TLS vulnerabilities

An issue was discovered in OpenSSH 7.9. Due to the scp implementation being derived from 1983 rcp, the server chooses which files/directories are sent to the client. However, the scp client only performs cursory validation of the object name returned (only directory traversal attacks are prevented). A malicious scp server (or Man-in-The-Middle attacker) can overwrite arbitrary files in the scp client target directory. If recursive operation (-r) is performed, the server can manipulate subdirectories as well (for example, to overwrite the .ssh/authorized_keys file).

In OpenSSH 7.9, due to accepting and displaying arbitrary stderr output from the server, a malicious server (or Man-in-The-Middle attacker) can manipulate the client output, for example to use ANSI control codes to hide additional files being transferred.

In OpenSSH 7.9, scp.c in the scp client allows remote SSH servers to bypass intended access restrictions via the filename of . or an empty filename. The impact is modifying the permissions of the target directory on the client side.

Remotely observable behaviour in auth-gss2.c in OpenSSH through 7.8 could be used by remote attackers to detect existence of users on a target system when GSS2 is in use. NOTE: the discoverer states 'We understand that the OpenSSH developers do not want to treat such a username enumeration (or "oracle") as a vulnerability.'

OpenSSH through 7.7 is prone to a user enumeration vulnerability due to not delaying bailout for an invalid authenticating user until after the packet containing the request has been fully parsed, related to auth2-gss.c, auth2-hostbased.c, and auth2-pubkey.c.

curl version curl 7.20.0 to and including curl 7.59.0 contains a CWE-126: Buffer Over-read vulnerability in denial of service that can result in curl can be tricked into reading data beyond the end of a heap based buffer used to store downloaded RTSP content.. This vulnerability appears to have been fixed in curl < 7.20.0 and curl >= 7.60.0.

The FTP wildcard function in curl and libcurl before 7.57.0 allows remote attackers to cause a denial of service (out-of-bounds read and application crash) or possibly have unspecified other impact via a string that ends with an '[' character.

The NTLM authentication feature in curl and libcurl before 7.57.0 on 32-bit platforms allows attackers to cause a denial of service (integer overflow and resultant buffer overflow, and application crash) or possibly have unspecified other impact via vectors involving long user and password fields.

An IMAP FETCH response line indicates the size of the returned data, in number of bytes. When that response says the data is zero bytes, libcurl would pass on that (non-existing) data with a pointer and the size (zero) to the deliver-data function. libcurl's deliver-data function treats zero as a magic number and invokes strlen() on the data to figure out the length. The strlen() is called on a heap based buffer that might not be zero terminated so libcurl might read beyond the end of it into whatever memory lies after (or just crash) and then deliver that to the application as if it was actually downloaded.

curl supports "globbing" of URLs, in which a user can pass a numerical range to have the tool iterate over those numbers to do a sequence of transfers. In the globbing function that parses the numerical range, there was an omission that made curl read a byte beyond the end of the URL if given a carefully crafted, or just wrongly written, URL. The URL is stored in a heap based buffer, so it could then be made to wrongly read something else instead of crashing. An example of a URL that triggers the flaw would be `http://ur%20[0-60000000000000000000`.

When doing a TFTP transfer and curl/libcurl is given a URL that contains a very long file name (longer than about 515 bytes), the file name is truncated to fit within the buffer boundaries, but the buffer size is still wrongly updated to use the untruncated length. This too large value is then used in the sendto() call, making curl attempt to send more data than what is actually put into the buffer. The endto() function will then read beyond the end of the heap based buffer. A malicious HTTP(S) server could redirect a vulnerable libcurl-using client to a crafted TFTP URL (if the client hasn't restricted which protocols it allows redirects to) and trick it to send private memory contents to a remote server over UDP. Limit curl's redirect protocols with --proto-redir and libcurl's with CURLOPT_REDIR_PROTOCOLS.

The ourWriteOut function in tool_writeout.c in curl 7.53.1 might allow physically proximate attackers to obtain sensitive information from process memory in opportunistic circumstances by reading a workstation screen during use of a --write-out argument ending in a '%' character, which leads to a heap-based buffer over-read.

The (1) TLS and (2) DTLS implementations in OpenSSL 1.0.1 before 1.0.1g do not properly handle Heartbeat Extension packets, which allows remote attackers to obtain sensitive information from process memory via crafted packets that trigger a buffer over-read, as demonstrated by reading private keys, related to d1_both.c and t1_lib.c, aka the Heartbleed bug.

Memory leak in the zlib_stateful_finish function in crypto/comp/c_zlib.c in OpenSSL 0.9.8l and earlier and 1.0.0 Beta through Beta 4 allows remote attackers to cause a denial of service (memory consumption) via vectors that trigger incorrect calls to the CRYPTO_cleanup_all_ex_data function, as demonstrated by use of SSLv3 and PHP with the Apache HTTP Server, a related issue to CVE-2008-1678.

The TLS protocol, and the SSL protocol 3.0 and possibly earlier, as used in Microsoft Internet Information Services (IIS) 7.0, mod_ssl in the Apache HTTP Server 2.2.14 and earlier, OpenSSL before 0.9.8l, GnuTLS 2.8.5 and earlier, Mozilla Network Security Services (NSS) 3.12.4 and earlier, multiple Cisco products, and other products, does not properly associate renegotiation handshakes with an existing connection, which allows man-in-the-middle attackers to insert data into HTTPS sessions, and possibly other types of sessions protected by TLS or SSL, by sending an unauthenticated request that is processed retroactively by a server in a post-renegotiation context, related to a "plaintext injection" attack, aka the "Project Mogul" issue.

A certain Red Hat modification to the ChrootDirectory feature in OpenSSH 4.8, as used in sshd in OpenSSH 4.3 in Red Hat Enterprise Linux (RHEL) 5.4 and Fedora 11, allows local users to gain privileges via hard links to setuid programs that use configuration files within the chroot directory, related to requirements for directory ownership.

lib/ssluse.c in cURL and libcurl 7.4 through 7.19.5, when OpenSSL is used, does not properly handle a '\0' character in a domain name in the subject's Common Name (CN) field of an X.509 certificate, which allows man-in-the-middle attackers to spoof arbitrary SSL servers via a crafted certificate issued by a legitimate Certification Authority, a related issue to CVE-2009-2408.

The Network Security Services (NSS) library before 3.12.3, as used in Firefox; GnuTLS before 2.6.4 and 2.7.4; OpenSSL 0.9.8 through 0.9.8k; and other products support MD2 with X.509 certificates, which might allow remote attackers to spoof certificates by using MD2 design flaws to generate a hash collision in less than brute-force time. NOTE: the scope of this issue is currently limited because the amount of computation required is still large.

The dtls1_retrieve_buffered_fragment function in ssl/d1_both.c in OpenSSL before 1.0.0 Beta 2 allows remote attackers to cause a denial of service (NULL pointer dereference and daemon crash) via an out-of-sequence DTLS handshake message, related to a "fragment bug."

ssl/s3_pkt.c in OpenSSL before 0.9.8i allows remote attackers to cause a denial of service (NULL pointer dereference and daemon crash) via a DTLS ChangeCipherSpec packet that occurs before ClientHello.

Use-after-free vulnerability in the dtls1_retrieve_buffered_fragment function in ssl/d1_both.c in OpenSSL 1.0.0 Beta 2 allows remote attackers to cause a denial of service (openssl s_client crash) and possibly have unspecified other impact via a DTLS packet, as demonstrated by a packet from a server that uses a crafted server certificate.

Multiple memory leaks in the dtls1_process_out_of_seq_message function in ssl/d1_both.c in OpenSSL 0.9.8k and earlier 0.9.8 versions allow remote attackers to cause a denial of service (memory consumption) via DTLS records that (1) are duplicates or (2) have sequence numbers much greater than current sequence numbers, aka "DTLS fragment handling memory leak."

The dtls1_buffer_record function in ssl/d1_pkt.c in OpenSSL 0.9.8k and earlier 0.9.8 versions allows remote attackers to cause a denial of service (memory consumption) via a large series of "future epoch" DTLS records that are buffered in a queue, aka "DTLS record buffer limitation bug."

OpenSSL before 0.9.8k on WIN64 and certain other platforms does not properly handle a malformed ASN.1 structure, which allows remote attackers to cause a denial of service (invalid memory access and application crash) by placing this structure in the public key of a certificate, as demonstrated by an RSA public key.

The CMS_verify function in OpenSSL 0.9.8h through 0.9.8j, when CMS is enabled, does not properly handle errors associated with malformed signed attributes, which allows remote attackers to repudiate a signature that originally appeared to be valid but was actually invalid.

The ASN1_STRING_print_ex function in OpenSSL before 0.9.8k allows remote attackers to cause a denial of service (invalid memory access and application crash) via vectors that trigger printing of a (1) BMPString or (2) UniversalString with an invalid encoded length.

The redirect implementation in curl and libcurl 5.11 through 7.19.3, when CURLOPT_FOLLOWLOCATION is enabled, accepts arbitrary Location values, which might allow remote HTTP servers to (1) trigger arbitrary requests to intranet servers, (2) read or overwrite arbitrary files via a redirect to a file: URL, or (3) execute arbitrary commands via a redirect to an scp: URL.

OpenSSL, probably 0.9.6, does not verify the Basic Constraints for an intermediate CA-signed certificate, which allows remote attackers to spoof the certificates of trusted sites via a man-in-the-middle attack, a related issue to CVE-2002-0970.

OpenSSL 0.9.8i and earlier does not properly check the return value from the EVP_VerifyFinal function, which allows remote attackers to bypass validation of the certificate chain via a malformed SSL/TLS signature for DSA and ECDSA keys.

Error handling in the SSH protocol in (1) SSH Tectia Client and Server and Connector 4.0 through 4.4.11, 5.0 through 5.2.4, and 5.3 through 5.3.8; Client and Server and ConnectSecure 6.0 through 6.0.4; Server for Linux on IBM System z 6.0.4; Server for IBM z/OS 5.5.1 and earlier, 6.0.0, and 6.0.1; and Client 4.0-J through 4.3.3-J and 4.0-K through 4.3.10-K; and (2) OpenSSH 4.7p1 and possibly other versions, when using a block cipher algorithm in Cipher Block Chaining (CBC) mode, makes it easier for remote attackers to recover certain plaintext data from an arbitrary block of ciphertext in an SSH session via unknown vectors.

A certain Debian patch for OpenSSH before 4.3p2-9etch3 on etch; before 4.6p1-1 on sid and lenny; and on other distributions such as SUSE uses functions that are not async-signal-safe in the signal handler for login timeouts, which allows remote attackers to cause a denial of service (connection slot exhaustion) via multiple login attempts. NOTE: this issue exists because of an incorrect fix for CVE-2006-5051.

Certain Red Hat Enterprise Linux (RHEL) 4 and 5 packages for OpenSSH, as signed in August 2008 using a legitimate Red Hat GPG key, contain an externally introduced modification (Trojan Horse) that allows the package authors to have an unknown impact. NOTE: since the malicious packages were not distributed from any official Red Hat sources, the scope of this issue is restricted to users who may have obtained these packages through unofficial distribution points. As of 20080827, no unofficial distributions of this software are known.

OpenSSH before 5.1 sets the SO_REUSEADDR socket option when the X11UseLocalhost configuration setting is disabled, which allows local users on some platforms to hijack the X11 forwarding port via a bind to a single IP address, as demonstrated on the HP-UX platform.

sshd in OpenSSH 4 on Debian GNU/Linux, and the 20070303 OpenSSH snapshot, allows remote authenticated users to obtain access to arbitrary SELinux roles by appending a :/ (colon slash) sequence, followed by the role name, to the username.

Memory leak in the zlib_stateful_init function in crypto/comp/c_zlib.c in libssl in OpenSSL 0.9.8f through 0.9.8h allows remote attackers to cause a denial of service (memory consumption) via multiple calls, as demonstrated by initial SSL client handshakes to the Apache HTTP Server mod_ssl that specify a compression algorithm.

OpenSSL 0.9.8f and 0.9.8g allows remote attackers to cause a denial of service (crash) via a TLS handshake that omits the Server Key Exchange message and uses "particular cipher suites," which triggers a NULL pointer dereference.

Double free vulnerability in OpenSSL 0.9.8f and 0.9.8g, when the TLS server name extensions are enabled, allows remote attackers to cause a denial of service (crash) via a malformed Client Hello packet. NOTE: some of these details are obtained from third party information.

OpenSSL 0.9.8c-1 up to versions before 0.9.8g-9 on Debian-based operating systems uses a random number generator that generates predictable numbers, which makes it easier for remote attackers to conduct brute force guessing attacks against cryptographic keys.

OpenSSH 4.4 up to versions before 4.9 allows remote authenticated users to bypass the sshd_config ForceCommand directive by modifying the .ssh/rc session file.

OpenSSH 4.3p2, and probably other versions, allows local users to hijack forwarded X connections by causing ssh to set DISPLAY to :10, even when another process is listening on the associated port, as demonstrated by opening TCP port 6010 (IPv4) and sniffing a cookie sent by Emacs.

Unspecified vulnerability in the linux_audit_record_event function in OpenSSH 4.3p2, as used on Fedora Core 6 and possibly other systems, allows remote attackers to write arbitrary characters to an audit log via a crafted username. NOTE: some of these details are obtained from third party information.

Off-by-one error in the DTLS implementation in OpenSSL 0.9.8 before 0.9.8f allows remote attackers to execute arbitrary code via unspecified vectors.

Off-by-one error in the SSL_get_shared_ciphers function in OpenSSL 0.9.7 up to 0.9.7l, and 0.9.8 up to 0.9.8f, might allow remote attackers to execute arbitrary code via a crafted packet that triggers a one-byte buffer underflow. NOTE: this issue was introduced as a result of a fix for CVE-2006-3738. As of 20071012, it is unknown whether code execution is possible.

ssh in OpenSSH before 4.7 does not properly handle when an untrusted cookie cannot be created and uses a trusted X11 cookie instead, which allows attackers to violate intended policy and gain privileges by causing an X client to be treated as trusted.

Unspecified vulnerability in SSHield 1.6.1 with OpenSSH 3.0.2p1 on Cisco WebNS 8.20.0.1 on Cisco Content Services Switch (CSS) series 11000 devices allows remote attackers to cause a denial of service (connection slot exhaustion and device crash) via a series of large packets designed to exploit the SSH CRC32 attack detection overflow (CVE-2001-0144), possibly a related issue to CVE-2002-1024.

The BN_from_montgomery function in crypto/bn/bn_mont.c in OpenSSL 0.9.8e and earlier does not properly perform Montgomery multiplication, which might allow local users to conduct a side-channel attack and retrieve RSA private keys.

libcurl 7.14.0 through 7.16.3, when built with GnuTLS support, does not check SSL/TLS certificate expiration or activation dates, which allows remote attackers to bypass certain access restrictions.

OpenSSH, when using OPIE (One-Time Passwords in Everything) for PAM, allows remote attackers to determine the existence of certain user accounts, which displays a different response if the user account exists and is configured to use one-time passwords (OTP), a similar issue to CVE-2007-2243.

OpenSSH 4.6 and earlier, when ChallengeResponseAuthentication is enabled, allows remote attackers to determine the existence of user accounts by attempting to authenticate via S/KEY, which displays a different response if the user account exists, a similar issue to CVE-2001-1483.

Unspecified vulnerability in the sshd Privilege Separation Monitor in OpenSSH before 4.5 causes weaker verification that authentication has been successful, which might allow attackers to bypass authentication. NOTE: as of 20061108, it is believed that this issue is only exploitable by leveraging vulnerabilities in the unprivileged process, which are not known to exist.