Perf Events and tool security

Overview

Usage of Performance Counters for Linux (perf_events) [1] , [2] , [3] can impose a considerable risk of leaking sensitive data accessed by monitored processes. The data leakage is possible both in scenarios of direct usage of perf_events system call API [2] and over data files generated by Perf tool user mode utility (Perf) [3] , [4] . The risk depends on the nature of data that perf_events performance monitoring units (PMU) [2] and Perf collect and expose for performance analysis. Collected system and performance data may be split into several categories:

  1. System hardware and software configuration data, for example: a CPU model and its cache configuration, an amount of available memory and its topology, used kernel and Perf versions, performance monitoring setup including experiment time, events configuration, Perf command line parameters, etc.

  2. User and kernel module paths and their load addresses with sizes, process and thread names with their PIDs and TIDs, timestamps for captured hardware and software events.

  3. Content of kernel software counters (e.g., for context switches, page faults, CPU migrations), architectural hardware performance counters (PMC) [8] and machine specific registers (MSR) [9] that provide execution metrics for various monitored parts of the system (e.g., memory controller (IMC), interconnect (QPI/UPI) or peripheral (PCIe) uncore counters) without direct attribution to any execution context state.

  4. Content of architectural execution context registers (e.g., RIP, RSP, RBP on x86_64), process user and kernel space memory addresses and data, content of various architectural MSRs that capture data from this category.

Data that belong to the fourth category can potentially contain sensitive process data. If PMUs in some monitoring modes capture values of execution context registers or data from process memory then access to such monitoring capabilities requires to be ordered and secured properly. So, perf_events/Perf performance monitoring is the subject for security access control management [5] .

perf_events/Perf access control

To perform security checks, the Linux implementation splits processes into two categories [6] : a) privileged processes (whose effective user ID is 0, referred to as superuser or root), and b) unprivileged processes (whose effective UID is nonzero). Privileged processes bypass all kernel security permission checks so perf_events performance monitoring is fully available to privileged processes without access, scope and resource restrictions.

Unprivileged processes are subject to a full security permission check based on the process’s credentials [5] (usually: effective UID, effective GID, and supplementary group list).

Linux divides the privileges traditionally associated with superuser into distinct units, known as capabilities [6] , which can be independently enabled and disabled on per-thread basis for processes and files of unprivileged users.

Unprivileged processes with enabled CAP_SYS_ADMIN capability are treated as privileged processes with respect to perf_events performance monitoring and bypass scope permissions checks in the kernel.

Unprivileged processes using perf_events system call API is also subject for PTRACE_MODE_READ_REALCREDS ptrace access mode check [7] , whose outcome determines whether monitoring is permitted. So unprivileged processes provided with CAP_SYS_PTRACE capability are effectively permitted to pass the check.

Other capabilities being granted to unprivileged processes can effectively enable capturing of additional data required for later performance analysis of monitored processes or a system. For example, CAP_SYSLOG capability permits reading kernel space memory addresses from /proc/kallsyms file.

perf_events/Perf privileged users

Mechanisms of capabilities, privileged capability-dumb files [6] and file system ACLs [10] can be used to create a dedicated group of perf_events/Perf privileged users who are permitted to execute performance monitoring without scope limits. The following steps can be taken to create such a group of privileged Perf users.

  1. Create perf_users group of privileged Perf users, assign perf_users group to Perf tool executable and limit access to the executable for other users in the system who are not in the perf_users group:

# groupadd perf_users
# ls -alhF
-rwxr-xr-x  2 root root  11M Oct 19 15:12 perf
# chgrp perf_users perf
# ls -alhF
-rwxr-xr-x  2 root perf_users  11M Oct 19 15:12 perf
# chmod o-rwx perf
# ls -alhF
-rwxr-x---  2 root perf_users  11M Oct 19 15:12 perf

  1. Assign the required capabilities to the Perf tool executable file and enable members of perf_users group with performance monitoring privileges [6] :

# setcap "cap_sys_admin,cap_sys_ptrace,cap_syslog=ep" perf
# setcap -v "cap_sys_admin,cap_sys_ptrace,cap_syslog=ep" perf
perf: OK
# getcap perf
perf = cap_sys_ptrace,cap_sys_admin,cap_syslog+ep

As a result, members of perf_users group are capable of conducting performance monitoring by using functionality of the configured Perf tool executable that, when executes, passes perf_events subsystem scope checks.

This specific access control management is only available to superuser or root running processes with CAP_SETPCAP, CAP_SETFCAP [6] capabilities.

perf_events/Perf unprivileged users

perf_events/Perf scope and access control for unprivileged processes is governed by perf_event_paranoid [2] setting:

-1:

Impose no scope and access restrictions on using perf_events performance monitoring. Per-user per-cpu perf_event_mlock_kb [2] locking limit is ignored when allocating memory buffers for storing performance data. This is the least secure mode since allowed monitored scope is maximized and no perf_events specific limits are imposed on resources allocated for performance monitoring.

>=0:

scope includes per-process and system wide performance monitoring but excludes raw tracepoints and ftrace function tracepoints monitoring. CPU and system events happened when executing either in user or in kernel space can be monitored and captured for later analysis. Per-user per-cpu perf_event_mlock_kb locking limit is imposed but ignored for unprivileged processes with CAP_IPC_LOCK [6] capability.

>=1:

scope includes per-process performance monitoring only and excludes system wide performance monitoring. CPU and system events happened when executing either in user or in kernel space can be monitored and captured for later analysis. Per-user per-cpu perf_event_mlock_kb locking limit is imposed but ignored for unprivileged processes with CAP_IPC_LOCK capability.

>=2:

scope includes per-process performance monitoring only. CPU and system events happened when executing in user space only can be monitored and captured for later analysis. Per-user per-cpu perf_event_mlock_kb locking limit is imposed but ignored for unprivileged processes with CAP_IPC_LOCK capability.

perf_events/Perf resource control

Open file descriptors

The perf_events system call API [2] allocates file descriptors for every configured PMU event. Open file descriptors are a per-process accountable resource governed by the RLIMIT_NOFILE [11] limit (ulimit -n), which is usually derived from the login shell process. When configuring Perf collection for a long list of events on a large server system, this limit can be easily hit preventing required monitoring configuration. RLIMIT_NOFILE limit can be increased on per-user basis modifying content of the limits.conf file [12] . Ordinarily, a Perf sampling session (perf record) requires an amount of open perf_event file descriptors that is not less than the number of monitored events multiplied by the number of monitored CPUs.

Memory allocation

The amount of memory available to user processes for capturing performance monitoring data is governed by the perf_event_mlock_kb [2] setting. This perf_event specific resource setting defines overall per-cpu limits of memory allowed for mapping by the user processes to execute performance monitoring. The setting essentially extends the RLIMIT_MEMLOCK [11] limit, but only for memory regions mapped specifically for capturing monitored performance events and related data.

For example, if a machine has eight cores and perf_event_mlock_kb limit is set to 516 KiB, then a user process is provided with 516 KiB * 8 = 4128 KiB of memory above the RLIMIT_MEMLOCK limit (ulimit -l) for perf_event mmap buffers. In particular, this means that, if the user wants to start two or more performance monitoring processes, the user is required to manually distribute the available 4128 KiB between the monitoring processes, for example, using the –mmap-pages Perf record mode option. Otherwise, the first started performance monitoring process allocates all available 4128 KiB and the other processes will fail to proceed due to the lack of memory.

RLIMIT_MEMLOCK and perf_event_mlock_kb resource constraints are ignored for processes with the CAP_IPC_LOCK capability. Thus, perf_events/Perf privileged users can be provided with memory above the constraints for perf_events/Perf performance monitoring purpose by providing the Perf executable with CAP_IPC_LOCK capability.

Bibliography