Use the PATCH subcommand to repair data residing in a RECFM=F or
RECFM=FBS data set or to manage the list of patches in effect for
a dump.
Patching may impact IPCS performance and is intended to be used
very sparingly. The reason that a patching capability has been included
is the following scenario:
- You attempt to run a high level report against a dump. The report
is important for your analysis.
- The report writer encounters a block that appears to be damaged.
Rather than using the contents of the damaged block and risking the
production of a misleading report, the report writer identifies the
block and the damage detected.
- You examine the damaged block, verify that its damage is not the
root problem that you sought, and are able to determine values that
repair damage to it.
- You use the PATCH subcommand to identify the repairs to IPCS.
IPCS does not alter the dump data set in any way. The alterations
are stored in your dump directory.
Patching storage that IPCS knows
can be seen from multiple perspectives, such as both common virtual
storage and real storage visible to each CPU in the dumped system,
affects all perspectives.
- Restrictions
- IPCS may access dump data before application of a patch, recording
conclusions regarding that data in the dump directory before application
of a patch. The PATCH subcommand does not attempt to locate and alter
any such data. Some of this data may be affected using other subcommands
such as
- DROPDUMP RECORDS(TRANSLATION)
- DROPMAP
- DROPSYM
- The current implementation of PATCH support directly uses data
in dump records for most information associated with DISPLAY(MACHINE)
output and the related data that may be extracted from a dump using
the EVALUATE subcommand. Processing of storage by EVALUATE does honor
PATCH requests.
- Storage may be added to what was dumped, such as from ASID(75),
through PATCH processing, but PATCH will not attempt to identify the
absolute or real storage locations where that storage would have resided
in the dumped system. If this is important to your analysis, you
must use PATCH to add it from all perspectives important to your analysis.
- Qualifier
The following qualifiers distinguish the functions
performed by the PATCH subcommand:
- Qualifier
- Function
- ADD
- Causes the PATCH subcommand to store a new patch. See Adding or replacing a patch for more information. Existing, overlapping
patches are considered to be an error and cause the new patch to be
rejected.
- DELETE
- Causes the PATCH subcommand to delete patches. See Deleting patches for more information.
- LIST
- Causes the PATCH subcommand to list patches. See Listing patches for more information.
- REPLACE
- Causes the PATCH subcommand to store a patch, replacing one or
more existing ones whose descriptions overlap the new one. See Adding or replacing a patch for more information. At least one existing,
overlapping patch is expected. If there is none, it is considered
to be an error, and the new patch is rejected.
- STORE
- Causes the PATCH subcommand to store a patch, replacing any existing
ones whose descriptions overlap the new one. See Adding or replacing a patch for more information.