Re: [PATCH] ath11k: fix fallthrough warnings

[Kalle Valo <kvalo@codeaurora.org>]

msinada@codeaurora.org writes:

> On 2019-06-06 08:13, Kalle Valo wrote:
>> Muna Sinada <msinada@codeaurora.org> writes:
>>
>>> Marked expected switch case fall-through.
>>>
>>> Fall through warnings were generated when running ath11k-check script
>>> with --extra flag:
>>>
>>>     drivers/net/wireless/ath/ath11k/core.c:741:14:
>>> 	warning: this statement may fall through Wimplicit-fallthrough=]
>>>     drivers/net/wireless/ath/ath11k/dp.c:133:6:
>>> 	warning: this statement may fall through [-Wplicit-fallthrough=]
>>>
>>> Signed-off-by: Muna Sinada <msinada@codeaurora.org>
>>
>> [...]
>>
>>> --- a/drivers/net/wireless/ath/ath11k/dp.c
>>> +++ b/drivers/net/wireless/ath/ath11k/dp.c
>>> @@ -138,6 +138,7 @@ int ath11k_dp_srng_setup(struct ath11k_base
>>> *ab, struct dp_srng *ring,
>>>  			break;
>>>  		}
>>>  		/* follow through when ring_num >= 3 */
>>> +		/* fall through */
>>
>> Duplicate comments saying the same thing, can't you merge those?
>
> ath11k_check did not recognize "fall through" in the first comment
> causing the initial fall through warning. It seems that fall through
> comment needs to be on a separate comment to avoid getting warning
> from ath11k_check. Should I get rid of first comment or leave things
> how they are now?

Ah, I tested this myself now and indeed GCC does not regocnise '/* fall
through when ring_num >= 3 */' as a valid fallthrough comment even the
recommendation implied otherwise:

https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html

Oh well, so I guess we just need to have two comments even it's IMHO
ugly. Unless someone has other ideas?

-- 
Kalle Valo

_______________________________________________
ath11k mailing list
ath11k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath11k