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;
Subject: Re: [PATCH v2 00/14] x86: Add support for Clang CFI
To:     Sami Tolvanen <samitolvanen@google.com>
Cc:     X86 ML <x86@kernel.org>, Kees Cook <keescook@chromium.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>
References: <20210823171318.2801096-1-samitolvanen@google.com>
 <1706ee8e-c21c-f867-c0be-24814a92b853@redhat.com>
 <CABCJKufrpx9arM-hfX_bR-efO+13VBMFNBTe4ff036VEZi1LZQ@mail.gmail.com>
From:   Tom Stellard <tstellar@redhat.com>
Message-ID: <9349a92d-f2a7-9ee4-64db-98d30eadc505@redhat.com>
Date:   Tue, 24 Aug 2021 10:26:16 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.6.1
MIME-Version: 1.0
In-Reply-To: <CABCJKufrpx9arM-hfX_bR-efO+13VBMFNBTe4ff036VEZi1LZQ@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 8/23/21 10:20 AM, Sami Tolvanen wrote:
> On Mon, Aug 23, 2021 at 10:16 AM Tom Stellard <tstellar@redhat.com> wrote:
>>
>> On 8/23/21 10:13 AM, 'Sami Tolvanen' via Clang Built Linux wrote:
>>> This series adds support for Clang's Control-Flow Integrity (CFI)
>>> checking to x86_64. With CFI, the compiler injects a runtime
>>> check before each indirect function call to ensure the target is
>>> a valid function with the correct static type. This restricts
>>> possible call targets and makes it more difficult for an attacker
>>> to exploit bugs that allow the modification of stored function
>>> pointers. For more details, see:
>>>
>>>     https://clang.llvm.org/docs/ControlFlowIntegrity.html
>>>
>>> Version 2 depends on Clang >=14, where we fixed the issue with
>>> referencing static functions from inline assembly. Based on the
>>> feedback for v1, this version also changes the declaration of
>>> functions that are not callable from C to use an opaque type,
>>> which stops the compiler from replacing references to them. This
>>> avoids the need to sprinkle function_nocfi() macros in the kernel
>>> code.
>>
>> How invasive are the changes in clang 14 necessary to make CFI work?
>> Would it be possible to backport them to LLVM 13?
> 
> I'm not sure what the LLVM backport policy is, but this specific fix
> was quite simple:
> 
> https://reviews.llvm.org/rG7ce1c4da7726
> 

That looks like something we could backport, I filed a bug to track
the backport: https://bugs.llvm.org/show_bug.cgi?id=51588.

Do you have any concerns about backporting it or do you think it's pretty
safe?

-Tom


> Sami
>