[ECOS] why only this address in pci

Stewart Heitmann stewart@heitmann.id.au
Thu Mar 2 22:32:00 GMT 2006

Hi Sumanth,
you are not alone. I am having similar problems on a board based on
the IXDPG425 (Montejade) eval board.
Unfortunately I have no answers and am only getting very odd
results that I cannot make sense of.

At first I was bombing out at the next line from you, at
   HAL_PCI_CFG_WRITE_UINT32(0, 0, CYG_PCI_CFG_BAR_2, 0x02000000); 
This was using the ixdpg425 redboot ROM configuration (without
network NPE) shipped with redboot-intel-ixp4xx-050425.

Thinking I had a problem with the PCI bus (my board has the PCI
bus disabled in hardware) I hacked out most of the platform
specific code relating to the PCI bus.
During this work I did manage to see some PCI debug messages
on the console, so I thought I was making progress.

Since then I have lost the console again and am getting
even weirder results. Moreover, the weird stuff happens before
any of the code that I hacked gets executed!

I am now finding the call to hal_xscale_core_init (from ixp425_misc.c)
doesn't jump t the correct address.
	5000ac0c:   eb004ffa   bl 5001ebfc <hal_xscale_core_init>
It ends up jumping to 0x5001_ac14 instead of the intended 0x5000_ebfc
address!!!! I have no idea why that happens!
Stewart Heitmann

On Thursday 02 March 2006 10:51 pm, sumanth wrote:
> Hi all, 
>              We are porting ecos for ixp425 evalution board, we are using
> the grg 
>              for the purpose , 
>               now we are able to go up to plf_hardware_init 
>               we are strucking up at cyg_hal_plf_pci_init 
> in pci  it is using this address  are there any specific reasons 
> *IXP425_PCI_PCIMEMBASE = 0x48494A4B 
> *IXP425_PCI_AHBMEMBASE = 0x00010203 
> i am strucking at 
> HAL_PCI_CFG_WRITE_UINT32(0, 0, CYG_PCI_CFG_BAR_1, 0x01000000); 
> can any one help me what is the reason or any pointers at this stage of
> porting 
> thanks & regards, 
> Sumanth 

Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

More information about the Ecos-discuss mailing list