[PATCH 03/13] gpg: Limit the size of key packets to a sensible value.

[Daniel Kahn Gillmor]

From: Werner Koch <wk at gnupg.org>

* g10/parse-packet.c (MAX_KEY_PACKET_LENGTH): New.
(MAX_UID_PACKET_LENGTH): New.
(MAX_COMMENT_PACKET_LENGTH): New.
(MAX_ATTR_PACKET_LENGTH): New.
(parse_key): Limit the size of a key packet to 256k.
(parse_user_id): Use macro for the packet size limit.
(parse_attribute): Ditto.
(parse_comment): Ditto.
--

Without that it is possible to force gpg to allocate large amounts of
memory by using a bad encoded MPI.  This would be an too easy DoS.
Another way to mitigate would be to change the MPI read function to
allocate memory dynamically while reading the MPI.  However, that
complicates and possibly slows down the code.  A too large key packet
is in any case a sign for broken data and thus gpg should not use it.

Reported-by: Hanno Böck
GnuPG-bug-id: 1823
Signed-off-by: Werner Koch <wk at gnupg.org>

(back ported from commit 382ba4b137b42d5f25a7e256bb7c053ee5ac7b64)

[dkg: rebased to STABLE-BRANCH-1-4]
Signed-off-by: Daniel Kahn Gillmor <dkg at fifthhorseman.net>
---
 g10/parse-packet.c | 30 ++++++++++++++++++++++++++++--
 1 file changed, 28 insertions(+), 2 deletions(-)

diff --git a/g10/parse-packet.c b/g10/parse-packet.c
index e4e524c..c0b6ad6 100644
--- a/g10/parse-packet.c
+++ b/g10/parse-packet.c
@@ -40,6 +40,12 @@
 #define MAX_EXTERN_MPI_BITS 16384
 #endif
 
+/* Maximum length of packets to avoid excessive memory allocation.  */
+#define MAX_KEY_PACKET_LENGTH     (256 * 1024)
+#define MAX_UID_PACKET_LENGTH     (  2 * 1024)
+#define MAX_COMMENT_PACKET_LENGTH ( 64 * 1024)
+#define MAX_ATTR_PACKET_LENGTH    ( 16 * 1024*1024)
+
 
 static int mpi_print_mode;
 static int list_mode;
@@ -1663,6 +1669,13 @@ parse_key( IOBUF inp, int pkttype, unsigned long pktlen,
         rc = G10ERR_INVALID_PACKET;
 	goto leave;
     }
+    else if (pktlen > MAX_KEY_PACKET_LENGTH) {
+        log_error ("packet(%d) too large\n", pkttype);
+        if (list_mode)
+            fputs (":key packet: [too large]\n", listfp);
+        rc = G10ERR_INVALID_PACKET;
+        goto leave;
+    }
 
     timestamp = read_32(inp); pktlen -= 4;
     if( is_v4 ) {
@@ -2083,7 +2096,7 @@ parse_user_id( IOBUF inp, int pkttype, unsigned long pktlen, PACKET *packet )
        allocatable, and a very large pktlen could actually cause our
        allocation to wrap around in xmalloc to a small number. */
 
-    if(pktlen>2048)
+    if (pktlen > MAX_UID_PACKET_LENGTH)
       {
 	log_error("packet(%d) too large\n", pkttype);
 	iobuf_skip_rest(inp, pktlen, 0);
@@ -2152,6 +2165,19 @@ parse_attribute( IOBUF inp, int pkttype, unsigned long pktlen, PACKET *packet )
 {
     byte *p;
 
+    (void)pkttype;
+
+    /* We better cap the size of an attribute packet to make DoS not
+       too easy.  16MB should be more then enough for one attribute
+       packet (ie. a photo).  */
+    if (pktlen > MAX_ATTR_PACKET_LENGTH) {
+        log_error ("packet(%d) too large\n", pkttype);
+        if (list_mode)
+          fprintf (listfp, ":attribute packet: [too large]\n");
+        iobuf_skip_rest (inp, pktlen, 0);
+        return G10ERR_INVALID_PACKET;
+      }
+
 #define EXTRA_UID_NAME_SPACE 71
     packet->pkt.user_id = xmalloc_clear(sizeof *packet->pkt.user_id
 					+ EXTRA_UID_NAME_SPACE);
@@ -2186,7 +2212,7 @@ parse_comment( IOBUF inp, int pkttype, unsigned long pktlen, PACKET *packet )
        overflow in the malloc below.  Comment packets are actually not
        anymore define my OpenPGP and we even stopped to use our
        private comment packet. */
-    if (pktlen>65536)
+    if (pktlen > MAX_COMMENT_PACKET_LENGTH)
       {
 	log_error ("packet(%d) too large\n", pkttype);
 	iobuf_skip_rest (inp, pktlen, 0);
-- 
2.1.4