CVE-2026-72280

In the Linux kernel KVM arm64 code, CVE-2026-72280 concerns a bogus WARN emitted when a guest could write to the ZCR_EL2 sysreg alias in a nested context (featuring FEAT_NV2). The root cause is the warning leading to a hyp_panic after hyp vector setup; the fix is to drop that bogus WARN. The vuln...

CVE-2026-68455 liveupdate: validate session type before performing operation

In the Linux kernel, the following vulnerability has been resolved: liveupdate: validate session type before performing operation The sessions ioctls are not applicable to all session types. PRESERVEFD is only applicable to outgoing sessions. RETRIEVEFD and FINISH are only valid for incoming...

CVE-2026-68455

The CVE-2026-68455 entry concerns the Linux kernel liveupdate path, where certain session ioctls were allowed on mismatched session types. A user could create an outgoing session, preserve a memfd, then retrieve without a kexec, causing the memfd retrieve handler to run in an invalid post-kexec c...

CVE-2026-68437

In the Linux kernel, the following vulnerability has been resolved: drm/imagination: Fit paired fragment job in the correct CCCB For geometry jobs with a paired fragment job, at the moment, the DRM scheduler's preparejob callback: - checks for internal driver dependencies for the geometry job; -...

CVE-2026-68250

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/sdma5.2: replace BUGON with WARNON There's no need to crash the kernel for these cases. cherry picked from commit ae658afc7f47f6147371ec42cc6b1a793dfdb5af...

CVE-2026-68204

In the Linux kernel, the following vulnerability has been resolved: media: vivid: check for vb2isbusy when toggling caps The vividupdateformatcap/out functions must only be called if the capture/output queue are not busy. But for the controls that select the CROP/COMPOSE/SCALE capability that is...

CVE-2026-68181

In the Linux kernel, the following vulnerability has been resolved: mei: bus: access meidevice under devicelock on cleanup Fix couple of problems in meiclbusdevrelease: meiclflushqueues is running without lock. bus-filelist access after meidevbusputbus can become a use-after-free if this was the...

CVE-2026-68157

In the Linux kernel, the following vulnerability has been resolved: libceph: guard missing CRUSH type name lookup Localized read selection can walk a parent bucket whose name exists in the CRUSH map while its type has no matching entry in typenames. getimmediateparent then dereferences a NULL...

CVE-2026-68131

In the Linux kernel, the following vulnerability has been resolved: rbd: Reset positive result codes to zero in object map update path In a reply message to an RBD request, a positive result code indicates a data payload, which is not allowed for writes. While rbdosdreqcallback already resets a...

CVE-2026-68114

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/gfx12.1: replace BUGON with WARNON There's no need to crash the kernel for these cases. cherry picked from commit e4d99e04b2e9b13b97d3b17804c735f62689db23...

CVE-2026-68111

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/gfx9: replace BUGON with WARNON There's no need to crash the kernel for these cases. cherry picked from commit b71604f8685b0eba07866f4e8dc30f93e1931054...

EUVD-2026-55602

In the Linux kernel, the following vulnerability has been resolved: mtd: fix double free and WARNON in addmtddevice error paths When deviceregister or mtdnvmemadd fails inside addmtddevice for a partition, the error handling triggers mtdrelease via putdevice or deviceunregister. mtdrelease calls...

CVE-2026-68416

Linux kernel CVE-2026-68416 fixes a double-free in add_mtd_device() for MTD partitions and prevents a WARN_ON by overriding dev->type and dev->release before put_device() in error paths. The patch ensures mtd_release() is not called during error handling, avoiding freeing mtd_info twice. Fo...

CVE-2026-68416 mtd: fix double free and WARN_ON in add_mtd_device() error paths

In the Linux kernel, the following vulnerability has been resolved: mtd: fix double free and WARNON in addmtddevice error paths When deviceregister or mtdnvmemadd fails inside addmtddevice for a partition, the error handling triggers mtdrelease via putdevice or deviceunregister. mtdrelease calls...

CVE-2026-68252

The CVE affects the Linux kernel’s DRM AMDGPU SDMA7.0 code path, where the vulnerable usage was a BUG_ON() that could crash the kernel. It has been fixed by replacing BUG_ON() with WARN_ON(), indicating that such checks should warn rather than crash. Affected component: drm/amdgpu/sdma7.0. Root c...

CVE-2026-68250

In the Linux kernel component drm/amdgpu/sdma5.2, CVE-2026-68250 addresses a path where a BUG_ON() could crash the kernel. The vulnerability is resolved by replacing BUG_ON() with WARN_ON(), avoiding a kernel crash in these cases. The fix was cherry-picked from commit ae658afc7f47f6147371ec42cc6b...

EUVD-2026-55347

In the Linux kernel, the following vulnerability has been resolved: drm/amdgpu/gfx11: replace BUGON with WARNON There's no need to crash the kernel for these cases. cherry picked from commit daa62107452d2451787c4248ca38fa2d1a0cbefd...

EUVD-2026-55318

In the Linux kernel, the following vulnerability has been resolved: media: pwc: Drain fillbuf on startstreaming failure pwcisocinit submits its isochronous URBs with usbsubmiturb.., GFPKERNEL in a loop. After the first URB is submitted, its completion handler pwcisochandler can run on another CPU...

CVE-2026-68204

CVE-2026-68204 affects the Linux kernel media subsystem (vivid driver). The vulnerability arises because vivid_update_format_cap/out() may be invoked to change formats while the capture/output queue is still busy, due to missing vb2_is_busy() checks when toggling CROP/COMPOSE/SCALE capabilities. ...

CVE-2026-68204 media: vivid: check for vb2_is_busy() when toggling caps

In the Linux kernel, the following vulnerability has been resolved: media: vivid: check for vb2isbusy when toggling caps The vividupdateformatcap/out functions must only be called if the capture/output queue are not busy. But for the controls that select the CROP/COMPOSE/SCALE capability that is...