Am 31.07.2012 22:10, schrieb Tito:
> On Tuesday 31 July 2012 18:08:15 walter harms wrote:
>> compile tested,
>> refactor correct_password.c to avoid one #if
>>
>> Signed-off-by: wharms <[email protected]>
>> ---
>>  libbb/correct_password.c |    9 +++------
>>  1 files changed, 3 insertions(+), 6 deletions(-)
>>
>> diff --git a/libbb/correct_password.c b/libbb/correct_password.c
>> index 6301589..7cabd33 100644
>> --- a/libbb/correct_password.c
>> +++ b/libbb/correct_password.c
>> @@ -41,12 +41,6 @@ int FAST_FUNC correct_password(const struct passwd *pw)
>>      char *unencrypted, *encrypted;
>>      const char *correct;
>>      int r;
>> -#if ENABLE_FEATURE_SHADOWPASSWDS
>> -    /* Using _r function to avoid pulling in static buffers */
>> -    struct spwd spw;
>> -    char buffer[256];
>> -#endif
>> -
>>      /* fake salt. crypt() can choke otherwise. */
>>      correct = "aa";
>>      if (!pw) {
>> @@ -55,7 +49,10 @@ int FAST_FUNC correct_password(const struct passwd *pw)
>>      }
>>      correct = pw->pw_passwd;
>>  #if ENABLE_FEATURE_SHADOWPASSWDS
>> +    /* Using _r function to avoid pulling in static buffers */
>>      if ((correct[0] == 'x' || correct[0] == '*') && !correct[1]) {
>> +            struct spwd spw;
>> +            char buffer[256];
>>              /* getspnam_r may return 0 yet set result to NULL.
>>               * At least glibc 2.4 does this. Be extra paranoid here. */
>>              struct spwd *result = NULL;
>>
> 
> Hi,
> I personally I'm not for mixing variable declarations and code,
> I admit tough that in this case as it are just a few lines it could be ok.
> 

me too,
actualy i was thinking to move that all into a special function but ....
maybe you will ?

re,
 wh


_______________________________________________
busybox mailing list
[email protected]
http://lists.busybox.net/mailman/listinfo/busybox

Reply via email to