diff mbox series

[v6,12/16] PCI/pwrctl: add PCI power control core code

Message ID 20240325131624.26023-13-brgl@bgdev.pl (mailing list archive)
State Not Applicable
Headers show
Series power: sequencing: implement the subsystem and add first users | expand

Checks

Context Check Description
netdev/series_format fail Series longer than 15 patches (and no cover letter)
netdev/tree_selection success Guessed tree name to be net-next
netdev/ynl success Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present success Fixes tag not required for -next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 950 this patch: 950
netdev/build_tools success Errors and warnings before: 0 this patch: 0
netdev/cc_maintainers success CCed 3 of 3 maintainers
netdev/build_clang success Errors and warnings before: 956 this patch: 956
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success No Fixes tag
netdev/build_allmodconfig_warn success Errors and warnings before: 962 this patch: 962
netdev/checkpatch warning WARNING: added, moved or deleted file(s), does MAINTAINERS need updating?
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc fail Errors and warnings before: 0 this patch: 4
netdev/source_inline success Was 0 now: 0

Commit Message

Bartosz Golaszewski March 25, 2024, 1:16 p.m. UTC
From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Some PCI devices must be powered-on before they can be detected on the
bus. Introduce a simple framework reusing the existing PCI OF
infrastructure.

The way this works is: a DT node representing a PCI device connected to
the port can be matched against its power control platform driver. If
the match succeeds, the driver is responsible for powering-up the device
and calling pcie_pwrctl_device_set_ready() which will trigger a PCI bus
rescan as well as subscribe to PCI bus notifications.

When the device is detected and created, we'll make it consume the same
DT node that the platform device did. When the device is bound, we'll
create a device link between it and the parent power control device.

Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
---
 drivers/pci/Kconfig         |   1 +
 drivers/pci/Makefile        |   1 +
 drivers/pci/pwrctl/Kconfig  |   8 +++
 drivers/pci/pwrctl/Makefile |   4 ++
 drivers/pci/pwrctl/core.c   | 136 ++++++++++++++++++++++++++++++++++++
 include/linux/pci-pwrctl.h  |  51 ++++++++++++++
 6 files changed, 201 insertions(+)
 create mode 100644 drivers/pci/pwrctl/Kconfig
 create mode 100644 drivers/pci/pwrctl/Makefile
 create mode 100644 drivers/pci/pwrctl/core.c
 create mode 100644 include/linux/pci-pwrctl.h

Comments

Simon Horman March 27, 2024, 4:16 p.m. UTC | #1
On Mon, Mar 25, 2024 at 02:16:20PM +0100, Bartosz Golaszewski wrote:
> From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> 
> Some PCI devices must be powered-on before they can be detected on the
> bus. Introduce a simple framework reusing the existing PCI OF
> infrastructure.
> 
> The way this works is: a DT node representing a PCI device connected to
> the port can be matched against its power control platform driver. If
> the match succeeds, the driver is responsible for powering-up the device
> and calling pcie_pwrctl_device_set_ready() which will trigger a PCI bus
> rescan as well as subscribe to PCI bus notifications.
> 
> When the device is detected and created, we'll make it consume the same
> DT node that the platform device did. When the device is bound, we'll
> create a device link between it and the parent power control device.
> 
> Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>

Hi Bartosz,

some minor Kernel doc nits from my side.

...

> diff --git a/drivers/pci/pwrctl/core.c b/drivers/pci/pwrctl/core.c

...

> +/**
> + * devm_pci_pwrctl_device_set_ready - Managed variant of
> + * pci_pwrctl_device_set_ready().
> + *

nit: @dev should be documented here

> + * @pwrctl: PCI power control data
> + *
> + * Returns:
> + * 0 on success, negative error number on error.
> + */
> +int devm_pci_pwrctl_device_set_ready(struct device *dev,
> +				     struct pci_pwrctl *pwrctl)
> +{
> +	int ret;
> +
> +	ret = pci_pwrctl_device_set_ready(pwrctl);
> +	if (ret)
> +		return ret;
> +
> +	return devm_add_action_or_reset(dev,
> +					devm_pci_pwrctl_device_unset_ready,
> +					pwrctl);
> +}
> +EXPORT_SYMBOL_GPL(devm_pci_pwrctl_device_set_ready);
> +
> +MODULE_AUTHOR("Bartosz Golaszewski <bartosz.golaszewski@linaro.org>");
> +MODULE_DESCRIPTION("PCI Device Power Control core driver");
> +MODULE_LICENSE("GPL");
> diff --git a/include/linux/pci-pwrctl.h b/include/linux/pci-pwrctl.h
> new file mode 100644
> index 000000000000..ae8324ea7eeb
> --- /dev/null
> +++ b/include/linux/pci-pwrctl.h
> @@ -0,0 +1,51 @@
> +/* SPDX-License-Identifier: GPL-2.0-only */
> +/*
> + * Copyright (C) 2024 Linaro Ltd.
> + */
> +
> +#ifndef __PCI_PWRCTL_H__
> +#define __PCI_PWRCTL_H__
> +
> +#include <linux/notifier.h>
> +
> +struct device;
> +struct device_link;
> +
> +/*
> + * This is a simple framework for solving the issue of PCI devices that require
> + * certain resources (regulators, GPIOs, clocks) to be enabled before the
> + * device can actually be detected on the PCI bus.
> + *
> + * The idea is to reuse the platform bus to populate OF nodes describing the
> + * PCI device and its resources, let these platform devices probe and enable
> + * relevant resources and then trigger a rescan of the PCI bus allowing for the
> + * same device (with a second associated struct device) to be registered with
> + * the PCI subsystem.
> + *
> + * To preserve a correct hierarchy for PCI power management and device reset,
> + * we create a device link between the power control platform device (parent)
> + * and the supplied PCI device (child).
> + */
> +
> +/**
> + * struct pci_pwrctl - PCI device power control context.
> + * @dev - Address of the power controlling device.

nit: I think this should be "@dev: " rather than "@dev - "
     As is, "./scripts/kernel-doc -none" complains.
> + *
> + * An object of this type must be allocated by the PCI power control device and
> + * passed to the pwrctl subsystem to trigger a bus rescan and setup a device
> + * link with the device once it's up.
> + */
> +struct pci_pwrctl {
> +	struct device *dev;
> +
> +	/* Private, don't use. */

I think Private needs to be followed by a ':' rather than a ',' to keep
kernel-doc happy.

> +	struct notifier_block nb;
> +	struct device_link *link;
> +};
> +
> +int pci_pwrctl_device_set_ready(struct pci_pwrctl *pwrctl);
> +void pci_pwrctl_device_unset_ready(struct pci_pwrctl *pwrctl);
> +int devm_pci_pwrctl_device_set_ready(struct device *dev,
> +				     struct pci_pwrctl *pwrctl);
> +
> +#endif /* __PCI_PWRCTL_H__ */
> -- 
> 2.40.1
> 
>
diff mbox series

Patch

diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig
index d35001589d88..aa4d1833f442 100644
--- a/drivers/pci/Kconfig
+++ b/drivers/pci/Kconfig
@@ -296,5 +296,6 @@  source "drivers/pci/hotplug/Kconfig"
 source "drivers/pci/controller/Kconfig"
 source "drivers/pci/endpoint/Kconfig"
 source "drivers/pci/switch/Kconfig"
+source "drivers/pci/pwrctl/Kconfig"
 
 endif
diff --git a/drivers/pci/Makefile b/drivers/pci/Makefile
index 175302036890..8ddad57934a6 100644
--- a/drivers/pci/Makefile
+++ b/drivers/pci/Makefile
@@ -9,6 +9,7 @@  obj-$(CONFIG_PCI)		+= access.o bus.o probe.o host-bridge.o \
 
 obj-$(CONFIG_PCI)		+= msi/
 obj-$(CONFIG_PCI)		+= pcie/
+obj-$(CONFIG_PCI)		+= pwrctl/
 
 ifdef CONFIG_PCI
 obj-$(CONFIG_PROC_FS)		+= proc.o
diff --git a/drivers/pci/pwrctl/Kconfig b/drivers/pci/pwrctl/Kconfig
new file mode 100644
index 000000000000..96195395af69
--- /dev/null
+++ b/drivers/pci/pwrctl/Kconfig
@@ -0,0 +1,8 @@ 
+# SPDX-License-Identifier: GPL-2.0-only
+
+menu "PCI Power control drivers"
+
+config PCI_PWRCTL
+	tristate
+
+endmenu
diff --git a/drivers/pci/pwrctl/Makefile b/drivers/pci/pwrctl/Makefile
new file mode 100644
index 000000000000..52ae0640ef7b
--- /dev/null
+++ b/drivers/pci/pwrctl/Makefile
@@ -0,0 +1,4 @@ 
+# SPDX-License-Identifier: GPL-2.0-only
+
+obj-$(CONFIG_PCI_PWRCTL)		+= pci-pwrctl-core.o
+pci-pwrctl-core-y			:= core.o
diff --git a/drivers/pci/pwrctl/core.c b/drivers/pci/pwrctl/core.c
new file mode 100644
index 000000000000..11bd19519841
--- /dev/null
+++ b/drivers/pci/pwrctl/core.c
@@ -0,0 +1,136 @@ 
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2024 Linaro Ltd.
+ */
+
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/pci-pwrctl.h>
+#include <linux/property.h>
+#include <linux/slab.h>
+
+static int pci_pwrctl_notify(struct notifier_block *nb, unsigned long action,
+			     void *data)
+{
+	struct pci_pwrctl *pwrctl = container_of(nb, struct pci_pwrctl, nb);
+	struct device *dev = data;
+
+	if (dev_fwnode(dev) != dev_fwnode(pwrctl->dev))
+		return NOTIFY_DONE;
+
+	switch (action) {
+	case BUS_NOTIFY_ADD_DEVICE:
+		/*
+		 * We will have two struct device objects bound to two different
+		 * drivers on different buses but consuming the same DT node. We
+		 * must not bind the pins twice in this case but only once for
+		 * the first device to be added.
+		 *
+		 * If we got here then the PCI device is the second after the
+		 * power control platform device. Mark its OF node as reused.
+		 */
+		dev->of_node_reused = true;
+		break;
+	case BUS_NOTIFY_BOUND_DRIVER:
+		pwrctl->link = device_link_add(dev, pwrctl->dev,
+					       DL_FLAG_AUTOREMOVE_CONSUMER);
+		if (!pwrctl->link)
+			dev_err(pwrctl->dev, "Failed to add device link\n");
+		break;
+	case BUS_NOTIFY_UNBOUND_DRIVER:
+		if (pwrctl->link)
+			device_link_del(pwrctl->link);
+		break;
+	}
+
+	return NOTIFY_DONE;
+}
+
+/**
+ * pci_pwrctl_device_set_ready() - Notify the pwrctl subsystem that the PCI
+ * device is powered-up and ready to be detected.
+ *
+ * @pwrctl: PCI power control data
+ *
+ * Returns:
+ * 0 on success, negative error number on error.
+ *
+ * Note:
+ * This function returning 0 doesn't mean the device was detected. It means,
+ * that the bus rescan was successfully started. The device will get bound to
+ * its PCI driver asynchronously.
+ */
+int pci_pwrctl_device_set_ready(struct pci_pwrctl *pwrctl)
+{
+	int ret;
+
+	if (!pwrctl->dev)
+		return -ENODEV;
+
+	pwrctl->nb.notifier_call = pci_pwrctl_notify;
+	ret = bus_register_notifier(&pci_bus_type, &pwrctl->nb);
+	if (ret)
+		return ret;
+
+	pci_lock_rescan_remove();
+	pci_rescan_bus(to_pci_dev(pwrctl->dev->parent)->bus);
+	pci_unlock_rescan_remove();
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(pci_pwrctl_device_set_ready);
+
+/**
+ * pci_pwrctl_device_unset_ready() - Notify the pwrctl subsystem that the PCI
+ * device is about to be powered-down.
+ *
+ * @pwrctl: PCI power control data
+ */
+void pci_pwrctl_device_unset_ready(struct pci_pwrctl *pwrctl)
+{
+	/*
+	 * We don't have to delete the link here. Typically, this function
+	 * is only called when the power control device is being detached. If
+	 * it is being detached then the child PCI device must have already
+	 * been unbound too or the device core wouldn't let us unbind.
+	 */
+	bus_unregister_notifier(&pci_bus_type, &pwrctl->nb);
+}
+EXPORT_SYMBOL_GPL(pci_pwrctl_device_unset_ready);
+
+static void devm_pci_pwrctl_device_unset_ready(void *data)
+{
+	struct pci_pwrctl *pwrctl = data;
+
+	pci_pwrctl_device_set_ready(pwrctl);
+}
+
+/**
+ * devm_pci_pwrctl_device_set_ready - Managed variant of
+ * pci_pwrctl_device_set_ready().
+ *
+ * @pwrctl: PCI power control data
+ *
+ * Returns:
+ * 0 on success, negative error number on error.
+ */
+int devm_pci_pwrctl_device_set_ready(struct device *dev,
+				     struct pci_pwrctl *pwrctl)
+{
+	int ret;
+
+	ret = pci_pwrctl_device_set_ready(pwrctl);
+	if (ret)
+		return ret;
+
+	return devm_add_action_or_reset(dev,
+					devm_pci_pwrctl_device_unset_ready,
+					pwrctl);
+}
+EXPORT_SYMBOL_GPL(devm_pci_pwrctl_device_set_ready);
+
+MODULE_AUTHOR("Bartosz Golaszewski <bartosz.golaszewski@linaro.org>");
+MODULE_DESCRIPTION("PCI Device Power Control core driver");
+MODULE_LICENSE("GPL");
diff --git a/include/linux/pci-pwrctl.h b/include/linux/pci-pwrctl.h
new file mode 100644
index 000000000000..ae8324ea7eeb
--- /dev/null
+++ b/include/linux/pci-pwrctl.h
@@ -0,0 +1,51 @@ 
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2024 Linaro Ltd.
+ */
+
+#ifndef __PCI_PWRCTL_H__
+#define __PCI_PWRCTL_H__
+
+#include <linux/notifier.h>
+
+struct device;
+struct device_link;
+
+/*
+ * This is a simple framework for solving the issue of PCI devices that require
+ * certain resources (regulators, GPIOs, clocks) to be enabled before the
+ * device can actually be detected on the PCI bus.
+ *
+ * The idea is to reuse the platform bus to populate OF nodes describing the
+ * PCI device and its resources, let these platform devices probe and enable
+ * relevant resources and then trigger a rescan of the PCI bus allowing for the
+ * same device (with a second associated struct device) to be registered with
+ * the PCI subsystem.
+ *
+ * To preserve a correct hierarchy for PCI power management and device reset,
+ * we create a device link between the power control platform device (parent)
+ * and the supplied PCI device (child).
+ */
+
+/**
+ * struct pci_pwrctl - PCI device power control context.
+ * @dev - Address of the power controlling device.
+ *
+ * An object of this type must be allocated by the PCI power control device and
+ * passed to the pwrctl subsystem to trigger a bus rescan and setup a device
+ * link with the device once it's up.
+ */
+struct pci_pwrctl {
+	struct device *dev;
+
+	/* Private, don't use. */
+	struct notifier_block nb;
+	struct device_link *link;
+};
+
+int pci_pwrctl_device_set_ready(struct pci_pwrctl *pwrctl);
+void pci_pwrctl_device_unset_ready(struct pci_pwrctl *pwrctl);
+int devm_pci_pwrctl_device_set_ready(struct device *dev,
+				     struct pci_pwrctl *pwrctl);
+
+#endif /* __PCI_PWRCTL_H__ */