| Description: | The kernel-rt packages provide the Real Time Linux Kernel, which enables fine-tuning for systems with extremely high determinism requirements.
Security Fix(es):
Linux kernel built with the Kernel-based Virtual Machine (CONFIG_KVM) support is vulnerable to a null pointer dereference flaw. It could occur on x86 platform, when emulating an undefined instruction. An attacker could use this flaw to crash the host kernel resulting in DoS. (CVE-2016-8630, Important)
A race condition issue leading to a use-after-free flaw was found in the way the raw packet sockets implementation in the Linux kernel networking subsystem handled synchronization while creating the TPACKET_V3 ring buffer. A local user able to open a raw packet socket (requires the CAP_NET_RAW capability) could use this flaw to elevate their privileges on the system. (CVE-2016-8655, Important)
A flaw was discovered in the Linux kernel's implementation of VFIO. An attacker issuing an ioctl can create a situation where memory is corrupted and modify memory outside of the expected area. This may overwrite kernel memory and subvert kernel execution. (CVE-2016-9083, Important)
The use of a kzalloc with an integer multiplication allowed an integer overflow condition to be reached in vfio_pci_intrs.c. This combined with CVE-2016-9083 may allow an attacker to craft an attack and use unallocated memory, potentially crashing the machine. (CVE-2016-9084, Moderate)
Red Hat would like to thank Philip Pettersson for reporting CVE-2016-8655.
Bug Fix(es):
Previously, the asynchronous page fault woke code references spinlocks, which were actually sleeping locks in the RT kernel. Because of this, when the code was executed from the exception context, a bug warning appeared on the console. With this update, the regular wait queue and spinlock code in this area has been modified to use simple-wait-queue and raw-spinlocks. This code change enables the asynchronous page fault code to run in a non-preemptable state without bug warnings. (BZ#1418035)
|