From: Steve Dickson Subject: Re: [PATCH 13/19] NFSD: Server implementation of MAC Labeling Date: Tue, 23 Apr 2013 15:51:24 -0400 Message-ID: <5176E63C.50303@RedHat.com> References: <1364939160-20874-1-git-send-email-SteveD@redhat.com> <1364939160-20874-14-git-send-email-SteveD@redhat.com> <20130410151953.GC24404@pad.fieldses.org> <5176E3BB.4020508@RedHat.com> <20130423194404.GH20622@pad.fieldses.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Trond Myklebust , "David P. Quigley" , Linux NFS list , Linux FS devel list , Linux Security List , SELinux List To: "J. Bruce Fields" Return-path: In-Reply-To: <20130423194404.GH20622@pad.fieldses.org> Sender: linux-security-module-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org On 23/04/13 15:44, J. Bruce Fields wrote: > On Tue, Apr 23, 2013 at 03:40:43PM -0400, Steve Dickson wrote: >> >> >> On 10/04/13 11:19, J. Bruce Fields wrote: >>> On Tue, Apr 02, 2013 at 05:45:54PM -0400, Steve Dickson wrote: >>>> +__be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp, >>>> + struct nfs4_label *label) >>>> +{ >>>> + return -EOPNOTSUPP; >>> >>> Anything that returns a __be32 is expecting nfserr_something, not >>> -ERRNO. >>> >> I think it makes sense when label are not configured for this to >> return nfserr_notsupp which will then get turned into a -EOPNOTSUPP, >> agreed? > > Right though as I've said on previous review really this should be > caught by fs/nfsd/nfs4proc.c:check_attr_support() which is called just > before nfsd4_set_nfs4_label() in nfsd4_setattr(). It will be once I make nfsd_suppattrs2() only give v4.2 attrs on 4.2 mounts... steved. > > --b. >