Security update for Linux kernel
Knowledgebase
(Last modified: 15DEC2006)
solutions Security update for Linux kernel SuSE Linux Maintenance Web (36b4bba8bf8a44877f22acb24254f105)
kernel-source
kernel-syms
Product(s): SUSE CORE 9 for IBM S/390 31bit
Patch: patch-11325
Release: 20061215
Obsoletes: d501672f20f1f468d24344550dbc92e0
- When AppArmor allows execution of another application via "x" profiles it is possible by attackers within the current profile to inject environment variables into the called binary, like for instance LD_PRELOAD. Should the execution by "ux" (unconfined) it is possible to break out of the confinement irregardless of checks done by the called application. New execution flags have been added, including "Ux" and "Px" which make the called application clear environment variables same as for setuid applications and avoid easy break outs. To activate this change a userland update of the subdomain-* packages is required. [#172061]
- If a profile only allowed read ("r") access to a resource the confined application could use mmap with PROT_EXEC flag to load this resource and execute code in it. A special "m" modifier for "r" was added to allow such mmap() calls, it is disabled by default. This change also requires new userland subdomain-* packages, also supplied via maintenance update. [#175388]
- A race condition in ptrace_check_attach() could be used by local attackers to hang the machine. [#180423]
- CVE-2006-4145: A bug within the UDF filesystem that caused machine hangs when truncating files on the filesystem was fixed. [#186226]
- A potential crash when receiving IPX packets was fixed. This problem is thought not to be exploitable. [#197809]
- CVE-2006-3741: A struct file leak was fixed in the perfmon(2) system call on the Itanium architecture. [#202269]
- CVE-2006-4538: A malformed ELF image can be used on the Itanium architecture to trigger a kernel crash (denial of service) when a local attacker can supply it to be started. [#203822]
- CVE-2006-4997: A problem in the ATM protocol handling clip_mkip function could be used by remote attackers to potentially crash the machine. [#205383]
- CVE-2006-5757/ CVE-2006-6060: A problem in the grow_buffers function could be used to crash or hang the machine using a corrupted filesystem. This affects filesystem types ISO9660 and NTFS. [#205384]
- CVE-2006-4813: A information leak in __block_prepare_write was fixed, which could disclose private information of previously unlinked files. [#208460]
- CVE-2006-5174: On the S/390 architecture copy_from_user() could be used by local attackers to read kernel memory. [#209880]
- CVE-2006-5648: On the PowerPC architecture a syscall has been wired without the proper futex implementation that can be exploited by a local attacker to hang the machine. [#217295]
- CVE-2006-5649: On the PowerPC architecture the proper futex implementation was missing a fix for alignment check which could be used by a local attacker to crash the machine. [#217295]
- CVE-2006-5823: A problem in cramfs could be used to crash the machine during mounting a crafted cramfs image. This requires an attacker to supply such a crafted image and have a user mount it. [#218237]
- CVE-2006-6053: A problem in the ext3 filesystem could be used by attackers able to supply a crafted ext3 image to cause a denial of service or further data corruption if a user mounts this image. [#220288]
- CVE-2006-6054: A problem in the ext2 filesystem could be used by attackers supplying crafted ext2 images to users could crash the machine during mount. [#220330]
- CVE-2006-6056: Missing return code checking in the HFS could be used to crash machine when a user complicit attacker is able to supply a specially crafted HFS image. [#221230]
- patches.fixes/quota-reiser-deadlock.diff, patches.fixes/reiserfs-defacl-xattr-lock-fix.diff: re-enabled these reiserfs fixes previously suspected to cause #170162 below [#115460]
- patches.fixes/dm-bio_list_merge-fix.diff: device-mapper snapshot: bio_list fix [#117435]
- patches.suse/ocfs2-configfs-20060707, patches.suse/ocfs2-configfs-20060809: Update to OCFS2 1.2.3 [#138287]
- patches.drivers/icadd-license_gpl.patch: Add license to ICA drivers [#140103]
- patches.fixes/vma-defrag: avoid fragmentation [#141653]
- patches.fixes/within-logging-rate-limit: Scaling fix for simultaneous unaligned accesses [#145599]
- patches.fixes/statd-refcount-fix: Fix refcounting problems in host management in lockd. [#148009]
- patches.fixes/statd-regular-gc: Make lockd garbage collect hosts more often [#148009]
- patches.fixes/wait_on_page_writeback_range_wq.patch: fix oops in __aio_run_iocbs [#165140]
- patches.fixes/i8042-reentry: Prevents i8042_interrupt() from being reentered. [#167187]
- patches.fixes/reiserfs-nfs-lock-fix.diff: Add forgotten lock in reiserfs_get_dentry [#170162]
- CKRM fails to enforce CPU resource guarantees [#173437]:
- patches.suse/calculate-task-pressure-properly
- patches.suse/do_not_allow_a_limit_of_zero
- patches.suse/doc_update
- patches.suse/dont_remove_if_guar_zero
- patches.suse/fix_savings
- patches.suse/ignore-rq-with-1proc-in-load-balancing
- patches.suse/make_softlimit_proper
- patches.suse/mult_overflow_fix.patch
- patches.suse/no_undue_importance_for_low_grt
- patches.suse/pressure_zero_when_nr_running_is_zero
- patches.suse/rename_total_grt_to_total_esl
- patches.suse/savings_accum_recalc
- patches.suse/use_busiest_load_instead_of_avg_load
- patches.suse/use_grt_in_sort_instead_of_inc_limit
- patches.suse/use_limit_instead_of_grt_when_consuming_surplus
- patches.suse/fix_demand_calc.patch1
- patches.suse/be_less_aggressive_in_load_balancing
- patches.fixes/dev-mem-false-positive: avoid false positive BUG_ON on /dev/mem page aliases. [#176958]
- patches.fixes/nfsd-type-3-fh: Fix nfsd support for filesystems on devices with large minors [#182552]
- patches.suse/audit-realpath-enoent: open(/foo/bar) would be logged whenever /foo didn't exist [#200118]
- patches.fixes/bonding-arpmon: bonding driver: verify incoming packets in arp_mon mode [#202326]
- patches.suse/lkcd-struct-runqueue-export: cannot find per cpu symbol 'runqueues' when analyzing dump [#202697]
- patches.fixes/xfs-unlink-recovery-fix: [XFS] unlink recovery fix. [#202877]
- patches.fixes/libata-eh-data-corruption: Fix data corruption when invoking error handling [#202998]
- disabled patches.arch/s390-35-06-april2004.diff because it breaks kernel-um [#203196]
- patches.fixes/i386-cpuinfo-amd-dualcore-fix.diff: /proc/cpuinfo: display number of cores [#203749]
- Mark qla2400 as supported [#204078]
- patches.fixes/sys_flock-race-fix.diff: Fix sys_flock() race [#205323]
- patches.arch/ia64-mca-ignore-valid.fr: [ia64] avoid KDB errors from kdba_mca_init. [#209660]
- patches.drivers/qla1280-bus-reset-handling: performance slowdown after bus reset on qla12160 HBA [#213717]
- patches.drivers/qla1280-scb-timeout: qla1280 times out on long operations such as tape rewind [#214695]
- patches.fixes/pfm_record_tm-locking-fix: Add list entry with write lock protection in perfmon [#214936]
- fix error in: patches.drivers/usb-add-raritan-kvm-usb-dongle-to-the-hid_quirk_noget-blacklist.patch: USB: add Raritan KVM USB Dongle to the HID_QUIRK_NOGET blacklist [#216225]
- patches.fixes/cifs-inval-dentry-sles9.patch: [CIFS] When file is deleted locally but later recreated on the server they don't get noticed. [#220540]
- Added latest patchset from IBM: Patchcluster 40
- Problem-ID: 26648 - cio: incorrect detection of unsol. interrupts during sense pgid
- Problem-ID: 26756 - qeth: After cable pull (out/in) device does not work
- Problem-ID: 27254 - cio: subchannel scan loop for reappearing channel paths ends early
- Problem-ID: 27255 - cio: incorrect no-path indication after machine check
- Problem-ID: 27256 - cio: race condition leaves device in inaccessible state
- Problem-ID: 27257 - cio: path group not updated by CHPID vary operation
- Problem-ID: 27258 - cio: path verification ignores reappearing channel paths
- Problem-ID: 27068 - qeth: VLAN header reordering does not work on packets received through qeth interface in layer 2 mode.
- Problem-ID: 25692 - z90crypt: Logfile flooding when no crypto hardware is available.
- Problem-ID: 28177 - cio: incorrect device operational notification
- Problem-ID: 27654 - cio: 0 is a valid chpid.
- Problem-ID: 26754 - kernel: incorrect copy_in_user.
- Problem-ID: 27706 - kernel: user readable uninitialised kernel memory.
- Problem-ID: 21059 - dasd: Failing partition detection during device bring-up.
- Problem-ID: 28120 - qeth: do not manipulate outgoing cloned skbs
- Problem-ID: 28522 - cio: css_probe_device() must be called enabled.
- Problem-ID: 28547 - cio: inaccessible device after CHPID deactivation
- Problem-ID: 28548 - cio: I/O request failure after CHPID deactivation
- Added latest patchset from IBM: Patchcluster 41
- Problem-ID: 28507 - dasd: Fixed handling of invalid values when writing into sysfs-attribute.
- Problem-ID: 29173 - cio: reipl fails after channel path reset
- Problem-ID: 29341 - kernel: Misaligned wait-PSW.
- Problem-ID: 28121 - qeth: device functions are not callable in atomic context.
rpm -Fvh kernel-s390x.rpm kernel-source.rpm kernel-syms.rpm
zipl
When rebooting the Linux on zSeries z/VM guests, please ensure that you have installed the PTFs for APAR VM63742:
- z/VM 4.4: UM31426
- z/VM 5.1: UM31428
Finally, reboot the system with
shutdown -r now
to load the new kernel (replace "now" with the appropriate amount of time to allow local users to cleanly log out, for example "+5" for five minutes.)
Download Source Packages
Download the source code of the patches for maintained products.
Disclaimer
The Origin of this information may be internal or external to Novell. Novell makes all reasonable efforts to verify this information. However, the information provided in this document is for your information only. Novell makes no explicit or implied claims to the validity of this information.
Any trademarks referenced in this document are the property of their respective owners. Consult your product manuals for complete trademark information.