From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Mick Subject: Re: [PATCH 1/4] rbd: verify rbd image order value Date: Mon, 22 Oct 2012 15:43:55 -0700 Message-ID: <5085CC2B.3060404@inktank.com> References: <5085791C.9010205@inktank.com> <5085798C.8010605@inktank.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-pa0-f46.google.com ([209.85.220.46]:35048 "EHLO mail-pa0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932319Ab2JVWn7 (ORCPT ); Mon, 22 Oct 2012 18:43:59 -0400 Received: by mail-pa0-f46.google.com with SMTP id hz1so2275566pad.19 for ; Mon, 22 Oct 2012 15:43:58 -0700 (PDT) In-Reply-To: <5085798C.8010605@inktank.com> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Alex Elder Cc: ceph-devel@vger.kernel.org Personally I find that sizeof(int) doesn't really help here, but otherwise it's fine by me On 10/22/2012 09:51 AM, Alex Elder wrote: > This adds a verification that an rbd image's object order is > within the upper and lower bounds supported by this implementation. > > It must be at least 9 (SECTOR_SHIFT), because the Linux bio system > assumes that minimum granularity. > > It also must be less than 32 (at the moment anyway) because there > exist spots in the code that store the size of a "segment" (object > backing an rbd image) in a signed int variable, which can be 32 bits > including the sign. We should be able to relax this limit once > we've verified the code uses 64-bit types where needed. > > Note that the CLI tool already limits the order to the range 12-25. > > Signed-off-by: Alex Elder > --- > drivers/block/rbd.c | 10 ++++++++++ > 1 file changed, 10 insertions(+) > > diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c > index d032883..4734446 100644 > --- a/drivers/block/rbd.c > +++ b/drivers/block/rbd.c > @@ -533,6 +533,16 @@ static bool rbd_dev_ondisk_valid(struct > rbd_image_header_ondisk *ondisk) > if (memcmp(&ondisk->text, RBD_HEADER_TEXT, sizeof (RBD_HEADER_TEXT))) > return false; > > + /* The bio layer requires at least sector-sized I/O */ > + > + if (ondisk->options.order < SECTOR_SHIFT) > + return false; > + > + /* If we use u64 in a few spots we may be able to loosen this */ > + > + if (ondisk->options.order > 8 * sizeof (int) - 1) > + return false; > + > /* > * The size of a snapshot header has to fit in a size_t, and > * that limits the number of snapshots. >