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
