On 05/04/2012 08:56 AM, Daniel Vetter wrote:
>> +static int i915_error_state_release(struct inode *inode, struct file *file)
>> +{
>> + struct seq_file *m = file->private_data;
>> + struct i915_error_state_file_priv *error_priv = m->private;
>> +
>> + if (error_priv->error)
>> + kref_put(&error_priv->error->ref, i915_error_state_free);
>> + kfree(error_priv);
Maybe a stupid question, but shouldn't we hold the error_lock here as well?
But apart from that:
Reviewed-by: Eugeni Dodonov <eugeni.dodonov at intel.com>
