Linux Kernel 5.18

Found Linux Kernel flaw in the i740 driver. The Userspace program could pass any values to the driver through ioctl() interface. The driver doesn't check the value of 'pixclock', so it may cause a divide by zero error.

mm/huge_memory: don't unpoison huge_zero_folio

drm/amdkfd: Check for potential null return of kmalloc_array()

irqchip/gic-v3: Fix GICR_CTLR.RWP polling

veth: Ensure eth header is in skb's linear part

memory: renesas-rpc-if: fix platform-device leak in error path

gpio: Restrict usage of GPIO chip irq members before initialization

net/smc: Fix NULL pointer dereference in smc_pnet_find_ib()

net: atlantic: remove aq_nic_deinit() when resume

i2c: dev: check return value when calling dev_set_name()

net: openvswitch: fix leak of nested actions

powerpc: Fix virt_addr_valid() for 64-bit Book3E & 32-bit

drm/panel: ili9341: fix optional regulator handling

net: ethernet: stmmac: fix altr_tse_pcs function when using a fixed-link

scsi: zorro7xx: Fix a resource leak in zorro7xx_remove_one()

powerpc/memhotplug: Add add_pages override for PPC

btrfs: release correct delalloc amount in direct IO write path

drm/amd/display: Fix by adding FPU protection for dcn30_internal_validate_bw

mm/secretmem: fix panic when growing a memfd_secret

Drivers: hv: vmbus: Fix potential crash on module unload

mmmremap.c: avoid pointless invalidate_range_start/end on mremap(old_size=0)

drm/imx: Fix memory leak in imx_pd_connector_get_modes

dpaa2-ptp: Fix refcount leak in dpaa2_ptp_probe

highmem: fix checks in __kmap_local_sched_{in,out}

iommu/omap: Fix regression in probe for NULL pointer dereference

arch/arm64: Fix topology initialization for core scheduling

ARM: davinci: da850-evm: Avoid NULL pointer dereference

There is a sleep-in-atomic bug in /net/nfc/netlink.c that allows an attacker to crash the Linux kernel by simulating a nfc device from user-space.

crypto: qcom-rng - fix infinite loop on requests not multiple of WORD_SZ

block: null_blk: end timed out poll request

cachefiles: unmark inode in use in error path

platform/x86: thinkpad_acpi: Fix a memory leak of EFCH MMIO resource

A flaw that boot CPU could be vulnerable for the speculative execution behavior kind of attacks in the Linux kernel X86 CPU Power management options functionality was found in the way user resuming CPU from suspend-to-RAM. A local user could use this flaw to potentially get unauthorized access to some memory of the CPU similar to the speculative execution behavior kind of attacks.

Drivers: hv: vmbus: Fix initialization of device object in vmbus_device_register()

drm/amdgpu: Fix call trace warning and hang when removing amdgpu device

btrfs: zoned: traverse devices under chunk_mutex in btrfs_can_activate_zone

parisc: Do not reprogram affinitiy on ASP chip

io_uring: fix fget leak when fs don't support nowait buffered read

mm/memory-failure: fix hugetlb_lock AA deadlock in get_huge_page_for_hwpoison

A flaw was found in the MCTP protocol in the Linux kernel. The function mctp_unregister() reclaims the device's relevant resource when a netcard detaches. However, a running routine may be unaware of this and cause the use-after-free of the mdev->addrs object, potentially leading to a denial of service.