|
Bugzilla – Full Text Bug Listing |
| Summary: | libgssapi's error handling improvement | ||
|---|---|---|---|
| Product: | [openSUSE] SUSE LINUX 10.0 | Reporter: | Wolfgang Rosenauer <wolfgang.rosenauer> |
| Component: | Basesystem | Assignee: | Olaf Kirch <okir> |
| Status: | RESOLVED FIXED | QA Contact: | E-mail List <qa-bugs> |
| Severity: | Normal | ||
| Priority: | P5 - None | CC: | mc |
| Version: | Final | ||
| Target Milestone: | --- | ||
| Hardware: | All | ||
| OS: | SuSE Linux 10.0 | ||
| Whiteboard: | |||
| Found By: | Other | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
|
Description
Wolfgang Rosenauer
2006-02-03 08:29:32 UTC
If I understand the problem right, we need to symlink libgssapi_krb5.so.2 to libgssapi_krb5.so? Not anymore. Firefox 2.0 will use libgssapi_krb5.so.2 automatically as first try. What should be fixed within this bug is the behaviour of our libgssapi.so to just exit if not configured properly. IMHO a module shouldn't exit the whole application. For Firefox it isn't needed to change anything for future products (including CODE10). But maybe there are other users of libgssapi.so which get the same problem. There are, indeed. The entire NFS RPCSEC_GSS code seems to rely on this. Michael, can the krb5 package be changed to include the libgssapi_krb5.so symlink? Sure, but does this solve any problem? I think they search at compile time for this and at compile time krb5-devel might be installed. The problem is, that they want a libgssapi library. libgssapi_krb5 is only the implementation of one mechanism for gssapi. If you want that I do the change send me a "doit" :-) Michael, it's slightly more complicated than that. The NFS security stuff is linked against libgssapi which provides the mechanism switching. If the caller requests a krb5 mechanism, libgssapi will try to open libgssapi_krb5.so I could also change the mechglue.conf file to go and try to load libgssapi_krb5.so.2, but I'm not sure whether that's the best possible solution. Ahh, ok. Should I submitt it to STABLE? I have submitted it. Thanks, Michael! So I'm closing this bug. |