[U-Boot] [PATCH] mpc8xxx: lc_common_dimm_params: make less verbose

[York Sun]

On Thu, 2011-07-28 at 20:45 +0200, Wolfgang Denk wrote:
> Dear York Sun,
> 
> In message <1311875176.29459.14.camel at oslab-l1> you wrote:
> >
> > > -				printf("Detected RDIMM %s\n",
> > > +				debug("Detected RDIMM %s\n",
> > >  					dimm_params[i].mpart);
> > >  			} else {
> > >  				temp2 = 1;
> > > -				printf("Detected UDIMM %s\n",
> > > +				debug("Detected UDIMM %s\n",
> > >  					dimm_params[i].mpart);
> > >  			}
> > >  		}
> > 
> > NAK.
> 
> NAK the NAK.
> 
> > We need to log module part number for testing and verification,
> 
> debug() is exactly what is intended to be used for testing.

I didn't mean testing the U-boot. I meant to test the board and overall
software. As we keep the test logs for later comparison, it is helpful
to keep the DDR module information.

> 
> I don't know what exactly you mean by "verification", but I guess that
> covers it, too.
> 
> Keeping this as prontf() corrupts the format of the boot messages, and
> adds to the boot time without real value.

I don't agree one line with the module information corrupts the boot
messages. Do we have fixed format? Am I missing something?

I didn't count the time but don't think the one line message will add
much to boot time. And it does have real value even you don't use it.

> 
> > especially when comparing with different parts. It is also helpful for
> > support, without asking customers to recompile to enable debugging.
> 
> Then feel free to add a custom command to display this specific
> information, but keep it off the regular, unconditionally printed boot
> messages.

That's another way to do it. But we don't have to add a new command if a
simple message does the job.

York