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;
From:   Thomas Gleixner <tglx@linutronix.de>
To:     Sami Tolvanen <samitolvanen@google.com>, x86@kernel.org
Cc:     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, linux-kernel@vger.kernel.org,
        clang-built-linux@googlegroups.com,
        Sami Tolvanen <samitolvanen@google.com>
Subject: Re: [PATCH 04/15] static_call: Use global functions for the self-test
In-Reply-To: <20210416203844.3803177-5-samitolvanen@google.com>
References: <20210416203844.3803177-1-samitolvanen@google.com> <20210416203844.3803177-5-samitolvanen@google.com>
Date:   Fri, 16 Apr 2021 23:37:16 +0200
Message-ID: <87v98lud0z.ffs@nanos.tec.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On Fri, Apr 16 2021 at 13:38, Sami Tolvanen wrote:

> With CONFIG_CFI_CLANG, the compiler renames static functions. This
> breaks static_call users using static functions, because the current
> implementation assumes functions have stable names by hardcoding them
> in inline assembly. Make the self-test functions global to prevent the
> compiler from renaming them.

Sorry, no.

> Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
> ---
>  kernel/static_call.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/static_call.c b/kernel/static_call.c
> index 723fcc9d20db..d09f500c2d2a 100644
> --- a/kernel/static_call.c
> +++ b/kernel/static_call.c
> @@ -503,12 +503,12 @@ long __static_call_return0(void)
>  
>  #ifdef CONFIG_STATIC_CALL_SELFTEST
>  
> -static int func_a(int x)
> +int func_a(int x)
>  {
>  	return x+1;
>  }
>  
> -static int func_b(int x)
> +int func_b(int x)
>  {
>  	return x+2;
>  }

Did you even compile that?

Global functions without a prototype are generating warnings, but we can
ignore them just because of sekurity, right?

Aside of that polluting the global namespace with func_a/b just to work
around a tool shortcoming is beyond hillarious.

Fix the tool not the perfectly correct code.

Thanks,

        tglx