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:   Sat, 29 Jul 2023 11:34:38 +0100
From:   Conor Dooley <conor@kernel.org>
To:     Jisheng Zhang <jszhang@kernel.org>
Cc:     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
Subject: Re: [PATCH] riscv: enable DEBUG_FORCE_FUNCTION_ALIGN_64B
Message-ID: <20230729-shucking-slang-370c6baa358e@spud>
References: <20230727160356.3874-1-jszhang@kernel.org>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha256;
        protocol="application/pgp-signature"; boundary="LxHC1VpAZFy1zAHq"
Content-Disposition: inline
In-Reply-To: <20230727160356.3874-1-jszhang@kernel.org>
Precedence: bulk


--LxHC1VpAZFy1zAHq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hey,

On Fri, Jul 28, 2023 at 12:03:56AM +0800, Jisheng Zhang wrote:
> Allow to force all function address 64B aligned as it is possible for
> other architectures. This may be useful when verify if performance
> bump is caused by function alignment changes.
>=20
> Before commit 1bf18da62106 ("lib/Kconfig.debug: add ARCH dependency
> for FUNCTION_ALIGN option"), riscv supports enabling the
> DEBUG_FORCE_FUNCTION_ALIGN_64B option, but after that commit, each
> arch needs to claim the support explicitly. I tried the config file in
> [1] for both RV64 and RV32, I can't reproduce the build error as [1],
> there is no reason for not allowing to enforce this function alignment.
>=20
> Link: https://lore.kernel.org/lkml/202202271612.W32UJAj2-lkp@intel.com/ [=
1]

This is a CSKY randconfig, is there any particular reason that running
that randconfig (over a year later) and on a different architecture
would trigger whatever the condition was?

The original commit here seems far too penal - why was it not just
disabled on CSKY??? I tried looking a bit on lore, but didn't see
anything explaining the subset of supported archs they picked.
I did see Guo Ren wondering if rv32 would be similarly problematic - but
since this is something likely to just trip up randconfigs, I think we
should go for it and if rv32 becomes a problem, restrict this to 64-bit.

Reviewed-by: Conor Dooley <conor.dooley@microchip.com>

Thanks,
Conor.

> Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
> ---
>  lib/Kconfig.debug | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>=20
> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> index fbc89baf7de6..39ffd218e960 100644
> --- a/lib/Kconfig.debug
> +++ b/lib/Kconfig.debug
> @@ -502,7 +502,7 @@ config SECTION_MISMATCH_WARN_ONLY
> =20
>  config DEBUG_FORCE_FUNCTION_ALIGN_64B
>  	bool "Force all function address 64B aligned"
> -	depends on EXPERT && (X86_64 || ARM64 || PPC32 || PPC64 || ARC || S390)
> +	depends on EXPERT && (X86_64 || ARM64 || PPC32 || PPC64 || ARC || RISCV=
 || S390)
>  	select FUNCTION_ALIGNMENT_64B
>  	help
>  	  There are cases that a commit from one domain changes the function
> --=20
> 2.40.1
>=20

--LxHC1VpAZFy1zAHq
Content-Type: application/pgp-signature; name="signature.asc"

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

iHUEABYIAB0WIQRh246EGq/8RLhDjO14tDGHoIJi0gUCZMTrPQAKCRB4tDGHoIJi
0iGdAQDyN75O2rQ1WW8cAOtDv2dka5BxnhOAnhZQPU4T3q4cpgD+LfZKzuDdbrub
ytNDwsKc2dgDCxjr1N0tLkggxyp44ws=
=mi/7
-----END PGP SIGNATURE-----

--LxHC1VpAZFy1zAHq--