On 7/28/2023 11:11 PM, Konrad Dybcio wrote:
> On 28.07.2023 15:23, Vikash Garodia wrote:
>> This implements common helper functions for v4l2 to vidc and
>> vice versa conversion for different enums.
>> Add helpers for state checks, buffer management, locks etc.
>>
>> Signed-off-by: Dikshita Agarwal <quic_dikshita@xxxxxxxxxxx>
>> Signed-off-by: Vikash Garodia <quic_vgarodia@xxxxxxxxxxx>
>> ---
> [...]
>
>> +
>> +#define is_odd(val) ((val) % 2 == 1)
>> +#define in_range(val, min, max) (((min) <= (val)) && ((val) <= (max)))
>> +#define COUNT_BITS(a, out) { \
> hweight.* functions?
>
> [...]
>
sure, will replace with hweight.
>> +
>> +const char *cap_name(enum msm_vidc_inst_capability_type cap_id)
>> +{
>> + const char *name = "UNKNOWN CAP";
> Perhaps it'd be worth to include the unknown cap id here
>
could you please elaborate more on this.
>> +
>> + if (cap_id >= ARRAY_SIZE(cap_name_arr))
>> + goto exit;
>> +
>> + name = cap_name_arr[cap_id];
>> +
>> +exit:
>> + return name;
>> +}
> [...]
>
>> +
>> +const char *buf_name(enum msm_vidc_buffer_type type)
>> +{
>> + const char *name = "UNKNOWN BUF";
> Similarly here
>
could you please elaborate more on this.
>> +
>> + if (type >= ARRAY_SIZE(buf_type_name_arr))
>> + goto exit;
>> +
>> + name = buf_type_name_arr[type];
>> +
>> +exit:
>> + return name;
>> +}
> [...]
>
>> +const char *v4l2_type_name(u32 port)
>> +{
>> + switch (port) {
> switch-case seems a bit excessive here.
>
>> + case INPUT_MPLANE: return "INPUT";
>> + case OUTPUT_MPLANE: return "OUTPUT";
>> + }
>> +
>> + return "UNKNOWN";
>> +}
that's right, will fix in next version
> [...]
>
> There's some more stuff I'd comment on, but 4500 lines in a single patch
> is way too much to logically follow.
>
> Couple more style suggestions:
> - use Reverse-Christmas-tree sorting for variable declarations
> - some oneliner functions could possibly become preprocessor macros
> - when printing giant debug messages, you may want to use loops
> - make sure your indentation is in order, 100 chars per line is
> totally fine
> - generally inline magic hex values are discouraged, but if they're
> necessary, the hex should be lowercase
Nice suggestions! will take care of these comments in next version.
Thanks,
Dikshita
>
> Konrad
