Received-SPF: pass (google.com: domain of linux-crypto+bounces-1822-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
MIME-Version: 1.0
References: <20240202101311.it.893-kees@kernel.org> <20240202101642.156588-2-keescook@chromium.org>
 <CANpmjNPPbTNPJfM5MNE6tW-jCse+u_RB8bqGLT3cTxgCsL+x-A@mail.gmail.com> <202402020405.7E0B5B3784@keescook>
In-Reply-To: <202402020405.7E0B5B3784@keescook>
From: Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>
Date: Fri, 2 Feb 2024 17:08:48 +0100
Message-ID: <CANiq72ku9wsHtnPAh5G71Y_pbsftrPPyV5wmDCcZRM+WB6KVjA@mail.gmail.com>
Subject: Re: [PATCH v2 2/6] ubsan: Reintroduce signed and unsigned overflow sanitizers
To: Kees Cook <keescook@chromium.org>
Cc: Marco Elver <elver@google.com>, linux-hardening@vger.kernel.org, 
	Justin Stitt <justinstitt@google.com>, Miguel Ojeda <ojeda@kernel.org>, 
	Nathan Chancellor <nathan@kernel.org>, Nick Desaulniers <ndesaulniers@google.com>, 
	Peter Zijlstra <peterz@infradead.org>, Hao Luo <haoluo@google.com>, 
	Przemek Kitszel <przemyslaw.kitszel@intel.com>, Fangrui Song <maskray@google.com>, 
	Masahiro Yamada <masahiroy@kernel.org>, Nicolas Schier <nicolas@fjasle.eu>, 
	Bill Wendling <morbo@google.com>, Andrey Konovalov <andreyknvl@gmail.com>, 
	Jonathan Corbet <corbet@lwn.net>, x86@kernel.org, linux-kernel@vger.kernel.org, 
	linux-kbuild@vger.kernel.org, llvm@lists.linux.dev, linux-doc@vger.kernel.org, 
	netdev@vger.kernel.org, linux-crypto@vger.kernel.org, 
	kasan-dev@googlegroups.com, linux-acpi@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Fri, Feb 2, 2024 at 1:17=E2=80=AFPM Kees Cook <keescook@chromium.org> wr=
ote:
>
> Perhaps I should hold off on bringing the unsigned sanitizer back? I was
> hoping to work in parallel with the signed sanitizer, but maybe this
> isn't the right approach?

If you can do anything to keep it in-tree, I think it would be nice so
that others can easily use it to test the tooling and to start to
clean up cases. A per-subsystem opt-in like Marco says could be a way,
and you could perhaps do one very small subsystem or similar to see
how it would look like.

Something that could also help would be to split the cases even
further (say, only overflows and not underflows), but is that a
possibility with the current tooling?

Thanks for working on this, Kees!

Cheers,
Miguel