[PATCH] Don't reject name server due to empty answer

[PATCH] Don't reject name server due to empty answer

[Andreas Schwab]

The resolver rejects a name server when it responds with an empty reply
(both AA and RA flags cleared) like this:

$ dig aaaa @134.178.14.1 www.bom.gov.au.

; <<>> DiG 9.6.1-P3-RedHat-9.6.1-16.P3.fc12 <<>> aaaa @134.178.14.1 www.bom.gov.au.
; (1 server found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 58652
;; flags: qr; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;www.bom.gov.au.                        IN      AAAA

This only happens when using UDP, the TCP query code happily accepts
that as a negative reply.

Andreas.

* resolv/res_send.c (send_dg): Don't reject name server due to
empty answer.
---
 resolv/res_send.c |    9 ---------
 1 files changed, 0 insertions(+), 9 deletions(-)

diff --git a/resolv/res_send.c b/resolv/res_send.c
index 845b658..d164b55 100644
--- a/resolv/res_send.c
+++ b/resolv/res_send.c
@@ -1294,15 +1294,6 @@ send_dg(res_state statp,
 			if (!statp->pfcode)
 				return (0);
 		}
-		if (anhp->rcode == NOERROR && anhp->ancount == 0
-		    && anhp->aa == 0 && anhp->ra == 0 && anhp->arcount == 0) {
-			DprintQ(statp->options & RES_DEBUG,
-				(stdout, "referred query:\n"),
-				*thisansp,
-				(*thisresplenp > *thisanssizp)
-				? *thisanssizp : *thisresplenp);
-			goto next_ns;
-		}
 		if (!(statp->options & RES_IGNTC) && anhp->tc) {
 			/*
 			 * To get the rest of answer,
-- 
1.7.0.1


-- 
Andreas Schwab, schwab@redhat.com
GPG Key fingerprint = D4E8 DBE3 3813 BB5D FA84  5EC7 45C6 250E 6F00 984E
"And now for something completely different."

Re: [PATCH] Don't reject name server due to empty answer

[Ulrich Drepper]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 03/16/2010 10:17 AM, Andreas Schwab wrote:
> The resolver rejects a name server when it responds with an empty reply
> (both AA and RA flags cleared) like this:

This code was added for a reason:

https://bugzilla.redhat.com/show_bug.cgi?id=162625


Why should this be reverted?  Yes, the TCP code should likely be fixed
up as well but that's a different story.

- -- 
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAkug5SwACgkQ2ijCOnn/RHRvPACeKkzg3lXy7czgzn7M24IoEAdu
6iIAn3THopqQmmooHpKZVvjeGaklSqqE
=5G4u
-----END PGP SIGNATURE-----