[SECURITY] [DSA 5096-1] linux security update
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
- -------------------------------------------------------------------------
Debian Security Advisory DSA-5096-1 security@debian.org
https://www.debian.org/security/ Salvatore Bonaccorso
March 09, 2022 https://www.debian.org/security/faq
- -------------------------------------------------------------------------
Package : linux
CVE ID : CVE-2020-29374 CVE-2020-36322 CVE-2021-3640 CVE-2021-3744
CVE-2021-3752 CVE-2021-3760 CVE-2021-3764 CVE-2021-3772
CVE-2021-4002 CVE-2021-4083 CVE-2021-4135 CVE-2021-4155
CVE-2021-4203 CVE-2021-20317 CVE-2021-20321 CVE-2021-20322
CVE-2021-22600 CVE-2021-28711 CVE-2021-28712 CVE-2021-28713
CVE-2021-28714 CVE-2021-28715 CVE-2021-28950 CVE-2021-38300
CVE-2021-39685 CVE-2021-39686 CVE-2021-39698 CVE-2021-39713
CVE-2021-41864 CVE-2021-42739 CVE-2021-43389 CVE-2021-43975
CVE-2021-43976 CVE-2021-44733 CVE-2021-45095 CVE-2021-45469
CVE-2021-45480 CVE-2022-0001 CVE-2022-0002 CVE-2022-0322
CVE-2022-0330 CVE-2022-0435 CVE-2022-0487 CVE-2022-0492
CVE-2022-0617 CVE-2022-0644 CVE-2022-22942 CVE-2022-24448
CVE-2022-24959 CVE-2022-25258 CVE-2022-25375
Debian Bug : 988044 989285 990411 994050
Several vulnerabilities have been discovered in the Linux kernel that
may lead to a privilege escalation, denial of service or information
leaks.
CVE-2020-29374
Jann Horn of Google reported a flaw in Linux's virtual memory
management. A parent and child process initially share all their
memory, but when either writes to a shared page, the page is
duplicated and unshared (copy-on-write). However, in case an
operation such as vmsplice() required the kernel to take an
additional reference to a shared page, and a copy-on-write occurs
during this operation, the kernel might have accessed the wrong
process's memory. For some programs, this could lead to an
information leak or data corruption.
This issue was already fixed for most architectures, but not on
MIPS and System z. This update corrects that.
CVE-2020-36322, CVE-2021-28950
The syzbot tool found that the FUSE (filesystem-in-user-space)
implementation did not correctly handle a FUSE server returning
invalid attributes for a file. A local user permitted to run a
FUSE server could use this to cause a denial of service (crash).
The original fix for this introduced a different potential denial
of service (infinite loop in kernel space), which has also been
fixed.
CVE-2021-3640
Lin Ma discovered a race condiiton in the Bluetooth protocol
implementation that can lead to a use-after-free. A local
user could exploit this to cause a denial of service (memory
corruption or crash) or possibly for privilege escalation.
CVE-2021-3744, CVE-2021-3764
minihanshen reported bugs in the ccp driver for AMD
Cryptographic Coprocessors that could lead to a resource leak.
On systems using this driver, a local user could exploit this to
cause a denial of service.
CVE-2021-3752
Likang Luo of NSFOCUS Security Team discovered a flaw in the
Bluetooth L2CAP implementation that can lead to a user-after-free.
A local user could exploit this to cause a denial of service
(memory corruption or crash) or possibly for privilege escalation.
CVE-2021-3760, CVE-2021-4202
Lin Ma discovered race conditions in the NCI (NFC Controller
Interface) driver, which could lead to a use-after-free. A local
user could exploit this to cause a denial of service (memory
corruption or crash) or possibly for privilege escalation.
This driver is not enabled in Debian's official kernel
configurations.
CVE-2021-3772
A flaw was found in the SCTP protocol implementation, which would
allow a networked attacker to break an SCTP association. The
attacker would only need to know or guess the IP addresses and
ports for the association.
CVE-2021-4002
It was discovered that hugetlbfs, the virtual filesystem used by
applications to allocate huge pages in RAM, did not flush the
CPU's TLB in one case where it was necessary. In some
circumstances a local user would be able to read and write huge
pages after they are freed and reallocated to a different process.
This could lead to privilege escalation, denial of service or
information leaks.
CVE-2021-4083
Jann Horn reported a race condition in the local (Unix) sockets
garbage collector, that can lead to use-after-free. A local user
could exploit this to cause a denial of service (memory corruption
or crash) or possibly for privilege escalation.
CVE-2021-4135
A flaw was found in the netdevsim driver which would lead to an
information leak.
This driver is not enabled in Debian's official kernel
configurations.
CVE-2021-4155
Kirill Tkhai discovered a data leak in the way the XFS_IOC_ALLOCSP
IOCTL in the XFS filesystem allowed for a size increase of files
with unaligned size. A local attacker can take advantage of this
flaw to leak data on the XFS filesystem.
CVE-2021-4203
Jann Horn reported a race condition in the local (Unix) sockets
implementation that can lead to a use-after-free. A local user
could exploit this to leak sensitive information from the kernel.
CVE-2021-20317
It was discovered that the timer queue structure could become
corrupt, leading to waiting tasks never being woken up. A local
user with certain privileges could exploit this to cause a denial
of service (system hang).
CVE-2021-20321
A race condition was discovered in the overlayfs filesystem
driver. A local user with access to an overlayfs mount and to its
underlying upper directory could exploit this for privilege
escalation.
CVE-2021-20322
An information leak was discovered in the IPv4 implementation. A
remote attacker could exploit this to quickly discover which UDP
ports a system is using, making it easier for them to carry out a
DNS poisoning attack against that system.
CVE-2021-22600
The syzbot tool found a flaw in the packet socket (AF_PACKET)
implementation which could lead to incorrectly freeing memory. A
local user with CAP_NET_RAW capability (in any user namespace)
could exploit this for denial of service (memory corruption or
crash) or possibly for privilege escalation.
CVE-2021-28711, CVE-2021-28712, CVE-2021-28713 (XSA-391)
Juergen Gross reported that malicious PV backends can cause a denial
of service to guests being serviced by those backends via high
frequency events, even if those backends are running in a less
privileged environment.
CVE-2021-28714, CVE-2021-28715 (XSA-392)
Juergen Gross discovered that Xen guests can force the Linux
netback driver to hog large amounts of kernel memory, resulting in
denial of service.
CVE-2021-38300
Piotr Krysiuk discovered a flaw in the classic BPF (cBPF) JIT
compiler for MIPS architectures. A local user could exploit
this to excute arbitrary code in the kernel.
This issue is mitigated by setting sysctl
net.core.bpf_jit_enable=0, which is the default. It is *not*
mitigated by disabling unprivileged use of eBPF.
CVE-2021-39685
Szymon Heidrich discovered a buffer overflow vulnerability in the
USB gadget subsystem, resulting in information disclosure, denial of
service or privilege escalation.
CVE-2021-39686
A race condition was discovered in the Android binder driver, that
could lead to incorrect security checks. On systems where the
binder driver is loaded, a local user could exploit this for
privilege escalation.
CVE-2021-39698
Linus Torvalds reported a flaw in the file polling implementation,
which could lead to a use-after-free. A local user could exploit
this for denial of service (memory corruption or crash) or
possibly for privilege escalation.
CVE-2021-39713
The syzbot tool found a race condition in the network scheduling
subsystem which could lead to a use-after-free. A local user
could exploit this for denial of service (memory corruption or
crash) or possibly for privilege escalation.
CVE-2021-41864
An integer overflow was discovered in the Extended BPF (eBPF)
subsystem. A local user could exploit this for denial of service
(memory corruption or crash), or possibly for privilege
escalation.
This can be mitigated by setting sysctl
kernel.unprivileged_bpf_disabled=1, which disables eBPF use by
unprivileged users.
CVE-2021-42739
A heap buffer overflow was discovered in the firedtv driver for
FireWire-connected DVB receivers. A local user with access to a
firedtv device could exploit this for denial of service (memory
corruption or crash), or possibly for privilege escalation.
CVE-2021-43389
The Active Defense Lab of Venustech discovered a flaw in the CMTP
subsystem as used by Bluetooth, which could lead to an
out-of-bounds read and object type confusion. A local user with
CAP_NET_ADMIN capability in the initial user namespace could
exploit this for denial of service (memory corruption or crash),
or possibly for privilege escalation.
CVE-2021-43975
Brendan Dolan-Gavitt reported a flaw in the
hw_atl_utils_fw_rpc_wait() function in the aQuantia AQtion ethernet
device driver which can result in denial of service or the execution
of arbitrary code.
CVE-2021-43976
Zekun Shen and Brendan Dolan-Gavitt discovered a flaw in the
mwifiex_usb_recv() function of the Marvell WiFi-Ex USB Driver. An
attacker able to connect a crafted USB device can take advantage of
this flaw to cause a denial of service.
CVE-2021-44733
A race condition was discovered in the Trusted Execution
Environment (TEE) subsystem for Arm processors, which could lead
to a use-after-free. A local user permitted to access a TEE
device could exploit this for denial of service (memory corruption
or crash) or possibly for privilege escalation.
CVE-2021-45095
It was discovered that the Phone Network protocol (PhoNet) driver
has a reference count leak in the pep_sock_accept() function.
CVE-2021-45469
Wenqing Liu reported an out-of-bounds memory access in the f2fs
implementation if an inode has an invalid last xattr entry. An
attacker able to mount a specially crafted image can take advantage
of this flaw for denial of service.
CVE-2021-45480
A memory leak flaw was discovered in the __rds_conn_create()
function in the RDS (Reliable Datagram Sockets) protocol subsystem.
CVE-2022-0001 (INTEL-SA-00598)
Researchers at VUSec discovered that the Branch History Buffer in
Intel processors can be exploited to create information side-
channels with speculative execution. This issue is similar to
Spectre variant 2, but requires additional mitigations on some
processors.
This can be exploited to obtain sensitive information from a
different security context, such as from user-space to the kernel,
or from a KVM guest to the kernel.
CVE-2022-0002 (INTEL-SA-00598)
This is a similar issue to CVE-2022-0001, but covers exploitation
within a security context, such as from JIT-compiled code in a
sandbox to hosting code in the same process.
This can be partly mitigated by disabling eBPF for unprivileged
users with the sysctl: kernel.unprivileged_bpf_disabled=2. This
update does that by default.
CVE-2022-0322
Eiichi Tsukata discovered a flaw in the sctp_make_strreset_req()
function in the SCTP network protocol implementation which can
result in denial of service.
CVE-2022-0330
Sushma Venkatesh Reddy discovered a missing GPU TLB flush in the
i915 driver, resulting in denial of service or privilege escalation.
CVE-2022-0435
Samuel Page and Eric Dumazet reported a stack overflow in the
networking module for the Transparent Inter-Process Communication
(TIPC) protocol, resulting in denial of service or potentially the
execution of arbitrary code.
CVE-2022-0487
A use-after-free was discovered in the MOXART SD/MMC Host Controller
support driver. This flaw does not impact the Debian binary packages
as CONFIG_MMC_MOXART is not set.
CVE-2022-0492
Yiqi Sun and Kevin Wang reported that the cgroup-v1 subsystem does
not properly restrict access to the release-agent feature. A local
user can take advantage of this flaw for privilege escalation and
bypass of namespace isolation.
CVE-2022-0617
butt3rflyh4ck discovered a NULL pointer dereference in the UDF
filesystem. A local user that can mount a specially crafted UDF
image can use this flaw to crash the system.
CVE-2022-0644
Hao Sun reported a missing check for file read permission in the
finit_module() and kexec_file_load() system calls. The security
impact of this is unclear, since these system calls are usually
only available to the root user.
CVE-2022-22942
It was discovered that wrong file file descriptor handling in the
VMware Virtual GPU driver (vmwgfx) could result in information leak
or privilege escalation.
CVE-2022-24448
Lyu Tao reported a flaw in the NFS implementation in the Linux
kernel when handling requests to open a directory on a regular file,
which could result in a information leak.
CVE-2022-24959
A memory leak was discovered in the yam_siocdevprivate() function of
the YAM driver for AX.25, which could result in denial of service.
CVE-2022-25258
Szymon Heidrich reported the USB Gadget subsystem lacks certain
validation of interface OS descriptor requests, resulting in memory
corruption.
CVE-2022-25375
Szymon Heidrich reported that the RNDIS USB gadget lacks validation
of the size of the RNDIS_MSG_SET command, resulting in information
leak from kernel memory.
For the oldstable distribution (buster), these problems have been
fixed in version 4.19.232-1. This update additionally includes many
more bug fixes from stable updates 4.19.209-4.19.232 inclusive.
We recommend that you upgrade your linux packages.
For the detailed security status of linux please refer to
its security tracker page at:
https://security-tracker.debian.org/tracker/linux
Further information about Debian Security Advisories, how to apply
these updates to your system and frequently asked questions can be
found at: https://www.debian.org/security/
Mailing list: debian-security-announce@lists.debian.org
-----BEGIN PGP SIGNATURE-----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=5fQT
-----END PGP SIGNATURE-----
Reply to: