On 08/28/2008 04:06 AM, John Summerfield wrote: > Steffen Maier wrote: > >> We should reboot, if (and only if) the reipl configuration went fine. >> Then the customer would see improved usability, since the system reboots >> automatically from the correct device. > > At present, one can specify whether to automatically reboot (on any > platform) or sit there looking stupid. It shouldn't be hard to add an > option to halt instead at that point. Linuxrc.s390 (or probably init respectively on the other platforms) already handles the choice between reboot and halt. There's a do_reboot and a do_shutdown function, which are called as trap on receiving SIGUSR2 or SIGUSR1 respectively. During installation, the user logs in through ssh, which starts the loader and in turn anaconda. The configuration of reipl and hence the detection if it was successful happens in anaconda. Loader waits for its child anaconda to exit. On anaconda's exit, loader decides how to proceed. On s390, loader has been sending a SIGUSR1 to linuxrc.s390 to initiate a halt unconditionally so far. This is exactly the place where Mark's posted patch introduces the functionality of doing a dependent reboot or halt: Anaconda writes the number 1 or 2 as ASCII text into the file /var/run/sigusr depending on the success of reipl configuration. Loader reads this file and sends the appropriate signal to linuxrc.s390. So this is our proposed and tested mechanism for doing a dependent reboot or halt. Of course, we are open to suggestions on how to best model this "backwards interprocess communication" between anaconda and loader or linuxrc.s390 respectively. > Maybe on S390, the choice would be > between reboot and halt? Or should there be three choices for all? Exactly, it's just between the two choices of reboot and halt. Steffen Linux on System z Development IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschäftsführung: Erich Baier Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294 _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/anaconda-devel-list
