3 insertions(+)
>
> diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c
> index 8df9abfa947b..9d139727f164 100644
> --- a/drivers/acpi/osl.c
> +++ b/drivers/acpi/osl.c
> @@ -1129,6 +1129,7 @@ void acpi_os_wait_events_complete(void)
> flush_workqueue(kacpid_wq);
> flush_workqueue(kacpi_notify_wq);
> }
> +EXPORT_SYMBOL(acpi_os_wait_events_complete);
>
> struct acpi_hp_work {
> struct work_struct work;
> diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c
> index 7a3431018e0a..5008ead4609a 100644
> --- a/drivers/acpi/sbshc.c
> +++ b/drivers/acpi/sbshc.c
> @@ -196,6 +196,7 @@ int acpi_smbus_unregister_callback(struct acpi_smb_hc *hc)
> hc->callback = NULL;
> hc->context = NULL;
> mutex_unlock(&hc->lock);
> + acpi_os_wait_events_complete();
> return 0;
> }
>
> @@ -292,6 +293,7 @@ static int acpi_smbus_hc_remove(struct acpi_device *device)
>
> hc = acpi_driver_data(device);
> acpi_ec_remove_query_handler(hc->ec, hc->query_bit);
> + acpi_os_wait_events_complete();
> kfree(hc);
> device->driver_data = NULL;
> return 0;
>

Applied, thanks!