GNOME Bugzilla – Bug 169145
It was crashed couse I didn't give the password of my key
Last modified: 2005-05-04 16:02:56 UTC
Distribution: Debian 3.1 Package: seahorse Severity: normal Version: GNOME2.9.92 0.7.5 Gnome-Distributor: Ubuntu Synopsis: It was crashed couse I didn't give the password of my key Bugzilla-Product: seahorse Bugzilla-Component: general Bugzilla-Version: 0.7.5 BugBuddy-GnomeVersion: 2.0 (2.9.2) Description: Description of the crash: It was crashed couse I didn't give the password of my key Steps to reproduce the crash: 1. 2. 3. Expected Results: How often does this happen? At once Additional Information: Debugging Information: Backtrace was generated from '/usr/bin/seahorse' (no debugging symbols found) Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1". (no debugging symbols found) `system-supplied DSO at 0xffffe000' has disappeared; keeping its symbols. (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) [Thread debugging using libthread_db enabled] [New Thread -1222688640 (LWP 16434)] (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) 0xffffe410 in __kernel_vsyscall ()
+ Trace 56402
Thread 1 (Thread -1222688640 (LWP 16434))
------- Bug moved to this database by unknown@bugzilla.gnome.org 2005-03-03 17:20 ------- Unknown version 0.7.5 in product seahorse. Setting version to "0.7.x". Unknown platform unknown. Setting to default platform "Other". Unknown milestone "unknown" in product "seahorse". Setting to default milestone for this product, '---' Setting to default status "UNCONFIRMED". Setting qa contact to the default for this product. This bug either had no qa contact or an invalid one.
I discovered this same behavior in seahorse-agent. Steps to duplicate: Start seahorse-agent Perform task that requires passphrase Hit "OK" without entering text into passphrase window Cue the assertion. seahorse-agent -d ** ERROR **: file seahorse-agent-cache.c: line 308 (seahorse_agent_cache_set): assertion failed: (pass && pass[0]) aborting... Aborted I'll work on a patch later. There must be a similar bug in the seahorse key manager's code. I'll try and get a patch for that too.
Created attachment 45997 [details] [review] Fixes bug It turns out seahorse is great for finals studying avoidance. This bug solves the problem by not entering the cache function if a null string is returned. It instead behaves as if the cancel button was clicked. Alternately, we could be a pain and insert a different string on null that is impossible to enter into a passphrase so that it just fails out and requeries the user. I think this solution works fine because it follows the method of least surprise.
Hmmm, I wonder if we should support zero length passwords. Does GPG support zero-length passwords, and if so, does it prompt for them?
Yes, GPG supports zero length passphrases. No, it doesn't prompt for them. It doesn't trigger seahorse-agent either. Sure we should support them. The advanced key generator should allow them, and currently doesn't. The assistant should keep requiring matching passphrases.
Created attachment 45998 [details] [review] Proposed patch Does this patch work? It may make more sense in the long run... Or the other approach would be to not allow empty passphrases in seahorse-agent, and disable the 'OK' button etc... Given that GPG doesn't prompt for them, that is.
Yes, it works. I think the error message is sufficient and we don't have to disable the 'OK' button.
Committed patch. In addition, the bug that the bug that the original reporter experienced should now be fixed due to changes in the expiry date modification code. (see stack trace)