Security issues: * Using PCI passthrough without interrupt remapping support allowed KVM guests to generate MSI interrupts and thus potentially inject traps. A privileged guest user could use this flaw to crash the host or possibly escalate their privileges on the host. The fix for this issue can prevent PCI passthrough working and guests starting. Refer to Red Hat Bugzilla bug 715555 for details. * Flaw in the client-side NLM implementation could allow a local, unprivileged user to cause a denial of service. * Integer underflow in the Bluetooth implementation could allow a remote attacker to cause a denial of service or escalate their privileges by sending a specially-crafted request to a target system via Bluetooth. * Buffer overflows in the netlink-based wireless configuration interface implementation could allow a local user, who has the CAP_NET_ADMIN capability, to cause a denial of service or escalate their privileges on systems that have an active wireless interface. * Flaw in the way the maximum file offset was handled for ext4 file systems could allow a local, unprivileged user to cause a denial of service. * Flaw allowed napi_reuse_skb to be called on VLAN packets. An attacker on the local network could use this flaw to send crafted packets to a target, possibly causing a denial of service. * Integer signedness error in next_pidmap could allow a local, unprivileged user to cause a denial of service. * Race condition in the memory merging support could allow a local, unprivileged user to cause a denial of service. KSM is off by default, but on systems running VDSM, or on KVM hosts, it is likely turned on by the ksm/ksmtuned services. * Flaw in inet_diag_bc_audit could allow a local, unprivileged user to cause a denial of service. * Flaw in the way space was allocated in the Global File System 2 implementation. If the file system was almost full, and a local, unprivileged user made an fallocate request, it could result in a denial of service. Setting quotas to prevent users from using all available disk space would prevent exploitation of this flaw. * Local, unprivileged users could send signals via the sigqueueinfo system call, with si_code set to SI_TKILL and with spoofed process and user IDs, to other processes. This flaw does not allow existing permission checks to be bypassed; signals can only be sent if your privileges allow you to already do so. * Heap overflow in the EFI GUID Partition Table implementation could allow a local attacker to cause a denial of service by mounting a disk containing crafted partition tables. * Structure padding in two structures in the Bluetooth implementation was not initialized properly before being copied to user-space, possibly allowing local, unprivileged users to leak kernel stack memory to user-space. * /proc/[PID]/io is world-readable by default. Previously, these files could be read without any further restrictions. A local, unprivileged user could read these files, belonging to other, possibly privileged processes to gather confidential information, such as the length of a password used in a process. Red Hat would like to thank Vasily Averin for reporting CVE-2011-2491; Dan Rosenberg for reporting CVE-2011-2497 and CVE-2011-2213; Ryan Sweat for reporting CVE-2011-1576; Robert Swiecki for reporting CVE-2011-1593; Andrea Righi for reporting CVE-2011-2183; Julien Tinnes of the Google Security Team for reporting CVE-2011-1182; Timo Warns for reporting CVE-2011-1776; Marek Kroemeke and Filip Palian for reporting CVE-2011-2492; and Vasiliy Kulikov of Openwall for reporting CVE-2011-2495.