=>Testing Environment: Qemux86 sato ,Qemux86 sdk Image from http://yoctobuild01.jf.intel.com/milestone/20101014/qemu/qemux86/poky-image- sdk-qemux86-20101014212122.rootfs.ext3 http://yoctobuild01.jf.intel.com/milestone/20101014/qemu/qemux86/poky-image-sato-qemux86-20101014213758.rootfs.ext3 =>Bug Descript fail to reboot =>Reproduce Steps: 1.run poky-qemu; 2.launch terminal; 3.run command "reboot".
Is this failure to reboot expected following this commit: "Scott has made sure qemu will terminate if the guest reboots in commit: http://git.pokylinux.org/cgit.cgi/poky/commit/?id=9a2ae44380a5ef81fb099075af99351db06a3252 The -no-reboot flag is unconditionally applied in the poky-qemu-internal; removing it allows qemux86 to reboot as expected.
Since bugs #100 and this one seemed to be in conflict with each other, I discussed the matter with Richard on Jabber, who feels that reboot capability for x86 is not a requirement: (09:23:52 AM) sgarman_laptop: RP: I have a question regarding the bugfix I submitted for bug #100, which added -no-reboot to the qemu default options. Apparently all the non-x86 qemu architectures already had -no-reboot in their default options (set in poky-qemu-internal), so by making it the universal default I only added it to the x86 arches. But now bug #475 has been added, which states that it is a bug that the x86 arches cannot reboot. So I need to know what the intended behavior is. (09:25:05 AM) RP: sgarman_laptop: Does that bug mean that qemu exits when a reboot is requested? (09:25:16 AM) sgarman_laptop: RP: correct (09:25:21 AM) RP: If it does exit ok, I'd probably call it NOTABUG (09:25:40 AM) sgarman_laptop: RP: ok, just verifying that the fix I did for #100 was actually what was intended (09:26:17 AM) tom.zanussi: RP: rebooting works ok in qemux86, so that will remove it (09:26:55 AM) tom.zanussi: RP: I mean, everything works fine without the -no-reboot flag on qemux86, so why remove it? (09:27:24 AM) tom.zanussi: RP: It's just a problem for arm and ppc, from the description in bug 100 (09:29:18 AM) RP: tom.zanussi: I'd really like the qemu* machines to have consistent behaviour (09:29:35 AM) RP: If we define it that both reboot and halt exit qemu, I think there will be less surprises for people
Verified we have the intended fix, 2010-12-13