| Description: | The kernel packages contain the Linux kernel, the core of any Linux operating system.
This update fixes the following security issues:
a flaw was found in the Unidirectional Lightweight Encapsulation (ULE) implementation. A remote attacker could send a specially-crafted ISO MPEG-2 Transport Stream (TS) frame to a target system, resulting in an infinite loop (denial of service). (CVE-2010-1086, Important)
on AMD64 systems, it was discovered that the kernel did not ensure the ELF interpreter was available before making a call to the SET_PERSONALITY macro. A local attacker could use this flaw to cause a denial of service by running a 32-bit application that attempts to execute a 64-bit application. (CVE-2010-0307, Moderate)
a flaw was found in the kernel connector implementation. A local, unprivileged user could trigger this flaw by sending an arbitrary number of notification requests using specially-crafted netlink messages, resulting in a denial of service. (CVE-2010-0410, Moderate)
a flaw was found in the Memory-mapped I/O (MMIO) instruction decoder in the Xen hypervisor implementation. An unprivileged guest user could use this flaw to trick the hypervisor into emulating a certain instruction, which could crash the guest (denial of service). (CVE-2010-0730, Moderate)
a divide-by-zero flaw was found in the azx_position_ok() function in the driver for Intel High Definition Audio, snd-hda-intel. A local, unprivileged user could trigger this flaw to cause a kernel crash (denial of service). (CVE-2010-1085, Moderate)
This update also fixes the following bugs:
in some cases, booting a system with the "iommu=on" kernel parameter resulted in a Xen hypervisor panic. (BZ#580199)
the fnic driver flushed the Rx queue instead of the Tx queue after fabric login. This caused crashes in some cases. (BZ#580829)
"kernel unaligned access" warnings were logged to the dmesg log on some systems. (BZ#580832)
the "Northbridge Error, node 1, core: -1 K8 ECC error" error occurred on some systems using the amd64_edac driver. (BZ#580836)
in rare circumstances, when using kdump and booting a kernel with "crashkernel=128M@16M", the kdump kernel did not boot after a crash. (BZ#580838)
TLB page table entry flushing was done incorrectly on IBM System z, possibly causing crashes, subtle data inconsistency, or other issues. (BZ#580839)
iSCSI failover times were slower than in Red Hat Enterprise Linux 5.3. (BZ#580840)
fixed floating point state corruption after signal. (BZ#580841)
in certain circumstances, under heavy load, certain network interface cards using the bnx2 driver and configured to use MSI-X, could stop processing interrupts and then network connectivity would cease. (BZ#587799)
cnic parts resets could cause a deadlock when the bnx2 device was enslaved in a bonding device and that device had an associated VLAN. (BZ#581148)
some BIOS implementations initialized interrupt remapping hardware in a way the Xen hypervisor implementation did not expect. This could have caused a system hang during boot. (BZ#581150)
AMD Magny-Cours systems panicked when booting a 32-bit kernel. (BZ#580846)
Users should upgrade to these updated packages, which contain backported patches to correct these issues. The system must be rebooted for this update to take effect.
|