[Letux-kernel] Fwd: N900 off mode

[H. Nikolaus Schaller]

Hi,
it looks as if others are also trying to check power consumption of sgx.

Unfortunately we are still a little fragmented with mailing lists...

BR,
Nikolaus


> Anfang der weitergeleiteten Nachricht:
> 
> Von: Tony Lindgren <tony at atomide.com>
> Betreff: Aw: N900 off mode
> Datum: 9. Dezember 2019 um 16:25:31 MEZ
> An: Merlijn Wajer <merlijn at wizzup.org>
> Kopie: linux-omap <linux-omap at vger.kernel.org>, maemo-leste at lists.dyne.org
> 
> Hi,
> 
> * Merlijn Wajer <merlijn at wizzup.org> [191209 02:32]:
>> I'm trying to figure out how I can try get my Nokia N900 to go into off
>> mode.
> ...
> 
>> BTW, the kernel is patched with SGX support, so that may have something
>> to do with it, but it seems to actually be idle, if I am reading
>> pm_debug/count correctly. [2]
> 
> OK that's good to hear :)
> 
>> # cat idle.sh
>> #!/bin/sh
>> 
>> export UART_IDLE_MS=5000
>> idle_uarts() {
>>        # Enable autosuspend
>>        uarts=$(find /sys/bus/platform/devices/*.serial/power/ -type d)
>>        for uart in ${uarts}; do
>>                echo -n ${UART_IDLE_MS} > ${uart}/autosuspend_delay_ms
>>                echo -n enabled > ${uart}/wakeup
>>                echo -n auto > ${uart}/control
>>        done
>> 
>>        # Configure wake-up from suspend
>>        uarts=$(find /sys/class/tty/tty[SO]*/power/ -type d 2>/dev/null)
>>        for uart in ${uarts}; do
>>                echo -n enabled > ${uart}/wakeup
>>        done
>> }
>> 
>> idle_uarts
>> 
>> mount -t debugfs debugfs /sys/kernel/debug
>> echo 1 > /sys/kernel/debug/pm_debug/enable_off_mode
> 
> The above should do the right thing.
> 
>> [2] # cat /sys/kernel/debug/pm_debug/count
>> usbhost_pwrdm
>> (ON),OFF:0,RET:670,INA:0,ON:671,RET-LOGIC-OFF:0,RET-MEMBANK1-OFF:0
>> sgx_pwrdm (OFF),OFF:24,RET:0,INA:0,ON:24,RET-LOGIC-OFF:0,RET-MEMBANK1-OFF:0
> 
> Yup sgx_pwrdm above looks good.
> 
>> core_pwrdm
>> (ON),OFF:0,RET:0,INA:0,ON:1,RET-LOGIC-OFF:0,RET-MEMBANK1-OFF:0,RET-MEMBANK2-OFF:0
>> per_pwrdm (ON),OFF:0,RET:0,INA:0,ON:1,RET-LOGIC-OFF:0,RET-MEMBANK1-OFF:0
> 
> Here per_pwrdm has some device(s) blocking idle for both per and core.
> 
> Below is my old debug patch that provides more info on the blocking
> bits. It's probably some cm_idlest_per bit(s) that's blocking compared
> to the known good 00010000 and 00000042 values.
> 
> Care to try it and reply with the cm_idlest_per and cm_idlest1_core
> values you're seeing?
> 
> I've checked that n900 idles fine for me with omap2plus_defconfig
> and minimal user space for v5.4.
> 
> Regards,
> 
> Tony
> 
> 8< -----------------------
> Allows seeing the deeper idle state blockers in
> /sys/kernel/debug/pm_debug/count. For example, when
> off idle is working on beaglboard xm, this is what
> i see:
> 
> # sleep 5; cat /sys/kernel/debug/pm_debug/count
> ...
> 0006ffff 48005020 (fa005020) cm_idlest_per blocking bits: 00010000
> e7ffffbd 48004a20 (fa004a20) cm_idlest1_core blocking bits: 00000042
> 0000000d 48004a28 (fa004a28) cm_idlest3_core
> ---
> arch/arm/mach-omap2/pm-debug.c | 68 ++++++++++++++++++++++++++++++++++
> 1 file changed, 68 insertions(+)
> 
> diff --git a/arch/arm/mach-omap2/pm-debug.c b/arch/arm/mach-omap2/pm-debug.c
> --- a/arch/arm/mach-omap2/pm-debug.c
> +++ b/arch/arm/mach-omap2/pm-debug.c
> @@ -136,10 +136,78 @@ static int pwrdm_dbg_show_timer(struct powerdomain *pwrdm, void *user)
> 	return 0;
> }
> 
> +#include "iomap.h"
> +
> +struct dregs {
> +	const char	*desc;
> +	u32		phys;
> +	void __iomem	*virt;
> +	u32		mask;
> +};
> +
> +#define PER_CM_BASE	0x48005000
> +#define PER_CM_REG(name, offset, mask)				\
> +	{ name, PER_CM_BASE + offset,				\
> +	OMAP2_L4_IO_ADDRESS(PER_CM_BASE + offset), mask, }
> +
> +static struct dregs cm_per[] = {
> +	PER_CM_REG("cm_idlest_per", 0x20, 0xfff80000), /* p 513 */
> +	{ NULL, },
> +};
> +
> +#define CORE_CM_BASE	0x48004a00
> +#define CORE_CM_REG(name, offset, mask)				\
> +	{ name, CORE_CM_BASE + offset,				\
> +	OMAP2_L4_IO_ADDRESS(CORE_CM_BASE + offset), mask, }
> +
> +static struct dregs cm_core[] = {
> +	CORE_CM_REG("cm_idlest1_core", 0x20, 0x9c800109), /* p 467 */
> +	CORE_CM_REG("cm_idlest3_core", 0x28, 0xfffffffb),
> +	{ NULL, },
> +};
> +
> +void __dregs_dump(struct dregs *dregs, struct seq_file *s)
> +{
> +	for (; dregs->desc; dregs++) {
> +		u32 val, blockers;
> +
> +		val = __raw_readl(dregs->virt);
> +
> +		seq_printf(s, "%08x %08x (%p) %s",
> +			   val, dregs->phys, dregs->virt,
> +			   dregs->desc);
> +
> +		if (dregs->mask) {
> +			blockers = ~val;
> +			blockers &= ~dregs->mask;
> +
> +			if (blockers)
> +				seq_printf(s, " blocking bits: %08x",
> +					   blockers);
> +		}
> +
> +		seq_printf(s, "\n");
> +	}
> +}
> +
> +void cm_per_dump(struct seq_file *s)
> +{
> +	__dregs_dump(cm_per, s);
> +}
> +
> +void cm_core_dump(struct seq_file *s)
> +{
> +	__dregs_dump(cm_core, s);
> +}
> +
> static int pm_dbg_counters_show(struct seq_file *s, void *unused)
> {
> 	pwrdm_for_each(pwrdm_dbg_show_counter, s);
> 	clkdm_for_each(clkdm_dbg_show_counter, s);
> +	if (cpu_is_omap34xx()) {
> +		cm_per_dump(s);
> +		cm_core_dump(s);
> +	}
> 
> 	return 0;
> }
> -- 
> 2.24.0