arch/mips/net/bpf_jit.c in the Linux kernel before 5.4.10 can generate undesirable machine code when transforming unprivileged cBPF programs allowing execution of arbitrary code within the kernel context. This occurs because conditional branches can exceed the 128 KB limit of the MIPS architecture.

Go before 1.16.9 and 1.17.x before 1.17.2 has a Buffer Overflow via large arguments in a function invocation from a WASM module when GOARCH=wasm GOOS=js is used.

net/netfilter/nf_conntrack_standalone.c in the Linux kernel before 5.12.2 allows observation of changes in any net namespace because these changes are leaked into all other net namespaces. This is related to the NF_SYSCTL_CT_MAX NF_SYSCTL_CT_EXPECT_MAX and NF_SYSCTL_CT_BUCKETS sysctls.

net/nfc/llcp_sock.c in the Linux kernel before 5.12.10 allows local unprivileged users to cause a denial of service (NULL pointer dereference and BUG) by making a getsockname call after a certain type of failure of a bind call.

drivers/net/ethernet/xilinx/ll_temac_main.c in the Linux kernel before 5.12.13 allows remote attackers to cause a denial of service (buffer overflow and lockup) by sending heavy network traffic for about ten minutes.

The mac80211 subsystem in the Linux kernel before 5.12.13 when a device supporting only 5 GHz is used allows attackers to cause a denial of service (NULL pointer dereference in the radiotap parser) by injecting a frame with 802.11a rates.

drivers/net/ethernet/xilinx/xilinx_emaclite.c in the Linux kernel before 5.13.3 makes it easier for attackers to defeat an ASLR protection mechanism because it prints a kernel pointer (i.e. the real IOMEM pointer).

drivers/usb/host/max3421-hcd.c in the Linux kernel before 5.13.6 allows physically proximate attackers to cause a denial of service (use-after-free and panic) by removing a MAX-3421 USB device in certain situations.

btrfs in the Linux kernel before 5.13.4 allows attackers to cause a denial of service (deadlock) via processes that trigger allocation of new system chunks during times when there is a shortage of free space in the system space_info.

fs/nfsd/trace.h in the Linux kernel before 5.13.4 might allow remote attackers to cause a denial of service (out-of-bounds read in strlen) by sending NFS traffic when the trace event framework is being used for nfsd.

net/sunrpc/xdr.c in the Linux kernel before 5.13.4 allows remote attackers to cause a denial of service (xdr_set_page_base slab-out-of-bounds access) by performing many NFS 4.2 READ_PLUS operations.

arch/powerpc/perf/core-book3s.c in the Linux kernel before 5.12.13 on systems with perf_event_paranoid=-1 and no specific PMU driver support registered allows local users to cause a denial of service (perf_instruction_pointer NULL pointer dereference and OOPS) via a "perf record" command.

fs/nfs/nfs4client.c in the Linux kernel before 5.13.4 has incorrect connection-setup ordering which allows operators of remote NFSv4 servers to cause a denial of service (hanging of mounts) by arranging for those servers to be unreachable during trunking detection.

arch/x86/kvm/mmu/paging_tmpl.h in the Linux kernel before 5.12.11 incorrectly computes the access permissions of a shadow page leading to a missing guest protection page fault.

An issue was discovered in the tokio crate before 1.8.1 for Rust. Upon a JoinHandle::abort, a Task may be dropped in the wrong thread.

GNU cpio through 2.13 allows attackers to execute arbitrary code via a crafted pattern file because of a dstring.c ds_fgetstr integer overflow that triggers an out-of-bounds heap write. NOTE: it is unclear whether there are common cases where the pattern file associated with the -E option is untrusted data.

In kernel/bpf/hashtab.c in the Linux kernel through 5.13.8 there is an integer overflow and out-of-bounds write when many elements are placed in a single bucket. NOTE: exploitation might be impractical without the CAP_SYS_ADMIN capability.