RE: [PATCH 00/14] usb: typec: UCSI driver overhaul

[Ajay Gupta <ajayg@nvidia.com>]

Hi Heikki,

> -----Original Message-----
> From: Heikki Krogerus <heikki.krogerus@linux.intel.com>
> Sent: Friday, September 27, 2019 5:53 AM
> To: Ajay Gupta <ajayg@nvidia.com>
> Cc: linux-usb@vger.kernel.org
> Subject: Re: [PATCH 00/14] usb: typec: UCSI driver overhaul
> 
> On Fri, Sep 27, 2019 at 12:13:57AM +0000, Ajay Gupta wrote:
> > Hi Heikki,
> > > -----Original Message-----
> > > From: Heikki Krogerus <heikki.krogerus@linux.intel.com>
> > > Sent: Thursday, September 26, 2019 3:07 AM
> > > To: Ajay Gupta <ajayg@nvidia.com>
> > > Cc: linux-usb@vger.kernel.org
> > > Subject: [PATCH 00/14] usb: typec: UCSI driver overhaul
> > >
> > > Hi Ajay,
> > >
> > > Here's the pretty much complete rewrite of the I/O handling that I
> > > was talking about. The first seven patches are not actually related
> > > to this stuff, but I'm including them here because the rest of the
> > > series is made on top of them. I'm including also that fix patch I
> > > send you earlier.
> > >
> > > After this it should be easier to handle quirks. My idea how to
> > > handle the multi-instance connector alt modes is that we "emulate"
> > > the PPM in ucsi_ccg.c in order to handle them, so ucsi.c is not touched at
> all.
> > >
> > > We can now get the connector alternate modes that the actual
> > > controller supplies during probe - before registering the ucsi
> > > interface - and squash all alt modes with the same SVID into one
> > > that we supply to the ucsi.c when ever it sends GET_ALTERNATE_MODES
> > > command. Also other alt mode commands like SET_NEW_CAM can have
> > > special processing in ucsi_ccg.c and ucsi_ccg.c alone. There should
> > > not be any problem with that anymore.
> > I took the changes and loaded on my GPU system and do not see altmode
> > devices under /sys/bus/typec/devices/*. Its empty.
> >
> > Below error is seen
> > "ucsi_ccg 4-0008: con1: failed to register alternate modes"
> >
> > ucsi_run_command() is returning -16.
> >
> > I will review the ccg changes and try to debug above issue.
> 
> I tested this series with the NVIDIA GPU PCI card that we have. The controller
> is reporting BUSY as responce to the GET_ALTERNATE_MODES command. I
> added a loop to ucsi_exec_command() where I read the CCI until the there is
> no UCSI_CCI_BUSY bit set, and it seems to work.
> 
> Here are both changes:
> 
> diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c index
> accf54d987ad..d70ee8006c34 100644
> --- a/drivers/usb/typec/ucsi/ucsi.c
> +++ b/drivers/usb/typec/ucsi/ucsi.c
> @@ -102,6 +102,7 @@ static int ucsi_read_error(struct ucsi *ucsi)
> 
>  static int ucsi_exec_command(struct ucsi *ucsi, u64 cmd)  {
> +       unsigned long timeout;
>         u32 cci;
>         int ret;
> 
> @@ -109,12 +110,15 @@ static int ucsi_exec_command(struct ucsi *ucsi, u64
> cmd)
>         if (ret)
>                 return ret;
> 
> -       ret = ucsi->ops->read(ucsi, UCSI_CCI, &cci, sizeof(cci));
> -       if (ret)
> -               return ret;
> +       timeout = jiffies + msecs_to_jiffies(UCSI_TIMEOUT_MS);
> +       do {
> +               if (time_is_before_jiffies(timeout))
> +                       return -ETIMEDOUT;
> 
> -       if (cci & UCSI_CCI_BUSY)
> -               return -EBUSY;
> +               ret = ucsi->ops->read(ucsi, UCSI_CCI, &cci, sizeof(cci));
> +               if (ret)
> +                       return ret;
> +       } while (cci & UCSI_CCI_BUSY);
> 
>         if (!(cci & UCSI_CCI_COMMAND_COMPLETE))
>                 return -EIO;
> @@ -330,7 +334,7 @@ static int ucsi_register_altmodes(struct
> ucsi_connector *con, u8 recipient)
>                 command |= UCSI_GET_ALTMODE_RECIPIENT(recipient);
>                 command |= UCSI_GET_ALTMODE_CONNECTOR_NUMBER(con-
> >num);
>                 command |= UCSI_GET_ALTMODE_OFFSET(i);
> -               command |= UCSI_GET_ALTMODE_NUM_ALTMODES(1); /* One at a
> time */
> +
>                 len = ucsi_run_command(con->ucsi, command, alt, sizeof(alt));
>                 if (len <= 0)
>                         return len;
> 
> Let me know if that works.
It works now for me too. I will review rest of changes anyways and also need
to update my change on altmode squashing.

Thanks
>nvpublic
> 
> thanks,
> 
> --
> heikki