RE: [PATCH v2 2/2] misc: Add iop driver for Sunplus SP7021

[Tony Huang 黃懷厚]

Dear Arnd:

> -----Original Message-----
> From: Arnd Bergmann <arnd@xxxxxxxx>
> Sent: Monday, December 6, 2021 4:04 PM
> To: Tony Huang 黃懷厚 <tony.huang@xxxxxxxxxxx>
> Cc: Arnd Bergmann <arnd@xxxxxxxx>; Tony Huang
> <tonyhuang.sunplus@xxxxxxxxx>; Derek Kiernan <derek.kiernan@xxxxxxxxxx>;
> Dragan Cvetic <dragan.cvetic@xxxxxxxxxx>; gregkh
> <gregkh@xxxxxxxxxxxxxxxxxxx>; Linux Kernel Mailing List
> <linux-kernel@xxxxxxxxxxxxxxx>; Rob Herring <robh+dt@xxxxxxxxxx>; DTML
> <devicetree@xxxxxxxxxxxxxxx>; Wells Lu 呂芳騰 <wells.lu@xxxxxxxxxxx>
> Subject: Re: [PATCH v2 2/2] misc: Add iop driver for Sunplus SP7021
> 
> On Mon, Dec 6, 2021 at 4:42 AM Tony Huang 黃懷厚
> <tony.huang@xxxxxxxxxxx> wrote:
> > > Subject: Re: [PATCH v2 2/2] misc: Add iop driver for Sunplus SP7021
> > > On Fri, Dec 3, 2021 at 4:48 AM Tony Huang
> <tonyhuang.sunplus@xxxxxxxxx> wrote:
> > > > +
> > > > +static const struct file_operations sp_iop_fops = {
> > > > +       .owner                  = THIS_MODULE,
> > > > +       .open                   = sp_iop_open,
> > > > +       .read                   = sp_iop_read,
> > > > +       .write                  = sp_iop_write,
> > > > +       .release                = sp_iop_release,
> > > > +};
> > >
> > > This does nothing because all the callbacks are empty. You removed
> > > the inappropriate user space interfaces as I asked you to, but if
> > > there is no way for either kernel or user space to interact with the
> > > hardware, I don't see a point in merging the driver until you add a new
> interface that is usable.
> > >
> >
> > I will modify sp_iop_read() to monitor IOP mailbox data.
> 
> Why is this a useful interface to have? If this is only for debugging, a tracepoint
> may be more useful than a full character device.
> 

Yes, it is for debugging.

> > > Something looks wrong here, maybe reread the documentation for
> > > runtime power management to find a way of putting the device into
> > > low-power mode when it is unused.
> > >
> >
> > When the poweroff command is executed, the run
> > sp_iop_platform_driver_poweroff(void)
> > function will enter the standby mode. The power off will be executed.
> > In the system, IOP can continue to work when other modules in the
> > system enter standby / power down modes to monitor whether the system
> wakes up through RTC.
> 
> Ok, in that case you can probably just remove the empty callback functions.
> 
>          Arnd