MIME-Version: 1.0
References: <20230114-bpf-v1-1-f836695a8b62@pefoley.com> <701abf4bbf5b7957a24d2f164c643e1d9f586fad.camel@gmail.com>
 <4dc6a571-8564-b38c-31df-0d9741dfc592@meta.com>
In-Reply-To: <4dc6a571-8564-b38c-31df-0d9741dfc592@meta.com>
From:   Peter Foley <pefoley2@pefoley.com>
Date:   Sun, 22 Jan 2023 20:28:15 -0800
Message-ID: <CAOFdcFOh_rNt3pfkNHPPeK=Kojro==Kpgmmu-VnLuaMYfFoiwg@mail.gmail.com>
Subject: Re: [PATCH] tools: bpf: Disable stack protector
To:     Yonghong Song <yhs@meta.com>
Cc:     Eduard Zingerman <eddyz87@gmail.com>,
        Quentin Monnet <quentin@isovalent.com>,
        Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Andrii Nakryiko <andrii@kernel.org>,
        Martin KaFai Lau <martin.lau@linux.dev>,
        Song Liu <song@kernel.org>, Yonghong Song <yhs@fb.com>,
        John Fastabend <john.fastabend@gmail.com>,
        KP Singh <kpsingh@kernel.org>,
        Stanislav Fomichev <sdf@google.com>,
        Hao Luo <haoluo@google.com>, Jiri Olsa <jolsa@kernel.org>,
        Nathan Chancellor <nathan@kernel.org>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Tom Rix <trix@redhat.com>, bpf@vger.kernel.org,
        linux-kernel@vger.kernel.org, llvm@lists.linux.dev
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, Jan 18, 2023 at 11:34 PM Yonghong Song <yhs@meta.com> wrote:
> On 1/18/23 11:28 AM, Eduard Zingerman wrote:
> >
> > While working on clang patch to disable stack protector
> > for BPF target I've noticed that there is an option to
> > disable default configuration file altogether [1]:
> >
> >    --no-default-config
> >
> > Should we consider it instead of -fno-stack-protector
> > to shield ourselves from any potential distro-specific
> > changes?
>
> Peter, could you help check whether adding --no-default-config works
> in your environment or not?
>
> >
> > [1] https://clang.llvm.org/docs/ClangCommandLineReference.html#cmdoption-clang-no-default-config

I guess I could, but I'm not convinced that's the right thing to do.
Ideally problems with distro-specific configs would cause loud
failures (like this one) and result in fixes like the changes being
made to upstream clang/gcc.
Simply unconditionally disabling distro configs seems to be the wrong
way to approach this and makes it less likely that future problems
will be reported in the first place.