RE: [ECOS] Assabet flash problem

[Gary Thomas <gthomas at cambridge dot redhat dot com>]

On 01-Jun-2001 Nicolas Pitre wrote:
> 
> I have an Assabet for which the precompiled Redboot binary from current CVS
> seems to work just fine.  It is just a bit outdated compared to the current
> state of the source, but still it works.
> 
> However if I recompile redboot myself, the resulting binary just doesn't
> want to recognize the flash on the board anymore:
> 
>|Can't identify FLASH, sorry, man b0, dev d0, id [QRY.] stat 0
>|00018780: B0D0 0002 0000 0000  0000 0000 0000 0000   |................|
>|00018790: 5152 5901 0031 0000  0000 0030 5530 5503   |QRY..1.....0U0U.|
>|000187A0: 060A 0F04 0404 0415  0200 0500 011F 0000   |................|
>|000187B0: 0150 5249                                  |.PRI            |
>|FLASH: driver init failed!, status: 0x8
> 
> Since it worked with older version of Redboot I guess the problem should be
> quite dumb and someone will tell me the quick trick without me having to
> dive deep into the code...  Right?  :-)

Actually, we've completely restructured the Flash drivers in the meantime :-)
I'm sure that the differences are in here.

Look in the file "devs/flash/intel/strata/current/src/strata.c".  Just to see
if everything else works, try changing the line which reads
    if ( (qp->manuf_code == FLASH_Intel_code)
to
    if ( 1

If that works, then you could try this (take the other change out).
In the file "devs/flash/intel/strata/current/src/flash_query.c", change the
macro CNT to be a larger number.

Note: I've seen some of the older StrataFlash devices that went gaga for
a while after issuing the "Read Query" command - that's what the delay is for.

Note 2: If you have a working RedBoot in Flash, you can do this with a RAM
based version of course, for faster & easier testing.

Note 3: The latest Assabet RedBoot code does handle Linux boot in the same
way as on the iPAQ.