Fail properly when cloning from invalid HTTP URL

Currently, when cloning from invalid HTTP URL, git clone will possibly
return curl error, then a confusing message about remote HEAD and then
return success and leave an empty repository behind, confusing either
the end-user or the automated service calling it (think repo.or.cz).

This patch changes the error() calls in get_refs_via_curl() to die()s,
akin to the other get_refs_*() functions.

Cc: Daniel Barkalow <barkalow@iabervon.org>
Signed-off-by: Petr Baudis <pasky@suse.cz>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

transport.c

@@ -463,17 +463,14 @@ static struct ref *get_refs_via_curl(struct transport *transport)
 		run_active_slot(slot);
 		if (results.curl_result != CURLE_OK) {
 			strbuf_release(&buffer);
-			if (missing_target(&results)) {
-				return NULL;
-			} else {
-				error("%s", curl_errorstr);
-				return NULL;
-			}
+			if (missing_target(&results))
+				die("%s not found: did you run git update-server-info on the server?", refs_url);
+			else
+				die("%s download error - %s", refs_url, curl_errorstr);
 		}
 	} else {
 		strbuf_release(&buffer);
-		error("Unable to start request");
-		return NULL;
+		die("Unable to start HTTP request");
 	}
 
 	data = buffer.buf;