RHSA-2017:0086 Important: kernel security, bug fix, and enhancement update

Updated

An update for kernel is now available for Red Hat Enterprise Linux 7.

Red Hat Product Security has rated this update as having a security impact of Important. A Common Vulnerability Scoring System (CVSS) base score, which gives
a detailed severity rating, is available for each vulnerability from the CVE links.

The kernel packages contain the Linux kernel, the core of any Linux operating system.

  • A use-after-free vulnerability was found in the kernels socket recvmmsg subsystem. This may allow remote attackers to corrupt memory and may allow execution of arbitrary code. This corruption takes place during the error handling routines within __sys_recvmmsg() function. This content is not included.CVE-2016-7117, Important

  • A use-after-free vulnerability was found in tcp_xmit_retransmit_queue and other tcp_* functions. This condition could allow an attacker to send an incorrect selective acknowledgment to existing connections, possibly resetting a connection. This content is not included.CVE-2016-6828, Moderate

  • A flaw was found in the Linux kernels implementation of sctp protocol in which a remote attacker can trigger an out of bounds read with an offset of up to 64kB. This may panic the machine with a page-fault. This content is not included.CVE-2016-9555, Moderate

This update also fixes the following bugs:

  • Previously, the performance of Internet Protocol over InfiniBand (IPoIB) was suboptimal due to a conflict of IPoIB with Generic Receive Offload (GRO). This update moves the data cached by the IPoIB driver from a control block into the IPoIB hard header, thus avoiding the GRO problem and the corruption of IPoIB address information. As a result, the performance of IPoIB has been improved. This content is not included.BZ#1390668

  • Previously, a NFS server with Red Hat Enterprise Linux 7 sometimes terminated unexpectedly when the nfsd4_process_open2() function was processing an OPEN request. This was happening due to a race condition which could corrupt the data structures representing the NFSv4 server state. This update fixes the state management to not corrupt the NFSv4 server state, and the NFS server no longer crashes under the described circumstances. This content is not included.BZ#1390672

  • When a virtual machine (VM) with 2 PCI-Passthrough Ethernet interfaces attached was created, deleted and recreated, the operating system terminated unexpectedly and rebooted during the recreation. This update fixes the race condition between the eventfd and virqfd signaling mechanisms in the vfio driver. As a result, the operating system tnow boots without crashing in the described situation. This content is not included.BZ#1391611

  • Previously, a packet loss occurred when the team driver in round-robin mode was sending a large number of packets. This update fixes counting of the packets in the round-robin runner of the team driver, and the packet loss no longer occurs in the described situation. This content is not included.BZ#1392023

  • Previously, deleting a network namespace (netns) occasionally ended by a kernel oops. This was happening because the virtual network devices contained in the deleted namespace could be deleted in any order. If the loopback device was not deleted as the last item, others netns devices, such as vxlan devices, could end up with dangling references to the loopback device. With this update, the underlying source code has been fixed to ensure the correct order when deleting the virtual network devices on netns deletion. As a result, the kernel oops no longer occurs under the described circumstances. This content is not included.BZ#1392024

  • Previously, a Kabylake system with a Sunrise Point Platform Controller Hub (PCH) with a PCI device ID of 0xA149, incorrectly showed the following warning message during the boot:

    Unknown Intel PCH (0xa149) detected.
Warning: Intel Kabylake processor with unknown PCH - this hardware has not undergone testing by Red Hat and might not be certified. Please consult https://hardware.redhat.com for certified hardware.

    This update fixes the underlying source code to properly recognize this PCH. As a result, the operating system now boots without displaying the warning messages. This content is not included.BZ#1392033

  • Previously, the operating system occasionally locked up after a long run. This was caused by a race condition between the try_to_wake_up() function and a woken up task in the core scheduler. With this update, the race condition has been fixed, and the operating system no longer locks up in the described scenario. This content is not included.BZ#1393719

  • Previously, non-root users were not allowed to perform data compression with the genwqe driver, because the access rights for the /dev/genwqe* files were set to 600. This update changes the permissions for /dev/genwqe* to 666, and non-root users are now able to perform data compression with genwqe. This content is not included.BZ#1393723

  • Previously, an occasional kernel oops occurred when a PCI adapter was hotplugged. With this update, the pci_host_bridge.release_fn() function has been fixed, and the kernel oops no longer occurs under described circumstances. This content is not included.BZ#1393724

  • Previously, data corruption or data loss sometimes occurred for data that were written by Direct Access (DAX) IO. This was happening due to a race condition between DAX IO and a page fault, which converted unwritten extents and zeroed them out. With this update, pre-zeroed blocks are used for DAX IO so that conversion is no longer necessary in the fault path, which eliminates the potential race condition. As a result, neither data loss nor data corruption occurs under the described circumstances. This content is not included.BZ#1394707

  • When a change of some ethtool features, such as Transmission Control Protocol Segmentation Offload/Generic Segmentation Offload (TSO/GSO), was followed by unloading the i40e driver, a kernel crash sometimes occurred on systems with more than 127 online CPUs. This was happening because all interrupt vectors on i40e are allocated to transmitter/receiver queues and Flow Director SideBand (FD_SB) is kept disabled. Changing the ethtool features enables FD_SB without any vectors available. This update adds a check to verify whether the MSI-X vector is available after FD_SB has been enabled. As a result, the kernel no longer crashes under the described circumstances. This content is not included.BZ#1394708

  • Previously, memory corruption by copying data into the wrong memory locations sometimes occurred, because the __copy_tofrom_user() function was returning incorrect values. This update fixes the __copy_tofrom_user() function so that it no longer returns larger values than the number of bytes it was asked to copy. As a result, memory corruption no longer occurs in he described scenario. This content is not included.BZ#1395565

  • Previously, a bonded network interface in a virtual machine did not pass traffic after a failover, and the ixgbe module on the host reported the following warning message:

    ixgbe 0000:01:00.1 em2: VF 0 requested MACVLAN filter but is administratively denied

    This update fixes the problem by adding a check for the vfinfo[].trusted bit into the IXGBE_VF_SET_MACVLAN message handler. As a result, connectivity continues to work and the error message no longer occurs after a failover. This content is not included.BZ#1395572

  • Some hardware configurations, such as HPE Synergy 680 or other similar platforms with 5 or more PCI cards, contain small memory regions, which are omitted by the kernel when allocating memory for kexec Executable and Lincable Format (ELF). Consequently, the operating system previously terminated unexpectedly when loading the kdump kernel. With this update, the underlying source code has been fixed to allocate enough ELF header space for memory ranges. As a result, the operating system now loads kdump kernel as expected under the described circumstances. This content is not included.BZ#1395573

  • Previously, virtual machines (VM) with Single Root I/O Virtualization (SR-IOV) interfaces failed to boot on a host running the Red Hat Enterprise Linux 7.3 operating system with the following libvirt error message:

    error : virNetDevLinkDump:2013 : error dumping ens1f0 (2) interface: Message too long

    This update fixes the rtnetlink message size computation, so that dumping interfaces with SR-IOV virtual functions succeed, and the VMs now boot as expected in the described situation. This content is not included.BZ#1395811

  • When Red Hat Enterprise Linux 7.3 was booted on an Intel Xeon Phi architecture server with Hyper-Threading Technology disabled, the operating system did not enumerate all identified sockets and cores. With this update, the underlying source code has been fixed, and the operating system now enumerates the correct number of sockets and cores in the described scenario. This content is not included.BZ#1398173

  • When one of the drives became unresponsive, all other drives intermittently hung, because the megaraid_sas driver incorrectly sent a reset request to the PowerEdge RAID Controller (PERC). This update fixes megaraid_sas, and thus the hang of one drive no longer leads to intermittent loss of access to all drives on the system. This content is not included.BZ#1398175

  • Previously, an XFS corruption in some cases occurred on Seagate 8TB drive based volumes after a planned system shutdown or reboot, when a disk write back cache was used. With this update, the megaraid_sas driver has been fixed and the XFS corruption no longer occurs in the described scenario. This content is not included.BZ#1398179

  • Previously, certain values calculated for the /proc/<PID>/stat files, were truncated to zero and used as a divisor. As a consequence, a kernel panic could occur during reading /proc/<PID>/stat. This update defines a new div64_ul() division and corrects the affected calculation in the proc_sched_show_task() function. As a result, the kernel panic no longer occurs in the described situation. This content is not included.BZ1398361

  • Previously, the mount.cifs command was unable to mount a Server Message Block (SMB) share folder using the CIFS kernel module in case that the user mounting did not have access to all the sub-folders along the share path. With this update, a set of patches has been applied to the CIFS kernel module. As a result, mount.cifs now mounts a SMB share folder as expected in the described scenario. This content is not included.BZ#1400055

  • When the operating system was suspended to sleep state (S3) by using the pm-suspend command, the operating system became unresponsive until it was forced to power down or to reboot. This update fixes the Advanced Configuration and Power Interface (ACPI) to not save Non Volatile Sleeping (NVS) memory region to accelerate S3. As a result, the operating system now resumes from S3 as expected. This content is not included.BZ#1402326

Article Type