[PATCH 04/13] scpi: Add legacy SCP functions calling legacy_scpi_send_message
Sudeep Holla
sudeep.holla at arm.com
Tue Aug 23 07:47:02 PDT 2016
On 23/08/16 09:19, Neil Armstrong wrote:
> On 08/19/2016 06:22 PM, Sudeep Holla wrote:
>>
>>
>> On 18/08/16 11:10, Neil Armstrong wrote:
>>> In order to support legacy SCP functions from kernel-wide driver, add legacy
>>> functions using the legacy command enums and calling legacy_scpi_send_message.
>>>
>>> Signed-off-by: Neil Armstrong <narmstrong at baylibre.com>
>>> ---
>>> drivers/firmware/arm_scpi.c | 118 ++++++++++++++++++++++++++++++++++++++++++++
>>> 1 file changed, 118 insertions(+)
>>>
>>> diff --git a/drivers/firmware/arm_scpi.c b/drivers/firmware/arm_scpi.c
>>> index 50b1297..bb9965f 100644
>>> --- a/drivers/firmware/arm_scpi.c
>>> +++ b/drivers/firmware/arm_scpi.c
>>> @@ -578,6 +578,8 @@ scpi_clk_get_range(u16 clk_id, unsigned long *min, unsigned long *max)
>>> return ret;
>>> }
>>>
>>> +/* scpi_clk_get_range not available for legacy */
>>> +
>>> static unsigned long scpi_clk_get_val(u16 clk_id)
>>> {
>>> int ret;
>>> @@ -589,6 +591,18 @@ static unsigned long scpi_clk_get_val(u16 clk_id)
>>> return ret ? ret : le32_to_cpu(clk.rate);
>>> }
>>>
>>> +static unsigned long legacy_scpi_clk_get_val(u16 clk_id)
>>> +{
>>> + int ret;
>>> + struct clk_get_value clk;
>>> + __le16 le_clk_id = cpu_to_le16(clk_id);
>>> +
>>> + ret = legacy_scpi_send_message(LEGACY_SCPI_CMD_GET_CLOCK_VALUE,
>>> + &le_clk_id, sizeof(le_clk_id),
>>> + &clk, sizeof(clk));
>>> + return ret ? ret : le32_to_cpu(clk.rate);
>>> +}
>>> +
>>> static int scpi_clk_set_val(u16 clk_id, unsigned long rate)
>>> {
>>> int stat;
>>> @@ -601,6 +615,19 @@ static int scpi_clk_set_val(u16 clk_id, unsigned long rate)
>>> &stat, sizeof(stat));
>>> }
>>>
>>> +static int legacy_scpi_clk_set_val(u16 clk_id, unsigned long rate)
>>> +{
>>> + int stat;
>>> + struct legacy_clk_set_value clk = {
>>> + .id = cpu_to_le16(clk_id),
>>> + .rate = cpu_to_le32(rate)
>>> + };
>>> +
>>> + return legacy_scpi_send_message(LEGACY_SCPI_CMD_SET_CLOCK_VALUE,
>>> + &clk, sizeof(clk),
>>> + &stat, sizeof(stat));
>>
>> Except this one which has a different structure format, why do we need
>> to define legacy versions of other functions ? Can't we play with
>> function pointer or have a boolean in drvinfo structure and use then in
>> the existing functions as I had shown in one of the earlier emails.
>>
>
> The main problem is that the command indexes deviates starting at
> SCPI_CMD_SET_CSS_PWR_STATE, I'll be pleased to know how to implement it.
>
Yes, I was thinking of some kind of mapping to new index using an array.
--
Regards,
Sudeep
More information about the linux-amlogic
mailing list