[coreboot] Flashrom support for EPIA-N(L) - patch for review

Luc Verhaegen libv at skynet.be
Fri Jun 19 12:31:53 CEST 2009


On Thu, Jun 18, 2009 at 01:05:40PM +0100, Harrison, Jon (SELEX GALILEO, UK) wrote:
> I didn't mean to suggest that there is anything wroung with flashrom
> itself, there is either some issue with reading the coreboot tables from
> the coreboot flash after booting from the stock flash (which I have to
> do all of the time at the moment) or my coreboot image is broken in some
> way ... I've not had the hex editor out (yet) as I'm working more on
> getting the initial CN400 work done and personally I am quite happy with
> having to -m at this stage of development.
> 
> I was keener to get the flashrom functionality that does work out there,
> so that there is a minimum level of functionality available for anyone
> else out there that wants to start contributing to the development
> effort on this MB.
> 
> Of course this may all turn to be linked ..........

> On Thu, Jun 18, 2009 at 11:36:49AM +0100, Harrison, Jon (SELEX GALILEO,
> UK) wrote:
> > 
> > Auto detection doesn't seem to be working properly and at this stage
> I'm
> > breaking things so much that the -m switch is necessary for reliable
> > operation.
> 
> Why doesn't it seem to work properly?
> 
> Luc Verhaegen.

Ok. communication error here.

You will only have the coreboot tables when coreboot has succesfully 
booted and your config has USE_OPTION_TABLE.

But... like with many of the other epia boards, you will never disable 
rom access from coreboot and therefor flashrom does not need to enable
specifically for this board.

So:

* original bios has an extra write disable, but also has subsystem ids.
  here flashrom should use the subsystem ids to autodetect this board.
* coreboot, from reset, should not touch set the board specific 
  disable. So flashrom doesn't need to know or touch this.
* when coreboot is ran after the normal bios, the subsystem ids usually
  still are there, and so is the write protection (if you're plugging 
  chips). Flashrom then should still autodetect through subsystem ids 
  still.

So, do you really need flashrom to run the board specific enable when 
you've booted coreboot? If not, remove the coreboot ids, and also make 
it impossible for people to randomly try board enables (through the 
command line) and accidentally hit this code.

I am quite happy with the rest of the code though.

Luc Verhaegen.




More information about the coreboot mailing list