Re: dwc3 PHY autosuspend feature

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi,

On Wed, Mar 06, 2013 at 08:14:34PM +0000, Paul Zimmerman wrote:
> > From: Felipe Balbi [mailto:balbi@xxxxxx]
> > Sent: Wednesday, March 06, 2013 12:11 PM
> > On Wed, Mar 06, 2013 at 07:22:33PM +0000, Paul Zimmerman wrote:
> > > > From: Felipe Balbi [mailto:balbi@xxxxxx]
> > > > Sent: Wednesday, March 06, 2013 2:36 AM
> > > > To: Paul Zimmerman; Kishon Vijay Abraham I; Linux USB Mailing List
> > > > Subject: dwc3 PHY autosuspend feature
> > > >
> > > > Hi Paul,
> > > >
> > > > Kishon has uncovered a possible 'situation' with dwc3 which we need to
> > > > clear up.
> > > >
> > > > According to your commit 802fde98 (usb: dwc3: support new revisions of
> > > > DWC3 core), we don't need to suspend/resume the PHYs at all with any
> > > > version > 1.94a, but what we observe is that SUSPHY bits are never set.
> > > >
> > > > What does that mean ? Would PHY autosuspend be an optional feature by
> > > > any chance ? Also, I can't find anywhere in the databook saying that
> > > > fiddling with SUSPHY should be avoided due to core IP controlling that
> > > > automatically.
> > > >
> > > > Can you enlighten us ?
> > >
> > > You still need to set the SusPHY bits, but you only need to do it once.
> > > After that, the core will take care of suspending/resuming the PHYs
> > > at the right time. You will not see the SusPHY bits changing, though.
> > >
> > > Actually, for GUSB2PHYCFG.SusPHY you can't just set it once. You
> > > need to clear it before any register operation that is marked in the
> > > databook with the note "If GUSB2PHYCFG[6] is set to '1', it must be
> > > set to '0' prior to issuing this command and may be set to '1' after
> > > the command completes". Then when the operation completes
> > > you can set it back to 1.
> > 
> > What if I set both to zero during driver probe ? Can I then completely
> > forget about them ?
> 
> Yes. But then you will be wasting power by having both PHYs always active.

fair enough, but wouldn't the core take care of the PHYs after probe() ?

-- 
balbi

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux