On Mon, May 17, 2021 at 6:14 PM Bjorn Andersson
[off-list ref] wrote:

On Mon 17 May 04:13 CDT 2021, Andy Shevchenko wrote:

quoted

On Sun, May 16, 2021 at 6:47 AM Bjorn Andersson
[off-list ref] wrote:

quoted

In typec_mux_match() "nval" is assigned the number of elements in the
"svid" fwnode property, then the variable is used to store the success
of the read and finally attempts to loop between 0 and "success" - i.e.
not at all - and the code returns indicating that no match was found.

Fix this by using a separate variable to track the success of the read,
to allow the loop to get a change to find a match.

...

quoted

quoted

-       nval = fwnode_property_read_u16_array(fwnode, "svid", val, nval);
-       if (nval < 0) {
+       ret = fwnode_property_read_u16_array(fwnode, "svid", val, nval);
+       if (ret < 0) {
                kfree(val);
-               return ERR_PTR(nval);
+               return ERR_PTR(ret);
        }

This changes the behaviour of the original code, i.e. nval can be
still positive but less than we got from previous call. Some fwnode
backends in some cases potentially can _successfully_ read less than
asked.

Perhaps

  nval = ret;

or drop the patch.

Per the kerneldoc of fwnode_property_read_u16_array:

 * Return: number of values if @val was %NULL,
 *         %0 if the property was found (success),

@val is not NULL, as we just checked for that, so the function will
always return 0 on success.

I don't see anything indicating that the number of elements can be
different from what fwnode_property_count_u16() returned.

Okay, I have checked the backends of fwnode and indeed, OF case (from
where I remember such behaviour) deliberately does

if (ret >= 0)
  return 0;

Otherwise the rest return 0 directly / explicitly.

The only exception is _read_string_array().

quoted

quoted

        for (i = 0; i < nval; i++) {

-- 
With Best Regards,
Andy Shevchenko