From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hans de Goede Date: Tue, 26 Feb 2008 14:11:45 +0000 Subject: Re: [lm-sensors] Patch: don't start initscript by default on Message-Id: <47C41E21.4060805@hhs.nl> MIME-Version: 1 Content-Type: multipart/mixed; boundary="------------090500050005000409060300" List-Id: References: <47B04B1C.9070405@hhs.nl> In-Reply-To: <47B04B1C.9070405@hhs.nl> To: lm-sensors@vger.kernel.org This is a multi-part message in MIME format. --------------090500050005000409060300 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Jean Delvare wrote: > Hi Hans, > > On Mon, 11 Feb 2008 14:18:20 +0100, Hans de Goede wrote: >> The attached patch modifies the initscript included with lm_sensors so that it >> doesn't start by default on chkconfig using systems. >> >> The patch also modifies sensors-detect to activate the initscript once >> lm_sensors has been successfully configured. >> >> If I don't get any objections I will be committing this to svn in a couple of days. > > Overall I am fine with this change (it doesn't affect openSuse.) > >> diff -up lm_sensors-3.0.1/prog/detect/sensors-detect.foo lm_sensors-3.0.1/prog/detect/sensors-detect >> --- lm_sensors-3.0.1/prog/detect/sensors-detect.foo 2008-01-22 13:22:47.000000000 +0100 >> +++ lm_sensors-3.0.1/prog/detect/sensors-detect 2008-02-11 13:42:04.000000000 +0100 >> @@ -5943,6 +5943,15 @@ EOT >> >> if (-x "/sbin/insserv" && -f "/etc/init.d/lm_sensors") { >> system("/sbin/insserv", "/etc/init.d/lm_sensors"); >> + } elsif (-x "/sbin/chkconfig" && -f "/etc/init.d/lm_sensors") { >> + system("/sbin/chkconfig", "lm_sensors", "on"); >> + if (-x "/sbin/service") { >> + if (-f "/var/lock/subsys/lm_sensors") { >> + system("/sbin/service", "lm_sensors", "restart"); >> + } else { >> + system("/sbin/service", "lm_sensors", "start"); >> + } > > My only worry is to see the lock file name "leak" to sensors-detect. > I'd prefer it to be only mentioned in the initialization script > itself... Is there any problem to calling "restart" unconditionally? Or > maybe even better, shouldn't we call "stop" at the beginning of > sensors-detect? Probing for devices while the hwmon drivers are loaded > has side effects we probably want to avoid anyway. > Calling stop at the start of sensors-detect is a good idea, new version attached. This once again shouldn't effect any non RedHat derived distro's, but still a quick review would be a good idea I think. Regards, Hans --------------090500050005000409060300 Content-Type: text/plain; name="lm_sensors-3.0.1-service-default-off.patch" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="lm_sensors-3.0.1-service-default-off.patch" --- lm_sensors-3.0.1/prog/init/lm_sensors.init.default-off 2007-04-27 14:34:17.000000000 +0200 +++ lm_sensors-3.0.1/prog/init/lm_sensors.init 2008-02-26 14:55:58.000000000 +0100 @@ -1,6 +1,6 @@ #!/bin/sh # -# chkconfig: 2345 26 74 +# chkconfig: - 26 74 # description: sensors is used for monitoring motherboard sensor values. # config: /etc/sysconfig/lm_sensors # --- lm_sensors-3.0.1/prog/detect/sensors-detect.default-off 2008-01-22 13:22:47.000000000 +0100 +++ lm_sensors-3.0.1/prog/detect/sensors-detect 2008-02-26 15:00:07.000000000 +0100 @@ -5673,6 +5673,11 @@ sub main exit -1; } + if (-x "/sbin/service" && -f "/etc/init.d/lm_sensors" && + -f "/var/lock/subsys/lm_sensors") { + system("/sbin/service", "lm_sensors", "stop"); + } + initialize_kernel_version(); initialize_conf; initialize_proc_pci; @@ -5943,6 +5948,11 @@ EOT if (-x "/sbin/insserv" && -f "/etc/init.d/lm_sensors") { system("/sbin/insserv", "/etc/init.d/lm_sensors"); + } elsif (-x "/sbin/chkconfig" && -f "/etc/init.d/lm_sensors") { + system("/sbin/chkconfig", "lm_sensors", "on"); + if (-x "/sbin/service") { + system("/sbin/service", "lm_sensors", "start"); + } } else { print "You should now start the lm_sensors service to load the required\n". "kernel modules.\n\n"; --------------090500050005000409060300 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors --------------090500050005000409060300--