[Frage] Fritz!Box 7530 fwmod firmware unpacking

[gedia]

Hello all,

Is there any way to unpack firmware intended for FB7530? I know it's not supported by freetz currently, but I tried to unpack anyway (with config for 7590, which I now know is quite different despite having similar specs).

I really need to install the international firmware on the German version, and the only way seems to be with a modified image to persistently change the firmware_version variable to "avme" from "avm". But I can't modify it if I can't unpack it, so I was wondering if anyone knew of any workarounds.

This firmware fragmentation thing with AVM is very annoying anyway, the german version ends up in various e-shops all over Europe and the merchants fail to properly communicate the fact that the device is intended for a foreign market to buyers. Now I'm stuck with Annex B and DSL won't sync...

Thanks in advance.

### Zusammenführung Doppempost by stoney ###

Also do you think I could have better luck with modfs? Am I right in thinking that this is an unpacker/patcher/repacker? I haven't been able to find use instructions for it...

Does it support the 7530?

 

[er13]

7530 is a little endian box (i.e. little endian CPU, little endian SquashFS format). 7590 is however big endian.

So if you use 7590 as a baseline you need to somehow trick fwmod into using little endian SquashFS.

Overriding FREETZ_TARGET_ARCH_LE/FREETZ_TARGET_ARCH_BE won't work probably - there is no mipsel toolchain for 7590 (even though it's not needed for pure repacking Freetz build system will still try to download/create it).

Copying tools/mksquashfs4-avm-le over tools/mksquashfs4-avm-be and tools/unsquashfs4-avm-le over tools/unsquashfs4-avm-be might however work. I however never tested it myself. So no guarantee/liability or whatever.

p.s. the (highly experimental) support for 7530 might become available in Freetz in the next couple of days. 7530 uses almost the same hardware as 4040, which is almost supported in my local working copy here.

 

[gedia]

Thanks! That's good to know. I'll wait for 7530/4040 preliminary support and try my luck then...

 

[gedia]

So I'm able to unpack successfully using latest git master, but repacking seems to fail:

~/avm/image $ ../freetz/fwmod -p -d unpack ~/Downloads/FRITZ.Box_7530.en-de-es-it-fr-pl-nl.164.07.02.image
detected firmware 7530_en-de-es-fr-it-nl-pl 164.07.02 rev62314 (16.10.2018 17:51:45)

STEP 3: PACK
WARNING: Modifications (STEP 2) and this step should never
         ever be run with different configurations!
         This can result in invalid images!!!
WARNING: firmware does not seem to be modified by the script
invoking custom script
  restoring support for /var/flash/debug.cfg
    patching ./filesystem/etc/init.d/rc.tail.sh
  checking for left over Subversion directories
packing var.tar
creating filesystem image (SquashFS4-xz)
  SquashFS block size: 64 kB (65536 bytes)
copying kernel image
  kernel image size: 2.8 MB, max 4.0 MB, free 1.2 MB (1258240 bytes)
ERROR: creation of filesystem failed

Any ideas?

 

[Shirocco88]

Gibt es auch eine ".config" Datei, ?
dadurch wird erst eine exakte Reproduzierung des Problems möglich

 

[gedia]

Hello,

I'm guessing you're asking if I run make menuconfig first, which I did. There is a .config file under ~/avm/freetz, and the unpacked image is under ~/avm/image/unpack.

 

[er13]

No ideas as you haven't provided any information making it possible for us to reproduce the problem. Which .config did you try it with? Was it a one for 7590 or for 7530. What exactly did you do? All steps please.

In either case, if you want to use a German box on an Annex A line then you have two options:

1. set kernel_args to annex=A as already pointed out by @KunterBunter in #4
2. activate the "Unhide Annex selection" option (available since yesterday) in Freetz menuconfig and use Freetz in a completely automated "no freetz"-mode with German firmware. I.e. there is no need to reflash the German box with the international firmware.

 

[gedia]

I repeated the procedure under ~/avm/freetz and it seems to have worked! Thanks for the hint. Will post here the results