CVE-2013-1798

Andrew Honig reported a flaw in the way KVM (Kernel-based Virtual Machine) emulated the IOAPIC. A privileged guest user could exploit this flaw to read host memory or cause a denial of service (crash the host). (CVE-2013-1798)

It was discovered that the KVM implementation in the Linux kernel, when paravirtual TLB flushes are enabled in guests, the hypervisor in some situations could miss deferred TLB flushes or otherwise mishandle them. An attacker in a guest VM could use this to expose sensitive information (read memory from another guest VM). (CVE-2019-3016)

Al Viro discovered that the vfs layer in the Linux kernel contained a use- after-free vulnerability. A local attacker could use this to cause a denial of service (system crash) or possibly expose sensitive information (kernel memory). (CVE-2020-8428)

The SUSE Linux Enterprise 11 Service Pack 2 kernel was updated to fix a regression introduced by the previous update:

* scsi_dh_alua: Incorrect reference counting in the SCSI ALUA
  initialization code lead to system crashes on boot (bnc#858831).

As the update introducing the regression was marked security, this is also marked security even though this bug is not security relevant.

The SUSE Linux Enterprise 11 SP1 Teradata kernel was updated to fix bugs and security issues.

Following security issues were fixed:

CVE-2011-1083: Limit the path length users can build using epoll() to avoid local attackers consuming lots of kernel CPU time.

CVE-2011-4086: Fixed a oops in jbd/jbd2 that could be caused by specific filesystem access patterns.

CVE-2011-4622: KVM: Prevent starting PIT timers in the absence of irqchip support.

CVE-2012-0045: KVM: Extend 'struct x86_emulate_ops' with 'get_cpuid' and fix missing checks in syscall emulation.

CVE-2012-0879: Fix io_context leak after clone with CLONE_IO.

CVE-2012-1090: Fixed a dentry refcount leak in the CIFS file system that could lead to a crash on unmount.

CVE-2012-1097: The regset common infrastructure assumed that regsets would always have .get and .set methods, but necessarily .active methods. Unfortunately people have since written regsets without .set method, so NULL pointer dereference attacks were possible.

Following non-security issues were fixed:

* SCSI inquiry doesn't return data on SLES 11-SP1 Xen VMs (bnc#745929).
* FC transport driver killing off the timers/work queues (bnc#734300).
* The driver ixgbevf doesn't work on newer SLES 11-SP1 kernels
  (bnc#752972).
* Pack sparsemem memmap sections closer together and in higher zones
  (bnc#743870).

Following feature was implemented:

* The megaraid_sas driver update to version 5.40-LSI (bnc#736813).

Security Issues:

* CVE-2011-1083
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-1083>
* CVE-2011-4086
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-4086>
* CVE-2011-4622
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-4622>
* CVE-2012-0045
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2012-0045>
* CVE-2012-0879
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2012-0879>
* CVE-2012-1090
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2012-1090>
* CVE-2012-1097
  <http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2012-1097>

The ioapic_read_indirect function in virt/kvm/ioapic.c in the Linux kernel through 3.8.4 does not properly handle a certain combination of invalid IOAPIC_REG_SELECT and IOAPIC_REG_WINDOW operations, which allows guest OS users to obtain sensitive information from host OS memory or cause a denial of service (host OS OOPS) via a crafted application.