linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Jernej Škrabec" <jernej.skrabec@gmail.com>
To: daniel.lezcano@linaro.org, tglx@linutronix.de, wens@csie.org,
	samuel@sholland.org, maxime@cerno.tech,
	Victor Hassan <victor@allwinnertech.com>
Cc: linux-kernel@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-sunxi@lists.linux.dev
Subject: Re: [RESEND] clocksource: sun4i: Fix the bug that tick_resume stucks
Date: Wed, 12 Oct 2022 23:46:41 +0200	[thread overview]
Message-ID: <14741195.tv2OnDr8pf@kista> (raw)
In-Reply-To: <20221009032507.39145-1-victor@allwinnertech.com>

Hi Victor,

Dne nedelja, 09. oktober 2022 ob 05:25:07 CEST je Victor Hassan napisal(a):
> Currently syscore_resume() will stuck on tick_resume().
> Fix this by changing  `.tick_resume` from
> sun4i_clkevt_shutdown() to a new function sun4i_tick_resume().

This is pretty involved change. Can you split it in more manageable chunks? 
Also you didn't state why it stucks and what does new approach differently so 
it doesn't stuck.

Best regards,
Jernej

> 
> Signed-off-by: Victor Hassan <victor@allwinnertech.com>
> ---
>  drivers/clocksource/timer-sun4i.c | 131 ++++++++++++++++++++++--------
>  1 file changed, 96 insertions(+), 35 deletions(-)
> 
> diff --git a/drivers/clocksource/timer-sun4i.c
> b/drivers/clocksource/timer-sun4i.c index 94dc6e42e983..574398c35a22 100644
> --- a/drivers/clocksource/timer-sun4i.c
> +++ b/drivers/clocksource/timer-sun4i.c
> @@ -38,6 +38,19 @@
> 
>  #define TIMER_SYNC_TICKS	3
> 
> +/* Registers which needs to be saved and restored before and after sleeping
> */ +static u32 sun4i_timer_regs_offset[] = {
> +	TIMER_IRQ_EN_REG,
> +	TIMER_CTL_REG(0),
> +	TIMER_INTVAL_REG(0),
> +	TIMER_CNTVAL_REG(0),
> +	TIMER_CTL_REG(1),
> +	TIMER_INTVAL_REG(1),
> +	TIMER_CNTVAL_REG(1),
> +};
> +
> +static void __iomem *sun4i_timer_sched_base __read_mostly;
> +
>  /*
>   * When we disable a timer, we need to wait at least for 2 cycles of
>   * the timer source clock. We will use for that the clocksource timer
> @@ -79,10 +92,41 @@ static void sun4i_clkevt_time_start(void __iomem *base,
> u8 timer, base + TIMER_CTL_REG(timer));
>  }
> 
> +static inline void sun4i_timer_save_regs(struct timer_of *to)
> +{
> +	void __iomem *base = timer_of_base(to);
> +	int i;
> +	u32 *regs_backup = (u32 *)to->private_data;
> +
> +	for (i = 0; i < ARRAY_SIZE(sun4i_timer_regs_offset); i++)
> +		regs_backup[i] = readl(base + 
sun4i_timer_regs_offset[i]);
> +}
> +
> +static inline void sun4i_timer_restore_regs(struct timer_of *to)
> +{
> +	void __iomem *base = timer_of_base(to);
> +	int i;
> +	u32 *regs_backup = (u32 *)to->private_data;
> +
> +	for (i = 0; i < ARRAY_SIZE(sun4i_timer_regs_offset); i++)
> +		writel(regs_backup[i], base + 
sun4i_timer_regs_offset[i]);
> +}
> +
>  static int sun4i_clkevt_shutdown(struct clock_event_device *evt)
>  {
>  	struct timer_of *to = to_timer_of(evt);
> 
> +	sun4i_timer_save_regs(to);
> +	sun4i_clkevt_time_stop(timer_of_base(to), 0);
> +
> +	return 0;
> +}
> +
> +static int sun4i_tick_resume(struct clock_event_device *evt)
> +{
> +	struct timer_of *to = to_timer_of(evt);
> +
> +	sun4i_timer_restore_regs(to);
>  	sun4i_clkevt_time_stop(timer_of_base(to), 0);
> 
>  	return 0;
> @@ -137,45 +181,54 @@ static irqreturn_t sun4i_timer_interrupt(int irq, void
> *dev_id) return IRQ_HANDLED;
>  }
> 
> -static struct timer_of to = {
> -	.flags = TIMER_OF_IRQ | TIMER_OF_CLOCK | TIMER_OF_BASE,
> -
> -	.clkevt = {
> -		.name = "sun4i_tick",
> -		.rating = 350,
> -		.features = CLOCK_EVT_FEAT_PERIODIC | 
CLOCK_EVT_FEAT_ONESHOT,
> -		.set_state_shutdown = sun4i_clkevt_shutdown,
> -		.set_state_periodic = sun4i_clkevt_set_periodic,
> -		.set_state_oneshot = sun4i_clkevt_set_oneshot,
> -		.tick_resume = sun4i_clkevt_shutdown,
> -		.set_next_event = sun4i_clkevt_next_event,
> -		.cpumask = cpu_possible_mask,
> -	},
> -
> -	.of_irq = {
> -		.handler = sun4i_timer_interrupt,
> -		.flags = IRQF_TIMER | IRQF_IRQPOLL,
> -	},
> -};
> +static void __init sun4i_clockevent_init(struct timer_of *to)
> +{
> +	to->clkevt.name = "sun4i_tick";
> +	to->clkevt.features = CLOCK_EVT_FEAT_PERIODIC | 
CLOCK_EVT_FEAT_ONESHOT;
> +	to->clkevt.set_state_shutdown = sun4i_clkevt_shutdown;
> +	to->clkevt.set_state_periodic = sun4i_clkevt_set_periodic;
> +	to->clkevt.set_state_oneshot = sun4i_clkevt_set_oneshot;
> +	to->clkevt.tick_resume = sun4i_tick_resume;
> +	to->clkevt.set_next_event = sun4i_clkevt_next_event;
> +	to->clkevt.cpumask = cpu_possible_mask;
> +	to->of_irq.flags = IRQF_TIMER | IRQF_IRQPOLL;
> +
> +	sun4i_timer_sched_base = timer_of_base(to) + TIMER_CNTVAL_REG(1);
> +}
> 
>  static u64 notrace sun4i_timer_sched_read(void)
>  {
> -	return ~readl(timer_of_base(&to) + TIMER_CNTVAL_REG(1));
> +	return (u64)~readl(sun4i_timer_sched_base);
>  }
> 
>  static int __init sun4i_timer_init(struct device_node *node)
>  {
> +	struct timer_of *to;
>  	int ret;
>  	u32 val;
> 
> -	ret = timer_of_init(node, &to);
> +	to = kzalloc(sizeof(*to), GFP_KERNEL);
> +	if (!to)
> +		return -ENOMEM;
> +
> +	to->flags = TIMER_OF_IRQ | TIMER_OF_CLOCK | TIMER_OF_BASE;
> +	to->of_irq.handler = sun4i_timer_interrupt;
> +	ret = timer_of_init(node, to);
>  	if (ret)
> -		return ret;
> +		goto err;
> 
> -	writel(~0, timer_of_base(&to) + TIMER_INTVAL_REG(1));
> +	sun4i_clockevent_init(to);
> +
> +	to->private_data = kcalloc(ARRAY_SIZE(sun4i_timer_regs_offset),
> sizeof(u32), GFP_KERNEL); +	if (!to->private_data) {
> +		ret = -ENOMEM;
> +		goto err1;
> +	}
> +
> +	writel(~0, timer_of_base(to) + TIMER_INTVAL_REG(1));
>  	writel(TIMER_CTL_ENABLE | TIMER_CTL_RELOAD |
>  	       TIMER_CTL_CLK_SRC(TIMER_CTL_CLK_SRC_OSC24M),
> -	       timer_of_base(&to) + TIMER_CTL_REG(1));
> +	       timer_of_base(to) + TIMER_CTL_REG(1));
> 
>  	/*
>  	 * sched_clock_register does not have priorities, and on sun6i and
> @@ -186,32 +239,40 @@ static int __init sun4i_timer_init(struct device_node
> *node) of_machine_is_compatible("allwinner,sun5i-a10s") ||
>  	    of_machine_is_compatible("allwinner,suniv-f1c100s"))
>  		sched_clock_register(sun4i_timer_sched_read, 32,
> -				     timer_of_rate(&to));
> +				     timer_of_rate(to));
> 
> -	ret = clocksource_mmio_init(timer_of_base(&to) + 
TIMER_CNTVAL_REG(1),
> -				    node->name, 
timer_of_rate(&to), 350, 32,
> +	ret = clocksource_mmio_init(timer_of_base(to) + 
TIMER_CNTVAL_REG(1),
> +				    node->name, 
timer_of_rate(to), 350, 32,
>  				    clocksource_mmio_readl_down);
>  	if (ret) {
>  		pr_err("Failed to register clocksource\n");
> -		return ret;
> +		goto err2;
>  	}
> 
>  	writel(TIMER_CTL_CLK_SRC(TIMER_CTL_CLK_SRC_OSC24M),
> -	       timer_of_base(&to) + TIMER_CTL_REG(0));
> +	       timer_of_base(to) + TIMER_CTL_REG(0));
> 
>  	/* Make sure timer is stopped before playing with interrupts */
> -	sun4i_clkevt_time_stop(timer_of_base(&to), 0);
> +	sun4i_clkevt_time_stop(timer_of_base(to), 0);
> 
>  	/* clear timer0 interrupt */
> -	sun4i_timer_clear_interrupt(timer_of_base(&to));
> +	sun4i_timer_clear_interrupt(timer_of_base(to));
> 
> -	clockevents_config_and_register(&to.clkevt, timer_of_rate(&to),
> +	clockevents_config_and_register(&to->clkevt, timer_of_rate(to),
>  					TIMER_SYNC_TICKS, 
0xffffffff);
> 
>  	/* Enable timer0 interrupt */
> -	val = readl(timer_of_base(&to) + TIMER_IRQ_EN_REG);
> -	writel(val | TIMER_IRQ_EN(0), timer_of_base(&to) + 
TIMER_IRQ_EN_REG);
> +	val = readl(timer_of_base(to) + TIMER_IRQ_EN_REG);
> +	writel(val | TIMER_IRQ_EN(0), timer_of_base(to) + 
TIMER_IRQ_EN_REG);
> +
> +	return ret;
> 
> +err2:
> +	kfree(to->private_data);
> +err1:
> +	timer_of_cleanup(to);
> +err:
> +	kfree(to);
>  	return ret;
>  }
>  TIMER_OF_DECLARE(sun4i, "allwinner,sun4i-a10-timer",
> --
> 2.29.0



  reply	other threads:[~2022-10-12 21:46 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-09  3:25 [RESEND] clocksource: sun4i: Fix the bug that tick_resume stucks Victor Hassan
2022-10-12 21:46 ` Jernej Škrabec [this message]
2022-10-15  9:04 ` Thomas Gleixner
2022-10-25  6:23 ` Victor Hassan
2022-10-26  8:21   ` maxime
  -- strict thread matches above, loose matches on Subject: below --
2022-09-28  8:48 Victor Hassan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=14741195.tv2OnDr8pf@kista \
    --to=jernej.skrabec@gmail.com \
    --cc=daniel.lezcano@linaro.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-sunxi@lists.linux.dev \
    --cc=maxime@cerno.tech \
    --cc=samuel@sholland.org \
    --cc=tglx@linutronix.de \
    --cc=victor@allwinnertech.com \
    --cc=wens@csie.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).