This report was co-authored by @timwolla
Summary
The random byte generation function used in the SOAP HTTP Digest authentication code is not checked for failure. This can result in a stack information leak. Furthermore, there's an insufficient number of random bytes used.
Details
If php_random_bytes_throw fails, the nonce will be uninitialized, but
still sent to the server. The client nonce is intended to protect
against a malicious server. See section 5.10 and 5.12 of RFC 7616,
and bullet point 2 below.
Tim pointed out that even though it's the MD5 of the nonce that gets sent, enumerating 31 bits is trivial. So we have still a stack information leak of 31 bits.
Furthermore, Tim found the following issues:
-
The small size of cnonce might cause the server to erroneously reject a request due to a repeated (cnonce, nc) pair. As per the birthday problem 31 bits of randomness will return a duplication with 50% chance after less than 55000 requests and nc always starts counting at 1.
-
The cnonce is intended to protect the client and password against a malicious server that returns a constant server nonce where the server precomputed a rainbow table between passwords and correct client response. As storage is fairly cheap, a server could precompute the client responses for (a subset of) client nonces and still have a chance of reversing the client response with the same probability as the cnonce duplication.
Precomputing the rainbow table for all 2^31 cnonces increases the rainbow table size by factor 2 billion, which is infeasible. But precomputing it for 2^14 cnonces only increases the table size by factor 16k and the server would still have a 10% chance of successfully reversing a password with a single client request.
PoC
We do not have a proof-of-concept.
Impact
The weak randomness affects applications that use SOAP with HTTP Digest authentication against a possibly malicious server over HTTP. The stack information leak applies to both HTTP and HTTPS, but is only a few bytes.
Proposed patch
This patch fixes the issues by increasing the nonce size, and checking the return value of php_random_bytes_throw(). In the process we also get rid of the MD5 hashing of the nonce.
From f26fcd3a152a5c6b2d140cb35a5040671696f6e1 Mon Sep 17 00:00:00 2001
From: Niels Dossche <7771979+nielsdos@users.noreply.github.com>
Date: Sun, 16 Apr 2023 15:05:03 +0200
Subject: [PATCH] Fix missing randomness check and insufficient random bytes
for SOAP HTTP Digest
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
If php_random_bytes_throw fails, the nonce will be uninitialized, but
still sent to the server. The client nonce is intended to protect
against a malicious server. See section 5.10 and 5.12 of RFC 7616 [1],
and bullet point 2 below.
Tim pointed out that even though it's the MD5 of the nonce that gets sent,
enumerating 31 bits is trivial. So we have still a stack information leak
of 31 bits.
Furthermore, Tim found the following issues:
* The small size of cnonce might cause the server to erroneously reject
a request due to a repeated (cnonce, nc) pair. As per the birthday
problem 31 bits of randomness will return a duplication with 50%
chance after less than 55000 requests and nc always starts counting at 1.
* The cnonce is intended to protect the client and password against a
malicious server that returns a constant server nonce where the server
precomputed a rainbow table between passwords and correct client response.
As storage is fairly cheap, a server could precompute the client responses
for (a subset of) client nonces and still have a chance of reversing the
client response with the same probability as the cnonce duplication.
Precomputing the rainbow table for all 2^31 cnonces increases the rainbow
table size by factor 2 billion, which is infeasible. But precomputing it
for 2^14 cnonces only increases the table size by factor 16k and the server
would still have a 10% chance of successfully reversing a password with a
single client request.
This patch fixes the issues by increasing the nonce size, and checking
the return value of php_random_bytes_throw(). In the process we also get
rid of the MD5 hashing of the nonce.
[1] RFC 7616: https://www.rfc-editor.org/rfc/rfc7616
Co-authored-by: Tim Düsterhus ---
ext/soap/php_http.c | 21 +++++++++++++--------
1 file changed, 13 insertions(+), 8 deletions(-)
diff --git a/ext/soap/php_http.c b/ext/soap/php_http.c
index db3c97b647..a619949c69 100644
--- a/ext/soap/php_http.c
+++ b/ext/soap/php_http.c
@@ -657,18 +657,23 @@ try_again:
has_authorization = 1;
if (Z_TYPE_P(digest) == IS_ARRAY) {
char HA1[33], HA2[33], response[33], cnonce[33], nc[9];
- zend_long nonce;
+ unsigned char nonce[16];
PHP_MD5_CTX md5ctx;
unsigned char hash[16];
- php_random_bytes_throw(&nonce, sizeof(nonce));
- nonce &= 0x7fffffff;
+ if (UNEXPECTED(php_random_bytes_throw(&nonce, sizeof(nonce)) != SUCCESS)) {
+ ZEND_ASSERT(EG(exception));
+ php_stream_close(stream);
+ convert_to_null(Z_CLIENT_HTTPURL_P(this_ptr));
+ convert_to_null(Z_CLIENT_HTTPSOCKET_P(this_ptr));
+ convert_to_null(Z_CLIENT_USE_PROXY_P(this_ptr));
+ smart_str_free(&soap_headers_z);
+ smart_str_free(&soap_headers);
+ return FALSE;
+ }
- PHP_MD5Init(&md5ctx);
- snprintf(cnonce, sizeof(cnonce), ZEND_LONG_FMT, nonce);
- PHP_MD5Update(&md5ctx, (unsigned char*)cnonce, strlen(cnonce));
- PHP_MD5Final(hash, &md5ctx);
- make_digest(cnonce, hash);
+ php_hash_bin2hex(cnonce, nonce, sizeof(nonce));
+ cnonce[32] = 0;
if ((tmp = zend_hash_str_find(Z_ARRVAL_P(digest), "nc", sizeof("nc")-1)) != NULL &&
Z_TYPE_P(tmp) == IS_LONG) {
--
2.40.0