All of lore.kernel.org
 help / color / mirror / Atom feed
From: Hans de Goede <hdegoede@redhat.com>
To: Philipp Jungkamp <p.jungkamp@gmx.net>,
	Armin Wolf <W_Armin@gmx.de>, Mark Gross <markgross@kernel.org>
Cc: platform-driver-x86@vger.kernel.org
Subject: Re: [PATCH v3] platform/x86: ideapad-laptop: support for more special keys in WMI
Date: Mon, 14 Nov 2022 17:41:33 +0100	[thread overview]
Message-ID: <7e263715-e759-9293-ec53-27033a79a2bc@redhat.com> (raw)
In-Reply-To: <20221114144140.5054-1-p.jungkamp@gmx.net>

Hi,

On 11/14/22 15:41, Philipp Jungkamp wrote:
> The event data of the WMI event 0xD0, which is assumed to be the
> fn_lock, is used to indicate several special keys on newer Yoga 7/9
> laptops.
> 
> The notify_id 0xD0 is non-unique in the DSDT of the Yoga 9 14IAP7, this
> causes wmi_get_event_data() to report wrong values.
> Port the ideapad-laptop WMI code to the wmi bus infrastructure which
> does not suffer from the shortcomings of wmi_get_event_data().
> 
> Signed-off-by: Philipp Jungkamp <p.jungkamp@gmx.net>
> ---
> Hello,
> 
> is this about right? It works for me.
> 
> What I don't really like here is the dev_set_drvdata() which takes a non-const
> void * and I pass it a const pointer. I do cast the value of dev_get_drvdata()
> back to a const pointer, but this seems rather ugly.
> I preferred it over allocating a single int for the device or casting an enum
> to a void *. This additionally removes the need for a remove funtion.

I decided just take quick peek and I think the cleanest solution here would
be to add a driver-data struct with the lookup result of:

	struct ideapad_private *priv = container_of(wdrv,
						    struct ideapad_private,
						    wmi_drv);

Stored in there + the enum value and then alloc it with devm_kzalloc
to avoid the need for a remove callback.

I'm also wondering if we could then maybe not move some other variables
only used in the wmi_notify callback to that driver-data struct ?
(I did not check if there are any candidates).

Also this is going to need a big fat warning (comment) that the cute trick
with registering a wmi_driver struct embedded inside the platform_driver
data struct very much relies on there being only one platform_device
instance to which the platform_driver will bind ever, otherwise
we will get multiple wmi_driver's registered for the same WMI GUIDs
and then the container-off will likely return the driver-data of the
first platform device ...

Which makes me wonder if it would not be cleaner to just use a global
pointer for this ?   Note this is an honest open question.

Actually since the platform_device gets instantiated by the ACPI
core there is no guarantee there will be only 1. So I think that
the container_of on the wmi-driver trick needs to go, instead
introduce:

1. A global ideapad_private_data_mutex mutex
2. A global ideapad_private_data pointer

And:

1. In ideapad_acpi_add:
lock the mutex
check that ideapad_private_data is not already set and if it is bail with an error
set ideapad_private_data
unlock the mutex

2. in ideapad_acpi_remove:
lock the mutex
clear the golbal pointer
unlock the mutex

3. In the wmi-driver's notify method:
lock the mutex
check ideapad_private_data is not NULL
process event
unlock the mutex

4. replace module_platform_driver with normal module init + exit
functions which register both drivers / unregister both drivers.

I believe that this will be a more clean approach then the embedded
wmi_driver struct cuteness.

Regards,

Hans




> 
> Regards,
> Philipp
> 
>  drivers/platform/x86/ideapad-laptop.c | 109 +++++++++++++++++++-------
>  1 file changed, 80 insertions(+), 29 deletions(-)
> 
> diff --git a/drivers/platform/x86/ideapad-laptop.c b/drivers/platform/x86/ideapad-laptop.c
> index 33b3dfdd1b08..6d35a9e961cf 100644
> --- a/drivers/platform/x86/ideapad-laptop.c
> +++ b/drivers/platform/x86/ideapad-laptop.c
> @@ -30,6 +30,7 @@
>  #include <linux/seq_file.h>
>  #include <linux/sysfs.h>
>  #include <linux/types.h>
> +#include <linux/wmi.h>
> 
>  #include <acpi/video.h>
> 
> @@ -38,10 +39,19 @@
>  #define IDEAPAD_RFKILL_DEV_NUM	3
> 
>  #if IS_ENABLED(CONFIG_ACPI_WMI)
> -static const char *const ideapad_wmi_fnesc_events[] = {
> -	"26CAB2E5-5CF1-46AE-AAC3-4A12B6BA50E6", /* Yoga 3 */
> -	"56322276-8493-4CE8-A783-98C991274F5E", /* Yoga 700 */
> -	"8FC0DE0C-B4E4-43FD-B0F3-8871711C1294", /* Legion 5 */
> +enum ideapad_wmi_event_type {
> +	IDEAPAD_WMI_EVENT_ESC,
> +	IDEAPAD_WMI_EVENT_FN_KEYS,
> +};
> +
> +enum ideapad_wmi_event_type ideapad_wmi_esc = IDEAPAD_WMI_EVENT_ESC,
> +enum ideapad_wmi_event_type ideapad_wmi_fn_keys = IDEAPAD_WMI_EVENT_FN_KEYS;
> +
> +static const struct wmi_device_id ideapad_wmi_id_table[] = {
> +	{ "26CAB2E5-5CF1-46AE-AAC3-4A12B6BA50E6", &ideapad_wmi_esc }, /* Yoga 3 */
> +	{ "56322276-8493-4CE8-A783-98C991274F5E", &ideapad_wmi_esc }, /* Yoga 700 */
> +	{ "8FC0DE0C-B4E4-43FD-B0F3-8871711C1294", &ideapad_wmi_fn_keys }, /* Legion 5 */
> +	{}
>  };
>  #endif
> 
> @@ -130,7 +140,7 @@ struct ideapad_private {
>  	struct ideapad_dytc_priv *dytc;
>  	struct dentry *debug;
>  	unsigned long cfg;
> -	const char *fnesc_guid;
> +	struct wmi_driver wmi_drv;
>  	struct {
>  		bool conservation_mode    : 1;
>  		bool dytc                 : 1;
> @@ -1074,6 +1084,7 @@ static void ideapad_sysfs_exit(struct ideapad_private *priv)
>  /*
>   * input device
>   */
> +#define IDEAPAD_WMI_KEY 0x100
>  static const struct key_entry ideapad_keymap[] = {
>  	{ KE_KEY,   6, { KEY_SWITCHVIDEOMODE } },
>  	{ KE_KEY,   7, { KEY_CAMERA } },
> @@ -1087,6 +1098,26 @@ static const struct key_entry ideapad_keymap[] = {
>  	{ KE_KEY,  66, { KEY_TOUCHPAD_OFF } },
>  	{ KE_KEY,  67, { KEY_TOUCHPAD_ON } },
>  	{ KE_KEY, 128, { KEY_ESC } },
> +
> +	/*
> +	 * WMI keys
> +	 */
> +
> +	/* FnLock (handled by the firmware) */
> +	{ KE_IGNORE,	0x02 | IDEAPAD_WMI_KEY },
> +	/* Customizable Lenovo Hotkey ("star" with 'S' inside) */
> +	{ KE_KEY,	0x01 | IDEAPAD_WMI_KEY, { KEY_FAVORITES } },
> +	/* Dark mode toggle */
> +	{ KE_KEY,	0x13 | IDEAPAD_WMI_KEY, { KEY_PROG1 } },
> +	/* Sound profile switch */
> +	{ KE_KEY,	0x12 | IDEAPAD_WMI_KEY, { KEY_PROG2 } },
> +	/* Lenovo Virtual Background application */
> +	{ KE_KEY,	0x28 | IDEAPAD_WMI_KEY, { KEY_PROG3 } },
> +	/* Lenovo Support */
> +	{ KE_KEY,	0x27 | IDEAPAD_WMI_KEY, { KEY_HELP } },
> +	/* Refresh Rate Toggle */
> +	{ KE_KEY,	0x0a | IDEAPAD_WMI_KEY, { KEY_DISPLAYTOGGLE } },
> +
>  	{ KE_END },
>  };
> 
> @@ -1491,25 +1522,47 @@ static void ideapad_acpi_notify(acpi_handle handle, u32 event, void *data)
>  }
> 
>  #if IS_ENABLED(CONFIG_ACPI_WMI)
> -static void ideapad_wmi_notify(u32 value, void *context)
> +static int ideapad_wmi_probe(struct wmi_device *wdev, const void *context)
>  {
> -	struct ideapad_private *priv = context;
> +	dev_set_drvdata(&wdev->dev, (void *) context);
> +	return 0;
> +}
> +
> +static void ideapad_wmi_notify(struct wmi_device *wdev, union acpi_object *data)
> +{
> +	struct wmi_driver *wdrv = container_of(wdev->dev.driver,
> +					       struct wmi_driver,
> +					       driver);
> +	struct ideapad_private *priv = container_of(wdrv,
> +						    struct ideapad_private,
> +						    wmi_drv);
> +	const enum ideapad_wmi_event_type *event = dev_get_drvdata(&wdev->dev);
>  	unsigned long result;
> 
> -	switch (value) {
> -	case 128:
> -		ideapad_input_report(priv, value);
> +	switch (*event) {
> +	case IDEAPAD_WMI_EVENT_ESC:
> +		ideapad_input_report(priv, 128);
>  		break;
> -	case 208:
> +	case IDEAPAD_WMI_EVENT_FN_KEYS:
>  		if (!eval_hals(priv->adev->handle, &result)) {
>  			bool state = test_bit(HALS_FNLOCK_STATE_BIT, &result);
> 
>  			exec_sals(priv->adev->handle, state ? SALS_FNLOCK_ON : SALS_FNLOCK_OFF);
>  		}
> +
> +		if (data->type != ACPI_TYPE_INTEGER) {
> +			dev_warn(&wdev->dev,
> +				 "WMI event data is not an integer\n");
> +			break;
> +		}
> +
> +		dev_dbg(&wdev->dev, "WMI fn-key event: 0x%llx\n",
> +			data->integer.value);
> +
> +		ideapad_input_report(priv,
> +				     data->integer.value | IDEAPAD_WMI_KEY);
> +
>  		break;
> -	default:
> -		dev_info(&priv->platform_device->dev,
> -			 "Unknown WMI event: %u\n", value);
>  	}
>  }
>  #endif
> @@ -1671,25 +1724,24 @@ static int ideapad_acpi_add(struct platform_device *pdev)
>  	}
> 
>  #if IS_ENABLED(CONFIG_ACPI_WMI)
> -	for (i = 0; i < ARRAY_SIZE(ideapad_wmi_fnesc_events); i++) {
> -		status = wmi_install_notify_handler(ideapad_wmi_fnesc_events[i],
> -						    ideapad_wmi_notify, priv);
> -		if (ACPI_SUCCESS(status)) {
> -			priv->fnesc_guid = ideapad_wmi_fnesc_events[i];
> -			break;
> -		}
> -	}
> +	priv->wmi_drv = (struct wmi_driver) {
> +		.driver = {
> +			.name = "ideapad-wmi-fn-keys",
> +		},
> +		.id_table = ideapad_wmi_id_table,
> +		.probe = ideapad_wmi_probe,
> +		.notify = ideapad_wmi_notify,
> +	};
> 
> -	if (ACPI_FAILURE(status) && status != AE_NOT_EXIST) {
> -		err = -EIO;
> -		goto notification_failed_wmi;
> -	}
> +	err = wmi_driver_register(&priv->wmi_drv);
> +	if (err)
> +		goto register_failed_wmi;
>  #endif
> 
>  	return 0;
> 
>  #if IS_ENABLED(CONFIG_ACPI_WMI)
> -notification_failed_wmi:
> +register_failed_wmi:
>  	acpi_remove_notify_handler(priv->adev->handle,
>  				   ACPI_DEVICE_NOTIFY,
>  				   ideapad_acpi_notify);
> @@ -1720,8 +1772,7 @@ static int ideapad_acpi_remove(struct platform_device *pdev)
>  	int i;
> 
>  #if IS_ENABLED(CONFIG_ACPI_WMI)
> -	if (priv->fnesc_guid)
> -		wmi_remove_notify_handler(priv->fnesc_guid);
> +	wmi_driver_unregister(&priv->wmi_drv);
>  #endif
> 
>  	acpi_remove_notify_handler(priv->adev->handle,
> --
> 2.38.1
> 


  parent reply	other threads:[~2022-11-14 16:42 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-11 16:04 [PATCH] Add IdeaPad WMI Fn Keys driver Philipp Jungkamp
2022-09-11 16:18 ` Hans de Goede
2022-09-19  7:54   ` Hans de Goede
2022-11-10 20:02     ` Philipp Jungkamp
2022-11-10 21:02       ` Hans de Goede
2022-11-13 12:12         ` [PATCH v2 1/2] platform/x86: wmi: Disambiguate WMI event data on some ACPI tables Philipp Jungkamp
2022-11-13 12:12           ` [PATCH v2 2/2] platform/x86: ideapad-laptop: support for more special keys in WMI Philipp Jungkamp
2022-11-13 20:30           ` [PATCH v2 1/2] platform/x86: wmi: Disambiguate WMI event data on some ACPI tables Armin Wolf
2022-11-13 21:42             ` Philipp Jungkamp
2022-11-13 23:02               ` Armin Wolf
2022-11-14 14:41                 ` [PATCH v3] platform/x86: ideapad-laptop: support for more special keys in WMI Philipp Jungkamp
2022-11-14 16:15                   ` Hans de Goede
2022-11-14 16:41                   ` Hans de Goede [this message]
2022-11-15  0:09                     ` [PATCH v4] " Philipp Jungkamp
2022-11-15 21:19                       ` Hans de Goede
2022-11-16 11:06                         ` [PATCH v5] " Philipp Jungkamp
2022-11-16 14:43                           ` Hans de Goede
2022-11-15  2:27                     ` [PATCH v3] " Armin Wolf
2022-11-15 11:19                       ` Hans de Goede

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=7e263715-e759-9293-ec53-27033a79a2bc@redhat.com \
    --to=hdegoede@redhat.com \
    --cc=W_Armin@gmx.de \
    --cc=markgross@kernel.org \
    --cc=p.jungkamp@gmx.net \
    --cc=platform-driver-x86@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.