-----Ursprüngliche Nachricht----- Von: Sylvain Munaut 246tnt@gmail.com
Damnit, I forgot to enable TX ... sorry about that.
Sylvain,
now it behaves much more stable!! That is awesome. Thank you. What have you changed / what was the issue?
Now I'm in the network only with a single disconnect since a few minutes. I also had the chance to make a phone-call now.
Stefan
now it behaves much more stable!! That is awesome. Thank you. What have you changed / what was the issue?
Now I'm in the network only with a single disconnect since a few minutes. I also had the chance to make a phone-call now.
I didn't change anything, I just compiled the git.
For some unknown reason, some ARM toolchain produce "half working" binaries ... we have no idea why that happens.
What's the toolchain you use ? Could you send be the binary you compiled ?
Cheers,
Sylvain
On 9 January 2012 13:45, Sylvain Munaut 246tnt@gmail.com wrote:
I didn't change anything, I just compiled the git.
For some unknown reason, some ARM toolchain produce "half working" binaries ... we have no idea why that happens.
What's the toolchain you use ? Could you send be the binary you compiled ?
Cheers,
Sylvain
Hi,
I have had this with my Pirelli DP-L10. Steve|m on IRC gave me his binaries and it worked. I am at work at the moment but can send through some logs/details later if it helps as well.
Cheers,
Paul
On 9 January 2012 13:54, Paul Dart pauldart@gmail.com wrote:
Hi,
I have had this with my Pirelli DP-L10. Steve|m on IRC gave me his binaries and it worked. I am at work at the moment but can send through some logs/details later if it helps as well.
Cheers,
Paul
Just for completeness here are my results:
1. Compiled myself with gnuarm-3.4.3 (as suggested here http://bb.osmocom.org/trac/wiki/GettingStarted ). http://twigman.com/dev/layer1.highram.bin MD5: 682d115436bf14a3abdd60b1032eb1aa layer1.highram.bin
Results on running with osmocon (end of log):
Preparing the last block, filling 534 bytes, block checksum is 0x8a handle_write_block(): 1024 bytes (1024/1024) handle_write_block(): Block 51 finished Finished, sent 51 blocks in total Received block ack from phone Sending checksum: 0x7c Checksum on phone side matches, let's branch to your code Branching to 0x00820000 Received branch ack, your code is running now!
and it just sits there.
2. Using steve|m's bin file http://twigman.com/dev/pirelli.l1.highram.bin MD5: 6b6a84199e9a1b272fc0e954bd2f1cef pirelli.l1.highram.bin
Results on running with osmocon (end of log):
Preparing the last block, filling 1012 bytes, block checksum is 0xb6 handle_write_block(): 1024 bytes (1024/1024) handle_write_block(): Block 52 finished Finished, sent 52 blocks in total Received block ack from phone Sending checksum: 0xdc Checksum on phone side matches, let's branch to your code Branching to 0x00820000 Received branch ack, your code is running now!
OSMOCOM Layer 1 (revision osmocon_v0.0.0-1208-g506a344-modified) ====================================================================== Device ID code: 0xb496 Device Version code: 0x0000 ARM ID code: 0xfff3 cDSP ID code: 0x0128 Die ID code: e5140a2ee514551d ====================================================================== REG_DPLL=0x2413 CNTL_ARM_CLK=0xf0a1 CNTL_CLK=0xff91 CNTL_RST=0xfff3 CNTL_ARM_DIV=0xfff9 ====================================================================== Power up simcard: Assert DSP into Reset Releasing DSP from Reset Setting some dsp_api.ndb values Setting API NDB parameters DSP Download Status: 0x0001 DSP API Version: 0x0000 0x0000 Finishing download phase DSP Download Status: 0x0002 DSP API Version: 0x3606 0x0000 LOST 7333!
and then it seems to work.
So this seems to be some compilation error (clearly as they are different sized files). I imagine it's due to that GettingStarted page pointing to
http://bb.osmocom.org/trac/wiki/toolchain and not http://bb.osmocom.org/trac/wiki/GnuArmToolchain
Is this just a documentation problem? I think so. If everyone agrees we should update the wiki...
Comments?
Paul
On 9 janv. 2012, at 14:45, Sylvain Munaut wrote:
now it behaves much more stable!! That is awesome. Thank you. What have you changed / what was the issue?
Now I'm in the network only with a single disconnect since a few minutes. I also had the chance to make a phone-call now.
I didn't change anything, I just compiled the git.
For some unknown reason, some ARM toolchain produce "half working" binaries ... we have no idea why that happens.
FYI During 28C3, I found out that the firmware compiled on Mac OS X with arm-none-eabi-gcc (from macport) was not working for me, whereas arm-elf-gcc was working just fine.
iZsh
baseband-devel@lists.osmocom.org