On Thu, Aug 14, 2014 at 01:28:15AM +0200, Lennart Poettering wrote: > On Wed, 13.08.14 16:35, Dave Reisner ([email protected]) wrote: > > Looks good. The code is certainly not any more complicated than the > current strapenda3(), so it sounds like something to apply. > > > /* If the passed init is actually the same as the > > * systemd binary, then let's suppress it. */ > > diff --git a/src/test/test-util.c b/src/test/test-util.c > > index 16f89b4..8776899 100644 > > --- a/src/test/test-util.c > > +++ b/src/test/test-util.c > > @@ -907,6 +907,12 @@ static void test_strshorten(void) { > > assert_se(strlen(strshorten(s, 0)) == 0); > > } > > > > +static void test_strappenda(void) { > > + assert_se(streq(strappenda("", "foo", "bar"), "foobar")); > > + assert_se(streq(strappenda("foo", "bar", "baz"), "foobarbaz")); > > + assert_se(streq(strappenda("foo", "", "bar", "baz"), "foobarbaz")); > > +} > > it's not portable to avoid alloca() when invoking a function (which the > strlen() in streq() is). This is documented in the alloca(3) man page, > see section "BUGS".
Ah, thanks! > Can you change the test to first place the result of strappenda() in a > variable, and then pass that on to streq()? > > Please commit then! Fixed up and pushed, thanks for the quick review! d _______________________________________________ systemd-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/systemd-devel
