On Tue, Jul 31, 2012 at 6:08 PM, walter harms <[email protected]> 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;
> --
> 1.6.0.2

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

Reply via email to