#3247 BLOC MP Star: NAND rot
Zarro Boogs per Child
bugtracker at laptop.org
Thu Sep 13 00:37:53 EDT 2007
#3247: NAND rot
--------------------------------+-------------------------------------------
Reporter: wmb at firmworks.com | Owner: mlj
Type: defect | Status: new
Priority: blocker | Milestone: MP Start
Component: hardware | Version:
Resolution: | Keywords:
Verified: 0 |
--------------------------------+-------------------------------------------
Comment(by wmb at firmworks.com):
Replying to [comment:10 Luna]:
> Dear Mitch,
> What condition will OFW show the massage,"Unsupported NAND FLASH Device
Code AD"?
That message happens if the second byte of the ID data is not one of the
supported values. The expected values are either 0xDC for 512 MB chips,
or 0xD3 for 1 GB chips (some 1GB parts are organized as two 512 MB chips
in the same package). See "configure-size" in
openfirmware/dev/olpc/cafenand/configure.fth
The value 0xAD is not reasonable for that second byte. However, 0xAD *is*
the expected value for the first byte of the ID data for Hynix chips.
Perhaps the read-id command is returning an extra junk byte at the
beginning of the ID data. I do not know what would cause that to happen.
You can look at the ID data as follows:
{{{
ok begin-select /nandflash
ok debug configure-size
ok open
<debugger activates>
: configure-size ( ffaed000 )
dup f
Type 'resume' to return to debugger
ok dup 5 cdump
ad dc 80 95 ad
ok
}}}
That will display the ID buffer. The values shown are for a Hynix part on
a B4.
AD is the Hynix device ID code, DC is the size code.
If you then do:
{{{
ok resume
( ffaed000 )
dup ( ffaed000 ffaed000 )
1 ( ffaed000 ffaed000 1 )
<continue typing spaces>
40000 ( ffaed000 40000 )
endof ( ffaed000 40000 )
to ( ffaed000 )
;
: configure-size ( ffaed000 )
dup f
ok dup 5 cdump
}}}
The second call to "configure-size" is for the second 512 MB chip inside
the 1 GB package.
--
Ticket URL: <https://dev.laptop.org/ticket/3247#comment:12>
One Laptop Per Child <https://dev.laptop.org>
OLPC bug tracking system
More information about the Bugs
mailing list