| Summary: | net-misc/networkmanager-0.6.5 executes non-existing /usr/bin/killall | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | João Vale <jpvale> |
| Component: | Current packages | Assignee: | Steev Klimaszewski (RETIRED) <steev> |
| Status: | RESOLVED FIXED | ||
| Severity: | minor | CC: | kevin |
| Priority: | High | ||
| Version: | 2007.0 | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | Fix path to 'killall' | ||
So re-emerge it? Tried re-emerging sys-process/psmisc and then net-misc/networkmanager-0.6.5, but the results were the same. Sorry, should've said that in the report.
Anyway, I snooped around and I found this in line 240 of src/backends/NetworkManagerGentoo.c
nm_spawn_process("/usr/bin/killall -q nscd");
Not sure if this is the right place to change it though, so I didn't make a patch. :P
The location of a binary like killall should be a configure option, or detected at build time by configure. My inclination would be to put a local patch on it to detect for configure, and sent that patch upstream. After re-reading it, I'd say patch the Gentoo file, and send it upstream. Seems like a strange way to handle everything, but i guess upstream knows best ?? Created attachment 124537 [details, diff]
Fix path to 'killall'
Shall I submit this upstream?
Unfortunately, I don't experience this - and really, you would need to fix it some other way, on my hardened machine, i get steev@keuhi /bin $ which killall /usr/bin/killall steev@keuhi /bin $ ls -l /usr/bin/killall -rwxr-xr-x 1 root root 18K 2007-08-13 15:25 /usr/bin/killall on my amd64 semi-stable i get steev@equilibrium sigma6 $ which killall /usr/bin/killall steev@equilibrium sigma6 $ ls -l /usr/bin/killall lrwxrwxrwx 1 root root 12 2007-04-19 08:31 /usr/bin/killall -> /bin/killall on x86 stable I get steev@ariandorix ~ $ which killall /usr/bin/killall steev@ariandorix ~ $ ls -l /usr/bin/killall -rwxr-xr-x 1 root root 14328 2007-08-14 07:40 /usr/bin/killall Seems like the problem was (silently?) fixed, probably in the sys-process/psmisc ebuild. Oh well. Scratch one from the bug count. |
When I connect to a network, I get this in the syslog: NetworkManager: <WARN> nm_spawn_process(): nm_spawn_process('/usr/bin/killall -q nscd'): could not spawn process. (Failed to execute child process "/usr/bin/killall" (No such file or directory)) Found a few bugs concerning killall changing from /usr/bin/killall to /bin/killall , but none related to NetworkManager.