Re: [PATCH 1/4] clocksource: Add support for the Mediatek SoCs

From: Stephen Boyd
Date: Wed Apr 09 2014 - 16:59:08 EST


On 04/09, Matthias Brugger wrote:
> diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig
> index 96918e1..bb29321 100644
> --- a/drivers/clocksource/Kconfig
> +++ b/drivers/clocksource/Kconfig
> @@ -144,6 +144,10 @@ config VF_PIT_TIMER
> config SYS_SUPPORTS_SH_CMT
> bool
>
> +config MTK_TIMER
> + bool
> +
> +

Why two newlines?

> diff --git a/drivers/clocksource/mtk_timer.c b/drivers/clocksource/mtk_timer.c
> new file mode 100644
> index 0000000..bf901e3
> --- /dev/null
> +++ b/drivers/clocksource/mtk_timer.c
> @@ -0,0 +1,248 @@
> +
> +#include <linux/clk.h>
> +#include <linux/clockchips.h>
> +#include <linux/interrupt.h>
> +#include <linux/irq.h>
> +#include <linux/irqreturn.h>
> +#include <linux/sched_clock.h>

Were you planning on registering a sched_clock source?

> +#include <linux/of.h>
> +#include <linux/of_address.h>
> +#include <linux/of_irq.h>
> +
[...]
> +
> +static struct clock_event_device mtk_clockevent = {
> + .name = "mtk_tick",
> + .rating = 300,
> + .shift = 32,

This is unnecessary as it's handled by
clockevents_config_and_register().

> + .features = CLOCK_EVT_FEAT_PERIODIC | CLOCK_EVT_FEAT_ONESHOT,
> + .set_mode = mtk_clkevt_mode,
> + .set_next_event = mtk_clkevt_next_event,
> +};
> +
> +
> +static irqreturn_t mtk_timer_interrupt(int irq, void *dev_id)
> +{
> + struct clock_event_device *evt = (struct clock_event_device *)dev_id;

Unnecessary cast from void.

> +
> + /* Acknowledge timer0 irq */
> + writel(GPT_IRQ_ACK(GPT_CLK_EVT), gpt_base + GPT_IRQ_ACK_REG);
> + evt->event_handler(evt);
> +
> + return IRQ_HANDLED;
> +}
> +
[...]
> +
> +static struct irqaction mtk_timer_irq = {
> + .name = "mtk_timer0",
> + .flags = IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL,

IRQF_DISABLED is a nop. Remove it.

> + .handler = mtk_timer_interrupt,
> + .dev_id = &mtk_clockevent,
> +};
> +
> +static u32 mtk_timer_sched_read(void)
> +{
> + return readl(gpt_base + TIMER_CNT_REG(GPT_CLK_SRC));
> +}

This is unused unless you register a sched_clock source.

> +
> +static void __init mtk_timer_init(struct device_node *node)
> +{
> + unsigned long rate = 0;
> + struct clk *clk;
> + int ret, irq;
> + u32 val;
> +
> + gpt_base = of_iomap(node, 0);
> + if (!gpt_base)
> + panic("Can't map registers");
> +
> + irq = irq_of_parse_and_map(node, 0);
> + if (irq <= 0)
> + panic("Can't parse IRQ");
> +
> + clk = of_clk_get_by_name(node, "sys_clk");
> + if (IS_ERR(clk))
> + panic("Can't get timer clock");
> + clk_prepare_enable(clk);
> +
> + rate = clk_get_rate(clk);
> +
> + mtk_timer_global_reset();
> +
> + /* Configure clock source */
> + mtk_timer_reset(GPT_CLK_SRC);
> +
> + writel(TIMER_CLK_SRC(TIMER_CLK_SRC_SYS13M) | TIMER_CLK_DIV1,
> + gpt_base + TIMER_CLK_REG(GPT_CLK_SRC));
> +
> + writel(TIMER_CTRL_OP(TIMER_CTRL_OP_FREERUN) | TIMER_CTRL_ENABLE,
> + gpt_base + TIMER_CTRL_REG(GPT_CLK_SRC));
> +
> + clocksource_mmio_init(gpt_base + TIMER_CNT_REG(GPT_CLK_SRC), node->name,
> + rate, 300, 32, clocksource_mmio_readl_up);
> +
> + ticks_per_jiffy = DIV_ROUND_UP(rate, HZ);
> +
> + /* Configure clock event */
> + mtk_timer_reset(GPT_CLK_EVT);
> +
> + writel(TIMER_CLK_SRC(TIMER_CLK_SRC_SYS13M) | TIMER_CLK_DIV1,
> + gpt_base + TIMER_CLK_REG(GPT_CLK_EVT));
> + writel(0, gpt_base + TIMER_CMP_REG(GPT_CLK_EVT));
> +
> + writel(TIMER_CTRL_OP(TIMER_CTRL_OP_REPEAT) | TIMER_CTRL_ENABLE,
> + gpt_base + TIMER_CTRL_REG(GPT_CLK_EVT));
> +
> + ret = setup_irq(irq, &mtk_timer_irq);

Most clocksource drivers use request_irq() nowadays. Can you use
that?

> + if (ret)
> + pr_warn("failed to setup irq %d\n", irq);
> +
> + /* Enable timer0 interrupt */
> + val = readl(gpt_base + GPT_IRQ_EN_REG);
> + writel(val | GPT_IRQ_ENABLE(GPT_CLK_EVT), gpt_base + GPT_IRQ_EN_REG);
> +
> + mtk_clockevent.cpumask = cpumask_of(0);

Is it possible for this timer to be used on SMP hardware? If so,
this should probably be cpu_all_mask. Please assign the .irq
member here as well.

> +
> + clockevents_config_and_register(&mtk_clockevent, rate, 0x3,
> + 0xffffffff);
> +}
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/