[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAJZ5v0g9HjLr8n3OQwMY0EK5GdCc+8CJnO3mEUXom3g2sz9jXQ@mail.gmail.com>
Date: Fri, 17 Dec 2021 18:01:48 +0100
From: "Rafael J. Wysocki" <rafael@...nel.org>
To: Heikki Krogerus <heikki.krogerus@...ux.intel.com>
Cc: "Rafael J. Wysocki" <rafael@...nel.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
Sakari Ailus <sakari.ailus@...ux.intel.com>,
Prashant Malani <pmalani@...omium.org>,
ACPI Devel Maling List <linux-acpi@...r.kernel.org>,
"open list:ULTRA-WIDEBAND (UWB) SUBSYSTEM:"
<linux-usb@...r.kernel.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 1/4] acpi: Store the known device locations
On Fri, Dec 17, 2021 at 2:24 PM Heikki Krogerus
<heikki.krogerus@...ux.intel.com> wrote:
>
> This adds a list that will hold all the detected device
> locations. The location of a device is known if it has
> Physical Location of Device (_PLD) object within its scope.
This paragraph isn't really accurate any more, because the location
information is not stored in that list.
To be precise, the new list is a list of entries that each contain a
list of devices sharing the same physical location information (and
thus presumably being physically located in the same place or
physically overlapping so to speak).
Honestly, I would change the terminology and naming to reflect that
concept (see below).
> Each entry in the list represents a known location, and each
> of those locations can then have a list of devices that are
> currently assigned to those locations.
>
> The location entry that contains the current location of a
> device can be acquired with a new function
> acpi_device_get_location(). The location structure returned
> by this function contains the list of devices sharing it.
>
> The knowledge of the other devices that share a location
> can be used in device drivers that need to know the
> connections to other components inside a system. USB3 ports
> will for example always share their location with a USB2
> port.
>
> For now, the device locations can not be updated, so they
> will only contain lists the devices that are initially in
> those locations. But that can later be easily changed if
> needed by adding API that can be used to update the
> locations.
>
> Signed-off-by: Heikki Krogerus <heikki.krogerus@...ux.intel.com>
> ---
> drivers/acpi/scan.c | 77 +++++++++++++++++++++++++++++++++++++++++
> include/acpi/acpi_bus.h | 19 ++++++++++
> 2 files changed, 96 insertions(+)
>
> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
> index 5991dddbc9ceb..f147c0ad5f944 100644
> --- a/drivers/acpi/scan.c
> +++ b/drivers/acpi/scan.c
> @@ -19,6 +19,7 @@
> #include <linux/dma-map-ops.h>
> #include <linux/platform_data/x86/apple.h>
> #include <linux/pgtable.h>
> +#include <linux/crc32.h>
>
> #include "internal.h"
>
> @@ -42,6 +43,8 @@ static LIST_HEAD(acpi_scan_handlers_list);
> DEFINE_MUTEX(acpi_device_lock);
> LIST_HEAD(acpi_wakeup_device_list);
> static DEFINE_MUTEX(acpi_hp_context_lock);
> +static LIST_HEAD(acpi_location_list);
I would call this "acpi_location_sharing_list".
> +static DEFINE_MUTEX(acpi_location_lock);
And this "acpi_location_sharing_lock".
>
> /*
> * The UART device described by the SPCR table is the only object which needs
> @@ -485,6 +488,7 @@ static void acpi_device_del(struct acpi_device *device)
> break;
> }
>
> + list_del(&device->location_list);
> list_del(&device->wakeup_list);
> mutex_unlock(&acpi_device_lock);
>
> @@ -654,6 +658,76 @@ static int acpi_tie_acpi_dev(struct acpi_device *adev)
> return 0;
> }
>
> +static void acpi_store_device_location(struct acpi_device *adev)
This can be called "acpi_dev_save_location_sharing_info()".
> +{
> + struct acpi_device_location *location;
> + struct acpi_pld_info *pld;
> + acpi_status status;
> + u32 crc;
> +
> + status = acpi_get_physical_device_location(adev->handle, &pld);
> + if (ACPI_FAILURE(status))
> + return;
> +
> + crc = crc32(~0, pld, sizeof(*pld));
> +
> + mutex_lock(&acpi_location_lock);
> +
> + list_for_each_entry(location, &acpi_location_list, node)
> + if (location->pld_crc == crc)
> + goto out_add_to_location;
> +
> + /* The location does not exist yet so creating it. */
> +
> + location = kzalloc(sizeof(*location), GFP_KERNEL);
> + if (!location) {
> + acpi_handle_err(adev->handle, "Unable to store location\n");
> + goto err_unlock;
> + }
> +
> + list_add_tail(&location->node, &acpi_location_list);
> + INIT_LIST_HEAD(&location->devices);
> + location->pld_crc = crc;
> +
> +out_add_to_location:
> + list_add_tail(&adev->location_list, &location->devices);
> +
> +err_unlock:
> + ACPI_FREE(pld);
> + mutex_unlock(&acpi_location_lock);
> +}
> +
> +/**
> + * acpi_device_get_location - Get the device location
> + * @adev: ACPI device handle
> + *
> + * Return a pointer to a struct acpi_device_location object containing the
> + * location information obtained by evaluating _PLD (Physical Location of
> + * Device) for @adev when it is available, along with the list of devices
> + * sharing the same location information (if any), or NULL otherwise.
> + */
> +struct acpi_device_location *acpi_device_get_location(struct acpi_device *adev)
And this "acpi_dev_get_location_sharing_info()".
> +{
> + struct acpi_device_location *location;
> + struct list_head *tmp;
> +
> + mutex_lock(&acpi_location_lock);
> +
> + list_for_each_entry(location, &acpi_location_list, node) {
> + list_for_each(tmp, &location->devices) {
> + if (tmp == &adev->location_list)
> + goto out_unlock;
> + }
> + }
> + location = NULL;
> +
> +out_unlock:
> + mutex_unlock(&acpi_location_lock);
> +
> + return location;
> +}
> +EXPORT_SYMBOL_GPL(acpi_device_get_location);
> +
> static int __acpi_device_add(struct acpi_device *device,
> void (*release)(struct device *))
> {
> @@ -670,6 +744,7 @@ static int __acpi_device_add(struct acpi_device *device,
> INIT_LIST_HEAD(&device->wakeup_list);
> INIT_LIST_HEAD(&device->physical_node_list);
> INIT_LIST_HEAD(&device->del_list);
> + INIT_LIST_HEAD(&device->location_list);
> mutex_init(&device->physical_node_lock);
>
> mutex_lock(&acpi_device_lock);
> @@ -712,6 +787,8 @@ static int __acpi_device_add(struct acpi_device *device,
> if (device->wakeup.flags.valid)
> list_add_tail(&device->wakeup_list, &acpi_wakeup_device_list);
>
> + acpi_store_device_location(device);
> +
> mutex_unlock(&acpi_device_lock);
>
> if (device->parent)
> diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h
> index d6fe27b695c3d..9123884e4e7ec 100644
> --- a/include/acpi/acpi_bus.h
> +++ b/include/acpi/acpi_bus.h
> @@ -354,6 +354,18 @@ struct acpi_device_data {
> struct list_head subnodes;
> };
>
> +/*
> + * struct acpi_device_location - Device location based on _PLD
> + * @devices: List of devices that share this location
> + * @node: Entry in the internal list of locations
> + * @pld_crc: CRC-32 hash of the _PLD
> + */
> +struct acpi_device_location {
"acpi_dev_location_sharing_info" ?
> + struct list_head devices;
> + struct list_head node;
> + u32 pld_crc;
> +};
> +
> struct acpi_gpio_mapping;
>
> /* Device */
> @@ -366,6 +378,7 @@ struct acpi_device {
> struct list_head node;
> struct list_head wakeup_list;
> struct list_head del_list;
> + struct list_head location_list;
"location_sharing_list" ?
> struct acpi_device_status status;
> struct acpi_device_flags flags;
> struct acpi_device_pnp pnp;
> @@ -731,11 +744,17 @@ static inline void acpi_bus_put_acpi_device(struct acpi_device *adev)
> {
> acpi_dev_put(adev);
> }
> +
> +struct acpi_device_location *acpi_device_get_location(struct acpi_device *adev);
> #else /* CONFIG_ACPI */
>
> static inline int register_acpi_bus_type(void *bus) { return 0; }
> static inline int unregister_acpi_bus_type(void *bus) { return 0; }
>
> +static inline struct acpi_device_location *acpi_device_get_location(struct acpi_device *adev)
> +{
> + return NULL;
> +}
> #endif /* CONFIG_ACPI */
>
> #endif /*__ACPI_BUS_H__*/
> --
And overall I'm wondering if this can be achieved by storing the
pld_crc directly in struct acpi_device and doing a
bus_for_each_dev(&acpi_bus_type, ...) walk every time a list of
devices sharing a _PLD is needed?
It looks like typec_link_ports() is the only user of this and it can
easily afford doing a walk like the above if I'm not mistaken.
Powered by blists - more mailing lists