Received-SPF: pass (google.com: domain of linux-kernel+bounces-179708-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Date: Wed, 15 May 2024 10:32:46 +0100
From: Conor Dooley <conor.dooley@microchip.com>
To: Andrew Jones <ajones@ventanamicro.com>
CC: "Wang, Xiao W" <xiao.w.wang@intel.com>, "paul.walmsley@sifive.com"
	<paul.walmsley@sifive.com>, "palmer@dabbelt.com" <palmer@dabbelt.com>,
	"aou@eecs.berkeley.edu" <aou@eecs.berkeley.edu>, "luke.r.nels@gmail.com"
	<luke.r.nels@gmail.com>, "xi.wang@gmail.com" <xi.wang@gmail.com>,
	"bjorn@kernel.org" <bjorn@kernel.org>, "ast@kernel.org" <ast@kernel.org>,
	"daniel@iogearbox.net" <daniel@iogearbox.net>, "andrii@kernel.org"
	<andrii@kernel.org>, "martin.lau@linux.dev" <martin.lau@linux.dev>,
	"eddyz87@gmail.com" <eddyz87@gmail.com>, "song@kernel.org" <song@kernel.org>,
	"yonghong.song@linux.dev" <yonghong.song@linux.dev>,
	"john.fastabend@gmail.com" <john.fastabend@gmail.com>, "kpsingh@kernel.org"
	<kpsingh@kernel.org>, "sdf@google.com" <sdf@google.com>, "haoluo@google.com"
	<haoluo@google.com>, "jolsa@kernel.org" <jolsa@kernel.org>,
	"linux-riscv@lists.infradead.org" <linux-riscv@lists.infradead.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"bpf@vger.kernel.org" <bpf@vger.kernel.org>, "pulehui@huawei.com"
	<pulehui@huawei.com>, "Li, Haicheng" <haicheng.li@intel.com>,
	"conor@kernel.org" <conor@kernel.org>, Ben Dooks <ben.dooks@codethink.co.uk>
Subject: Re: [PATCH v2] riscv, bpf: Optimize zextw insn with Zba extension
Message-ID: <20240515-jogger-pummel-19fe4e9e8314@wendy>
References: <20240511023436.3282285-1-xiao.w.wang@intel.com>
 <20240513-5c6f04fb4a29963c63d09aa2@orel>
 <DM8PR11MB575179A3EB8D056B3EEECA74B8E32@DM8PR11MB5751.namprd11.prod.outlook.com>
 <20240514-944dec90b2c531d8b6c783f7@orel>
 <20240515-cone-getting-d17037b51e97@wendy>
Precedence: bulk
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha256;
	protocol="application/pgp-signature"; boundary="es0X3vLG906/aXNt"
Content-Disposition: inline
In-Reply-To: <20240515-cone-getting-d17037b51e97@wendy>

--es0X3vLG906/aXNt
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, May 15, 2024 at 09:19:46AM +0100, Conor Dooley wrote:
> On Tue, May 14, 2024 at 03:37:02PM +0200, Andrew Jones wrote:
> > On Tue, May 14, 2024 at 07:36:04AM GMT, Wang, Xiao W wrote:
> > > > From: Andrew Jones <ajones@ventanamicro.com>
> >> > > > +config RISCV_ISA_ZBA
> > > > > +	bool "Zba extension support for bit manipulation instructions"
> > > > > +	depends on TOOLCHAIN_HAS_ZBA
> > > >=20
> > > > We handcraft the instruction, so why do we need toolchain support?
> > >=20
> > > Good point, we don't need toolchain support for this bpf jit case.
> > >=20
> > > >=20
> > > > > +	depends on RISCV_ALTERNATIVE
> > > >=20
> > > > Also, while riscv_has_extension_likely() will be accelerated with
> > > > RISCV_ALTERNATIVE, it's not required.
> > >=20
> > > Agree, it's not required. For this bpf jit case, we should drop these=
 two dependencies.
> > >=20
> > > BTW, Zbb is used in bpf jit, the usage there also doesn't depend on t=
oolchain and
> > > RISCV_ALTERNATIVE, but the Kconfig for RISCV_ISA_ZBB has forced the d=
ependencies
> > > due to Zbb assembly programming elsewhere.
> > > Maybe we could just dynamically check the existence of RISCV_ISA_ZB* =
before jit code
> > > emission? or introduce new config options for bpf jit? I prefer the f=
irst method and
> > > welcome any comments.
> >=20
> > My preferences is to remove as much of the TOOLCHAIN_HAS_ stuff as
> > possible. We should audit the extensions which have them to see if
> > they're really necessary.
>=20
> While I think it is reasonable to allow the "RISCV_ISA_ZBB" option to
> control whether or not bpf is allowed to use it for optimisations, only
> allowing bpf to do that if there's toolchain support feels odd to me..
> Maybe we need to sorta steal from Charlie's patchset and introduce
> some hidden options that have the toolchain dep that are used by the
> alternative macros etc?
>=20
> I'll have a poke at how bad that looks I think.

I don't love this, in particular my option naming, but it would allow
the Zbb optimisations in the kernel to not depend on toolchain support
while not muddying the Kconfig waters for users:
https://git.kernel.org/pub/scm/linux/kernel/git/conor/linux.git/commit/?h=
=3Driscv-zbb_split
A similar model could be followed if there were to be some
optimisations for Zba in the future that do require toolchain support:

--es0X3vLG906/aXNt
Content-Type: application/pgp-signature; name="signature.asc"

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

iHUEABYIAB0WIQRh246EGq/8RLhDjO14tDGHoIJi0gUCZkSBPgAKCRB4tDGHoIJi
0nCmAQCsoI65PD/ah3I73wtpDrzK+PCNiu2WNVR/dUBJlyt29AD9Gw6ZED7Qb42h
L7QgyIUuoEoC+s9FgNEV/wpZUPvx9Qo=
=gTNF
-----END PGP SIGNATURE-----

--es0X3vLG906/aXNt--