On Thu, 20.02.14 21:11, Daniel Buch (boogiewasth...@gmail.com) wrote:
>
> log_debug("Looking up address info for %s:%s", node,
> service);
> - r = getaddrinfo(node, service, &hints, &result);
> - if (r != 0) {
> + r = sd_resolve_getaddrinfo(resolve, &q, node, service,
> &hints);
> + if (r < 0)
> + log_error("Error: %s %d\n", gai_strerror(r), r);
> +
> + while (!sd_resolve_is_done(q)) {
> + r = sd_resolve_wait(resolve, (uint64_t) -1);
> + if (r < 0) {
> + log_error("Error: %s\n", strerror(-r));
> + assert_not_reached("sd_resolve_wait()
> failed");
> + }
> + }
> +
> + r = sd_resolve_getaddrinfo_done(q, &result);
> + if (r < 0) {
> log_error("Failed to resolve host %s:%s: %s", node,
> service, gai_strerror(r));
> return -EHOSTUNREACH;
> }
Yupp, as David already pointed out, this takes the async API and makes
it sync again...
I figure before we can implement this part in the socket proxy we need
to enhance sd-resolve to provide hooks into the event loop. And that
probably means we also need to teach it callbacks...
So I figure this is a bit more difficult to do...
Lennart
--
Lennart Poettering, Red Hat
_______________________________________________
systemd-devel mailing list
systemd-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/systemd-devel
