Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:8 as permitted sender) client-ip=2620:137:e000::3:8;
MIME-Version: 1.0
References: <20231009224347.2076221-1-mmaurer@google.com> <20231010081220.GD377@noisy.programming.kicks-ass.net>
In-Reply-To: <20231010081220.GD377@noisy.programming.kicks-ass.net>
From:   Matthew Maurer <mmaurer@google.com>
Date:   Tue, 10 Oct 2023 07:06:32 -0700
Message-ID: <CAGSQo01=NRosJCzRCWcqZfvdHzH+d6dThN02SpvkF2EHsFXD-g@mail.gmail.com>
Subject: Re: [PATCH] x86: Enable IBT in Rust if enabled in C
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        Dave Hansen <dave.hansen@linux.intel.com>, x86@kernel.org,
        Miguel Ojeda <ojeda@kernel.org>,
        Alex Gaynor <alex.gaynor@gmail.com>,
        Wedson Almeida Filho <wedsonaf@gmail.com>,
        "H. Peter Anvin" <hpa@zytor.com>,
        Boqun Feng <boqun.feng@gmail.com>, Gary Guo <gary@garyguo.net>,
        =?UTF-8?Q?Bj=C3=B6rn_Roy_Baron?= <bjorn3_gh@protonmail.com>,
        Benno Lossin <benno.lossin@proton.me>,
        Andreas Hindborg <a.hindborg@samsung.com>,
        Alice Ryhl <aliceryhl@google.com>,
        linux-kernel@vger.kernel.org, rust-for-linux@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Oct 10, 2023 at 1:12=E2=80=AFAM Peter Zijlstra <peterz@infradead.or=
g> wrote:
>
> On Mon, Oct 09, 2023 at 10:42:54PM +0000, Matthew Maurer wrote:
> > These flags are not made conditional on compiler support because at the
> > moment exactly one version of rustc supported, and that one supports
> > these flags.
> >
> > Building without these additional flags will manifest as objtool
> > printing a large number of errors about missing ENDBR and if CFI is
> > enabled (not currently possible) will result in incorrectly structured
> > function prefixes.
>
> Well, I would also imagine running it on actual IBT enabled hardware
> will get you a non-booting kernel.
>
> > Signed-off-by: Matthew Maurer <mmaurer@google.com>
> > ---
> >
> > Split out the IBT additions as per
> > https://lkml.kernel.org/linux-fsdevel/CANiq72kK6ppBE7j=3Dz7uua1cJMKaLoR=
5U3NUAZXT5MrNEs9ZhfQ@mail.gmail.com/
> >
> > arch/x86/Makefile | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/arch/x86/Makefile b/arch/x86/Makefile
> > index 5bfe5caaa444..941f7abf6dbf 100644
> > --- a/arch/x86/Makefile
> > +++ b/arch/x86/Makefile
> > @@ -81,6 +81,7 @@ ifeq ($(CONFIG_X86_KERNEL_IBT),y)
> >  #   https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D104816
> >  #
> >  KBUILD_CFLAGS +=3D $(call cc-option,-fcf-protection=3Dbranch -fno-jump=
-tables)
> > +KBUILD_RUSTFLAGS +=3D -Zcf-protection=3Dbranch -Zno-jump-tables
>
> One question, -Zcf-protection=3Dbranch, will that ever emit NOTRACK
> prefix? The kernel very explicitly does not support (enable) NOTRACK.
rustc does this via LLVM, so its code generation works very similarly to cl=
ang.
It does not create its own explicit NOTRACKs, but LLVM will by default
with just -Zcf-protection-branch.
I've linked a godbolt showing that at least for the basic case, your
no-jump-tables approach from clang ports over.
https://godbolt.org/z/bc4n6sq5q
Whether rust generates NOTRACK should end up being roughly equivalent
to whether clang generates it, and if LLVM gains a code generation
flag for NOTRACK being disallowed some day, we can pass that through
as well.
>
> >  else
> >  KBUILD_CFLAGS +=3D $(call cc-option,-fcf-protection=3Dnone)
> >  endif
> > --
> > 2.42.0.609.gbb76f46606-goog
> >