linux-kernel - Re: [PATCH v1 4/6] Input: mt6779-keypad

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <b2676b5c-14b3-2058-9fb8-d6d78cc5d29c@collabora.com>
Date:   Thu, 21 Jul 2022 10:34:00 +0200
From:   AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@...labora.com>
To:     Mattijs Korpershoek <mkorpershoek@...libre.com>,
        Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>,
        Matthias Brugger <matthias.bgg@...il.com>
Cc:     linux-kernel@...r.kernel.org, linux-input@...r.kernel.org,
        Fabien Parent <parent.f@...il.com>, devicetree@...r.kernel.org,
        linux-mediatek@...ts.infradead.org,
        Fabien Parent <fparent@...libre.com>,
        linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH v1 4/6] Input: mt6779-keypad - support double keys matrix

Il 20/07/22 16:48, Mattijs Korpershoek ha scritto:
> MediaTek keypad has 2 modes of detecting key events:
> - single key: each (row, column) can detect one key
> - double key: each (row, column) is a group of 2 keys
> 
> Double key support exists to minimize cost, since it reduces the number
> of pins required for physical keys.
> 
> Double key is configured by setting BIT(0) of the KP_SEL register.
> 
> Enable double key matrix support based on the mediatek,double-keys
> device tree property.
> 
> Signed-off-by: Mattijs Korpershoek <mkorpershoek@...libre.com>
> Reviewed-by: Matthias Brugger <matthias.bgg@...il.com>
> 
> diff --git a/drivers/input/keyboard/mt6779-keypad.c b/drivers/input/keyboard/mt6779-keypad.c
> index bf447bf598fb..9a5dbd415dac 100644
> --- a/drivers/input/keyboard/mt6779-keypad.c
> +++ b/drivers/input/keyboard/mt6779-keypad.c
> @@ -18,6 +18,7 @@
>   #define MTK_KPD_DEBOUNCE_MASK	GENMASK(13, 0)
>   #define MTK_KPD_DEBOUNCE_MAX_MS	256
>   #define MTK_KPD_SEL		0x0020
> +#define MTK_KPD_SEL_DOUBLE_KP_MODE	BIT(0)
>   #define MTK_KPD_SEL_COL	GENMASK(15, 10)
>   #define MTK_KPD_SEL_ROW	GENMASK(9, 4)
>   #define MTK_KPD_SEL_COLMASK(c)	GENMASK((c) + 9, 10)
> @@ -31,6 +32,7 @@ struct mt6779_keypad {
>   	struct clk *clk;
>   	u32 n_rows;
>   	u32 n_cols;
> +	bool double_keys;
>   	DECLARE_BITMAP(keymap_state, MTK_KPD_NUM_BITS);
>   };
>   
> @@ -67,8 +69,13 @@ static irqreturn_t mt6779_keypad_irq_handler(int irq, void *dev_id)
>   			continue;
>   
>   		key = bit_nr / 32 * 16 + bit_nr % 32;
> -		row = key / 9;
> -		col = key % 9;
> +		if (keypad->double_keys) {
> +			row = key / 13;
> +			col = (key % 13) / 2;
> +		} else {
> +			row = key / 9;
> +			col = key % 9;
> +		}

I don't fully like this if branch permanently evaluating true or false, as no
runtime can actually change this result...

In practice, it's fine, but I was wondering if anyone would disagree with the
following proposal...

struct mt6779_keypad {
	.......
	void (*calc_row_col)(unsigned int *row, unsigned int *col);
};

In mt6779_keypad_irq_handler:

	key = bit_nr / 32 * 16 + bit_nr % 32;
	keypad->calc_row_col(&row, &col);

and below...

>   
>   		scancode = MATRIX_SCAN_CODE(row, col, row_shift);
>   		/* 1: not pressed, 0: pressed */
> @@ -150,6 +157,8 @@ static int mt6779_keypad_pdrv_probe(struct platform_device *pdev)
>   
>   	wakeup = device_property_read_bool(&pdev->dev, "wakeup-source");
>   
> +	keypad->double_keys = device_property_read_bool(&pdev->dev, "mediatek,double-keys");
> +
>   	dev_dbg(&pdev->dev, "n_row=%d n_col=%d debounce=%d\n",
>   		keypad->n_rows, keypad->n_cols, debounce);
>   
> @@ -166,6 +175,10 @@ static int mt6779_keypad_pdrv_probe(struct platform_device *pdev)
>   	regmap_write(keypad->regmap, MTK_KPD_DEBOUNCE,
>   		     (debounce * (1 << 5)) & MTK_KPD_DEBOUNCE_MASK);
>   
> +	if (keypad->double_keys)

		keypad->calc_row_col = mt6779_keypad_calc_row_col_double_kp;

> +		regmap_update_bits(keypad->regmap, MTK_KPD_SEL,
> +				   MTK_KPD_SEL_DOUBLE_KP_MODE, MTK_KPD_SEL_DOUBLE_KP_MODE);
> +

	} else {
		keypad->calc_row_col = mt6779_keypad_calc_row_col_single_kp;
	}

>   	regmap_update_bits(keypad->regmap, MTK_KPD_SEL, MTK_KPD_SEL_ROW,
>   			   MTK_KPD_SEL_ROWMASK(keypad->n_rows));
>   	regmap_update_bits(keypad->regmap, MTK_KPD_SEL, MTK_KPD_SEL_COL,

what do you think?

Cheers,
Angelo