Summary

Terrapin is a prefix truncation attack targeting the SSH protocol. More precisely, Terrapin breaks the integrity of SSH's secure channel. By carefully adjusting the sequence numbers during the handshake, an attacker can remove an arbitrary amount of messages sent by the client or server at the beginning of the secure channel without the client or server noticing it.

Mitigations

To mitigate this protocol vulnerability, OpenSSH suggested a so-called "strict kex" which alters the SSH handshake to ensure a Man-in-the-Middle attacker cannot introduce unauthenticated messages as well as convey sequence number manipulation across handshakes.

Warning: To take effect, both the client and server must support this countermeasure.

As a stop-gap measure, peers may also (temporarily) disable the affected algorithms and use unaffected alternatives like AES-GCM instead until patches are available.

Details

The SSH specifications of ChaCha20-Poly1305 (chacha20-poly1305@openssh.com) and Encrypt-then-MAC (-etm@openssh.com MACs) are vulnerable against an arbitrary prefix truncation attack (a.k.a. Terrapin attack). This allows for an extension negotiation downgrade by stripping the SSHMSGEXTINFO sent after the first message after SSHMSGNEWKEYS, downgrading security, and disabling attack countermeasures in some versions of OpenSSH. When targeting Encrypt-then-MAC, this attack requires the use of a CBC cipher to be practically exploitable due to the internal workings of the cipher mode. Additionally, this novel attack technique can be used to exploit previously unexploitable implementation flaws in a Man-in-the-Middle scenario.

The attack works by an attacker injecting an arbitrary number of SSHMSGIGNORE messages during the initial key exchange and consequently removing the same number of messages just after the initial key exchange has concluded. This is possible due to missing authentication of the excess SSHMSGIGNORE messages and the fact that the implicit sequence numbers used within the SSH protocol are only checked after the initial key exchange.

In the case of ChaCha20-Poly1305, the attack is guaranteed to work on every connection as this cipher does not maintain an internal state other than the message's sequence number. In the case of Encrypt-Then-MAC, practical exploitation requires the use of a CBC cipher; while theoretical integrity is broken for all ciphers when using this mode, message processing will fail at the application layer for CTR and stream ciphers.

For more details see https://terrapin-attack.com.

Impact

This attack targets the specification of ChaCha20-Poly1305 (chacha20-poly1305@openssh.com) and Encrypt-then-MAC (-etm@openssh.com), which are widely adopted by well-known SSH implementations and can be considered de-facto standard. These algorithms can be practically exploited; however, in the case of Encrypt-Then-MAC, we additionally require the use of a CBC cipher. As a consequence, this attack works against all well-behaving SSH implementations supporting either of those algorithms and can be used to downgrade (but not fully strip) connection security in case SSH extension negotiation (RFC8308) is supported. The attack may also enable attackers to exploit certain implementation flaws in a man-in-the-middle (MitM) scenario.

A missing bounds check was found in the way OpenSSL handled TLS heartbeat extension packets. This flaw could be used to reveal up to 64k of memory from a connected client or server.

Only 1.0.1 releases of OpenSSL are affected including 1.0.1f (and 1.0.2 betas)

The following upstream commit introduced TLS/DTLS heatbeat support and also this issue:

http://git.openssl.org/gitweb/?p=openssl.git;a=commitdiff;h=4817504

FileZilla 3.40.0 contains a denial of service vulnerability in the local search functionality that allows local attackers to crash the application by supplying a malformed path string. Attackers can trigger the crash by entering a crafted path containing 384 'A' characters followed by 'BBBB' and 'CCCC' sequences in the search directory field and initiating a local search operation.

Buffer overflow in FileZilla Server before 0.9.31 allows remote attackers to cause a denial of service via unspecified vectors related to SSL/TLS packets.

It was found that OpenSSL was vulnerable to a SSL/TLS MITM vulnerability. An attacker using a carefully crafted handshake can force the use of weak keying material in OpenSSL SSL/TLS clients and servers. This can be exploited by a Man-in-the-middle (MITM) attack where the attacker can decrypt and modify traffic from the attacked client and server.

As per the upstream advisory:

The attack can only be performed between a vulnerable client and server. OpenSSL clients are vulnerable in all versions of OpenSSL. Servers are only known to be vulnerable in OpenSSL 1.0.1 and 1.0.2-beta1. Users of OpenSSL servers earlier than 1.0.1 are advised to upgrade as a precaution.

OpenSSL 0.9.8 SSL/TLS users (client and/or server) should upgrade to 0.9.8za. OpenSSL 1.0.0 SSL/TLS users (client and/or server) should upgrade to 1.0.0m. OpenSSL 1.0.1 SSL/TLS users (client and/or server) should upgrade to 1.0.1h.

Acknowledgements:

Red Hat would like to thank the OpenSSL project for reporting this issue. Upstream acknowledges KIKUCHI Masashi of Lepidum as the original reporter of this issue.

FileZilla Server before 0.9.22 allows remote attackers to cause a denial of service (crash) via a wildcard argument to the (1) LIST or (2) NLST commands, which results in a NULL pointer dereference, a different set of vectors than CVE-2006-6564. NOTE: CVE analysis suggests that the problem might be due to a malformed PORT command.

FileZilla Client 3.63.1 contains a DLL hijacking vulnerability that allows attackers to execute malicious code by placing a crafted TextShaping.dll in the application directory. Attackers can generate a reverse shell payload using msfvenom and replace the missing DLL to achieve remote code execution when the application launches.

In PuTTY 0.68 through 0.80 before 0.81, biased ECDSA nonce generation allows an attacker to recover a user's NIST P-521 secret key via a quick attack in approximately 60 signatures. This is especially important in a scenario where an adversary is able to read messages signed by PuTTY or Pageant. One scenario is that the adversary is an operator of an SSH server to which the victim authenticates (for remote login or file copy), even though this server is not fully trusted by the victim, and the victim uses the same private key for SSH connections to other services operated by other entities. Here, the rogue server operator (who would otherwise have no way to determine the victim's private key) can derive the victim's private key, and then use it for unauthorized access to those other services. Because SSH is sometimes used to authenticate to Git services, it is possible that this vulnerability could be leveraged for supply-chain attacks on software maintained in Git. It is also conceivable that signed messages from PuTTY or Pageant are readable by adversaries more easily in other scenarios, but none have yet been disclosed.

Affected Products

- PuTTY 0.68 - 0.80

The following (not necessarily complete) list of products bundle an affected PuTTY version and are therefore vulnerable as well:

- FileZilla 3.24.1 - 3.66.5 - WinSCP 5.9.5 - 6.3.2 - TortoiseGit 2.4.0.2 - 2.15.0 - TortoiseSVN 1.10.0 - 1.14.6

Impact

The nonce bias allows for full secret key recovery of NIST P-521 keys after a malicious actor has seen roughly 60 valid ECDSA signatures generated by any PuTTY component under the same key. Luckily, client signatures are transmitted within the secure channel of SSH, requiring a malicious server to acquire such signatures. If the key has been used to sign arbitrary data (e.g., git commits by forwarding Pageant to a development host), the publicly available signatures (e.g., on GitHub) can be used as well.

All NIST P-521 client keys used with PuTTY must be considered compromised, given that the attack can be carried out even after the root cause has been fixed in the source code (assuming that ~60 pre-patch signatures are available to an adversary).

Mitigations

This vulnerability has been fixed in PuTTY 0.81, FileZilla 3.67.0, WinSCP 6.3.3, and TortoiseGit 2.15.0.1. Users of TortoiseSVN are advised to configure TortoiseSVN to use Plink from the latest PuTTY 0.81 release when accessing a SVN repository via SSH until a patch becomes available.

ECDSA NIST-P521 keys used with any vulnerable product / component should be considered compromised and consequently revoked by removing them from authorizedkeys, GitHub, ...

References: https://www.chiark.greenend.org.uk/~sgtatham/putty/changes.html https://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/vuln-p521-bias.html

A vulnerability, which was classified as problematic, was found in FileZilla Server up to 0.9.50. This affects an unknown part of the component PORT Handler. The manipulation leads to unintended intermediary. It is possible to initiate the attack remotely. Upgrading to version 0.9.51 is able to address this issue. It is recommended to upgrade the affected component.

A vulnerability has been found in FileZilla Client 3.17.0.0 and classified as problematic. This vulnerability affects unknown code of the file C:\Program Files\FileZilla FTP Client\uninstall.exe of the component Installer. The manipulation leads to unquoted search path. The attack can be initiated remotely. The exploit has been disclosed to the public and may be used.

Untrusted search path in FileZilla before 3.41.0-rc1 allows an attacker to gain privileges via a malicious 'fzsftp' binary in the user's home directory.

FileZilla FTP server before 0.9.6, when using MODE Z (zlib compression), allows remote attackers to cause a denial of service (infinite loop) via certain file uploads or directory listings.

FileZilla FTP server before 0.9.6 allows remote attackers to cause a denial of service via a request for a filename containing an MS-DOS device name such as CON, NUL, COM1, LPT1, and others.

DISPUTED FileZilla v3.59.0 allows attackers to obtain cleartext passwords of connected SSH or FTP servers via a memory dump.- NOTE: the vendor does not consider this a vulnerability.