From: Junio C Hamano Date: Tue, 09 Jan 2018 20:57:43 GMT Subject: Re: [PATCH v1 1/2] submodule: port submodule subcommand 'sync' from shell to C Message-ID: In-Reply-To: <20180109175703.4793-2-pc44800@gmail.com> Prathamesh Chavan writes: > +static int print_default_remote(int argc, const char **argv, const char *prefix) > +{ > + const char *remote; > + > + if (argc != 1) > + die(_("submodule--helper print-default-remote takes no arguments")); > + > + remote = get_default_remote(); > + if (remote) > + printf("%s\n", remote); > + > + return 0; > +} This is called directly from main and return immediately after printing, so a small leak of remote does not matter, I guess. > +static void sync_submodule(const char *path, const char *prefix, > + unsigned int flags) > +{ > + const struct submodule *sub; > + char *remote_key = NULL; > + char *sub_origin_url, *super_config_url, *displaypath; > + struct strbuf sb = STRBUF_INIT; > + struct child_process cp = CHILD_PROCESS_INIT; > + char *sub_config_path = NULL; > + > + if (!is_submodule_active(the_repository, path)) > + return; > + > + sub = submodule_from_path(&null_oid, path); > + > + if (sub && sub->url) { > + if (starts_with_dot_dot_slash(sub->url) || starts_with_dot_slash(sub->url)) { Not a big deal, but other codepaths seem to fold this pattern into two lines, i.e. if (starts_with_dot_dot_slash(sub->url) || starts_with_dot_slash(sub->url)) { > + sub_origin_url = relative_url(remote_url, sub->url, up_path); > + super_config_url = relative_url(remote_url, sub->url, NULL); On this side, these two are allocated memory that need to be freed. > + } else { > + sub_origin_url = xstrdup(sub->url); > + super_config_url = xstrdup(sub->url); This side as well. > + } > + } else { > + sub_origin_url = ""; > + super_config_url = ""; But not these. You have free() of these two at the end of this function, which will break things.