From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:45813) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VCBwe-0004UH-FK for qemu-devel@nongnu.org; Wed, 21 Aug 2013 13:11:49 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1VCBwZ-0006dM-7E for qemu-devel@nongnu.org; Wed, 21 Aug 2013 13:11:44 -0400 Received: from mx1.redhat.com ([209.132.183.28]:8488) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VCBwY-0006d7-VE for qemu-devel@nongnu.org; Wed, 21 Aug 2013 13:11:39 -0400 Message-ID: <5214F496.7050407@redhat.com> Date: Wed, 21 Aug 2013 19:10:46 +0200 From: Paolo Bonzini MIME-Version: 1.0 References: <1377103396-24307-1-git-send-email-pbonzini@redhat.com> <20130821164817.GC10012@redhat.com> <5214EFFF.3060804@redhat.com> <5214F2C0.8050203@redhat.com> In-Reply-To: <5214F2C0.8050203@redhat.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [RFC PATCH v2 0/3] Start fixing the pvpanic mess List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: pkrempa@redhat.com, marcel.a@redhat.com, libvir-list@redhat.com, hutao@cn.fujitsu.com, mst@redhat.com, qemu-devel@nongnu.org, armbru@redhat.com, rhod@redhat.com, kraxel@redhat.com, anthony@codemonkey.ws, lcapitulino@redhat.com, lersek@redhat.com, afaerber@suse.de -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Il 21/08/2013 19:02, Eric Blake ha scritto: > So, this boils down to a question of what SHOULD the valid states > for be? Generically, we want > destroy to not invalidate a guest, but also to > not instantiate a pvpanic device; since that covers the libvirt > defaults. We also want restart to not > invalidate a guest, but also to not instantiate a pvpanic device, > since so many existing guests have that setting thanks to > virt-install. > > Maybe that means we add attributes/sub-elements to that > express whether pvpanic device is permitted; and the absence of > that attribute means the status quo (the tag is > effectively ignored because without pvpanic device, there is no way > for libvirt to learn if a guest panicked). Or does it mean we > expose a new sub-element of , similar to how we have a > subelement that controls whether the memballoon device > is show to the guest, and just document that for qemu, > is a no-op without the subelement? Perhaps is better? Note that for s390 works without . Anyway yes, that's a possibility. More precisely, you could still use for internal errors even without having anything in (Xen conflates panics and internal errors). But then, "pause" and "ignore" are useful on-panic policies, yet they do not make sense for internal errors. Hence my suggestion to introduce a new element . We can still make a no-op without the appropriate element under . Paolo -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.19 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBAgAGBQJSFPSWAAoJEBvWZb6bTYbywvMP/3MKlED6Y69QM7Xt0aAKEtZp RmcKUt1m4MAuV91eBmN5/fv+ui0yKzrnZWz0mgF+V3eWCJdnEiewkGocetpx+Mw7 V4wuGVkYUWXV/O8A6x3thENOoxaHYO4OP8dUgkWQLGHXRNTljAd4iyVxBiIbITod fZvhVEbk8n9Mk3U61JxeMRB5PDjXRHcjgLgpR7htujpVBMTBFAsqxLzflsxsd/p7 UOZ0D3vk6m00DHdIzcJ5pc0dyqaylEaljs3Lf1MNbC/fN8I1sgrMWYMYnukU+moC GRKS6OB1ySeq2MkMwe73RimtE8M8MNmtVUKle94bmymPdGD3V+qKmBq6o7Hzd+b7 l8Rkht9gWP7Z4T22ZOYVqHpRXaDivkHuRCL2Va3BpyYv48Atyk/G77uB1uSaGTj+ ooRNoEqO61e49JOM3NiEYRI6Gl2YJ5O560j7dK59mQ6OIrLMtuN6Wo1ZiubdKCOa HB3e6qF0drAEQIBSdqQiU83F58ta634Rqy5R5kc1ad9cuiLMtUDPNbHlKsJtf+Th Dyu301fxt/1IIxPheoBQNVLRoXtAfoqpxM1nasjRphQqnULpnl7q7MipAclEUadQ Q7KE0YFPw38BYxl2FWhZOuTNI2kN921PGNiqYFFVpOWCf/aW/uqxU86LnJVSdvZs T9sRlLpVL4LTGHbFDooI =nCgT -----END PGP SIGNATURE-----