- Due to incorrect order of initialization, Anaconda failed to detect the presence of the zSeries Linux fibre-channel adapter (zFCP) disks. To fix this bug, the cio_settle kernel interface has been implemented, which waits for the zFCP devices to become online. Now, Anaconda detects the zFCp devices as intended.
- For each possible CPU, the zfcpdump kernel consumed all memory for the per-CPU data structures. Because only 32 MB are available for zfcpdump, zfcpdump could run out of memory. This update adds a new kernel parameter "possible_cpus=1", and the zfcpdump system no longer runs out of memory.
- Previously, when the "fdasd -c" command was called with a configuration file that contained only one parameter for a partition, the fdasd daemon terminated unexpectedly with a segmentation fault during configuration file parsing. This update adds a new function to parse configuration file lines, and fdasd no longer crashes in the described situation.
- Previously, removing a device that was currently offline resulted in an error. The znetconf protocol has been fixed to handle removal of an offline ccwgroup devices correctly, and the "znetconf -r" command now removes a currently offline device as intended.
- The output of the lsqeth command depends on the presence of a file named "?". Due to a bug in the grep command regular expression (regex), the system qeth devices failed to be detected when a file named "?" was present in the current working directory. To fix this bug, the regex argument of grep has been put in single quotes, and the system qeth devices are now detected successfully.
- Due to incomplete dependencies specified in the s390utils packages, various command-line tools did not work when invoked or gathered incomplete data. This update adds the missing dependencies to the tools, which now work as expected.
- BZ#1017854, BZ#1031143, BZ#1032061, BZ#1088328
- With this update, various information gathered by the dbginfo.sh utility has been expanded, along with the relevant manual page.
- This update introduces a new interface that enables Linux applications such as Data Stage to access and process read-only data in physical sequential data sets owned by IBM System z without interfering with System z. By avoiding FTP or NFS transfer of data from System z, the turnaround time for batch processing is significantly reduced.