From: Mimi Zohar Subject: Re: [PATCH] security, crypto: LLVMLinux: Remove VLAIS from ima_crypto.c Date: Mon, 08 Sep 2014 09:43:41 -0400 Message-ID: <1410183821.8054.27.camel@dhcp-9-2-203-236.watson.ibm.com> References: <1409958360-8061-1-git-send-email-behanw@converseincode.com> <540A590F.1030407@converseincode.com> <540BBD9E.6050708@converseincode.com> <540D73C9.3000906@samsung.com> <540DA051.2030701@converseincode.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Dmitry Kasatkin , Thomas Gleixner , james.l.morris@oracle.com, linux-ima-devel@lists.sourceforge.net, linux-ima-user@lists.sourceforge.net, linux-kernel@vger.kernel.org, linux-security-module@vger.kernel.org, serge@hallyn.com, torvalds@linux-foundation.org, Mark Charlebois , Jan-Simon =?ISO-8859-1?Q?M=F6ller?= , Linux Crypto Mailing List , Herbert Xu To: Behan Webster Return-path: In-Reply-To: <540DA051.2030701@converseincode.com> Sender: linux-security-module-owner@vger.kernel.org List-Id: linux-crypto.vger.kernel.org On Mon, 2014-09-08 at 07:25 -0500, Behan Webster wrote:=20 > On 09/08/14 04:15, Dmitry Kasatkin wrote: > > On 07/09/14 05:06, Behan Webster wrote: > >> On 09/06/14 03:11, Thomas Gleixner wrote: > >>> On Fri, 5 Sep 2014, Behan Webster wrote: > >>>> On 09/05/14 17:18, Thomas Gleixner wrote: > >>>>>> Signed-off-by: Behan Webster > >>>>>> Signed-off-by: Mark Charlebois > >>>>>> Signed-off-by: Jan-Simon M=C3=B6ller > >>>>> This SOB chain is completely ass backwards. See Documentation/.= =2E. > >>>> "The Signed-off-by: tag indicates that the signer was involved i= n the > >>>> development of the patch, or that he/she was in the patch's deli= very > >>>> path." > >>>> > >>>> All three of us were involved. Does that not satisfy this rule? > >>> No. Read #12 > >>> > >>> The sign-off is a simple line at the end of the explanation for t= he > >>> patch, which certifies that you wrote it or otherwise have the ri= ght to > >>> pass it on as an open-source patch. > >>> > >>> So the above chain says: > >>> > >>> Written-by: Behan > >>> Passed-on-by: Mark > >>> Passed-on-by: Jan > >>> > >>> That would be correct if you sent the patch to Mark, Mark sent it= to > >>> Jan and Jan finally submitted it to LKML. > >> I suppose "Reviewed-by" is probably more appropriate for the last = 2 > >> then. Will fix. > >> > >>>>>> - struct { > >>>>>> - struct shash_desc shash; > >>>>>> - char ctx[crypto_shash_descsize(tfm)]; > >>>>>> - } desc; > >>>>>> + char desc[sizeof(struct shash_desc) + > >>>>>> + crypto_shash_descsize(tfm)] CRYPTO_MINALIGN_ATTR; > >>>>>> + struct shash_desc *shash =3D (struct shash_desc *)desc; > >>>>> That anon struct should have never happened in the first place. > >>>> Sadly this is a design pattern used in many places through out t= he > >>>> kernel, and > >>>> appears to be fundamental to the crypto system. I was advised *n= ot* > >>>> to change > >>>> it, so we haven't. > >>>> > >>>> I agree that it's not a good practice. > >>>> > >>>>> Not > >>>>> your problem, but you are not making it any better. You replace= open > >>>>> coded crap with even more unreadable crap. > >>>>> > >>>>> Whats wrong with > >>>>> > >>>>> SHASH_DESC_ON_STACK(shash, tfm); > >>>> Nothing is wrong with that. I would have actually preferred that= =2E > >>>> But it would > >>>> have fundamentally changed a lot more code. > >>> Errm. Why is > >>> > >>> #define SHASH_DESC_ON_STACK(shash, tfm) \ > >>> char __shash[sizeof(.....)]; \ > >>> struct shash_desc *shash =3D (struct shash_desc *) __shash > >>> > >>> requiring more fundamental than open coding the same thing a gazi= llion > >>> times. You still need to change ALL usage sides of the anon struc= t. > >>> > >>> So in fact you could avoid the whole code change by making it > >>> > >>> SHASH_DESC_ON_STACK(desc, tfm); > >>> > >>> and do the anon struct or a proper struct magic in the macro. > >> I see. I thought you meant a more fundamental change to the crypto > >> system API. My misunderstanding. > >> > >> Ironically we tried to stay away from macros since the last time w= e > >> tried to replace VLAIS using macros (we've attempted patches to re= move > >> VLAIS a few times) we were told *not* to hide the implementation w= ith > >> macro magic. Though, to be fair, we were using more pointer math i= n > >> our other macro-based effort, and the non-crypto uses of VLAIS are= a > >> lot more complex to replace. > >> > >> Like I said I'm actually a fan of hiding ugliness in macros. Will = fix. > >> > >> Again, thanks for the feedback, > >> > >> Behan > >> > > Hi, > > > > Despite if it is crap or not, it was said already in this thread, > > following "design pattern" is heavily used through out the kernel -= by > > crypto core itself and by many widely used clients. > > > > struct { > > struct shash_desc shash; > > char ctx[crypto_shash_descsize(tfm)]; > > } desc; > > > > > > My question why do you want to change this particular piece of code= ? > Because it employs Variable Length Arrays in Structs. A construct whi= ch=20 > is explicitly forbidden by the C standard (C89, C99, C11). Because th= e=20 > vast majority of kernel developers I've talked to about this have bee= n=20 > unaware of the use of VLAIS in the kernel and most find its use=20 > objectionable (there is a similar objection to the use of nested=20 > functions). Because implementing VLAIS in a compiler can severely imp= act=20 > the generated instructions surrounding its use, which is why most=20 > compilers don't implement VLAIS as a feature. Because using such a=20 > construct precludes standards based compilers from competing with the= =20 > incumbent (my interest is enabling the use of clang and LLVM based=20 > technologies as a toolchain choice to compile and develop the kernel)= =2E >=20 > > What about rest of the kernel? > The LLVMLinux project is systematically working to remove the use of=20 > VLAIS from the kernel (already removed from ext4, USB Gadget, netfilt= er,=20 > mac802.11, apparmor, bluetooth, etc). Users of the crpyto subsystem a= re=20 > one of the last and heaviest users of VLAIS. >=20 > > To solve your problem you probably need to change everything. > Essentially yes. Though I like to think of it as finding alternatives= to=20 > where ever it is still used. "Changing everything" implies much large= r=20 > changes which aren't necessary in most cases. Sometimes the alternati= ve=20 > is merely using a flexible member (zero length array at the end of th= e=20 > struct, instead of a VLA in the struct). In several places several VL= As=20 > are used in the same struct. And recently we found that exofs is usin= g a=20 > VLAIS inside VLAIS (second order VLAIS) in one of its structures. So = not=20 > finished yet. >=20 > > If we are going to change it and introduce any macros, it is better= to > > do with the guidance from crypto folks. > Absolutely. Most of the crypto related patches have been sent to them= =2E I=20 > am absolutely looking for their input. >=20 > > I added CC:linux-crypto@vger.kernel.org mailing list and Herbert Xu= , > > crypto maintainer. > I suppose this specific patch may not have CC that list. However, mos= t=20 > of the other VLAIS removal patches were copied to linux-crypto, Herbe= rt=20 > Xu and David Miller. Behan, thank you for the explanation. The same snippet of code used here, and elsewhere in the kernel, is taken from the crypto subsystem. Once it is resolved in the crypto subsystem, the same solution should b= e propogated. Mimi -- To unsubscribe from this list: send the line "unsubscribe linux-securit= y-module" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html