========================================================================== The checksum's (found through sum -r) of the files that you have received (other than this README) are as follows: 15567 3 patchSG0001418 46721 72 patchSG0001418.eoe_man 12616 5926 patchSG0001418.eoe_sw 22595 24 patchSG0001418.idb ========================================================================== - 1 - 1. Patch_SG0001418_Release_Note This release note describes patch SG0001418 to IRIX 6.2. 1.1 Supported_Hardware_Platforms This patch contains networking related bug fixes for all platforms running IRIX 6.2. This software can be installed on any configurations supported under Irix 6.2. 1.2 Supported_Software_Platforms This patch contains bug fixes for IRIX 6.2 pertaining to the networking subsystems. This patch also contains support necessary for running NFS over TCP. The software can be installed on any configurations supported under Irix 6.2. 1.3 Bugs_Fixed_by_Patch_SG0001418 This patch contains fixes for the following bugs in IRIX 6.2 or other improvements. Bug numbers from Silicon Graphics bug tracking system are included for reference. o Bug 439983: O2-R10K PANIC:tlbmiss: invalid kptbl entry. (see notes below) o Bug 441987: double trip on socket lock (see notes below) o Bug 440572: svr4net still broken on ficus-ssg (see notes below) o Bug 438858: race in tpisocket tests (see notes below) o Bug 284409: tpisocket race conditions cause kernel to panic (see notes below) o Bug 254795: panic in tpitcp_function o Bug 291184: netstat -a reports UNIX domain socket types incorrectly o Bug 351600: Host routes not deleted when IP aliases deleted o Bug 352269: sysctl_iflist() does not report alias addresses o Bug 353160: Routing broken for PPP links with not all 1's netmasks - 2 - o Bug 353174: Cannot add multicast groups to PPP links o Bug 354880: del_vif() can return wil holding mroute lock o Bug 360200: Badly-behaved web clients can leave connections in closing state on server o Bug 360309: Kernel panic in tcp_fasttimo at NTT o Bug 360697: Multicast rate limiting does not use MP locks o Bug 361330: Increase IP_MAX_MEMBERSHIPS to > 20 (Say 300) o Bug 362346: Required fixes for runing RSVP on machines running mrouted o Bug 363009: Invalid mbuf causing tlbmiss in m_freem o Bug 364172: ipaliases causes problems with .rhosts o Bug 364653: ipaliases kernel panic o Bug 365048: MP deadlock in IP multicast code o Bug 366431: System could panic if >327 permanent ARP entries added o Bug 368408: TCP RST packets could be sent with invalid checksums on boards that do hardware checksumming. o Bug 370077: MP deadlock in hash table code under IP multicasting o Bug 372087: netstat -f unix would report no output on 32-bit systems o Bug 374988: Kernel crashes with unused network interface using new routed which supports RIP Version 2 o Bug 375172: Panic: kernel fault o Bug 375230: routed panic("ifafree: ifa- >ifa_start_inifaddr NULL"); o Bug 375298: ip-aliasing bug o Bug 376741: MP system can crash in accept() if connection terminated prematurely; They crash with the - 3 - message "panic: accept". o Bug 379703: Multiprocessor systems could crash in in_pcbnotify() o Bug 380174: IP forwarding performance optimization o Bug 382081: bug in 'netstat -C' on SIGCONT o Bug 387519: livelock in soclose() could result in sockd hanging o Bug 389819: race condition in select() could cause new connections to be missed. o Bug 389824: sosend() could leave socket locked if uiomove() failed. o /var/sysgen/boot/bsd.a includes support for the portions of the 4.4BSD-Lite sysctl() facility used by Patch SG0001371 and Patch SG0001227. o Bug 365622: the manual page for routing socket support was missing from IRIX 6.2. o Bug 382134: netstat -C could dump core when displaying the multicast routing screen (#8). o Bug 391121: tli program crashes/hangs 6.2 machine o Bug 394867: some systems incorrectly send ICMP messages in response to multicast datagrams, and these error reports would confuse multicast applications. IRIX now ignores ICMP error reports sent in response to a datagram that was multicast o Bug 399569: processes could sleep in sosend() and be unkillable. o Bug 408070: with patch 1247 installed, broadcast packets were not looped back correctly. o Bug 409947: `route add 255.255.255.255 10.1` fails & `route get 255.255.255.255` crashed the system. o Bug 415809: multicast rate limiting problem with Cosmo Mediabase o Bug 416381: host unreachable errors were ignored on connections in the SYN_RCVD state, leaving congested listen queues on busy servers - 4 - o Bug 421674: tpisocket could leave data structures locked in some cases. o Bug 422437: The system could run out of mbufs when using Lan Emulation over ATM cards. o Bug 422441: Notification of ICMP errors was very expensive even in cases where it did not need to be. o Bug 423124: rtnetd could hang sleeping in tcp_close() due to a bad interaction between TCP connection processing and NFS page-faulting. o Bug 427433: defend against denial-of-service attack consisting of a stream of TCP SYNs. o Bug 428441: under certain circumstances, getsockopt() could free an mbuf twice. o Bug 428841: if the system ran out of mbufs, UDP could crash the kernel. o Bug 430154: snoop sockets on the loopback interface, lo0, and so netsnooping on local traffic was broken in IRIX 6.2. o Bug 429599: deadlock in ARP o Bug 433439: problem with SO_REUSEADDR and UDP o Bug 438125: under certain circumstances, large ping packets could crash the system. o Bug 443482: MP systems could crash if an application attempted to connect an AF_UNIX socket to itself. o Bug 450354: mbuf leak. 1.4 Subsystems_Included_in_Patch_SG0001418 This patch release includes these subsystems: o patchSG0001418.eoe_sw.svr4net o patchSG0001418.eoe_sw.unix o patchSG0001418.eoe_sw.ipgate - 5 - 1.5 Installation_Instructions Because you want to install only the patches for problems you have encountered, patch software is not installed by default. After reading the descriptions of the bugs fixed in this patch (see Section 1.3), determine the patches that meet your specific needs. If, after reading Sections 1.1 and 1.2 of these release notes, you are unsure whether your hardware and software meet the requirements for installing a particular patch, run inst. The inst program does not allow you to install patches that are incompatible with your hardware or software. Patch software is installed like any other Silicon Graphics software product. Follow the instructions in your Software Installation Administrator's Guide to bring up the miniroot form of the software installation tools. Follow these steps to select a patch for installation: 1. At the Inst> prompt, type install patchSGxxxxxxx where xxxxxxx is the patch number. 2. Initiate the installation sequence. Type Inst> go 3. You may find that two patches have been marked as incompatible. (The installation tools reject an installation request if an incompatibility is detected.) If this occurs, you must deselect one of the patches. Inst> keep patchSGxxxxxxx where xxxxxxx is the patch number. 4. After completing the installation process, exit the inst program by typing Inst> quit - 6 - 1.6 Patch_Removal_Instructions To remove a patch, use the versions remove command as you would for any other software subsystem. The removal process reinstates the original version of software unless you have specifically removed the patch history from your system. versions remove patchSGxxxxxxx where xxxxxxx is the patch number. To keep a patch but increase your disk space, use the versions removehist command to remove the patch history. versions removehist patchSGxxxxxxx where xxxxxxx is the patch number. 1.7 Known_Problems The fixes for bugs: 441987, 440572, 438858, and 284409 require that patch #1572 also be installed. patch #1418 and patch #1572 may be installed in either order. If patch #1572 is not installed, then the complete fixes for the four bugs listed above will not be present and there is the possibility that the kernel may panic when svr4net is enabled (rpcbind) and used under load.