[U-Boot] arch/powerpc/lib/board.c:get_effective_memsize() for 4GB+ systems

Robert Thorhuus robert.thorhuus at ericsson.com
Fri Sep 21 15:56:27 CEST 2012


> -----Original Message-----
> From: Timur Tabi [mailto:timur at freescale.com] 
> Sent: den 21 september 2012 15:29
> To: Robert Thorhuus
> Cc: Kumar Gala; u-boot at lists.denx.de
> Subject: Re: [U-Boot] 
> arch/powerpc/lib/board.c:get_effective_memsize() for 4GB+ systems
> 
> Robert Thorhuus wrote:
> > No I will not be using much memory at all for 
> functionality. But the 
> > memory needs to be tested. What is your proposal for that then?
> 
> We have that already.  Look at CONFIG_SYS_POST_MEMORY.  It 
> uses sliding 2GB TLBs to test all of DDR.

Ok. I'll look at that. Thanks.

> 
> > And I see U-Boot as the first software place for test and 
> debug. Maybe I want to read out RAM contents?
> 
> It's a boot loader, not a testing platform.

Sorry. But this is were I disagree with you. Of course its prime function is to boot an operating system. And usually in a desktop environment that is what you need. But if you have newly developed hardware it is very seldom everything works all the time. There are a lot of debugging hardware sessions before you have stable hardware. So in embedded systems I would say U-Boot can very well be a primary choice of testing platform in the beginning of a development. Mainly because sucessfully booting U-Boot demands less of the hardware than sucessfully boot an OS. Say you have a NOR and 1 MB L3 cache but no DDR3, U-Boot is set, OS not.

And what test platform would you suggest? And what if your extended testing needs to be a viable option at every boot and that you have boot time requirements?

If you have suggestions I will be happy because I'd like to move my testing from U-Boot for several reasons but currently the restraints are tough.

BR


More information about the U-Boot mailing list