[RFC] Bootscript changes
dbn.lists at gmail.com
Wed Apr 11 11:23:32 PDT 2007
On 4/11/07, Alexander E. Patrakov <patrakov at ums.usu.ru> wrote:
> Dan Nicholson wrote:
> > There is still another copy of pppd running when this occurs, right?
> > I think when I was testing things out I never actually had two copies of
> > the process running. Could you see if the attached patch against SVN
> > works correctly? Hopefully, this fixes the whole codepath to do the
> > right thing.
> The patch passes my testcase.
Later I realized what the [ -z "$killsig" ] was for. The subsequent
tests check whether the process is still running, but this isn't
appropriate is you just sent SIGHUP, for example. With my previous
patch, this could break some bootscripts that send a non terminating
signal through killproc, e.g.:
killproc process HUP
Although, in that case, it would probably be more appropriate to use
reloadproc. Still, I don't want to break the existing behavior, so I
changed the function a bit so that it will only check if the process
is still running if the signal is SIGTERM or SIGKILL. I don't know if
this is totally appropriate since I'm no signal expert.
Attached is the updated patch. I'd appreciate if you could test again.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 2366 bytes
Desc: not available
More information about the lfs-dev