[U-Boot-Users] [PATCH] New NAND subsystem: mtd like commands support jffs2 and bad blocks

[Stefan Roese]

Hi Guido,

On Sunday 08 October 2006 15:49, Guido Classen wrote:
> >> +		if ( !opts->scrub && bbtest ) {
> >
> >No spaces after opening brace and before closing brace.
>
> Now I finally know what you meant. AFAIK the round ones are called
> parenthesis, not brace! Braces are the curly and square onse.

Thanks for clarifying this.

> >Now it compiles clean for my testplatform. Thanks.
> >
> >As it seems you have set the "quiet" option for all commands. Because of
> >this we don't see the skipping of bad blocks message for example. So
> > please disable the quiet option.
>
> You're right. I set the "quiet" option for all commands. I have done some
> testing again and get an impression that the output of progress messages on
> the serial console takes too much time. It seems that they slow down the
> operations especially on small page devices (four times more messages).
> Hence I think the best solution would be to make them switchable on command
> line, so they can be switched on for testing and off for normal operation.

Hmmm. Not sure here. Perhaps you could try to enhance the output so that it 
there printend that much output. And I also noticed that the output is not 
correct:

=> nand erase 0 1000000

NAND erase: device 0 offset 0x0, size 0x1000000
Erasing 128 Kibyte at fe0000 --  1% complete.
OK

The 1% is for the complete FLASH size, but I was erasing just a small part of 
it. I would really like to see 100%, even when erasing a small part.

So this is another reason to modify the output of the routines, at least of 
the erase routine.

> Maybe a "quiet" option at command line for the related commands (like
> "clean" option) will be the right way to do it. What do you think about
> this?
>
> I will rework the patch with your suggestions as soon as provisions to
> switch on/off the "quiet" option and resubmit it the next days.

Please see my comment above. I don't want the commands to become too complex 
with lots of options (they are allready complex enough). So my suggestion is 
to change the code that it doesn't slow down the operation anymore.

Best regards,
Stefan