OS/2 Shareware BBS: 29 Fixes

home *** CD-ROM | disk | FTP | other *** search

/ OS/2 Shareware BBS: 29 Fixes_o / 29-Fixes_o.zip / us6025b1.dsk / WR06025.INF (.txt) < prev

Wrap

OS/2 Help File | 1993-11-12 | 70KB | 1,810 lines

ΓòÉΓòÉΓòÉ 1. Introduction ΓòÉΓòÉΓòÉ IBM OS/2 Extended Services 1.0 ServicePak WR06025 IBM PTF WR06025 This ServicePak applies to: IBM OS/2 Extended Services Version 1.0 - Communications Manager Warning: 1 - If Communications Manager was CID installed, files may have been left in the OS2 directory tree which can cause unpredictable results. Double Click Here for a list of files. THESE FILES SHOULD BE DELETED. Hint: Use "Services - Copy to file" to produce a file with these names that could be made into a command file to automate the deletions. Warning: 2 - If asked for permission to replace APPNV.CTL, respond Yes. It is recommended to Have appropriate Bootable Standalone Service Diskettes. ΓòÉΓòÉΓòÉ 1.1. Affected SYSLEVEL Files ΓòÉΓòÉΓòÉ SYSLEVEL.ACS IBM OS/2 Communications Manager ΓòÉΓòÉΓòÉ 1.2. Trademarks ΓòÉΓòÉΓòÉ The following are trademarks of IBM Corporation: o IBM o OS/2 o Operating System/2 o Presentation Manager ΓòÉΓòÉΓòÉ 2. Installation ΓòÉΓòÉΓòÉ Installing the ServicePak is essentially a three phase process: Phase 1 - Quiesce the Target System Render the system serviceable by insuring that there will be no Locked Files during the install phase. Phase 2 - Install the ServicePak Select one of the following methods to replace system files with those provided on this ServicePak: Service.exe: A Presentation Manager Installation Aid that checks for Locked Files and allows user selectivity. FService.exe: A standalone tool that can be run from a minimal (eg. diskette based) OS/2 system. Redirected (CID) Install: A technique that allows distribution of service across a Local Area Network. Phase 3 - Return the Serviced System to Normal Operation Perform a normal system boot. Warning: 1 - If Communications Manager was CID installed, files may have been left in the OS2 directory tree which can cause unpredictable results. Double Click Here for a list of files. THESE FILES SHOULD BE DELETED. Hint: Use "Services - Copy to file" to produce a file with these names that could be made into a command file to automate the deletions. Warning: 2 - If asked for permission to replace APPNV.CTL, respond Yes. ΓòÉΓòÉΓòÉ 2.1. Installation using SERVICE.EXE ΓòÉΓòÉΓòÉ This method will apply fixes to selected subsystems, partitions, and directories. SERVICE.EXE will check for Locked Files before continuing with the application of service. If SERVICE.EXE cannot proceed, it will present the following options: "Retry", "Reboot", or "Cancel". 1. Boot the system from fixed-disk. 2. Insert ServicePak diskette #1 into the A: drive. 3. Establish A: as the Current Directory. 4. From an OS/2 session, type A:SERVICE and press Enter. 5. Make selections from the panel. If the Locked Files panel appears, choose from the options below. Retry - Quiesce the subsystem being serviced and retry the application of service. a) Stop Communications Manager if it is running. b) Move to another OS/2 window (OS/2 command prompt). c) With ServicePak diskette 1 in the A: drive, run "CMKFMSMI" with a parameter of 0 (zero) to unlock frame manager. Type A:\CMKFMSMI 0 and press Enter. d) Return to the Service window and Choose "Retry". Reboot - A response file containing the current selections will be written to the boot drive's root directory. FSERVICE.EXE can then be employed, after reboot, to complete the service process. a) Choose "Reboot". b) Perform the system Shutdown procedure. c) Boot from the Bootable Standalone Service Diskettes. d) Establish A:\ as the Current Directory. e) Replace the Boot Diskette with ServicePak diskette #1. f) At the A: prompt type FSERVICE and press Enter. g) At the panel titled "Updating Default Directories" press Enter (Option 1). h) Follow the prompts provided by the Installation Aid. Cancel - Abort the service process without updating any files. a) This safely terminates the service procedure. b) Skip the next (Reboot) step. 6. Reboot the system normally. ΓòÉΓòÉΓòÉ 2.2. Installation using FSERVICE.EXE ΓòÉΓòÉΓòÉ This method applies fixes using a response file. The default response file will apply fixes to a default set of partitions and directories. The defaults established by product development are generally dependent on the SYSLEVEL.xxx file where each product has a unique "xxx". This is the simplest and safest method to install ServicePaks to systems that do not have multiple versions of the product being serviced. It does require booting from diskette to insure that there are no Locked Files during the install phase. 1. Boot from the Bootable Standalone Service Diskettes 2. Establish A: as the Current Directory. 3. Replace the Boot Diskette with ServicePak diskette #1. 4. At the A: prompt type "FSERVICE" and press Enter. 5. At the panel titled "Updating Default Directories" press Enter (Option 1) 6. Follow the prompts provided by the Installation Aid. 7. Reboot the system normally. ΓòÉΓòÉΓòÉ 2.3. Redirected Installation (CID) Method ΓòÉΓòÉΓòÉ This method is useful for those who are using the CID (Configuration, Installation, and Distribution) Services as provided by IBM NTS/2. Installation using SERVICE.EXE or Installation using FSERVICE.EXE are recommended for Servers exploiting the 386 HPFS advanced functions: Fault Tolerance and/or Access Control Lists. While setting up a CID server is beyond the scope of this document, a sample CID Utility REXX Command file (CID_6025.CMD) is included that may prove useful in setting up a CID server for this ServicePak. The sample CID_6025.CMD assumes the following: o CID_6025.CMD resides in the ..\CLIENT subdirectory. o The ServicePak files are placed on the code server using "XCOPY A:\*.* /S" (or equivalent) at 'x:\csd\cid_6025' where 'x:' is the drive seen by the client. o The following directory structure contains the ServicePak contents. csd\cid_6025\ * Root directory of ServicePak csd\cid_6025\FIX * No files csd\cid_6025\FIX\ACS * Communications Manager Fixes o A log directory exists at 'x:\log\csd\cid_6025' for the log files. ΓòÉΓòÉΓòÉ 2.4. Bootable Standalone Service Diskettes ΓòÉΓòÉΓòÉ Sometimes it becomes necessary to boot from diskette in order to perform certain operations on fixed-disk, like installing this ServicePak or running CHKDSK /f. Hint:It can be useful to copy CHKDSK.COM, EDLIN.COM, and other small utilities to your standalone service diskette (SA#2), (providing there is sufficient space available). All necessary files must be contained on diskette. Files from the fixed-disk should not be referenced in A:\CONFIG.SYS or A:\STARTUP.CMD. Otherwise Locked Files can still exist on the fixed-disk. Should the need arise to boot from diskette for any reason, you may: o Use the Install diskette from OS/2 1.3. Escape to A: o Use the Install diskette and Diskette #1 of OS/2 2.x. Escape to A: o Create custom Bootable Standalone Service Diskettes from your licensed materials. The makeup of these is dependent on the system to be serviced. Select the instructions for: Extended Services 1.0 on OS/2 1.3 Extended Services 1.0 on OS/2 2.x Without IBM LAN Server Advanced. Advanced Version Servers. CAUTION: LAN Servers that utilize Advanced features like Access Control Lists or Fault Tolerance, must use custom Bootable Standalone Service Diskettes. Regular HPFS boot diskettes, such as the OS/2 Installation diskettes, CAN NOT mirror changes to Fault Tolerance partitions or access files/directories which are controlled by access control profiles. ΓòÉΓòÉΓòÉ 2.4.1. Making OS/2 1.3 Bootable Diskettes. ΓòÉΓòÉΓòÉ 1. Make a copy of the OS/2 1.3 Installation Diskette. Call it SA#1. 2. Modify CONFIG.SYS on SA#1 as follows: buffers=50 memman=noswap protshell=a:\cmd.exe protectonly=yes libpath=a:\; ifs=hpfs.ifs /c:64 set path=a:\; set dpath=a:\; ΓòÉΓòÉΓòÉ 2.4.2. Making OS/2 2.x Bootable Diskettes ΓòÉΓòÉΓòÉ 1. Make a copy of the OS/2 2.x Installation Diskette & Diskette #1. Call them SA#1 and SA#2 respectively. 2. Delete SYSINST2.EXE from Diskette SA#2. 3. Delete BUNDLE from Diskette SA#2 (this will be on OS/2 2.1 only) 4. Modify CONFIG.SYS on SA#2 as follows: buffers=50 memman=noswap protshell=a:\cmd.exe protectonly=yes libpath=a:\; ifs=hpfs.ifs /c:64 set path=a:\; set dpath=a:\; 5. Copy UHPFS.DLL from OS/2 Diskette #2 or C:\OS2\DLL to SA#2. ΓòÉΓòÉΓòÉ 2.4.3. Making Advanced Server Bootable Diskettes ΓòÉΓòÉΓòÉ MAKEDISK is an Advanced Server utility designed to make the necessary Bootable Standalone Service Diskettes used in periodic maintenance of 386 HPFS systems. 1. Make a copy of the OS/2 2.x Installation Diskette & Diskette #1. Call them SA#1 and SA#2 respectively. 2. Delete SYSINST2.EXE from Diskette SA#2. 3. Delete BUNDLE from Diskette SA#2 (this will be on OS/2 2.1 only) 4. From an OS/2 prompt run MAKEDISK. Note: Syntax is MAKEDISK /BOOTDRIVE:d where d is the drive on which OS/2 2.x is installed. 5. Insert Diskette SA#2 into a diskette drive when prompted. Note: For additional Information see the Network Administrators Reference: Creating a 386 HPFS Boot Diskette, or LAN Server Utilities. 6. If using Fault Tolerance, add the Fault Tolerance files to the Bootable Diskette. 7. Verify that the 386 HPFS Bootable Standalone Service Diskette works properly. o Shutdown the system. o Insert the Bootable Standalone Service Diskette into the A: drive. o Reboot the system (Ctrl+Alt+Del, or other method). o Determine if Local Security is working. o Determine if Fault Tolerance is working. ΓòÉΓòÉΓòÉ 2.4.4. Adding Fault Tolerance to 386 HPFS Standalone Boot Diskette ΓòÉΓòÉΓòÉ 386 HPFS Standalone Boot diskettes require additional customizing in order to properly update files on partitions which are mirrored. The following steps may have to be done for each LAN Server with Fault Tolerance since the configuration information transferred in step 3 can be unique for each such system. 1. Copy the following four Fault Tolerance files from the first diskette of the ServicePak to your Bootable Standalone Service Diskette. 1) DISKFT.SYS 2) FTD.MSG 3) FTATTRIB.EXE 4) FTCOPY.CMD 2. Add the following line to the CONFIG.SYS on the Bootable Standalone Service Diskette SA#2. DEVICE=DISKFT.SYS 3. At an OS/2 command prompt, make the current drive A:, and have the 386 HPFS Bootable Standalone Service Diskette in drive A. Type "FTCOPY" to copy the customized Fault Tolerance configuration information to the 386 HPFS Bootable Standalone Service Diskette. Warning: Test your 386 HPFS Bootable Standalone Service Diskette before you make any alterations to your system. This Diskette is the Master Key should your system ever fail to boot from its hard drive. ΓòÉΓòÉΓòÉ <hidden> Locked Files ΓòÉΓòÉΓòÉ Sometimes it is essential that programs own (become the master of) system resources such as Memory Blocks, Communication Ports, and Files. When a program stakes claim to a file, that file is said to be Locked; no other program is allowed to alter (replace or write to) it until the owning program explicitly removes the lock. .DLL Files can become Locked by residing in a directory that is included in the LIBPATH string of CONFIG.SYS. For example, NETAPI.DLL can be locked when \MUGLIB\DLL exists in the LIBPATH of CONFIG.SYS. ΓòÉΓòÉΓòÉ <hidden> Getting the A: Prompt from OS/2 1.3 Install Diskettes ΓòÉΓòÉΓòÉ To obtain the familiar A: prompt when using the OS/2 1.3 Product installation diskettes as Bootable Standalone Service Diskettes: o Boot the system with Install Diskette in drive A. o At the IBM Logo Screen - Press Esc to Exit The A: prompt should appear. OS/2 Programs, including FSERVICE.EXE, can be run from here. ΓòÉΓòÉΓòÉ <hidden> Getting the A: Prompt from OS/2 2.x Install Diskettes ΓòÉΓòÉΓòÉ To obtain the familiar A: prompt when using the OS/2 2.x Product installation diskettes as Bootable Standalone Service Diskettes: o Boot the system with Install Diskette in drive A. o At the IBM Logo Screen - Insert Diskette #1, press Enter o At the Welcome Screen - Press Esc to Exit The A: prompt should appear. OS/2 Programs, including FSERVICE.EXE, can be run from here. ΓòÉΓòÉΓòÉ <hidden> Creating a 386 HPFS Boot Diskette ΓòÉΓòÉΓòÉ Refer to appropriate Appendix in: Operating System/2 Local Area Network Server Network Administrator Reference Volume1: Planning and Installation ΓòÉΓòÉΓòÉ <hidden> LAN Server Utilities ΓòÉΓòÉΓòÉ Refer to Appendix C, Miscellaneous Utilities MAKEDISK in: Operating System/2 Local Area Network Server Network Administrator Reference Volume3: Network Administrator Tasks Version 2.0 and Version 3.0 ΓòÉΓòÉΓòÉ 3. Readme (New Function) ΓòÉΓòÉΓòÉ Sample programs have not been shipped in this version of Communications Manager. A problem has been identified when compiling the APPCTELL sample program, or any user written application program that uses a call to VioPopUp with the C Set/2 compiler. Cprintf's must be used instead of printf's. This version of the Communications Manager contains the following enhancements: o 5250 Emulator Automatic Session Recovery o 3270 Host Print EBCDIC Passthru Option o Handle Type 1 UNBIND Similar to Type 2 UNBIND o User SDT Responses in LU0 SLI Interface o LUA Correlation Table Limit Overrides o EHLLAPI Virtual Device Driver Windows support These enhancements are described below and are not included in the current publications. ΓòÉΓòÉΓòÉ 3.1. 5250 Emulator Automatic Session Recovery ΓòÉΓòÉΓòÉ This version of the Communications Manager 5250 emulator contains an enhancement that enables the 5250 emulator to automatically re-establish emulation sessions when the sessions have been reset. Sessions are reset when the Emulator can no longer communicate to the host (AS/400 or S/36) system, usually due to a communications line failure. A reset session is noted by the blanking of the screen with the cursor placed in the upper right corner of the screen, and the System Available (SA) indicator on the emulator status line turned 'off'. When auto-recovery is initiated, the System Available indicator on the emulation screen will change to indicate recovery mode is active. The emulator will continuously try, every 30 seconds, to restart the session(s) until it is successful, an unrecoverable error occurs, or until the session(s) is(are) manually stopped. If recovery is successful, the host sign-on screen will reappear and the System Available indicator will resume its normal state. In most cases, the 5250 emulator will not know that the communications line has failed until the user initiates an action, such as pressing Enter on the keyboard, that causes the emulator to exchange information with the host. The 5250 emulator session recovery is disabled by default. To enable the feature, an environment variable named "EM5250R' must be SET to "AUTO" before the 5250 emulator is started. This can be done by following the procedure below: 1. Locate the Communications Manager command file named "STARTCM.CMD". This is located in the directory named "CMLIB" on the drive that was selected during the OS/2 Extended Services installation process. 2. Edit the file named STARTCM.CMD with a text editor. 3. Insert a line that contains "SET EM5250R=AUTO" after the line "SET EM5250=WIN". 4. Save the file. 5. Restart Communications Manager. ΓòÉΓòÉΓòÉ 3.2. 3270 Host Print EBCDIC Passthru Option ΓòÉΓòÉΓòÉ A new function, EBCDIC Passthru, has been added to 3270 host print in this version of the Communications Manager. This new function works exactly like the ASCII Passthru function except that the data will be translated from EBCDIC to ASCII. To enable the EBCDIC Passthru function, place the following string anywhere in the Queue Description field: EBCDIC PASSTHRU ΓòÉΓòÉΓòÉ 3.3. Handle Type 1 UNBIND Similar to Type 2 UNBIND ΓòÉΓòÉΓòÉ A new option has been added to the SLI_OPEN verb; lua_session_type. This field, formerly reserved, is at offset 65 in the LUA verb control block. When this field is set to the value 1 (LUA_SESSION_TYPE_DEDICATED) the SLI processes UNBIND type X'01', normal UNBIND, in almost the same manner it processes UNBIND with BIND forthcoming. When the normal UNBIND is received, a positive response is sent and the SLI waits for a new BIND from the Primary LU. NOTIFY DISABLED is not sent to the SSCP as it is in normal processing. If an UNBIND type X'01' is received during SLI_OPEN processing following BIND, the LU-LU session is suspended until a new BIND, optional CRV and STSN, and SDT are received. At this time the SLI_OPEN verb will complete. The SLI_OPEN verb may be terminated prematurely by issuing SLI_CLOSE abend. If an UNBIND type X'01' is received in the middle of an SLI session (after SLI_OPEN has completed and before SLI_CLOSE has been issued), the LU-LU session is suspended until a new BIND, optional CRV and STSN, and SDT are received. If an SLI_RECEIVE or SLI_BID is active when the UNBIND arrives, it will complete with primary return code STATUS, secondary return code NOT_READY. If no SLI_RECEIVE or SLI_BID is active the next SLI_RECEIVE or SLI_BID issued will complete with these return codes. After SDT +RSP has flowed, the next active SLI_RECEIVE or SLI_BID will complete with primary return code STATUS, secondary return code READY. SLI_CLOSE abend will terminate the SLI session prematurely. If UNBIND type X'01' is received during primary or secondary initiated SLI_CLOSE normal processing, the SLI_CLOSE verb completes with primary return code CANCELLED, secondary return code RECEIVED_UNBIND_NORMAL. The LU-LU session is suspended until a new BIND, optional CRV and STSN, and SDT are received. After SDT +RSP has flowed, the next active SLI_RECEIVE or SLI_BID will complete with primary return code STATUS, secondary return code READY. SLI_CLOSE abend will terminate the SLI session prematurely. Field description: lua_session_type Defines how the SLI should process UNBIND type X'01', UNBIND NORMAL. The valid values follow: LUA_SESSION_TYPE_NORMAL (= 0) When an UNBIND NORMAL is received from the Primary Logical Unit, the SLI sends a positive response and issues RUI_TERM which causes a NOTIFY DISABLED to flow to the SSCP. The SSCP-LU flow is disabled. This is the default value for this parameter. LUA_SESSION_TYPE_DEDICATED (= 1) When an UNBIND NORMAL is received from the Primary Logical Unit, the SLI sends a positive response and the SLI session is suspended until a new BIND, optional CRV and STSN, and SDT commands are received. The SLI does NOT issue RUI_TERM in this case and NOTIFY DISABLED does NOT flow to the SSCP. The optional lua_session_type parameter tells the SLI how to process UNBIND type X'01', UNBIND NORMAL. This parameter takes effect after the SLI_OPEN verb passes initial parameter checking and stays in effect until SLI_CLOSE abend is issued or until the SLI issues RUI_TERM. o Standard UNBIND NORMAL Processing Set the lua_session_type parameter to LUA_SESSION_TYPE_NORMAL for this option. This is the default value. With this option the SLI sends a positive response to an UNBIND NORMAL sent by the Primary Logical Unit and issues RUI_TERM which causes a NOTIFY DISABLED to flow to the SSCP. These actions: 1. Destroy the LU-LU session. 2. Indicate to the SSCP and PLU that the SLU is unable to process new BINDs. New BINDs which are received are rejected. 3. Prevent data from flowing on the SSCP-LU session. The SLI will issue RUI_TERM when it receives any UNBIND except type X'02', UNBIND with BIND forthcoming. o Dedicated UNBIND NORMAL Processing Set the lua_session_type parameter to LUA_SESSION_TYPE_DEDICATED for this option. With this option the SLI sends a positive response to an UNBIND NORMAL sent by the Primary Logical Unit but does NOT issue RUI_TERM. The status of the SSCP-LU session is unchanged (enabled). The SLI session is suspended until a BIND, optional CRV and STSN, and SDT commands are received. An SLI session which is waiting for a new BIND can be terminated by issuing SLI_CLOSE abend. The SLI will issue RUI_TERM when it receives any UNBIND except type X'02', UNBIND with BIND forthcoming, or UNBIND type '01', normal UNBIND. This option is useful when the Primary Logical Unit is unable to send an UNBIND with BIND forthcoming, but desires this type of behavior when UNBIND NORMAL is sent. Two new return codes have been added as follows: o 0000004F INVALID_SESSION_TYPE Explanation: An SLI_OPEN verb contained an invalid value in the lua_session_type parameter. Programmer response: Correct the SLI_OPEN verb and use one of the valid choices for the session type: - LUA_SESSION_TYPE_NORMAL - LUA_SESSION_TYPE_DEDICATED o 00000057 RECEIVED_UNBIND_NORMAL Explanation: During primary or secondary initiated SLI_CLOSE normal processing for a session started with a SLI_OPEN verb which specified lua_session_type equal to LUA_SESSION_TYPE_DEDICATED, SLI received an UNBIND type X'01'. The session is suspended until a BIND, optional CRV and STSN, and SDT commands are received. Any user extension routines that were supplied by the original SLI_OPEN verb are called again; these routines must be reentrant. After the SLI processes the SDT command, the SLI session resumes. Programmer Response: Issue an SLI_BID or SLI_RECEIVE verb to receive the READY status when the SLI session resumes. Note that the application can continue to issue SLI_SEND and SLI_RECEIVE verbs for SSCP normal flow data. ΓòÉΓòÉΓòÉ 3.4. User SDT Responses in LU0 SLI Interface ΓòÉΓòÉΓòÉ In the LU0 subsystem the SLI interface of LUA will now allow users to send their own SDT response. This is done by the addition of a new SLI extension routine entry point, SLI_SDT_ROUTINE, which can be activated by defining a user supplied DLL and routine via the SLI_OPEN verb with routine type set to LUA_ROUTINE_TYPE_SDT (X'04'). If an SDT routine is not supplied and an SDT request is received, the SLI issues a positive response to the SDT. The documentation for the SLI_CRV_ROUTINE may be used to describe the new SLI_SDT_ROUTINE verb by simply substituting "SDT" for "CRV" throughout that documentation. The SLI_SDT_ROUTINE is completed when the user's extension routine returns with the lua_prim_rc and the lua_data_length parameters set in the SLI_SDT_ROUTINE verb control block. Overwrite the SDT RU with the SDT response. A primary return code of OK indicates that the SDT was accepted. If the user's routine rejects the SDT, it should set the primary return code to NEGATIVE_RSP and put the sense code in the SDT buffer. A negative response from this routine cancels the SLI_OPEN verb. The SLI returns a primary return code of X'000F', SESSION_FAILURE, X'0000003D', NEG_RSP_FROM_SDT_ROUTINE. NEW ERROR DOCUMENTATION: 0000003D NEG_RSP_FROM_SDT_ROUTINE Explanation: The user-provided SLI SDT routine generated a negative response to an SDT. This condition causes the SLI_OPEN verb to end. Programmer Response: If the SDT is unacceptable, correct the host problem. If the SLI SDT routine returned a negative response in error, correct the SLI SDT routine. ΓòÉΓòÉΓòÉ 3.5. LUA Correlation Table Limit Overrides ΓòÉΓòÉΓòÉ The correlation tables used to track request and responses maintained within the LUA storage are limited in size. This size restriction is reflected in the maximum number of outstanding request chains that may be sent or received on each flow. LUA applications that send many SNA LU-LU normal exception response chains to the primary logical unit (PLU) may be affected by the LU-LU normal send correlation table limits. Whenever a correlation table becomes full the RUI_WRITE (SLI_SEND) verb will fail with lua_prim_rc=UNSUCCESSFUL and lua_sec_rc=SEND_CORRELATION_TABLE_FULL. Two options have been added to RUI_INIT and SLI_OPEN to relieve the restrictions for the LU-LU send normal correlation table. These options are LUA_CT_MODE and LUA_CT_LU_NORM_SEND_SIZE. Setting LUA_CT_MODE=LUA_WRAP_LU_NORM_SEND_CT will make the LU-LU normal send correlation table appear to have an unlimited size (the table will not become full). Using this option will only have an effect if the application sends exception response chains to the PLU. Sending many definite response chains to the PLU will fill the correlation table until the responses are received from the PLU. The LUA_CT_LU_NORM_SEND_SIZE option allows the application to size the LU-LU normal send correlation table for its particular need. This option will have the same effect when sending either definite or exception response chains. Both options may be used simultaneously. LUA_CT_MODE=LUA_WRAP_LU_NORM_SEND_CT places the LUA LU-LU normal send correlation table in a "wrap" mode. The default is not to wrap. When the table is full, the "wrap" mode reuses the oldest correlation table entry with the following characteristics: o RU_CTGY = FMD o exception response chain o complete chain has been sent to the PLU When LUA receives a response to a request where the corresponding correlation table entry has been reused, the LU-LU session is terminated because of a protocol violation. All outstanding LUA verbs (synchronous and asynchronous) for this LU are terminated with lua_prim_rc=SESSION_FAILURE and lua_sec_rc=RSP_CORRELATION_ERROR. A new session must be established to continue communications. The LUA_CT_LU_NORM_SEND_SIZE option is a value indicating the maximum number of chains the correlation table can hold. This value is multiplied by a factor of fifty to yield the actual size of the correlation table. For example: o If lua_ct_lu_norm_send_size=O (the default) 50 correlation table entries o If lua_ct_lu_norm_send_size=1 LUA allocates 50 correlation table entries o If lua_ct_lu_norm_send_size=2 LUA allocates 100 correlation table entries o If lua_ct_lu_norm_send_size=10 LUA allocates 500 correlation table entries LUA does not allocate one table large enough to hold all the entries. Instead, smaller blocks of storage are allocated to "build up" to the maximum size requested. For example, if the LUA application specifies LUA_CT_LU_NORM_SEND_SIZE=20 (20 x 50 = 1000 entries) and only 200 chains are sent, then enough storage will be dynamically allocated to hold 200 entries even though 1000 were requested. The default LU-LU normal send correlation table size is 50 entries (chains). The correlation table blocks are freed when a RUI_TERM verb is issued or when a BIND is returned for an RUI_READ. ΓòÉΓòÉΓòÉ 3.5.1. Guidelines for using LUA_CT_MODE and LUA_CT_LU_NORM_SEND_SIZE: ΓòÉΓòÉΓòÉ o Use lua_ct_mode=LUA_WRAP_LU_NORM_SEND_CT when error recovery is not important in the context of a session and to minimize the storage requirments of the correlation tables. o Use lua_ct_lu_norm_send_size to size the correlation table when the LUA application needs to process all responses for error recovery. Both options may be used for a session. For example, to reduce the risk of a negative response received after the corresponding request has been removed from the correlation table, the table can be increased in size while allowing it to "wrap". ΓòÉΓòÉΓòÉ 3.6. EHLLAPI Virtual Device Driver ΓòÉΓòÉΓòÉ Windows support. The EHLLAPI virtual device driver (VHAPI) allows both existing DOS and Windows EHLLAPI applications to run in the OS/2 virtual DOS machine (VDM) environment. This applet provides a Windows EHLLAPI application support in addition to the ES 1.0 VHAPI equivalent support. This document contains the following information. o Installation o Usage o Windows EHLLAPI function call syntax o VHAPI system structure ΓòÉΓòÉΓòÉ 3.6.1. Installation ΓòÉΓòÉΓòÉ 1. Make a backup file for existing VHAPI.OS2 if necessary. 2. Copy VHAPI.OS2 from the diskette to a CMLIB sub-directory. 3. Add the following statement to the CONFIG.SYS file: DEVICE=d:\CMLIB\VHAPI.OS2 Where d is the drive on which Communication Manager is installed. To automatically add this statement to your CONFIG.SYS file, type the following command at the OS/2 command prompt: VHAPINST Note: Use the appropriate path indicating where CMLIB has been installed. You must restart your workstation after making changes to the CONFIG.SYS. To remove the DEVICE= statement from your CONFIG.SYS file, type the following command at the OS/2 command prompt: VHAPINST /D Note: You must restart your workstation after making changes to the CONFIG.SYS. ΓòÉΓòÉΓòÉ 3.6.2. Usage ΓòÉΓòÉΓòÉ To execute a Windows EHLLAPI application in the WIN-OS/2 session, PCSHLL.DLL provided by this package is prerequisite. It needs to be called by a Windows EHLLAPI application running in the WIN-OS/2 session. VHAPI accepts up to 16 hllapi sessions at a time. If an EHLLAPI application attempts to establish the 17th hllapi session, VHAPI returns the system error code (rc=9). Note that this DLL file is the interface module between a Windows EHLLAPI application and the OS/2 Communications Manager. If running in the pure Windows environment, use the DLL path indicating where the Personal Communications/3270 has been installed. ΓòÉΓòÉΓòÉ 3.6.3. Windows EHLLAPI function call syntax ΓòÉΓòÉΓòÉ PCSHLL.DLL only accepts the same syntax as Personal Communications/3270 Version 2.0 provides, though the Version 3.0 offers the additional syntax for EHLLAPI function call. void FAR PASCAL hllapi(LPINT, LPSTR, LPINT, LPINT); /*import hllapi*/ hllapi ((LPINT) &iHLLAPIFunction, /*pointer to EHLLAPI function number */ (LPSTR) zDataString, /*pointer to Data String */ (LPINT) &iStringLength, /*pointer to Data String length */ (LPINT) &iReturnCode );/*pointer to return code */ ΓòÉΓòÉΓòÉ 3.6.4. VHAPI system structure ΓòÉΓòÉΓòÉ The following illustrates the VHAPI system structure. VDM_1 (WIN-OS/2 Desktop) VDM_2(WIN-OS/2 Window) VDM_3(DOS Window) ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé Γöé Γöé Windows ΓöéΓöé Windows Γöé Γöé Γöé Γöé Windows Γöé Γöé Γöé Γöé DOS Γöé Γöé Γöé Γöé EHLLAPI ΓöéΓöé EHLLAPI Γöé... Γöé Γöé Γöé EHLLAPI Γöé Γöé Γöé Γöé EHLLAPI Γöé Γöé Γöé Γöé appl.[1]ΓöéΓöé appl.[2]Γöé Γöé Γöé Γöé appl.[3]Γöé Γöé Γöé Γöé appl.[4]Γöé Γöé Γöé ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé ΓööΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé Γöé Γöé Γöé Γöé Γöé PCSHLL.DLL Γöé Γöé Γöé Γöé PCSHLL.DLLΓöé Γöé Γöé Γöé Γöé Γöé Γöé ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé ΓööΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé Γöé Γöé Γöé ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöéΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöéΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöéΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöéΓöÇΓöÇΓöÇΓöÇΓöÿ Γöé Γöé Γöé Γöé Γöé Γöé INT 7FhΓöé ΓöéIRET INT 7FhΓöé ΓöéIRET INT 7FhΓöé ΓöéIRET AX=0104hΓöé Γöé AX=0104hΓöé Γöé AX=0104hΓöé Γöé Γöé Γöé Γöé ΓöîΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÉ Γöé EHLLAPI Virtual Device Driver (VHAPI) Γöé Γö£ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö¼ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöñ Γöé EHLLAPI[1]Γöé ΓöéEHLLAPI[2]Γöé ΓöéEHLLAPI[3]Γöé ΓöéEHLLAPI[4]Γöé........... Γöé Γö£ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö┤ΓöÇΓöÇ ΓöÇ ΓöÇ ΓöÇ ΓöÇ ΓöÇ ΓöÇ Γöé Γöé LLAPI Γöé Γö£ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöñ Γöé 3270 Emulator Γöé ΓööΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇ[Communications Manager]ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÿ ΓòÉΓòÉΓòÉ 4. Things that Changed ΓòÉΓòÉΓòÉ The following sections describe changes that have been made to the product's executables, publications, and messages/helps. ΓòÉΓòÉΓòÉ 4.1. CM SNA Services ΓòÉΓòÉΓòÉ o JR05659: SYS0032 Attempting To Copy PROTOCOL.INI SYS0032 error when OS/2 Extended Services V1.0 programmable configuration utility attempts to copy PROTOCOL.INI. o JR05679: SYS1201 on Boot of SDLCDD.SYS Using the driver SDLCDD.SYS that was supplied with Extended Services, boot of the system gets SYS1201 'Device Driver SDLCDD.SYS was not Installed' and start of Communications Manager gets ACS0121 'Hardware Error on SDLC Adapter' when using the model 645111A MPA card. o JR05777: CTRL+BRK OR SLI_CLOSE NOT ENDING LU0 SESSION When a workstation has an LU0 session (going through an OS/2 ES gateway) with the host (IMS) and the user needs to reboot the system without a normal close of the LU0 application, the sessions always remain active with IMS. o JR05830: Programmable Configuration Default RAM Address When using the programmable configuration utility, if the parameter RAM=blank (no address) is coded, it automatically adds RAM=D800 as default in the PROTOCOL.INI file. o JR05856: Improper Update Of Alert Hierarchy Resource List HRL in an alert should not be updated if the resource name is only the local CP name and Resource Type=CP/SP/PU. o JR05857: Out-of-Memory Condition Polling For SSCP-PU Session Comm Mgr abends with an error log indicating that it is out of unlocked storage when PNA polls for an active SSCP-PU and multiple GET_PCIDS are enqueued. o JR05864: Access Violation Using Programmable Config Access violation when executing the following programmable configuration command: rc = lan(get, ieee8022 eq "0",, 'adapt_type_' eq adapt_type) The access violation does not occur if LAN Adapter and Protocol Support is configured for IBM token ring but does occur if LAPS is configured for Western Digital Ethercard and Ungermann-Bass. o JR05865: BCH0039 Using Programmable Configuration When using programmable configuration, the verification fails and messages are logged in the programmable configuration log: BCH0039: 3270SNA Verify Session 5 Field mnemonic AT_KEYBOARD must be specified. BCH0039: 3270SNA Verify Session 5 Field mnemonic ENH_KEYBOARD must be specified. o JR05901: Error Msg SYS1201 Booting OS/2 2.0 Error msg SYS1201 when booting OS/2 2.0 with one of the following devices installed: MicroGate SDLC modems models SV32, MV32, MG96, MG144 or USA IBM ASYNC/SDLC V.32 modem/A Part #93F0973. o JR05902: Trap D Specifying Existing File Name CM ES 1.0 programmable configuration receives Trap D when specifying an already existing file name. The trap occurs when using a REXX procedure with the programmable configuration API of CM to create a configuration file. o JR05989: Trap D SYS3175 Creating Configuration File Programmable configuration can run short of memory during execution. PC should detect the problem and report it instead of trapping. PC will generate message BCH0048E and BCH0057E as well as REXX return codes 48 and 57 when a memory problem occurs. o JR06038: Trap D Handling MS Application A Customer Application communicating between multiple end nodes (EN) and a network node (NN) using the SEND_MDS_MU API of ES 1.0 receives multiple Trap D's in the end node and the network node usually hangs. o JR06063: OS/2 Must be Rebooted After DB2 is Shutdown OS/2 must be rebooted after recycling DB2 or a communications error will occur when trying to reconnect. o JR06065: TG Defined Incorrectly in Topology Database ACT_ON_DEMAND link to NN reported as local not network in toplogy database. o JR06072: Lu_Name not LU_Alias of Default_Local_LU Returned The DISPLAY VERB documentation indicates that the System default Information should return the default_local_lu_alias when it is selected. However, it returns the lu_name of the default LU rather than the lu_alias of the default LU. o JR06121: Defect in Wildcard Routing in an End Node ES 1.0 fails to go on looking for wildcard definition if NN server search cannot find requested LU. o JR06122: Server Hangs with 30+ OS/2 Requestors LAN Transport was changed to allow only 8 data buffers in ES 1.0 because of changes to support the NDIS interface. The MAX RU size on the server and workstations must match. These changes affect the support for maximum conversations supported per adapter and possibly performance. o JR06123: Unsolicited IPM Causes Lost Data When ES receives an unsolicited IPM with NWS=1, it resets its windows and never sends out anything else to VTAM. The window that ES should have is 1, so ES should have been able to send out one RU, but it didn't. o JR06127: National Chars $ and @ Not Accepted in PLU Name National chars $ and @ are not accepted in PLU name. When attempting to enter the fully-qualified partner LU name on the 'Creating LU to LU Security' panel in SNA feature profile, SNA network definitions, advanced SNA features. The only national character accepted is '#'. The HELP panel states $ and @ are also valid. o JR06142: No Bind Acknowledgement if Notify Off-Line Received CM ES 1.0 gateway is not sending a negative response to a bind in progress if a notify off-line is received. o JR06144: APPC Link not Re-established If an APPC link is configured to the PREFERRED_NN server and that link is lost, then an attempt should be made to reconnect. o JR06199: Unable to Use DMA Port After CM Used After Communications Manager is used, the highspeed DMA port cannot be used. If Communications Manager is never started, there is no problem using the port with other communications packages. If they start CM they can use the port with CM but if CM is stopped and they try to access the port there are tildes inserted between each character in the echo back of the modem command that was sent. o JR06207: FSI Hangs After Using Programmable Configuration Communications Manager full screen interface (FSI) will hang after using the verify function via programmable configuration. User must shut CM down and reboot the machine and bring CM FSI back up and check the errors. o JR06241: X.25 Adapter Failure With Message ACS0786C Random error message ACS0786c issued (the X.25 API detected an error on adapter) and gateway and X.25 hang. The DCE continues sending RR frames but the gateway stops responding. o JR06245: XID Failure Occurs With 10160005 XID failure occurs between a S/36 and ES/1. The XID failure indicates that the adjacent node does not support bind segment generation but does support receipt of bind segments. Any T2.1 node supporting receipt of bind segments must also support generation of bind segments. o JR06276: SYS1201 - Unable to Load SDLCDD.SYS SYS1201 (unable to load SLDCDD.SYS driver for ES 1.0) on PS/1 2123 because the Model 2123 is not in the Model Table, and because this machine has a PC-AT (ISA) bus. o JR06289: System Will Not Boot After Verify of a CM .CFG Verification of a Communications Manager Configuration Caused Config.Sys to be rearranged such that the system will not boot. o JR06300: CM Fails to Unbind the SNASVCMG Session In a CM configuration for a DEPENDENT LOCAL LU, CM is soliciting a BIND (INIT_SELF) for SNASVCMG. Next CM presents a CNOS to the INDEPENDENT PARTNER LU. Then CM fails to UNBIND the SNASVCMG session. This causes the ALLOCATION REQUEST to hang. o JR06305: Input Lost from CM SNA Profile Config Panel CM SNA Features configuration panel 'loses' node information added for a second DLC profile. o JR06318: Trap 000D, Exception in Device Driver LPODD$01 MIS server trapping with 000D. Trap followed by exception in device driver LPODD$01. At the time of the failure the customer had 10 data bases running on the server and was in the process of adding two more data bases. o JR06319: Internal Processing Error At Boot Up Time An internal processing error occurs at boot up time when RDSLANDD.SYS is included in a CONFIG.SYS device statement. o JR06329: VHAPI Returns 200 Bytes when Specifying 100 Using VHAPI supplied with ES 1.0, EHLLAPI DOS application call Copy_PS_To_String with a 100 byte parameter. VHAPI returns 200 bytes. The customer's buffer is set up for 100 bytes, and this causes an overlay of other data. o JR06343: Config File Unverified After Set Keylock If in a REXX file for Programmable Configuration, you verify the config file and then set the Keylock, the file becomes unverified according to Communications Manager. A subsequent attempt to use COPYCFG to copy this configuration fails. o JR06354: Sense 080F6051 on Bind Request When the PLU and the SLU are configured with the same name, ES rejects the bind request with Sense Code 080F6051. Other symptoms may include Sense codes 083500XX (XX points to the primary LU or secondary LU field), 08410001 or 08410002. o JR06356: Sense 0835FFFF on Bind Request CICS rejects a bind request from ES for a single session LU with a sense code of 0835FFFF because the session ID control vector is not included in the user data of the bind request sent to CICS. o JR06359: Trap D at NN Doing Directed Search Trap D when establishing an intermediate session. Trap occurred when processing the LOCATE_REPLY when the NN on each end sets the KEEP flag in the LOCATE GDS variable (byte 4) and when the reply is complete reply indicator is set (also byte 4 of the LOCATE GDS variable). o JR06360: ISR Trap Routing Session Control Response Receiving a Start Data Traffic response signal on an intermediate session, Trap is received in ICPPMU (Session Connector Manager) when trying to process the signal. o JR06361: ES Changing CHAIN_FLOW_REQUEST Discard When a LOCATE/CDINIT/FIND is sent with the byte after the LOCATE GDS variable (X'12C4') equal to X'10', meaning that the LOCATE is a CHAIN-FLOW with the KEEP_LOCATE_CHAIN field set to DISCARD, ES is changing the REQUEST_REPLY_CHAIN_STATUS to COMPLETE_REPLY causing a protocol error. o JR06363: LOCATE Rejected With 08900010 LOCATE rejected with 08900010 after LOCATE with FLAGS='C0' because an ES node thinks it is no longer in session with the adjacent control point. o JR06365: Padded Blanks Not Allowed in SLU Name ES 1.0 issues a BIND -rsp with sense code 80040000 after receiving a BIND with a secondary LU name which is padded with blanks. If the SLU name is not padded with blanks, the BIND is accepted. o JR06364: Trap Processing LOCATE/CDINIT Attempting to establish conversation from NN through ES1.0 NN to NN, the session for the user-node is successfully bound. However, a Locate/Request is received with the flags byte after the X'12C4' of the LOCATE GDS equal to X'10'. This is not responded to or forwarded by the ES1.0 NN. Eventually the ES1.0 NN traps in DNP149. o JR06401: LUA Large File XFER Stops With No Errors Sending large files via LUA over SDLC from host, the transmission suddenly stops. Traces available indicate that the session hangs during an outstanding LUA API request. o JR06409: Sense 80050000 When CICS Recycled Gateway does not reconnect LU0 when the CICS host is recycled. CM SNA trace shows the LU0 sending NOTIFY ENABLE to the GW and the GW machine rejects with sense 80050000. CICS host does not receive the NOTIFY from the remote LU0 and the session cannot be established again. o JR06428: Error Type 0014 Sense Code 80050000 When a link is established via connection network, an error type 0014 with Sense Code 80050000 is issued each time the application ends. This results in a performance degradation. Same error type occurs using APPCTELL sample. o JR06457: FMTTRACE Traps Formatting Bind When formatting a BIND request with LU type 0 in byte 15 (in the PS Usage field), FMTTRACE abends at cs:ip of 0047:017f (trap 000d). o JR06460: SYS1201 Loading SDLC Driver The SDLCDD.SYS device driver is not loading during system boot on some OEM/PCM machines. The result is a SYS1201 error message. o JR06522: Sense Code 80140001 Routing to LEN When connecting from NN/EN ---> NN ---> LEN, if the link had been active, was deactivated, and then reactivated, the allocation of the session fails with a sense code of 80140001 (resource found, but no route exists). This can also happen when routing from the LEN to NN to another node. o JR06546: Invalid FQ_NAME in PLU List FQ_NAME in PLU list of connections panel gets the 2nd NETID when the partner network id is changed. This problem exists when the FQ_PARTNER_LU_NAME in DEFINE_PARTNER_LU and DEFINE_PARTNER_LU_LOCATION only contain an LU_NAME versus a NETWORKID.LUNAME. o JR06551: Config.Sys Garbage The order of the Config.Sys statements during VERIFY are rearranged when a configuration change is made in Communication Manager that does not add or remove features (file with no changes). o JR06556: NN Fails to Update Topology Properly ES NN fails to update topology properly when connected to DPPX/370 network. When an attempt is made to establish an LU-LU session with an LU on one of the DPPX/370 NNs, sense code 80140001 is received. The ES NN is able to use the DPPX/370 NNs to search the NN, but is not able to use them to route a session since the topology database is not properly updated. o JR06557: TRAP D at CS:IP=XXXX:2151 in APPC When running CP-CP sessions between ES and VTAM V4R1, ES may Trap D at CS:IP=XXXX:2151 in APPC (DS Component), CSLIM=E28A. The trap occurs in DNP059, where APPC tries to remove the CV81 and CV82 from the FIND GDS Variable. The LOCATE GDS Variable is X'12C4'. o JR06562: Start of CM Fails with 232+ LUs Configured If more than 232 pools are defined in Host Connection of Gateway Profile, then a record that is longer than 1024 bytes is created in the .CFG file for CM. The result is CM fails to start. Symptoms include: APPC Internal Error (ACS0109C), error log X'0026', X'00000001'. o JR06563: Null GDS Variable Returns Invalid Return Code Null GDS variable with conditional end bracket (Dealloc_Normal) returns Bad_State return code, not Dealloc_Normal return code. Symptoms include Sense Code 000412F1, Return Code 0002 000000B1. o JR06567: Downstream Workstation Connection Fails Downstream workstations attached to a CM ES 1.0 Gateway will not connect with a VTAM host session. Workstations get a 2B? on their screen with no error messages logged. VTAM on the host shows the 3270 LUs in a PACTL state. o JR06571: X25.CFG Migration Fails Migration of X25 configuration file from EE 1.3 to ES 1.0 fails using the CMUPGRAD command, ESINST, or REINST. The X25 routing tables are discarded. This only works when using the ES 1.0 migration menu if different names are given as INPUT and OUTPUT configuration files. o JR06574: Sense FFFF0003 or 80140001 on ALLOCATE Verb When an application program receives an attach from an implicit partner LU, it cannot then ALLOCATE a session back to the implicit partner LU. o JR06576: CANT_RAISE_SESSION_LIMITS Return Code on CNOS Attempt to reset sesson limits from a non-zero value to a zero value back to a non-zero value using CNOS for an intra-LU session fails with RC-STATE_CHECK, CANT_RAISE_SESSION_LIMITS (X'0002', X'00000151'). An intra-LU session is a session from a local LU back to itself, the same LU. o JR06608: MSG APN0009C for LINK_NAME Beginning With # If the LINK_NAME of the DEFINE_LOGICAL_LINK verb starts with a '#' sign, the system returns a message APN0009C when you start CM with this NDF file. Internal error found in configuration file c:\cmlib\appn\xxx.cf2. Verb code:3600 return code:0001 000001c1. o JR06615: Correlation Table Limit Overrides Return code of 0014, sense code 08780002 response when LUA Correlation Table exceeds its limits. This forces a user to use definite response mode on the request which affects the performance of the system. o JR06619: LUA Application Loops Between SLI_OPEN & CLOSE LUA application gets stuck in following loop:SLI_OPEN (completes 0014, UNSUCCESSFUL, 00000035, SESSION_ALREADY_OPEN) and SLI_CLOSE (completes 0002, STATE CHECK, 00000034, NO_SLI_SESSION) o JR06622: GDS Data Corrupt Following FMH5 Length 0E While processing an FMH5 of length 0E with following GDS data stream in same RU, CM appears to be corrupting the first byte of the LL field in the ensuing GDS data stream. o JR06628: Vital Product Data Incorrect When Netview queries OS/2 1.3 for vital product data, OS/2 shows ACS0448W (timeout) and Type FFFF 00000006 and 0017. o JR06634: -RSP Sent by LUA Causes Memory Corruption LUA RUI stops responding after a -RSP is sent. RUI_WRITE requests returns with 0 return code without sending data. LUA trace ends abruptly and SDLC Trap 000D occurs. o JR06635: CMC-006 Running CM Batch Configuration Utility Message CMC-006 received on an IEEE 802.2 (LAN) adapter profile record (identified by the characters LN2) that start the record. Message received when running CM Batch Configuration Utility. o JR06639: SLI_SEND of a Large RU Fails SLI_SEND of a large RU fails with 0014 20020000 when the max_ru_size specified in the BIND forces the SLI component to split the RU into a chain of more than 256 elements. o JR06643: Dependent LUs Registered by EN Rejected When dependent LUs are registered by EN, VTAM rejects as a duplicate LU. ES allows LUs to be both dependent and independent and will accept NAU addresses that are non-zero and respond positively when a LOCATE request is issued. o JR06644: Sense Code 10030023 Not Forwarded by ES Sense code of X'10030023' was received in the LOCATE GDS variable of a LOCATE_REPLY. This is not forwarded on the LOCATE GDS variable sent to the originating CP of the LOCATE because it is not in the list of expected return codes for LOCATE processing. This hangs the session that issued the LOCATE. o JR06646: Sense Code 08050000 on BIND Request BIND request sent down by the Host for a single session fails with 08050000 because the CNOS to set the session limits has not completed. This is a timing problem due to the order of the CNOS verbs in the .CFG file. o JR06649: Unexpected String Value 'PIP_ALLOWED' Migration from CM definitions used with the OS/2 EE 1.3 + NS/2 to ES 1.0 fails with the following messages if the VERIFY is performed via the panel functions: ACS0148C The function in progress ended due to a system error in MESSAGE.DAT: APN0403E Command syntax error detected on line xx of the ndf. APN0517E Unexpected string value 'PIP_ALLOWED'. If using the CMUPGRAD.EXE instead of the panel functions, the migration will hang, in which case no messages will be reported. o JR06655: DEFINE-TP FileSpec Padded Incorrectly A FileSpec specified on a DEFINE_TP verb that is padded with X'00' is accepted but not handled correctly when attach is received. APPC assumes that it will be padded with ASCII blanks (X'20') in determining the end of the program name. o JR06662: SLI_SEND Fails with 0030 200D0000 SLI_SEND fails with 0030 200D0000 when secondary LU receives a BID, BEGIN_BRACKET, or BEGIN_CHAIN. The SLI_SEND cannot continue sending the chain and issues RUI_TERM. This causes NOTIFY disabled and UNBIND to be sent to terminate the session. o JR06663: Outstanding Notify Command Not Cleared ACTLU type ERP does not issue HIERARCHICAL_RESET commands to clear outstanding NOTIFY commands. Type COLD issues the reset but this cannot be selected under VTAM configuration. Using SLI_CLOSE to cancel an SLI_OPEN operation will hang the cancelling operation. o JR06664: SDLC Stops Sending/Receiving Test Frames CM SDLC device driver stops responding to test frames while using an OEM device to maintain an active link over SDLC. This would force a re-IPL of the PS/2. o JR06701: Trap D When Re-IPLing the Host Trap D when the DACTLU request is sent down from the host to take down the SSCP-LU session and the normal DACTLUs in the RU are 9 bytes or larger. o JR06707: Message BCH0037 During Verify Incorrect verify on the definition of link_name in LDIR results in message BCH0037. Correct verify is to disallow the definition of the same link_name more than once in a local directory. o JR06715: Trap D Shutting Down DBM at Server Using KILLDBM utility supplied by Database Manager to stop database manager at the Database Manager server at the same time a START USING DBM at the requestor is being done, causes the server to trap (CSLIM=F8FC, IP=6050). o JR06736: ATTACH Rejected with Sense 10086011 Sense 10086011 is issued when the ACCESS_SECURITY length field following the TP name in the ATTACH is missing or contains a value too large to be handled correctly. o JR06747: Trap D When 080A0000 Received Trap D occurs at IP-A094 in code segment with CSLIM-F14C. The CM trace indicates that an FMH-7 is received with a sense code of 080A0000. This occurred after Receive_and_Wait was issued to force data to be sent. o JR06789: SLI Close Call with Normal Flag Fails SLI_CLOSE normal verb completes with primary return code x'0014', UNSUCCESSFUL, secondary return code x'0000007F' SLI_LOGIC_ERROR. Error Type 0049 Subtype 000060007 is logged by CM RUI_WRITE verb. o JR06790: Handle Type 1 UNBIND Similar to Type 2 LUA SLI_OPEN has an added option: Type 1 UNBIND can be handled similarly to a Type 2 UNBIND. The new option allows UNBIND X'3201' to be processed by SLI in the same fashion as an UNBIND x'3202'. o JR06792: Repeated UNBIND/BIND Sequence Fails Repeated UNBIND X'3202'/BIND sequence fails. No BIND +RSP is sent by LUA and the SLI application hangs. o JR06836: No Retry Limit Set if CLEAR 00 From X25 Network On X25 network (SVC) sends continuous CALL without any possibilities to set a retry limit. If for any reason the host comes down (hardware trouble...etc), the Gateway receives a CLEAR from TRANSPAC (SVC circuit) then Commmunications Manager issues a CALL to the host. As the request can't be delivered to it's destination, TRANSPAC sends a CLEAR 00 to the Gateway and Communications Manager reissues a CALL. o JR06840: Shared RAM Location Written Incorrectly CID or programmable configuration used to define LAN IEEE8022 has a parameter 'Shared RAM Location' for which the naming syntax allows only 4 hexadecimal characters, A000-F000. When this gets written to PROTOCOL.INI the format should be '0xNNNN' but writes NNNN instead. o JR06899: BCH0001 (RC=1) Running Programmable Config Cmd File Running programmable configuration command file results in BCH0001 message (return code = 1). The FILE ADD fails with the following in the MYCONFIG.LG: Job MYCONFIG Processing Has Started. BCH0001: File Add MYCONFIG.INI The file MYCONFIG.INI was not found. o JR06946: PC for 802.2 Adds NETBEUI Section in Error If Programmable Configuration is used to configure 802.2 then a NETBEUI section is also added to the PROTOCOL.INI though it should not be. All the keywords in the NETBEUI section are ending with just the =character (no parameters). After doing a REINST with this PROTOCOL.INI, the DEVICE= statements are added to CONFIG.SYS, but the device drivers are not copied to the system. This results in not found messages for the device drivers when the system is booted. o JR06948: Large File Transfer Using NvDM/2 Hangs When transferring a large file from NvDM/2 to NvDM on the Host using CM, the file transfer stops after having transferred about 1MB of data. The RU size does not seem to matter. This is observed using adaptive pacing between CM and NCP. One of the contributing factors is the window size given by NCP to CM. Adaptive pacing is used between NCP and CM. Additional symptoms include: slow response time, link outages, sessions lost, one CM waiting on an IPM from another CM that is waiting on an available frame to respond to the IPM, CM waiting on an available frame to respond to an IPM, and unable to send large file (1meg+) from the NvDM/2 node to the host after installing Release 4 at the host and XR20249 on the PC. o JR07012: Free Unused Link Parm Default Incorrect CM CID install uses the default (ACSCFGUS) configuration file as the model configuration. This model does not contain a Token Ring DLC profile and the customer's CID response file did not contain the FREE_LINK keyword. This causes CM CID to default the free unused link parm of the output DLC profile to YES when it should have been NO. o JR07289: Loop Between SLI_OPEN and SLI_CLOSE SLI_OPEN verb failure causes an LUA application to loop between SLI_OPEN and SLI_CLOSE. The LUA application is unable to recover the session. o JR07338: Trap When Receiving Attach After BIS Reply Trap in APPC at offset 4EC0 by receiving machine during file transfer following an Attach. The trace shows a BIS request being received from the machine that is receiving the files followed by a BIS request (reply). The BIS request is followed by an Attach received from the machine receiving the files. The trap occurs during processing of this Attach. o JR07339: Rejected Incoming PIP Data When PIP Data comes in from the host, it is rejected. The PIP data should be allowed and received by the user on the next RECEIVE and WAIT after being notified on a RECEIVE ALLOCATE that it is present. o JR07361: Traps in RDSLAN on DDCS Gateway using SQLL00 when Multiple Clients Use Database. Random TRAP D's can occur when using SQLL00 with multiple clients accessing a database. The number of clients being able to log on varies with the RU size of the server. With an RU size of 4096, up to 30 clients could use the database before the problem occurred. With an rusize of 1024 up to 45 clients could use the database. Similarly, increasing RAM on the server may increase the number of clients to log on. ΓòÉΓòÉΓòÉ 4.2. 3270 Services ΓòÉΓòÉΓòÉ o JR05829: RC=16 During Programmable Config Verify Customer is running PC and receives msg: BCH0116 START_3270_APPC is selected to autostart but no session is selected to autostart. The verify step fails though the configuration will verify successfully under CM. o JR05926: Begium Keyboard Profile Copy Problem Within the keyboard profile choice list found in ACSCFG, the second choice is normally for Belgium but appears incorrectly the first line overlapping the --ALL-- choice. Selecting the first line results in copying all the keyboard profiles for the specific emulation EXCEPT the Belgian one. This happens for both the 3270 and the 5250 keyboard profiles. o JR05959: First Line Of Page Missing When Printed First line on page is missing when printing host print file to HP LASERJET SERIES II because application is printing beyond the printing space allowed by HP SERIES II printer. o JR05964: 3270 Emulation Window Restored to Incorrect Size This problem occurs as follows: 1. Size and position 3270 emulation window 2. Minimize window (to minimized window viewer) 3. Restore the window (double click on icon in minimized window view or window list). The window will be restored with the upper left corner of the window in its initial position, but the size will be the size of a full screen window. o JR06073: Title Bar Moved out of Viewing Area When a 3270 session is full screen size, the title bar goes to the top of the screen. When the pointer is 'clicked' mid screen, the screen moves up about 1/3 the height of the title bar line causing a portion of it to move up out of the viewing area. Problem occurs using OS/2 2.0 and ES 1.0. o JR06227: Msg IKT00405I issued in CM 3270 Emulation IKT00405I message is issued when in CM 3270 emulator session to an application via VTAM, TSO, and GDDM. The IKT00405I is presented by the host because CM rejected a PIU with a sense x'1003'. The PIU contained an outbound 3270 data stream but the 3270 data stream appears to be correctly built. o JR06243: ADMCHART Hangs on PF3 Exit ADMCHART hangs when attempting to exit via a PF3 key after file transfer receive is executed. o JR06244: Incorrect Output on Underscored Text Host directed print to d/t4019 printers get incorrect output on underscored text. A few characters of a word will be underscored, but the rest of the underscore will print on the next line. o JR06254: PROG753 in OIA PROG753 (repeated character is invalid) appears in OIA and host data was not displayed. o JR06269: Host Print Causes Line Wrapping CM 3270 Host Print causes line wrapping with 4029 printer at 600 DPI and 17 pitch. Output appears with 1.5 inches line width instead of 8 inches or more. o JR06314: 3270 Emulator Window not Closed from Task List When user selects 'close' option in the task list window on OS/2 2.0, the 3270 Emulator window is not closed. o JR06357: Alternate APL Character Set Ignored When using CM EHLLAPI, the alternate APL character set is ignored when specifying a horizontal line or vertical bar. A lower case 's' is returned and displayed for a horizontal line and a lower case 'e' is returned and displayed for a vertical bar. o JR06414: Customization lost on System Re-boot When the system is booted up, it does not remember the customization that was set up before shutdown. This seems to happen only with the first session. o JR06420: SYS3175 on 3270 Print Control Print server for 3270 print control fails with SYS3175 trap. Problem occurs when two or more LP sessions try to send and receive messages and they are using the same memory location. The result is one message sends an invalid message in a different LP. The LP processes the invalid message which usually has the HDFG_INBOUND bit set. This causes the LP to take an error path and bypass notifying PRD to close the job. The Trap D results when the TQ header for RestoreKBBuf was corrupted. o JR06507: Invalid Find Field Length The EHLLAPI call Find Field Position (32) on a field with 2 spaces separating 2 attributes results in the pointer being 1 less than the correct value, thus providing invalid results when the Find Field Length (31) is requested. o JR06520: 209 Error When 3174 Port Hangs When CM is stopped or a shutdown is performed, CM will go through a terminate routine. When connected to a 3174 via DFT, this routine terminates and the adapter closes before the 3174 has a chance to respond to the AEDV offline. This causes the port on the 3174 to go offline or hang which requires an IML of the control unit. This normally occurs on faster PS/2s. The 3174 will display a 209 error. o JR06528: EHLLAPI Function 10 Returns System Error (9) Using EHLLAPI Function 10, QUERY SESSIONS, to provide session information on the active 3270, sessions correctly returns 0 active sessions if CM is active but without sessions. If a session is started and then stops, QUERY SESSIONS returns 0 active sessions with a RC9 system error. Issuing the SYSTEM RESET (21) the session appears but EHLLAPI QUERY SESSIONS still gives RC9 incorrectly. o JR06585: WS Binds as MOD2 but Acts Like MOD5 Running a CICS transaction which performs DYNAMIC QUERY to determine the presentation space size of the workstation, the workstation is configured as a MOD5, bound as a MOD2, but acts as a MOD5 instead of a MOD2. o JR06653: No Print Jobs Sent to Printer Servers Intermittently EMAIL print jobs are not being received by the print server; the printer server is not hung and will continue to print LAN based PM prints, but no host directed prints. The only way to get the jobs to restart printing is by bouncing the associated LU for that printer. If you look at the EMAIL print job on the host you will see a number of jobs lined up in the queue for printing. o JR06677: RC24 for EHLLAPI Search Presentation Space Search presentation space (6) fails with return code 24 (the search string was not found) running sample program HLLSAMP. On PC/3270 a string containing blanks or nulls as the match string is found when doing a Search_PS(6) call but the same string on CM/2 or ES 1.0 CM is not found. o JR06679: Print Control Window Ends After SYS3175 SYS3175 (Access Violation) causes 3270 print control window to end after it is responded to. o JR06690: EHLLAPI fails on a START HOST NOTIFICATION(23) After many START HOST NOTIFICATION commands are sent to the host session, EHLLAPI fails with a system error(return code 9) and the host session is hung. o JR06691: Incorrect code page value return on Print Query The EBCDIC code page value in the LP character set QUERY REPLY is byte reversed. o JR06692: Host print fails after INFOPAK job prints Host directed print fails after printing from INFOPAK. Various host jobs can be printed (CICS, NETPASS, INFOPAK...), but print jobs fail following an INFOPAK job. Error and message logs are empty. o JR06693: 2002 With CICS Host Directed Print During some print jobs 3270 LU1 printer session sends a -RSP 2002 to the host erroneously indicating a SNA chaining error has occurred. When CICS receives the -RSP, it will UNBIND the LU and no longer use it until manual host operator intervention occurs. o JR06694: Problems printing using GPI mechanisms When a line being processed is longer than what can be held in the buffer (maximum length = 200), a buffer overflow can occur wiping out data. Symptoms include corruption of print spooler files, and multiple page ejects within a job. o JR06695: Sense 1001 on CM host print Host printing to VM 3268 is successful but printing via CM fails with sense code 1001. Symptoms include print error "Forms Mismatch", 2002 bracket error and loss of data. o JR06730: Updates to PS or OIA Not Sent DOS PC/3270 sends the data to the EHLLAPI application for every update on the PS or OIA (Host update or Terminal update), but OS/2 CM Query_Host_Update EHLLAPI commands send the data to the EHLLAPI application for updates on the PS or OIA to the host only. o JR06745: SDLC Lines will drop and recycle without cause on 50MHZ DX machines. During normal operations the SDLC link to the host will drop. The host PU will indicate an INOP condition and the workstation will have errors in the ERROR.DAT file (Type 17, Subtype 00000002, Alert ID 0E2DDF11). "Secondary Receive Inactivity timer". NCP Line traces show periods of inactivity from the workstation in the for of multiple retries on RR frames in which there is no response. Typically the links will eventually recover via normal link level recovery (XID exchange) after the alert is logged by CM. The problem can also have Type 17 Subtype 00000002 Alert ID 15C2CCE5 log entries. o JR06759: 3270 Print Control Hangs and Traps SYS3175 or trap in ACS3ETQ.DLL while performing Host Print. Hangs may occur with following symptoms: 3270 Print Control shows STARTED but no files are sent or 3270 Print Control shows STARTED yet Host Print files are queued on the Host waiting to be sent. o JR06767: Extra Page Eject on LU3 Printers Host print sends X'0C0D' form feed/carriage return at the end of each job in order to reposition print for next job. The extraneous form feed causes an extra page eject. o JR06775: Extra Blank Lines Added to 3270 Hostprint On a hostprint job directed by CICS to an OS/2 2.0 ES/LS 3270 non-DFT printer session, hostprint emulation adds several lines of spaces where there should be only one line of spaces. The same hostprint job when directed to a 3287 printer or via ATTACHMATE (instead of OS/2 ES) prints as intended. o JR06787: Trap 000D When Installing GDDM-OS/2 SYS3175 trap 000D if REXX is not installed on a PS/2 and running GQFINST to install GDDM-OS/2 link. o JR06799: Trap 000D with LS 3.0 386 HPFS Installed When starting Communications Manager, ACS3EINI gives a Trap 000D on a system which has Lan Server 3.0 and 386 HPFS Local Security installed. o JR06803: ES Loses 3270 Window Sizing and Positioning The problem shows up in two steps: 1. 3270 terminal session windows are sized and placed on the desktop, then all the windows are minimized, and when the sessions are taken from the window list or the minimized icons window, they are restored to the positions and sizes set prior to minimizing; and, 2. if the session windows are minimized and the workstation 'shutdown' , when the workstation is rebooted, the previous sizing and placement of the session windows is lost. o JR06838: IKT00405I for 3270EM Session With Geomanager IKT00405I screen erasure caused by error recovery procedure. This message is issued when in CM 3270 emulator session to an application via VTAM, TSO, and GDDM. CM rejects the PIU with sense code x'1003'. The PIU contains an outbound 3270 data stream which is correctly built. o JR07000: IKT00405I for 3270EM Session See APAR comments for JR06838. Additional symptoms include segmentation problem of separated length field of structured fields. o JR07011: 3270 Print control fails in ACS3ETQ.DLL This APAR addresses two problems, a Trap or SYS3175 occurs in the ACS3ETQ.DLL, and 3270 Hangs while using 3270 Print Control. The print control problem has two symptoms: 1. A job Started but nothing is sent to the workstation. 2. Or nothing has started but yet jobs are queued on the host This APAR supersedes JR06420. o JR07337: CM 3270EM Returns Sense Code 2003 Bracket error, sense code 2003, returned by CM/3270EM when other 3270 program doesn't return end of bracket at the same data flow. Both use 'Bracket Termination Rule-1'. ΓòÉΓòÉΓòÉ 4.3. 5250 Services ΓòÉΓòÉΓòÉ o JR05646: BCH0113,BCH0116,BCN0016 Using Programmable Config When not all 5250 sessions are configured for autostart using ES 1.0 programmable config utility, verification fails with errors, BCH0113, BCH0116, and BCN0016 in the job log. But if the config file is set to autostart all sessions, then the verification runs clean. o JR05926: Begium Keyboard Profile Copy Problem Within the keyboard profile choice list found in ACSCFG, the second choice is normally for Belgium but appears wrongly on the first line overlapping the --ALL-- choice. Selecting the first line results in copying all the keyboard profiles for the specific emulation EXCEPT the Belgian one. This happens for both the 3270 and the 5250 keyboard profiles. o JR05942: SYS3175 Printing to 5250WSF Printer SYS3175 access violation at 00016FE4 ACS5PRNT.EXE 0001 CSLIM=000091D7 when printing from AS/400. o JR06160: Caps Lock Light Reversed in 5250WSF Caps Lock enabling becomes reversed when a user switches the sessions after pressing caps lock key. o JR06476: TWINAX Connection to AS400 Fails Twinax connection to an AS400 fails when using a machine with an ISA or AT bus. User gets a 400 signon screen, enters id and password, but can't get 5250WSF connection completed. ACS0797, ACS0653 are received and error log contains 0017 00000002 ALERT 7F1C7AF8. o JR06489: 5250 Automatic Session Recovery 5250 work station sessions to the AS/400 are not re- established when the link to the AS/400 is recovered. The 5250 emulator session recovery is described in more detail in the README file. o JR06536: Problem with Numlock in 5250 Emulation This problem occurs in Windowed Mode only. Num lock function does not work after Shift lock is ON when French keyboard AZERTY 102 keys remapped. o JR06604: SYS3175 Printing with Unsupported Font When 5250WSF receives an unsupported font (in this case font 254), message SYS3175 is issued, the AS/400 job ends, and the AS/400 device descriptions get damaged and must be recreated. o JR06625: 5250 WSF Data Shifts Positions to the Right Incorrect formatting occurs when printing the first print line of the first print job at 15 CPI. This error occurs only when attempting to print at 15 CPI on a PC (OS/2) connected printer. o JR06626: Escape sequence inserted Print files spooled from AS/400 to OS/2 workstation gets an escape sequence inserted from CM in error. o JR06630: Copy_to_PS Function Generates Wrong Mnemonic EHLLAPI call COPYSTRINGTOPS interprets a "@1" as a keyboard mnemonic giving RC=05 when CM running in 5250 Emulator. Works correctly in 3270 Emulation. o JR06632: French 5250 Keyboard Remapping Incorrect The French OS/2 ES 1.0 5250 Keyboard Mapping for keyboard ID 189 displays the English pound sign incorrectly as "*". The key can be reconfigured to map correctly, but shipped defaults are incorrect. o JR06877: 5250WSF Hang Running TWINAX to AS/400 5250WSF hangs while using various AS/400 functions. Failure occurs while using the PF4 key (PROMPT) within SEU (System Editor Utility) only on workstations using the ISA BUS and with the TWINAX APAR JR06476 applied. ΓòÉΓòÉΓòÉ <hidden> ΓòÉΓòÉΓòÉ \OS2\INSTALL\APPNV.CTL \OS2\DLL\ACS3ECT.DLL \OS2\DLL\ACSGCBLD.DLL \OS2\DLL\ACSNLDBC.DLL \OS2\DLL\ACSSVC.DLL \OS2\DLL\APPNCFG.DLL \OS2\DLL\APPNCFG2.DLL \OS2\DLL\APPNCFG3.DLL \OS2\DLL\APPNW_M.DLL \OS2\DLL\APPNUTL.DLL \OS2\DLL\CMISPFPC.DLL \OS2\DLL\RCFGFIL1.DLL \OS2\DLL\RCFGFILE.DLL \OS2\DLL\RCOMAIN.DLL \OS2\DLL\RCVMAIN.DLL \OS2\DLL\RPSMAIN.DLL \OS2\DLL\RCVERMIG.DLL \OS2\DLL\RCVSHELL.DLL ΓòÉΓòÉΓòÉ <hidden> ΓòÉΓòÉΓòÉ The item in the title line is a trademark of IBM Corporation ΓòÉΓòÉΓòÉ <hidden> ΓòÉΓòÉΓòÉ The item in the title line is a trademark of Microsoft Corporation