Updated: 24-SEP-2003 (Use your browsers' Reload button to ensure you're viewing the most recent version)
ALPBASE02_071 OpenVMS Alpha V7.1 Y2K Compliance ECO Summary
Copyright (c) Compaq Computer Corporation 1998, 1999. All rights reserved.
**********************< NOTES >****************************
* *
* If you have previously installed ALPBASE01_071, you do *
* not need to install ALPBASE02_071. 17-JUN-1998 *
* *
* If you are running volume shadowing on your system, *
* ALPSHAD05_071 must also be installed to avoid problems *
* with writing crashdumps to shadowed disks and Boot *
* driver initialization failure errors after running *
* shutdown. *
* *
***********************************************************
New Kit Date: 05-MAY-1998
Modification Date: 25-MAR-2002
Modification Type: Made available on ftp site.
OP/SYS: OpenVMS Alpha
COMPONENTS: BACKUPSHR SYS$DE500BADRIVER
DECC$SHR EXCEPTION
DUMP EXCEPTION_MON
EXCHANGE EXEC_INIT
F11AACP CLUE$SDA
F11BXQP IO_ROUTINES
LIBRSHR IO_ROUTINES_MO
MTAAACP LAN
STABACKUP LANACP
TECOSHR LANCP
VERIFY PROCESS_MANAGEMENT
VMS$REMEDIAL_ID PROCESS_MANAGEMENT_MON
STARLET.OLB SDA
SYSBOOT SDA$SHARE
VERIFY VMS$REMEDIAL_ID
SOURCE: Compaq Computer Corporation
ECO INFORMATION:
ECO Kit Name: ALPBASE02_071
ECO Kits Superseded by This ECO Kit: ALPBASE01_071
ALPACRT04_071 ALPF11X02_071 ALPLAN01_H1071 ALPSYS01_071
ALPACRT03_071 ALPF11X01_071 ALPSYS02_071
ALPACRT02_071
ALPACRT01_071
ECO Kit Approximate Size: 36,900 Blocks
Kit Applies To: OpenVMS Alpha V7.1, V7.1-1H1, V7.1-1H2
System/Cluster Reboot Necessary: Yes
Rolling Re-boot Supported: Yes
Installation Rating: 1 - To be installed on all systems running
the listed versions of OpenVMS.
NOTE: Inclusion of ALPY2K01_071 Functionality
The ALPBASE01_071 remedial kit includes all functionality
that was shipped with the ALPY2K01_071 remedial kit.
Therefore, after the installation of the ALPBASE02_071
remedial kit, your system will be Y2K compliant. There
is no need to install the ALPY2K01_071 remedial kit.
In addition, configurations that have Volume Shadowing
must also install the ALPSHAD04_071 or its subsequent kit
to ensure that all dependent images are installed.
Kit Dependencies:
The following remedial kit(s) must be installed BEFORE
installation of this kit:
None
In order to receive all the corrections listed in this
kit, the following remedial kits should also be installed:
None
ECO KIT SUMMARY:
An ECO kit exists for Multiple Corrections and Year 2000 Compliance
on OpenVMS Alpha V7.1 through V7.1-1H2.
Known Problems:
o A regression has been found in the ALPACRT04_071 and
ALPBASE01_071 kit where a DECC read() function fails to read an
entire buffer. The read() function will now only read the
maximum of bytes as determined by the SYSGEN parameter
DEFMBXMXMSG in a single call. By default, this is 256 bytes.
Expecting reads larger than DEFMBXMXMSG was not previously
assured, according to the DEC C RTL Reference Manual, which
states:
"This function returns the number of bytes read. The return
value does not necessarily equal nbytes. For example, if the
input is from a terminal, at most one line of characters is
read. The read function does not span record boundaries in a
record file and, therefore, reads at most one record. A
separate read must be done for each record."
If this problem is encountered, the workaround is to change the
dynamic SYSGEN parameter DEFMBXMXMSG to be larger than the
largest expected read, or to recode the application to properly
handle the return value from read().
A fix for this problem is currently being designed and will be
available in the ALPACRT05_071 and later remedial kits.
o DECWINDOWS Motif users that do not use the new interface,
(i.e, decw$start_new_desktop = "FALSE") must also install
the ALPMOTF02_U4012, or later, remedial kit to insure that
the workstation will start correctly after installation of
the ALPBASE01_071 kit. Failure to do so will result in a
DECW$SERVER crash that will prevent you from using your
system.
If the decw$start_new_desktop logical is set to TRUE, the new
desktop will start correctly, even after the installation of
this kit.
o You must NOT attempt to re-install the ALPACRT01_071 through
ALPACRT04_071 nor the ALPY2K01_071 remedial kits after
installing the ALPBASE01_071 kit. Installation of these kits
will lead to a regression in the CRTL object library. If these
kits are installed after ALPBASE01_071, then you must reinstall
ALPBASE01_071 or ALPACRTL05_071 (or later).
Problems Addressed in ALPBASE02_071:
o The EXCHANGE.EXE, VERIFY.EXE and F11BXQP.EXE images in the
ALPBASE01_071 remedial kit had installation conflicts with
the ALPY2K01_071 remedial kit. This kit corrects those
conflicts.
Problems Addressed in ALPBASE01_071:
o If a user mode program causes a user mode exception, rather
than the image simply exiting with an ACCVIO, the system
crashes with a SSRVEXCEPT bugcheck.
o The display of the signal array on image exit was garbled.
o A multithreaded process with upcalls enabled may hang in HIB
due to a missing event flag upcall. If a program makes heavy
use of event flags for synchronization, notification of an
event flag being set can sometimes be lost. This can result in
a thread waiting forever for an event which already took place.
o The SDA$SHARE and SYS$BASE_IMAGE images which were shipped in
ALPSHAD03_071 result in an SDA warning when SDA is used.
o Additional system routine vectors to support new ATM connection
manager support shipped with V7.1-1H1.
o File system integrity could be compromised if a device is
mounted /NOSHARE on one system and as the member of a shadow
system disk on another system.
No "Alloc. lock ID" is setup on the booting system.
o SYSBOOT is enhanced to check foreign boot media, when the
foreign boot flag is set, for SYS$CPU_ROUTINES_xxxx.EXE before
looking on on the system disk. This allows the SHIP/3PB
process to be used to supply new platform-specific code that
replaces existing code.
o Bugcheck if CONFIG.DAT is missing on a foreign boot.
o Bugcheck if CONFIG.DAT is empty on a foreign boot.
o Boot HALTs on some platforms with the floppy as a secondary
boot device.
o Error message text that was not referenced by code has been
removed.
o If a system with a shadowed system disk is shutdown and
rebooted, the same CLUE HISTORY output (and AUTOCLUE data
mailed via DSNlink) can be produced multiple times.
o Corruption can occur in the heap storage used by SDA to old its
binary tree of symbol names and values. The result is an
ACCVIO, apparently from the first CLUE command issued.
o SDA doesn't handle relocatable global (non-universal) symbols
correctly if they are in resident images.
o System Crash due to use of section table offset in CCB$L_WIND.
o Because UCB structures might vary in size, CLUE first reads in
an octaword to determine how large the UCB really is, followed
by another read of the whole UCB into the local buffer.
Currently, this local buffer is 8KB in size. However, on a
system with some laser disks connected to SCSI, the UCB is over
11KB in size. This will result in CLUE corrupting some
internal data storage, causing CLUE to report %SDA-E-BADMAJVER
error in subsequent commands.
o HELP in Alpha SDA mostly reflects pre-V6.1 functionality.
o SSRVECEXPT Bugcheck during Boot at F11FR$CACHE_LOOKUP_C+00010.
Current process name is LANACP.
o In an OpenVMS cluster, some clients would configure a SCSI
device, some would not.
o Synchronization problem in CONFIGURE opens a crash timing
window during DEV_SCAN routine. This causes an ACCVIO
exception bugcheck.
o OpenVMS could potentially create processes in the same group
UIC with the same process name.
o When a Satellite is booted over the DE425, and a system
shutdown is requested, the system will hang on the way down
waiting for access to the system disk. This problem can affect
any Alpha system with an EISA bus and a DE425.
o Fix for corruption of LEC ARP event queues resulting in
INVEXCEPTN bugchecks.
o The 3C589 Ethernet Card, revision D, requires more time for
reset than the original revision C card.
o Some NCL commands in DECNET/PLUS cause systems to crash
with INVEXCEPTN, Exception while above ASTDEL at location
SYS$NETWORK_SERVICES offset 0000EA8C, EMAA$BUILD_RESPONSE_C+00B1.
o Applications that start/stop rapidly and repeatedly on the LAN
driver, cause the network management code to create additional
data structures with CSMACD$EW_xxxxxxxx (where xxxxxxx is that
new UCB address created). After running the application,
non-paged pool is consumed and never returned. This is only a
problem using network management NET$CSMACD and NET$FDDI.
While starting/stopping an application interface to LAN driver,
after 100 start/stops, with client name structures having
CSMACD$EW_xxxxxxxx, you can see that pool is consumed. This is
only a problem using network management NET$CSMACD, NET$FDDI,
etc.
o A multi-segment transmit that has a buffer that crosses the page
boundary between unmapped and mapped physical memory may result
in a DMA error.
o Oversize packets are being issued by the EWDRIVER devices.
This packet would most likely be discarded by the receiving
adapter as oversized and the node the packet was being
transmitted to may timeout waiting to receive it.
o VCI (VAX Communication Interface) users are not able to
determine if a port has become unuseable. This prevents them
from switching to another adaptor.
o After transmit timeout (soft error handling), processing of the
alias addresses is not re-enabled in the SET_FILTERS routine.
No packets would be received by that adapter.
o DECnet IV starts but does not work, (i.e., no adjacency message
results.
o During error handling it is possible for the device to be
started with the correct address but for the driver to think
that a different address is in use. If that happens, the
driver throws away receive packets and applications don't work
unless they are transmit-only. This only happens if DECnet is
being used on the system (or some other application that changes
the address.
o This kit adds DE205 boot support for the following configurations:
Ethernet/FDDI adapters on EISA bus or PCI bus with the
following CPU types:
ALPHAserver 800 5/xxx EV56
ALPHAserver 1000A 5/xxx EV56
Realtime PCI board EV56
o The printf family of functions incorrectly uses the same
scratch buffers when running in AST and non-AST modes. This
may cause incorrect results if the non-AST execution is
interrupted by an AST routine which calls a function requiring
the same buffers.
o A DCL command, EXCHANGE/NETWORK with /transfer_mode=convert and
file size greater than 255 blocks results in an ACCVIO
o XQPERR Bugcheck:
"Found a stale referenced or non directory FCB in FCB queue"
o The machine will crash with an XQPERR bugcheck while trying to
allocate space on a disk. The problem occurs on stand alone
machines attempting explicit LBN or contiguous placement.
o VERIFY hangs in an infinite loop during its scan of the
directory structure.
o Verify does not flag files which are directly back linked to
themselves as having invalid backlinks.
o Analyze/disk goes into an infinite loop.
o The ATM driver loses too many receive packets causing a system
crash.
Problems Addressed in ALPACRT04_071:
o The mmap function no longer fails if a file is opened with read
access and the calling program has specified both the
PROT_WRITE and MAP_PRIVATE flags. The correct behavior of
using the MAP_PRIVATE flag is to disregard the access mode of
the opened file.
o The opendir function no longer fails if a file having the same
name, but no extension, exists in the same directory as the
directory being opened. Prior to this change, having an
extensionless file in the same directory as the directory being
opened would fail, even if the full name and extension was used
in the opendir call.
o Changes made in OpenVMS V7.0 have affected the processing of
files having "FORTRAN Carriage Control" record attributes. The
following changes were introduced in OpenVMS V7.0 and are
repaired in this ECO kit:
o The DEC C RTL now prepends (instead of appends) linefeed
characters to records with the single-space carriage
control. This was the behavior prior to OpenVMS V7.0.
Prefixing this type of record with linefeeds is what RMS
requires and allows overprinting to work better than when
the linefeed was appended.
The only known problem with the new version is that the
carriage-control character is not generated after the last file
record. This problem will be fixed in the next release of
CRTL.
o A regression was introduced in the VAXACRT03_071 ECO kit such
that calls to the stat function which use the "DNA=STRING"
parameter access violate. Calls to fopen may also access
violate when the fopen function calls stat to determine if the
file exists.
o A request was made that the sysconf function be enhanced to
return information about the processors in a system.
Specifically the _SC_NPROC_CONF (200) and _SC_NPROC_ONLN (201)
options are now supported. Modifications to the UNISTD.H
header file to define these constants will be made available in
a future release of the DEC C compiler. It is recommended that
application developers add the code
#ifndef _SC_NPROC_CONF
# define _SC_NPROC_CONF 200
# define _SC_NPROC_ONLN 201
#endif
until these definitions are available.
o A problem related to the size of an intermediate pipe buffer
may lead to a write operation receiving a %SYSTEM-F-MBTOOSML.
The size of the buffer has been corrected.
o The access function would incorrectly examine the SYSTEM
protection codes and use those codes in determining if access
would be allowed.
The access function required that all requested access modes be
available in a single group of OpenVMS protection codes. The
access modes may now be spread across protection code groups
such as world, group, owner, and system.
o A problem related to the size of an intermediate pipe buffer
may lead to a write operation receiving a %SYSTEM-F-MBTOOSML.
The size of the buffer has been corrected.
o The access function would incorrectly examine the SYSTEM
protection codes and use those codes in determining if access
would be allowed.
The access function required that all requested access modes be
available in a single group of OpenVMS protection codes. The
access modes may now be spread across protection code groups
such as world, group, owner, and system.
o A user reported that the sprintf function, when called using a
simple "%s" format specifier performed worse in a threaded
application than the equivalent call to strcpy. We have
changed the processing of "%s" to perform in line with the
strcpy function for this format specifier.
o Beginning in the ALPACRT03_071 and ALPY2K01_071 ECO kits,
calling the getenv function for the "TERM" variable began
returning the string "name-80" instead of correctly returning
the terminal type such as "VT320-80".
o Beginning in the ALPACRT03_071 remedial kit, the ioctl function
defaults to using the software available as the
UCX$IOCTL_ROUTINES example code. An error in the original code
is that if an error were to occur, a system service status code
(such as SS$_ACCVIO) would be returned as an errno value. This
has been corrected to return either the corresponding errno
value or EVMSERR with vaxc$errno being assigned this system
service status code.
Problems Addressed in ALPACRT03_071:
o Users have requested that kill support the POSIX semantics of
"if the process ID is negative but not -1, the signal will be
sent to all processes whose group ID is equal to the absolute
value of the process ID, and for which the process has
permission to send a signal." This has been added with the
restrictions that the process is executing on the same node and
does not have a SYSTEM UIC. The errno value is set to ESRCH if
no processes are found which match the condition specified.
o An enhancement was made in the DEC C V5.6 compiler to optimize
certain format strings passed to the printf family of library
functions. This ECO kit adds the runtime support to the
shareable image. Prior to this, the symbols were resolved via
objects added to STARLET by the compiler. Full details of this
support can be found in the DEC C V5.6 release notes.
o The performance of DEC C sprintf was much slower than VAX C.
An analysis of the printf engine resulted in changes which
brings DEC C within 10% of VAX C.
o The functions fopen and freopen were mapping invalid access
modes to read mode. Invalid modes now cause errno to be set to
EINVAL and the open call to fail.
o The times function was changed in OpenVMS V7.0 to return the
number of clock ticks since boot time. Performing year 2000
testing by setting the system time forward causes this return
value to overflow. The times function has been changed to
return the number of clock ticks since login time, which is
less likely to overflow.
o The lseek function may position incorrectly with repeated calls
to seek, in a file containing fixed length records of odd
length. This problem does not occur with even length or
variable length records.
o Mailbox devices are record oriented devices, except when
created by the pipe function where they are opened as stream
devices. Applications which use mailboxes can now force the
library to treat all mailboxes as stream devices by creating an
environment variable named DECC$MAILBOX_CTX_STM.
o The runtime library has been corrected to treat UNIX directory
specifications identically in each of the routines which accept
a directory specification as a parameter. These runtime
library functions are access, opendir, mkdir, and rmdir. Prior
to this change, one could call opendir with "/dev/dir", but was
forced to append ".dir" to this when calling the access
function.
o This ECO kit includes major performance improvements when using
time related functions along with Universal Coordinated Time.
o A cache of values has been added to the getenv function to
avoid the library making repeated calls to translate a logical
name or to obtain a symbol value for environment variables
which are not set. If your application makes direct calls to
set logical names, this caching can be disabled by defining
DECC$DISABLE_GETENV_CACHE prior to calling any runtime library
functions.
o The ANSI standard states that streams opened in update mode may
read and write to the stream. It further states that reads
must be followed by file positioning prior to writing to the
stream. The problem corrected was that positioning functions
would fail when the file was a terminal. Applications may now
position such streams back to the beginning using either rewind
or seek.
o It was reported that opendir overflowed the stack when running
in a threaded application. While correcting this problem, the
opendir successful return value was changed from one to zero to
align with the X/Open Specifications.
o A problem introduced in OpenVMS V7.1 causes the first record of
a file to be overwritten when the file is opened in append
mode. The correct behavior is that all write operations are
done at the end of the file.
o Porting code to OpenVMS is hampered by the difference between
command procedures and executable images and the mechanisms
necessary to invoke them. When passed the string "TEST", the
exec functions now search for "TEST.","TEST.EXE", and
"TEST.COM". If found, it is executed as either an image or a
command procedure, depending on information in the file header.
o Several new universals have been added to the DECC$SHR image
shipped with this ECO kit. The presence of these universals
may affect application developers who compile using this image.
If a developer begins to get errors of the form
%LINK-W-MULDEF, symbol DECC$XXX multiply defined
in module DECC$SHR file SYS$COMMON:[SYSLIB]DECC$SHR.EXE;1
while linking the application, the compile command should be
modified to include "/PREFIX=EXCEPT=XXX". This instructs the
compiler to exclude this function when doing name prefixing,
which is equivalent to the behavior prior to this ECO kit.
o The printf function was enhanced to print "(null)" when passed
a null pointer to be used with the "%S" format specifier.
Prior to this, the DEC C RTL would issue an access violation
error.
o Several functions were found to not accept the angle bracket
form of directory specifications. These functions include
decc$translate_vms and stat, which now accept either square
brackets or angle brackets in the directory portion of file
specifications.
o The exec functions use a mailbox to coordinate open file
information between the parent and child processes. A user
reported their system hung after 1000 successful invocations of
the same child process. This hang was caused by the parent
process failing to release an exclusive mode lock being used to
coordinate access to the mailbox. The hang would occur when a
mailbox was assigned to the parent process for a second time.
o A user reports calling ioctl sets errno to ENOSYS (Function not
implemented). Beginning with OpenVMS V7.0, the library looks
for support in the underlying TCPIP stack and sets this error
if the support is not found. The function has been enhanced to
execute UCX$EXAMPLES:UCX$IOCTL_ROUTINES under these conditions.
o The decc$to_vms function has been enhanced to recognize names
found on other systems, converting "/dev/null" to "NLA0:",
"/tmp" to "SYS$SCRATCH:", and "/bin" to "SYS$SYSTEM:".
o The getpwnam function now uppercases the username parameter if
it is not found in its original form. Prior to this change,
the function would fail.
Problems Addressed in ALPACRT02_071:
o Those functions, such as printf, which have thread specific
data incorrectly use the threads interface to release that
memory when a thread is being destroyed. The result is that
all thread specific memory is lost.
o The read, fread, write, and fwrite DEC C RTL functions now
return unsuccessful status with errno set to EINVAL if one of
the arguments of size_t type or total number of bytes to be
transmitted is not in the range 0 to INT_MAX. The total number
of bytes is the product of the size_of_item and number_items
arguments for fread and fwrite functions. The value of INT_MAX
is defined in the <limits.h> header file.
o The interval timer function, setitimer, fails to reset itself
when used in a multithreaded application. The result is a
single firing of the timer as opposed to repeated firing of the
timer at fixed intervals.
o A change made in OpenVMS V7.1 and remedial kits to other
versions cause calls to the mktemp function using templates of
the form "dumpXXXXXX.txt" to no longer substitute the pattern
with the process id. While a change is necessary to prohibit
substituting the directory portion in a template such as
"[XX]dumpXXXXXX", the change that was made was overly harsh,
forcing all substitution to the end of the pattern.
o The strstr function accesses memory beyond the ends of the
strings passed. In cases where the next page is not
accessible, the result is an access violation. The problem was
reported against the ADA compiler, which uses the strstr
function in this way.
o Although files in general are correctly inherited after a
fork/exec function call, files which are opened in any sort of
sharing mode are not.
o The ECO kits ALPACRT01_071 and VAXACRT01_071 made changes to
the return value of the puts and fputs routines. While these
changes were in line with the documentation, which states that
they return non-negative numbers on success, specific
applications were coded to expect zero as success. Since this
zero return value was documented with the VAXC product, we have
restored the original behavior.
o When accessing files in stream mode, closing the file may
result in an extra byte being written to the file. While this
byte is not seen using the type command, it may be seen when
using the dump utility.
o A customer reports that the functions atof, strtod, and wcstod
incorrectly return HUGE_VAL values when compiling with the
IEEE_FLOAT and an ieee_mode qualifier of DENORM_RESULTS. The
results returned have been modified to take the compilation
mode of the calling program into account. The return value can
now be compared against HUGE_VAL.
o The following code segment demonstrates a problem in the printf
family of functions running on OpenVMS for Alpha.
double n;
for (n = 9.0e16; n < 11.0e17; n += 1.0e+17)
printf ("%20.0f\n", n);
The result is that the final zero is missing in the display of
all but the first and last line.
Problems Addressed in ALPACRT01_071:
o The stat function now uses a thread specific buffer to store
data. Prior to this correction, calling stat from two separate
threads would interfere with one another.
o An ISV reports that extra <lf> characters are seen on occasion
when using a subprocess which sends data back to the parent
process using a mailbox.
o A case was found where the fseek function failed, correctly
returned a -1 value, but failed to set errno properly.
o One of the arguments to the decc$to_vms function is
"allow_wild" which is documented to accept the values zero and
one. If wildcards are used in the file specification, they are
either rejected or expanded into the resultant file
specifications. Passing a value of -1 for the allow_wild
parameter now returns the file specifications with the
wildcards intact, but after having prepared to the point of
doing a sys$search.
o Unlike Digital UNIX, the fsync(socket_id) call results in an
access violation instead of returning an EINVAL status.
o A user reports and demonstrates that opening and closing
sockets does not properly release mutexes eventually causing
this resource to be exhausted. This problem was introduced in
OpenVMS V7.0.
o Applications which call opendir and readdir recursively to
traverse subdirectories may end up in an infinite loop when
reading the directory which includes returning
[000000]000000.DIR. The readdir function no longer returns an
entry which is equivalent to the directory being read.
o OpenVMS V7.1 changed the behavior of the readdir function in
that it no longer retains the ".dir" extension when returning a
directory using the UNIX file syntax. Applications which rely
on this extension can now define a DECC$READDIR_KEEPDOTDIR
logical to restore this behavior.
o Beginning with OpenVMS V7.1, the DEC C Runtime Library does not
read the first record while opening the file if such preloading
will cause RMS to lock the record. Several users have reported
that images which used to work fine now fail. We have tuned
this record preloading code to take these user programs into
account.
Problems Addressed in ALPF11X02_071:
o An XQPERR bugcheck occurs when trying to create a file.
o Bad FID bugcheck when trying to mark a file header free in the
index file bitmap.
o There are multiply allocated blocks and file headers on the
disk.
o The contents of a header or bitmap block could be corrupted
within the block buffer cache.
o Failure to take an allocation lock could be ignored.
o If a DEACCESS request failed with a SS$_DEADLOCK error, a
process could be left in RWAST state indefinitely.
o If a large file is created on a fragmented disk with quotas
enabled and the user needs to use EXQUOTA privilege to allocate
the necessary disk space, an internal XQP table can become
corrupted leading to SECAUDERR, Fatal error attempting to
perform a security audit bug-checks.
o Superseding a file with a version limit set can remove the
oldest version of a file even if that version is undeletable by
the user according to its protection mask or ACL. The file is
removed but not deleted, and can be recovered by
ANALYZE/DISK/REPAIR.
o Attempting to queue a maximal length (39.39;5) filename to the
XQP for spooling to a symbiont would cause either an infinite
CPU loop or a FILCNTNONZ, Open file count nonzero after process
rundown bugcheck.
Problems Addressed in ALPF11X01_071:
o The problem occurs when a file is deleted while still being
accessed by someone. This produces an XQPERR bugcheck when an
attempt is made to access the deleted file.
o The problem may result in an XQPERR bugcheck which claims that:
"all the index buffers are active" during the processing of a
directory file.
The problem occurs when no free directory index BFRD's are
found on the first pass through MAKE_DIRINDX. The thread then
stalls to allow some of the BFRD's to be freed, but doesn't
release the cache lock which would allow others to do this.
This means that if no free BFRD was found on the first try then
none will be found on subsequent tries either, and the bugcheck
will occur.
o The fault occurs as a UNXSIGNAL Bugcheck after running AUTOGEN,
and rebooting with very large SYSGEN ACP cache parameters.
The problem occurs when the calculation of DIRINDX in
RELEASE_LOCKBASIS overflows. This is due to a multiplication
taking place before a division, which overflows when more than
87381 blocks have been allocated between the ACP_MAPCACHE,
ACP_DIRCACHE and ACP_HDRCACHE caches.
Problems Addressed in ALPLAN01_H1071:
o The Personal Workstation 433A with the 100 Mbit Medium
Attachment Unit (MAU) and a 21143 chip will not operate in 100
Mbit mode.
Problems Addressed in ALPSYS01_071:
o Problem was isolated to a specific $UNWIND call not
transferring control to the correct PC on OpenVMS Alpha V7.1.
Problems Addressed in ALPSYS02_071:
o Reducing the value of the PHYSICAL_MEMORY sysgen parameter on
an Alphaserver 2100 with console version V4.7 or later may fail
to boot. The failure is a boot crash just after the PFN
database is initialized.
RELATED ARTICLES:
Detailed articles describing the problems listed above may exist in
the OPENVMS database. To view these articles, open the appropriate
product database and perform a query using either of the following
search strings: 'ALPBASE02_071', 'ALPBASE', 'ALPACRT', 'ALPF11X',
'ALPLAN', or 'ALPSYS'.
ECO KIT ORDERING INSTRUCTIONS:
If after an evaluation you wish to obtain this kit, request it
electronically using the appropriate Advanced Electronic Services
(AES) Service Tool. If you are not familiar with how to request
kits electronically, open the DIA, WIS or DSNLINK database and
review the article entitled:
[AES] How To Electronically Request ECO Kits Using Service Tools
INSTALLATION NOTES:
The images in this kit will not take effect until the system is
rebooted. If there are other nodes in the VMScluster, they must
also be rebooted in order to make use of the new image(s).
If it is not possible or convenient to reboot the entire cluster at
this time, a rolling re-boot may be performed.
==========================================================================
| Table of Kit Image Information |
+----------------------------+----------+-----------------+--------------+
| | Overall | Image File | Image Link |
| Image Name | Checksum | Identification | Date/Time |
+----------------------------+----------+-----------------+--------------+
| ALPHA_CLUE$SDA.EXE | 5DCD8452 | X-5 | 19-JAN-1998 |
| | | 22:57:18.18 |
+----------------------------+----------+-----------------+--------------+
| APB.EXE | 87E5F94B | X-8 | 7-JAN-1998 |
| | | 12:53:09.39 |
+----------------------------+----------+-----------------+--------------+
| CLUE$SDA.EXE | 4F55E9FC | X-5 | 19-JAN-1998 |
| | | 22:59:18.61 |
+----------------------------+----------+-----------------+--------------+
| CONFIGURE.EXE | A85F449A | X-4 | 7-JAN-1998 |
| | | 11:43:09.04 |
+----------------------------+----------+-----------------+--------------+
| DEBUG_APB.EXE | BAB40063 | X-8 | 7-JAN-1998 |
| | | 12:53:15.15 |
+----------------------------+----------+-----------------+--------------+
| DECC$SHR.EXE | 7A896DEE | X07.1-4 | 29-JAN-1998 |
| | | 12:27:14.91 |
+----------------------------+----------+-----------------+--------------+
| DUMP.EXE | 73175CB7 | X-12 | 18-FEB-1998 |
| | | 10:43:57.85 |
+----------------------------+----------+-----------------+--------------+
| EXCEPTION.EXE | 30CCEC7B | X-3 | 8-JAN-1998 |
| | | 15:13:03.15 |
+----------------------------+----------+-----------------+--------------+
| EXCEPTION_MON.EXE | 027AA098 | X-3 | 8-JAN-1998 |
| | | 15:13:14.97 |
+----------------------------+----------+-----------------+--------------+
| EXCHANGE.EXE | 3423FC84 | X-3 | 15-OCT-1997 |
| | | 23:45:45.49 |
+----------------------------+----------+-----------------+--------------+
| EXEC_INIT.EXE | CA0A91D6 | X-1 | 8-JAN-1998 |
| | | 15:13:26.58 |
+----------------------------+----------+-----------------+--------------+
| F11BXQP.EXE | E6F7C3DB | XQP V7.1R-0002 | 15-OCT-1997 |
| | | 23:48:51.29 |
+----------------------------+----------+-----------------+--------------+
| IMAGE_MANAGEMENT.EXE | 0EAEBFE1 | X-2 | 8-JAN-1998 |
| | | 15:13:38.65 |
+----------------------------+----------+-----------------+--------------+
| IO_ROUTINES.EXE | 5FEE1FE3 | X-3 | 8-JAN-1998 |
| | | 15:13:52.28 |
+----------------------------+----------+-----------------+--------------+
| IO_ROUTINES_MON.EXE | F62D753F | X-3 | 8-JAN-1998 |
| | | 15:14:17.56 |
+----------------------------+----------+-----------------+--------------+
| LANACP.EXE | 0A4E1018 | X-24 | 14-JAN-1998 |
| | | 11:18:05.24 |
+----------------------------+----------+-----------------+--------------+
| LANCP.EXE | D123BBF7 | X-33 | 14-JAN-1998 |
| | | 11:18:02.33 |
+----------------------------+----------+-----------------+--------------+
| LBRSHR.EXE | B1AB785E | A09-22 | 18-FEB-1998 |
| | | 10:43:50.43 |
+----------------------------+----------+-----------------+--------------+
| MTAAACP.EXE | FE8E3C75 | X-7 | 18-FEB-1998 |
| | | 10:44:01.84 |
+----------------------------+----------+-----------------+--------------+
| CSMACD.EXE | 083B874C | X-13 | 14-JAN-1998 |
| | | 11:17:57.72 |
+----------------------------+----------+-----------------+--------------+
| FDDI.EXE | 3381F286 | X-13 | 14-JAN-1998 |
| | | 11:17:59.32 |
+----------------------------+----------+-----------------+--------------+
| SDA$SHARE.EXE | 67335C12 | X-2 | 8-JAN-1998 |
| | | 16:39:57.22 |
+----------------------------+----------+-----------------+--------------+
| SDA.EXE | A4054495 | X-1 | 8-JAN-1998 |
| | | 16:40:05.02 |
+----------------------------+----------+-----------------+--------------+
| SHOW.EXE | 313CCA52 | X-1 | 10-DEC-1997 |
| | | 08:00:22.43 |
+----------------------------+----------+-----------------+--------------+
| STACONFIG.EXE | E9178A9D | X-26 | 29-JAN-1998 |
| | | 12:50:32.11 |
+----------------------------+----------+-----------------+--------------+
| SYS$ATMWORKS351.EXE | 4DE9DC8F | X-8 | 14-JAN-1998 |
| | | 11:18:23.42 |
+----------------------------+----------+-----------------+--------------+
| SYS$BASE_IMAGE.EXE | 7784A978 | ALPHA X6C7-G5N | 8-JAN-1998 |
| | | 15:12:24.26 |
+----------------------------+----------+-----------------+--------------+
| SYS$CLDRIVER.EXE | 755A888F | X-10 | 14-JAN-1998 |
| | | 11:16:58.62 |
+----------------------------+----------+-----------------+--------------+
| SYS$ECDRIVER.EXE | 5C28A829 | X-3 | 14-JAN-1998 |
| | | 11:16:55.51 |
+----------------------------+----------+-----------------+--------------+
| SYS$ELDRIVER.EXE | 2F78A02A | X-24 | 14-JAN-1998 |
| | | 11:17:01.82 |
+----------------------------+----------+-----------------+--------------+
| SYS$EODRIVER.EXE | 63CDF50D | X-3 | 14-JAN-1998 |
| | | 11:17:21.67 |
+----------------------------+----------+-----------------+--------------+
| SYS$ERBTDRIVER.EXE | AF0768BA | X-4 | 7-JAN-1998 |
| | | 12:35:39.43 |
+----------------------------+----------+-----------------+--------------+
| SYS$ERDRIVER.EXE | 69C1EDFE | X-3 | 14-JAN-1998 |
| | | 11:17:04.89 |
+----------------------------+----------+-----------------+--------------+
| SYS$ESDRIVER.EXE | A2B0FCAF | X-3 | 14-JAN-1998 |
| | | 11:17:13.50 |
+----------------------------+----------+-----------------+--------------+
| SYS$EWBTDRIVER.EXE | 47F20F84 | X-4 | 7-JAN-1998 |
| | | 12:35:43.86 |
+----------------------------+----------+-----------------+--------------+
| SYS$EWDRIVER.EXE | 0824339E | X-3 | 14-JAN-1998 |
| | | 11:17:07.93 |
+----------------------------+----------+-----------------+--------------+
| SYS$EWDRIVER_DE500BA.EXE | 713E2868 | X-3 | 14-JAN-1998 |
| | | 11:17:10.62 |
+----------------------------+----------+-----------------+--------------+
| SYS$EXDRIVER.EXE | B52684B3 | X-3 | 14-JAN-1998 |
| | | 11:17:16.22 |
+----------------------------+----------+-----------------+--------------+
| SYS$EZDRIVER.EXE | C217FB2B | X-3 | 14-JAN-1998 |
| | | 11:17:18.95 |
+----------------------------+----------+-----------------+--------------+
| SYS$FADRIVER.EXE | E1EDAC4C | X-3 | 14-JAN-1998 |
| | | 11:17:24.44 |
+----------------------------+----------+-----------------+--------------+
| SYS$FCBTDRIVER.EXE | F8B3D1F0 | X-1 | 7-JAN-1998 |
| | | 12:35:57.12 |
+----------------------------+----------+-----------------+--------------+
| SYS$FCDRIVER.EXE | 8B83029C | X-3 | 14-JAN-1998 |
| | | 11:17:27.27 |
+----------------------------+----------+-----------------+--------------+
| SYS$FRBTDRIVER.EXE | 37E2B971 | X-1 | 7-JAN-1998 |
| | | 12:35:59.39 |
+----------------------------+----------+-----------------+--------------+
| SYS$FRDRIVER.EXE | B618D3EA | X-3 | 14-JAN-1998 |
| | | 11:17:30.11 |
+----------------------------+----------+-----------------+--------------+
| SYS$FWBTDRIVER.EXE | 6D2B629F | X-1 | 7-JAN-1998 |
| | | 12:36:01.81 |
+----------------------------+----------+-----------------+--------------+
| SYS$FWDRIVER.EXE | 66C65CB0 | X-3 | 14-JAN-1998 |
| | | 11:17:33.00 |
+----------------------------+----------+-----------------+--------------+
| SYS$FXDRIVER.EXE | 14324C22 | X-3 | 14-JAN-1998 |
| | | 11:17:35.91 |
+----------------------------+----------+-----------------+--------------+
| SYS$HCDRIVER.EXE | 93528048 | X-27 | 14-JAN-1998 |
| | | 11:17:49.91 |
+----------------------------+----------+-----------------+--------------+
| SYS$HWDRIVER.EXE | 5A348692 | X-27 | 14-JAN-1998 |
| | | 11:17:54.41 |
+----------------------------+----------+-----------------+--------------+
| SYS$ICDRIVER.EXE | 3EE28071 | X-26 | 14-JAN-1998 |
| | | 11:17:43.07 |
+----------------------------+----------+-----------------+--------------+
| SYS$IRDRIVER.EXE | 4BC41221 | X-26 | 14-JAN-1998 |
| | | 11:17:38.84 |
+----------------------------+----------+-----------------+--------------+
| SYS$IWDRIVER.EXE | 6069C49E | X-26 | 14-JAN-1998 |
| | | 11:17:46.27 |
+----------------------------+----------+-----------------+--------------+
| SYS$LAN.EXE | 0B222F4F | X-3 | 14-JAN-1998 |
| | | 11:16:32.80 |
+----------------------------+----------+-----------------+--------------+
| SYS$LAN_ATM.EXE | F865A684 | X-26 | 14-JAN-1998 |
| | | 11:16:41.29 |
+----------------------------+----------+-----------------+--------------+
| SYS$LAN_CSMACD.EXE | 6EDF7945 | X-3 | 14-JAN-1998 |
| | | 11:16:35.81 |
+----------------------------+----------+-----------------+--------------+
| SYS$LAN_FDDI.EXE | 1F18D0E9 | X-3 | 14-JAN-1998 |
| | | 11:16:49.40 |
+----------------------------+----------+-----------------+--------------+
| SYS$LAN_TR.EXE | 19D9A8A3 | X-3 | 14-JAN-1998 |
| | | 11:16:52.09 |
+----------------------------+----------+-----------------+--------------+
| SYS$VM.EXE | D215DDE9 | X-3 | 8-JAN-1998 |
| | | 15:16:19.33 |
+----------------------------+----------+-----------------+--------------+
| SYSBOOT.EXE | 4C3F79D2 | X-82a1 | 15-DEC-1997 |
| | | 13:37:04.43 |
+----------------------------+----------+-----------------+--------------+
| SYSINIT.EXE | C45F3FE2 | X-66 | 18-AUG-1997 |
| | | 11:23:23.22 |
+----------------------------+----------+-----------------+--------------+
| TECOSHR_TV.EXE | AAEBC836 | TECOSHR V40.37 | 2-SEP-1997 |
| | | 23:42:05.32 |
+----------------------------+----------+-----------------+--------------+
| VERIFY.EXE | 5E46D48F | X-23 | 15-OCT-1997 |
| | | 23:47:58.03 |
+----------------------------+----------+-----------------+--------------+
| VMS$REMEDIAL_ID.EXE | 3EC6483F | V1.0 | 14-APR-1998 |
| | | 10:15:06.93 |
+----------------------------+----------+-----------------+--------------+
|
