diff mbox

[1/9] ARM: imx: move timer resources into a structure

Message ID 1431677507-27420-2-git-send-email-shawnguo@kernel.org
State New
Headers show

Commit Message

Shawn Guo May 15, 2015, 8:11 a.m. UTC
From: Shawn Guo <shawn.guo@linaro.org>

Instead of passing around as argument, let's move timer resources like
irq and clocks together with base address into a data structure, and
reference the resources from the struct variable directly to simplify
the function call interface.

Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
---
 arch/arm/mach-imx/time.c | 119 ++++++++++++++++++++++++-----------------------
 1 file changed, 61 insertions(+), 58 deletions(-)

Comments

Shenwei Wang May 15, 2015, 4:36 p.m. UTC | #1
> -----Original Message-----

> From: shawnguo@kernel.org [mailto:shawnguo@kernel.org]

> Sent: 2015年5月15日 3:12

> To: linux-arm-kernel@lists.infradead.org

> Cc: kernel@pengutronix.de; Daniel Lezcano; Wang Shenwei-B38339; Shawn Guo

> Subject: [PATCH 1/9] ARM: imx: move timer resources into a structure

> 

> From: Shawn Guo <shawn.guo@linaro.org>

> 

> Instead of passing around as argument, let's move timer resources like irq and

> clocks together with base address into a data structure, and reference the

> resources from the struct variable directly to simplify the function call interface.

> 

> Signed-off-by: Shawn Guo <shawn.guo@linaro.org>

> ---

>  arch/arm/mach-imx/time.c | 119 ++++++++++++++++++++++++-----------------------

>  1 file changed, 61 insertions(+), 58 deletions(-)

> 

> diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c index

> ab5ee1c445f3..8ad7cb2a7f08 100644

> --- a/arch/arm/mach-imx/time.c

> +++ b/arch/arm/mach-imx/time.c

> @@ -84,27 +84,34 @@

>  static struct clock_event_device clockevent_mxc;  static enum

> clock_event_mode clockevent_mode = CLOCK_EVT_MODE_UNUSED;

> 

> -static void __iomem *timer_base;

> +struct imx_timer {

> +	void __iomem *base;

> +	int irq;

> +	struct clk *clk_per;

> +	struct clk *clk_ipg;

> +};

> +

> +static struct imx_timer imxtm;

> 

Since the changes still used a global variable to hold the information, it doesn't take advantage of what the clocksource framework provides, and make the driver unable to support multi-instances.

Regards,
Shenwei

>  static inline void gpt_irq_disable(void)  {

>  	unsigned int tmp;

> 

>  	if (timer_is_v2())

> -		__raw_writel(0, timer_base + V2_IR);

> +		__raw_writel(0, imxtm.base + V2_IR);

>  	else {

> -		tmp = __raw_readl(timer_base + MXC_TCTL);

> -		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, timer_base + MXC_TCTL);

> +		tmp = __raw_readl(imxtm.base + MXC_TCTL);

> +		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, imxtm.base + MXC_TCTL);

>  	}

>  }

> 

>  static inline void gpt_irq_enable(void)  {

>  	if (timer_is_v2())

> -		__raw_writel(1<<0, timer_base + V2_IR);

> +		__raw_writel(1<<0, imxtm.base + V2_IR);

>  	else {

> -		__raw_writel(__raw_readl(timer_base + MXC_TCTL) |

> MX1_2_TCTL_IRQEN,

> -			timer_base + MXC_TCTL);

> +		__raw_writel(__raw_readl(imxtm.base + MXC_TCTL) |

> MX1_2_TCTL_IRQEN,

> +			imxtm.base + MXC_TCTL);

>  	}

>  }

> 

> @@ -112,12 +119,12 @@ static void gpt_irq_acknowledge(void)  {

>  	if (timer_is_v1()) {

>  		if (cpu_is_mx1())

> -			__raw_writel(0, timer_base + MX1_2_TSTAT);

> +			__raw_writel(0, imxtm.base + MX1_2_TSTAT);

>  		else

>  			__raw_writel(MX2_TSTAT_CAPT | MX2_TSTAT_COMP,

> -				timer_base + MX1_2_TSTAT);

> +				imxtm.base + MX1_2_TSTAT);

>  	} else if (timer_is_v2())

> -		__raw_writel(V2_TSTAT_OF1, timer_base + V2_TSTAT);

> +		__raw_writel(V2_TSTAT_OF1, imxtm.base + V2_TSTAT);

>  }

> 

>  static void __iomem *sched_clock_reg;

> @@ -134,10 +141,10 @@ static unsigned long imx_read_current_timer(void)

>  	return __raw_readl(sched_clock_reg);

>  }

> 

> -static int __init mxc_clocksource_init(struct clk *timer_clk)

> +static int __init mxc_clocksource_init(void)

>  {

> -	unsigned int c = clk_get_rate(timer_clk);

> -	void __iomem *reg = timer_base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);

> +	unsigned int c = clk_get_rate(imxtm.clk_per);

> +	void __iomem *reg = imxtm.base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);

> 

>  	imx_delay_timer.read_current_timer = &imx_read_current_timer;

>  	imx_delay_timer.freq = c;

> @@ -157,11 +164,11 @@ static int mx1_2_set_next_event(unsigned long evt,

> {

>  	unsigned long tcmp;

> 

> -	tcmp = __raw_readl(timer_base + MX1_2_TCN) + evt;

> +	tcmp = __raw_readl(imxtm.base + MX1_2_TCN) + evt;

> 

> -	__raw_writel(tcmp, timer_base + MX1_2_TCMP);

> +	__raw_writel(tcmp, imxtm.base + MX1_2_TCMP);

> 

> -	return (int)(tcmp - __raw_readl(timer_base + MX1_2_TCN)) < 0 ?

> +	return (int)(tcmp - __raw_readl(imxtm.base + MX1_2_TCN)) < 0 ?

>  				-ETIME : 0;

>  }

> 

> @@ -170,12 +177,12 @@ static int v2_set_next_event(unsigned long evt,  {

>  	unsigned long tcmp;

> 

> -	tcmp = __raw_readl(timer_base + V2_TCN) + evt;

> +	tcmp = __raw_readl(imxtm.base + V2_TCN) + evt;

> 

> -	__raw_writel(tcmp, timer_base + V2_TCMP);

> +	__raw_writel(tcmp, imxtm.base + V2_TCMP);

> 

>  	return evt < 0x7fffffff &&

> -		(int)(tcmp - __raw_readl(timer_base + V2_TCN)) < 0 ?

> +		(int)(tcmp - __raw_readl(imxtm.base + V2_TCN)) < 0 ?

>  				-ETIME : 0;

>  }

> 

> @@ -206,11 +213,11 @@ static void mxc_set_mode(enum clock_event_mode

> mode,

>  	if (mode != clockevent_mode) {

>  		/* Set event time into far-far future */

>  		if (timer_is_v2())

> -			__raw_writel(__raw_readl(timer_base + V2_TCN) - 3,

> -					timer_base + V2_TCMP);

> +			__raw_writel(__raw_readl(imxtm.base + V2_TCN) - 3,

> +					imxtm.base + V2_TCMP);

>  		else

> -			__raw_writel(__raw_readl(timer_base + MX1_2_TCN) - 3,

> -					timer_base + MX1_2_TCMP);

> +			__raw_writel(__raw_readl(imxtm.base + MX1_2_TCN) - 3,

> +					imxtm.base + MX1_2_TCMP);

> 

>  		/* Clear pending interrupt */

>  		gpt_irq_acknowledge();

> @@ -259,9 +266,9 @@ static irqreturn_t mxc_timer_interrupt(int irq, void

> *dev_id)

>  	uint32_t tstat;

> 

>  	if (timer_is_v2())

> -		tstat = __raw_readl(timer_base + V2_TSTAT);

> +		tstat = __raw_readl(imxtm.base + V2_TSTAT);

>  	else

> -		tstat = __raw_readl(timer_base + MX1_2_TSTAT);

> +		tstat = __raw_readl(imxtm.base + MX1_2_TSTAT);

> 

>  	gpt_irq_acknowledge();

> 

> @@ -284,49 +291,48 @@ static struct clock_event_device clockevent_mxc = {

>  	.rating		= 200,

>  };

> 

> -static int __init mxc_clockevent_init(struct clk *timer_clk)

> +static int __init mxc_clockevent_init(void)

>  {

>  	if (timer_is_v2())

>  		clockevent_mxc.set_next_event = v2_set_next_event;

> 

>  	clockevent_mxc.cpumask = cpumask_of(0);

>  	clockevents_config_and_register(&clockevent_mxc,

> -					clk_get_rate(timer_clk),

> +					clk_get_rate(imxtm.clk_per),

>  					0xff, 0xfffffffe);

> 

>  	return 0;

>  }

> 

> -static void __init _mxc_timer_init(int irq,

> -				   struct clk *clk_per, struct clk *clk_ipg)

> +static void __init _mxc_timer_init(void)

>  {

>  	uint32_t tctl_val;

> 

> -	if (IS_ERR(clk_per)) {

> +	if (IS_ERR(imxtm.clk_per)) {

>  		pr_err("i.MX timer: unable to get clk\n");

>  		return;

>  	}

> 

> -	if (!IS_ERR(clk_ipg))

> -		clk_prepare_enable(clk_ipg);

> +	if (!IS_ERR(imxtm.clk_ipg))

> +		clk_prepare_enable(imxtm.clk_ipg);

> 

> -	clk_prepare_enable(clk_per);

> +	clk_prepare_enable(imxtm.clk_per);

> 

>  	/*

>  	 * Initialise to a known state (all timers off, and timing reset)

>  	 */

> 

> -	__raw_writel(0, timer_base + MXC_TCTL);

> -	__raw_writel(0, timer_base + MXC_TPRER); /* see datasheet note */

> +	__raw_writel(0, imxtm.base + MXC_TCTL);

> +	__raw_writel(0, imxtm.base + MXC_TPRER); /* see datasheet note */

> 

>  	if (timer_is_v2()) {

>  		tctl_val = V2_TCTL_FRR | V2_TCTL_WAITEN | MXC_TCTL_TEN;

> -		if (clk_get_rate(clk_per) == V2_TIMER_RATE_OSC_DIV8) {

> +		if (clk_get_rate(imxtm.clk_per) == V2_TIMER_RATE_OSC_DIV8) {

>  			tctl_val |= V2_TCTL_CLK_OSC_DIV8;

>  			if (cpu_is_imx6dl() || cpu_is_imx6sx()) {

>  				/* 24 / 8 = 3 MHz */

>  				__raw_writel(7 << V2_TPRER_PRE24M,

> -					timer_base + MXC_TPRER);

> +					imxtm.base + MXC_TPRER);

>  				tctl_val |= V2_TCTL_24MEN;

>  			}

>  		} else {

> @@ -336,47 +342,44 @@ static void __init _mxc_timer_init(int irq,

>  		tctl_val = MX1_2_TCTL_FRR | MX1_2_TCTL_CLK_PCLK1 |

> MXC_TCTL_TEN;

>  	}

> 

> -	__raw_writel(tctl_val, timer_base + MXC_TCTL);

> +	__raw_writel(tctl_val, imxtm.base + MXC_TCTL);

> 

>  	/* init and register the timer to the framework */

> -	mxc_clocksource_init(clk_per);

> -	mxc_clockevent_init(clk_per);

> +	mxc_clocksource_init();

> +	mxc_clockevent_init();

> 

>  	/* Make irqs happen */

> -	setup_irq(irq, &mxc_timer_irq);

> +	setup_irq(imxtm.irq, &mxc_timer_irq);

>  }

> 

>  void __init mxc_timer_init(unsigned long pbase, int irq)  {

> -	struct clk *clk_per = clk_get_sys("imx-gpt.0", "per");

> -	struct clk *clk_ipg = clk_get_sys("imx-gpt.0", "ipg");

> +	imxtm.clk_per = clk_get_sys("imx-gpt.0", "per");

> +	imxtm.clk_ipg = clk_get_sys("imx-gpt.0", "ipg");

> 

> -	timer_base = ioremap(pbase, SZ_4K);

> -	BUG_ON(!timer_base);

> +	imxtm.base = ioremap(pbase, SZ_4K);

> +	BUG_ON(!imxtm.base);

> 

> -	_mxc_timer_init(irq, clk_per, clk_ipg);

> +	_mxc_timer_init();

>  }

> 

>  static void __init mxc_timer_init_dt(struct device_node *np)  {

> -	struct clk *clk_per, *clk_ipg;

> -	int irq;

> -

> -	if (timer_base)

> +	if (imxtm.base)

>  		return;

> 

> -	timer_base = of_iomap(np, 0);

> -	WARN_ON(!timer_base);

> -	irq = irq_of_parse_and_map(np, 0);

> +	imxtm.base = of_iomap(np, 0);

> +	WARN_ON(!imxtm.base);

> +	imxtm.irq = irq_of_parse_and_map(np, 0);

> 

> -	clk_ipg = of_clk_get_by_name(np, "ipg");

> +	imxtm.clk_ipg = of_clk_get_by_name(np, "ipg");

> 

>  	/* Try osc_per first, and fall back to per otherwise */

> -	clk_per = of_clk_get_by_name(np, "osc_per");

> -	if (IS_ERR(clk_per))

> -		clk_per = of_clk_get_by_name(np, "per");

> +	imxtm.clk_per = of_clk_get_by_name(np, "osc_per");

> +	if (IS_ERR(imxtm.clk_per))

> +		imxtm.clk_per = of_clk_get_by_name(np, "per");

> 

> -	_mxc_timer_init(irq, clk_per, clk_ipg);

> +	_mxc_timer_init();

>  }

>  CLOCKSOURCE_OF_DECLARE(mx1_timer, "fsl,imx1-gpt", mxc_timer_init_dt);

> CLOCKSOURCE_OF_DECLARE(mx25_timer, "fsl,imx25-gpt", mxc_timer_init_dt);

> --

> 1.9.1
Daniel Lezcano May 18, 2015, 10:43 a.m. UTC | #2
On 05/15/2015 10:11 AM, shawnguo@kernel.org wrote:
> From: Shawn Guo <shawn.guo@linaro.org>
>
> Instead of passing around as argument, let's move timer resources like
> irq and clocks together with base address into a data structure, and
> reference the resources from the struct variable directly to simplify
> the function call interface.
>
> Signed-off-by: Shawn Guo <shawn.guo@linaro.org>

I would be nice to embed the clockevent_mxc variable in the structure as 
well and use container_of to retrieve the structure from the struct 
clockevent.

The struct clockevent can be the unified parameter across the different 
functions.

Refer to drivers/clocksource/rockchip_timer.c as an example.

> ---
>   arch/arm/mach-imx/time.c | 119 ++++++++++++++++++++++++-----------------------
>   1 file changed, 61 insertions(+), 58 deletions(-)
>
> diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c
> index ab5ee1c445f3..8ad7cb2a7f08 100644
> --- a/arch/arm/mach-imx/time.c
> +++ b/arch/arm/mach-imx/time.c
> @@ -84,27 +84,34 @@
>   static struct clock_event_device clockevent_mxc;
>   static enum clock_event_mode clockevent_mode = CLOCK_EVT_MODE_UNUSED;
>
> -static void __iomem *timer_base;
> +struct imx_timer {
> +	void __iomem *base;
> +	int irq;
> +	struct clk *clk_per;
> +	struct clk *clk_ipg;
> +};
> +
> +static struct imx_timer imxtm;
>
>   static inline void gpt_irq_disable(void)
>   {
>   	unsigned int tmp;
>
>   	if (timer_is_v2())
> -		__raw_writel(0, timer_base + V2_IR);
> +		__raw_writel(0, imxtm.base + V2_IR);
>   	else {
> -		tmp = __raw_readl(timer_base + MXC_TCTL);
> -		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, timer_base + MXC_TCTL);
> +		tmp = __raw_readl(imxtm.base + MXC_TCTL);
> +		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, imxtm.base + MXC_TCTL);
>   	}
>   }
>
>   static inline void gpt_irq_enable(void)
>   {
>   	if (timer_is_v2())
> -		__raw_writel(1<<0, timer_base + V2_IR);
> +		__raw_writel(1<<0, imxtm.base + V2_IR);
>   	else {
> -		__raw_writel(__raw_readl(timer_base + MXC_TCTL) | MX1_2_TCTL_IRQEN,
> -			timer_base + MXC_TCTL);
> +		__raw_writel(__raw_readl(imxtm.base + MXC_TCTL) | MX1_2_TCTL_IRQEN,
> +			imxtm.base + MXC_TCTL);
>   	}
>   }
>
> @@ -112,12 +119,12 @@ static void gpt_irq_acknowledge(void)
>   {
>   	if (timer_is_v1()) {
>   		if (cpu_is_mx1())
> -			__raw_writel(0, timer_base + MX1_2_TSTAT);
> +			__raw_writel(0, imxtm.base + MX1_2_TSTAT);
>   		else
>   			__raw_writel(MX2_TSTAT_CAPT | MX2_TSTAT_COMP,
> -				timer_base + MX1_2_TSTAT);
> +				imxtm.base + MX1_2_TSTAT);
>   	} else if (timer_is_v2())
> -		__raw_writel(V2_TSTAT_OF1, timer_base + V2_TSTAT);
> +		__raw_writel(V2_TSTAT_OF1, imxtm.base + V2_TSTAT);
>   }
>
>   static void __iomem *sched_clock_reg;
> @@ -134,10 +141,10 @@ static unsigned long imx_read_current_timer(void)
>   	return __raw_readl(sched_clock_reg);
>   }
>
> -static int __init mxc_clocksource_init(struct clk *timer_clk)
> +static int __init mxc_clocksource_init(void)
>   {
> -	unsigned int c = clk_get_rate(timer_clk);
> -	void __iomem *reg = timer_base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);
> +	unsigned int c = clk_get_rate(imxtm.clk_per);
> +	void __iomem *reg = imxtm.base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);
>
>   	imx_delay_timer.read_current_timer = &imx_read_current_timer;
>   	imx_delay_timer.freq = c;
> @@ -157,11 +164,11 @@ static int mx1_2_set_next_event(unsigned long evt,
>   {
>   	unsigned long tcmp;
>
> -	tcmp = __raw_readl(timer_base + MX1_2_TCN) + evt;
> +	tcmp = __raw_readl(imxtm.base + MX1_2_TCN) + evt;
>
> -	__raw_writel(tcmp, timer_base + MX1_2_TCMP);
> +	__raw_writel(tcmp, imxtm.base + MX1_2_TCMP);
>
> -	return (int)(tcmp - __raw_readl(timer_base + MX1_2_TCN)) < 0 ?
> +	return (int)(tcmp - __raw_readl(imxtm.base + MX1_2_TCN)) < 0 ?
>   				-ETIME : 0;
>   }
>
> @@ -170,12 +177,12 @@ static int v2_set_next_event(unsigned long evt,
>   {
>   	unsigned long tcmp;
>
> -	tcmp = __raw_readl(timer_base + V2_TCN) + evt;
> +	tcmp = __raw_readl(imxtm.base + V2_TCN) + evt;
>
> -	__raw_writel(tcmp, timer_base + V2_TCMP);
> +	__raw_writel(tcmp, imxtm.base + V2_TCMP);
>
>   	return evt < 0x7fffffff &&
> -		(int)(tcmp - __raw_readl(timer_base + V2_TCN)) < 0 ?
> +		(int)(tcmp - __raw_readl(imxtm.base + V2_TCN)) < 0 ?
>   				-ETIME : 0;
>   }
>
> @@ -206,11 +213,11 @@ static void mxc_set_mode(enum clock_event_mode mode,
>   	if (mode != clockevent_mode) {
>   		/* Set event time into far-far future */
>   		if (timer_is_v2())
> -			__raw_writel(__raw_readl(timer_base + V2_TCN) - 3,
> -					timer_base + V2_TCMP);
> +			__raw_writel(__raw_readl(imxtm.base + V2_TCN) - 3,
> +					imxtm.base + V2_TCMP);
>   		else
> -			__raw_writel(__raw_readl(timer_base + MX1_2_TCN) - 3,
> -					timer_base + MX1_2_TCMP);
> +			__raw_writel(__raw_readl(imxtm.base + MX1_2_TCN) - 3,
> +					imxtm.base + MX1_2_TCMP);
>
>   		/* Clear pending interrupt */
>   		gpt_irq_acknowledge();
> @@ -259,9 +266,9 @@ static irqreturn_t mxc_timer_interrupt(int irq, void *dev_id)
>   	uint32_t tstat;
>
>   	if (timer_is_v2())
> -		tstat = __raw_readl(timer_base + V2_TSTAT);
> +		tstat = __raw_readl(imxtm.base + V2_TSTAT);
>   	else
> -		tstat = __raw_readl(timer_base + MX1_2_TSTAT);
> +		tstat = __raw_readl(imxtm.base + MX1_2_TSTAT);
>
>   	gpt_irq_acknowledge();
>
> @@ -284,49 +291,48 @@ static struct clock_event_device clockevent_mxc = {
>   	.rating		= 200,
>   };
>
> -static int __init mxc_clockevent_init(struct clk *timer_clk)
> +static int __init mxc_clockevent_init(void)
>   {
>   	if (timer_is_v2())
>   		clockevent_mxc.set_next_event = v2_set_next_event;
>
>   	clockevent_mxc.cpumask = cpumask_of(0);
>   	clockevents_config_and_register(&clockevent_mxc,
> -					clk_get_rate(timer_clk),
> +					clk_get_rate(imxtm.clk_per),
>   					0xff, 0xfffffffe);
>
>   	return 0;
>   }
>
> -static void __init _mxc_timer_init(int irq,
> -				   struct clk *clk_per, struct clk *clk_ipg)
> +static void __init _mxc_timer_init(void)
>   {
>   	uint32_t tctl_val;
>
> -	if (IS_ERR(clk_per)) {
> +	if (IS_ERR(imxtm.clk_per)) {
>   		pr_err("i.MX timer: unable to get clk\n");
>   		return;
>   	}
>
> -	if (!IS_ERR(clk_ipg))
> -		clk_prepare_enable(clk_ipg);
> +	if (!IS_ERR(imxtm.clk_ipg))
> +		clk_prepare_enable(imxtm.clk_ipg);
>
> -	clk_prepare_enable(clk_per);
> +	clk_prepare_enable(imxtm.clk_per);
>
>   	/*
>   	 * Initialise to a known state (all timers off, and timing reset)
>   	 */
>
> -	__raw_writel(0, timer_base + MXC_TCTL);
> -	__raw_writel(0, timer_base + MXC_TPRER); /* see datasheet note */
> +	__raw_writel(0, imxtm.base + MXC_TCTL);
> +	__raw_writel(0, imxtm.base + MXC_TPRER); /* see datasheet note */
>
>   	if (timer_is_v2()) {
>   		tctl_val = V2_TCTL_FRR | V2_TCTL_WAITEN | MXC_TCTL_TEN;
> -		if (clk_get_rate(clk_per) == V2_TIMER_RATE_OSC_DIV8) {
> +		if (clk_get_rate(imxtm.clk_per) == V2_TIMER_RATE_OSC_DIV8) {
>   			tctl_val |= V2_TCTL_CLK_OSC_DIV8;
>   			if (cpu_is_imx6dl() || cpu_is_imx6sx()) {
>   				/* 24 / 8 = 3 MHz */
>   				__raw_writel(7 << V2_TPRER_PRE24M,
> -					timer_base + MXC_TPRER);
> +					imxtm.base + MXC_TPRER);
>   				tctl_val |= V2_TCTL_24MEN;
>   			}
>   		} else {
> @@ -336,47 +342,44 @@ static void __init _mxc_timer_init(int irq,
>   		tctl_val = MX1_2_TCTL_FRR | MX1_2_TCTL_CLK_PCLK1 | MXC_TCTL_TEN;
>   	}
>
> -	__raw_writel(tctl_val, timer_base + MXC_TCTL);
> +	__raw_writel(tctl_val, imxtm.base + MXC_TCTL);
>
>   	/* init and register the timer to the framework */
> -	mxc_clocksource_init(clk_per);
> -	mxc_clockevent_init(clk_per);
> +	mxc_clocksource_init();
> +	mxc_clockevent_init();
>
>   	/* Make irqs happen */
> -	setup_irq(irq, &mxc_timer_irq);
> +	setup_irq(imxtm.irq, &mxc_timer_irq);
>   }
>
>   void __init mxc_timer_init(unsigned long pbase, int irq)
>   {
> -	struct clk *clk_per = clk_get_sys("imx-gpt.0", "per");
> -	struct clk *clk_ipg = clk_get_sys("imx-gpt.0", "ipg");
> +	imxtm.clk_per = clk_get_sys("imx-gpt.0", "per");
> +	imxtm.clk_ipg = clk_get_sys("imx-gpt.0", "ipg");
>
> -	timer_base = ioremap(pbase, SZ_4K);
> -	BUG_ON(!timer_base);
> +	imxtm.base = ioremap(pbase, SZ_4K);
> +	BUG_ON(!imxtm.base);
>
> -	_mxc_timer_init(irq, clk_per, clk_ipg);
> +	_mxc_timer_init();
>   }
>
>   static void __init mxc_timer_init_dt(struct device_node *np)
>   {
> -	struct clk *clk_per, *clk_ipg;
> -	int irq;
> -
> -	if (timer_base)
> +	if (imxtm.base)
>   		return;
>
> -	timer_base = of_iomap(np, 0);
> -	WARN_ON(!timer_base);
> -	irq = irq_of_parse_and_map(np, 0);
> +	imxtm.base = of_iomap(np, 0);
> +	WARN_ON(!imxtm.base);
> +	imxtm.irq = irq_of_parse_and_map(np, 0);
>
> -	clk_ipg = of_clk_get_by_name(np, "ipg");
> +	imxtm.clk_ipg = of_clk_get_by_name(np, "ipg");
>
>   	/* Try osc_per first, and fall back to per otherwise */
> -	clk_per = of_clk_get_by_name(np, "osc_per");
> -	if (IS_ERR(clk_per))
> -		clk_per = of_clk_get_by_name(np, "per");
> +	imxtm.clk_per = of_clk_get_by_name(np, "osc_per");
> +	if (IS_ERR(imxtm.clk_per))
> +		imxtm.clk_per = of_clk_get_by_name(np, "per");
>
> -	_mxc_timer_init(irq, clk_per, clk_ipg);
> +	_mxc_timer_init();
>   }
>   CLOCKSOURCE_OF_DECLARE(mx1_timer, "fsl,imx1-gpt", mxc_timer_init_dt);
>   CLOCKSOURCE_OF_DECLARE(mx25_timer, "fsl,imx25-gpt", mxc_timer_init_dt);
>
Shawn Guo May 19, 2015, 7:57 a.m. UTC | #3
On Fri, May 15, 2015 at 04:36:18PM +0000, Shenwei Wang wrote:
> 
> 
> > -----Original Message-----
> > From: shawnguo@kernel.org [mailto:shawnguo@kernel.org]
> > Sent: 2015年5月15日 3:12
> > To: linux-arm-kernel@lists.infradead.org
> > Cc: kernel@pengutronix.de; Daniel Lezcano; Wang Shenwei-B38339; Shawn Guo
> > Subject: [PATCH 1/9] ARM: imx: move timer resources into a structure
> > 
> > From: Shawn Guo <shawn.guo@linaro.org>
> > 
> > Instead of passing around as argument, let's move timer resources like irq and
> > clocks together with base address into a data structure, and reference the
> > resources from the struct variable directly to simplify the function call interface.
> > 
> > Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
> > ---
> >  arch/arm/mach-imx/time.c | 119 ++++++++++++++++++++++++-----------------------
> >  1 file changed, 61 insertions(+), 58 deletions(-)
> > 
> > diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c index
> > ab5ee1c445f3..8ad7cb2a7f08 100644
> > --- a/arch/arm/mach-imx/time.c
> > +++ b/arch/arm/mach-imx/time.c
> > @@ -84,27 +84,34 @@
> >  static struct clock_event_device clockevent_mxc;  static enum
> > clock_event_mode clockevent_mode = CLOCK_EVT_MODE_UNUSED;
> > 
> > -static void __iomem *timer_base;
> > +struct imx_timer {
> > +	void __iomem *base;
> > +	int irq;
> > +	struct clk *clk_per;
> > +	struct clk *clk_ipg;
> > +};
> > +
> > +static struct imx_timer imxtm;
> > 
> Since the changes still used a global variable to hold the information, it doesn't take advantage of what the clocksource framework provides, and make the driver unable to support multi-instances.
> 

The goal of the series is to clean up cpu_is_xxx usages, and move the
driver into drivers/clocksource.  The multi-instances can be supported
later when we see a need for that.

Shawn
Shawn Guo May 19, 2015, 7:58 a.m. UTC | #4
On Mon, May 18, 2015 at 12:43:10PM +0200, Daniel Lezcano wrote:
> On 05/15/2015 10:11 AM, shawnguo@kernel.org wrote:
> >From: Shawn Guo <shawn.guo@linaro.org>
> >
> >Instead of passing around as argument, let's move timer resources like
> >irq and clocks together with base address into a data structure, and
> >reference the resources from the struct variable directly to simplify
> >the function call interface.
> >
> >Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
> 
> I would be nice to embed the clockevent_mxc variable in the
> structure as well and use container_of to retrieve the structure
> from the struct clockevent.
> 
> The struct clockevent can be the unified parameter across the
> different functions.
> 
> Refer to drivers/clocksource/rockchip_timer.c as an example.

Okay, thanks for the example.  I will add a patch to do that.

Shawn
diff mbox

Patch

diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c
index ab5ee1c445f3..8ad7cb2a7f08 100644
--- a/arch/arm/mach-imx/time.c
+++ b/arch/arm/mach-imx/time.c
@@ -84,27 +84,34 @@ 
 static struct clock_event_device clockevent_mxc;
 static enum clock_event_mode clockevent_mode = CLOCK_EVT_MODE_UNUSED;
 
-static void __iomem *timer_base;
+struct imx_timer {
+	void __iomem *base;
+	int irq;
+	struct clk *clk_per;
+	struct clk *clk_ipg;
+};
+
+static struct imx_timer imxtm;
 
 static inline void gpt_irq_disable(void)
 {
 	unsigned int tmp;
 
 	if (timer_is_v2())
-		__raw_writel(0, timer_base + V2_IR);
+		__raw_writel(0, imxtm.base + V2_IR);
 	else {
-		tmp = __raw_readl(timer_base + MXC_TCTL);
-		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, timer_base + MXC_TCTL);
+		tmp = __raw_readl(imxtm.base + MXC_TCTL);
+		__raw_writel(tmp & ~MX1_2_TCTL_IRQEN, imxtm.base + MXC_TCTL);
 	}
 }
 
 static inline void gpt_irq_enable(void)
 {
 	if (timer_is_v2())
-		__raw_writel(1<<0, timer_base + V2_IR);
+		__raw_writel(1<<0, imxtm.base + V2_IR);
 	else {
-		__raw_writel(__raw_readl(timer_base + MXC_TCTL) | MX1_2_TCTL_IRQEN,
-			timer_base + MXC_TCTL);
+		__raw_writel(__raw_readl(imxtm.base + MXC_TCTL) | MX1_2_TCTL_IRQEN,
+			imxtm.base + MXC_TCTL);
 	}
 }
 
@@ -112,12 +119,12 @@  static void gpt_irq_acknowledge(void)
 {
 	if (timer_is_v1()) {
 		if (cpu_is_mx1())
-			__raw_writel(0, timer_base + MX1_2_TSTAT);
+			__raw_writel(0, imxtm.base + MX1_2_TSTAT);
 		else
 			__raw_writel(MX2_TSTAT_CAPT | MX2_TSTAT_COMP,
-				timer_base + MX1_2_TSTAT);
+				imxtm.base + MX1_2_TSTAT);
 	} else if (timer_is_v2())
-		__raw_writel(V2_TSTAT_OF1, timer_base + V2_TSTAT);
+		__raw_writel(V2_TSTAT_OF1, imxtm.base + V2_TSTAT);
 }
 
 static void __iomem *sched_clock_reg;
@@ -134,10 +141,10 @@  static unsigned long imx_read_current_timer(void)
 	return __raw_readl(sched_clock_reg);
 }
 
-static int __init mxc_clocksource_init(struct clk *timer_clk)
+static int __init mxc_clocksource_init(void)
 {
-	unsigned int c = clk_get_rate(timer_clk);
-	void __iomem *reg = timer_base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);
+	unsigned int c = clk_get_rate(imxtm.clk_per);
+	void __iomem *reg = imxtm.base + (timer_is_v2() ? V2_TCN : MX1_2_TCN);
 
 	imx_delay_timer.read_current_timer = &imx_read_current_timer;
 	imx_delay_timer.freq = c;
@@ -157,11 +164,11 @@  static int mx1_2_set_next_event(unsigned long evt,
 {
 	unsigned long tcmp;
 
-	tcmp = __raw_readl(timer_base + MX1_2_TCN) + evt;
+	tcmp = __raw_readl(imxtm.base + MX1_2_TCN) + evt;
 
-	__raw_writel(tcmp, timer_base + MX1_2_TCMP);
+	__raw_writel(tcmp, imxtm.base + MX1_2_TCMP);
 
-	return (int)(tcmp - __raw_readl(timer_base + MX1_2_TCN)) < 0 ?
+	return (int)(tcmp - __raw_readl(imxtm.base + MX1_2_TCN)) < 0 ?
 				-ETIME : 0;
 }
 
@@ -170,12 +177,12 @@  static int v2_set_next_event(unsigned long evt,
 {
 	unsigned long tcmp;
 
-	tcmp = __raw_readl(timer_base + V2_TCN) + evt;
+	tcmp = __raw_readl(imxtm.base + V2_TCN) + evt;
 
-	__raw_writel(tcmp, timer_base + V2_TCMP);
+	__raw_writel(tcmp, imxtm.base + V2_TCMP);
 
 	return evt < 0x7fffffff &&
-		(int)(tcmp - __raw_readl(timer_base + V2_TCN)) < 0 ?
+		(int)(tcmp - __raw_readl(imxtm.base + V2_TCN)) < 0 ?
 				-ETIME : 0;
 }
 
@@ -206,11 +213,11 @@  static void mxc_set_mode(enum clock_event_mode mode,
 	if (mode != clockevent_mode) {
 		/* Set event time into far-far future */
 		if (timer_is_v2())
-			__raw_writel(__raw_readl(timer_base + V2_TCN) - 3,
-					timer_base + V2_TCMP);
+			__raw_writel(__raw_readl(imxtm.base + V2_TCN) - 3,
+					imxtm.base + V2_TCMP);
 		else
-			__raw_writel(__raw_readl(timer_base + MX1_2_TCN) - 3,
-					timer_base + MX1_2_TCMP);
+			__raw_writel(__raw_readl(imxtm.base + MX1_2_TCN) - 3,
+					imxtm.base + MX1_2_TCMP);
 
 		/* Clear pending interrupt */
 		gpt_irq_acknowledge();
@@ -259,9 +266,9 @@  static irqreturn_t mxc_timer_interrupt(int irq, void *dev_id)
 	uint32_t tstat;
 
 	if (timer_is_v2())
-		tstat = __raw_readl(timer_base + V2_TSTAT);
+		tstat = __raw_readl(imxtm.base + V2_TSTAT);
 	else
-		tstat = __raw_readl(timer_base + MX1_2_TSTAT);
+		tstat = __raw_readl(imxtm.base + MX1_2_TSTAT);
 
 	gpt_irq_acknowledge();
 
@@ -284,49 +291,48 @@  static struct clock_event_device clockevent_mxc = {
 	.rating		= 200,
 };
 
-static int __init mxc_clockevent_init(struct clk *timer_clk)
+static int __init mxc_clockevent_init(void)
 {
 	if (timer_is_v2())
 		clockevent_mxc.set_next_event = v2_set_next_event;
 
 	clockevent_mxc.cpumask = cpumask_of(0);
 	clockevents_config_and_register(&clockevent_mxc,
-					clk_get_rate(timer_clk),
+					clk_get_rate(imxtm.clk_per),
 					0xff, 0xfffffffe);
 
 	return 0;
 }
 
-static void __init _mxc_timer_init(int irq,
-				   struct clk *clk_per, struct clk *clk_ipg)
+static void __init _mxc_timer_init(void)
 {
 	uint32_t tctl_val;
 
-	if (IS_ERR(clk_per)) {
+	if (IS_ERR(imxtm.clk_per)) {
 		pr_err("i.MX timer: unable to get clk\n");
 		return;
 	}
 
-	if (!IS_ERR(clk_ipg))
-		clk_prepare_enable(clk_ipg);
+	if (!IS_ERR(imxtm.clk_ipg))
+		clk_prepare_enable(imxtm.clk_ipg);
 
-	clk_prepare_enable(clk_per);
+	clk_prepare_enable(imxtm.clk_per);
 
 	/*
 	 * Initialise to a known state (all timers off, and timing reset)
 	 */
 
-	__raw_writel(0, timer_base + MXC_TCTL);
-	__raw_writel(0, timer_base + MXC_TPRER); /* see datasheet note */
+	__raw_writel(0, imxtm.base + MXC_TCTL);
+	__raw_writel(0, imxtm.base + MXC_TPRER); /* see datasheet note */
 
 	if (timer_is_v2()) {
 		tctl_val = V2_TCTL_FRR | V2_TCTL_WAITEN | MXC_TCTL_TEN;
-		if (clk_get_rate(clk_per) == V2_TIMER_RATE_OSC_DIV8) {
+		if (clk_get_rate(imxtm.clk_per) == V2_TIMER_RATE_OSC_DIV8) {
 			tctl_val |= V2_TCTL_CLK_OSC_DIV8;
 			if (cpu_is_imx6dl() || cpu_is_imx6sx()) {
 				/* 24 / 8 = 3 MHz */
 				__raw_writel(7 << V2_TPRER_PRE24M,
-					timer_base + MXC_TPRER);
+					imxtm.base + MXC_TPRER);
 				tctl_val |= V2_TCTL_24MEN;
 			}
 		} else {
@@ -336,47 +342,44 @@  static void __init _mxc_timer_init(int irq,
 		tctl_val = MX1_2_TCTL_FRR | MX1_2_TCTL_CLK_PCLK1 | MXC_TCTL_TEN;
 	}
 
-	__raw_writel(tctl_val, timer_base + MXC_TCTL);
+	__raw_writel(tctl_val, imxtm.base + MXC_TCTL);
 
 	/* init and register the timer to the framework */
-	mxc_clocksource_init(clk_per);
-	mxc_clockevent_init(clk_per);
+	mxc_clocksource_init();
+	mxc_clockevent_init();
 
 	/* Make irqs happen */
-	setup_irq(irq, &mxc_timer_irq);
+	setup_irq(imxtm.irq, &mxc_timer_irq);
 }
 
 void __init mxc_timer_init(unsigned long pbase, int irq)
 {
-	struct clk *clk_per = clk_get_sys("imx-gpt.0", "per");
-	struct clk *clk_ipg = clk_get_sys("imx-gpt.0", "ipg");
+	imxtm.clk_per = clk_get_sys("imx-gpt.0", "per");
+	imxtm.clk_ipg = clk_get_sys("imx-gpt.0", "ipg");
 
-	timer_base = ioremap(pbase, SZ_4K);
-	BUG_ON(!timer_base);
+	imxtm.base = ioremap(pbase, SZ_4K);
+	BUG_ON(!imxtm.base);
 
-	_mxc_timer_init(irq, clk_per, clk_ipg);
+	_mxc_timer_init();
 }
 
 static void __init mxc_timer_init_dt(struct device_node *np)
 {
-	struct clk *clk_per, *clk_ipg;
-	int irq;
-
-	if (timer_base)
+	if (imxtm.base)
 		return;
 
-	timer_base = of_iomap(np, 0);
-	WARN_ON(!timer_base);
-	irq = irq_of_parse_and_map(np, 0);
+	imxtm.base = of_iomap(np, 0);
+	WARN_ON(!imxtm.base);
+	imxtm.irq = irq_of_parse_and_map(np, 0);
 
-	clk_ipg = of_clk_get_by_name(np, "ipg");
+	imxtm.clk_ipg = of_clk_get_by_name(np, "ipg");
 
 	/* Try osc_per first, and fall back to per otherwise */
-	clk_per = of_clk_get_by_name(np, "osc_per");
-	if (IS_ERR(clk_per))
-		clk_per = of_clk_get_by_name(np, "per");
+	imxtm.clk_per = of_clk_get_by_name(np, "osc_per");
+	if (IS_ERR(imxtm.clk_per))
+		imxtm.clk_per = of_clk_get_by_name(np, "per");
 
-	_mxc_timer_init(irq, clk_per, clk_ipg);
+	_mxc_timer_init();
 }
 CLOCKSOURCE_OF_DECLARE(mx1_timer, "fsl,imx1-gpt", mxc_timer_init_dt);
 CLOCKSOURCE_OF_DECLARE(mx25_timer, "fsl,imx25-gpt", mxc_timer_init_dt);