Re: [PATCH v2] i2c: rcar: add SMBus block read support

[Wolfram Sang]

> >  	/* If next received data is the _LAST_, go to new phase. */
> > -	if (priv->pos + 1 == msg->len) {
> > +	if (priv->pos + 1 == msg->len && !recv_len_init) {
> 
> If a message contains a single byte after the length byte,
> when we come here after processing the length (in the same function call),
> "pos" is 1, "len" is 2, and we indeed are going to process the last byte.
> However, "recv_len_init" is still "true", and we skip these corresponding
> register writes, which is probably incorrect.
> The flag in this case should be re-set back to "false" after length
> processing and "pos" moving, but I think the variant in my patch

Confirmed. Tests fail with only one extra byte and clearing
'recv_len_init' fixes the issue. I don't think this is the proper
solution, though. I think it will create more readable code if we update
the checks. So people will understand what we are aiming for. The
current code is already implicit enough.

Attachment: signature.asc
Description: PGP signature