The cipher_bin2hexstr has an offbyone in malloc - leading to a missing string-terminator. Whoever wrote this, please put a brown paper bag on your head or bring a case of franconian beer to my office (preferred).

Maybe the return value of malloc also should be subject to further scrutiny, but i am not familiar with g_malloc0, maybe this already takes care of everything.

BTW: after analyzing my logs, this might account to many of the mysterious random failures, where i always thought that just the timeout was too small.

Created attachment 66704 [details]
Trivial patch, this should fix it.

Bart, after looking at the code, this might also help with your WEP problems.

Nice catch, seife.  Also nice work tracking down the specific bug.

Submitted fix to CVS and updated package (20060207) to autobuild.

Everyone: Please retest any problems you have add with the previous versions and update your bugs!

Sorry.  This didn't fix my bug (#144268)

works in 10.1 final