fix an out-of-bound access in start_cgi

Long time ago, client->req was a static buffer so the memcpy was safe.
However, it's been since moved to a dynamically allocated string, so
it's very often smaller than sizeof(req.buf) (1024), hence the out of
bound access which results in a SIGSEGV very often on OpenBSD thanks to
Otto' malloc.

The situation with the iri parser, client->req and how the request is
forwarded to the other process needs to be improved: this is just a fix
to address the issue quickly, a better one would be to restructure the
iri parser APIs and rethink how the info is forwarded to the ex process.

gmid.h

@@ -229,6 +229,7 @@ struct client {
 	uint32_t	 id;
 	struct tls	*ctx;
 	char		*req;
+	size_t		 reqlen;
 	struct iri	 iri;
 	char		 domain[DOMAIN_NAME_LEN];
 

server.c

@@ -743,7 +743,7 @@ start_cgi(const char *spath, const char *relpath, struct client *c)
 
 	memset(&req, 0, sizeof(req));
 
-	memcpy(req.buf, c->req, sizeof(req.buf));
+	memcpy(req.buf, c->req, c->reqlen);
 
 	req.iri_schema_off = c->iri.schema - c->req;
 	req.iri_host_off = c->iri.host - c->req;
@@ -1024,6 +1024,12 @@ client_read(struct bufferevent *bev, void *d)
 		bufferevent_enable(bev, EVBUFFER_READ);
 		return;
 	}
+	c->reqlen = strlen(c->req);
+	if (c->reqlen > 1024+2) {
+		log_err(c, "URL too long");
+		start_reply(c, BAD_REQUEST, "bad request");
+		return;
+	}
 
 	if (!parse_iri(c->req, &c->iri, &parse_err) ||
 	    !puny_decode(c->iri.host, decoded, sizeof(decoded), &parse_err)) {