Re: [PATCH] rds: send: mark expected switch fall-through in rds_rm_size

["Gustavo A. R. Silva" <garsilva@embeddedor.com>]

Hi Santosh,

On 02/20/2018 11:54 AM, Santosh Shilimkar wrote:
> Hi,
> 
> 2/19/2018 10:10 AM, Gustavo A. R. Silva wrote:
>> In preparation to enabling -Wimplicit-fallthrough, mark switch cases
>> where we are expecting to fall through.
>>
>> Addresses-Coverity-ID: 1465362 ("Missing break in switch")
>> Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
>> ---
>>   net/rds/send.c | 2 ++
>>   1 file changed, 2 insertions(+)
>>
>> diff --git a/net/rds/send.c b/net/rds/send.c
>> index 028ab59..79d158b 100644
>> --- a/net/rds/send.c
>> +++ b/net/rds/send.c
>> @@ -902,6 +902,8 @@ static int rds_rm_size(struct msghdr *msg, int 
>> num_sgs)
>>           case RDS_CMSG_ZCOPY_COOKIE:
>>               zcopy_cookie = true;
>> +            /* fall through */
>> +
>>           case RDS_CMSG_RDMA_DEST:
>>           case RDS_CMSG_RDMA_MAP:
>>               cmsg_groups |= 2;
>>
> So coverity greps for commet as "fall through" for
> -Wimplicit-fallthrough build ?
> 

No. Basically, Coverity only reports cases in which a break, return or 
continue statement is missing.

Now, if the statements I mention above are missing and if you add the 
following line to your Makefile:

KBUILD_CFLAGS  += $(call cc-option,-Wimplicit-fallthrough)

You will get a warning if a fall-through comment is missing.

> Adding that comments itself if fine but was curious
> about it if some one makes a spell error in this
> comment what happens ;-)
> 

In this case, Coverity would still report the same "Missing break in 
switch" error, but you'll get a GCC warning.

> For patch itself,
> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com>

--
Gustavo