On 09/14/2018 11:12 AM, Phil Elwell wrote:
> On 14/09/2018 18:03, Florian Fainelli wrote:
>> On 09/14/2018 06:22 AM, Phil Elwell wrote:
>>> Use the compatible string in the DTB to select the correct cache line
>>> size for the SoC - 32 for BCM2835, and 64 for BCM2836 and BCM2837.
>>>
>>> Signed-off-by: Phil Elwell <phil@xxxxxxxxxxxxxxx>
>>> ---
>>
>> [snip]
>>
>>> @@ -170,6 +170,14 @@ static struct device *vchiq_dev;
>>> static DEFINE_SPINLOCK(msg_queue_spinlock);
>>> static struct platform_device *bcm2835_camera;
>>> +static struct vchiq_drvdata bcm2835_drvdata = {
>>> + .cache_line_size = 32,
>>> +};
>>> +
>>> +static struct vchiq_drvdata bcm2836_drvdata = {
>>> + .cache_line_size = 64,
>>> +};
>>
>> Those two structures could probably be marked const. Other than that,
>> the approach definitively looks good to me.
>
> The mutability is intentional - the structure pointer to the firmware is
> also
> stored there. This isn't the only piece of mutable static data in a driver
> that will only have a single instance, so allocating and initialising a
> per-instance structure seems needlessly complicated.
Fair enough, thanks for the explanation.
--
Florian
