Hi,

> Ah, so what I meant above is there is no 'STOP' bit between each msg in
> i2c_msgs,
> but 'REAPEATED_START' still holds true.  We are sending 'START' bit for each
> msg.
> So these is how each msg in i2c_msg is sent,
>
> |------MSG1--------|-----MSG2---------|------MSG3------------|
>
>   |START|DATA|------|START|DATA|---|START|DATA|STOP|
>
> If my commit text does not make this clear, I  can reword that ?

OK, now this looks to me perfectly fine: A number of *messages*
concatenated into one *transfer* by repeated start. That's the way it
should be.

So, I'd simply remove these words:

"The QUP i2c hardware has no way to inform that there should not be a
'STOP' at the end of transaction. The only way to implement this is to
coalesce all the i2c_msg in i2c_msgs in to one transaction and transfer
them."

This sounded like the HW needed a special handling, so I was under the
impression REP_START was broken. However, unless I misunderstood
something again, this now sounds like the standard case and we can keep
the commit message simple. If you are okay with that, I can update it
here, no need to resend.

Thanks,

   Wolfram