Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 2 Aug 2023 09:52:45 +0100
From:   Conor Dooley <conor.dooley@microchip.com>
To:     Greg KH <gregkh@linuxfoundation.org>
CC:     Minda Chen <minda.chen@starfivetech.com>,
        Conor Dooley <conor@kernel.org>,
        Jisheng Zhang <jszhang@kernel.org>,
        Samuel Holland <samuel@sholland.org>,
        Dao Lu <daolu@rivosinc.com>, Heiko Stuebner <heiko@sntech.de>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Palmer Dabbelt <palmer@dabbelt.com>,
        Albert Ou <aou@eecs.berkeley.edu>,
        <linux-kernel@vger.kernel.org>, <linux-riscv@lists.infradead.org>,
        Mason Huo <mason.huo@starfivetech.com>,
        <stable@vger.kernel.org>
Subject: Re: [PATCH v1] riscv: Using TOOLCHAIN_HAS_ZIHINTPAUSE marco replace
 zihintpause
Message-ID: <20230802-decibel-unshaved-fde1cfba2d20@wendy>
References: <20230802064215.31111-1-minda.chen@starfivetech.com>
 <20230802-sharpness-spoon-f9b8804fb66f@wendy>
 <d64874cb-8628-a6d2-d2f4-8af4d0ebf8b2@starfivetech.com>
 <20230802-seismic-gallstone-fca0f4b17076@wendy>
 <c42bd997-8795-8bf7-eee1-3ac8b153371a@starfivetech.com>
 <2023080253-headache-moneybags-a833@gregkh>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha256;
        protocol="application/pgp-signature"; boundary="FW1Kj+nV7cDQRX/N"
Content-Disposition: inline
In-Reply-To: <2023080253-headache-moneybags-a833@gregkh>
Precedence: bulk

--FW1Kj+nV7cDQRX/N
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, Aug 02, 2023 at 10:33:27AM +0200, Greg KH wrote:
> On Wed, Aug 02, 2023 at 04:17:51PM +0800, Minda Chen wrote:
> >=20
> >=20
> > On 2023/8/2 15:48, Conor Dooley wrote:
> > > On Wed, Aug 02, 2023 at 03:32:15PM +0800, Minda Chen wrote:
> > >>=20
> > >>=20
> > >> On 2023/8/2 14:54, Conor Dooley wrote:
> > >> > Hey Minda,
> > >> >=20
> > >> > On Wed, Aug 02, 2023 at 02:42:15PM +0800, Minda Chen wrote:
> > >> >> Actually it is a part of Conor's
> > >> >> commit aae538cd03bc ("riscv: fix detection of toolchain
> > >> >> Zihintpause support").
> > >> >> It is looks like a merge issue.
> > >> >=20
> > >> > Yup, spot on.
> > >> >=20
> > >> >> Samuel's
> > >> >> commit 0b1d60d6dd9e ("riscv: Fix build with
> > >> >> CONFIG_CC_OPTIMIZE_FOR_SIZE=3Dy") do not base on Conor's commit a=
nd
> > >> >> revert to __riscv_zihintpause. So this patch can fix it.
> > >> >>=20
> > >> >> Signed-off-by: Minda Chen <minda.chen@starfivetech.com>
> > >> >=20
> > >> > Did you actually manage to trigger this, or was this by inspection?
> > >> > clang-15 + binutils 2.35 was, IIRC, how we spotted this because th=
at's
> > >> > what the clang-built-linux CI uses to test the LTS kernels from be=
fore
> > >> > LLVM's IAS was supported for RISC-V. Seemingly all that needs to be
> > >> > satisfied there is that zihintpause doesn't appear in -march so th=
is has
> > >> > gone unnoticed.
> > >> >=20
> > >> > Fixes: 3c349eacc559 ("Merge patch "riscv: Fix build with CONFIG_CC=
_OPTIMIZE_FOR_SIZE=3Dy"")
> > >> > Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
> > >> >=20
> > >> > Thanks,
> > >> > Conor.
> > >> >=20
> > >> Thanks, Conor. I found this just by inspection. I found a issue that=
 vdso.so call cpu_relax
> > >> cause application core dump in kernel 6.1.31. I need Samuel'patch to=
 fix this. And I search the log
> > >> of processor.h found this issue.
> > >=20
> > > That doesn't look like it is fixed in later stable kernels (we are at
> > > 6.1.42-rcN right now I think). It sounds we should ask Greg to backpo=
rt
> > > 0b1d60d6dd9e ("riscv: Fix build with CONFIG_CC_OPTIMIZE_FOR_SIZE=3Dy")
> > > to 6.1. Does that make sense to you?
> > Yes. 6.1 is lts kernel. Starfive will use this kernel for a long time. =
Thanks.
>=20
> What is preventing you from moving to a newer kernel version?  All of
> your kernel changes are already properly merged into Linus's tree,
> right?

Regardless of their reasons, "vdso.so call cpu_relax cause application
core dump" is something that we should fix in stable kernels, no?

--FW1Kj+nV7cDQRX/N
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQRh246EGq/8RLhDjO14tDGHoIJi0gUCZMoZXQAKCRB4tDGHoIJi
0oRWAP9BTGH3q07+XeBW8yRULD9PA1MB/TiLZFZTyMSVwZ+k7gEAkN5bfTlTj2Bu
4J2aZnx6QsteTBSDwjAzsIcBaqYmGQU=
=deFA
-----END PGP SIGNATURE-----

--FW1Kj+nV7cDQRX/N--