DSM-G600, DNS-3xx and NSA-220 Hack Forum

Actually, I haven't had  the time yet to try it out. Also, the features listed there don't do _that_ much for me. But, since I'm an upgrade geek, I might try it tonight if I have time.

I've tried it and turned back to 1.03final, and i can confirm that this release has that issue (telnetd from ffp0.5 doesn't start, you can see a cxa_atexit related message in the ffp.log, on my box dropbear & openssh didn't work any more either).
To fix it, the kernel of the firmware should be compiled with cxa_atexit enabled i think... or perhaps we could compile a busybox for ffp0.5 with that option disabled and see if this is enough or if that error masks other errors...

Thanks fonz for the fast reply, here is a piece of log i've posted in another topic about 1.04alpha1 & ffp0.5 (as of april 20th), to give you an idea of the error displayed:

* /ffp/start/sysinit.sh ...
mknod: can't resolve symbol '__cxa_atexit'
mknod: can't resolve symbol '__cxa_atexit'
mkdir: can't resolve symbol '__cxa_atexit'
mount: mounting devpts on /dev/pts failed: No such file or directory
chmod: can't resolve symbol '__cxa_atexit'
* /ffp/start/syslogd.sh inactive
.......
* /ffp/start/telnetd.sh ...
basename: can't resolve symbol '__cxa_atexit'
tr: can't resolve symbol '__cxa_atexit'
basename: can't resolve symbol '__cxa_atexit'
Starting /ffp/sbin/telnetd
* /ffp/start/rsyncd.sh inactive

Tonight i'll post something more up-to-date, with firmware 1-04rc1 and the latest ffp binaries.
I've googled for __cxa_atexit and i found that a --disable__cxa_atexit option exists, and looks like that:
- It can be used while building gcc, and is related to the use of atexit() from libc,uclibc,ecc.. in the binaries produced by gcc.
- http://busybox.net/lists/buildroot/2007 … 4788.html,  this error is displayed when a binary and its uclibc have a different __cxa_atexit  settings (Really weird... this should not happen, if everything here, uclibc,busybox,etc... is from the the ffp and then obviusly compiled with the same configuration).
Fonz, i'm surely missing something BIG here... if i recompile everything, even gcc, with --disable__cxa_atexit the issue should be solved, what do you think? Should i give a try?

Hmm... looks like that the big thing i was missing was actually the search button ...  i saw your note on the changelist but didn't associate it with my issue, tnx.

About the new firmware, as described in the changelist: minor changes in the user/group administration, new clear print queue button, major changes to the upnp server ( that i personally use instead of mediatob,twonky,etc... for some issues with .sub files) that now supports some new format and has a new smart update mechanism, no more based on fs scan, it now uses inotify.
Seems to work, no issues found so far.