In the Linux kernel, the following vulnerability has been resolved:
tracing: Have show_event_filters/triggers files take trace array ref
The newly added files show_event_filters and show_event_triggers that show
all filters or triggers that are set within the trace array do not take a
reference for the trace array it is showing. Without taking a reference,
the trace_array may be freed via "rmdir" while a task is reading one of
theses files. Those files iterate all the events within an instance
(trace_array) and nothing prevents that instance from being freed while
its data is being read. This causes a use-after-free crash.
Have the open of both those files take the trace_array reference via the
trace_array_get() that prevents the trace_array from being freed while the
files are opened.
CVSS Vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
CVSS Score: 7.8
AV:L - The UAF is reached only via local tracefs: open/read of instances/<name>/show_event_filters or show_event_triggers (ftrace_event_show_*_open -> seq_read -> t_start) concurrent with rmdir(2) of that instance (tracefs_syscall_rmdir -> instance_rmdir -> __remove_instance). There is no network, adjacent-radio, or physical path.
AC:L - The attacker owns both sides and can trigger it sequentially: mkdir an instance, open show_event_filters/triggers (no trace_array ref), rmdir it (__remove_instance sees tr->ref==1 and kfree()s tr), then read the still-open fd. No victim state or uninfluenced layout is required; CONFIG_TRACING is standard.
PR:L - The open path checks only LOCKDOWN_TRACEFS and DAC (files are 0440; instance mkdir/rmdir has no capable() gate). Tracefs uid=/gid=/mode= mounts on Android/Perfetto, ChromeOS, and distro tracing groups grant this to unprivileged users, matching CVE-2024-50207, CVE-2025-38101, and CVE-2026-74634.
UI:N - The attacker creates the instance, opens the seq file, rmdirs it, and reads the fd from their own threads. No separate victim mount, click, or cooperating process is required.
S:U - The use-after-free corrupts the host kernel's trace_array and event-file heap objects within the same OS security authority. It enables local privilege escalation but does not cross a VM, IOMMU, sandbox, or hardware trust boundary.
C:H - After rmdir kfree()s the trace_array and its trace_event_file objects, t_start/t_next walk the freed tr->events list and t_show_filters copies filter->filter_string to userspace via seq_printf, disclosing recycled kernel heap. Use-after-free is scored C:H.
I:H - t_show_triggers invokes data->cmd_ops->print() on use-after-free trigger objects, and walking the freed tr->events list is a heap UAF write/control-flow primitive after spraying. Use-after-free is scored I:H.
A:H - syzkaller reported a KASAN slab-use-after-free in t_start/t_next when the instance is removed while the file is read, which oopses or panics the kernel. Any use-after-free of the trace_array is a reliable whole-system crash even without full exploitation.
| Attack Vector |
Local |
Scope |
Unchanged |
| Attack Complexity |
Low |
Confidentiality Impact |
High |
| Privileges Required |
Low |
Integrity Impact |
High |
| User Interaction |
None |
Availability Impact |
High |
AV:L - The UAF is reached only via local tracefs: open/read of instances/<name>/show_event_filters or show_event_triggers (ftrace_event_show_*_open -> seq_read -> t_start) concurrent with rmdir(2) of that instance (tracefs_syscall_rmdir -> instance_rmdir -> __remove_instance). There is no network, adjacent-radio, or physical path.
AC:L - The attacker owns both sides and can trigger it sequentially: mkdir an instance, open show_event_filters/triggers (no trace_array ref), rmdir it (__remove_instance sees tr->ref==1 and kfree()s tr), then read the still-open fd. No victim state or uninfluenced layout is required; CONFIG_TRACING is standard.
PR:L - The open path checks only LOCKDOWN_TRACEFS and DAC (files are 0440; instance mkdir/rmdir has no capable() gate). Tracefs uid=/gid=/mode= mounts on Android/Perfetto, ChromeOS, and distro tracing groups grant this to unprivileged users, matching CVE-2024-50207, CVE-2025-38101, and CVE-2026-74634.
UI:N - The attacker creates the instance, opens the seq file, rmdirs it, and reads the fd from their own threads. No separate victim mount, click, or cooperating process is required.
S:U - The use-after-free corrupts the host kernel's trace_array and event-file heap objects within the same OS security authority. It enables local privilege escalation but does not cross a VM, IOMMU, sandbox, or hardware trust boundary.
C:H - After rmdir kfree()s the trace_array and its trace_event_file objects, t_start/t_next walk the freed tr->events list and t_show_filters copies filter->filter_string to userspace via seq_printf, disclosing recycled kernel heap. Use-after-free is scored C:H.
I:H - t_show_triggers invokes data->cmd_ops->print() on use-after-free trigger objects, and walking the freed tr->events list is a heap UAF write/control-flow primitive after spraying. Use-after-free is scored I:H.
A:H - syzkaller reported a KASAN slab-use-after-free in t_start/t_next when the instance is removed while the file is read, which oopses or panics the kernel. Any use-after-free of the trace_array is a reliable whole-system crash even without full exploitation.
CVSS 3.1