Re: [PATCH 2/2] rtc-tps65910: enable RTC power domain on initialization

[Andrew Morton <akpm@linux-foundation.org>]

On Thu, 29 Nov 2012 23:11:37 +0000
"Kim, Milo" <Milo.Kim@ti.com> wrote:

> Hi Venu
> 
> > >  Enabling RTC HW block depends on the default value of TPS65910
> > register.
> > >  In some mode, RTC block is disabled by default.(eg. AM3517
> > Craneboard)
> > >  In this case, RTC_PWDN(RTC power down) bit should be cleared to
> > enable
> > >  the RTC HW block.
> > 
> > From the description of RTC_PWDN bit of DEVCTRL_REG in TPS65910 data
> > sheet
> > it is very evident that the default value of RTC_PWDN is 0.
> 
> According to the datasheet(http://www.ti.com/lit/ds/swcs046q/swcs046q.pdf),
> the default value RTC_PWDN is 1 which means power down.
> 
> The default values are loaded from the EEPROM with BOOT_MODE 0,1 pin connection.
> The RTC is disabled by default when BOOT_MODE = 00.
> The Craneboard has the BOOT_MODE 00.
> 
> You may have other EEPROM settings,
> however the official datasheet shows the RTC block is off by default.
> 
> Could you check the silicon version number? (0x80 register - JTAGVERNUM_REG)
> In my case, the read value is 0x01.
> 
> > Probably on "AM3517 Craneboard", some code is running prior to the RTC
> > driver
> > which might be writing 1 on to this bit. IMO you must disable that
> > write operation
> > instead of just writing default value into a register.
> 
> Thank you for your comment.
> I've read the DEVCTRL register (0x3F) in mfd tps65910 driver
> as soon as the regmap registration is done, however RTC_PWDN is always 1.
> That means the RTC is disabled by default.
> 

I've merged this patch into -mm for 3.8 along with a note-to-self that
there might still be open issues.  Venu, please let us know if the
above settles things?