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:   Tue, 07 Jun 2022 07:56:52 +0100
Message-ID: <87tu8xyn9n.wl-maz@kernel.org>
From:   Marc Zyngier <maz@kernel.org>
To:     Jianmin Lv <lvjianmin@loongson.cn>
Cc:     Thomas Gleixner <tglx@linutronix.de>, linux-kernel@vger.kernel.org,
        Xuefeng Li <lixuefeng@loongson.cn>,
        Huacai Chen <chenhuacai@gmail.com>,
        Jiaxun Yang <jiaxun.yang@flygoat.com>,
        Huacai Chen <chenhuacai@loongson.cn>,
        Hanjun Guo <guohanjun@huawei.com>,
        Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Subject: Re: [PATCH RFC V2 02/10] irqchip: Add LoongArch CPU interrupt controller support
In-Reply-To: <0c6b88f8-c100-4068-5370-c30e95365b3c@loongson.cn>
References: <1653649335-11998-1-git-send-email-lvjianmin@loongson.cn>
        <1653649335-11998-3-git-send-email-lvjianmin@loongson.cn>
        <87bkvf56wg.wl-maz@kernel.org>
        <64990891.8322.18119c6d212.Coremail.lvjianmin@loongson.cn>
        <87a6ax6c0n.wl-maz@kernel.org>
        <1c7b1ee5-3f1e-c090-fb2e-65741de76a9f@loongson.cn>
        <871qw25ctt.wl-maz@kernel.org>
        <0c6b88f8-c100-4068-5370-c30e95365b3c@loongson.cn>
User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue)
 FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/27.1
 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)
MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue")
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, 07 Jun 2022 04:41:22 +0100,
Jianmin Lv <lvjianmin@loongson.cn> wrote:
>=20
>=20
> On 2022/6/6 =E4=B8=8B=E5=8D=886:02, Marc Zyngier wrote:
> > + Lorenzo and Hanjun who maintain the ACPI irq code
> >=20
> > On Thu, 02 Jun 2022 04:16:30 +0100,
> > Jianmin Lv <lvjianmin@loongson.cn> wrote:

[...]

> >> I'm so sorry, I really don't find a way to reuse driver/acpi/irq.c
> >> after my humble work.
> > I don't think reimplementing ACPI is the solution. What could be a
> > reasonable approach is a way to overload the retrieval of the
> > acpi_gsi_domain_id fwnode with a GSI parameter.
> >=20
> > I hacked the following patch, which will give you an idea of what I
> > have in mind (only compile-tested).
>=20
>=20
> Hi, Marc, thanks so much for your patch. I have verified it on my
> LoongArch machine and it works well.
>=20
>=20
> BTW, in acpi_get_irq_source_fwhandle(), maybe
> acpi_get_gsi_domain_id(ctx->index) is needed to changed to
> acpi_get_gsi_domain_id(irq->interrupts[ctx->index])?

Yes, absolutely. Thanks for spotting it.

> I have another question, for LoongArch, acpi_isa_irq_to_gsi is
> required to implemente, but no common version, do we need to
> implemente an weak version in driver/acpi/irq.c as following?
>=20
>=20
> int __weak acpi_isa_irq_to_gsi(unsigned int isa_irq, u32 *gsi)
> {
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (gsi)
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0 *gsi =3D isa_irq;
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return 0;
> }

Do you actually have CONFIG_ISA? In 2022? For a brand new architecture?

If you really have to, then this needs to be a bit more involved:

#ifdef CONFIG_ISA
int __weak acpi_isa_irq_to_gsi(unsigned int isa_irq, u32 *gsi)
{
	if (irq < nr_legacy_irqs()) {
		*gsi =3D isa_irq;
		return 0;
	}

	return -1;
}
#endif

But I'd rather you get rid of any such legacy if this can be avoided.

> I'll use the way you provided here to reuse driver/acpi/irq.c in next
> version. How should I do next? Should I integrate your patch into my
> next version or wait for you to merge it first?

Please pick up the patch (with the above fix), and use it as a prefix
to your series. It needs to be reviewed by the relevant maintainers
anyway.

Thanks,

	M.

--=20
Without deviation from the norm, progress is not possible.