Re: [PATCH] query_krb5_ccache: Removed dead code that was flagged by a covscan

[Steve Dickson <SteveD@RedHat.com>]

On 2/7/20 12:25 PM, Olga Kornievskaia wrote:
> On Fri, Feb 7, 2020 at 10:22 AM Steve Dickson <steved@redhat.com> wrote:
>>
>> Signed-off-by: Steve Dickson <steved@redhat.com>
>> ---
>>  utils/gssd/krb5_util.c | 2 --
>>  1 file changed, 2 deletions(-)
>>
>> diff --git a/utils/gssd/krb5_util.c b/utils/gssd/krb5_util.c
>> index bff759f..a1c43d2 100644
>> --- a/utils/gssd/krb5_util.c
>> +++ b/utils/gssd/krb5_util.c
>> @@ -1066,8 +1066,6 @@ query_krb5_ccache(const char* cred_cache, char **ret_princname,
>>                             *ret_realm = strdup(str+1);
>>                     }
>>                     k5_free_unparsed_name(context, princstring);
>> -               } else {
>> -                       found = 0;
>>                 }
> 
> Uhm, sorry wasn't fast enough for you commit decision but I don't see
> that this a dead code? krb5_unparse_string() could return an error so
> "else" is a valid condition. I mean it's probably unlikely that
> check_for_tgt() returns found and they you can't parse the principal
> name out of it. But things like memory errors could still be valid
> error conditions?
Sorry for being so quick with the commit... 

The covscan complained "warning: Value stored to 'found' is never read"
which was true... after setting found = 0, found was never used. 

Yes, the "else" is a valid condition but not necessary since
setting 'found' to zero does not do anything... 

steved.

> 
> 
>>         }
>>         krb5_free_principal(context, principal);
>> --
>> 2.24.1
>>
>