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, 29 Jun 2022 14:15:01 +0100
Message-ID: <87sfnn1ui2.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,
        Hanjun Guo <guohanjun@huawei.com>,
        Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
        Jiaxun Yang <jiaxun.yang@flygoat.com>,
        Huacai Chen <chenhuacai@loongson.cn>
Subject: Re: [PATCH V13 07/13] irqchip/loongson-pch-msi: Add ACPI init support
In-Reply-To: <1656329997-20524-8-git-send-email-lvjianmin@loongson.cn>
References: <1656329997-20524-1-git-send-email-lvjianmin@loongson.cn>
        <1656329997-20524-8-git-send-email-lvjianmin@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 Mon, 27 Jun 2022 12:39:51 +0100,
Jianmin Lv <lvjianmin@loongson.cn> wrote:
>=20
> From: Huacai Chen <chenhuacai@loongson.cn>
>=20
> We are preparing to add new Loongson (based on LoongArch, not compatible
> with old MIPS-based Loongson) support. LoongArch use ACPI other than DT
> as its boot protocol, so add ACPI init support.

Same rant about the content-less paragraph, which I asked you to drop
a few version ago.

>=20
> PCH-PIC/PCH-MSI stands for "Interrupt Controller" that described in
> Section 5 of "Loongson 7A1000 Bridge User Manual". For more information
> please refer Documentation/loongarch/irq-chip-model.rst.
>=20
> For systems with two chipsets, there are two related msi irqdomains,
> each of which has the same node id as its parent irqdomain. So we
> use a structure to mantain the relation of node and it's parent
> irqdomain as pch irqdomin, and add a pci_segment field to it for
> matching the pci segment of a pci device when setting msi irqdomain
> for the device.
>=20
> struct acpi_vector_group {
>         int node;
>         int pci_segment;
>         struct irq_domain *parent;
> };
>=20
> The field 'pci_segment' and 'node' are initialized from MCFG, and
> the parent irqdomain will set field 'parent' by matching same
> 'node'.
>=20
> Co-developed-by: Jianmin Lv <lvjianmin@loongson.cn>
> Signed-off-by: Jianmin Lv <lvjianmin@loongson.cn>
> Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
> ---
>  arch/loongarch/include/asm/irq.h       |   7 +-
>  arch/loongarch/kernel/irq.c            |  30 ++++++-
>  drivers/irqchip/irq-loongson-pch-msi.c | 147 +++++++++++++++++++++++----=
------
>  3 files changed, 138 insertions(+), 46 deletions(-)
>=20
> diff --git a/arch/loongarch/include/asm/irq.h b/arch/loongarch/include/as=
m/irq.h
> index a444dc6..367aa44 100644
> --- a/arch/loongarch/include/asm/irq.h
> +++ b/arch/loongarch/include/asm/irq.h
> @@ -50,9 +50,11 @@ static inline bool on_irq_stack(int cpu, unsigned long=
 sp)
> =20
>  struct acpi_vector_group {
>  	int node;
> +	int pci_segment;
>  	struct irq_domain *parent;
>  };
>  extern struct acpi_vector_group pch_group[MAX_IO_PICS];
> +extern struct acpi_vector_group msi_group[MAX_IO_PICS];
> =20
>  #define CORES_PER_EIO_NODE	4
> =20
> @@ -112,9 +114,8 @@ struct irq_domain *htvec_acpi_init(struct irq_domain =
*parent,
>  					struct acpi_madt_ht_pic *acpi_htvec);
>  int pch_lpc_acpi_init(struct irq_domain *parent,
>  					struct acpi_madt_lpc_pic *acpi_pchlpc);
> -struct irq_domain *pch_msi_acpi_init(struct irq_domain *parent,
> -					struct acpi_madt_msi_pic *acpi_pchmsi);
>  int find_pch_pic(u32 gsi);
> +struct fwnode_handle *get_pch_msi_handle(int pci_segment);
> =20
>  extern struct acpi_madt_lio_pic *acpi_liointc;
>  extern struct acpi_madt_eio_pic *acpi_eiointc[MAX_IO_PICS];
> @@ -127,7 +128,7 @@ struct irq_domain *pch_msi_acpi_init(struct irq_domai=
n *parent,
>  extern struct irq_domain *cpu_domain;
>  extern struct irq_domain *liointc_domain;
>  extern struct fwnode_handle *pch_lpc_handle;
> -extern struct irq_domain *pch_msi_domain[MAX_IO_PICS];
> +extern struct fwnode_handle *pch_msi_handle[MAX_IO_PICS];
>  extern struct fwnode_handle *pch_pic_handle[MAX_IO_PICS];
> =20
>  extern irqreturn_t loongson3_ipi_interrupt(int irq, void *dev);
> diff --git a/arch/loongarch/kernel/irq.c b/arch/loongarch/kernel/irq.c
> index f2115be..82bc6c7 100644
> --- a/arch/loongarch/kernel/irq.c
> +++ b/arch/loongarch/kernel/irq.c
> @@ -27,8 +27,8 @@
> =20
>  struct irq_domain *cpu_domain;
>  struct irq_domain *liointc_domain;
> -struct irq_domain *pch_msi_domain[MAX_IO_PICS];
>  struct acpi_vector_group pch_group[MAX_IO_PICS];
> +struct acpi_vector_group msi_group[MAX_IO_PICS];
> =20
>  /*
>   * 'what should we do if we get a hw irq event on an illegal vector'.
> @@ -55,6 +55,33 @@ int arch_show_interrupts(struct seq_file *p, int prec)
>  	return 0;
>  }
> =20
> +static int early_pci_mcfg_parse(struct acpi_table_header *header)

Shouldn't this be __init as well?

> +{
> +	struct acpi_table_mcfg *mcfg;
> +	struct acpi_mcfg_allocation *mptr;
> +	int i, n;
> +
> +	if (header->length < sizeof(struct acpi_table_mcfg))
> +		return -EINVAL;
> +
> +	n =3D (header->length - sizeof(struct acpi_table_mcfg)) /
> +					sizeof(struct acpi_mcfg_allocation);
> +	mcfg =3D (struct acpi_table_mcfg *)header;
> +	mptr =3D (struct acpi_mcfg_allocation *) &mcfg[1];
> +
> +	for (i =3D 0; i < n; i++, mptr++) {
> +		msi_group[i].pci_segment =3D mptr->pci_segment;
> +		msi_group[i].node =3D (mptr->address >> 44) & 0xf;
> +	}
> +
> +	return 0;
> +}
> +
> +void __init init_msi_parent_group(void)
> +{
> +	acpi_table_parse(ACPI_SIG_MCFG, early_pci_mcfg_parse);
> +}
> +
>  void __init init_IRQ(void)
>  {
>  	int i;
> @@ -68,6 +95,7 @@ void __init init_IRQ(void)
>  	clear_csr_ecfg(ECFG0_IM);
>  	clear_csr_estat(ESTATF_IP);
> =20
> +	init_msi_parent_group();

The changes in this file should be a separate patch, as it is
completely independent.

>  	irqchip_init();
>  #ifdef CONFIG_SMP
>  	ipi_irq =3D EXCCODE_IPI - EXCCODE_INT_START;
> diff --git a/drivers/irqchip/irq-loongson-pch-msi.c b/drivers/irqchip/irq=
-loongson-pch-msi.c
> index e3801c4..5db4a68 100644
> --- a/drivers/irqchip/irq-loongson-pch-msi.c
> +++ b/drivers/irqchip/irq-loongson-pch-msi.c
> @@ -15,14 +15,30 @@
>  #include <linux/pci.h>
>  #include <linux/slab.h>
> =20
> +static int nr_pics;
> +
>  struct pch_msi_data {
>  	struct mutex	msi_map_lock;
>  	phys_addr_t	doorbell;
>  	u32		irq_first;	/* The vector number that MSIs starts */
>  	u32		num_irqs;	/* The number of vectors for MSIs */
>  	unsigned long	*msi_map;
> +	struct fwnode_handle *domain_handle;
>  };
> =20
> +static struct pch_msi_data *pch_msi_priv[2];

What is this '2'?

> +
> +struct fwnode_handle *get_pch_msi_handle(int pci_segment)
> +{
> +	int i;
> +
> +	for (i =3D 0; i < MAX_IO_PICS; i++) {
> +		if (msi_group[i].pci_segment =3D=3D pci_segment)

AFAICT, this driver is supposed to compile on MIPS too. Clearly, you
haven't bothered checking that it was still building:

drivers/irqchip/irq-loongson-pch-msi.c: In function =E2=80=98get_pch_msi_ha=
ndle=E2=80=99:
drivers/irqchip/irq-loongson-pch-msi.c:35:18: error: =E2=80=98MAX_IO_PICS=
=E2=80=99 undeclared (first use in this function)
   35 |  for (i =3D 0; i < MAX_IO_PICS; i++) {
      |                  ^~~~~~~~~~~
drivers/irqchip/irq-loongson-pch-msi.c:35:18: note: each undeclared identif=
ier is reported only once for each function it appears in
drivers/irqchip/irq-loongson-pch-msi.c:36:7: error: =E2=80=98msi_group=E2=
=80=99 undeclared (first use in this function); did you mean =E2=80=98task_=
group=E2=80=99?
   36 |   if (msi_group[i].pci_segment =3D=3D pci_segment)
      |       ^~~~~~~~~
      |       task_group

I've stopped here.

	M.

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