Hi,
> -----Original Message-----
> From: Dasgupta, Romit
> Sent: Thursday, November 26, 2009 2:33 PM
> To: Gupta, Ajay Kumar
> Cc: linux-...@vger.kernel.org; linux-omap@vger.kernel.org;
> felipe.ba...@nokia.com; Gadiyar, Anand
> Subject: Re: [PATCH 1/2 v2] musb: Add context save and restore support
>
> > --- a/drivers/usb/musb/musb_core.c
> > +++ b/drivers/usb/musb/musb_core.c
> > @@ -2167,21 +2167,163 @@ static int __devexit musb_remove(struct
> platform_device *pdev)
> >
> > #ifdef CONFIG_PM
> >
> > +static struct musb_context_registers musb_context;
> > +
> > +void musb_save_context(void __iomem *musb_base)
> > +{
> > + int i;
> > +
> > + musb_context.faddr = musb_readb(musb_base, MUSB_FADDR);
> > + musb_context.power = musb_readb(musb_base, MUSB_POWER);
> > + musb_context.intrtxe = musb_readw(musb_base, MUSB_INTRTXE);
> > + musb_context.intrrxe = musb_readw(musb_base, MUSB_INTRRXE);
> > + musb_context.intrusbe = musb_readb(musb_base, MUSB_INTRUSBE);
> > + musb_context.frame = musb_readw(musb_base, MUSB_FRAME);
> Not necessary for gadget
> > + musb_context.testmode = musb_readb(musb_base, MUSB_TESTMODE);
> Not sure if it is necessary for gadget.
Right!
..so we can all #ifdefs to save/restore gadget or host only registers.
> > + musb_context.index = musb_readb(musb_base, MUSB_INDEX);
>
>
<snip>
> > + musb_read_txfifosz(musb_base);
> > + musb_context.index_regs[i].rxfifosz =
> > + musb_read_rxfifosz(musb_base);
>
> If MUSB_CONFIGDATA_DYNFIFO is not set then saving FIFO address and sizes
> are probably not necessary. Not sure if they are detrimental as well.
We can add it within DYNFIFO check.
> > +
> > + musb_context.index_regs[i].txfunaddr =
> > + musb_read_txfunaddr(musb_base, i);
> > + musb_context.index_regs[i].txhubaddr =
<snip.>
> > */
> > + reg = musb_readb(musb->mregs, MUSB_POWER);
> > + reg &= ~MUSB_POWER_SOFTCONN;
> > + musb_writeb(musb->mregs, MUSB_POWER, reg);
> After this softdisconnect when we subsequently get reconnected followed by
> a reset we could have changed our speed (LS,FS, HS).
I didn't get it completely. I think as a peripheral USB speed will remain
same after resume for the gadget driver it was before suspend?
If we intend to change the gadget driver then we may have support for a
different speed but it would anyways happen after resume is completed.
Regards,
Ajay
> In your resume
> routine we are writing back our previous speed. I am not sure here but is
> it not possible that we have reset the controller and then restored back
> the old speed overwriting the new speed?
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
