This is the OpenBSD 3.5 release errata & patch list:

The patches below are available in CVS via the OPENBSD_3_5 patch branch.

For more detailed information on how to install patches to OpenBSD, please consult the OpenBSD FAQ.

• 033: SECURITY FIX: April 28, 2005   All architectures
  Fix a buffer overflow, memory leaks, and NULL pointer dereference in cvs(1) . None of these issues are known to be exploitable. CAN-2005-0753 .
  A source code patch exists which remedies this problem.

• 032: RELIABILITY FIX: April 4, 2005   All architectures
  Handle an edge condition in tcp(4) timestamps.
  A source code patch exists which remedies this problem.

• 031: SECURITY FIX: March 30, 2005   All architectures
  Due to buffer overflows in telnet(1) , a malicious server or man-in-the-middle attack could allow execution of arbitrary code with the privileges of the user invoking telnet(1) .
  A source code patch exists which remedies this problem.

• 030: RELIABILITY FIX: March 30, 2005   All architectures
  Bugs in the tcp(4) stack can lead to memory exhaustion or processing of TCP segments with invalid SACK options and cause a system crash.
  A source code patch exists which remedies this problem.

• 029: SECURITY FIX: March 16, 2005   amd64 only
  More stringent checking should be done in the copy(9) functions to prevent their misuse.
  A source code patch exists which remedies this problem.

• 028: SECURITY FIX: February 28, 2005   i386 only
  More stringent checking should be done in the copy(9) functions to prevent their misuse.
  A source code patch exists which remedies this problem.

• 027: RELIABILITY FIX: January 11, 2005   All architectures
  A bug in the tcp(4) stack allows an invalid argument to be used in calculating the TCP retransmit timeout. By sending packets with specific values in the TCP timestamp option, an attacker can cause a system panic.
  A source code patch exists which remedies this problem.

• 026: SECURITY FIX: January 12, 2005   All architectures
  httpd(8) 's mod_include module fails to properly validate the length of user supplied tag strings prior to copying them to a local buffer, causing a buffer overflow.
  This would require enabling the XBitHack directive or server-side includes and making use of a malicious document.
  A source code patch exists which remedies this problem.

• 025: RELIABILITY FIX: January 6, 2005   All architectures
  The getcwd(3) library function contains a memory management error, which causes failure to retrieve the current working directory if the path is very long.
  A source code patch exists which remedies this problem.

• 024: SECURITY FIX: December 14, 2004   All architectures
  On systems running isakmpd(8) it is possible for a local user to cause kernel memory corruption and system panic by setting ipsec(4) credentials on a socket.
  A source code patch exists which remedies this problem.

• 023: RELIABILITY FIX: November 10, 2004   All architectures
  Due to a bug in lynx(1) it is possible for pages such as this to cause lynx(1) to exhaust memory and then crash when parsing such pages.
  A source code patch exists which remedies this problem.

• 022: RELIABILITY FIX: November 10, 2004   All architectures
  pppd(8) contains a bug that allows an attacker to crash his own connection, but it cannot be used to deny service to other users.
  A source code patch exists which remedies this problem.

• 021: RELIABILITY FIX: November 10, 2004   All architectures
  BIND contains a bug which results in BIND trying to contact nameservers via IPv6, even in cases where IPv6 connectivity is non-existent. This results in unnecessary timeouts and thus slow DNS queries.
  A source code patch exists which remedies this problem.

• 020: SECURITY FIX: September 20, 2004   All architectures
  Eilko Bos reported that radius authentication, as implemented by login_radius(8), was not checking the shared secret used for replies sent by the radius server. This could allow an attacker to spoof a reply granting access to the attacker. Note that OpenBSD does not ship with radius authentication enabled.
  A source code patch exists which remedies this problem.

• 019: SECURITY FIX: September 16, 2004   All architectures
  Chris Evans reported several flaws (stack and integer overflows) in the Xpm library code that parses image files (CAN-2004-0687, CAN-2004-0688). Some of these would be exploitable when parsing malicious image files in an application that handles XPM images, if they could escape ProPolice.
  A source code patch exists which remedies this problem.

• 018: SECURITY FIX: September 10, 2004   All architectures
  httpd(8) 's mod_rewrite module can be made to write one zero byte in an arbitrary memory position outside of a char array, causing a DoS or possibly buffer overflows. This would require enabling dbm for mod_rewrite and making use of a malicious dbm file.
  A source code patch exists which remedies this problem.

• 017: RELIABILITY FIX: August 29, 2004   All architectures
  Due to incorrect error handling in zlib an attacker could potentially cause a Denial of Service attack. CAN-2004-0797 .
  A source code patch exists which remedies this problem.

• 016: RELIABILITY FIX: August 26, 2004   All architectures
  As reported by Vafa Izadinia bridge(4) with IPsec processing enabled can be crashed remotely by a single ICMP echo traversing the bridge.
  A source code patch exists which remedies this problem.

• 015: RELIABILITY FIX: August 25, 2004   All architectures
  Improved verification of ICMP errors in order to minimize the impact of ICMP attacks against TCP.
  http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html
  A source code patch exists which remedies this problem.

• 014: RELIABILITY FIX: July 25, 2004   All architectures
  Under a certain network load the kernel can run out of stack space. This was encountered in an environment using CARP on a VLAN interface. This issue initially manifested itself as a FPU related crash on boot up.
  A source code patch exists which remedies this problem.

• 013: SECURITY FIX: June 12, 2004   All architectures
  Multiple vulnerabilities have been found in httpd(8) / mod_ssl. CAN-2003-0020, CAN-2003-0987, CAN-2004-0488, CAN-2004-0492.
  A source code patch exists which remedies this problem.

• 012: SECURITY FIX: June 10, 2004   All architectures
  As disclosed by Thomas Walpuski isakmpd(8) is still vulnerable to unauthorized SA deletion. An attacker can delete IPsec tunnels at will.
  A source code patch exists which remedies this problem.

• 011: SECURITY FIX: June 9, 2004   All architectures
  Multiple remote vulnerabilities have been found in the cvs(1) server that allow an attacker to crash the server or possibly execute arbitrary code with the same privileges as the CVS server program.
  A source code patch exists which remedies this problem.

• 010: RELIABILITY FIX: June 9, 2004   All architectures
  A FIFO bug was introduced in OpenBSD 3.5 that occurs when a FIFO is opened in non-blocking mode for writing when there are no processes reading the FIFO. One program affected by this is the qmail mail server which could go into an infinite loop and consume all CPU.
  A source code patch exists which remedies this problem.

• 009: SECURITY FIX: May 30, 2004   All architectures
  A flaw in the Kerberos V kdc(8) server could result in the administrator of a Kerberos realm having the ability to impersonate any principal in any other realm which has established a cross-realm trust with their realm. The flaw is due to inadequate checking of the "transited" field in a Kerberos request. For more details see Heimdal's announcement.
  A source code patch exists which remedies this problem.

• 008: SECURITY FIX: May 26, 2004   All architectures
  With the introduction of IPv6 code in xdm(1), one test on the 'requestPort' resource was deleted by accident. This makes xdm create the chooser socket even if xdmcp is disabled in xdm-config, by setting requestPort to 0. See XFree86 bugzilla for details.
  A source code patch exists which remedies this problem.

• 007: SECURITY FIX: May 20, 2004   All architectures
  A heap overflow in the cvs(1) server has been discovered that can be exploited by clients sending malformed requests, enabling these clients to run arbitrary code with the same privileges as the CVS server program.
  A source code patch exists which remedies this problem.

• 006: SECURITY FIX: May 13, 2004   All architectures
  Check for integer overflow in procfs. Use of procfs is not recommended.
  A source code patch exists which remedies this problem.

• 005: RELIABILITY FIX: May 6, 2004   All architectures
  Reply to in-window SYN with a rate-limited ACK.
  A source code patch exists which remedies this problem.

• 004: RELIABILITY FIX: May 5, 2004   All architectures
  Restore the ability to negotiate tags/wide/sync with some SCSI controllers ( i.e. siop(4), trm(4), iha(4) ).
  A source code patch exists which remedies this problem.

• 003: RELIABILITY FIX: May 5, 2004   All architectures
  Under load "recent model" gdt(4) controllers will lock up.
  A source code patch exists which remedies this problem.

• 002: SECURITY FIX: May 5, 2004   All architectures
  Pathname validation problems have been found in cvs(1), allowing malicious clients to create files outside the repository, allowing malicious servers to overwrite files outside the local CVS tree on the client and allowing clients to check out files outside the CVS repository.
  A source code patch exists which remedies this problem.

• 001: BROKEN PACKAGE ON CD: May 4, 2004  macppc only
  The powerpc autobook-1.3.tgz package found on CD2 has been found to be corrupt, and will not extract. A replacement package can be found on the ftp sites.