Removed from his working MGTSnogo mtdblocks.
It is V4.1 (like the one I messed up, the quote is ".F660v4.1 .... error! Wrong board !! You may upgrade (6818) F660v4 version on new Board (6818C0)."); That is, 2.2 firmware should eat.
mtdblock0 16384, similar to nvram
mtdblock1 14320K, it looks like CFE (incl micro_httpd), but not 14M? The tail clearly contains something compounded.
mtdblock2 14320K, there is also something similar, but the files are different
mtdblock3 == mtdblock4 1048576, contains something like "BootImageNum = 0
DownloadState = 3
LastBooting = 0
IsOmciUpgrading = 0 "
In short, nothing is clear to me by the results.
The device that I screwed up behaves strangely. On the 1st and 2nd internet, he has a carrier and he blinks an LED if there is activity, on 3 and 4 - no. It also lights up with a WPS light. Somehow it does not look like a completely killed CFE, maybe it's a killed nvram?
About reset type 30-30-30 I know that does not help.
You will find the PDFs from BCM6818 (but it seems that it is not needed and do you need to flash a flash drive?), As well as from the 17th and 16th.
FLASH has NAND (which is read by sectors, not ROM), do I understand correctly that there should be something that forces it to be read first in RAM, then run it?
UPD: By the way, a smart article about JTAGing of a similar device:
http: // huaweihg612hack ... roadcom-bcm6368-hg612 /on BCM6368, there is also NAND flash
777L @ 09/18/2014, 10:35 PM
You can copy it to a USB flash drive, there you can redo it and copy it back to the device.
It just won’t work, because read only file system. And for this we need deeper research .. I hope that dreadbit will be able to collect jtag
In general, jtag will not help here at all. As in the article about ZTE831 there it is necessary to tear squashfs, and then assemble it back.
Post has been editedShoore - 16.10.14, 00:13Reason for editing: combined