Jérôme Marant wrote:
Le mercredi 15 novembre 2006 01:04, Attilio Fiandrotti a écrit :I think what you did was correct, and i think this also may be a proof of the fact that gtkdfb is not guilty.But we both agree it comes from signal, don't we?
This is what i tought some times ago, but now i'm no longer that sure.You've proved DFB's insternal signaling mechanism works correctly on AMD64, and doesn't interfere with cdebconf's signal handlers.
Jerome, please look at bug #397678: do you see any similarity with the crash that you experienced in g-i ?I took a look at this bug report but failed to understand how it is related to the VT crash. What do you have in mind?
i was wondering if another d-i component, like main-menu, could be responsable for the crash, but this was just an idea. ATM i really don't have a clue, maybe attaching a debugger to cdebconf could make some light?
cheers attilio