diff mbox series

[linux,dev-4.17,3/7] mmc: Aspeed: Add Aspeed sdhci core driver

Message ID 1531812378-14316-4-git-send-email-ryanchen.aspeed@gmail.com
State Changes Requested, archived
Headers show
Series mmc/host: Add Aspeed SDIO driver | expand

Commit Message

Ryan Chen July 17, 2018, 7:26 a.m. UTC
In Aspeed SoC's sdhci have two slots and have a interrupt status and
general information in top for register. add sdhci core driver
defore sdhci driver probe

V0->V1
move irqchip/irq-aspeed-sdhci-ic.c to drivers/mmc/host/aspeed-sdhci-core.c

Signed-off-by: Ryan Chen <ryanchen.aspeed@gmail.com>
---
 drivers/irqchip/Makefile              |   2 +-
 drivers/mmc/host/Makefile             |   1 +
 drivers/mmc/host/aspeed-sdhci-core.c  | 147 ++++++++++++++++++++++++++++++++++
 include/linux/mmc/sdhci-aspeed-data.h |  28 +++++++
 4 files changed, 177 insertions(+), 1 deletion(-)
 create mode 100644 drivers/mmc/host/aspeed-sdhci-core.c
 create mode 100644 include/linux/mmc/sdhci-aspeed-data.h

Comments

Benjamin Herrenschmidt July 26, 2018, 1:41 a.m. UTC | #1
On Tue, 2018-07-17 at 15:26 +0800, Ryan Chen wrote:
> In Aspeed SoC's sdhci have two slots and have a interrupt status and
> general information in top for register. add sdhci core driver
> defore sdhci driver probe
> 
> V0->V1
> move irqchip/irq-aspeed-sdhci-ic.c to drivers/mmc/host/aspeed-sdhci-core.c
> 
> Signed-off-by: Ryan Chen <ryanchen.aspeed@gmail.com>
> ---
>  drivers/irqchip/Makefile              |   2 +-
>  drivers/mmc/host/Makefile             |   1 +
>  drivers/mmc/host/aspeed-sdhci-core.c  | 147 ++++++++++++++++++++++++++++++++++
>  include/linux/mmc/sdhci-aspeed-data.h |  28 +++++++
>  4 files changed, 177 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/mmc/host/aspeed-sdhci-core.c
>  create mode 100644 include/linux/mmc/sdhci-aspeed-data.h
> 
> diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
> index 5ed465a..d99cb1d 100644
> --- a/drivers/irqchip/Makefile
> +++ b/drivers/irqchip/Makefile
> @@ -78,7 +78,7 @@ obj-$(CONFIG_MVEBU_ODMI)		+= irq-mvebu-odmi.o
>  obj-$(CONFIG_MVEBU_PIC)			+= irq-mvebu-pic.o
>  obj-$(CONFIG_LS_SCFG_MSI)		+= irq-ls-scfg-msi.o
>  obj-$(CONFIG_EZNPS_GIC)			+= irq-eznps.o
> -obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o
> +obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o 
>  obj-$(CONFIG_STM32_EXTI) 		+= irq-stm32-exti.o
>  obj-$(CONFIG_QCOM_IRQ_COMBINER)		+= qcom-irq-combiner.o
>  obj-$(CONFIG_IRQ_UNIPHIER_AIDET)	+= irq-uniphier-aidet.o
> diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
> index 6aead24..c3e5a1f 100644
> --- a/drivers/mmc/host/Makefile
> +++ b/drivers/mmc/host/Makefile
> @@ -78,6 +78,7 @@ obj-$(CONFIG_MMC_SDHCI_ESDHC_IMX)	+= sdhci-esdhc-imx.o
>  obj-$(CONFIG_MMC_SDHCI_DOVE)		+= sdhci-dove.o
>  obj-$(CONFIG_MMC_SDHCI_TEGRA)		+= sdhci-tegra.o
>  obj-$(CONFIG_MMC_SDHCI_OF_ARASAN)	+= sdhci-of-arasan.o
> +obj-$(CONFIG_MMC_SDHCI_OF_ASPEED)	+= aspeed-sdhci-core.o
>  obj-$(CONFIG_MMC_SDHCI_OF_AT91)		+= sdhci-of-at91.o
>  obj-$(CONFIG_MMC_SDHCI_OF_ESDHC)	+= sdhci-of-esdhc.o
>  obj-$(CONFIG_MMC_SDHCI_OF_HLWD)		+= sdhci-of-hlwd.o
> diff --git a/drivers/mmc/host/aspeed-sdhci-core.c b/drivers/mmc/host/aspeed-sdhci-core.c
> new file mode 100644
> index 0000000..c3d9d45
> --- /dev/null
> +++ b/drivers/mmc/host/aspeed-sdhci-core.c
> @@ -0,0 +1,147 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * SDHCI IRQCHIP driver for the Aspeed SoC
> + *
> + * Copyright (C) ASPEED Technology Inc.
> + * Ryan Chen <ryan_chen@aspeedtech.com>
> + *
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or (at
> + * your option) any later version.
> + *
> + */
> +
> +#include <linux/platform_device.h>
> +#include <linux/module.h>
> +#include <linux/delay.h>
> +#include <linux/irq.h>
> +#include <linux/irqchip.h>
> +#include <linux/irqchip/chained_irq.h>
> +#include <linux/irqdomain.h>
> +#include <linux/of_address.h>
> +#include <linux/of_irq.h>
> +#include <linux/of.h>
> +#include <linux/io.h>
> +#include <linux/clk.h>
> +#include <linux/mmc/sdhci-aspeed-data.h>
> +
> +#define ASPEED_SDHCI_SLOT_NUM			2
> 

Change the "irq" prefixes on the functions.. Dont' do a chained
handler, that will be too much overhead for no benefit. Your SoC aren't
very fast and you want to avoid that overhead.

Just do a normal interrupt handler, and have it call each port
interrupts.

This isn't really an interrupt controller, it has no
enable/disable/masking ability, I would just create the ports directly
from the same driver and process the interrupts.

I don't think there's benefit in keeping the ports as separate drivers.

> +static void aspeed_sdhci_irq_handler(struct irq_desc *desc)
> +{
> +	struct aspeed_sdhci_irq *sdhci_irq = irq_desc_get_handler_data(desc);
> +	struct irq_chip *chip = irq_desc_get_chip(desc);
> +	unsigned long bit, status;
> +	unsigned int slot_irq;
> +
> +	chained_irq_enter(chip, desc);
> +	status = readl(sdhci_irq->regs + ASPEED_SDHCI_ISR);
> +	for_each_set_bit(bit, &status, ASPEED_SDHCI_SLOT_NUM) {
> +		slot_irq = irq_find_mapping(sdhci_irq->irq_domain, bit);
> +		generic_handle_irq(slot_irq);
> +	}
> +	chained_irq_exit(chip, desc);
> +}
> +
> +static void noop(struct irq_data *data) { }
> +
> +static unsigned int noop_ret(struct irq_data *data)
> +{
> +	return 0;
> +}
> +
> +struct irq_chip sdhci_irq_chip = {
> +	.name		= "sdhci-ic",
> +	.irq_startup	= noop_ret,
> +	.irq_shutdown	= noop,
> +	.irq_enable	= noop,
> +	.irq_disable	= noop,
> +	.irq_ack	= noop,
> +	.irq_mask	= noop,
> +	.irq_unmask	= noop,
> +	.flags		= IRQCHIP_SKIP_SET_WAKE,
> +};
> +
> +static int ast_sdhci_map_irq_domain(struct irq_domain *domain,
> +				    unsigned int irq, irq_hw_number_t hwirq)
> +{
> +	irq_set_chip_and_handler(irq, &sdhci_irq_chip, handle_simple_irq);
> +	irq_set_chip_data(irq, domain->host_data);
> +
> +	return 0;
> +}
> +
> +static const struct irq_domain_ops aspeed_sdhci_irq_domain_ops = {
> +	.map = ast_sdhci_map_irq_domain,
> +};
> +
> +static int irq_aspeed_sdhci_probe(struct platform_device *pdev)
> +{
> +	struct aspeed_sdhci_irq *sdhci_irq;
> +	struct clk *sdclk;
> +
> +	sdhci_irq = kzalloc(sizeof(*sdhci_irq), GFP_KERNEL);
> +	if (!sdhci_irq)
> +		return -ENOMEM;
> +
> +	platform_set_drvdata(pdev, sdhci_irq);
> +	//node->data = sdhci_irq;
> +	pdev->dev.of_node->data = sdhci_irq;
> +
> +	sdhci_irq->regs = of_iomap(pdev->dev.of_node, 0);
> +	if (IS_ERR(sdhci_irq->regs))
> +		return PTR_ERR(sdhci_irq->regs);
> +
> +	sdclk = devm_clk_get(&pdev->dev, NULL);
> +	if (IS_ERR(sdclk)) {
> +		dev_err(&pdev->dev, "no sdclk clock defined\n");
> +		return PTR_ERR(sdclk);
> +	}
> +
> +	clk_prepare_enable(sdclk);
> +
> +	sdhci_irq->parent_irq = irq_of_parse_and_map(pdev->dev.of_node, 0);
> +	if (sdhci_irq->parent_irq < 0)
> +		return sdhci_irq->parent_irq;
> +
> +	sdhci_irq->irq_domain = irq_domain_add_linear(
> +					pdev->dev.of_node, ASPEED_SDHCI_SLOT_NUM,
> +					&aspeed_sdhci_irq_domain_ops, NULL);
> +	if (!sdhci_irq->irq_domain)
> +		return -ENOMEM;
> +
> +	sdhci_irq->irq_domain->name = "aspeed-sdhci-irq";
> +
> +	irq_set_chained_handler_and_data(sdhci_irq->parent_irq,
> +					 aspeed_sdhci_irq_handler, sdhci_irq);
> +
> +	pr_info("sdhci irq controller registered, irq %d\n", sdhci_irq->parent_irq);
> +
> +	return 0;
> +}
> +
> +static const struct of_device_id irq_aspeed_sdhci_dt_ids[] = {
> +	{ .compatible = "aspeed,aspeed-sdhci-irq", },
> +	{},
> +};

You should call this aspeed,ast2{4,5}00-sdhci-core. I would change your
device-tree representation as follow (add back all the reg &
properties):

	sdhci-core {
		reg = <common regs>
		ranges = either empty prop or the 2 ranges goign to the children
		compatible = <aspeed,ast2400-sdhci-core>;

		sdhci-slot@xxx {
			compatible = <aspeed,ast2400-sdhci-slot>;
		};

		sdhci-slot@yyy {
			compatible = <aspeed,ast2400-sdhci-slot>;
		};
	}

> +MODULE_DEVICE_TABLE(of, irq_aspeed_sdhci_dt_ids);
> +
> +static struct platform_driver irq_aspeed_sdhci_device_driver = {
> +	.probe		= irq_aspeed_sdhci_probe,
> +	.driver		= {
> +		.name   = KBUILD_MODNAME,
> +		.of_match_table	= irq_aspeed_sdhci_dt_ids,
> +	}
> +};
> +
> +static int __init irq_aspeed_sdhci_init(void)
> +{
> +	return platform_driver_register(&irq_aspeed_sdhci_device_driver);
> +}
> +core_initcall(irq_aspeed_sdhci_init);

This should be a normal device init call, thus module_platform_driver
is all you need here.

> +
> +MODULE_AUTHOR("Ryan Chen");
> +MODULE_DESCRIPTION("ASPEED SOC SDHCI IRQ Driver");
> +MODULE_LICENSE("GPL v2");
> diff --git a/include/linux/mmc/sdhci-aspeed-data.h b/include/linux/mmc/sdhci-aspeed-data.h
> new file mode 100644
> index 0000000..fba2bf2
> --- /dev/null
> +++ b/include/linux/mmc/sdhci-aspeed-data.h
> @@ -0,0 +1,28 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +#ifndef LINUX_MMC_SDHCI_ASPEED_DATA_H
> +#define LINUX_MMC_SDHCI_ASPEED_DATA_H
> +
> +#include <linux/io.h>
> +
> +#define ASPEED_SDHCI_INFO			0x00
> +#define  ASPEED_SDHCI_S1MMC8			BIT(25)
> +#define  ASPEED_SDHCI_S0MMC8			BIT(24)
> +#define ASPEED_SDHCI_BLOCK			0x04
> +#define ASPEED_SDHCI_CTRL			0xF0
> +#define ASPEED_SDHCI_ISR			0xFC
> +
> +struct aspeed_sdhci_irq {
> +	void __iomem *regs;
> +	int parent_irq;
> +	struct irq_domain *irq_domain;
> +};
> +
> +static inline void aspeed_sdhci_set_8bit_mode(struct aspeed_sdhci_irq *sdhci_irq, int mode)
> +{
> +	if (mode)
> +		writel(ASPEED_SDHCI_S0MMC8 | readl(sdhci_irq->regs), sdhci_irq->regs);
> +	else
> +		writel(~ASPEED_SDHCI_S0MMC8 & readl(sdhci_irq->regs), sdhci_irq->regs);
> +}
> +
> +#endif
Benjamin Herrenschmidt July 26, 2018, 1:47 a.m. UTC | #2
On Thu, 2018-07-26 at 11:41 +1000, Benjamin Herrenschmidt wrote:
> Change the "irq" prefixes on the functions.. Dont' do a chained
> handler, that will be too much overhead for no benefit. Your SoC aren't
> very fast and you want to avoid that overhead.
> 
> Just do a normal interrupt handler, and have it call each port
> interrupts.
> 
> This isn't really an interrupt controller, it has no
> enable/disable/masking ability, I would just create the ports directly
> from the same driver and process the interrupts.
> 
> I don't think there's benefit in keeping the ports as separate drivers.

Look at how shdci-pci does it, sdhci-acpi as well, they have multiple
slots. You can just register multiple slots from a single driver
I think. You can still use device-tree sub-nodes and iterate them
from the core driver to create the slots if you want, in case you may
add slots in the future, but it's not even that a big deal.

Ben.
Ryan Chen July 26, 2018, 3:07 a.m. UTC | #3
On Thu, Jul 26, 2018 at 11:47:40AM +1000, Benjamin Herrenschmidt wrote:
> On Thu, 2018-07-26 at 11:41 +1000, Benjamin Herrenschmidt wrote:
> > Change the "irq" prefixes on the functions.. Dont' do a chained
> > handler, that will be too much overhead for no benefit. Your SoC aren't
> > very fast and you want to avoid that overhead.
> > 
> > Just do a normal interrupt handler, and have it call each port
> > interrupts.
> > 
> > This isn't really an interrupt controller, it has no
> > enable/disable/masking ability, I would just create the ports directly
> > from the same driver and process the interrupts.
> > 
> > I don't think there's benefit in keeping the ports as separate drivers.
> 
> Look at how shdci-pci does it, sdhci-acpi as well, they have multiple
> slots. You can just register multiple slots from a single driver
> I think. You can still use device-tree sub-nodes and iterate them
> from the core driver to create the slots if you want, in case you may
> add slots in the future, but it's not even that a big deal.
> 
> Ben.
> 
>
Thanks, after look into sdhci-pci, it should only request one pci_irq, 
But, i am wondering where is the slot dispatch. could you help me point out it?
Benjamin Herrenschmidt July 26, 2018, 3:14 a.m. UTC | #4
On Thu, 2018-07-26 at 11:07 +0800, Ryan Chen wrote:
> On Thu, Jul 26, 2018 at 11:47:40AM +1000, Benjamin Herrenschmidt wrote:
> > On Thu, 2018-07-26 at 11:41 +1000, Benjamin Herrenschmidt wrote:
> > > Change the "irq" prefixes on the functions.. Dont' do a chained
> > > handler, that will be too much overhead for no benefit. Your SoC aren't
> > > very fast and you want to avoid that overhead.
> > > 
> > > Just do a normal interrupt handler, and have it call each port
> > > interrupts.
> > > 
> > > This isn't really an interrupt controller, it has no
> > > enable/disable/masking ability, I would just create the ports directly
> > > from the same driver and process the interrupts.
> > > 
> > > I don't think there's benefit in keeping the ports as separate drivers.
> > 
> > Look at how shdci-pci does it, sdhci-acpi as well, they have multiple
> > slots. You can just register multiple slots from a single driver
> > I think. You can still use device-tree sub-nodes and iterate them
> > from the core driver to create the slots if you want, in case you may
> > add slots in the future, but it's not even that a big deal.
> > 
> > Ben.
> > 
> > 
> 
> Thanks, after look into sdhci-pci, it should only request one pci_irq, 
> But, i am wondering where is the slot dispatch. could you help me point out it?

So the PCI one just uses the same IRQ number for all the slots in
sdhci_pci_probe_slot(), so the IRQ ends up shared.

So on every irq it will jsut check all slots.

Not sure if that's a problem. Due to how sdhci is organized, if you
really want to make sure it only accesses the one slot, then you indeed
do have to use a cascaded interrupt controller as you are doing, but
I'm not sure it's really worthwhile.

Cheers,
Ben.
Ryan Chen July 26, 2018, 3:26 a.m. UTC | #5
On Thu, Jul 26, 2018 at 01:14:58PM +1000, Benjamin Herrenschmidt wrote:
> On Thu, 2018-07-26 at 11:07 +0800, Ryan Chen wrote:
> > On Thu, Jul 26, 2018 at 11:47:40AM +1000, Benjamin Herrenschmidt wrote:
> > > On Thu, 2018-07-26 at 11:41 +1000, Benjamin Herrenschmidt wrote:
> > > > Change the "irq" prefixes on the functions.. Dont' do a chained
> > > > handler, that will be too much overhead for no benefit. Your SoC aren't
> > > > very fast and you want to avoid that overhead.
> > > > 
> > > > Just do a normal interrupt handler, and have it call each port
> > > > interrupts.
> > > > 
> > > > This isn't really an interrupt controller, it has no
> > > > enable/disable/masking ability, I would just create the ports directly
> > > > from the same driver and process the interrupts.
> > > > 
> > > > I don't think there's benefit in keeping the ports as separate drivers.
> > > 
> > > Look at how shdci-pci does it, sdhci-acpi as well, they have multiple
> > > slots. You can just register multiple slots from a single driver
> > > I think. You can still use device-tree sub-nodes and iterate them
> > > from the core driver to create the slots if you want, in case you may
> > > add slots in the future, but it's not even that a big deal.
> > > 
> > > Ben.
> > > 
> > > 
> > 
> > Thanks, after look into sdhci-pci, it should only request one pci_irq, 
> > But, i am wondering where is the slot dispatch. could you help me point out it?
> 
> So the PCI one just uses the same IRQ number for all the slots in
> sdhci_pci_probe_slot(), so the IRQ ends up shared.
> 
> So on every irq it will jsut check all slots.
> 
> Not sure if that's a problem. Due to how sdhci is organized, if you
> really want to make sure it only accesses the one slot, then you indeed
> do have to use a cascaded interrupt controller as you are doing, but
> I'm not sure it's really worthwhile.
> 
> Cheers,
> Ben.
> 
> 
Yes, there is two way, to handle irq.
One is use share irq, so my sdhci driver have following outcome when share 
irq, 
	cat /proc/interrupt
	----------
	34:         82      AVIC  26 Edge      mmc0, mmc1
	---------
when use cascaded interrupt. 
	.........
	292:         35  sdhci-ic   0 Edge      mmc0
	293:         42  sdhci-ic   1 Edge      mmc1
	---------
use separate slot (actually it is) interrupt, it have advantage for each slot 
debug. if use share. the debug will not easy.
Benjamin Herrenschmidt July 26, 2018, 3:55 a.m. UTC | #6
On Thu, 2018-07-26 at 11:26 +0800, Ryan Chen wrote:
> > 
> > 
> 
> Yes, there is two way, to handle irq.
> One is use share irq, so my sdhci driver have following outcome when share 
> irq, 
>         cat /proc/interrupt
>         ----------
>         34:         82      AVIC  26 Edge      mmc0, mmc1
>         ---------
> when use cascaded interrupt. 
>         .........
>         292:         35  sdhci-ic   0 Edge      mmc0
>         293:         42  sdhci-ic   1 Edge      mmc1
>         ---------
> use separate slot (actually it is) interrupt, it have advantage for each slot 
> debug. if use share. the debug will not easy.

I see ... well, you can then continue doing a cascaded controller if
you wish then. I sitll don't think you need to have separate drivers
however, but that's up to you.

You definitely don't want to use core initcalls. Also having a single
driver might make it easier to implement the card detect stuff which is
done by the "common" part, isn't it ?

Cheers,
Ben.
Ryan Chen July 26, 2018, 9:10 a.m. UTC | #7
On Thu, Jul 26, 2018 at 11:41:49AM +1000, Benjamin Herrenschmidt wrote:
> On Tue, 2018-07-17 at 15:26 +0800, Ryan Chen wrote:
> > In Aspeed SoC's sdhci have two slots and have a interrupt status and
> > general information in top for register. add sdhci core driver
> > defore sdhci driver probe
> > 
> > V0->V1
> > move irqchip/irq-aspeed-sdhci-ic.c to drivers/mmc/host/aspeed-sdhci-core.c
> > 
> > Signed-off-by: Ryan Chen <ryanchen.aspeed@gmail.com>
> > ---
> >  drivers/irqchip/Makefile              |   2 +-
> >  drivers/mmc/host/Makefile             |   1 +
> >  drivers/mmc/host/aspeed-sdhci-core.c  | 147 ++++++++++++++++++++++++++++++++++
> >  include/linux/mmc/sdhci-aspeed-data.h |  28 +++++++
> >  4 files changed, 177 insertions(+), 1 deletion(-)
> >  create mode 100644 drivers/mmc/host/aspeed-sdhci-core.c
> >  create mode 100644 include/linux/mmc/sdhci-aspeed-data.h
> > 
> > diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
> > index 5ed465a..d99cb1d 100644
> > --- a/drivers/irqchip/Makefile
> > +++ b/drivers/irqchip/Makefile
> > @@ -78,7 +78,7 @@ obj-$(CONFIG_MVEBU_ODMI)		+= irq-mvebu-odmi.o
> >  obj-$(CONFIG_MVEBU_PIC)			+= irq-mvebu-pic.o
> >  obj-$(CONFIG_LS_SCFG_MSI)		+= irq-ls-scfg-msi.o
> >  obj-$(CONFIG_EZNPS_GIC)			+= irq-eznps.o
> > -obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o
> > +obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o 
> >  obj-$(CONFIG_STM32_EXTI) 		+= irq-stm32-exti.o
> >  obj-$(CONFIG_QCOM_IRQ_COMBINER)		+= qcom-irq-combiner.o
> >  obj-$(CONFIG_IRQ_UNIPHIER_AIDET)	+= irq-uniphier-aidet.o
> > diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
> > index 6aead24..c3e5a1f 100644
> > --- a/drivers/mmc/host/Makefile
> > +++ b/drivers/mmc/host/Makefile
> > @@ -78,6 +78,7 @@ obj-$(CONFIG_MMC_SDHCI_ESDHC_IMX)	+= sdhci-esdhc-imx.o
> >  obj-$(CONFIG_MMC_SDHCI_DOVE)		+= sdhci-dove.o
> >  obj-$(CONFIG_MMC_SDHCI_TEGRA)		+= sdhci-tegra.o
> >  obj-$(CONFIG_MMC_SDHCI_OF_ARASAN)	+= sdhci-of-arasan.o
> > +obj-$(CONFIG_MMC_SDHCI_OF_ASPEED)	+= aspeed-sdhci-core.o
> >  obj-$(CONFIG_MMC_SDHCI_OF_AT91)		+= sdhci-of-at91.o
> >  obj-$(CONFIG_MMC_SDHCI_OF_ESDHC)	+= sdhci-of-esdhc.o
> >  obj-$(CONFIG_MMC_SDHCI_OF_HLWD)		+= sdhci-of-hlwd.o
> > diff --git a/drivers/mmc/host/aspeed-sdhci-core.c b/drivers/mmc/host/aspeed-sdhci-core.c
> > new file mode 100644
> > index 0000000..c3d9d45
> > --- /dev/null
> > +++ b/drivers/mmc/host/aspeed-sdhci-core.c
> > @@ -0,0 +1,147 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * SDHCI IRQCHIP driver for the Aspeed SoC
> > + *
> > + * Copyright (C) ASPEED Technology Inc.
> > + * Ryan Chen <ryan_chen@aspeedtech.com>
> > + *
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License as published by
> > + * the Free Software Foundation; either version 2 of the License, or (at
> > + * your option) any later version.
> > + *
> > + */
> > +
> > +#include <linux/platform_device.h>
> > +#include <linux/module.h>
> > +#include <linux/delay.h>
> > +#include <linux/irq.h>
> > +#include <linux/irqchip.h>
> > +#include <linux/irqchip/chained_irq.h>
> > +#include <linux/irqdomain.h>
> > +#include <linux/of_address.h>
> > +#include <linux/of_irq.h>
> > +#include <linux/of.h>
> > +#include <linux/io.h>
> > +#include <linux/clk.h>
> > +#include <linux/mmc/sdhci-aspeed-data.h>
> > +
> > +#define ASPEED_SDHCI_SLOT_NUM			2
> > 
> 
> Change the "irq" prefixes on the functions.. Dont' do a chained
> handler, that will be too much overhead for no benefit. Your SoC aren't
> very fast and you want to avoid that overhead.
> 
> Just do a normal interrupt handler, and have it call each port
> interrupts.
> 
> This isn't really an interrupt controller, it has no
> enable/disable/masking ability, I would just create the ports directly
> from the same driver and process the interrupts.
> 
> I don't think there's benefit in keeping the ports as separate drivers.
> 
> > +static void aspeed_sdhci_irq_handler(struct irq_desc *desc)
> > +{
> > +	struct aspeed_sdhci_irq *sdhci_irq = irq_desc_get_handler_data(desc);
> > +	struct irq_chip *chip = irq_desc_get_chip(desc);
> > +	unsigned long bit, status;
> > +	unsigned int slot_irq;
> > +
> > +	chained_irq_enter(chip, desc);
> > +	status = readl(sdhci_irq->regs + ASPEED_SDHCI_ISR);
> > +	for_each_set_bit(bit, &status, ASPEED_SDHCI_SLOT_NUM) {
> > +		slot_irq = irq_find_mapping(sdhci_irq->irq_domain, bit);
> > +		generic_handle_irq(slot_irq);
> > +	}
> > +	chained_irq_exit(chip, desc);
> > +}
> > +
> > +static void noop(struct irq_data *data) { }
> > +
> > +static unsigned int noop_ret(struct irq_data *data)
> > +{
> > +	return 0;
> > +}
> > +
> > +struct irq_chip sdhci_irq_chip = {
> > +	.name		= "sdhci-ic",
> > +	.irq_startup	= noop_ret,
> > +	.irq_shutdown	= noop,
> > +	.irq_enable	= noop,
> > +	.irq_disable	= noop,
> > +	.irq_ack	= noop,
> > +	.irq_mask	= noop,
> > +	.irq_unmask	= noop,
> > +	.flags		= IRQCHIP_SKIP_SET_WAKE,
> > +};
> > +
> > +static int ast_sdhci_map_irq_domain(struct irq_domain *domain,
> > +				    unsigned int irq, irq_hw_number_t hwirq)
> > +{
> > +	irq_set_chip_and_handler(irq, &sdhci_irq_chip, handle_simple_irq);
> > +	irq_set_chip_data(irq, domain->host_data);
> > +
> > +	return 0;
> > +}
> > +
> > +static const struct irq_domain_ops aspeed_sdhci_irq_domain_ops = {
> > +	.map = ast_sdhci_map_irq_domain,
> > +};
> > +
> > +static int irq_aspeed_sdhci_probe(struct platform_device *pdev)
> > +{
> > +	struct aspeed_sdhci_irq *sdhci_irq;
> > +	struct clk *sdclk;
> > +
> > +	sdhci_irq = kzalloc(sizeof(*sdhci_irq), GFP_KERNEL);
> > +	if (!sdhci_irq)
> > +		return -ENOMEM;
> > +
> > +	platform_set_drvdata(pdev, sdhci_irq);
> > +	//node->data = sdhci_irq;
> > +	pdev->dev.of_node->data = sdhci_irq;
> > +
> > +	sdhci_irq->regs = of_iomap(pdev->dev.of_node, 0);
> > +	if (IS_ERR(sdhci_irq->regs))
> > +		return PTR_ERR(sdhci_irq->regs);
> > +
> > +	sdclk = devm_clk_get(&pdev->dev, NULL);
> > +	if (IS_ERR(sdclk)) {
> > +		dev_err(&pdev->dev, "no sdclk clock defined\n");
> > +		return PTR_ERR(sdclk);
> > +	}
> > +
> > +	clk_prepare_enable(sdclk);
> > +
> > +	sdhci_irq->parent_irq = irq_of_parse_and_map(pdev->dev.of_node, 0);
> > +	if (sdhci_irq->parent_irq < 0)
> > +		return sdhci_irq->parent_irq;
> > +
> > +	sdhci_irq->irq_domain = irq_domain_add_linear(
> > +					pdev->dev.of_node, ASPEED_SDHCI_SLOT_NUM,
> > +					&aspeed_sdhci_irq_domain_ops, NULL);
> > +	if (!sdhci_irq->irq_domain)
> > +		return -ENOMEM;
> > +
> > +	sdhci_irq->irq_domain->name = "aspeed-sdhci-irq";
> > +
> > +	irq_set_chained_handler_and_data(sdhci_irq->parent_irq,
> > +					 aspeed_sdhci_irq_handler, sdhci_irq);
> > +
> > +	pr_info("sdhci irq controller registered, irq %d\n", sdhci_irq->parent_irq);
> > +
> > +	return 0;
> > +}
> > +
> > +static const struct of_device_id irq_aspeed_sdhci_dt_ids[] = {
> > +	{ .compatible = "aspeed,aspeed-sdhci-irq", },
> > +	{},
> > +};
> 
> You should call this aspeed,ast2{4,5}00-sdhci-core. I would change your
> device-tree representation as follow (add back all the reg &
> properties):
> 
> 	sdhci-core {
> 		reg = <common regs>
> 		ranges = either empty prop or the 2 ranges goign to the children
> 		compatible = <aspeed,ast2400-sdhci-core>;
> 
> 		sdhci-slot@xxx {
> 			compatible = <aspeed,ast2400-sdhci-slot>;
> 		};
> 
> 		sdhci-slot@yyy {
> 			compatible = <aspeed,ast2400-sdhci-slot>;
> 		};
> 	}
> 

If i keep the interrupt controller for dispatch slot. 
Is the following is compromise way for dts? 
Each slot need interrupt-parent.

&sdhci {

	sdhci_core: interrupt-controller@0 {
		#interrupt-cells = <1>;
		compatible = "aspeed,ast2500-sdhci-core";
		reg = <0x0 0x100>;
		interrupts = <26>;
		interrupt-controller;
		clocks = <&syscon ASPEED_CLK_GATE_SDCLKCLK>;
	};

	sdhci_slot0: sdhci_slot@100 {
		compatible = "aspeed,ast2500-sdhci-slot";
		reg = <0x100 0x100>;
		interrupts = <0>;
		interrupt-parent = <&sdhci_core>;
		sdhci,auto-cmd12;
		clocks = <&syscon ASPEED_CLK_SDIO>;
		status = "disabled";
	};

	sdhci_slot1: sdhci_slot@200 {
		compatible = "aspeed,ast2500-sdhci-slot";
		reg = <0x200 0x100>;
		interrupts = <1>;
		interrupt-parent = <&sdhci_core>;
		sdhci,auto-cmd12;
		clocks = <&syscon ASPEED_CLK_SDIO>;
		status = "disabled";
	};

};



> > +MODULE_DEVICE_TABLE(of, irq_aspeed_sdhci_dt_ids);
> > +
> > +static struct platform_driver irq_aspeed_sdhci_device_driver = {
> > +	.probe		= irq_aspeed_sdhci_probe,
> > +	.driver		= {
> > +		.name   = KBUILD_MODNAME,
> > +		.of_match_table	= irq_aspeed_sdhci_dt_ids,
> > +	}
> > +};
> > +
> > +static int __init irq_aspeed_sdhci_init(void)
> > +{
> > +	return platform_driver_register(&irq_aspeed_sdhci_device_driver);
> > +}
> > +core_initcall(irq_aspeed_sdhci_init);
> 
> This should be a normal device init call, thus module_platform_driver
> is all you need here.
> 
> > +
> > +MODULE_AUTHOR("Ryan Chen");
> > +MODULE_DESCRIPTION("ASPEED SOC SDHCI IRQ Driver");
> > +MODULE_LICENSE("GPL v2");
> > diff --git a/include/linux/mmc/sdhci-aspeed-data.h b/include/linux/mmc/sdhci-aspeed-data.h
> > new file mode 100644
> > index 0000000..fba2bf2
> > --- /dev/null
> > +++ b/include/linux/mmc/sdhci-aspeed-data.h
> > @@ -0,0 +1,28 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +#ifndef LINUX_MMC_SDHCI_ASPEED_DATA_H
> > +#define LINUX_MMC_SDHCI_ASPEED_DATA_H
> > +
> > +#include <linux/io.h>
> > +
> > +#define ASPEED_SDHCI_INFO			0x00
> > +#define  ASPEED_SDHCI_S1MMC8			BIT(25)
> > +#define  ASPEED_SDHCI_S0MMC8			BIT(24)
> > +#define ASPEED_SDHCI_BLOCK			0x04
> > +#define ASPEED_SDHCI_CTRL			0xF0
> > +#define ASPEED_SDHCI_ISR			0xFC
> > +
> > +struct aspeed_sdhci_irq {
> > +	void __iomem *regs;
> > +	int parent_irq;
> > +	struct irq_domain *irq_domain;
> > +};
> > +
> > +static inline void aspeed_sdhci_set_8bit_mode(struct aspeed_sdhci_irq *sdhci_irq, int mode)
> > +{
> > +	if (mode)
> > +		writel(ASPEED_SDHCI_S0MMC8 | readl(sdhci_irq->regs), sdhci_irq->regs);
> > +	else
> > +		writel(~ASPEED_SDHCI_S0MMC8 & readl(sdhci_irq->regs), sdhci_irq->regs);
> > +}
> > +
> > +#endif
>
Benjamin Herrenschmidt July 26, 2018, 10:37 p.m. UTC | #8
On Thu, 2018-07-26 at 17:10 +0800, Ryan Chen wrote:
> On
> 
> If i keep the interrupt controller for dispatch slot. 
> Is the following is compromise way for dts? 
> Each slot need interrupt-parent.
> 
> &sdhci {
> 
> 	sdhci_core: interrupt-controller@0 {
> 		#interrupt-cells = <1>;
> 		compatible = "aspeed,ast2500-sdhci-core";
> 		reg = <0x0 0x100>;
> 		interrupts = <26>;
> 		interrupt-controller;
> 		clocks = <&syscon ASPEED_CLK_GATE_SDCLKCLK>;
> 	};
> 
> 	sdhci_slot0: sdhci_slot@100 {
> 		compatible = "aspeed,ast2500-sdhci-slot";
> 		reg = <0x100 0x100>;
> 		interrupts = <0>;
> 		interrupt-parent = <&sdhci_core>;
> 		sdhci,auto-cmd12;
> 		clocks = <&syscon ASPEED_CLK_SDIO>;
> 		status = "disabled";
> 	};
> 
> 	sdhci_slot1: sdhci_slot@200 {
> 		compatible = "aspeed,ast2500-sdhci-slot";
> 		reg = <0x200 0x100>;
> 		interrupts = <1>;
> 		interrupt-parent = <&sdhci_core>;
> 		sdhci,auto-cmd12;
> 		clocks = <&syscon ASPEED_CLK_SDIO>;
> 		status = "disabled";
> 	};

I think the "core" is more than just the interrupt controller. The
slots should be the children of it.

You don't necessarily need to treat the slots as being matched by a
separate driver, it complicates things imho, but up to you.

Ben.

> };
> 
> 
> 
> > > +MODULE_DEVICE_TABLE(of, irq_aspeed_sdhci_dt_ids);
> > > +
> > > +static struct platform_driver irq_aspeed_sdhci_device_driver = {
> > > +	.probe		= irq_aspeed_sdhci_probe,
> > > +	.driver		= {
> > > +		.name   = KBUILD_MODNAME,
> > > +		.of_match_table	= irq_aspeed_sdhci_dt_ids,
> > > +	}
> > > +};
> > > +
> > > +static int __init irq_aspeed_sdhci_init(void)
> > > +{
> > > +	return platform_driver_register(&irq_aspeed_sdhci_device_driver);
> > > +}
> > > +core_initcall(irq_aspeed_sdhci_init);
> > 
> > This should be a normal device init call, thus module_platform_driver
> > is all you need here.
> > 
> > > +
> > > +MODULE_AUTHOR("Ryan Chen");
> > > +MODULE_DESCRIPTION("ASPEED SOC SDHCI IRQ Driver");
> > > +MODULE_LICENSE("GPL v2");
> > > diff --git a/include/linux/mmc/sdhci-aspeed-data.h b/include/linux/mmc/sdhci-aspeed-data.h
> > > new file mode 100644
> > > index 0000000..fba2bf2
> > > --- /dev/null
> > > +++ b/include/linux/mmc/sdhci-aspeed-data.h
> > > @@ -0,0 +1,28 @@
> > > +/* SPDX-License-Identifier: GPL-2.0 */
> > > +#ifndef LINUX_MMC_SDHCI_ASPEED_DATA_H
> > > +#define LINUX_MMC_SDHCI_ASPEED_DATA_H
> > > +
> > > +#include <linux/io.h>
> > > +
> > > +#define ASPEED_SDHCI_INFO			0x00
> > > +#define  ASPEED_SDHCI_S1MMC8			BIT(25)
> > > +#define  ASPEED_SDHCI_S0MMC8			BIT(24)
> > > +#define ASPEED_SDHCI_BLOCK			0x04
> > > +#define ASPEED_SDHCI_CTRL			0xF0
> > > +#define ASPEED_SDHCI_ISR			0xFC
> > > +
> > > +struct aspeed_sdhci_irq {
> > > +	void __iomem *regs;
> > > +	int parent_irq;
> > > +	struct irq_domain *irq_domain;
> > > +};
> > > +
> > > +static inline void aspeed_sdhci_set_8bit_mode(struct aspeed_sdhci_irq *sdhci_irq, int mode)
> > > +{
> > > +	if (mode)
> > > +		writel(ASPEED_SDHCI_S0MMC8 | readl(sdhci_irq->regs), sdhci_irq->regs);
> > > +	else
> > > +		writel(~ASPEED_SDHCI_S0MMC8 & readl(sdhci_irq->regs), sdhci_irq->regs);
> > > +}
> > > +
> > > +#endif
diff mbox series

Patch

diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
index 5ed465a..d99cb1d 100644
--- a/drivers/irqchip/Makefile
+++ b/drivers/irqchip/Makefile
@@ -78,7 +78,7 @@  obj-$(CONFIG_MVEBU_ODMI)		+= irq-mvebu-odmi.o
 obj-$(CONFIG_MVEBU_PIC)			+= irq-mvebu-pic.o
 obj-$(CONFIG_LS_SCFG_MSI)		+= irq-ls-scfg-msi.o
 obj-$(CONFIG_EZNPS_GIC)			+= irq-eznps.o
-obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o
+obj-$(CONFIG_ARCH_ASPEED)		+= irq-aspeed-vic.o irq-aspeed-i2c-ic.o 
 obj-$(CONFIG_STM32_EXTI) 		+= irq-stm32-exti.o
 obj-$(CONFIG_QCOM_IRQ_COMBINER)		+= qcom-irq-combiner.o
 obj-$(CONFIG_IRQ_UNIPHIER_AIDET)	+= irq-uniphier-aidet.o
diff --git a/drivers/mmc/host/Makefile b/drivers/mmc/host/Makefile
index 6aead24..c3e5a1f 100644
--- a/drivers/mmc/host/Makefile
+++ b/drivers/mmc/host/Makefile
@@ -78,6 +78,7 @@  obj-$(CONFIG_MMC_SDHCI_ESDHC_IMX)	+= sdhci-esdhc-imx.o
 obj-$(CONFIG_MMC_SDHCI_DOVE)		+= sdhci-dove.o
 obj-$(CONFIG_MMC_SDHCI_TEGRA)		+= sdhci-tegra.o
 obj-$(CONFIG_MMC_SDHCI_OF_ARASAN)	+= sdhci-of-arasan.o
+obj-$(CONFIG_MMC_SDHCI_OF_ASPEED)	+= aspeed-sdhci-core.o
 obj-$(CONFIG_MMC_SDHCI_OF_AT91)		+= sdhci-of-at91.o
 obj-$(CONFIG_MMC_SDHCI_OF_ESDHC)	+= sdhci-of-esdhc.o
 obj-$(CONFIG_MMC_SDHCI_OF_HLWD)		+= sdhci-of-hlwd.o
diff --git a/drivers/mmc/host/aspeed-sdhci-core.c b/drivers/mmc/host/aspeed-sdhci-core.c
new file mode 100644
index 0000000..c3d9d45
--- /dev/null
+++ b/drivers/mmc/host/aspeed-sdhci-core.c
@@ -0,0 +1,147 @@ 
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * SDHCI IRQCHIP driver for the Aspeed SoC
+ *
+ * Copyright (C) ASPEED Technology Inc.
+ * Ryan Chen <ryan_chen@aspeedtech.com>
+ *
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or (at
+ * your option) any later version.
+ *
+ */
+
+#include <linux/platform_device.h>
+#include <linux/module.h>
+#include <linux/delay.h>
+#include <linux/irq.h>
+#include <linux/irqchip.h>
+#include <linux/irqchip/chained_irq.h>
+#include <linux/irqdomain.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of.h>
+#include <linux/io.h>
+#include <linux/clk.h>
+#include <linux/mmc/sdhci-aspeed-data.h>
+
+#define ASPEED_SDHCI_SLOT_NUM			2
+
+static void aspeed_sdhci_irq_handler(struct irq_desc *desc)
+{
+	struct aspeed_sdhci_irq *sdhci_irq = irq_desc_get_handler_data(desc);
+	struct irq_chip *chip = irq_desc_get_chip(desc);
+	unsigned long bit, status;
+	unsigned int slot_irq;
+
+	chained_irq_enter(chip, desc);
+	status = readl(sdhci_irq->regs + ASPEED_SDHCI_ISR);
+	for_each_set_bit(bit, &status, ASPEED_SDHCI_SLOT_NUM) {
+		slot_irq = irq_find_mapping(sdhci_irq->irq_domain, bit);
+		generic_handle_irq(slot_irq);
+	}
+	chained_irq_exit(chip, desc);
+}
+
+static void noop(struct irq_data *data) { }
+
+static unsigned int noop_ret(struct irq_data *data)
+{
+	return 0;
+}
+
+struct irq_chip sdhci_irq_chip = {
+	.name		= "sdhci-ic",
+	.irq_startup	= noop_ret,
+	.irq_shutdown	= noop,
+	.irq_enable	= noop,
+	.irq_disable	= noop,
+	.irq_ack	= noop,
+	.irq_mask	= noop,
+	.irq_unmask	= noop,
+	.flags		= IRQCHIP_SKIP_SET_WAKE,
+};
+
+static int ast_sdhci_map_irq_domain(struct irq_domain *domain,
+				    unsigned int irq, irq_hw_number_t hwirq)
+{
+	irq_set_chip_and_handler(irq, &sdhci_irq_chip, handle_simple_irq);
+	irq_set_chip_data(irq, domain->host_data);
+
+	return 0;
+}
+
+static const struct irq_domain_ops aspeed_sdhci_irq_domain_ops = {
+	.map = ast_sdhci_map_irq_domain,
+};
+
+static int irq_aspeed_sdhci_probe(struct platform_device *pdev)
+{
+	struct aspeed_sdhci_irq *sdhci_irq;
+	struct clk *sdclk;
+
+	sdhci_irq = kzalloc(sizeof(*sdhci_irq), GFP_KERNEL);
+	if (!sdhci_irq)
+		return -ENOMEM;
+
+	platform_set_drvdata(pdev, sdhci_irq);
+	//node->data = sdhci_irq;
+	pdev->dev.of_node->data = sdhci_irq;
+
+	sdhci_irq->regs = of_iomap(pdev->dev.of_node, 0);
+	if (IS_ERR(sdhci_irq->regs))
+		return PTR_ERR(sdhci_irq->regs);
+
+	sdclk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(sdclk)) {
+		dev_err(&pdev->dev, "no sdclk clock defined\n");
+		return PTR_ERR(sdclk);
+	}
+
+	clk_prepare_enable(sdclk);
+
+	sdhci_irq->parent_irq = irq_of_parse_and_map(pdev->dev.of_node, 0);
+	if (sdhci_irq->parent_irq < 0)
+		return sdhci_irq->parent_irq;
+
+	sdhci_irq->irq_domain = irq_domain_add_linear(
+					pdev->dev.of_node, ASPEED_SDHCI_SLOT_NUM,
+					&aspeed_sdhci_irq_domain_ops, NULL);
+	if (!sdhci_irq->irq_domain)
+		return -ENOMEM;
+
+	sdhci_irq->irq_domain->name = "aspeed-sdhci-irq";
+
+	irq_set_chained_handler_and_data(sdhci_irq->parent_irq,
+					 aspeed_sdhci_irq_handler, sdhci_irq);
+
+	pr_info("sdhci irq controller registered, irq %d\n", sdhci_irq->parent_irq);
+
+	return 0;
+}
+
+static const struct of_device_id irq_aspeed_sdhci_dt_ids[] = {
+	{ .compatible = "aspeed,aspeed-sdhci-irq", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, irq_aspeed_sdhci_dt_ids);
+
+static struct platform_driver irq_aspeed_sdhci_device_driver = {
+	.probe		= irq_aspeed_sdhci_probe,
+	.driver		= {
+		.name   = KBUILD_MODNAME,
+		.of_match_table	= irq_aspeed_sdhci_dt_ids,
+	}
+};
+
+static int __init irq_aspeed_sdhci_init(void)
+{
+	return platform_driver_register(&irq_aspeed_sdhci_device_driver);
+}
+core_initcall(irq_aspeed_sdhci_init);
+
+MODULE_AUTHOR("Ryan Chen");
+MODULE_DESCRIPTION("ASPEED SOC SDHCI IRQ Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/include/linux/mmc/sdhci-aspeed-data.h b/include/linux/mmc/sdhci-aspeed-data.h
new file mode 100644
index 0000000..fba2bf2
--- /dev/null
+++ b/include/linux/mmc/sdhci-aspeed-data.h
@@ -0,0 +1,28 @@ 
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef LINUX_MMC_SDHCI_ASPEED_DATA_H
+#define LINUX_MMC_SDHCI_ASPEED_DATA_H
+
+#include <linux/io.h>
+
+#define ASPEED_SDHCI_INFO			0x00
+#define  ASPEED_SDHCI_S1MMC8			BIT(25)
+#define  ASPEED_SDHCI_S0MMC8			BIT(24)
+#define ASPEED_SDHCI_BLOCK			0x04
+#define ASPEED_SDHCI_CTRL			0xF0
+#define ASPEED_SDHCI_ISR			0xFC
+
+struct aspeed_sdhci_irq {
+	void __iomem *regs;
+	int parent_irq;
+	struct irq_domain *irq_domain;
+};
+
+static inline void aspeed_sdhci_set_8bit_mode(struct aspeed_sdhci_irq *sdhci_irq, int mode)
+{
+	if (mode)
+		writel(ASPEED_SDHCI_S0MMC8 | readl(sdhci_irq->regs), sdhci_irq->regs);
+	else
+		writel(~ASPEED_SDHCI_S0MMC8 & readl(sdhci_irq->regs), sdhci_irq->regs);
+}
+
+#endif