2014-10-01 14:36 GMT+03:00 Ryo Munakata <[email protected]>: > Signed-off-by: Ryo Munakata <[email protected]> > --- > src/connection.c | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/src/connection.c b/src/connection.c > index f292853..424edd4 100644 > --- a/src/connection.c > +++ b/src/connection.c > @@ -931,6 +931,11 @@ wl_closure_invoke(struct wl_closure *closure, uint32_t > flags, > count + 2, &ffi_type_void, ffi_types); > > implementation = target->implementation; > + if (!implementation[opcode]) { > + wl_log("listener function for opcode %u of %s is NULL\n", > + opcode, target->interface->name); > + assert(implementation[opcode] != NULL);
Isn't it better to call abort() then, instead of assert? With the assert release builds will continue to segfault as before, and debug builds will check something which we already checked before. > + } > ffi_call(&cif, implementation[opcode], NULL, ffi_args); > } > > -- > 2.1.2 > > _______________________________________________ > wayland-devel mailing list > [email protected] > http://lists.freedesktop.org/mailman/listinfo/wayland-devel _______________________________________________ wayland-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/wayland-devel
