Re: [PATCH 07/12] hwrng: bcm2835-rng: Manage an optional clock

[Florian Fainelli]

Hi Stefan,

On 11/04/2017 06:50 AM, Stefan Wahren wrote:
> Hi Florian,
> 
>> Florian Fainelli <f.fainelli@xxxxxxxxx> hat am 2. November 2017 um 02:04 geschrieben:
>>
>>
>> One of the last steps before bcm63xx-rng can be eliminated is to manage
>> a clock during hwrng::init and hwrng::cleanup, so fetch it in the probe
>> function, and manage it during these two steps when valid.
>>
>> Signed-off-by: Florian Fainelli <f.fainelli@xxxxxxxxx>
>> ---
>>  drivers/char/hw_random/bcm2835-rng.c | 14 ++++++++++++++
>>  1 file changed, 14 insertions(+)
>>
>> diff --git a/drivers/char/hw_random/bcm2835-rng.c b/drivers/char/hw_random/bcm2835-rng.c
>> index ed20e0b6b7ae..35928efb52e7 100644
>> --- a/drivers/char/hw_random/bcm2835-rng.c
>> +++ b/drivers/char/hw_random/bcm2835-rng.c
>> @@ -15,6 +15,7 @@
>>  #include <linux/of_platform.h>
>>  #include <linux/platform_device.h>
>>  #include <linux/printk.h>
>> +#include <linux/clk.h>
>>  
>>  #define RNG_CTRL	0x0
>>  #define RNG_STATUS	0x4
>> @@ -33,6 +34,7 @@ struct bcm2835_rng_priv {
>>  	struct hwrng rng;
>>  	void __iomem *base;
>>  	bool mask_interrupts;
>> +	struct clk *clk;
>>  };
>>  
>>  static inline struct bcm2835_rng_priv *to_rng_priv(struct hwrng *rng)
>> @@ -67,6 +69,11 @@ static int bcm2835_rng_init(struct hwrng *rng)
>>  {
>>  	struct bcm2835_rng_priv *priv = to_rng_priv(rng);
>>  	u32 val;
>> +	int ret;
>> +
>> +	ret = clk_prepare_enable(priv->clk);
>> +	if (ret)
>> +		return ret;
>>  
>>  	if (priv->mask_interrupts) {
>>  		/* mask the interrupt */
>> @@ -88,6 +95,8 @@ static void bcm2835_rng_cleanup(struct hwrng *rng)
>>  
>>  	/* disable rng hardware */
>>  	__raw_writel(0, priv->base + RNG_CTRL);
>> +
>> +	clk_disable_unprepare(priv->clk);
>>  }
>>  
>>  struct bcm2835_rng_of_data {
>> @@ -130,6 +139,11 @@ static int bcm2835_rng_probe(struct platform_device *pdev)
>>  		return PTR_ERR(priv->base);
>>  	}
>>  
>> +	/* Clock is optional on most platforms */
>> +	priv->clk = devm_clk_get(dev, NULL);
>> +	if (IS_ERR(priv->clk))
>> +		priv->clk = NULL;
> 
> at least in case of EPROBE_DEFERED this isn't the expected behavior. Maybe we should better trigger on non-existing clock?

Good point, so more like:

if (IS_ERR(priv->clk) && PTR_ERR(priv->clk) == -ENODEV)?

-- 
Florian