kernel: list corruption of m_list or s_list if unused

The fs_pin implementation in the Linux kernel before 4.0.5 does not ensure the internal consistency of a certain list data structure, which allows local users to cause a denial of service (system crash) by leveraging user-namespace root access for an MNT_DETACH umount2 system call, related to fs/fs_pin.c and include/linux/fs_pin.h.

kernel: Crafted BPF filters may crash kernel during JIT optimisation

The bpf_int_jit_compile function in arch/x86/net/bpf_jit_comp.c in the Linux kernel before 4.0.6 allows local users to cause a denial of service (system crash) by creating a packet filter and then loading crafted BPF instructions that trigger late convergence by the JIT compiler.

gcc: Predictable randomness from std::random_device

The std::random_device class in libstdc++ in the GNU Compiler Collection (aka GCC) before 4.9.4 does not properly handle short reads from blocking sources, which makes it easier for context-dependent attackers to predict the random values via unspecified vectors.

kernel: Creating multiple sockets when SCTP module isn't loaded leads to kernel panic

The sctp_init function in net/sctp/protocol.c in the Linux kernel before 4.2.3 has an incorrect sequence of protocol-initialization steps, which allows local users to cause a denial of service (panic or memory corruption) by creating SCTP sockets before all of the steps have finished.

kernel: local privesc in key management

The KEYS subsystem in the Linux kernel before 4.4 allows local users to gain privileges or cause a denial of service (BUG) via crafted keyctl commands that negatively instantiate a key, related to security/keys/encrypted-keys/encrypted.c, security/keys/trusted.c, and security/keys/user_defined.c.

kernel: information leak in devio.c

The proc_connectinfo function in drivers/usb/core/devio.c in the Linux kernel through 4.6 does not initialize a certain data structure, which allows local users to obtain sensitive information from kernel stack memory via a crafted USBDEVFS_CONNECTINFO ioctl call.

kernel: Information leak in Linux sound module in timer.c

The snd_timer_user_params function in sound/core/timer.c in the Linux kernel through 4.6 does not initialize a certain data structure, which allows local users to obtain sensitive information from kernel stack memory via crafted use of the ALSA timer interface.

kernel: Information leak in events in timer.c

sound/core/timer.c in the Linux kernel through 4.6 does not initialize certain r1 data structures, which allows local users to obtain sensitive information from kernel stack memory via crafted use of the ALSA timer interface, related to the (1) snd_timer_user_ccallback and (2) snd_timer_user_tinterrupt functions.

kernel: Information leak in x25 module

The x25_negotiate_facilities function in net/x25/x25_facilities.c in the Linux kernel before 4.5.5 does not properly initialize a certain data structure, which allows attackers to obtain sensitive information from kernel stack memory via an X.25 Call Request.

The web server in Aternity before 9.0.1 does not require authentication for getMBeansFromURL loading of Java MBeans, which allows remote attackers to execute arbitrary Java code by registering MBeans.

Timing side-channel in github.com/robbert229/jwt

Token validation methods are susceptible to a timing side-channel during HMAC comparison. With a large enough number of requests over a low latency connection, an attacker may use this to determine the expected HMAC.

Elrond go Processing: fallback search of SCRs when not found in the main cache

Elrond-GO is a go implementation for the Elrond Network protocol. Versions prior to 1.3.50 are subject to a processing issue where nodes are affected when trying to process a cross-shard relayed transaction with a smart contract deploy transaction data. The problem was a bad correlation between the transaction caches and the processing component. If the above-mentioned transaction was sent with more gas than required, the smart contract result (SCR transaction) that should have returned the leftover gas, would have been wrongly added to a cache that the processing unit did not consider. The node stopped notarizing metachain blocks. The fix was actually to extend the SCR transaction search in all other caches if it wasn't found in the correct (expected) sharded-cache. There are no known workarounds at this time. This issue has been patched in version 1.3.50.

The ntpd_driver component before 1.3.0 and 2.x before 2.2.0 for Robot Operating System (ROS) allows attackers, who control the source code of a different node in the same ROS application, to change a robot's behavior. This occurs because a topic name depends on the attacker-controlled time_ref_topic parameter.

The br_parse_ip_options function in net/bridge/br_netfilter.c in the Linux kernel before 2.6.39 does not properly initialize a certain data structure, which allows remote attackers to cause a denial of service by leveraging connectivity to a network interface that uses an Ethernet bridge device.

kernel: emergency route cache flushing leads to node deadlock

A certain Red Hat patch for net/ipv4/route.c in the Linux kernel 2.6.18 on Red Hat Enterprise Linux (RHEL) 5 allows remote attackers to cause a denial of service (deadlock) via crafted packets that force collisions in the IPv4 routing hash table, and trigger a routing "emergency" in which a hash chain is too long. NOTE: this is related to an issue in the Linux kernel before 2.6.31, when the kernel routing cache is disabled, involving an uninitialized pointer and a panic.

gnome-screensaver: Race condition between shaking the unlock dialog and clearing the screen

gdk/gdkwindow.c in GTK+ before 2.18.5, as used in gnome-screensaver before 2.28.1, performs implicit paints on windows of type GDK_WINDOW_FOREIGN, which triggers an X error in certain circumstances and consequently allows physically proximate attackers to bypass screen locking and access an unattended workstation by pressing the Enter key many times.

kdebase: race condition may allow local attackers to bypass screen locking

Race condition in workspace/krunner/lock/lockdlg.cc in the KRunner lock module in kdebase in KDE SC 4.4.0 allows physically proximate attackers to bypass KScreenSaver screen locking and access an unattended workstation by pressing the Enter key at a certain time, related to multiple forked processes.

kernel: NFS: Fix an Oops when truncating a file

The nfs_wait_on_request function in fs/nfs/pagelist.c in Linux kernel 2.6.x through 2.6.33-rc5 allows attackers to cause a denial of service (Oops) via unknown vectors related to truncating a file and an operation that is not interruptible.

kvm: vulnerability in the hvc_console code that gets exposed via the new virtio_console functionality

Race condition in the hvc_close function in drivers/char/hvc_console.c in the Linux kernel before 2.6.34 allows local users to cause a denial of service or possibly have unspecified other impact by closing a Hypervisor Virtual Console device, related to the hvc_open and hvc_remove functions.

The pfs_getextattr function in FreeBSD 7.x before 7.3-RELEASE and 8.x before 8.0-RC1 unlocks a mutex that was not previously locked, which allows local users to cause a denial of service (kernel panic), overwrite arbitrary memory locations, and possibly execute arbitrary code via vectors related to opening a file on a file system that uses pseudofs.