From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:48865) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dcCRr-0001Xn-Li for qemu-devel@nongnu.org; Mon, 31 Jul 2017 11:17:36 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dcCRq-0004Lw-Nr for qemu-devel@nongnu.org; Mon, 31 Jul 2017 11:17:35 -0400 Date: Mon, 31 Jul 2017 11:17:24 -0400 From: Jeff Cody Message-ID: <20170731151724.GH6240@localhost.localdomain> References: <20170731125111.28052-1-kwolf@redhat.com> <20170731143809.GC22129@localhost.localdomain> <5e788182-b778-e027-2250-8e7c643786ae@amsat.org> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <5e788182-b778-e027-2250-8e7c643786ae@amsat.org> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH for-2.10] block/qapi: Remove redundat NULL check to silence Coverity List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Philippe =?iso-8859-1?Q?Mathieu-Daud=E9?= Cc: Kevin Wolf , peter.maydell@linaro.org, qemu-devel@nongnu.org, qemu-block@nongnu.org On Mon, Jul 31, 2017 at 11:54:57AM -0300, Philippe Mathieu-Daud=E9 wrote: > On 07/31/2017 11:38 AM, Jeff Cody wrote: > >On Mon, Jul 31, 2017 at 02:51:11PM +0200, Kevin Wolf wrote: > >>When skipping implicit nodes in bdrv_block_device_info(), we know tha= t > >>bs0 is always non-NULL; initially, because it's taken from a BdrvChil= d > > > >Not to mention, we deference bs0 in the chunk of code right above this= , so > >we'd segfault anyway if the initial value was NULL. >=20 > Yes, please move your assert before: >=20 > 137: if (bs0->drv && bs0->backing) { >=20 > Once there: > Reviewed-by: Philippe Mathieu-Daud=E9 > I don't think moving the assert() is the correct thing to do, as it is useful when iterating in the while() via backing_bs(). But maybe adding some asserts would be useful, if we are really concerned= . Looking at the code: 135 } 136 Maybe add an assert(bs0) here, as you suggest... 137 if (bs0->drv && bs0->backing) { 138 info->backing_file_depth++; 139 bs0 =3D bs0->backing->bs; But then maybe we want one here, too? Or maybe that is overkill :) 140 (*p_image_info)->has_backing_image =3D true; 141 p_image_info =3D &((*p_image_info)->backing_image); 142 } else { 143 break; 144 } 145 146 /* Skip automatically inserted nodes that the user isn't awar= e of for 147 * query-block (blk !=3D NULL), but not for query-named-block= -nodes */ 148 while (blk && bs0 && bs0->drv && bs0->implicit) { 149 bs0 =3D backing_bs(bs0); 150 } > > > >>and a BdrvChild never has a NULL bs, and after the first iteration > >>because implicit nodes always have a backing file. > >> > >>Remove the NULL check and add an assertion that the implicit node doe= s > >>indeed have a backing file. > >> > >>Signed-off-by: Kevin Wolf > > > > > >Reviewed-by: Jeff Cody > > > > > > > >>--- > >> block/qapi.c | 3 ++- > >> 1 file changed, 2 insertions(+), 1 deletion(-) > >> > >>diff --git a/block/qapi.c b/block/qapi.c > >>index d2b18ee9df..5f1a71f5d2 100644 > >>--- a/block/qapi.c > >>+++ b/block/qapi.c > >>@@ -145,8 +145,9 @@ BlockDeviceInfo *bdrv_block_device_info(BlockBack= end *blk, > >> /* Skip automatically inserted nodes that the user isn't aw= are of for > >> * query-block (blk !=3D NULL), but not for query-named-blo= ck-nodes */ > >>- while (blk && bs0 && bs0->drv && bs0->implicit) { > >>+ while (blk && bs0->drv && bs0->implicit) { > >> bs0 =3D backing_bs(bs0); > >>+ assert(bs0); > >> } > >> } > >>--=20 > >>2.13.3 > >> > >> > >