Veracode static analysis results
elb at pidgin.im
Wed Dec 5 16:22:29 EST 2012
Chris Wysopal spake unto us the following wisdom:
> A customer asked us to analyze Pidgin using our static analyzer. Our
> responsible disclosure policy is to inform you of any findings so that
> you may have the chance to review, comment, and/or fix the issues.
> I think the software performed very well on our analysis but there are
> a few issues we have found. Attached is our full report. You can find
> the description of the issues found on pages 10-15. We found 1 Very
> High criticality. 5 Medium, and 47 low. Here is a summary.
OK, here's my fifteen minute analysis of the bugs. There's only one I
think I'd really worry about. I've not Cc'd veracode, we can send
them our final conclusions.
* gtkpounce error is a false positive. Yes, we execute a user path
without verifying it, but that's the whole *point* of that feature.
It's not particularly safe, but only in an "enough rope to hang
yourself with" kind of way.
* NTLM session key -- I don't know enough about NTLM to say if this is a
real problem or not. Using a real RNG certainly wouldn't hurt.
* purple_core_migrate user-specified path is a false positive. Pidgin
can already be coerced to read any file the user can read, and in the
general sense *should* be able to do so. This class of bug simply
doesn't apply to Pidgin.
* PurpleDesktopItem creation from file -- I don't even know what this
* write_data_to_file path problem -- see purple_core_migrate
* write_data_to_file race -- this is real. We should be using open()
and fdopen() (or the g_ equivalents thereof?).
I'm not even going through these right now. Some of them probably merit
checks; the majority of the 47 are in imported code from glib or
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 482 bytes
Desc: Digital signature
More information about the security