Patch Name: PHNE_11674 Patch Description: s700_800 10.20 2.23 ACC X.25/9000 Support (EISA) Patch Creation Date: 97/07/09 Post Date: 97/07/14 Repost: 97/09/11 The superseding patch PHNE_12227 was removed from distribution. This patch will remain available till the new replacement patch PHNE_12526 is released. Hardware Platforms - OS Releases: s700: 10.20 s800: 10.20 Products: Z7401A 2.23; Z7436AA 2.23; Filesets: ACC-X25ST.ACC-X25ST-KRN X25CORE.X25CORE-KRN X25BSD.X25BSD-KRN X25IP.X25IP-KRN Automatic Reboot?: Yes Status: General Superseded Critical: No (superseded patches were critical) PHNE_11027: PANIC HANG Path Name: /hp-ux_patches/s700_800/10.X/PHNE_11674 Symptoms: PHNE_11674: SR 5003374884 / DTS TPO0h01431: Client Call request reject by the server test program with non-default X25 facility request. PHNE_11435: SR None: In an inbound flow control condition (especially when the remote system is continuing to send data packets after an RNR), in rare circumstances the ACC will send an RNR and never reply with an RR when the application reads all of its data. SR None: ENOSPC (no space) error in the Nettl logfile when an attempt is made to connect to a new SVC. PHNE_11027: SR 5003371195: - SVCs not getting cleared or hanging. - Duplicate packets arriving. - Netisr defect. - Defect in socket lock/unlock code will show up as: - Panic in sounlock() routine. - Random panic implicating malloc. - All networking on the system hangs. SR None: Also noticing in axin that when an error occurred on an outbound call, an unnecessary attempt would be made to disable the ZLU. SR None: 15 copies of x25stat(1M) were running. Customer did a "zmasterd kill". This caused the system to panic. SR 4701350231 / DTS TPO0h01210: If a user issued an outbound call request on the last free logical channel number at the same time that an inbound call arrived on that LCN, the outbound call would be lost causing the application to wait for ever for the call to complete. SR None: Application could get ENOSPC errors when only a single X.25 call was placed. X.25 SVCs still established when all of the applications using X.25 have terminated. SR None: When only a single VC is being used in the entire system, an inbound call indication or the call accepted packet could potentially not get delivered to the application. That is, the connect() call might hang. SR 4701352062 / DTS TPO0h01260: Running x25check test program on multiple 2-channel ACC cards causes the test to fail due to duplicate sequence number errors. SR 4701348086 / DTS TPO0h01132: PVC sockets that were closed by an application did not have a Reset Request issued on them as documented in the X.25/9000 manuals. SR 1653195677 / DTS TPO0h00908: A connection cannot be established using FTAM and OTS/9000 with the ACC when the ACC is connected to an older model Siemens switch. Defect Description: PHNE_11674: SR 5003374884 / DTS TPO0h01431: Problem in the axin driver. The axin driver is not handling the processing of the CLASS-D octect length for the facility packet. PHNE_11435: SR None: This problem occurs then the software has turned off inbound flow control and additional data is received that exceeds the size of the socket buffer. SR None This is a follow-up fix to the last superseded patch. Error due to code commented out for SVC recovery but flags governing code not commented out. PHNE_11027: SR 5003371195: - SVCs not getting cleared or hanging. Classic X.25 product was issuing a disconnect on the wrong SVC and then rejecting an inbound call accept. - Duplicate packets arriving. Classic X.25 product was issuing a disconnect on the wrong SVC and then rejecting an inbound call accept. - Workaround fix to accomodate netisr defect. Fixing unproctected Interrupt problem. The protection level around the calling of schednetisr() is dropped to splnet(). This opens up a small time frame to allow the ISR come, which causes the x25check out of sequence problem. - This patch is a comprehensive attempt to fix the socket lock/unlock problem. SR None: Fixed unnecessary disabling of the ZLU by commenting out code. SR None: 15 copies of x25stat(1M) were running. Customer did a "zmasterd kill". This caused the system to panic. Fixed spl levels, pointers in axin to handle this. SR 4701350231 / DTS TPO0h01210: When a call collision occured, both the inbound and outbound calls on that SVC were thrown away. SR None: If a clear request was issued by the application before the call was accepted (or cleared) from the remote end of the VC, the wrong call request table entry was being used. This caused the disconnect request to be ignored leaving an established SVC with no socket associated with it. SR None: Occassionally, netisr is called to schedule the processing of an inbound event, but netisr fails to call the driver's inbound event handler. Netisr loses schedule requests potentially caused an unprocessed "orphan" event to be left unprocessed. SR 4701352062 / DTS TPO0h01260: The ACC interface (DAM) driver queues the data packets and schedules the netisr to call axin_proc_eventq() to process incoming data packet in a more convinient time. The spl proctection around the code that set up the netisr was dropped from spl5 to splnet, which opens up a window for the ISR to preempt the axin driver before it has the chance to finish the staging of data packet; hence causes the same data packet to be processed twice. SR 4701348086 / DTS TPO0h01132: The X.25/9000 product deleted the functionality which issued the Reset Request when the socket close was issued. TPD was never informed of this change. SR 1653195677 / DTS TPO0h00908: OTS/9000 is setting up the facilities with the facility code 0x01 for reverse charging and fast select, but the associated facility parameter field is 0. This is allowed by the X.25 specification but is not accepted by some Siemens switches. SR: 5003374884 5003371195 4701350231 4701352062 4701348086 1653195677 Patch Files: /usr/conf/lib/libaxin.a /usr/conf/space.h.d/axin_space.h /usr/conf/lib/libx25.a /usr/conf/lib/libx25pa.a /usr/conf/lib/libx25ip.a what(1) Output: /usr/conf/lib/libaxin.a: ACC Rel B.02.23 for B.10.20 PHNE_11674 $Header: axi n_ssm.c@@/main/r2.23/2 05/23/97 00:31:30 $ /usr/conf/space.h.d/axin_space.h: $Header: axin_space.h@@/main/r2.23/1 04/30/97 09:43 :21 $ /usr/conf/lib/libx25.a: X.25: Version: B.10.20.00 $ Revision: TPD 2.7 PHNE_1 1027 (05 Jun 97 16:08) /usr/conf/lib/libx25pa.a: X.25: Version: B.10.20.00 $ Revision: TPD 2.7 PHNE_1 1027 (05 Jun 97 16:08) /usr/conf/lib/libx25ip.a: X.25: Version: B.10.20.00 $ Revision: TPD 2.7 PHNE_1 1027 (05 Jun 97 16:08) cksum(1) Output: 1641209185 132136 /usr/conf/lib/libaxin.a 1226570961 11050 /usr/conf/space.h.d/axin_space.h 2204106671 181428 /usr/conf/lib/libx25.a 1129683750 59864 /usr/conf/lib/libx25pa.a 4215227105 29896 /usr/conf/lib/libx25ip.a Patch Conflicts: None Patch Dependencies: None Hardware Dependencies: None Other Dependencies: None Supersedes: PHNE_11027 PHNE_11435 Equivalent Patches: None Patch Package Size: 470 KBytes Installation Instructions: Please review all instructions and the Hewlett-Packard SupportLine User Guide or your Hewlett-Packard support terms and conditions for precautions, scope of license, restrictions, and, limitation of liability and warranties, before installing this patch. ------------------------------------------------------------ 1. Back up your system before installing a patch. 2. Login as root. 3. Copy the patch to the /tmp directory. 4. Move to the /tmp directory and unshar the patch: cd /tmp sh PHNE_11674 5a. For a standalone system, run swinstall to install the patch: swinstall -x autoreboot=true -x match_target=true \ -s /tmp/PHNE_11674.depot 5b. For a homogeneous NFS Diskless cluster run swcluster on the server to install the patch on the server and the clients: swcluster -i -b This will invoke swcluster in the interactive mode and force all clients to be shut down. WARNING: All cluster clients must be shut down prior to the patch installation. Installing the patch while the clients are booted is unsupported and can lead to serious problems. The swcluster command will invoke an swinstall session in which you must specify: alternate root path - default is /export/shared_root/OS_700 source depot path - /tmp/PHNE_11674.depot To complete the installation, select the patch by choosing "Actions -> Match What Target Has" and then "Actions -> Install" from the Menubar. 5c. For a heterogeneous NFS Diskless cluster: - run swinstall on the server as in step 5a to install the patch on the cluster server. - run swcluster on the server as in step 5b to install the patch on the cluster clients. By default swinstall will archive the original software in /var/adm/sw/patch/PHNE_11674. If you do not wish to retain a copy of the original software, you can create an empty file named /var/adm/sw/patch/PATCH_NOSAVE. Warning: If this file exists when a patch is installed, the patch cannot be deinstalled. Please be careful when using this feature. It is recommended that you move the PHNE_11674.text file to /var/adm/sw/patch for future reference. To put this patch on a magnetic tape and install from the tape drive, use the command: dd if=/tmp/PHNE_11674.depot of=/dev/rmt/0m bs=2k Special Installation Instructions: SUBSYSTEM_SHUT Before installing this patch, use the following command to shutdown the ACC subsystem and kill the ACC daemons. /opt/acc/bin/zmasterd kill The above command kills any of the ACC daemons that are still running, including zmasterd, zx25d, watch, zmlog, zmon, and znode. The kernel will be regenerated and the system will reboot automatically during the patch installation.