Bug 616760

This issue has been proposed when we are only considering blocker
issues in the current Red Hat Enterprise Linux release.

** If you would still like this issue considered for the current
release, ask your support representative to file as a blocker on
your behalf. Otherwise ask that it be considered for the next
Red Hat Enterprise Linux release. **

It's not a bug, because init script is only used when you have installed daemon. I've moved init script into lm_sensors-sensord.

I don't see the lm_sensors init script in the lm_sensors-sensord package, either. Only the sensord init script is there, same as in the beta 1 package. The two are not the same. I don't think we are talking about the same init script.

Created attachment 433449 [details]
proposed patch

Damn, I forgot add a few lines into init script.

No, my point is that the two (sensors and sensord)are independent. The lm_sensors init script is supposed to load the drivers, while the sensord init script is supposed to run sensord. The two are separate.

lm_sensors doesn't require sensord, nor should it, but without the lm_sensors init script to load the drivers, sensors won't work.

You are making sensors dependent on sensord, which is non-sensical. Why should sensord be required to run for sensors to work, just for the sake of the same init script loading the drivers? Sensors won't work in the current configuration, which means at best sensors should have a dependency on sensord (for the init script without which it doesn't work). Plus, sensord is in the optional packages, not the core ones.

It wasn't broken, so why is it being "fixed"? What is the gain?

(In reply to comment #6)
> No, my point is that the two (sensors and sensord)are independent. The
> lm_sensors init script is supposed to load the drivers, while the sensord init
> script is supposed to run sensord. The two are separate.
> 
Aha, that's why there was almost two same scripts.

> lm_sensors doesn't require sensord, nor should it, but without the lm_sensors
> init script to load the drivers, sensors won't work.
> 
> You are making sensors dependent on sensord, which is non-sensical. Why should
> sensord be required to run for sensors to work, just for the sake of the same
> init script loading the drivers? Sensors won't work in the current
> configuration, which means at best sensors should have a dependency on sensord
> (for the init script without which it doesn't work). Plus, sensord is in the
> optional packages, not the core ones.
> 
> It wasn't broken, so why is it being "fixed"? What is the gain?    

Yes it was, sensord init script doesn't load/unload automatically modules when starts/stops at all.

> Yes it was, sensord init script doesn't load/unload automatically modules
> when starts/stops at all.

It isn't supposed to!

sensord service is dependant on lm_sensors service starting before it. In the same way that NFS is dependant on portmap and rpc services. What's wrong with the usual UNIX solution of having the lm_sensors service have a lower priority (earlier startup) than sensord? I still maintain that there was nothing broken to fix.

revert changes from lm_sensors-3.1.1-5.el6

Gordan: I verified lm_sensors init script is back in lm_sensors package. Once beta repositories are refreshed, could you also verify lm_sensors init script is back & working and that the new package satisfied your needs regarding this bug, please?

I certainly will, but it may be a few days before my local mirrors are updated.

*** Bug 611144 has been marked as a duplicate of this bug. ***

Red Hat Enterprise Linux 6.0 is now available and should resolve
the problem described in this bug report. This report is therefore being closed
with a resolution of CURRENTRELEASE. You may reopen this bug report if the
solution does not work for you.