Re: [microblaze-uclinux] Problem with 2nd stage bootloader (u-boot)

[John Williams <jwilliams@xxxxxxxxxxxxxx>]

Hi Jarno,

Radde, Jarno wrote:

> > If you read from the address (using u-boot's memory read 
> > command or xmd) do you get 0xffffffff at flash addresses (ie 
> > blank flash data)?
> This is what I get:
>
> U-Boot> md 0x28000000
> 28000000: ffff0008 aa99ffff 2000ffff 3000ffff    ........ ...0...
> 28000010: 0000ffff 2000ffff 2000ffff 3001ffff    .... ... ...0...
> 28000020: 0004ffff 3001ffff 0167ffff 3000ffff    ....0....g..0...
> 28000030: 0000ffff 2000ffff 3000ffff 0000ffff    .... ...0.......
> 28000040: 3000ffff 0000ffff 2000ffff 2000ffff    0....... ... ...
> 28000050: 2000ffff 2000ffff 2000ffff 2000ffff     ... ... ... ...

So, it looks like there's a problem with one of the flash chips - either 
in how it is wired up to the FPGA, how it is being accessed by software, 
or maybe physically with the chip (not very likely I don't think).

> The patterns of the upper 16 bits are familiar to me, thats 
> exactly what I read every time when I tried to write my own 
> bootloader. The lower 16 bits were always correct (half words of 
> the data I wrote to flash before).
>
> I think one of the chips is broken or something like that.

To be honest it seems pretty unlikely.  Or at least, there are many more 
likely options to rule out first!


> Ok, then I'll try. If it should not work, do you maybe have an 
> other idea what I could do? I want to do the following as some 
> kind of investigation:
>
> 1. Do something with the hardware using software running on 
> uCLinux (in eg text on the LCD, blinking leds, etc.)
> 2. Do the same thing but this time in hardware (custom IP)
>
> I want to generate both applications using Impulse C and split it 
> to two processes. One is generating sending data and the other 
> one receives it and does something with the hardware according to 
> the data it received. In the second case one of the process is a 
> hardware process.
>
> The "doing something in hardware" part of if should be as simple 
> as possible so if it is possible to avoid writing an own device 
> driver it would save me some time.

Sure - just remember that doing something trivial like blinking an LED 
is a good first step, not a good application to demonstrate some kind of 
conclusion about HW vs SW!

Regards,

John
___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/