regulator: core: Add option to prevent disabling unused regulators

  This may be useful for debugging and develompent purposes, when there are
drivers that depend on regulators to be enabled but do not request them.

It is inspired from the clk_ignore_unused and pd_ignore_unused parameters,
that are used to keep firmware-enabled clocks and power domains on even if
these are not used by drivers.

The parameter is not expected to be used in normal cases and should not be
needed on a platform with proper driver support.

Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
---

 Documentation/admin-guide/kernel-parameters.txt |  7 +++++++
 drivers/regulator/core.c                        | 17 +++++++++++++++++
 2 files changed, 24 insertions(+)
  

Hi,

On 10/28/23 03:24, Javier Martinez Canillas wrote:
> This may be useful for debugging and develompent purposes, when there are
> drivers that depend on regulators to be enabled but do not request them.
> 
> It is inspired from the clk_ignore_unused and pd_ignore_unused parameters,
> that are used to keep firmware-enabled clocks and power domains on even if
> these are not used by drivers.
> 
> The parameter is not expected to be used in normal cases and should not be
> needed on a platform with proper driver support.
> 
> Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>
> ---
> 
>  Documentation/admin-guide/kernel-parameters.txt |  7 +++++++
>  drivers/regulator/core.c                        | 17 +++++++++++++++++
>  2 files changed, 24 insertions(+)
> 
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index 0a1731a0f0ef..91b58d767c2c 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -5461,6 +5461,13 @@
>  			print every Nth verbose statement, where N is the value
>  			specified.
>  
> +	regulator_ignore_unused
> +			[REGULATOR]
> +			Prevents the regulator framework to disable regulators

			                                 from disabling regulators

> +			that are unused due not driver claiming them. This may

			                    no driver claiming them.

> +			be useful for debug and development, but should not be
> +			needed on a platform with proper driver support.
> +
>  	relax_domain_level=
>  			[KNL, SMP] Set scheduler's default relax_domain_level.
>  			See Documentation/admin-guide/cgroup-v1/cpusets.rst.
> diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
> index 3137e40fcd3e..220034ff0273 100644
> --- a/drivers/regulator/core.c
> +++ b/drivers/regulator/core.c
> @@ -6234,6 +6234,14 @@ static int regulator_late_cleanup(struct device *dev, void *data)
>  	return 0;
>  }
>  
> +static bool regulator_ignore_unused;
> +static int __init regulator_ignore_unused_setup(char *__unused)
> +{
> +	regulator_ignore_unused = true;
> +	return 1;
> +}
> +__setup("regulator_ignore_unused", regulator_ignore_unused_setup);
> +
>  static void regulator_init_complete_work_function(struct work_struct *work)
>  {
>  	/*
> @@ -6246,6 +6254,15 @@ static void regulator_init_complete_work_function(struct work_struct *work)
>  	class_for_each_device(&regulator_class, NULL, NULL,
>  			      regulator_register_resolve_supply);
>  
> +	/*
> +	 * For debugging purposes, it may be useful to prevent unused
> +	 * regulators to be disabled.

	              from being disabled.

> +	 */
> +	if (regulator_ignore_unused) {
> +		pr_warn("regulator: Not disabling unused regulators\n");
> +		return;
> +	}
> +
>  	/* If we have a full configuration then disable any regulators
>  	 * we have permission to change the status for and which are
>  	 * not in use or always_on.  This is effectively the default

HTH.
  

On Sat, Oct 28, 2023 at 12:24:03PM +0200, Javier Martinez Canillas wrote:
> This may be useful for debugging and develompent purposes, when there are
> drivers that depend on regulators to be enabled but do not request them.
> 
> It is inspired from the clk_ignore_unused and pd_ignore_unused parameters,
> that are used to keep firmware-enabled clocks and power domains on even if
> these are not used by drivers.
> 
> The parameter is not expected to be used in normal cases and should not be
> needed on a platform with proper driver support.
> 
> Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>

Thanks for posting this! I've had a need for this. With Randy's feedback
addressed:

Reviewed-by: Brian Masney <bmasney@redhat.com>
  

On Sat, Oct 28, 2023 at 12:24:03PM +0200, Javier Martinez Canillas wrote:
> This may be useful for debugging and develompent purposes, when there are
> drivers that depend on regulators to be enabled but do not request them.
> 
> It is inspired from the clk_ignore_unused and pd_ignore_unused parameters,
> that are used to keep firmware-enabled clocks and power domains on even if
> these are not used by drivers.
> 
> The parameter is not expected to be used in normal cases and should not be
> needed on a platform with proper driver support.
> 
> Signed-off-by: Javier Martinez Canillas <javierm@redhat.com>

Acked-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>

Thanks for looking into this. This patch is very handy to debug display related
issues on platforms without serial console.

On the other note, I'm wondering if we could use sync_state() for handling the
regulator_init_complete() work. This would ensure that the regulators are only
disabled when all the consumers are probed.

- Mani

> ---
> 
>  Documentation/admin-guide/kernel-parameters.txt |  7 +++++++
>  drivers/regulator/core.c                        | 17 +++++++++++++++++
>  2 files changed, 24 insertions(+)
> 
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index 0a1731a0f0ef..91b58d767c2c 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -5461,6 +5461,13 @@
>  			print every Nth verbose statement, where N is the value
>  			specified.
>  
> +	regulator_ignore_unused
> +			[REGULATOR]
> +			Prevents the regulator framework to disable regulators
> +			that are unused due not driver claiming them. This may
> +			be useful for debug and development, but should not be
> +			needed on a platform with proper driver support.
> +
>  	relax_domain_level=
>  			[KNL, SMP] Set scheduler's default relax_domain_level.
>  			See Documentation/admin-guide/cgroup-v1/cpusets.rst.
> diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
> index 3137e40fcd3e..220034ff0273 100644
> --- a/drivers/regulator/core.c
> +++ b/drivers/regulator/core.c
> @@ -6234,6 +6234,14 @@ static int regulator_late_cleanup(struct device *dev, void *data)
>  	return 0;
>  }
>  
> +static bool regulator_ignore_unused;
> +static int __init regulator_ignore_unused_setup(char *__unused)
> +{
> +	regulator_ignore_unused = true;
> +	return 1;
> +}
> +__setup("regulator_ignore_unused", regulator_ignore_unused_setup);
> +
>  static void regulator_init_complete_work_function(struct work_struct *work)
>  {
>  	/*
> @@ -6246,6 +6254,15 @@ static void regulator_init_complete_work_function(struct work_struct *work)
>  	class_for_each_device(&regulator_class, NULL, NULL,
>  			      regulator_register_resolve_supply);
>  
> +	/*
> +	 * For debugging purposes, it may be useful to prevent unused
> +	 * regulators to be disabled.
> +	 */
> +	if (regulator_ignore_unused) {
> +		pr_warn("regulator: Not disabling unused regulators\n");
> +		return;
> +	}
> +
>  	/* If we have a full configuration then disable any regulators
>  	 * we have permission to change the status for and which are
>  	 * not in use or always_on.  This is effectively the default
> -- 
> 2.41.0
>
  

On Wed, Nov 01, 2023 at 10:26:52AM +0530, Manivannan Sadhasivam wrote:

> On the other note, I'm wondering if we could use sync_state() for handling the
> regulator_init_complete() work. This would ensure that the regulators are only
> disabled when all the consumers are probed.

That assumes that everything defined in the DT both has a driver and has
the driver available for the currently running kernel neither of which
is a good assumption.
  

Mark Brown <broonie@kernel.org> writes:

> On Wed, Nov 01, 2023 at 10:26:52AM +0530, Manivannan Sadhasivam wrote:
>
>> On the other note, I'm wondering if we could use sync_state() for handling the
>> regulator_init_complete() work. This would ensure that the regulators are only
>> disabled when all the consumers are probed.
>
> That assumes that everything defined in the DT both has a driver and has
> the driver available for the currently running kernel neither of which
> is a good assumption.

Agreed. I believe the current logic of disabling all regulators using a
workqueue is the correct thing to do.

The only better option I think is to make user-space notify the kernel
that it won't load kernel modules anymore. But the delayed work would
sill be needed, since the kernel can't make an assumption on whether
user-space will notify of this or not.
  

On Sat, 28 Oct 2023 12:24:03 +0200, Javier Martinez Canillas wrote:
> This may be useful for debugging and develompent purposes, when there are
> drivers that depend on regulators to be enabled but do not request them.
> 
> It is inspired from the clk_ignore_unused and pd_ignore_unused parameters,
> that are used to keep firmware-enabled clocks and power domains on even if
> these are not used by drivers.
> 
> [...]

Applied to

   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator.git for-next

Thanks!

[1/1] regulator: core: Add option to prevent disabling unused regulators
      commit: c986968fe92f20f2db26fa6bce27795b2e9ebe22

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark