Re: FSUSB Register access

[Felipe Balbi <me@felipebalbi.com>]

On Tue, Nov 04, 2008 at 06:03:05PM -0700, York, Jeffrey-P56387 wrote:
> Balbi,
> 
> Thanks for your help.
> 
> See below for the answers.
> 
> -----Original Message-----
> From: Felipe Balbi [mailto:me@felipebalbi.com] 
> Sent: Tuesday, November 04, 2008 5:55 PM
> To: York, Jeffrey-P56387
> Cc: me@felipebalbi.com; linux-omap@vger.kernel.org
> Subject: Re: FSUSB Register access
> 
> Hi,
> 
> On Tue, Nov 04, 2008 at 05:38:28PM -0700, York, Jeffrey-P56387 wrote:
> > <JEFF SAYS:  I did this and I do see logging now.  It shows as
> expected
> > an external abort on non-linefetch when it tries to read from an FSUSB
> > register.  As far as I can tell from the TI spec, I should only have
> to
> > enable the 48 MHz clock...  And I know it is enabled via a read from
> the
> > CM_IDLEST_CKGEN register.>
> 
> hmm.. yeah, looks like it's clock issue. Do you call:
> 
> clk_get();
> clk_enable();
> 
> before reading the register ?? Although, if CM_IDLEST_CKGEN says it's
> on, it should be.
> 
> <JEFF SAYS:  I do when trying to request the "usb_fck" and "usb_l4_ick".
> Not sure what the code is doing for the 48 MHz clock.  That's done
> somewhere else in the Mobilinux kernel.  Given that the register says
> its on, I was assuming that clock was good.>

still, you were getting the "unable to enable clock usb_fck after 1000000
tries" message, right ? But then again, the register says it's good.

So might be something elsewhere. Could you share your code ? It'd be
much simpler to say what's going on by looking at your code.

-- 
balbi