Re: [PATCH] libnfsidmap: id_as_chars() fails zero value ids.

[Steve Dickson <SteveD@redhat.com>]

On 04/30/2014 12:30 PM, Jeff Layton wrote:
> On Wed, 30 Apr 2014 11:18:28 -0400
> Steve Dickson <steved@redhat.com> wrote:
> 
>> Root has a zero value id which is valid and
>> should not be mapped to nfsnobody
>>
>> Signed-off-by: Steve Dickson <steved@redhat.com>
>> ---
>>  libnfsidmap.c | 8 ++++++--
>>  1 file changed, 6 insertions(+), 2 deletions(-)
>>
>> diff --git a/libnfsidmap.c b/libnfsidmap.c
>> index 641d766..92bc493 100644
>> --- a/libnfsidmap.c
>> +++ b/libnfsidmap.c
>> @@ -99,8 +99,12 @@ static char * toupper_str(char *s)
>>  static int id_as_chars(char *name, int *id)
>>  {
>>  	long int value = strtol(name, NULL, 10);
>> -	if (value == 0)
>> -		return 0;
>> +
>> +	if (value == 0) {
>> +		/* zero value ids are valid */
>> +		if (strcmp(name, "0") != 0)
>> +			return 0;
>> +	}
>>  	*id = (int)value;
>>  	return 1;
>>  }
> 
> Well spotted. I think though that instead of doing the strcmp, you
> should instead just use the endptr value in the strtol call to determine
> whether the string was completely converted. If it isn't, we can just
> return 0 here. If it is completely converted then you know that you got
> a legit "0" string.
> 
Yeah that wold have been a good idea but I've already
committed and pushed this version... ;-)

I'm thinking it's six of one and a half a dozen of another
both version will do the same thing... but if it bothersome, 
send me a patch....

steved.