Intel H57 chipset (untested) and W25Q64 flash -V output

[Pontus Rodling]

Hi!
flashrom told me to send you guys this, so here it is.

I'm getting a transaction error when trying to read the (fully supported
according to the wiki) W25Q64 flash chip on a Intel H57 chipset which is
marked as untested.
I tried both using swseq and hwseq. Both results are attached.

// Pontus Rodling

Intel H57 chipset (untested) and W25Q64 flash -V output

[David Hendricks]

On Fri, Nov 25, 2011 at 8:10 AM, Pontus Rodling wrote:
> Hi!
> flashrom told me to send you guys this, so here it is.
>
> I'm getting a transaction error when trying to read the (fully supported
> according to the wiki) W25Q64 flash chip on a Intel H57 chipset which is
> marked as untested.
> I tried both using swseq and hwseq. Both results are attached.

Looks like the Management Engine region is totally locked (from both reads
and writes), so it will fail whenever flashrom tries to read/write in that
region;
0x5C: 0x05ff0001 (FREG2: Management Engine)
0x00001000-0x005fffff is locked

We need to do some rather ugly stuff to hack around that. I have a very
ugly hack that applies to the Chromium OS branch you can try:
# clone chromium os branch of flashrom
git clone http://git.chromium.org/chromiumos/third_party/flashrom.gitflashrom-cros
&& cd flashrom-cros

# apply patch from https://gerrit.chromium.org/gerrit/#change,12117
git pull
https://gerrit.chromium.org/gerrit/p/chromiumos/third_party/flashromrefs/cha...

With that patch, flashrom will fill in unreadable parts with 0xff bytes,
and will not attempt to erase/write those parts. Verbose output will show
"WD" (Write Denied) for regions that are unwriteable.

[Stefan Tauner]

On Fri, 25 Nov 2011 17:10:21 +0100
Pontus Rodling  wrote:
> 0x5C: 0x05ff0001 (FREG2: Management Engine)
> 0x00001000-0x005fffff is locked

Hello Pontus,

thanks for your report!

The problem is the locked ME region as quoted above and mentioned by
david before. We are working on unlocking it, but intel does not provide
us any documentation so please do not expect a real solution soon. The
workaround used in chromiumos' version of flashrom is to not touch the
locked region at all. But since it contains another hunk of the
firmware it might brick a board when one updates the bios, but not the
ME region. We do not know if this might really be the case, but i would
not want to try it without a safety net.

I have added the board (MS-7613 (Iona-GL8E)) to our list of
(un)supported boards (with an appropriate note) and will commit that
later together with other small changes.