Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Andy Lutomirski <luto@amacapital.net>
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH 05/15] x86: Implement function_nocfi
Date:   Sat, 17 Apr 2021 08:48:58 -0700
Message-Id: <23831847-4613-4CD7-9DBE-D0DFE5F9AFF0@amacapital.net>
References: <a8653ebf0a52498888f728524ebaa110@AcuMS.aculab.com>
Cc:     Kees Cook <keescook@chromium.org>,
        Andy Lutomirski <luto@kernel.org>,
        Borislav Petkov <bp@alien8.de>,
        Sami Tolvanen <samitolvanen@google.com>,
        X86 ML <x86@kernel.org>, Josh Poimboeuf <jpoimboe@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Nathan Chancellor <nathan@kernel.org>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Sedat Dilek <sedat.dilek@gmail.com>,
        linux-hardening@vger.kernel.org,
        LKML <linux-kernel@vger.kernel.org>,
        clang-built-linux <clang-built-linux@googlegroups.com>
In-Reply-To: <a8653ebf0a52498888f728524ebaa110@AcuMS.aculab.com>
To:     David Laight <David.Laight@aculab.com>
Precedence: bulk


> On Apr 17, 2021, at 7:20 AM, David Laight <David.Laight@aculab.com> wrote:=

>=20
> =EF=BB=BFFrom: Kees Cook
>> Sent: 16 April 2021 23:28
>>=20
>>> On Fri, Apr 16, 2021 at 03:06:17PM -0700, Andy Lutomirski wrote:
>>> On Fri, Apr 16, 2021 at 3:03 PM Borislav Petkov <bp@alien8.de> wrote:
>>>>=20
>>>> On Fri, Apr 16, 2021 at 02:49:23PM -0700, Sami Tolvanen wrote:
>>>>> __nocfi only disables CFI checking in a function, the compiler still
>>>>> changes function addresses to point to the CFI jump table, which is
>>>>> why we need function_nocfi().
>>>>=20
>>>> So call it __func_addr() or get_function_addr() or so, so that at least=

>>>> it is clear what this does.
>>>>=20
>>>=20
>>> This seems backwards to me.  If I do:
>>>=20
>>> extern void foo(some signature);
>>>=20
>>> then I would, perhaps naively, expect foo to be the actual symbol that
>>> gets called
>>=20
>> Yes.
>>=20
>>> and for the ABI to be changed to do the CFI checks.
>>=20
>> Uh, no? There's no ABI change -- indirect calls are changed to do the
>> checking.
>>=20
>>> The
>>> foo symbol would point to whatever magic is needed.
>>=20
>> No, the symbol points to the jump table entry. Direct calls get minimal
>> overhead and indirect calls can add the "is this function in the right
>> table" checking.
>=20
>=20
> Isn't this a bit like one of the PPC? ABI where function addresses
> aren't (always) the entry point.
> IIRC is causes all sorts of obscure grief.
>=20
> I'd also like to know how indirect calls are actually expected to work.
> The whole idea is that the potential targets might be in a kernel module
> that is loaded at run time.
>=20
> Scanning a list of possible targets has to be a bad design decision.
>=20
> If you are trying to check that the called function has the right
> prototype, stashing a hash of the prototype (or a known random number)
> before the entry point would give most of the benefits without most
> of the costs.
> The linker could be taught to do the same test (much like name mangling
> but without the crap user experience).
>=20
> That scheme only has the downside of a data cache miss and (hopefully)
> statically predicted correct branch (hmm - except you don't want to
> speculatively execute the wrong function) and polluting the data cache
> with code.

I admit I was quite surprised by the actual CFI implementation. I would have=
 expected a CFI=E2=80=99d function pointer to actually point to a little des=
criptor that contains a hash of the function=E2=80=99s type.  The whole bit v=
ector thing seems quite inefficient.

>=20
> This all seems like a ploy to force people to buy faster cpus.
>=20
>    David
>=20
> -
> Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1=
PT, UK
> Registration No: 1397386 (Wales)
>=20