From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:41622) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fpURp-0004SY-V2 for qemu-devel@nongnu.org; Tue, 14 Aug 2018 04:13:02 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fpURo-0006D8-QL for qemu-devel@nongnu.org; Tue, 14 Aug 2018 04:13:01 -0400 Date: Tue, 14 Aug 2018 16:12:52 +0800 From: Fam Zheng Message-ID: <20180814081252.GJ7549@lemon.usersys.redhat.com> References: <20180813023958.5352-1-famz@redhat.com> <20180813134206.GF4323@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180813134206.GF4323@localhost.localdomain> Subject: Re: [Qemu-devel] [PATCH v2] file-posix: Skip effectiveless OFD lock operations List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Kevin Wolf Cc: qemu-devel@nongnu.org, Max Reitz , qemu-block@nongnu.org On Mon, 08/13 15:42, Kevin Wolf wrote: > Am 13.08.2018 um 04:39 hat Fam Zheng geschrieben: > > If we know we've already locked the bytes, don't do it again; similarly > > don't unlock a byte if we haven't locked it. This doesn't change the > > behavior, but fixes a corner case explained below. > > > > Libvirt had an error handling bug that an image can get its (ownership, > > file mode, SELinux) permissions changed (RHBZ 1584982) by mistake behind > > QEMU. Specifically, an image in use by Libvirt VM has: > > > > $ ls -lhZ b.img > > -rw-r--r--. qemu qemu system_u:object_r:svirt_image_t:s0:c600,c690 b.img > > > > Trying to attach it a second time won't work because of image locking. > > And after the error, it becomes: > > > > $ ls -lhZ b.img > > -rw-r--r--. root root system_u:object_r:virt_image_t:s0 b.img > > > > Then, we won't be able to do OFD lock operations with the existing fd. > > In other words, the code such as in blk_detach_dev: > > > > blk_set_perm(blk, 0, BLK_PERM_ALL, &error_abort); > > > > can abort() QEMU, out of environmental changes. > > > > This patch is an easy fix to this and the change is regardlessly > > reasonable, so do it. > > > > Signed-off-by: Fam Zheng > > Thanks, applied to the block branch. Self-NACK. This breaks raw_abort_perm_update(). The extra bytes locked by raw_check_perm() are not tracked by s->perm (it is only updated in raw_set_perm()), thus will not get released. This patch is "misusing" s->perm and s->shared_perm. I'll revise the implementation and send another version together with dropping s->lock_fd. Fam