[flashrom] arm, chromebook and pci.h

[Stefan Tauner]

On Sun, 11 Jan 2015 20:43:23 -0800
bruce m beach <brucembeach at gmail.com> wrote:

> Here are some straightforward results
> 
> 1-----------------------------------------
> )flashrom -p linux_spi:dev=/dev/spidev1.0
> flashrom v0.9.7-r1711 on Linux 3.4.0 (armv7l)
> flashrom is free software, get the source code at
> http://www.flashrom.org
> 
> Calibrating delay loop... OK.
> Found Winbond flash chip "W25Q32.W" (4096 kB, SPI) on linux_spi.
> No operations were specified.
> )
> 
> 2-----------------------------------------
> )flashrom -r original_flash -p linux_spi:dev=/dev/spidev1.0
> flashrom v0.9.7-r1711 on Linux 3.4.0 (armv7l)
> flashrom is free software, get the source code at
> http://www.flashrom.org
> 
> Calibrating delay loop... OK.
> Found Winbond flash chip "W25Q32.W" (4096 kB, SPI) on linux_spi.
> Block protection could not be disabled!
> Reading flash... done.
> )

That's probably due to the hardware write protection enabled... as you
probably know the chrombooks have a screw or switch that does that.
There is public documentation about this for most/all chromebooks...

> >> Is the Chromebook not supported by this version of flashrom
> >That's exactly the case... google has forked flashrom in
> >chromiumos and I strongly recommend using their version at the
> >moment on those systems.
> 
> I'm aware of the chromiumos fork and spent an afternoon reading
> everything I could and decided flashrom was what I wanted (as
> apposed to flashrom_chrome say). I run a totally stock linux
> system. It is not an Ubuntu, debian or anything system. All the
> packages including yours I build out of the box with no
> modifications and no reference at all to the fact that this is an
> arm system. It is completely transparent. It is reasonable expect
> that even the low level operations work properly. It is all taken
> care of by the kernel.

Not necessarily. Flashrom uses interfaces that the kernel does not
provide but that reside directly in the hardware. (That's also the
reason why it does work on non-Linux systems.)

> Now far as getting flashrom_chrome, I
> don't have it. I have long since blown chrome away (I store my
> source files where it used to be) and don't want anything to do
> with it. (my hostname is system_zinc) Furthermore I don't think I
> could even build it on this machine without a great deal of work.

You don't need chromeos just to build chromoiumos' flashrom. It should
build just like vanilla flashrom (apart from some minor bugs like wrong
includes etc. that I am sure you can fix). I just build current
chromiumos' master of flashrom and it runs just fine.

> > a *truly* horrible thought is the inevitable work to be done to
> > converge the two flashroms again but that should not be your
> > problem except if you want to. ;)
> 
> This is not out of the question. I have been looking at that
> source but have not been able to do a 'git clone' on it, but
> might be able get it eventually. There are some addtional command
> line options that might be nice to have in addition to the
> chromebook stuff that they are probably concentrating on. I'm
> refering to the fact that there are a zillion different chromebooks out
> there. Intel based is one that comes immediately to mind, as
> apposed to arm.  Do you have a wish list?

Currently not. There are too many other open questions right now to
start even thinking about this yet.
-- 
Kind regards/Mit freundlichen Grüßen, Stefan Tauner