[U-Boot] [PATCH] p1020rdb-pd: platform support

[Scott Wood]

On 06/27/2013 08:55:26 PM, Zhang Haijun-B42677 wrote:
> > -----Original Message-----
> > From: Wood Scott-B07421
> > Sent: Friday, June 28, 2013 4:10 AM
> > To: Zhang Haijun-B42677
> > Cc: Wood Scott-B07421; Fleming Andy-AFLEMING; u-boot at lists.denx.de;  
> Huang
> > Changming-R66093; sun york-R58495
> > Subject: Re: [U-Boot] [PATCH] p1020rdb-pd: platform support
> >
> > On 06/27/2013 03:58:57 AM, Zhang Haijun-B42677 wrote:
> > > Hi, Scott
> > >
> > > Sorry to reply to so late.
> > > Pls see my reply below.
> > >
> > >
> > > Regards & Thanks
> > > Haijun.
> > >
> > > > -----Original Message-----
> > > > From: Wood Scott-B07421
> > > > Sent: Wednesday, May 22, 2013 6:24 AM
> > > > To: Zhang Haijun-B42677
> > > > Cc: Fleming Andy-AFLEMING; u-boot at lists.denx.de; Huang
> > > Changming-R66093;
> > > > Zhang Haijun-B42677; Zhang Haijun-B42677; sun york-R58495
> > > > Subject: Re: [U-Boot] [PATCH] p1020rdb-pd: platform support
> > > >
> > > > On 04/28/2013 01:17:34 AM, Haijun.Zhang wrote:
> > > > > From: "Haijun.Zhang" <haijun.zhang at freescale.com>
> > > > >
> > > > > Add platform support for p1020rdb-pd
> > > >
> > > > Explain what is different about "-pd" and why this new revision
> > > requires
> > > > a separate compile-time target.
> > > >
> > > [Haijun Wrote:] We use P1020RDB-PD to instead of P1020RDB-PC board
> > > later on.
> > > They are all supported. Nand, DDR, Nor and Spi Flash are different
> > > between them.
> > > So I thinks they should be two separate board as P1020UTM and  
> P1020MBG
> > > do.
> >
> > Put this information in the changelog.
> [Haijun Wrote:] change log you mean?

The git commit message.

> > > > If it's due to raw DDR config, say so in the changelog.  Is  
> there
> > > any way
> > > > we could read a revision register to select the proper DDR  
> config at
> > > > runtime?
> > > [Haijun Wrote:] Now we read DDR config from SPD fist, if not found
> > > read from config file.
> >
> > Is there any reason why SPD would not be found?
> [Haijun Wrote:] If someone erase the I2C area(SPD EEPROM) by mistake  
> or change by other possible reason.
> They will need to read from local config file to make sure uboot can  
> working.

I don't think we normally provide backup DDR config for such a case.   
There are many other, easier ways the user can accidentally screw up  
the system (e.g. they could erase U-Boot).  And it would get little to  
no testing, so likely bitrot or otherwise not work (e.g. the board  
starts shipping with different RAM, or the settings were never quite  
right to begin with).

> > > > > +#if (defined(CONFIG_P1020RDB) || defined(CONFIG_P1020RDB_PD))
> > > > >  #if defined(CONFIG_P1020RDB)
> > > > >  #define CONFIG_BOARDNAME "P1020RDB-PC"
> > > > > +#elif defined(CONFIG_P1020RDB_PD) #define CONFIG_BOARDNAME
> > > > > +"P1020RDB-PD"
> > > > > +#endif
> > > >
> > > > CONFIG_P1020RDB is not a good name to mean specifically the -PC
> > > revision.
> > > > Change it so that CONFIG_P1020RDB is set for all revisions of
> > > P1020RDB,
> > > > and introduce CONFIG_P1020RDB-PC for PC-specific things.
> >
> > Could you reply to this one?
> >
> > ...and of course I meant CONFIG_P1020RDB_PC rather than  
> CONFIG_P1020RDB-
> > PC. :-)
> [Haijun Wrote:] Oh, yes I agree. It a little strange to have "-PD",  
> but this patch was aim at add new board p1020rdb-pd board. If need  
> i'll send another patch to change the config name for PC,RDB and PD.
> I had send another patch(Mean v2 of this patch), pls give some advice.

Respin that patch so that CONFIG_P1020RDB is set for all revisions of  
p1020rdb, and either CONFIG_P1020RDB_PC or CONFIG_P1020RDB_PD is  
specified to indicated the revision.

-SCott