[U-Boot-Users] Re: Using U-BOOT in CSB272 board

[Lonsn]

Hi, Tolunay
  Thank you for your suggestions.
> I am the maintainer of CSB272 port of U-Boot. I've done the port 
> based on technical info of the board which is not very reliable. 
> Unfortunately, Cogent generally is not very responsive to emails 
> regarding the ambiguous/outdated technical information about their 
> boards.
> Some information is extracted by using BDI2000 after uMon 
> intitialized the hardware. Some information is based on datasheet of 
> the parts.
> 
> This is actually a discontinued board. In our application here, we do 
> not use the Epson controller.  We have switched to using CSB472 
> now which is smaller and does not have the parts we do not use.
Since we have chosen this board and I have to solve this problem.
> On 8 Sep 2004 at 22:23, Lonsn wrote:
> 
> 
>>then bootm, kernel panic at init EPSON1356 Display Controller, so I 
>>reset U-BOOT, I download kernel again
> 
> 
> To me it sounds like your Epson1356 driver issue in Linux kernel. 
> You should debug the driver to find out what exactly within the driver 
> is causing the panic.
> 
> Check the assumptions made by the driver. The particular driver 
> may be assuming that bootloader is doing a particular initialization 
> however a well written driver should not assume these and re-
> inititialize the hardware suitably. It could also be a race condition or 
> an uninitialized variable.
I will try to do it.

> 
>>Should I pay attention to SDRAM init sequence in U-BOOT as Wolfgang has 
>>said many times?
> 
> 
> As far as I know SDRAM initialization for CSB272 is correct. The 
> information supplied by Cogent in the tech. documentation is for 
> 32MB SDRAM (older) version of the board but your and my board 
> are 64MB.
I also compared the sdram init code between U-BOOT and uMon, there aren't
much differences except uMon delay 10ms after enable the memory bank at 
last.
So I think it's not the question of sdram init.

I will check the epson driver in kernel.