linux-hardening - Re: [RFC PATCH 09/21] arm64: Add CFI error handling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 5 May 2022 09:23:27 -0700
From:   Sami Tolvanen <samitolvanen@...gle.com>
To:     Mark Rutland <mark.rutland@....com>
Cc:     LKML <linux-kernel@...r.kernel.org>,
        Kees Cook <keescook@...omium.org>,
        Josh Poimboeuf <jpoimboe@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>, X86 ML <x86@...nel.org>,
        Catalin Marinas <catalin.marinas@....com>,
        Will Deacon <will@...nel.org>,
        Nathan Chancellor <nathan@...nel.org>,
        Nick Desaulniers <ndesaulniers@...gle.com>,
        Joao Moreira <joao@...rdrivepizza.com>,
        Sedat Dilek <sedat.dilek@...il.com>,
        Steven Rostedt <rostedt@...dmis.org>,
        linux-hardening@...r.kernel.org,
        linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
        llvm@...ts.linux.dev
Subject: Re: [RFC PATCH 09/21] arm64: Add CFI error handling

On Thu, May 5, 2022 at 8:45 AM Mark Rutland <mark.rutland@....com> wrote:
> It would be a bit nicer if we could encode the register index into the BRK
> immediate, i.e. allocate a range of 32 immediates (or 31 given BLR XZR is
> nonsensical), and have:
>
>         BRK #CFI_BRK_IMM + n
>
> ... where `n` is the Xn index.
>
> That way the kernel doesn't need to know the specific code sequence and
> wouldn't have to decode the instruction to find the relevant register -- we
> could determine that from the ESR alone. That would also avoid tying the
> compiler into a specific code sequence, and would allow that to change.
>
> Since the BRK immediate is 16 bits, we have enough space to also encode the
> index of the wB register, which would allow the kernel's BRK handler to recover
> and log the expected type value and the the value at the target of the branch
> (that latter we can recover from xN, so we don't need wA to be encoded into the
> immediate).

Sure, sounds like a good idea.

> ... does the compiler side of that sound possible?

Yes, this should be doable. I'll take a look and change this in the
next version.

Sami