On Sunday 20 March 2016 23:05:12 Christos Zoulas wrote:
> In article <1547312.w6y7ml9...@uberpc.marples.name>,
> Roy Marples <r...@marples.name> wrote:
>
> There is no need for pidfile_lock(), just fix pid_file() to return pid_t.
> I've audited the code in the tree and the code that checks, checks for -1.
> The compat code below is probably wrong anyway.
Ah, but it needs to check for != 0 as it can return the pid who has the lock,
which is itself an error.
>
> christos
>
> >+/* The old function.
> >+ * Historical behaviour is that pidfile is not re-written
> >+ * if path has not changed.
> >+ *
> >+ * Returns 0 on success, otherwise -1.
> >+ * As such we have no way of knowing the pid who owns the lock. */
> >
> > int
> > pidfile(const char *path)
> > {
> >
> >+ pid_t pid;
> >
> >- if (path == NULL || strchr(path, '/') == NULL) {
> >- char *default_path;
> >-
> >- if ((default_path = generate_varrun_path(path)) == NULL)
> >- return -1;
> >-
> >- if (create_pidfile(default_path) == -1) {
> >- free(default_path);
> >- return -1;
> >- }
> >-
> >- free(default_path);
> >- return 0;
> >- } else
> >- return create_pidfile(path);
> >+ pid = pidfile_lock(path);
> >+ return pid == 0 ? 0 : -1;
> >
> > }
> >
> >-=-=-=-=-=-
