I used the steps from here http://bb.osmocom.org/trac/wiki/GnuArmToolchain to build the toolchain. I also tried with and without xor, no results.
On Fri, Jan 7, 2011 at 4:34 PM, Bogdan Alecu b.alecu@yahoo.com wrote:
I used the steps from here http://bb.osmocom.org/trac/wiki/GnuArmToolchain to build the toolchain. I also tried with and without xor, no results.
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
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@gmail.com wrote:
From: Sylvain Munaut 246tnt@gmail.com Subject: Re: Osmocon not loading To: "Bogdan Alecu" b.alecu@yahoo.com Cc: baseband-devel@lists.osmocom.org Date: Friday, January 7, 2011, 4:21 PM
On Fri, Jan 7, 2011 at 4:34 PM, Bogdan Alecu b.alecu@yahoo.com wrote:
I used the steps from here http://bb.osmocom.org/trac/wiki/GnuArmToolchain to build the toolchain. I also tried with and without xor, no results.
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
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 used the steps from here http://bb.osmocom.org/trac/wiki/GnuArmToolchain to build the toolchain. I also tried with and without xor, no results.
Hi,
On 07.01.2011 17:30, Thatsit wrote:
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..
Well, it might be that the flash support is broken with that toolchain. But why would you need flash anyway, we completely run from SRAM at the moment, just load layer1.compalram.bin
Feel free to commit a fix for that though ;)
Regards, Steve
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:
Hi,
On 07.01.2011 17:30, Thatsit wrote:
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..
Well, it might be that the flash support is broken with that toolchain. But why would you need flash anyway, we completely run from SRAM at the moment, just load layer1.compalram.bin
Feel free to commit a fix for that though ;)
Regards, Steve
On 07.01.2011 18:01, Thatsit wrote:
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.
What do you define as 'does not initialize properly and stops working'? Can you provide the _full_ output of osmocon?
Regards, Steve
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:
On 07.01.2011 18:01, Thatsit wrote:
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.
What do you define as 'does not initialize properly and stops working'? Can you provide the _full_ output of osmocon?
Regards, Steve
Hi,
On 07.01.2011 18:15, Thatsit wrote:
Probably I try to find the problem behind it then, if nobody already knows what the root cause is..
Andreas Oberritter submitted a patch that fixes this issue, I just committed it to master.
See:
http://cgit.osmocom.org/cgit/osmocom-bb/commit/?id=0e130bf467923ddaff6000aa4...
Regards, Steve
To be more precise, it was the loader firmware. which didn't initialize properly.
Am 07.01.2011 um 18:09 schrieb Steve Markgraf:
On 07.01.2011 18:01, Thatsit wrote:
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.
What do you define as 'does not initialize properly and stops working'? Can you provide the _full_ output of osmocon?
Regards, Steve
baseband-devel@lists.osmocom.org
-
Bogdan Alecu -
Steve Markgraf -
Sylvain Munaut -
Thatsit