From:   Florian Weimer <fweimer@redhat.com>
To:     Andrei Vagin <avagin@gmail.com>
Cc:     Andy Lutomirski <luto@kernel.org>, linux-arch@vger.kernel.org,
        Linux API <linux-api@vger.kernel.org>,
        linux-x86_64@vger.kernel.org, kernel-hardening@lists.openwall.com,
        linux-mm@kvack.org, "the arch/x86 maintainers" <x86@kernel.org>,
        musl@lists.openwall.com, libc-alpha@sourceware.org,
        LKML <linux-kernel@vger.kernel.org>,
        Dave Hansen <dave.hansen@intel.com>,
        Kees Cook <keescook@chromium.org>
Subject: Re: [PATCH v2] x86: Implement arch_prctl(ARCH_VSYSCALL_CONTROL) to
 disable vsyscall
References: <878rwkidtf.fsf@oldenburg.str.redhat.com>
        <CANaxB-xpQr1mUUvWK5a53q49VK_HvR4Pws_NGKGa8-jihxkc_A@mail.gmail.com>
Date:   Mon, 27 Dec 2021 18:40:38 +0100
In-Reply-To: <CANaxB-xpQr1mUUvWK5a53q49VK_HvR4Pws_NGKGa8-jihxkc_A@mail.gmail.com>
        (Andrei Vagin's message of "Mon, 27 Dec 2021 08:49:38 -0800")
Message-ID: <87o8520wvd.fsf@oldenburg.str.redhat.com>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/27.2 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

* Andrei Vagin:

>> diff --git a/arch/x86/entry/vsyscall/vsyscall_64.c b/arch/x86/entry/vsyscall/vsyscall_64.c
>> index fd2ee9408e91..8eb3bcf2cedf 100644
>> --- a/arch/x86/entry/vsyscall/vsyscall_64.c
>> +++ b/arch/x86/entry/vsyscall/vsyscall_64.c
>> @@ -174,6 +174,12 @@ bool emulate_vsyscall(unsigned long error_code,
>>
>>         tsk = current;
>>
>> +       if (tsk->mm->context.vsyscall_disabled) {
>> +               warn_bad_vsyscall(KERN_WARNING, regs,
>> +                                 "vsyscall after lockout (exploit attempt?)");
>
> I don't think that we need this warning message. If we disable
> vsyscall, its address range is not differ from other addresses around
> and has to be handled the same way. For example, gVisor or any other
> sandbox engines may want to emulate vsyscall, but the kernel log will
> be full of such messages.

But with vsyscall=none, such messages are already printed.  That's why I
added the warning for the lockout case as well.

>> diff --git a/tools/testing/selftests/x86/vsyscall_control.c b/tools/testing/selftests/x86/vsyscall_control.c
>> new file mode 100644
>> index 000000000000..ee966f936c89
>> --- /dev/null
>> +++ b/tools/testing/selftests/x86/vsyscall_control.c
>
> I would move the test in a separate patch...

I can do that if it simplifies matters.

Thanks,
Florian