Index | Thread | Search

From:
Kirill A. Korinsky <kirill@korins.ky>
Subject:
Re: qcpas(4) mdt init: add retry
To:
Marcus Glocker <marcus@nazgul.ch>
Cc:
tech@openbsd.org
Date:
Sun, 17 May 2026 10:57:18 +0200

Download raw body.

Thread
On Sun, 17 May 2026 06:39:21 +0200,
Marcus Glocker <marcus@nazgul.ch> wrote:
> 
> On Sat, May 16, 2026 at 10:39:48PM +0200, Kirill A. Korinsky wrote:
> 
> > On Sat, 16 May 2026 18:12:25 +0200,
> > Marcus Glocker <marcus@nazgul.ch> wrote:
> > > 
> > > On the Samsung Galaxy Book4 Edge every since I am getting intermittent
> > > mdt initialization errors from qcpas(4) looking like
> > > 
> > > 	qcpas0: failed to shutdown lite firmware
> > > 	qcpas0: failed to receive ready signal
> > > 	qcpas0: failed to boot coprocessor
> > > 
> > > leaving me without the qcpas(4) HW sensors (they anyway ship no data
> > > on the Samsung currently, but that's a different story I want to tackle
> > > next).
> > > 
> > > Implementing a retry loop will help to overcome this situation:
> > > 
> > > 	qcpas0: failed to shutdown lite firmware
> > > 	qcpas0: failed to receive ready signal
> > > 	qcpas0: coprocessor boot attempt 2
> > > 	qcpas0: failed to receive ready signal
> > > 	qcpas0: coprocessor boot attempt 3
> > > 
> > > Feedback, regression testers, OKs?
> > >
> > 
> > Tested on my honor, it doesn't solve my issue, but it also do not introduce
> > any regression. If it fixes your machine I think it make sense to commit and
> > I OK with it, but I suggest to wait more tests and feedback.
> > 
> > On my device it errors like that:
> > 
> > qcpas0: failed to receive ready signal
> > qcpas0: coprocessor boot attempt 2
> > qcpas0: failed to receive ready signal
> > qcpas0: coprocessor boot attempt 3
> > qcpas0: failed to receive ready signal
> > qcpas0: failed to boot coprocessor
> > 
> > I've tried increase number of attempts, add pause... nothing.
> 
> Thanks for testing.
> 
> But I will need to draw back this diff for now;  It turned out in the
> meantime that while the HW sensors do come back after the retry, the
> chip still isn't fully initialized.
> 
> Without this diff, what situation do you have with qcpas(4) on your
> Honor?
>  

It works. But some boots it does.

But this is a bit chaotic device: some boots I may not have: USB-C ports,
multi touch in tracpad, qcpass or touchscreen.

-- 
wbr, Kirill