From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:48051) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Zf9Uo-0005Xl-8m for qemu-devel@nongnu.org; Thu, 24 Sep 2015 12:35:47 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Zf9Un-0001oJ-9M for qemu-devel@nongnu.org; Thu, 24 Sep 2015 12:35:46 -0400 From: Markus Armbruster References: <1443017389-7869-1-git-send-email-armbru@redhat.com> <1443017389-7869-7-git-send-email-armbru@redhat.com> <20150924153717.GA30243@thinpad.lan.raisama.net> Date: Thu, 24 Sep 2015 18:35:42 +0200 In-Reply-To: <20150924153717.GA30243@thinpad.lan.raisama.net> (Eduardo Habkost's message of "Thu, 24 Sep 2015 12:37:17 -0300") Message-ID: <871tdnbwrl.fsf@blackfin.pond.sub.org> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [Qemu-devel] [PATCH v2 6/7] qdev: Protect device-list-properties against broken devices List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eduardo Habkost Cc: Peter Maydell , thuth@redhat.com, Peter Crosthwaite , qemu-stable@nongnu.org, qemu-devel@nongnu.org, Christian Borntraeger , Alexander Graf , qemu-ppc@nongnu.org, Antony Pavlov , stefanha@redhat.com, Cornelia Huck , Paolo Bonzini , Alistair Francis , afaerber@suse.de, Li Guang , Richard Henderson Eduardo Habkost writes: > On Wed, Sep 23, 2015 at 04:09:48PM +0200, Markus Armbruster wrote: >> Several devices don't survive object_unref(object_new(T)): they crash >> or hang during cleanup, or they leave dangling pointers behind. >> >> This breaks at least device-list-properties, because >> qmp_device_list_properties() needs to create a device to find its >> properties. Broken in commit f4eb32b "qmp: show QOM properties in >> device-list-properties", v2.1. Example reproducer: >> >> $ qemu-system-aarch64 -nodefaults -display none -machine none -S >> -qmp stdio >> {"QMP": {"version": {"qemu": {"micro": 50, "minor": 4, "major": >> 2}, "package": ""}, "capabilities": []}} >> { "execute": "qmp_capabilities" } >> {"return": {}} >> { "execute": "device-list-properties", "arguments": { >> "typename": "pxa2xx-pcmcia" } } >> qemu-system-aarch64: /home/armbru/work/qemu/memory.c:1307: >> memory_region_finalize: Assertion `((&mr->subregions)->tqh_first == >> ((void *)0))' failed. >> Aborted (core dumped) >> [Exit 134 (SIGABRT)] >> >> Unfortunately, I can't fix the problems in these devices right now. >> Instead, add DeviceClass member cannot_even_create_with_object_new_yet >> to mark them: >> > [...] >> * Dangling pointers: most CPUs, plus "allwinner-a10", "digic", >> "fsl,imx25", "fsl,imx31", "xlnx,zynqmp", because they create such >> CPUs > > TYPE_TILEGX_CPU doesn't have cannot_even_create_with_object_new_yet set, > but it calls cpu_exec_init() on instance_init too. Rats, missed one! Test survived the dangling pointer somehow. Will fix in v3. > For the remaining changes: > > Reviewed-by: Eduardo Habkost Thanks!