Advisories ยป MGASA-2019-0098

Updated kernel-linus packages fix security vulnerabilities

Publication date: 20 Feb 2019
Type: security
Affected Mageia releases : 6
CVE: CVE-2018-1128 , CVE-2018-1129 , CVE-2018-14625 , CVE-2018-16862 , CVE-2018-16882 , CVE-2018-16884 , CVE-2018-18397 , CVE-2018-19824 , CVE-2018-19985 , CVE-2019-3701 , CVE-2019-3819 , CVE-2019-6974 , CVE-2019-7221 , CVE-2019-7222


This kernel-linus update is based on the upstream 4.14.100 and fixes at least
the following security issues:

Cross-hyperthread Spectre v2 mitigation is now provided by the Single
Thread Indirect Branch Predictors (STIBP) support. Note that STIBP also
requires the functionality be supported by the Intel microcode in use.

It was found that cephx authentication protocol did not verify ceph clients
correctly and was vulnerable to replay attack. Any attacker having access
to ceph cluster network who is able to sniff packets on network can use
this vulnerability to authenticate with ceph service and perform actions
allowed by ceph service (CVE-2018-1128).

A flaw was found in the way signature calculation was handled by cephx
authentication protocol. An attacker having access to ceph cluster network
who is able to alter the message payload was able to bypass signature
checks done by cephx protocol (CVE-2018-1129).

A flaw was found in the Linux Kernel where an attacker may be able to have
an uncontrolled read to kernel-memory from within a vm guest. A race
condition between connect() and close() function may allow an attacker
using the AF_VSOCK protocol to gather a 4 byte information leak or possibly
intercept or corrupt AF_VSOCK messages destined to other clients

A security flaw was found in the Linux kernel in a way that the cleancache
subsystem clears an inode after the final file truncation (removal). The
new file created with the same inode may contain leftover pages from
cleancache and the old file data instead of the new one (CVE-2018-16862).

A use-after-free issue was found in the way the Linux kernel's KVM
hypervisor processed posted interrupts when nested(=1) virtualization is
enabled. In nested_get_vmcs12_pages(), in case of an error while
processing posted interrupt address, it unmaps the 'pi_desc_page' without
resetting 'pi_desc' descriptor address, which is later used in
pi_test_and_clear_on(). A guest user/process could use this flaw to crash
the host kernel resulting in DoS or potentially gain privileged access to
a system (CVE-2018-16882).

A flaw was found in the Linux kernel's NFS41+ subsystem. NFS41+ shares
mounted in different network namespaces at the same time can make
bc_svc_process() use wrong back-channel IDs and cause a use-after-free
vulnerability. Thus a malicious container user can cause a host kernel
memory corruption and a system panic. Due to the nature of the flaw,
privilege escalation cannot be fully ruled out (CVE-2018-16884).

The userfaultfd implementation in the Linux kernel before 4.19.7 mishandles
access control for certain UFFDIO_ ioctl calls, as demonstrated by allowing
local users to write data into holes in a tmpfs file (if the user has
read-only access to that file, and that file contains holes)

In the Linux kernel through 4.19.6, a local user could exploit a
use-after-free in the ALSA driver by supplying a malicious USB Sound device
(with zero interfaces) (CVE-2018-19824).

A flaw was found in the Linux kernel in the function hso_probe() which
reads if_num value from the USB device (as an u8) and uses it without a
length check to index an array, resulting in an OOB memory read in
hso_probe() or hso_get_config_data(). An attacker with a forged USB
device and physical access to a system (needed to connect such a device)
can cause a system crash and a denial of service (CVE-2018-19985).

An issue was discovered in can_can_gw_rcv in net/can/gw.c in the Linux
kernel through 4.19.13. The CAN frame modification rules allow bitwise
logical operations that can be also applied to the can_dlc field. Because
of a missing check, the CAN drivers may write arbitrary content beyond
the data registers in the CAN controller's I/O memory when processing
can-gw manipulated outgoing frames. This is related to cgw_csum_xor_rel.
An unprivileged user can trigger a system crash (general protection fault)

A flaw was found in the Linux kernel in the function hid_debug_events_read()
in drivers/hid/hid-debug.c file which may enter an infinite loop with
certain parameters passed from a userspace. A local privileged user ("root")
can cause a system lock up and a denial of service (CVE-2019-3819).

In the Linux kernel before 4.20.8, kvm_ioctl_create_device in
virt/kvm/kvm_main.c mishandles reference counting because of a race
condition, leading to a use-after-free (CVE-2019-6974).

A use-after-free vulnerability was found in the way the Linux kernel's KVM
hypervisor emulates a preemption timer for L2 guests when nested (=1)
virtualization is enabled. This high resolution timer(hrtimer) runs when
a L2 guest is active. After VM exit, the sync_vmcs12() timer object is
stopped. The use-after-free occurs if the timer object is freed before
calling sync_vmcs12() routine. A guest user/process could use this flaw
to crash the host kernel resulting in a denial of service or, potentially,
gain privileged access to a system (CVE-2019-7221).

An information leakage issue was found in the way Linux kernel's KVM
hypervisor handled page fault exceptions while emulating instructions
like VMXON, VMCLEAR, VMPTRLD, and VMWRITE with memory address as an
operand. It occurs if the operand is a mmio address, as the returned
exception object holds uninitialized stack memory contents. A guest
user/process could use this flaw to leak host's stack memory contents
to a guest (CVE-2019-7222).

For other uptstream fixes in this update, see the referenced changelogs.