**** BEGIN LOGGING AT Mon May 12 02:59:59 2014
May 12 09:14:48 <ant_work> bluelightning: gm
May 12 09:17:16 <bluelightning> hi ant_work
May 12 09:17:37 <ant_work> so, thx for the logs
May 12 09:17:49 <ant_work> see: http://logs.nslu2-linux.org/livelogs/kexecboot-prev.txt
May 12 09:17:57 <bluelightning> seems the kernel buffer is too short to contain the whole log unfortunately
May 12 09:18:35 <ant_work> seems my chips are defective...one of them
May 12 09:18:39 <ant_work> ffff0039 vs. 390039
May 12 09:18:39 <bluelightning> do you know if I can fix that? (assuming it would help)
May 12 09:19:01 <ant_work> yours give readings 390039 at offset 15
May 12 09:19:18 <ant_work> mines ffff0039
May 12 09:19:46 <ant_work> I think your log would be 'as expected'..CFI only cares about 3 addresses
May 12 09:19:58 <ant_work> 10 - 11 - 12
May 12 09:20:57 <ant_work> to verify that I can send you a kernel w/out misdetection patch, failing on my device
May 12 09:22:12 <ant_work> I try to reflash to standard Sharp images but I can't think I've deleted the CFI table from one chip...
May 12 09:22:21 <ant_work> afaik it is ROM
May 12 09:23:34 <ant_work> anyway, good news
May 12 09:23:48 <ant_work> that misdete3ction-patch would hardly go upstream...
May 12 09:25:12 <ant_work> bluelightning: I think I've just had bad luck...fighting months vs. bugged hardware
May 12 09:25:41 <bluelightning> ant_work: :(
May 12 09:25:42 <ant_work> now if only lumag could confirm on a third collie...
May 12 09:26:16 <bluelightning> another alternative would be to send out a request to the mailing list for anyone else to try on theirs
May 12 09:26:34 <bluelightning> I'm sure there are other collie owners out there lurking :)
May 12 09:26:36 <lumag> ant_work, bluelightning hello
May 12 09:26:41 <bluelightning> hey lumag
May 12 09:26:47 <ant_work> hello
May 12 09:27:05 <lumag> Quite overloaded with the work. What image should I test on collie? Fresh OE?
May 12 09:27:24 <ant_work> is one kernel printing CFI readings
May 12 09:27:29 <lumag> I can probably try to do few tests today on in the next few days.
May 12 09:27:35 <ant_work> seems only mine gives ffffxxxx
May 12 09:27:47 <ant_work> or yours as well?
May 12 09:27:50 <lumag> Hmm. I remember also having this problem.
May 12 09:28:06 <lumag> ant_work, can you upload a kernel somewhere, I can run a test.
May 12 09:28:17 <ant_work> I'll send by mail now if you want
May 12 09:28:21 <lumag> yep.
May 12 09:31:42 <ant_work> lumag https://github.com/andrea-adami/collie-nor-flash/blob/master/docs/LH28FXXXBF_appendix_datasheet%282.44%29.pdf
May 12 09:31:50 <ant_work> page 7
May 12 09:32:09 <ant_work> in the Block Diagram I see "Query ROM"
May 12 09:32:46 <ant_work> I don't think it can be erased/flashed, though I'll reinstall original Sharp ROM full 16MB
May 12 09:50:57 <lumag> ant_work, It can not. At least I don't know a way to destroy flash internals that badly.
May 12 09:53:21 <bluelightning> I thought I remember reading from a long time ago that NOR flash could be damaged due to incorrect accesses...
May 12 11:43:40 <ant_work> fwiw  http://www.google.it/patents/US6118694
May 12 11:43:54 <ant_work> "In prior art flash memory, the CFI logic was incorporated into a separate portion of the chip as a read only memory ("ROM") array consisting of a plurality of rows and columns of transistors. "
May 12 11:44:50 <ant_work> lumag: ^^ they talk about X and y decoders, like in the tech sheet above
May 12 11:48:31 <ant_work> that patend is dated 2000 and introduces " the CFI data is stored in the CFI bit line decoder thereby eliminating the need for a separate area in the substrate of the memory device for the CFI "
May 12 14:01:26 <ant_work> and I read for contemporary models  http://www.sharpsma.com/download/LH28F160S5NS-l70-DSpdf   "Query database is stored in the 48 Byte ROM."
**** ENDING LOGGING AT Tue May 13 02:59:58 2014