mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs
@ 2024-11-20 10:37 Jacky Bai
  2025-03-05  6:31 ` Yuanjie Yang
  0 siblings, 1 reply; 4+ messages in thread
From: Jacky Bai @ 2024-11-20 10:37 UTC (permalink / raw)
  To: lpieralisi, sudeep.holla, rafael, daniel.lezcano, james.morse, d-gole
  Cc: linux-pm, linux-arm-kernel, linux-kernel, imx, khilman

With 'nosmp' or 'maxcpus=0' boot command line parameters,
the 'cpu_present_mask' may not be the same as 'cpu_possible_mask'.

In current psci cpuidle driver init, for_each_possible_cpu()
is used to init the cpuidle for each possible CPU. but in
drivers/base/cpu.c ->cpu_dev_register_generic(),
for_each_present_cpu() is used to register cpu device for
present CPUs.

When boot system with 'nosmp' or 'maxcpus=0', the cpuidle driver
init failed due to no valid CPU device sysfs node for non-boot CPUs.

[ 0.182993] Failed to register cpuidle device for cpu1

Use for_each_present_cpu() to register cpuidle only for present
CPUs.

Fixes: b0c69e1214bc ("drivers: base: Use present CPUs in GENERIC_CPU_DEVICES")
Signed-off-by: Jacky Bai <ping.bai@nxp.com>
Reviewed-by: Dhruva Gole <d-gole@ti.com>
---
 drivers/cpuidle/cpuidle-psci.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/cpuidle/cpuidle-psci.c b/drivers/cpuidle/cpuidle-psci.c
index 2562dc001fc1..00117e9b33e8 100644
--- a/drivers/cpuidle/cpuidle-psci.c
+++ b/drivers/cpuidle/cpuidle-psci.c
@@ -410,7 +410,7 @@ static int psci_cpuidle_probe(struct platform_device *pdev)
 	struct cpuidle_driver *drv;
 	struct cpuidle_device *dev;
 
-	for_each_possible_cpu(cpu) {
+	for_each_present_cpu(cpu) {
 		ret = psci_idle_init_cpu(&pdev->dev, cpu);
 		if (ret)
 			goto out_fail;
-- 
2.34.1


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs
  2024-11-20 10:37 [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs Jacky Bai
@ 2025-03-05  6:31 ` Yuanjie Yang
  2025-03-05 10:25   ` Sudeep Holla
  0 siblings, 1 reply; 4+ messages in thread
From: Yuanjie Yang @ 2025-03-05  6:31 UTC (permalink / raw)
  To: Jacky Bai, lpieralisi, sudeep.holla, rafael, daniel.lezcano,
	james.morse, d-gole
  Cc: linux-pm, linux-arm-kernel, linux-kernel, imx, khilman,
	quic_tingweiz, quic_yuanjiey

On Wed, Nov 20, 2024 at 06:37:49PM +0800, Jacky Bai wrote:
> With 'nosmp' or 'maxcpus=0' boot command line parameters,
> the 'cpu_present_mask' may not be the same as 'cpu_possible_mask'.
> 
> In current psci cpuidle driver init, for_each_possible_cpu()
> is used to init the cpuidle for each possible CPU. but in
> drivers/base/cpu.c ->cpu_dev_register_generic(),
> for_each_present_cpu() is used to register cpu device for
> present CPUs.
> 
> When boot system with 'nosmp' or 'maxcpus=0', the cpuidle driver
> init failed due to no valid CPU device sysfs node for non-boot CPUs.
> 
> [ 0.182993] Failed to register cpuidle device for cpu1
> 
> Use for_each_present_cpu() to register cpuidle only for present
> CPUs.
> 
> Fixes: b0c69e1214bc ("drivers: base: Use present CPUs in GENERIC_CPU_DEVICES")
> Signed-off-by: Jacky Bai <ping.bai@nxp.com>
> Reviewed-by: Dhruva Gole <d-gole@ti.com>
> ---
>  drivers/cpuidle/cpuidle-psci.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/cpuidle/cpuidle-psci.c b/drivers/cpuidle/cpuidle-psci.c
> index 2562dc001fc1..00117e9b33e8 100644
> --- a/drivers/cpuidle/cpuidle-psci.c
> +++ b/drivers/cpuidle/cpuidle-psci.c
> @@ -410,7 +410,7 @@ static int psci_cpuidle_probe(struct platform_device *pdev)
>  	struct cpuidle_driver *drv;
>  	struct cpuidle_device *dev;
>  
> -	for_each_possible_cpu(cpu) {
> +	for_each_present_cpu(cpu) {
>  		ret = psci_idle_init_cpu(&pdev->dev, cpu);
>  		if (ret)
>  			goto out_fail;
> -- 
> 2.34.1
> 

Tested-by: Yuanjie Yang <quic_yuanjiey@quicinc.com>

I meet the same problem, and I test your patch, it is good to me.

Thanks,
Yuanjie

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs
  2025-03-05  6:31 ` Yuanjie Yang
@ 2025-03-05 10:25   ` Sudeep Holla
  2025-03-05 10:29     ` Jacky Bai
  0 siblings, 1 reply; 4+ messages in thread
From: Sudeep Holla @ 2025-03-05 10:25 UTC (permalink / raw)
  To: Yuanjie Yang, Jacky Bai
  Cc: lpieralisi, rafael, daniel.lezcano, james.morse, d-gole,
	linux-pm, linux-arm-kernel, linux-kernel, imx, khilman,
	quic_tingweiz

Hi Jacky,

I wasn't Cc-ed on the original patch.

I was searching to provide feedback but couldn't find one.

On Wed, Mar 05, 2025 at 02:31:57PM +0800, Yuanjie Yang wrote:
> On Wed, Nov 20, 2024 at 06:37:49PM +0800, Jacky Bai wrote:
> > With 'nosmp' or 'maxcpus=0' boot command line parameters,
> > the 'cpu_present_mask' may not be the same as 'cpu_possible_mask'.
> > 
> > In current psci cpuidle driver init, for_each_possible_cpu()
> > is used to init the cpuidle for each possible CPU. but in
> > drivers/base/cpu.c ->cpu_dev_register_generic(),
> > for_each_present_cpu() is used to register cpu device for
> > present CPUs.
> > 
> > When boot system with 'nosmp' or 'maxcpus=0', the cpuidle driver
> > init failed due to no valid CPU device sysfs node for non-boot CPUs.
> > 
> > [ 0.182993] Failed to register cpuidle device for cpu1
> > 
> > Use for_each_present_cpu() to register cpuidle only for present
> > CPUs.
> > 

Can you reword it something like
"

for_each_possible_cpu() is currently used to initialize cpuidle in the
PSCI cpuidle driver.

However, in cpu_dev_register_generic(), for_each_present_cpu() is used
to register CPU devices which means the CPU devices are only registered
for present CPUs and not all possible CPUs.

With nosmp or maxcpus=0, only the boot CPU is present, leading to the
failure:

  |  Failed to register cpuidle device for cpu1

Change for_each_possible_cpu() to for_each_present_cpu() in the PSCI
cpuidle driver to ensure it only registers cpuidle devices for CPUs
that are actually present.

"

With that, you can add:

Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>

Please cc Ulf Hansson <ulf.hansson@linaro.org> as well. Either him or
Rafael can then pick up the patch.

-- 
Regards,
Sudeep

^ permalink raw reply	[flat|nested] 4+ messages in thread

* RE: [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs
  2025-03-05 10:25   ` Sudeep Holla
@ 2025-03-05 10:29     ` Jacky Bai
  0 siblings, 0 replies; 4+ messages in thread
From: Jacky Bai @ 2025-03-05 10:29 UTC (permalink / raw)
  To: Sudeep Holla, Yuanjie Yang
  Cc: lpieralisi, rafael, daniel.lezcano, james.morse, d-gole,
	linux-pm, linux-arm-kernel, linux-kernel, imx, khilman,
	quic_tingweiz

> Subject: Re: [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs
> 
> Hi Jacky,
> 
> I wasn't Cc-ed on the original patch.
> 
> I was searching to provide feedback but couldn't find one.
> 
> On Wed, Mar 05, 2025 at 02:31:57PM +0800, Yuanjie Yang wrote:
> > On Wed, Nov 20, 2024 at 06:37:49PM +0800, Jacky Bai wrote:
> > > With 'nosmp' or 'maxcpus=0' boot command line parameters, the
> > > 'cpu_present_mask' may not be the same as 'cpu_possible_mask'.
> > >
> > > In current psci cpuidle driver init, for_each_possible_cpu() is used
> > > to init the cpuidle for each possible CPU. but in drivers/base/cpu.c
> > > ->cpu_dev_register_generic(),
> > > for_each_present_cpu() is used to register cpu device for present
> > > CPUs.
> > >
> > > When boot system with 'nosmp' or 'maxcpus=0', the cpuidle driver
> > > init failed due to no valid CPU device sysfs node for non-boot CPUs.
> > >
> > > [ 0.182993] Failed to register cpuidle device for cpu1
> > >
> > > Use for_each_present_cpu() to register cpuidle only for present
> > > CPUs.
> > >
> 
> Can you reword it something like
> "
> 
> for_each_possible_cpu() is currently used to initialize cpuidle in the PSCI
> cpuidle driver.
> 
> However, in cpu_dev_register_generic(), for_each_present_cpu() is used to
> register CPU devices which means the CPU devices are only registered for
> present CPUs and not all possible CPUs.
> 
> With nosmp or maxcpus=0, only the boot CPU is present, leading to the
> failure:
> 
>   |  Failed to register cpuidle device for cpu1
> 
> Change for_each_possible_cpu() to for_each_present_cpu() in the PSCI
> cpuidle driver to ensure it only registers cpuidle devices for CPUs that are
> actually present.
> 
> "
> 
> With that, you can add:
> 
> Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
> 
> Please cc Ulf Hansson <ulf.hansson@linaro.org> as well. Either him or Rafael
> can then pick up the patch.
> 

Sorry, I should Cc you in the list. Thank you for your suggestion, I will resolve it in v3.

BR

> --
> Regards,
> Sudeep

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2025-03-05 10:29 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-11-20 10:37 [PATCH v2] cpuidle: psci: Init cpuidle only for present CPUs Jacky Bai
2025-03-05  6:31 ` Yuanjie Yang
2025-03-05 10:25   ` Sudeep Holla
2025-03-05 10:29     ` Jacky Bai

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®