The microcode on AMD 16h 00h through 0Fh processors does not properly handle the interaction between locked instructions and write-combined memory types, which allows local users to cause a denial of service (system hang) via a crafted application, aka the errata 793 issue.
https://exchange.xforce.ibmcloud.com/vulnerabilities/89335
https://bugzilla.redhat.com/show_bug.cgi?id=1035823
http://www.zdnet.com/blog/hardware/amd-owns-up-to-cpu-bug/18924
http://www.securitytracker.com/id/1029415
http://www.securityfocus.com/bid/63983
http://www.openwall.com/lists/oss-security/2013/12/02/1
http://www.debian.org/security/2015/dsa-3128
http://support.amd.com/TechDocs/51810_16h_00h-0Fh_Rev_Guide.pdf
http://security.gentoo.org/glsa/glsa-201407-03.xml
http://secunia.com/advisories/55840
http://rhn.redhat.com/errata/RHSA-2014-0285.html
http://openwall.com/lists/oss-security/2013/11/28/1
http://lists.opensuse.org/opensuse-security-announce/2014-04/msg00000.html
http://lists.opensuse.org/opensuse-security-announce/2014-03/msg00026.html
http://lists.opensuse.org/opensuse-security-announce/2014-03/msg00021.html
http://lists.opensuse.org/opensuse-security-announce/2014-03/msg00015.html
http://lists.opensuse.org/opensuse-security-announce/2014-03/msg00011.html
http://lists.opensuse.org/opensuse-security-announce/2014-03/msg00010.html
http://lists.fedoraproject.org/pipermail/package-announce/2013-December/124199.html
http://lists.fedoraproject.org/pipermail/package-announce/2013-December/124195.html
http://lists.fedoraproject.org/pipermail/package-announce/2013-December/123553.html
http://lists.dragonflybsd.org/pipermail/kernel/2011-December/046594.html