[U-Boot-Users] U-boot environment in Xilinx System ACE?

[U-Boot-Users] U-boot environment in Xilinx System ACE?

[Keith J Outwater]

Greetings everyone - 
I am working on a port of U-Boot to a custom Xilinx Virtex2 Pro FPGA based 
board which uses the PPC405 inside the FPGA.  The plan is to base the port 
on the existing ml300 port.

The nature of the design requires that the Xilinx System ACE compact flash 
card be the only nonvolatile memory on the board - no plain old FLASH 
memory or I2C EEPROM.  Therefore, I would like to store the U-Boot 
environment as a file on the compact flash card.

I have checked the latest CVS sources and it appears that no one has 
implemented this yet.

Before I take a crack at it, is anyone planning to do this or in the 
process of doing it?  Does anyone have any advice or caveats?


Thanks,
Keith

[U-Boot-Users] U-boot environment in Xilinx System ACE?

[Wolfgang Denk]

Dear Keith,

in message <OFD86B567B.6DB5DBFC-ON07257037.005D553D-07257037.005E4D94@mck.us.ray.com> you wrote:
>
> The nature of the design requires that the Xilinx System ACE compact flash 
> card be the only nonvolatile memory on the board - no plain old FLASH 
> memory or I2C EEPROM.  Therefore, I would like to store the U-Boot 
> environment as a file on the compact flash card.

This has been discussed before...

> I have checked the latest CVS sources and it appears that no one has 
> implemented this yet.

Right.

> Before I take a crack at it, is anyone planning to do this or in the 
> process of doing it?  Does anyone have any advice or caveats?

This is not trivial to handle. The environment needs to  be  accessed
very,  very  early  in  initialization - to read the console baudrate
when starting the console driver. IDE /  CompactFlash  on  the  other
hand  gets  initialized  much,  much later, long after relocating the
whole stuff to RAM.

I don't see how these two situations mix easily.  Can  you  attach  a
EEPROM via I2C or so?

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Life would be so much easier if we could  just  look  at  the  source
code.                                                   -- Dave Olson

[U-Boot-Users] Re: U-boot environment in Xilinx System ACE?

[Keith J Outwater]

Hi Wolfgang - 

wd at denx.de wrote on 07/07/2005 01:00:41 PM:
> Dear Keith,
> 
> in message <OFD86B567B.6DB5DBFC-ON07257037.005D553D-07257037.
> 005E4D94 at mck.us.ray.com> you wrote:
> >
> > The nature of the design requires that the Xilinx System ACE compact 
flash 
> > card be the only nonvolatile memory on the board - no plain old FLASH 
> > memory or I2C EEPROM.  Therefore, I would like to store the U-Boot 
> > environment as a file on the compact flash card.
> 
> This has been discussed before...

I thought as much.  I just makes too much sense :)

> 
> > I have checked the latest CVS sources and it appears that no one has 
> > implemented this yet.
> 
> Right.
> 
> > Before I take a crack at it, is anyone planning to do this or in the 
> > process of doing it?  Does anyone have any advice or caveats?
> 
> This is not trivial to handle. The environment needs to  be  accessed
> very,  very  early  in  initialization - to read the console baudrate
> when starting the console driver. IDE /  CompactFlash  on  the  other
> hand  gets  initialized  much,  much later, long after relocating the
> whole stuff to RAM.
> 
> I don't see how these two situations mix easily.  Can  you  attach  a
> EEPROM via I2C or so?

I would prefer to do it that way, but the hardware must not be able to
store any data on-board.  All nonvolatile storage must be "removable"
as in physical removal.

I did remember that the u-boot environment was read very early and it
sounds like the restricted runtime environment was the reason that
storing then environment in compact flash was not implemented.

If the problem is RAM, then I can solve that by dedicating some
internal RAM in the FPGA for stack and data.  Could we also use
the PPC405 caches?

> 
> Best regards,
> 
> Wolfgang Denk

[U-Boot-Users] Re: U-boot environment in Xilinx System ACE?

[Wolfgang Denk]

Dear Keith,

in message <OF203F9449.99AEB2AF-ON07257037.006EF267-07257037.0071F9E5@mck.us.ray.com> you wrote:
> 
> I did remember that the u-boot environment was read very early and it
> sounds like the restricted runtime environment was the reason that
> storing then environment in compact flash was not implemented.

Right, this is the big problem.

> If the problem is RAM, then I can solve that by dedicating some
> internal RAM in the FPGA for stack and data.  Could we also use
> the PPC405 caches?

It's not just having some memory, but all the other restrictions while
running from flash: no writable data segment, not initialized BSS.

Check the code where CFG_ENV_IS_NOWHERE is used; use a static  confi-
guration for the console baudrate etc. This can get you started...

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
The man on tops walks a lonely street;  the  "chain"  of  command  is
often a noose.

[U-Boot-Users] Re: U-boot environment in Xilinx System ACE?

[Peter Ryser]

Keith,

have a look at XAPP542 
(http://www.xilinx.com/bvdocs/appnotes/xapp542.pdf). We added support 
for System ACE CF and IIC and are working on getting a patch completed 
that goes back out into the U-Boot tree.

- Peter


Keith J Outwater wrote:

>Hi Wolfgang - 
>
>wd at denx.de wrote on 07/07/2005 01:00:41 PM:
>  
>
>>Dear Keith,
>>
>>in message <OFD86B567B.6DB5DBFC-ON07257037.005D553D-07257037.
>>005E4D94 at mck.us.ray.com> you wrote:
>>    
>>
>>>The nature of the design requires that the Xilinx System ACE compact 
>>>      
>>>
>flash 
>  
>
>>>card be the only nonvolatile memory on the board - no plain old FLASH 
>>>memory or I2C EEPROM.  Therefore, I would like to store the U-Boot 
>>>environment as a file on the compact flash card.
>>>      
>>>
>>This has been discussed before...
>>    
>>
>
>I thought as much.  I just makes too much sense :)
>
>  
>
>>>I have checked the latest CVS sources and it appears that no one has 
>>>implemented this yet.
>>>      
>>>
>>Right.
>>
>>    
>>
>>>Before I take a crack at it, is anyone planning to do this or in the 
>>>process of doing it?  Does anyone have any advice or caveats?
>>>      
>>>
>>This is not trivial to handle. The environment needs to  be  accessed
>>very,  very  early  in  initialization - to read the console baudrate
>>when starting the console driver. IDE /  CompactFlash  on  the  other
>>hand  gets  initialized  much,  much later, long after relocating the
>>whole stuff to RAM.
>>
>>I don't see how these two situations mix easily.  Can  you  attach  a
>>EEPROM via I2C or so?
>>    
>>
>
>I would prefer to do it that way, but the hardware must not be able to
>store any data on-board.  All nonvolatile storage must be "removable"
>as in physical removal.
>
>I did remember that the u-boot environment was read very early and it
>sounds like the restricted runtime environment was the reason that
>storing then environment in compact flash was not implemented.
>
>If the problem is RAM, then I can solve that by dedicating some
>internal RAM in the FPGA for stack and data.  Could we also use
>the PPC405 caches?
>
>  
>
>>Best regards,
>>
>>Wolfgang Denk
>>    
>>
>
>
>
>-------------------------------------------------------
>This SF.Net email is sponsored by the 'Do More With Dual!' webinar happening
>July 14 at 8am PDT/11am EDT. We invite you to explore the latest in dual
>core and dual graphics technology at this free one hour event hosted by HP,
>AMD, and NVIDIA.  To register visit http://www.hp.com/go/dualwebinar
>_______________________________________________
>U-Boot-Users mailing list
>U-Boot-Users at lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/u-boot-users
>
>
>  
>