openafs debug information CRC mismatch
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
openafs (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned | ||
Lucid |
Fix Released
|
Undecided
|
Unassigned | ||
pkg-create-dbgsym (Ubuntu) |
Fix Released
|
Undecided
|
Martin Pitt | ||
Lucid |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Binary package hint: openafs-dbg
We have successfully been using the debug symbols from openafs-dbg when debugging our Hardy fileservers. We are currently in the need of doing debugging of our lucid fileservers but it seems that the openafs-dbg package on lucid is broken.
> Description: Ubuntu 10.04.1 LTS
> Release: 10.04
>
> openafs-dbg:
> Installed: 1.4.12+dfsg-3
> Candidate: 1.4.12+dfsg-3
How to reproduce:
1) Install the debug package openafs-dbg
2) Run: gdb /usr/lib/
Instead of actually loading the debug symbols as expected the following message is printed:
> Reading symbols from /usr/lib/
> warning: the debug information found in "/usr/lib/
> not match "/usr/lib/
>
> warning: the debug information found in "/usr/lib/
> not match "/usr/lib/
> (no debugging symbols found)...done.
and consequently no debugging symbols are available for the debugger.
After searching LauncPad for similar problems I found among others https:/
Not having the possibility to attach a debugger and get symbols for openafs makes debugging afs tricky.
Related branches
Changed in pkg-create-dbgsym (Ubuntu): | |
assignee: | nobody → Martin Pitt (pitti) |
For those triaging or looking at this bug, just wanted to note that I believe this is fallout from Ubuntu's handling of debug symbols, which has diverged from Debian. I don't believe this is an issue in the packages from the Debian perspective. I am happy to modify the package so that it will work properly with Ubuntu if this doesn't break anything for Debian, but I don't know enough about how the Ubuntu debug generation works to be able to come up with such a patch myself. (This is assuming that a simple rebuild doesn't fix it, which it might for all I know.)