Libarchive multiple security issues

Buffer overflow in libarchive through 2.8.5 allows remote attackers to cause a denial of service (application crash) or possibly execute arbitrary code via a crafted TAR archive.

mount.vmhgfs in the VMware Host Guest File System (HGFS) in VMware Workstation 7.1.x before 7.1.4, VMware Player 3.1.x before 3.1.4, VMware Fusion 3.1.x before 3.1.3, VMware ESXi 3.5 through 4.1, and VMware ESX 3.0.3 through 4.1, when a Solaris or FreeBSD guest OS is used, allows guest OS users to modify arbitrary guest OS files via unspecified vectors, related to a "procedural error."

The Neighbor Discovery (ND) protocol implementation in the IPv6 stack in FreeBSD, NetBSD, and possibly other BSD-based operating systems allows remote attackers to cause a denial of service (CPU consumption and device hang) by sending many Router Advertisement (RA) messages with different source addresses, a similar vulnerability to CVE-2010-4670.

BSD compress LZW decoder buffer overflow

The LZW decompressor in (1) the BufCompressedFill function in fontfile/decompress.c in X.Org libXfont before 1.4.4 and (2) compress/compress.c in 4.3BSD, as used in zopen.c in OpenBSD before 3.8, FreeBSD, NetBSD 4.0.x and 5.0.x before 5.0.3 and 5.1.x before 5.1.1, FreeType 2.1.9, and other products, does not properly handle code words that are absent from the decompression table when encountered, which allows context-dependent attackers to trigger an infinite loop or a heap-based buffer overflow, and possibly execute arbitrary code, via a crafted compressed stream, a related issue to CVE-2006-1168 and CVE-2011-2896.

Buffer overflow in the kernel in FreeBSD 7.3 through 9.0-RC1 allows local users to cause a denial of service (panic) or possibly gain privileges via a bind system call with a long pathname for a UNIX socket.

Directory traversal vulnerability in openpam_configure.c in OpenPAM before r478 on FreeBSD 8.1 allows local users to load arbitrary DSOs and gain privileges via a .. (dot dot) in the service_name argument to the pam_start function, as demonstrated by a .. in the -c option to kcheckpass.

krb5: telnet client and server encrypt_keyid heap-based buffer overflow

Buffer overflow in libtelnet/encrypt.c in telnetd in FreeBSD 7.3 through 9.0, MIT Kerberos Version 5 Applications (aka krb5-appl) 1.0.2 and earlier, Heimdal 1.5.1 and earlier, GNU inetutils, and possibly other products allows remote attackers to execute arbitrary code via a long encryption key, as exploited in the wild in December 2011.

kernel: x86-64: avoid sysret to non-canonical address

The x86-64 kernel system-call functionality in Xen 4.1.2 and earlier, as used in Citrix XenServer 6.0.2 and earlier and other products; Oracle Solaris 11 and earlier; illumos before r13724; Joyent SmartOS before 20120614T184600Z; FreeBSD before 9.0-RELEASE-p3; NetBSD 6.0 Beta and earlier; Microsoft Windows Server 2008 R2 and R2 SP1 and Windows 7 Gold and SP1; and possibly other operating systems, when running on an Intel processor, incorrectly uses the sysret path in cases where a certain address is not a canonical address, which allows local users to gain privileges via a crafted application. NOTE: because this issue is due to incorrect use of the Intel specification, it should have been split into separate identifiers; however, there was some value in preserving the original mapping of the multi-codebase coordinated-disclosure effort to a single identifier.

crypt(): DES encrypted password weakness

The crypt_des (aka DES-based crypt) function in FreeBSD before 9.0-RELEASE-p2, as used in PHP, PostgreSQL, and other products, does not process the complete cleartext password if this password contains a 0x80 character, which makes it easier for context-dependent attackers to obtain access via an authentication attempt with an initial substring of the intended password, as demonstrated by a Unicode password.

The SCTP implementation in FreeBSD 8.2 allows remote attackers to cause a denial of service (NULL pointer dereference and kernel panic) via a crafted ASCONF chunk.

In FreeBSD 11.3-PRERELEASE before r345378, 12.0-STABLE before r345377, 11.2-RELEASE before 11.2-RELEASE-p10, and 12.0-RELEASE before 12.0-RELEASE-p4, a bug in pf does not check if the outer ICMP or ICMP6 packet has the same destination IP as the source IP of the inner protocol packet allowing a maliciously crafted ICMP/ICMP6 packet could bypass the packet filter rules and be passed to a host that would otherwise be unavailable.

Bhyve e82545 device emulation out-of-bounds write

The e1000 network adapters permit a variety of modifications to an Ethernet packet when it is being transmitted. These include the insertion of IP and TCP checksums, insertion of an Ethernet VLAN header, and TCP segmentation offload ("TSO"). The e1000 device model uses an on-stack buffer to generate the modified packet header when simulating these modifications on transmitted packets. When checksum offload is requested for a transmitted packet, the e1000 device model used a guest-provided value to specify the checksum offset in the on-stack buffer. The offset was not validated for certain packet types. A misbehaving bhyve guest could overwrite memory in the bhyve process on the host, possibly leading to code execution in the host context. The bhyve process runs in a Capsicum sandbox, which (depending on the FreeBSD version and bhyve configuration) limits the impact of exploiting this issue.

GELI silently omits the keyfile if read from stdin

When GELI reads a key file from standard input, it does not reuse the key file to initialize multiple providers at once resulting in the second and subsequent devices silently using a NULL key as the user key file. If a user only uses a key file without a user passphrase, the master key is encrypted with an empty key file allowing trivial recovery of the master key.

pf incorrectly handles multiple IPv6 fragment headers

In pf packet processing with a 'scrub fragment reassemble' rule, a packet containing multiple IPv6 fragment headers would be reassembled, and then immediately processed. That is, a packet with multiple fragment extension headers would not be recognized as the correct ultimate payload. Instead a packet with multiple IPv6 fragment headers would unexpectedly be interpreted as a fragmented packet, rather than as whatever the real payload is. As a result, IPv6 fragments may bypass pf firewall rules written on the assumption all fragments have been reassembled and, as a result, be forwarded or processed by the host.

NFS client data corruption and kernel memory disclosure

When a program running on an affected system appends data to a file via an NFS client mount, the bug can cause the NFS client to fail to copy in the data to be written but proceed as though the copy operation had succeeded. This means that the data to be written is instead replaced with whatever data had been in the packet buffer previously. Thus, an unprivileged user with access to an affected system may abuse the bug to trigger disclosure of sensitive information. In particular, the leak is limited to data previously stored in mbufs, which are used for network transmission and reception, and for certain types of inter-process communication. The bug can also be triggered unintentionally by system applications, in which case the data written by the application to an NFS mount may be corrupted. Corrupted data is written over the network to the NFS server, and thus also susceptible to being snooped by other hosts on the network. Note that the bug exists only in the NFS client; the version and implementation of the server has no effect on whether a given system is affected by the problem.

TCP spoofing vulnerability in pf(4)

In versions of FreeBSD 14.0-RELEASE before 14-RELEASE-p2, FreeBSD 13.2-RELEASE before 13.2-RELEASE-p7 and FreeBSD 12.4-RELEASE before 12.4-RELEASE-p9, the pf(4) packet filter incorrectly validates TCP sequence numbers.  This could allow a malicious actor to execute a denial-of-service attack against hosts behind the firewall.

Incorrect libcap_net limitation list manipulation

In versions of FreeBSD 13-RELEASE before 13-RELEASE-p5, under certain circumstances the cap_net libcasper(3) service incorrectly validates that updated constraints are strictly subsets of the active constraints.  When only a list of resolvable domain names was specified without setting any other limitations, an application could submit a new list of domains including include entries not previously listed.  This could permit the application to resolve domain names that were previously restricted.

libc stdio buffer overflow

In versions of FreeBSD 12.4-RELEASE prior to 12.4-RELEASE-p7 and FreeBSD 13.2-RELEASE prior to 13.2-RELEASE-p5 the __sflush() stdio function in libc does not correctly update FILE objects' write space members for write-buffered streams when the write(2) system call returns an error.  Depending on the nature of an application that calls libc's stdio functions and the presence of errors returned from the write(2) system call (or an overridden stdio write routine) a heap buffer overflow may occur. Such overflows may lead to data corruption or the execution of arbitrary code at the privilege level of the calling program.

arm64 boot CPUs may lack speculative execution protections

On CPU 0 the check for the SMCCC workaround is called before SMCCC support has been initialized. This resulted in no speculative execution workarounds being installed on CPU 0.

copy_file_range insufficient capability rights check

Before correction, the copy_file_range system call checked only for the CAP_READ and CAP_WRITE capabilities on the input and output file descriptors, respectively. Using an offset is logically equivalent to seeking, and the system call must additionally require the CAP_SEEK capability. This incorrect privilege check enabled sandboxed processes with only read or write but no seek capability on a file descriptor to read data from or write data to an arbitrary location within the file corresponding to that file descriptor.