Oliver Neukum <oneukum@xxxxxxx> writes:
> On Fri, 2015-03-20 at 15:02 +0100, Bjørn Mork wrote:
>> Oliver Neukum <oneukum@xxxxxxx> writes:
>>
>> > This makes sure the error handling path is the same for
>> > all error conditions, thus reducing code duplication.
>>
>> Great! This looks like I nice cleanup. Will test it out. Just a quick
>> comment based on the first glance:
>>
>> > outnl:
>> > return rv < 0 ? rv : count;
>>
>> How about changing the single remaining error without any rewinding into
>> a direct "return -ENOMEM", and then drop the "outnl" label and simplify
>> this to "return count"?
>
> That would be the direct opposite of what this patch tries to do.
> Direct returns in a function with lock are not what I prefer to do.
OK. I just noticed that you left the direct return just prior to this
one, after fixing up the error translation.
Anyway, the "outnl" is only ever used with rv < 0, and with your changes
all other cases where rv < 0 will jump to some other label. So I don't
think it makes much sense to keep the "return rv < 0 ? rv : count" even
if you don't want the direct return. Just move the "outnl" label
(possibly with a more describing name) down to the "return rv" at the
end.
I.e something like this:
bjorn@nemi:/usr/local/src/git/linux$ git diff
diff --git a/drivers/usb/class/cdc-wdm.c b/drivers/usb/class/cdc-wdm.c
index b78aa6f2a8c1..daa1ac300ac5 100644
--- a/drivers/usb/class/cdc-wdm.c
+++ b/drivers/usb/class/cdc-wdm.c
@@ -430,8 +430,7 @@ static ssize_t wdm_write
usb_autopm_put_interface(desc->intf);
mutex_unlock(&desc->wlock);
-outnl:
- return rv < 0 ? rv : count;
+ return count;
out_free_mem_pm:
usb_autopm_put_interface(desc->intf);
@@ -439,6 +438,7 @@ out_free_mem_lock:
mutex_unlock(&desc->wlock);
out_free_mem:
kfree(buf);
+outnl:
return rv;
}
Bjørn
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
