By the way.. If I use a tool chain built according to the mentioned wiki page, the resulting firmware does not work! It is impossible to initialize the flash.. Some weeks ago, I started debugging this. I found out that, as much I can remember, the CFI layer gives a bad protocol id (or version). I expected that it could be related to some padding or alignment differences... If I use binutils-2.17, gcc-4.1.1-c-c++, newlib-1.14.0, everything works fine Matthias Am 07.01.2011 um 16:34 schrieb Bogdan Alecu:

I didn't mean that I want to flash something. But any firmware simply does not initialize properly and stops working. By the way its on Ubuntu-10.10. Am 07.01.2011 um 17:52 schrieb Steve Markgraf:

It means that the firmware initialization function tries to initialize the flash memory which fails. I put some printfs in every place where it can fail and found that, as much I can remember, the protocol id (or version) returned 0 instead of (probably) 3.. But I will reproduce it and send the output.. It was with a C123. Probably I try to find the problem behind it then, if nobody already knows what the root cause is.. Regards Matthias Am 07.01.2011 um 18:09 schrieb Steve Markgraf:

Thank you Sylvain. After some talking on the IRC channel it seems that the phone is not sending the correct strings in HyperTerminal on Windows. I'll try with another cable and phone. It's really hard to find an already made USB cable. --- On Fri, 1/7/11, Sylvain Munaut <246tnt(a)gmail.com> wrote: From: Sylvain Munaut <246tnt(a)gmail.com> Subject: Re: Osmocon not loading To: "Bogdan Alecu" <b.alecu(a)yahoo.com> Cc: baseband-devel(a)lists.osmocom.org Date: Friday, January 7, 2011, 4:21 PM On Fri, Jan 7, 2011 at 4:34 PM, Bogdan Alecu <b.alecu(a)yahoo.com> wrote: Your cable is bad, don't look further ... if the loader doesn't see the PROMPT1, nothing will ever happen and only a bad cable/connection to the phone will do that. (or the phone serial port is damaged ...)    Sylvain