SEARCH CONTACT US SUPPORT SERVICES PRODUCTS STORE
United States    
COMPAQ STORE | PRODUCTS | SERVICES | SUPPORT | CONTACT US | SEARCH
gears
compaq support options
support home
software & drivers
ask Compaq
reference library
support forum
frequently asked questions
support tools
warranty information
service centers
contact support
product resources
parts for your system
give us feedback
associated links
.
} what's new
.
} contract access
.
} browse patch tree
.
} search patches
.
} join mailing list
.
} feedback
.
patches by topic
.
} DOS
.
} OpenVMS
.
} Security
.
} Tru64 Unix
.
} Ultrix 32
.
} Windows
.
} Windows NT
.
connection tools
.
} nameserver lookup
.
} traceroute
.
} ping
OpenVMS VAXSCSI03_061 VAX V5.5-2-V6.1 SCSI Drivers ECO Summary

Copyright (c) Digital Equipment Corporation 1994, 1995. All rights reserved. PRODUCT: OpenVMS VAX COMPONENTS: SCSI Drivers (DKDRIVER.EXE, MKDRIVER.EXE, PKBDRIVER.EXE, PKCDRIVER.EXE, PKIDRIVER.EXE, PKRDRIVER.EXE, PKSDRIVER.EXE) SPDTDEF.MAR SOURCE: Digital Equipment Corporation ECO INFORMATION: ECO Kit Name: VAXSCSI03_061 ECO Kits Superseded by This ECO Kit: VAXSCSI02_061 VAXSCSI01_061 (CSCPAT_1179) VAXDRIV09_U2055 VAXDRIV10_U2055 VAXDRI03_2H4055 VAXDRIV03_060 VAXDRIV04_060 VAXDRIV03_061 ECO Kit Approximate Size: Saveset A - 108 Blocks Saveset B - 288 Blocks Saveset C - 342 Blocks Saveset D - 288 Blocks Saveset E - 342 Blocks Total of 5 files - 1410 Blocks Kit Applies To: OpenVMS VAX V5.5-2, V5.5-2H4, V6.0, V6.1 System Reboot Necessary: Yes NOTE: OpenVMS VAX V5.5-2H4 is a limited hardware release, shipped only with the new systems (or system upgrades) listed below. It is not separately orderable and will not be distributed via Consolidated Distribution. o VAX 4000 Model 100A o VAX 4000 Model 500A o VAX 4000 Model 600A o VAX 4000 Model 700A NOTE: PKIDRIVER and PKSDRIVER are CPU port drivers. The CPUs that support these drivers are: o VAX 4000 Model 100 o VAX 4000 Model 200 o VAX 4000 Model 300 o VAX 4000 Model 300 o MicroVAX 3100 Model 90 To determine if these drivers exist on your system, perform the following steps. 1. Invoke SDA (VAX/VMS System Analyzer) by typing ANALYZE/SYSTEM at the DCL command prompt. 2. At the SDA> prompt, enter SHOW DEVICE PK. You will get a display that shows you the PK controllers and which PKxDRIVER you have on your system. 3. To return to the DCL command prompt, type EXIT at the SDA> prompt. ECO KIT SUMMARY: An ECO kit exists for SCSI Devices (DKDRIVER, MKDRIVER, PKBDRIVER, PKCDRIVER, PKIDRIVER, PKRDRIVER, PKSDRIVER) on OpenVMS VAX V5.5-2 through V6.1. Problems Addressed in the VAXSCSI03_061 Kit for OpenVMS VAX V6.0 and V5.5-2H4: o In certain heavily-loaded situations, TZ30 transfers can time out with the current DMA TIMEOUT of 3 seconds. The fix increases the DMA TIMEOUT in SCSI_CMD for the port driver, to 10 seconds. Problems Addressed in the VAXSCSI03_061 Kit for OpenVMS VAX V6.1 and V5.5-2H4: o Under heavy load, the fifteenth synchronous data transfer may not be counted by the NEC 53C94 chip. This will cause the process associated with the transfer to hang or for a SCSI bus reset to be issued and mount verification invoked. o Following a SCSI bus reset, read commands which can fail due to the bus reset cause system code to bugcheck. This is caused by a Page Read Error problem. Problems Addressed in the VAXSCSI02_061 Kit for OpenVMS VAX V5.5-2, V5.5-2H4, V6.0, and V6.1: o Some devices reply to Inquiry commands with the response data format field set to SCSI-1 (as opposed to CCS or SCSI-2). This setting causes these devices to be marked as offline. Problems Addressed in the VAXSCSI02_061 Kit for OpenVMS VAX V6.1: o Magnetic tapes with odd byte block sizes cause the DUMP utility to report data overrun errors. Problems Addressed in the VAXSCSI02_061 Kit for OpenVMS VAX V5.5-2H4 and V6.1: o The port driver does not properly detect and handle the queue full condition for TCQ operation. o Under heavy loads the command complete interrupt for an AEN finish command operation can occur prior to the fork level code releasing the port. This causes the interrupt to be misinterpreted and an INCONSTATE bugcheck occurs. o When a process executes a Diagnose command which results in a SCSI check condition and exits without requesting sense data, DKDRIVER will reject all future Diagnose commands with SS$_CHANINTLK errors. Problems Addressed in the VAXSCSI02_061 Kit for OpenVMS VAX V5.5-2: o The SCSI magtape class driver does not allow for the skipping of files by filemarks instead of by records. o The system may crash when I/O is canceled on SCSI tape devices. Problems Addressed in the VAXSCSI01_061 Kit for OpenVMS VAX V5.5-2, V5.5-2H4, V6.0, V6.1: o When multi-volume backups are performed using data compaction, the compaction switch is turned off after completion of the first volume. Problems Addressed in the VAXDRIV10_U2055 Kit for OpenVMS VAX V5.5-2 (PKIDRIVER and PKSDRIVER): o Data corruption might occur on read operations for segmented transfers. Problems Addressed in the VAXDRIV05_U2055 Kit for OpenVMS VAX V5.5-2 (PKIDRIVER and PKSDRIVER): o PKIDRIVER and PKSDRIVER have a number of odd-byte transfer problems, showing up as the last byte of a read being corrupted on certain transfer sizes and segmented transfers. Problems Addressed in the VAXDRIV09_U2055 Kit for OpenVMS VAX V5.5-2 (PKCDRIVER): o When the PKCDRIVER times out waiting for a re-selection interrupt, the correct value is not restored to R5. This causes the system to crash. Problems Addressed in the VAXDRIV08_U2055 Kit for OpenVMS VAX V5.5-2 (PKBDRIVER, PKCDRIVER, PKRDRIVER, and MKDRIVER): o Under a heavy SCSI load, the Receive Diagnostic Command to the TZK50 causes the SCSI bus to fail with a GROSS error. This command is used to retrieve the hardware/firmware revision level from the TZK50. o Even though data read or written over the SCSI bus is correct and the data read into the user buffer is correct, Write/Read Check errors occur for one byte data transfers. o If the PCKDRIVER times out while waiting for a reselection interrupt, the system will crash due to an invalid R5 value. Problems Addressed in the VAXDRIV06_U2055 Kit: o The disconnect timeout period specified in the MKDRIVER is not long enough to allow an INIT/ERASE operation to complete on a 90 meter tape (TLZ06). This causes ERASEFAIL and CTRLERR errors to be returned. For example: $ INIT/ERASE MKA500: test %INIT-W-ERASEFAIL, volume not completely erased -SYSTEM-F-CTRLERR, fatal controller error %INIT-F-CTRLERR, fatal controller error This problem is fixed in OpenVMS VAX V6.0. o When performing a BACKUP to a TK50 tape on a SCSI drive, POSITERR and DRVERR errors are returned if an attempt is made to add a second saveset on the same tape. For example: $ BACKUP/LOG R*.COM MKB500:ONE.SAV/SAVESET/NOREWIND/LABEL=TEST $ BACKUP/LOG S*.COM MKB500:TWO.SAV/SAVESET/NOREWIND/LABEL=TEST %BACKUP-F-POSITERR, error positioning MKB500:[000000]TWO.SAV; -SYSTEM-F-DRVERR, fatal drive error NOTE: When this condition occurs, the system error log also will contain the following port error: SCSI PORT SUB-SYSTEM, UNIT_MCY3$PKB0: ERROR TYPE 0906 SCSI BUS PHASE ERROR PHASE CHANGE TIMEOUT DURING COMMAND OUT SCSI ID 05 o A DRVERR error has always been reported in MK for BLANK CHECK errors. However, when initializing a new tape volume during a multi-volume BACKUP, this error causes the BACKUP operation to fail. To fix this problem, an OPINCOMPL error is now returned. This error message allows backup to completed and is also compatible with other non-SCSI drivers. o In certain heavily-loaded situations, TZ30 transfers could time-out with the current DMA TIMEOUT of 3 seconds. This fix increases the DMA timeout in SCSI_CMD for the port driver to 10 seconds. Problems Addressed in the VAXDRIV02_U2055 Kit for OpenVMS VAX V5.5-2 (PKBDRIVER, PKCDRIVER, PKRDRIVER, and MKDRIVER): o When multiple logical units (LUNs) are connected on a SCSI device, the device may enter into a Mount Verification state. o The status returned when issuing a QIO to a powered-off device (one which was powered on long enough to be configured properly) through PKCDRIVER (SS$_CTRLERR) is inconsistent with that returned by PKNDRIVER (SS$_TIMEOUT). o Coding errors in PKCDRIVER cause unexpected errors when used with user written class drivers. o The PKCDRIVER register dump routine increments the address of the error log buffer as a byte (INCB) instead of a longword (INCL). This makes it possible for the address to wrap around at the low byte boundary and corrupt the error log buffer. o A single TurboChannel interrupt occurs twice on the VAXstation 4000 Model 90. The VAXstation 4000-90 CPU latches the TurboChannel interrupt and the port driver does not clear the latched bit. As soon as the interrupt is serviced and the IPL is dropped, the interrupt shows up again. o The peak 5Mb/Second data rates aggravate signal integrity issues in certain maximum length cable configurations on the VAXstation 4000 Model 90 and the VAXstation 4000 Model 60. o Code from PKNDRIVER which is imported into PKCDRIVER at PK$ABORT causes various problems, particularly with command queueing enabled. Problems Addressed in the VAXDRI03_2H4055 Kit for OpenVMS VAX V5.5-2H4 (DKDRIVER, MKDRIVER, and PKxDRIVER): o The last byte is sometimes corrupted on read operations for segmented transfers. o Devices which reply to Inquiry commands with the response data format field set to SCSI-1 (as opposed to CCS or SCSI-2) are mistakenly marked off line. A SONY[R] optical disk drive has recently been found which responds with the SCSI-1 setting instead of the CCS setting. o Magnetic tapes with odd byte block sizes cause the DUMP utility to report data overrun errors. o The fork block at the end of the SPDT data structure is not receiving an initial fork IPL/lock value. This results in intermittent system crashes. Problems Addressed in the VAXDRI02_2H4055 Kit for OpenVMS VAX V5.5-2H4 (DKDRIVER, MKDRIVER, PKDRIVER, and PKSDRIVER): o DKDRIVER Mode Sense, Mode Select and other fixes: + Some newer Seagate drives are off-line at boot time. These drives worked with OpenVMS VAX V5.5-2 and OpenVMS VAX V6.0. + Mode sense data errors occur while configuring or mounting some SCSI devices, particularly those which support Tagged Command Queueing. + Multiple Mode Select failure device errors occur configuring and accessing disks through an HSC40. + I/O operations which complete with a sense key of 00 cause the I/O to fail and the device to be marked off-line. A SWAPPER bugcheck sometimes occurs. The error log will show a CHECK CONDITION returning a sense key of 00. + When using the IO$_DIAGNOSE interface to DKDRIVER, normal access to the device will sometimes stall even if no CHECK CONDITION has occurred. o MKDRIVER/PKCDRIVER Fixes: + MKDRIVER: - The Receive Diagnostic Command in the Packet Acknowledgment sequence to a TZK50 causes the SCSI bus to fail with a GROSS error during the DATA IN phase of the command from the target. This error condition is only reproducible with heavy SCSI load. - MKDRIVER does not check for canceled I/O before attempts are made to use the IRP in the IO_SENSECHAR: routine. + PKCDRIVER: - Write/Read Check errors occur for one-byte transfers even though the data read or written over the SCSI bus is correct and the data read into the user buffer is correct. - When the driver times out waiting for a re-selection interrupt, the correct value is not restored to R5. This causes the system to crash. o PKCDRIVER Fixes: + A SCSI command is read back as data but is not detected as such by the driver. + On an invalid map interrupt at the label INVALID_MAP_INT, the driver overwrites R1, which contains a valid value for EXE$GL_CPUNODSP. This causes a crash when the mainline driver code references an offset off R1. + When the DISP_INTERRUPT macro is invoked, the illegal command (ILL_CMD) branch is re-tried instead of taking the GROSS_ERR branch. o There is a PKSDRIVER odd byte problem in the DATAI_PHASE routine. o MKDRIVER: Provide a check for canceled I/O before attempting to use the IRP in the IO_SENSECHAR routine. o PKCDRIVER corrupts the first 24 bytes of the SPDT by using it as a fork block when it forks to flush an I/O queue on an error (SCSI reset). (SPDT$L_FQFL). Problems Addressed in the VAXDRIV04_060 Kit for OpenVMS VAX V6.0 (PKIDRIVER and PKSDRIVER): o Data corruption might occur on read operations for segmented transfers. Problems Addressed in the VAXDRIV03_060 Kit for OpenVMS VAX V6.0 (PKBDRIVER, PKCDRIVER, and PKRDRIVER): o When the PKCDRIVER times out waiting for a re-selection interrupt, the correct value is not restored to R5. This causes the system to crash. Problems Addressed in the VAXDRIV01_060 Kit for OpenVMS VAX V6.0 (PKBDRIVER, PKCDRIVER, and PKRDRIVER): o The peak 5Mb/Second data rates aggravate signal integrity issues in certain maximum length cable configurations on the VAXstation 4000 Model 90 and the VAXstation 4000 Model 60. o Code from PKNDRIVER which is imported into PKCDRIVER at PK$ABORT causes various problems, particularly with command queueing enabled. Problems Addressed in the VAXDRIV03_061 Kit for OpenVMS VAX V6.1 (DKDRIVER, PKIDRIVER, and PKSDRIVER): o I/O operations which complete with a sense key of 00 cause the I/O to fail and the device to be marked off-line. A SWAPPER bugcheck sometimes occurs. The error log will show a CHECK CONDITION returning a sense key of 00. Problems Addressed in the VAXDRIV02_061 Kit for OpenVMS VAX V6.1 (DKDRIVER, PKIDRIVER, and PKSDRIVER): o Data corruption might occur on read operations for segmented transfers. INSTALLATION NOTES: In order for the corrections in this kit to take effect, the system must be rebooted. If the system is a member of a VAXcluster, the entire cluster should be rebooted. [R] SONY is a registered trademark of SONY CORPORATION.



This patch can be found at any of these sites:

Colorado Site
Georgia Site



Files on this server are as follows:

vaxscsi03_061.README
vaxscsi03_061.CHKSUM
vaxscsi03_061.CVRLET_TXT
vaxscsi03_061.a-dcx_vaxexe
vaxscsi03_061.b-dcx_vaxexe
vaxscsi03_061.c-dcx_vaxexe
vaxscsi03_061.d-dcx_vaxexe
vaxscsi03_061.e-dcx_vaxexe

privacy and legal statement