ACPI / SBS: Fix rare oops when removing modules
authorRonald Tschalär <ronald@innovation.ch>
Mon, 1 Oct 2018 02:53:13 +0000 (19:53 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 25 Nov 2019 08:53:39 +0000 (09:53 +0100)
[ Upstream commit 757c968c442397f1249bb775a7c8c03842e3e0c7 ]

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 Tschalär <ronald@innovation.ch>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/acpi/osl.c
drivers/acpi/sbshc.c

index 416953a4251094ce2066d724c1d88583b8e89705..b9fade7a3bcf434050d593f4cb468c30e04a9e7b 100644 (file)
@@ -1126,6 +1126,7 @@ 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;
index 7a3431018e0ab4ce96dc055abf31444e498b04bb..5008ead4609a46edafe73b9481b9912c71efc464 100644 (file)
@@ -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;