schedule_work() returns a boolean (true on success, false if the work was already queued). This doesn't match the "negative number on error" return model that the function exports. Given that schedule_work() will always return true (we have an internal .pending check protected by a spinlock), just return false and ignore the return value of schedule_work(). Signed-off-by: Benjamin Tissoires <benjamin.tissoires@xxxxxxxxxx> --- no changes in v3 new in v2 --- drivers/i2c/i2c-smbus.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/i2c/i2c-smbus.c b/drivers/i2c/i2c-smbus.c index 819e5f2..f6780de 100644 --- a/drivers/i2c/i2c-smbus.c +++ b/drivers/i2c/i2c-smbus.c @@ -357,7 +357,10 @@ int i2c_handle_smbus_host_notify(struct smbus_host_notify *host_notify, host_notify->pending = true; spin_unlock_irqrestore(&host_notify->lock, flags); - return schedule_work(&host_notify->work); + /* schedule_work is called if .pending is false, so it can't fail. */ + schedule_work(&host_notify->work); + + return 0; } EXPORT_SYMBOL_GPL(i2c_handle_smbus_host_notify); -- 2.7.4 -- To unsubscribe from this list: send the line "unsubscribe linux-i2c" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
