[U-Boot] mx6: consolidating and extending anadig/analog/anatop register defs

[Marek Vasut]

> Hi Marek,
> 
> On 02/14/2012 01:37 PM, Marek Vasut wrote:
> > > Hi, Wolfgang,
> > > 
> > > > -----Original Message-----
> > > > From: u-boot-bounces at lists.denx.de
> > > > [mailto:u-boot-bounces at lists.denx.de] On Behalf Of Wolfgang
> > > > Grandegger Sent: Monday, February 13, 2012 8:08 PM
> > > > To: U-Boot
> > > > Subject: [U-Boot] mx6: consolidating and extending
> > > > anadig/analog/anatop register defs
> > > > 
> > > > Hello,
> > > > 
> > > > for the USB support on MX6Q, I'm trying to consolidate and extend
> > > > the register definitions for the ANADIG registers. Unfortunately,
> > > > three different names are used for the same registers: ANADIG,
> > > > analog and ANATOP. Any idea what the names do stand for and what
> > > > would be the most appropriate one. Apart from that, we should have
> > > > a separate structure and header file for these registers.
> > > > Currently, they are added to the CCM regs.
> > > 
> > > The register definition in the CCM according to the RM as followings:
> > > Addresses: CCM_analog_PLL_SYS is 0h base + 0h offset = 0000_0000h
> > > CCM_analog_PLL_SYS_SET is 0h base + 4h offset = 0000_0004h
> > > CCM_analog_PLL_SYS_CLR is 0h base + 8h offset = 0000_0008h
> > > CCM_analog_PLL_SYS_TOG is 0h base + Ch offset = 0000_000Ch
> > > 
> > > Thus, I put the register definition into the CCM parts.
> > 
> > This again resembles MX28 ... damn, can't we share some parts of MX28
> > for MX6Q ?
> > 
> > Fabio, I think I'll need one too, this is getting weird ;-)
> 
> Unfortunately, I did not get hold of a RM yet :(.

RM? Who's that?

 It's still under NDA,
> I believe. Therefore I will postpone my "consolidation" activities for
> the time being.

Sorry, I don't intent to hinder your efforts, I'm just seeing a pattern :)

The USB support patch I sent does work, as Dirk
> confirmed, and therefore I suggest to apply it as intermediate solution.

Well ... All right then. Though my sense for perfect solution is hurting a lot ;-)
> 
> Wolfgang.