linux-kernel - Re: [RFC][PATCH] rtc: sunxi: use external 32k oscillator if provided

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]

Date:   Thu, 27 Jul 2023 17:53:47 +0100
From:   Andre Przywara <andre.przywara@....com>
To:     Mans Rullgard <mans@...sr.com>,
        Alessandro Zummo <a.zummo@...ertech.it>,
        Alexandre Belloni <alexandre.belloni@...tlin.com>,
        Chen-Yu Tsai <wens@...e.org>,
        Jernej Skrabec <jernej.skrabec@...il.com>,
        Samuel Holland <samuel@...lland.org>
Cc:     linux-rtc@...r.kernel.org, linux-sunxi@...ts.linux.dev,
        linux-kernel@...r.kernel.org
Subject: Re: [RFC][PATCH] rtc: sunxi: use external 32k oscillator if provided

Hi Mans,

On 27/07/2023 16:01, Mans Rullgard wrote:
> Set the OSC32K_SRC_SEL bit in the LOSC control register if a clock is
> specified in the devicetree.
> 
> Signed-off-by: Mans Rullgard <mans@...sr.com>
> ---
> The newer sun6i rtc driver is a proper clk provider with parent
> selection.  Doing the same thing in this driver would be difficult
> while staying compatible with existing devicetrees.  For that reason,
> this simpler approach seems reasonable.
> ---
>   drivers/rtc/rtc-sunxi.c | 12 ++++++++++++
>   1 file changed, 12 insertions(+)
> 
> diff --git a/drivers/rtc/rtc-sunxi.c b/drivers/rtc/rtc-sunxi.c
> index 5d019e3a835a..4f1053eab778 100644
> --- a/drivers/rtc/rtc-sunxi.c
> +++ b/drivers/rtc/rtc-sunxi.c
> @@ -5,6 +5,7 @@
>    * Copyright (c) 2013, Carlo Caione <carlo.caione@...il.com>
>    */
>   
> +#include <linux/clk.h>
>   #include <linux/delay.h>
>   #include <linux/err.h>
>   #include <linux/fs.h>
> @@ -21,8 +22,10 @@
>   #include <linux/types.h>
>   
>   #define SUNXI_LOSC_CTRL				0x0000
> +#define SUNXI_LOSC_CTRL_KEY			(0x16aa << 16)
>   #define SUNXI_LOSC_CTRL_RTC_HMS_ACC		BIT(8)
>   #define SUNXI_LOSC_CTRL_RTC_YMD_ACC		BIT(7)
> +#define SUNXI_LOSC_CTRL_OSC32K_SRC_SEL		BIT(0)
>   
>   #define SUNXI_RTC_YMD				0x0004
>   
> @@ -422,6 +425,7 @@ MODULE_DEVICE_TABLE(of, sunxi_rtc_dt_ids);
>   static int sunxi_rtc_probe(struct platform_device *pdev)
>   {
>   	struct sunxi_rtc_dev *chip;
> +	struct clk *extclk;
>   	int ret;
>   
>   	chip = devm_kzalloc(&pdev->dev, sizeof(*chip), GFP_KERNEL);
> @@ -455,6 +459,14 @@ static int sunxi_rtc_probe(struct platform_device *pdev)
>   		return -ENODEV;
>   	}
>   
> +	/* use external oscillator if present */
> +	extclk = devm_clk_get_optional_enabled(&pdev->dev, NULL);
> +	if (IS_ERR(extclk))
> +		return PTR_ERR(extclk);
> +	if (extclk)
> +		writel(SUNXI_LOSC_CTRL_KEY | SUNXI_LOSC_CTRL_OSC32K_SRC_SEL,
> +		       chip->base + SUNXI_LOSC_CTRL);

This should be a read-modify-write operation, since we don't want to 
disturb other bits in this register.

In general this looks OK to me, but would need to be documented in the 
bindings docs, to allow an optional clocks property.

Cheers,
Andre

> +
>   	/* clear the alarm count value */
>   	writel(0, chip->base + SUNXI_ALRM_DHMS);
>