Hello Guenter, Brian

On 06/18/2016 01:09 PM, Guenter Roeck wrote:

On 06/17/2016 06:08 PM, Brian Norris wrote:

quoted

On Fri, Jun 17, 2016 at 02:41:51PM -0700, Guenter Roeck wrote:

quoted

On Fri, Jun 17, 2016 at 12:58:12PM -0700, Brian Norris wrote:

quoted

+int cros_ec_cmd_xfer_status(struct cros_ec_device *ec_dev,
+                struct cros_ec_command *msg)
+{
+    int ret;
+
+    ret = cros_ec_cmd_xfer(ec_dev, msg);
+    if (ret < 0)
+        dev_err(ec_dev->dev, "Command xfer error (err:%d)\n", ret);
+    else if (msg->result != EC_RES_SUCCESS)
+        return -EECRESULT - msg->result;

I have been wondering about the error return codes here, and if they should be
converted to standard Linux error codes. For example, I just hit error -1003
with a driver I am working on. This translates to EC_RES_INVALID_PARAM, or,
in Linux terms, -EINVAL. I think it would be better to use standard error
codes, especially since some of the errors are logged.

Agreed, specially since drivers may (wrongly) propagate whatever is returned
by this function to higher layers where the ChromeOS EC firmware error codes
makes no sense. So that will be a bug and can increase the cognitive load of
getting some weird error codes in core kernel code and developers may wonder
from where those came from until finally find that a EC driver returned that.

quoted

How do you propose we do that? Do all of the following become EINVAL?

Yes, I would just do that.

The idea of this helper is to remove duplicated code and AFAICT what most EC
drivers do is something similar to the following:

        ret = cros_ec_cmd_xfer(ec, msg);
        if (ret < 0)
	   	return ret;

        if (msg->result != EC_RES_SUCCESS) {
                dev_dbg(ec->dev, "EC result %d\n", msg->result);
                return -EINVAL;
        }

So in practice what most drivers really care is if the result was successful
or not, I don't see specific EC error handling in the EC drivers.  The real
EC error code is still in the message anyways so drivers that do cares about
the real EC error can look at msg->result instead.

quoted

         EC_RES_INVALID_COMMAND

    -EOPNOTSUPP

quoted

         EC_RES_INVALID_PARAM

    -EINVAL or -EBADMSG

quoted

         EC_RES_INVALID_VERSION

    -EPROTO or -EBADR or -EBADE or -EBADRQC or -EPROTOOPT

quoted

         EC_RES_INVALID_HEADER

    -EPROTO or -EBADR or -EBADE

Doesn't look that bad to me. Also, the raw error could still be logged,
for example with dev_dbg().

Yes, I think that adding a dev_dbg() with the real EC error code should
be enough, that's basically what drivers do since they can't propagate
the EC error to higher layers anyways.

Guenter

quoted

Best regards,
-- 
Javier Martinez Canillas
Open Source Group
Samsung Research America
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html