Re: [PATCH_v2 2/4] anrdoid/health: Fix wrong channel type defines

[Ravi kumar Veeramally <ravikumar.veeramally@linux.intel.com>]

Hi Szymon,

On 06/17/2014 12:29 PM, Szymon Janc wrote:
> Hi Ravi,
>
> On Tuesday 17 of June 2014 01:02:48 Ravi kumar Veeramally wrote:
>> Enums in bt_hl.h are defined in this order BTHL_CHANNEL_TYPE_RELIABLE,
>> BTHL_CHANNEL_TYPE_STREAMING, BTHL_CHANNEL_TYPE_ANY. But HDP_SPEC_V11(3.4)
>> has no-preference(any)-0, reliable-1, stream-2, reserved(0x03-0xff).
>> Providing utility to solve this.
>> ---
>>   android/health.c | 18 +++++++++++++++++-
>>   1 file changed, 17 insertions(+), 1 deletion(-)
>>
>> diff --git a/android/health.c b/android/health.c
>> index 7e2c5d6..5a87089 100644
>> --- a/android/health.c
>> +++ b/android/health.c
>> @@ -52,6 +52,10 @@
>>   #define HDP_VERSION			0x0101
>>   #define DATA_EXCHANGE_SPEC_11073	0x01
>>   
>> +#define CHANNEL_TYPE_ANY       0x00
>> +#define CHANNEL_TYPE_RELIABLE  0x01
>> +#define CHANNEL_TYPE_STREAM    0x02
>> +
> Isn't that already defined in hdp_types.h ?
   Android is not using hdp_types.h .
>>   static bdaddr_t adapter_addr;
>>   static struct ipc *hal_ipc = NULL;
>>   static struct queue *apps = NULL;
>> @@ -639,6 +643,18 @@ fail:
>>   							HAL_STATUS_FAILED);
>>   }
>>   
>> +static uint8_t android2channel_type(uint8_t type)
>> +{
>> +	switch (type) {
>> +	case HAL_HEALTH_CHANNEL_TYPE_RELIABLE:
>> +		return CHANNEL_TYPE_RELIABLE;
>> +	case HAL_HEALTH_CHANNEL_TYPE_STREAMING:
>> +		return CHANNEL_TYPE_STREAM;
>> +	}
>> +
>> +	return CHANNEL_TYPE_ANY;
> Move this under default case.
   Ok.
>> +}
>> +
>>   static void bt_health_mdep_cfg_data(const void *buf, uint16_t len)
>>   {
>>   	const struct hal_cmd_health_mdep *cmd = buf;
>> @@ -662,7 +678,7 @@ static void bt_health_mdep_cfg_data(const void *buf, uint16_t len)
>>   
>>   	mdep->role = cmd->role;
>>   	mdep->data_type = cmd->data_type;
>> -	mdep->channel_type = cmd->channel_type;
>> +	mdep->channel_type = android2channel_type(cmd->channel_type);
>>   	mdep->id = queue_length(app->mdeps) + 1;
>>   
>>   	if (cmd->descr_len > 0) {
>>
Thanks,
Ravi.