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;
Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 15.0 \(3693.40.0.1.81\))
Subject: Re: [PATCH] riscv: eliminate unreliable __builtin_frame_address(1)
From:   Jessica Clarke <jrtc27@jrtc27.com>
In-Reply-To: <87v8yg6lhf.fsf@igel.home>
Date:   Wed, 19 Jan 2022 19:05:13 +0000
Cc:     Changbin Du <changbin.du@gmail.com>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Palmer Dabbelt <palmer@dabbelt.com>,
        Albert Ou <aou@eecs.berkeley.edu>,
        linux-riscv <linux-riscv@lists.infradead.org>,
        linux-kernel@vger.kernel.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <AAAA7852-EBCA-47A3-B74E-A425023468C6@jrtc27.com>
References: <20220117154433.3124-1-changbin.du@gmail.com>
 <C2470F2D-9E45-49D7-A03B-E6A7BB4B9738@jrtc27.com> <87v8yg6lhf.fsf@igel.home>
To:     Andreas Schwab <schwab@linux-m68k.org>
Precedence: bulk

On 19 Jan 2022, at 10:58, Andreas Schwab <schwab@linux-m68k.org> wrote:
>=20
> On Jan 17 2022, Jessica Clarke wrote:
>=20
>> Yes, this is a bug, that is always wrong. LLVM gets this right.
>=20
> Is that an ABI requirement?  In case of a leaf function, gcc saves the
> caller's frame pointer in the first slot, not the second (it doesn't
> save the return address).

Leaf functions by definition don=E2=80=99t have callees that are trying =
to read
their frame pointer so aren=E2=80=99t relevant here. The stack frame =
layout
isn=E2=80=99t specified by the ABI, only that the in-memory outgoing =
arguments
be at the bottom when calling other functions. However, GCC knows what
layout it uses, so it should be consistent and follow that layout for
walking back up frames. Especially for __builtin_frame_address(1), that
just pertains to the current function=E2=80=99s frame, which it clearly =
knows
without a doubt, so there=E2=80=99s no reason to get that wrong. =
Accessing
0(s0) is just straight up wrong, that=E2=80=99s accessing past the top =
of the
stack frame, which is never going to make any sense.

Jess