Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 13 Jan 2020 21:05:05 -0800
User-Agent: K-9 Mail for Android
In-Reply-To: <20200114035858.GA2536335@rani.riverdale.lan>
References: <20200113161310.GA191743@rani.riverdale.lan> <20200113195337.604646-1-nivedita@alum.mit.edu> <202001131750.C1B8468@keescook> <261ae869-4169-296e-f673-5c08ff34bdde@zytor.com> <20200114035858.GA2536335@rani.riverdale.lan>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: [PATCH v3] x86/vmlinux: Fix vmlinux.lds.S with pre-2.23 binutils
To:     Arvind Sankar <nivedita@alum.mit.edu>
CC:     Kees Cook <keescook@chromium.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        x86@kernel.org, linux-kernel@vger.kernel.org,
        Thomas Lendacky <Thomas.Lendacky@amd.com>,
        Mauro Rossi <issor.oruam@gmail.com>,
        Michael Matz <matz@suse.de>
From:   hpa@zytor.com
Message-ID: <F7B2A25D-E6A2-4637-9802-8AEC543916E9@zytor.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On January 13, 2020 7:58:59 PM PST, Arvind Sankar <nivedita@alum=2Emit=2Eed=
u> wrote:
>On Mon, Jan 13, 2020 at 05:57:23PM -0800, H=2E Peter Anvin wrote:
>> On 2020-01-13 17:53, Kees Cook wrote:>>
>> >> diff --git a/arch/x86/kernel/vmlinux=2Elds=2ES
>b/arch/x86/kernel/vmlinux=2Elds=2ES
>> >> index 3a1a819da137=2E=2Ebad4e22384dc 100644
>> >> --- a/arch/x86/kernel/vmlinux=2Elds=2ES
>> >> +++ b/arch/x86/kernel/vmlinux=2Elds=2ES
>> >> @@ -144,10 +144,12 @@ SECTIONS
>> >>  		*(=2Etext=2E__x86=2Eindirect_thunk)
>> >>  		__indirect_thunk_end =3D =2E;
>> >>  #endif
>> >> +
>> >> +		/* End of text section */
>> >> +		_etext =3D =2E;
>> >>  	} :text =3D0xcccc
>> >> =20
>> >> -	/* End of text section, which should occupy whole number of
>pages */
>> >> -	_etext =3D =2E;
>> >> +	/* =2Etext should occupy whole number of pages */
>> >>  	=2E =3D ALIGN(PAGE_SIZE);
>> >=20
>> > NAK: linkers can add things at the end of =2Etext that will go
>missing from
>> > the kernel if _etext isn't _outside_ the =2Etext section, truly
>beyond the
>> > end of the =2Etext section=2E This patch will break Control Flow
>Integrity
>> > checking since the jump tables are at the end of =2Etext=2E
>> >=20
>> > Boris, we're always working around weird linker problems; I don't
>see a
>> > problem with the v2 patch to fix up old binutils=2E=2E=2E
>> >=20
>>=20
>> Why not add the marker into a separate section instead of leaving it
>as an
>> absolute "floater"? Very old binutils would botch that case, but I
>think that
>> has been long since addressed well below our current minimum version=2E
>>=20
>> 	-hpa
>>=20
>>=20
>>=20
>
>Kees, thanks, I noted in the other email that you had mentioned this in
>a since-reverted commit, but you did not mention in the most recent
>commit=2E
>
>hpa, I think this runs afoul of the bug you noted in commit
>fd952815307f
>("x86-32, relocs: Whitelist more symbols for ld bug workaround"), ld
>version 2=2E22=2E52=2E0=2E[12] can incorrectly promote relative symbols t=
o
>absolute, if the output section they appear in is otherwise empty=2E
>
>That's 2=2E22, which is more recent than the 2=2E21 that the kernel
>documents as supported=2E

Oh right=2E=2E=2E they introduced that little piece of drain bramage=2E
--=20
Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E