## File: README.md

# BPF Compiler Collection (BCC)

BCC is a toolkit for creating efficient kernel tracing and manipulation
programs, and includes several useful tools and examples. It makes use of
extended BPF (Berkeley Packet Filters), formally known as eBPF, a new feature
that was first added to Linux 3.15. Much of what BCC uses requires Linux 4.1
and above.

eBPF was [described by](https://lkml.org/lkml/2015/4/14/232) Ingo Molnár as:

> One of the more interesting features in this cycle is the ability to attach eBPF programs (user-defined, sandboxed bytecode executed by the kernel) to kprobes. This allows user-defined instrumentation on a live kernel image that can never crash, hang or interfere with the kernel negatively.

BCC makes BPF programs easier to write, with kernel instrumentation in C
(and includes a C wrapper around LLVM), and front-ends in Python and lua.
It is suited for many tasks, including performance analysis and network
traffic control.

## Screenshot

This example traces a disk I/O kernel function, and populates an in-kernel
power-of-2 histogram of the I/O size. For efficiency, only the histogram
summary is returned to user-level.

```Shell
# ./bitehist.py
Tracing... Hit Ctrl-C to end.
^C
     kbytes          : count     distribution
       0 -> 1        : 3        |                                      |
       2 -> 3        : 0        |                                      |
       4 -> 7        : 211      |**********                            |
       8 -> 15       : 0        |                                      |
      16 -> 31       : 0        |                                      |
      32 -> 63       : 0        |                                      |
      64 -> 127      : 1        |                                      |
     128 -> 255      : 800      |**************************************|
```

The above output shows a bimodal distribution, where the largest mode of
800 I/O was between 128 and 255 Kbytes in size.

See the source: [bitehist.py](examples/tracing/bitehist.py). What this traces,
what this stores, and how the data is presented, can be entirely customized.
This shows only some of many possible capabilities.

## Installing

See [INSTALL.md](INSTALL.md) for installation steps on your platform.

## FAQ

See [FAQ.txt](FAQ.txt) for the most common troubleshoot questions.

## Reference guide

See [docs/reference_guide.md](docs/reference_guide.md) for the reference guide to the bcc and bcc/BPF APIs.

## Contents

Some of these are single files that contain both C and Python, others have a
pair of .c and .py files, and some are directories of files.

### Tracing

#### Examples

- examples/tracing/[bitehist.py](examples/tracing/bitehist.py): Block I/O size histogram. [Examples](examples/tracing/bitehist_example.txt).
- examples/tracing/[disksnoop.py](examples/tracing/disksnoop.py): Trace block device I/O latency. [Examples](examples/tracing/disksnoop_example.txt).
- examples/[hello_world.py](examples/hello_world.py): Prints "Hello, World!" for new processes.
- examples/tracing/[mysqld_query.py](examples/tracing/mysqld_query.py): Trace MySQL server queries using USDT probes. [Examples](examples/tracing/mysqld_query_example.txt).
- examples/tracing/[nodejs_http_server.py](examples/tracing/nodejs_http_server.py): Trace Node.js HTTP server requests using USDT probes. [Examples](examples/tracing/nodejs_http_server_example.txt).
- examples/tracing/[stacksnoop](examples/tracing/stacksnoop.py): Trace a kernel function and print all kernel stack traces. [Examples](examples/tracing/stacksnoop_example.txt).
- tools/[statsnoop](tools/statsnoop.py): Trace stat() syscalls. [Examples](tools/statsnoop_example.txt).
- examples/tracing/[task_switch.py](examples/tracing/task_switch.py): Count task switches with from and to PIDs.
- examples/tracing/[tcpv4connect.py](examples/tracing/tcpv4connect.py): Trace TCP IPv4 active connections. [Examples](examples/tracing/tcpv4connect_example.txt).
- examples/tracing/[trace_fields.py](examples/tracing/trace_fields.py): Simple example of printing fields from traced events.
- examples/tracing/[undump.py](examples/tracing/undump.py): Dump UNIX socket packets. [Examples](examples/tracing/undump_example.txt)
- examples/tracing/[urandomread.py](examples/tracing/urandomread.py): A kernel tracepoint example, which traces random:urandom_read. [Examples](examples/tracing/urandomread_example.txt).
- examples/tracing/[vfsreadlat.py](examples/tracing/vfsreadlat.py) examples/tracing/[vfsreadlat.c](examples/tracing/vfsreadlat.c): VFS read latency distribution. [Examples](examples/tracing/vfsreadlat_example.txt).
- examples/tracing/[kvm_hypercall.py](examples/tracing/kvm_hypercall.py): Conditional static kernel tracepoints for KVM entry, exit and hypercall [Examples](examples/tracing/kvm_hypercall.txt).

#### Tools
[](images/bcc_tracing_tools_2019.png)


- tools/[argdist](tools/argdist.py): Display function parameter values as a histogram or frequency count. [Examples](tools/argdist_example.txt).
- tools/[bashreadline](tools/bashreadline.py): Print entered bash commands system wide. [Examples](tools/bashreadline_example.txt).
- tools/[bpflist](tools/bpflist.py): Display processes with active BPF programs and maps. [Examples](tools/bpflist_example.txt).
- tools/[capable](tools/capable.py): Trace security capability checks. [Examples](tools/capable_example.txt).
- tools/[compactsnoop](tools/compactsnoop.py): Trace compact zone events with PID and latency. [Examples](tools/compactsnoop_example.txt).
- tools/[criticalstat](tools/criticalstat.py): Trace and report long atomic critical sections in the kernel. [Examples](tools/criticalstat_example.txt)
- tools/[deadlock](tools/deadlock.py): Detect potential deadlocks on a running process. [Examples](tools/deadlock_example.txt).
- tools/[drsnoop](tools/drsnoop.py): Trace direct reclaim events with PID and latency. [Examples](tools/drsnoop_example.txt).
- tools/[funccount](tools/funccount.py): Count kernel function calls. [Examples](tools/funccount_example.txt).
- tools/[inject](tools/inject.py): Targeted error injection with call chain and predicates [Examples](tools/inject_example.txt).
- tools/[klockstat](tools/klockstat.py): Traces kernel mutex lock events and display locks statistics. [Examples](tools/klockstat_example.txt).
- tools/[opensnoop](tools/opensnoop.py): Trace open() syscalls. [Examples](tools/opensnoop_example.txt).
- tools/[readahead](tools/readahead.py): Show performance of read-ahead cache [Examples](tools/readahead_example.txt).
- tools/[reset-trace](tools/reset-trace.sh): Reset the state of tracing. Maintenance tool only. [Examples](tools/reset-trace_example.txt).
- tools/[stackcount](tools/stackcount.py): Count kernel function calls and their stack traces. [Examples](tools/stackcount_example.txt).
- tools/[syncsnoop](tools/syncsnoop.py): Trace sync() syscall. [Examples](tools/syncsnoop_example.txt).
- tools/[threadsnoop](tools/threadsnoop.py): List new thread creation. [Examples](tools/threadsnoop_example.txt).
- tools/[tplist](tools/tplist.py): Display kernel tracepoints or USDT probes and their formats. [Examples](tools/tplist_example.txt).
- tools/[trace](tools/trace.py): Trace arbitrary functions, with filters. [Examples](tools/trace_example.txt).
- tools/[ttysnoop](tools/ttysnoop.py): Watch live output from a tty or pts device. [Examples](tools/ttysnoop_example.txt).
- tools/[ucalls](tools/lib/ucalls.py): Summarize method calls or Linux syscalls in high-level languages. [Examples](tools/lib/ucalls_example.txt).
- tools/[uflow](tools/lib/uflow.py): Print a method flow graph in high-level languages. [Examples](tools/lib/uflow_example.txt).
- tools/[ugc](tools/lib/ugc.py): Trace garbage collection events in high-level languages. [Examples](tools/lib/ugc_example.txt).
- tools/[uobjnew](tools/lib/uobjnew.py): Summarize object allocation events by object type and number of bytes allocated. [Examples](tools/lib/uobjnew_example.txt).
- tools/[ustat](tools/lib/ustat.py): Collect events such as GCs, thread creations, object allocations, exceptions and more in high-level languages. [Examples](tools/lib/ustat_example.txt).
- tools/[uthreads](tools/lib/uthreads.py): Trace thread creation events in Java and raw pthreads. [Examples](tools/lib/uthreads_example.txt).

##### Memory and Process Tools

- tools/[execsnoop](tools/execsnoop.py): Trace new processes via exec() syscalls. [Examples](tools/execsnoop_example.txt).
- tools/[exitsnoop](tools/exitsnoop.py): Trace process termination (exit and fatal signals). [Examples](tools/exitsnoop_example.txt).
- tools/[killsnoop](tools/killsnoop.py): Trace signals issued by the kill() syscall. [Examples](tools/killsnoop_example.txt).
- tools/[kvmexit](tools/kvmexit.py): Display the exit_reason and its statistics of each vm exit. [Examples](tools/kvmexit_example.txt).
- tools/[memleak](tools/memleak.py): Display outstanding memory allocations to find memory leaks. [Examples](tools/memleak_example.txt).
- tools/[numasched](tools/numasched.py): Track the migration of processes between NUMAs. [Examples](tools/numasched_example.txt).
- tools/[oomkill](tools/oomkill.py): Trace the out-of-memory (OOM) killer. [Examples](tools/oomkill_example.txt).
- tools/[pidpersec](tools/pidpersec.py): Count new processes (via fork). [Examples](tools/pidpersec_example.txt).
- tools/[rdmaucma](tools/rdmaucma.py): Trace RDMA Userspace Connection Manager Access events. [Examples](tools/rdmaucma_example.txt).
- tools/[shmsnoop](tools/shmsnoop.py): Trace System V shared memory syscalls. [Examples](tools/shmsnoop_example.txt).
- tools/[slabratetop](tools/slabratetop.py): Kernel SLAB/SLUB memory cache allocation rate top. [Examples](tools/slabratetop_example.txt).

##### Performance and Time Tools

- tools/[dbslower](tools/dbslower.py): Trace MySQL/PostgreSQL queries slower than a threshold. [Examples](tools/dbslower_example.txt).
- tools/[dbstat](tools/dbstat.py): Summarize MySQL/PostgreSQL query latency as a histogram. [Examples](tools/dbstat_example.txt).
- tools/[funcinterval](tools/funcinterval.py): Time interval between the same function as a histogram. [Examples](tools/funcinterval_example.txt).
- tools/[funclatency](tools/funclatency.py): Time functions and show their latency distribution. [Examples](tools/funclatency_example.txt).
- tools/[funcslower](tools/funcslower.py): Trace slow kernel or user function calls. [Examples](tools/funcslower_example.txt).
- tools/[hardirqs](tools/hardirqs.py):  Measure hard IRQ (hard interrupt) event time. [Examples](tools/hardirqs_example.txt).
- tools/[mysqld_qslower](tools/mysqld_qslower.py): Trace MySQL server queries slower than a threshold. [Examples](tools/mysqld_qslower_example.txt).
- tools/[ppchcalls](tools/ppchcalls.py): Summarize ppc hcall counts and latencies. [Examples](tools/ppchcalls_example.txt).
- tools/[softirqs](tools/softirqs.py):  Measure soft IRQ (soft interrupt) event time. [Examples](tools/softirqs_example.txt).
- tools/[softirqslower](tools/softirqslower.py): Trace slow soft IRQ (interrupt). [Examples](tools/softirqslower_example.txt).
- tools/[syscount](tools/syscount.py): Summarize syscall counts and latencies. [Examples](tools/syscount_example.txt).

##### CPU and Scheduler Tools

- tools/[cpudist](tools/cpudist.py): Summarize on- and off-CPU time per task as a histogram. [Examples](tools/cpudist_example.txt)
- tools/[cpuunclaimed](tools/cpuunclaimed.py): Sample CPU run queues and calculate unclaimed idle CPU. [Examples](tools/cpuunclaimed_example.txt)
- tools/[llcstat](tools/llcstat.py): Summarize CPU cache references and misses by process. [Examples](tools/llcstat_example.txt).
- tools/[offcputime](tools/offcputime.py): Summarize off-CPU time by kernel stack trace. [Examples](tools/offcputime_example.txt).
- tools/[offwaketime](tools/offwaketime.py): Summarize blocked time by kernel off-CPU stack and waker stack. [Examples](tools/offwaketime_example.txt).
- tools/[profile](tools/profile.py): Profile CPU usage by sampling stack traces at a timed interval. [Examples](tools/profile_example.txt).
- tools/[runqlat](tools/runqlat.py): Run queue (scheduler) latency as a histogram. [Examples](tools/runqlat_example.txt).
- tools/[runqlen](tools/runqlen.py): Run queue length as a histogram. [Examples](tools/runqlen_example.txt).
- tools/[runqslower](tools/runqslower.py): Trace long process scheduling delays. [Examples](tools/runqslower_example.txt).
- tools/[wakeuptime](tools/wakeuptime.py): Summarize sleep to wakeup time by waker kernel stack. [Examples](tools/wakeuptime_example.txt).
- tools/[wqlat](tools/wqlat.py): Summarize work waiting latency on workqueue. [Examples](tools/wqlat_example.txt).

##### Network and Sockets Tools

- tools/[gethostlatency](tools/gethostlatency.py): Show latency for getaddrinfo/gethostbyname[2] calls. [Examples](tools/gethostlatency_example.txt).
- tools/[bindsnoop](tools/bindsnoop.py): Trace IPv4 and IPv6 bind() system calls (bind()). [Examples](tools/bindsnoop_example.txt).
- tools/[netqtop](tools/netqtop.py) tools/[netqtop.c](tools/netqtop.c): Trace and display packets distribution on NIC queues. [Examples](tools/netqtop_example.txt).
- tools/[sofdsnoop](tools/sofdsnoop.py): Trace FDs passed through unix sockets. [Examples](tools/sofdsnoop_example.txt).
- tools/[solisten](tools/solisten.py): Trace TCP socket listen. [Examples](tools/solisten_example.txt).
- tools/[sslsniff](tools/sslsniff.py): Sniff OpenSSL written and read data. [Examples](tools/sslsniff_example.txt).
- tools/[tcpaccept](tools/tcpaccept.py): Trace TCP passive connections (accept()). [Examples](tools/tcpaccept_example.txt).
- tools/[tcpconnect](tools/tcpconnect.py): Trace TCP active connections (connect()). [Examples](tools/tcpconnect_example.txt).
- tools/[tcpconnlat](tools/tcpconnlat.py): Trace TCP active connection latency (connect()). [Examples](tools/tcpconnlat_example.txt).
- tools/[tcpdrop](tools/tcpdrop.py): Trace kernel-based TCP packet drops with details. [Examples](tools/tcpdrop_example.txt).
- tools/[tcplife](tools/tcplife.py): Trace TCP sessions and summarize lifespan. [Examples](tools/tcplife_example.txt).
- tools/[tcpretrans](tools/tcpretrans.py): Trace TCP retransmits and TLPs. [Examples](tools/tcpretrans_example.txt).
- tools/[tcprtt](tools/tcprtt.py): Trace TCP round trip time. [Examples](tools/tcprtt_example.txt).
- tools/[tcpstates](tools/tcpstates.py): Trace TCP session state changes with durations. [Examples](tools/tcpstates_example.txt).
- tools/[tcpsubnet](tools/tcpsubnet.py): Summarize and aggregate TCP send by subnet. [Examples](tools/tcpsubnet_example.txt).
- tools/[tcpsynbl](tools/tcpsynbl.py): Show TCP SYN backlog. [Examples](tools/tcpsynbl_example.txt).
- tools/[tcptop](tools/tcptop.py): Summarize TCP send/recv throughput by host. Top for TCP. [Examples](tools/tcptop_example.txt).
- tools/[tcptracer](tools/tcptracer.py): Trace TCP established connections (connect(), accept(), close()). [Examples](tools/tcptracer_example.txt).
- tools/[tcpcong](tools/tcpcong.py): Trace TCP socket congestion control status duration. [Examples](tools/tcpcong_example.txt).
- tools/[mptcpify](tools/mptcpify.py): Force applications to use MPTCP instead of TCP. [Examples](tools/mptcpify_example.txt).

##### Storage and Filesystems Tools

- tools/[bitesize](tools/bitesize.py): Show per process I/O size histogram. [Examples](tools/bitesize_example.txt).
- tools/[cachestat](tools/cachestat.py): Trace page cache hit/miss ratio. [Examples](tools/cachestat_example.txt).
- tools/[cachetop](tools/cachetop.py): Trace page cache hit/miss ratio by processes. [Examples](tools/cachetop_example.txt).
- tools/[dcsnoop](tools/dcsnoop.py): Trace directory entry cache (dcache) lookups. [Examples](tools/dcsnoop_example.txt).
- tools/[dcstat](tools/dcstat.py): Directory entry cache (dcache) stats. [Examples](tools/dcstat_example.txt).
- tools/[biolatency](tools/biolatency.py): Summarize block device I/O latency as a histogram. [Examples](tools/biolatency_example.txt).
- tools/[biotop](tools/biotop.py): Top for disks: Summarize block device I/O by process. [Examples](tools/biotop_example.txt).
- tools/[biopattern](tools/biopattern.py): Identify random/sequential disk access patterns. [Examples](tools/biopattern_example.txt).
- tools/[biosnoop](tools/biosnoop.py): Trace block device I/O with PID and latency. [Examples](tools/biosnoop_example.txt).
- tools/[dirtop](tools/dirtop.py): File reads and writes by directory. Top for directories. [Examples](tools/dirtop_example.txt).
- tools/[filelife](tools/filelife.py): Trace the lifespan of short-lived files. [Examples](tools/filelife_example.txt).
- tools/[filegone](tools/filegone.py): Trace why file gone (deleted or renamed). [Examples](tools/filegone_example.txt).
- tools/[fileslower](tools/fileslower.py): Trace slow synchronous file reads and writes. [Examples](tools/fileslower_example.txt).
- tools/[filetop](tools/filetop.py): File reads and writes by filename and process. Top for files. [Examples](tools/filetop_example.txt).
- tools/[mdflush](tools/mdflush.py): Trace md flush events. [Examples](tools/mdflush_example.txt).
- tools/[mountsnoop](tools/mountsnoop.py): Trace mount and umount syscalls system-wide. [Examples](tools/mountsnoop_example.txt).
- tools/[virtiostat](tools/virtiostat.py): Show VIRTIO device IO statistics. [Examples](tools/virtiostat_example.txt).

###### Filesystems Tools

- tools/[btrfsdist](tools/btrfsdist.py): Summarize btrfs operation latency distribution as a histogram. [Examples](tools/btrfsdist_example.txt).
- tools/[btrfsslower](tools/btrfsslower.py): Trace slow btrfs operations. [Examples](tools/btrfsslower_example.txt).
- tools/[ext4dist](tools/ext4dist.py): Summarize ext4 operation latency distribution as a histogram. [Examples](tools/ext4dist_example.txt).
- tools/[ext4slower](tools/ext4slower.py): Trace slow ext4 operations. [Examples](tools/ext4slower_example.txt).
- tools/[nfsslower](tools/nfsslower.py): Trace slow NFS operations. [Examples](tools/nfsslower_example.txt).
- tools/[nfsdist](tools/nfsdist.py): Summarize NFS operation latency distribution as a histogram. [Examples](tools/nfsdist_example.txt).
- tools/[vfscount](tools/vfscount.py): Count VFS calls. [Examples](tools/vfscount_example.txt).
- tools/[vfsstat](tools/vfsstat.py): Count some VFS calls, with column output. [Examples](tools/vfsstat_example.txt).
- tools/[xfsdist](tools/xfsdist.py): Summarize XFS operation latency distribution as a histogram. [Examples](tools/xfsdist_example.txt).
- tools/[xfsslower](tools/xfsslower.py): Trace slow XFS operations. [Examples](tools/xfsslower_example.txt).
- tools/[zfsdist](tools/zfsdist.py): Summarize ZFS operation latency distribution as a histogram. [Examples](tools/zfsdist_example.txt).
- tools/[zfsslower](tools/zfsslower.py): Trace slow ZFS operations. [Examples](tools/zfsslower_example.txt).

### Networking

Examples:

- examples/networking/[distributed_bridge/](examples/networking/distributed_bridge): Distributed bridge example.
- examples/networking/[http_filter/](examples/networking/http_filter): Simple HTTP filter example.
- examples/networking/[simple_tc.py](examples/networking/simple_tc.py): Simple traffic control example.
- examples/networking/[simulation.py](examples/networking/simulation.py): Simulation helper.
- examples/networking/neighbor_sharing/[tc_neighbor_sharing.py](examples/networking/neighbor_sharing/tc_neighbor_sharing.py) examples/networking/neighbor_sharing/[tc_neighbor_sharing.c](examples/networking/neighbor_sharing/tc_neighbor_sharing.c): Per-IP classification and rate limiting.
- examples/networking/[tunnel_monitor/](examples/networking/tunnel_monitor): Efficiently monitor traffic flows.
- examples/networking/vlan_learning/[vlan_learning.py](examples/networking/vlan_learning/vlan_learning.py) examples/[vlan_learning.c](examples/networking/vlan_learning/vlan_learning.c): Demux Ethernet traffic into worker veth+namespaces.

### BPF Introspection

Tools that help to introspect BPF programs.

- introspection/[bps.c](introspection/bps.c): List all BPF programs loaded into the kernel. 'ps' for BPF programs. [Examples](introspection/bps_example.txt).

## Motivation

BPF guarantees that the programs loaded into the kernel cannot crash, and
cannot run forever, but yet BPF is general purpose enough to perform many
arbitrary types of computation. Currently, it is possible to write a program in
C that will compile into a valid BPF program, yet it is vastly easier to
write a C program that will compile into invalid BPF (C is like that). The user
won't know until trying to run the program whether it was valid or not.

With a BPF-specific frontend, one should be able to write in a language and
receive feedback from the compiler on the validity as it pertains to a BPF
backend. This toolkit aims to provide a frontend that can only create valid BPF
programs while still harnessing its full flexibility.

Furthermore, current integrations with BPF have a kludgy workflow, sometimes
involving compiling directly in a linux kernel source tree. This toolchain aims
to minimize the time that a developer spends getting BPF compiled, and instead
focus on the applications that can be written and the problems that can be
solved with BPF.

The features of this toolkit include:
* End-to-end BPF workflow in a shared library
  * A modified C language for BPF backends
  * Integration with llvm-bpf backend for JIT
  * Dynamic (un)loading of JITed programs
  * Support for BPF kernel hooks: socket filters, tc classifiers,
      tc actions, and kprobes
* Bindings for Python
* Examples for socket filters, tc classifiers, and kprobes
* Self-contained tools for tracing a running system

In the future, more bindings besides python will likely be supported. Feel free
to add support for the language of your choice and send a pull request!

## Tutorials

- [docs/tutorial.md](docs/tutorial.md): Using bcc tools to solve performance, troubleshooting, and networking issues.
- [docs/tutorial_bcc_python_developer.md](docs/tutorial_bcc_python_developer.md): Developing new bcc programs using the Python interface.

### Networking

At Red Hat Summit 2015, BCC was presented as part of a [session on BPF](http://www.devnation.org/#7784f1f7513e8542e4db519e79ff5eec).
A multi-host vxlan environment is simulated and a BPF program used to monitor
one of the physical interfaces. The BPF program keeps statistics on the inner
and outer IP addresses traversing the interface, and the userspace component
turns those statistics into a graph showing the traffic distribution at
multiple granularities. See the code [here](examples/networking/tunnel_monitor).

## Contributing

Already pumped up to commit some code? Here are some resources to join the
discussions in the [IOVisor](https://www.iovisor.org/) community and see
what you want to work on.

* _Mailing List:_ https://lists.iovisor.org/mailman/listinfo/iovisor-dev
* _IRC:_ #iovisor at irc.oftc.net
* _BCC Issue Tracker:_ [Github Issues](https://github.com/iovisor/bcc/issues)
* _A guide for contributing scripts:_ [CONTRIBUTING-SCRIPTS.md](CONTRIBUTING-SCRIPTS.md)

## External links

Looking for more information on BCC and how it's being used? You can find links to other BCC content on the web in [LINKS.md](LINKS.md).

---

## File: docs/kernel_config.md

# Kernel Configuration for BPF Features

## BPF Related Kernel Configurations

| Functionalities | Kernel Configuration | Description |
|:----------------|:---------------------|:------------|
| **Basic** | CONFIG_BPF_SYSCALL | Enable the bpf() system call |
|  | CONFIG_BPF_JIT | BPF programs are normally handled by a BPF interpreter. This option allows the kernel to generate native code when a program is loaded into the kernel. This will significantly speed-up processing of BPF programs |
|  | CONFIG_HAVE_BPF_JIT | Enable BPF Just In Time compiler |
|  | CONFIG_HAVE_EBPF_JIT | Extended BPF JIT (eBPF) |
|  | CONFIG_HAVE_CBPF_JIT | Classic BPF JIT (cBPF) |
|  | CONFIG_MODULES | Enable to build loadable kernel modules |
|  | CONFIG_BPF | BPF VM interpreter |
|  | CONFIG_BPF_EVENTS | Allow the user to attach BPF programs to kprobe, uprobe, and tracepoint events |
|  | CONFIG_PERF_EVENTS | Kernel performance events and counters |
|  | CONFIG_HAVE_PERF_EVENTS | Enable perf events |
|  | CONFIG_PROFILING | Enable the extended profiling support mechanisms used by profilers |
| **BTF** | CONFIG_DEBUG_INFO_BTF | Generate deduplicated BTF type information from DWARF debug info |
| | CONFIG_PAHOLE_HAS_SPLIT_BTF | Generate BTF for each selected kernel module |
| | CONFIG_DEBUG_INFO_BTF_MODULES | Generate compact split BTF type information for kernel modules |
| **Security** | CONFIG_BPF_JIT_ALWAYS_ON | Enable BPF JIT and removes BPF interpreter to avoid speculative execution |
| | CONFIG_BPF_UNPRIV_DEFAULT_OFF | Disable unprivileged BPF by default by setting |
| **Cgroup** | CONFIG_CGROUP_BPF | Support for BPF programs attached to cgroups |
| **Network** | CONFIG_BPFILTER | BPF based packet filtering framework (BPFILTER) |
| | CONFIG_BPFILTER_UMH | This builds bpfilter kernel module with embedded user mode helper |
| | CONFIG_NET_CLS_BPF | BPF-based classifier - to classify packets based on programmable BPF (JIT'ed) filters as an alternative to ematches |
| | CONFIG_NET_ACT_BPF | Execute BPF code on packets. The BPF code will decide if the packet should be dropped or not |
| | CONFIG_BPF_STREAM_PARSER | Enable this to allow a TCP stream parser to be used with BPF_MAP_TYPE_SOCKMAP |
| | CONFIG_LWTUNNEL_BPF | Allow to run BPF programs as a nexthop action following a route lookup for incoming and outgoing packets |
| | CONFIG_NETFILTER_XT_MATCH_BPF | BPF matching applies a linux socket filter to each packet and accepts those for which the filter returns non-zero |
| | CONFIG_IPV6_SEG6_BPF | To support  BPF seg6local hook. bpf: Add IPv6 Segment Routing helpersy. [Reference](https://github.com/torvalds/linux/commit/fe94cc290f535709d3c5ebd1e472dfd0aec7ee7) |
| **kprobes** | CONFIG_KPROBE_EVENTS | This allows the user to add tracing events (similar to tracepoints) on the fly via the ftrace interface |
|  | CONFIG_KPROBES | Enable kprobes-based dynamic events |
|  | CONFIG_HAVE_KPROBES | Check if krpobes enabled |
|  | CONFIG_HAVE_REGS_AND_STACK_ACCESS_API | This symbol should be selected by an architecture if it supports the API needed to access registers and stack entries from pt_regs. For example the kprobes-based event tracer needs this API. |
|  | CONFIG_KPROBES_ON_FTRACE | Have kprobes on function tracer if arch supports full passing of pt_regs to function tracing |
| **kprobe multi** | CONFIG_FPROBE | Enable fprobe to attach the probe on multiple functions at once |
| **kprobe override** | CONFIG_BPF_KPROBE_OVERRIDE | Enable BPF programs to override a kprobed function |
| **uprobes** | CONFIG_UPROBE_EVENTS | Enable uprobes-based dynamic events |
|  | CONFIG_ARCH_SUPPORTS_UPROBES | Arch specific uprobes support |
|  | CONFIG_UPROBES | Uprobes is the user-space counterpart to kprobes: they enable instrumentation applications (such as 'perf probe') to establish unintrusive probes in user-space binaries and libraries, by executing handler functions when the probes are hit by user-space applications. |
|  | CONFIG_MMU | MMU-based virtualised addressing space support by paged memory management |
| **Tracepoints** | CONFIG_TRACEPOINTS | Enable inserting tracepoints in the kernel and connect to proble functions |
|  | CONFIG_HAVE_SYSCALL_TRACEPOINTS | Enable syscall enter/exit tracing |
| **Raw Tracepoints** | Same as Tracepoints | |
| **LSM** | CONFIG_BPF_LSM | Enable instrumentation of the security hooks with BPF programs for implementing dynamic MAC and Audit Policies |
| **LIRC** | CONFIG_BPF_LIRC_MODE2 | Allow attaching BPF programs to a lirc device |

---

## File: docs/kernel-versions.md

# BPF Features by Linux Kernel Version

## eBPF support

Kernel version | Commit
---------------|-------
3.15 | [`bd4cf0ed331a`](https://github.com/torvalds/linux/commit/bd4cf0ed331a275e9bf5a49e6d0fd55dffc551b8)

## JIT compiling

The list of supported architectures for your kernel can be retrieved with:

    git grep HAVE_EBPF_JIT arch/

Feature / Architecture | Kernel version | Commit
-----------------------|----------------|-------
x86\_64                            | 3.16 | [`622582786c9e`](https://github.com/torvalds/linux/commit/622582786c9e041d0bd52bde201787adeab249f8)
ARM64                              | 3.18 | [`e54bcde3d69d`](https://github.com/torvalds/linux/commit/e54bcde3d69d40023ae77727213d14f920eb264a)
s390                               | 4.1  | [`054623105728`](https://github.com/torvalds/linux/commit/054623105728b06852f077299e2bf1bf3d5f2b0b)
Constant blinding for JIT machines | 4.7  | [`4f3446bb809f`](https://github.com/torvalds/linux/commit/4f3446bb809f20ad56cadf712e6006815ae7a8f9)
PowerPC64                          | 4.8  | [`156d0e290e96`](https://github.com/torvalds/linux/commit/156d0e290e969caba25f1851c52417c14d141b24)
Constant blinding - PowerPC64      | 4.9  | [`b7b7013cac55`](https://github.com/torvalds/linux/commit/b7b7013cac55d794940bd9cb7b7c55c9dececac4)
Sparc64                            | 4.12 | [`7a12b5031c6b`](https://github.com/torvalds/linux/commit/7a12b5031c6b947cc13918237ae652b536243b76)
MIPS                               | 4.13 | [`f381bf6d82f0`](https://github.com/torvalds/linux/commit/f381bf6d82f032b7410185b35d000ea370ac706b)
ARM32                              | 4.14 | [`39c13c204bb1`](https://github.com/torvalds/linux/commit/39c13c204bb1150d401e27d41a9d8b332be47c49)
x86\_32                            | 4.18 | [`03f5781be2c7`](https://github.com/torvalds/linux/commit/03f5781be2c7b7e728d724ac70ba10799cc710d7)
RISC-V RV64G                       | 5.1  | [`2353ecc6f91f`](https://github.com/torvalds/linux/commit/2353ecc6f91fd15b893fa01bf85a1c7a823ee4f2)
RISC-V RV32G                       | 5.7  | [`5f316b65e99f`](https://github.com/torvalds/linux/commit/5f316b65e99f109942c556dc8790abd4c75bcb34)
PowerPC32                          | 5.13 | [`51c66ad849a7`](https://github.com/torvalds/linux/commit/51c66ad849a703d9bbfd7704c941827aed0fd9fd)
LoongArch                          | 6.1  | [`5dc615520c4d`](https://github.com/torvalds/linux/commit/5dc615520c4dfb358245680f1904bad61116648e)

## Main features

Several (but not all) of these _main features_ translate to an eBPF program type.
The list of such program types supported in your kernel can be found in file
[`include/uapi/linux/bpf.h`](https://github.com/torvalds/linux/blob/master/include/uapi/linux/bpf.h):

    git grep -W 'bpf_prog_type {' include/uapi/linux/bpf.h

Feature | Kernel version | Commit
--------|----------------|-------
`AF_PACKET` (libpcap/tcpdump, `cls_bpf` classifier, netfilter's `xt_bpf`, team driver's load-balancing mode…) | 3.15 | [`bd4cf0ed331a`](https://github.com/torvalds/linux/commit/bd4cf0ed331a275e9bf5a49e6d0fd55dffc551b8)
Kernel helpers | 3.15 | [`bd4cf0ed331a`](https://github.com/torvalds/linux/commit/bd4cf0ed331a275e9bf5a49e6d0fd55dffc551b8)
`bpf()` syscall | 3.18 | [`99c55f7d47c0`](https://github.com/torvalds/linux/commit/99c55f7d47c0dc6fc64729f37bf435abf43f4c60)
Maps (_a.k.a._ Tables; details below) | 3.18 | [`99c55f7d47c0`](https://github.com/torvalds/linux/commit/99c55f7d47c0dc6fc64729f37bf435abf43f4c60)
BPF attached to sockets | 3.19 | [`89aa075832b0`](https://github.com/torvalds/linux/commit/89aa075832b0da4402acebd698d0411dcc82d03e)
BPF attached to `kprobes` | 4.1 | [`2541517c32be`](https://github.com/torvalds/linux/commit/2541517c32be2531e0da59dfd7efc1ce844644f5)
`cls_bpf` / `act_bpf` for `tc` | 4.1 | [`e2e9b6541dd4`](https://github.com/torvalds/linux/commit/e2e9b6541dd4b31848079da80fe2253daaafb549)
Tail calls | 4.2 | [`04fd61ab36ec`](https://github.com/torvalds/linux/commit/04fd61ab36ec065e194ab5e74ae34a5240d992bb)
Non-root programs on sockets | 4.4 | [`1be7f75d1668`](https://github.com/torvalds/linux/commit/1be7f75d1668d6296b80bf35dcf6762393530afc)
Persistent maps and programs (virtual FS) | 4.4 | [`b2197755b263`](https://github.com/torvalds/linux/commit/b2197755b2633e164a439682fb05a9b5ea48f706)
`tc`'s `direct-action` (`da`) mode | 4.4 | [`045efa82ff56`](https://github.com/torvalds/linux/commit/045efa82ff563cd4e656ca1c2e354fa5bf6bbda4)
`tc`'s `clsact` qdisc | 4.5 | [`1f211a1b929c`](https://github.com/torvalds/linux/commit/1f211a1b929c804100e138c5d3d656992cfd5622)
BPF attached to tracepoints | 4.7 | [`98b5c2c65c29`](https://github.com/torvalds/linux/commit/98b5c2c65c2951772a8fc661f50d675e450e8bce)
Direct packet access | 4.7 | [`969bf05eb3ce`](https://github.com/torvalds/linux/commit/969bf05eb3cedd5a8d4b7c346a85c2ede87a6d6d)
XDP (see below) | 4.8 | [`6a773a15a1e8`](https://github.com/torvalds/linux/commit/6a773a15a1e8874e5eccd2f29190c31085912c95)
BPF attached to perf events | 4.9 | [`0515e5999a46`](https://github.com/torvalds/linux/commit/0515e5999a466dfe6e1924f460da599bb6821487)
Hardware offload for `tc`'s `cls_bpf` | 4.9 | [`332ae8e2f6ec`](https://github.com/torvalds/linux/commit/332ae8e2f6ecda5e50c5c62ed62894963e3a83f5)
Verifier exposure and internal hooks | 4.9 | [`13a27dfc6697`](https://github.com/torvalds/linux/commit/13a27dfc669724564aafa2699976ee756029fed2)
BPF attached to cgroups for socket filtering | 4.10 | [`0e33661de493`](https://github.com/torvalds/linux/commit/0e33661de493db325435d565a4a722120ae4cbf3)
Lightweight tunnel encapsulation | 4.10 | [`3a0af8fd61f9`](https://github.com/torvalds/linux/commit/3a0af8fd61f90920f6fa04e4f1e9a6a73c1b4fd2)
**e**BPF support for `xt_bpf` module (iptables) | 4.10 | [`2c16d6033264`](https://github.com/torvalds/linux/commit/2c16d60332643e90d4fa244f4a706c454b8c7569)
BPF program tag | 4.10 | [`7bd509e311f4`](https://github.com/torvalds/linux/commit/7bd509e311f408f7a5132fcdde2069af65fa05ae)
Tracepoints to debug BPF | 4.11 (removed in 4.18) | [`a67edbf4fb6d`](https://github.com/torvalds/linux/commit/a67edbf4fb6deadcfe57a04a134abed4a5ba3bb5) [`4d220ed0f814`](https://github.com/torvalds/linux/commit/4d220ed0f8140c478ab7b0a14d96821da639b646)
Testing / benchmarking BPF programs | 4.12 | [`1cf1cae963c2`](https://github.com/torvalds/linux/commit/1cf1cae963c2e6032aebe1637e995bc2f5d330f4)
BPF programs and maps IDs | 4.13 | [`dc4bb0e23561`](https://github.com/torvalds/linux/commit/dc4bb0e2356149aee4cdae061936f3bbdd45595c)
BPF support for `sock_ops` | 4.13 | [`40304b2a1567`](https://github.com/torvalds/linux/commit/40304b2a1567fecc321f640ee4239556dd0f3ee0)
BPF support for skbs on sockets | 4.14 | [`b005fd189cec`](https://github.com/torvalds/linux/commit/b005fd189cec9407b700599e1e80e0552446ee79)
bpftool utility in kernel sources | 4.15 | [`71bb428fe2c1`](https://github.com/torvalds/linux/commit/71bb428fe2c19512ac671d5ee16ef3e73e1b49a8)
BPF attached to cgroups as device controller | 4.15 | [`ebc614f68736`](https://github.com/torvalds/linux/commit/ebc614f687369f9df99828572b1d85a7c2de3d92)
bpf2bpf function calls | 4.16 |  [`cc8b0b92a169`](https://github.com/torvalds/linux/commit/cc8b0b92a1699bc32f7fec71daa2bfc90de43a4d)
BPF used for monitoring socket RX/TX data | 4.17 | [`4f738adba30a`](https://github.com/torvalds/linux/commit/4f738adba30a7cfc006f605707e7aee847ffefa0)
BPF attached to raw tracepoints | 4.17 | [`c4f6699dfcb8`](https://github.com/torvalds/linux/commit/c4f6699dfcb8558d138fe838f741b2c10f416cf9)
BPF attached to `bind()` system call | 4.17 | [`4fbac77d2d09`](https://github.com/torvalds/linux/commit/4fbac77d2d092b475dda9eea66da674369665427) [`aac3fc320d94`](https://github.com/torvalds/linux/commit/aac3fc320d9404f2665a8b1249dc3170d5fa3caf)
BPF attached to `connect()` system call | 4.17 | [`d74bad4e74ee`](https://github.com/torvalds/linux/commit/d74bad4e74ee373787a9ae24197c17b7cdc428d5)
BPF Type Format (BTF) | 4.18 | [`69b693f0aefa`](https://github.com/torvalds/linux/commit/69b693f0aefa0ed521e8bd02260523b5ae446ad7)
AF_XDP | 4.18 |  [`fbfc504a24f5`](https://github.com/torvalds/linux/commit/fbfc504a24f53f7ebe128ab55cb5dba634f4ece8)
bpfilter | 4.18 |  [`d2ba09c17a06`](https://github.com/torvalds/linux/commit/d2ba09c17a0647f899d6c20a11bab9e6d3382f07)
End.BPF action for seg6local LWT | 4.18 |  [`004d4b274e2a`](https://github.com/torvalds/linux/commit/004d4b274e2a1a895a0e5dc66158b90a7d463d44)
BPF attached to LIRC devices | 4.18 |  [`f4364dcfc86d`](https://github.com/torvalds/linux/commit/f4364dcfc86df7c1ca47b256eaf6b6d0cdd0d936)
Pass map values to map helpers | 4.18 | [`d71962f3e627`](https://github.com/torvalds/linux/commit/d71962f3e627b5941804036755c844fabfb65ff5)
BPF socket reuseport | 4.19 | [`2dbb9b9e6df6`](https://github.com/torvalds/linux/commit/2dbb9b9e6df67d444fbe425c7f6014858d337adf)
BPF flow dissector | 4.20 | [`d58e468b1112`](https://github.com/torvalds/linux/commit/d58e468b1112dcd1d5193c0a89ff9f98b5a3e8b9)
BPF 1M insn limit | 5.2 | [`c04c0d2b968a`](https://github.com/torvalds/linux/commit/c04c0d2b968ac45d6ef020316808ef6c82325a82)
BPF cgroup sysctl | 5.2 | [`7b146cebe30c`](https://github.com/torvalds/linux/commit/7b146cebe30cb481b0f70d85779da938da818637)
BPF raw tracepoint writable | 5.2 | [`9df1c28bb752`](https://github.com/torvalds/linux/commit/9df1c28bb75217b244257152ab7d788bb2a386d0)
BPF bounded loop | 5.3 | [`2589726d12a1`](https://github.com/torvalds/linux/commit/2589726d12a1b12eaaa93c7f1ea64287e383c7a5)
BPF trampoline | 5.5 | [`fec56f5890d9`](https://github.com/torvalds/linux/commit/fec56f5890d93fc2ed74166c397dc186b1c25951)
BPF LSM hook | 5.7 | [`fc611f47f218`](https://github.com/torvalds/linux/commit/fc611f47f2188ade2b48ff6902d5cce8baac0c58) [`641cd7b06c91`](https://github.com/torvalds/linux/commit/641cd7b06c911c5935c34f24850ea18690649917)
BPF iterator | 5.8 | [`180139dca8b3`](https://github.com/torvalds/linux/commit/180139dca8b38c858027b8360ee10064fdb2fbf7)
BPF socket lookup hook | 5.9 | [`e9ddbb7707ff`](https://github.com/torvalds/linux/commit/e9ddbb7707ff5891616240026062b8c1e29864ca)
Sleepable BPF programs | 5.10 | [`1e6c62a88215`](https://github.com/torvalds/linux/commit/1e6c62a8821557720a9b2ea9617359b264f2f67c)
Mixing bpf2bpf function calls and tailcalls (x86\_64) | 5.10 | [`e411901c0b77`](https://github.com/torvalds/linux/commit/e411901c0b775a3ae7f3e2505f8d2d90ac696178)
Mixing bpf2bpf function calls and tailcalls (arm64) | 6.0 | [`d4609a5d8c70`](https://github.com/torvalds/linux/commit/d4609a5d8c70d21b4a3f801cf896a3c16c613fe1)
Mixing bpf2bpf function calls and tailcalls (s390) | 6.3 | [`dd691e847d28`](https://github.com/torvalds/linux/commit/dd691e847d28ac5f8b8e3005be44fd0e46722809)
Mixing bpf2bpf function calls and tailcalls (loongarch) | 6.4 | [`bb035ef0cc91`](https://github.com/torvalds/linux/commit/bb035ef0cc91e115faa80187ac8886a7f1914d06)


### Program types

Program type | Kernel version | Commit | Enum
-------------|----------------|--------|-----
Socket filter                  | 3.19 | [`ddd872bc3098`](https://github.com/torvalds/linux/commit/ddd872bc3098f9d9abe1680a6b2013e59e3337f7) | BPF_PROG_TYPE_SOCKET_FILTER
Kprobe                         | 4.1  | [`2541517c32be`](https://github.com/torvalds/linux/commit/2541517c32be2531e0da59dfd7efc1ce844644f5) | BPF_PROG_TYPE_KPROBE
traffic control (TC)           | 4.1  | [`96be4325f443`](https://github.com/torvalds/linux/commit/96be4325f443dbbfeb37d2a157675ac0736531a1) | BPF_PROG_TYPE_SCHED_CLS
traffic control (TC)           | 4.1  | [`94caee8c312d`](https://github.com/torvalds/linux/commit/94caee8c312d96522bcdae88791aaa9ebcd5f22c) | BPF_PROG_TYPE_SCHED_ACT
Tracepoint                     | 4.7  | [`98b5c2c65c29`](https://github.com/torvalds/linux/commit/98b5c2c65c2951772a8fc661f50d675e450e8bce) | BPF_PROG_TYPE_TRACEPOINT
XDP                            | 4.8  | [`6a773a15a1e8`](https://github.com/torvalds/linux/commit/6a773a15a1e8874e5eccd2f29190c31085912c95) | BPF_PROG_TYPE_XDP
Perf event                     | 4.9  | [`0515e5999a46`](https://github.com/torvalds/linux/commit/0515e5999a466dfe6e1924f460da599bb6821487) | BPF_PROG_TYPE_PERF_EVENT
cgroup socket filtering        | 4.10 | [`0e33661de493`](https://github.com/torvalds/linux/commit/0e33661de493db325435d565a4a722120ae4cbf3) | BPF_PROG_TYPE_CGROUP_SKB
cgroup sock modification       | 4.10 | [`610236587600`](https://github.com/torvalds/linux/commit/61023658760032e97869b07d54be9681d2529e77) | BPF_PROG_TYPE_CGROUP_SOCK
lightweight tunnel (IN)        | 4.10 | [`3a0af8fd61f9`](https://github.com/torvalds/linux/commit/3a0af8fd61f90920f6fa04e4f1e9a6a73c1b4fd2) | BPF_PROG_TYPE_LWT_IN
lightweight tunnel (OUT)       | 4.10 | [`3a0af8fd61f9`](https://github.com/torvalds/linux/commit/3a0af8fd61f90920f6fa04e4f1e9a6a73c1b4fd2) | BPF_PROG_TYPE_LWT_OUT
lightweight tunnel (XMIT)      | 4.10 | [`3a0af8fd61f9`](https://github.com/torvalds/linux/commit/3a0af8fd61f90920f6fa04e4f1e9a6a73c1b4fd2) | BPF_PROG_TYPE_LWT_XMIT
cgroup sock ops (per conn)     | 4.13 | [`40304b2a1567`](https://github.com/torvalds/linux/commit/40304b2a1567fecc321f640ee4239556dd0f3ee0) | BPF_PROG_TYPE_SOCK_OPS
stream parser / stream verdict | 4.14 | [`b005fd189cec`](https://github.com/torvalds/linux/commit/b005fd189cec9407b700599e1e80e0552446ee79) | BPF_PROG_TYPE_SK_SKB
cgroup device manager          | 4.15 | [`ebc614f68736`](https://github.com/torvalds/linux/commit/ebc614f687369f9df99828572b1d85a7c2de3d92) | BPF_PROG_TYPE_CGROUP_DEVICE
socket msg verdict             | 4.17 | [`4f738adba30a`](https://github.com/torvalds/linux/commit/4f738adba30a7cfc006f605707e7aee847ffefa0) | BPF_PROG_TYPE_SK_MSG
Raw tracepoint                 | 4.17 | [`c4f6699dfcb8`](https://github.com/torvalds/linux/commit/c4f6699dfcb8558d138fe838f741b2c10f416cf9) | BPF_PROG_TYPE_RAW_TRACEPOINT
socket binding                 | 4.17 | [`4fbac77d2d09`](https://github.com/torvalds/linux/commit/4fbac77d2d092b475dda9eea66da674369665427) | BPF_PROG_TYPE_CGROUP_SOCK_ADDR
LWT seg6local                  | 4.18 | [`004d4b274e2a`](https://github.com/torvalds/linux/commit/004d4b274e2a1a895a0e5dc66158b90a7d463d44) | BPF_PROG_TYPE_LWT_SEG6LOCAL
lirc devices                   | 4.18 | [`f4364dcfc86d`](https://github.com/torvalds/linux/commit/f4364dcfc86df7c1ca47b256eaf6b6d0cdd0d936) | BPF_PROG_TYPE_LIRC_MODE2
lookup SO_REUSEPORT socket     | 4.19 | [`2dbb9b9e6df6`](https://github.com/torvalds/linux/commit/2dbb9b9e6df67d444fbe425c7f6014858d337adf) | BPF_PROG_TYPE_SK_REUSEPORT
flow dissector                 | 4.20 | [`d58e468b1112`](https://github.com/torvalds/linux/commit/d58e468b1112dcd1d5193c0a89ff9f98b5a3e8b9) | BPF_PROG_TYPE_FLOW_DISSECTOR
cgroup sysctl                  | 5.2  | [`7b146cebe30c`](https://github.com/torvalds/linux/commit/7b146cebe30cb481b0f70d85779da938da818637) | BPF_PROG_TYPE_CGROUP_SYSCTL
writable raw tracepoints       | 5.2  | [`9df1c28bb752`](https://github.com/torvalds/linux/commit/9df1c28bb75217b244257152ab7d788bb2a386d0) | BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE
cgroup getsockopt/setsockopt   | 5.3  | [`0d01da6afc54`](https://github.com/torvalds/linux/commit/0d01da6afc5402f60325c5da31b22f7d56689b49) | BPF_PROG_TYPE_CGROUP_SOCKOPT
Tracing (BTF/BPF trampoline)   | 5.5  | [`f1b9509c2fb0`](https://github.com/torvalds/linux/commit/f1b9509c2fb0ef4db8d22dac9aef8e856a5d81f6) | BPF_PROG_TYPE_TRACING
struct ops                     | 5.6  | [`27ae7997a661`](https://github.com/torvalds/linux/commit/27ae7997a66174cb8afd6a75b3989f5e0c1b9e5a) | BPF_PROG_TYPE_STRUCT_OPS
extensions                     | 5.6  | [`be8704ff07d2`](https://github.com/torvalds/linux/commit/be8704ff07d2374bcc5c675526f95e70c6459683) | BPF_PROG_TYPE_EXT
LSM                            | 5.7  | [`fc611f47f218`](https://github.com/torvalds/linux/commit/fc611f47f2188ade2b48ff6902d5cce8baac0c58) | BPF_PROG_TYPE_LSM
lookup listening socket        | 5.9  | [`e9ddbb7707ff`](https://github.com/torvalds/linux/commit/e9ddbb7707ff5891616240026062b8c1e29864ca) | BPF_PROG_TYPE_SK_LOOKUP
Allow executing syscalls       | 5.15 | [`79a7f8bdb159`](https://github.com/torvalds/linux/commit/79a7f8bdb159d9914b58740f3d31d602a6e4aca8) | BPF_PROG_TYPE_SYSCALL

## Maps (_a.k.a._ Tables, in BCC lingo)

### Map types

The list of map types supported in your kernel can be found in file
[`include/uapi/linux/bpf.h`](https://github.com/torvalds/linux/blob/master/include/uapi/linux/bpf.h):

    git grep -W 'bpf_map_type {' include/uapi/linux/bpf.h

 Map type | Kernel version | Commit | Enum
----------|----------------|--------|------
Hash                            | 3.19 | [`0f8e4bd8a1fc`](https://github.com/torvalds/linux/commit/0f8e4bd8a1fc8c4185f1630061d0a1f2d197a475) | BPF_MAP_TYPE_HASH
Array                           | 3.19 | [`28fbcfa08d8e`](https://github.com/torvalds/linux/commit/28fbcfa08d8ed7c5a50d41a0433aad222835e8e3) | BPF_MAP_TYPE_ARRAY
Prog array                      | 4.2  | [`04fd61ab36ec`](https://github.com/torvalds/linux/commit/04fd61ab36ec065e194ab5e74ae34a5240d992bb) | BPF_MAP_TYPE_PROG_ARRAY
Perf events                     | 4.3  | [`ea317b267e9d`](https://github.com/torvalds/linux/commit/ea317b267e9d03a8241893aa176fba7661d07579) | BPF_MAP_TYPE_PERF_EVENT_ARRAY
Per-CPU hash                    | 4.6  | [`824bd0ce6c7c`](https://github.com/torvalds/linux/commit/824bd0ce6c7c43a9e1e210abf124958e54d88342) | BPF_MAP_TYPE_PERCPU_HASH
Per-CPU array                   | 4.6  | [`a10423b87a7e`](https://github.com/torvalds/linux/commit/a10423b87a7eae75da79ce80a8d9475047a674ee) | BPF_MAP_TYPE_PERCPU_ARRAY
Stack trace                     | 4.6  | [`d5a3b1f69186`](https://github.com/torvalds/linux/commit/d5a3b1f691865be576c2bffa708549b8cdccda19) | BPF_MAP_TYPE_STACK_TRACE
cgroup array                    | 4.8  | [`4ed8ec521ed5`](https://github.com/torvalds/linux/commit/4ed8ec521ed57c4e207ad464ca0388776de74d4b) | BPF_MAP_TYPE_CGROUP_ARRAY
LRU hash                        | 4.10 | [`29ba732acbee`](https://github.com/torvalds/linux/commit/29ba732acbeece1e34c68483d1ec1f3720fa1bb3) [`3a08c2fd7634`](https://github.com/torvalds/linux/commit/3a08c2fd763450a927d1130de078d6f9e74944fb) | BPF_MAP_TYPE_LRU_HASH
LRU per-CPU hash                | 4.10 | [`8f8449384ec3`](https://github.com/torvalds/linux/commit/8f8449384ec364ba2a654f11f94e754e4ff719e0) [`961578b63474`](https://github.com/torvalds/linux/commit/961578b63474d13ad0e2f615fcc2901c5197dda6) | BPF_MAP_TYPE_LRU_PERCPU_HASH
LPM trie (longest-prefix match) | 4.11 | [`b95a5c4db09b`](https://github.com/torvalds/linux/commit/b95a5c4db09bc7c253636cb84dc9b12c577fd5a0) | BPF_MAP_TYPE_LPM_TRIE
Array of maps                   | 4.12 | [`56f668dfe00d`](https://github.com/torvalds/linux/commit/56f668dfe00dcf086734f1c42ea999398fad6572) | BPF_MAP_TYPE_ARRAY_OF_MAPS
Hash of maps                    | 4.12 | [`bcc6b1b7ebf8`](https://github.com/torvalds/linux/commit/bcc6b1b7ebf857a9fe56202e2be3361131588c15) | BPF_MAP_TYPE_HASH_OF_MAPS
Netdevice references (array)    | 4.14 | [`546ac1ffb70d`](https://github.com/torvalds/linux/commit/546ac1ffb70d25b56c1126940e5ec639c4dd7413) | BPF_MAP_TYPE_DEVMAP
Socket references (array)       | 4.14 | [`174a79ff9515`](https://github.com/torvalds/linux/commit/174a79ff9515f400b9a6115643dafd62a635b7e6) | BPF_MAP_TYPE_SOCKMAP
CPU references                  | 4.15 | [`6710e1126934`](https://github.com/torvalds/linux/commit/6710e1126934d8b4372b4d2f9ae1646cd3f151bf) | BPF_MAP_TYPE_CPUMAP
AF_XDP socket (XSK) references  | 4.18 | [`fbfc504a24f5`](https://github.com/torvalds/linux/commit/fbfc504a24f53f7ebe128ab55cb5dba634f4ece8) | BPF_MAP_TYPE_XSKMAP
Socket references (hashmap)     | 4.18 | [`81110384441a`](https://github.com/torvalds/linux/commit/81110384441a59cff47430f20f049e69b98c17f4) | BPF_MAP_TYPE_SOCKHASH
cgroup storage                  | 4.19 | [`de9cbbaadba5`](https://github.com/torvalds/linux/commit/de9cbbaadba5adf88a19e46df61f7054000838f6) | BPF_MAP_TYPE_CGROUP_STORAGE
reuseport sockarray             | 4.19 | [`5dc4c4b7d4e8`](https://github.com/torvalds/linux/commit/5dc4c4b7d4e8115e7cde96a030f98cb3ab2e458c) | BPF_MAP_TYPE_REUSEPORT_SOCKARRAY
percpu cgroup storage           | 4.20 | [`b741f1630346`](https://github.com/torvalds/linux/commit/b741f1630346defcbc8cc60f1a2bdae8b3b0036f) | BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE
queue                           | 4.20 | [`f1a2e44a3aec`](https://github.com/torvalds/linux/commit/f1a2e44a3aeccb3ff18d3ccc0b0203e70b95bd92) | BPF_MAP_TYPE_QUEUE
stack                           | 4.20 | [`f1a2e44a3aec`](https://github.com/torvalds/linux/commit/f1a2e44a3aeccb3ff18d3ccc0b0203e70b95bd92) | BPF_MAP_TYPE_STACK
socket local storage            | 5.2  | [`6ac99e8f23d4`](https://github.com/torvalds/linux/commit/6ac99e8f23d4b10258406ca0dd7bffca5f31da9d) | BPF_MAP_TYPE_SK_STORAGE
Netdevice references (hashmap)  | 5.4  | [`6f9d451ab1a3`](https://github.com/torvalds/linux/commit/6f9d451ab1a33728adb72d7ff66a7b374d665176) | BPF_MAP_TYPE_DEVMAP_HASH
struct ops                      | 5.6  | [`85d33df357b6`](https://github.com/torvalds/linux/commit/85d33df357b634649ddbe0a20fd2d0fc5732c3cb) | BPF_MAP_TYPE_STRUCT_OPS
ring buffer                     | 5.8  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab) | BPF_MAP_TYPE_RINGBUF
inode storage                   | 5.10 | [`8ea636848aca`](https://github.com/torvalds/linux/commit/8ea636848aca35b9f97c5b5dee30225cf2dd0fe6) | BPF_MAP_TYPE_INODE_STORAGE
task storage                    | 5.11 | [`4cf1bc1f1045`](https://github.com/torvalds/linux/commit/4cf1bc1f10452065a29d576fc5693fc4fab5b919) | BPF_MAP_TYPE_TASK_STORAGE
Bloom filter                    | 5.16 | [`9330986c0300`](https://github.com/torvalds/linux/commit/9330986c03006ab1d33d243b7cfe598a7a3c1baa) | BPF_MAP_TYPE_BLOOM_FILTER
user ringbuf                    | 6.1  | [`583c1f420173`](https://github.com/torvalds/linux/commit/583c1f420173f7d84413a1a1fbf5109d798b4faa) | BPF_MAP_TYPE_USER_RINGBUF

### Map userspace API

Some (but not all) of these *API features* translate to a subcommand beginning with `BPF_MAP_`.
The list of subcommands supported in your kernel can be found in file
[`include/uapi/linux/bpf.h`](https://github.com/torvalds/linux/blob/master/include/uapi/linux/bpf.h):

    git grep -W 'bpf_cmd {' include/uapi/linux/bpf.h

Feature | Kernel version | Commit
--------|----------------|-------
Basic operations (lookup, update, delete, `GET_NEXT_KEY`) | 3.18 | [`db20fd2b0108`](https://github.com/torvalds/linux/commit/db20fd2b01087bdfbe30bce314a198eefedcc42e)
Pass flags to `UPDATE_ELEM` | 3.19 | [`3274f52073d8`](https://github.com/torvalds/linux/commit/3274f52073d88b62f3c5ace82ae9d48546232e72)
Pre-alloc map memory by default | 4.6 | [`6c9059817432`](https://github.com/torvalds/linux/commit/6c90598174322b8888029e40dd84a4eb01f56afe)
Pass `NULL` to `GET_NEXT_KEY` | 4.12 | [`8fe45924387b`](https://github.com/torvalds/linux/commit/8fe45924387be6b5c1be59a7eb330790c61d5d10)
Creation: select NUMA node | 4.14 | [`96eabe7a40aa`](https://github.com/torvalds/linux/commit/96eabe7a40aa17e613cf3db2c742ee8b1fc764d0)
Restrict access from syscall side | 4.15 | [`6e71b04a8224`](https://github.com/torvalds/linux/commit/6e71b04a82248ccf13a94b85cbc674a9fefe53f5)
Creation: specify map name | 4.15 | [`ad5b177bd73f`](https://github.com/torvalds/linux/commit/ad5b177bd73f5107d97c36f56395c4281fb6f089)
`LOOKUP_AND_DELETE_ELEM` | 4.20 | [`bd513cd08f10`](https://github.com/torvalds/linux/commit/bd513cd08f10cbe28856f99ae951e86e86803861)
Creation: `BPF_F_ZERO_SEED` | 5.0 | [`96b3b6c9091d`](https://github.com/torvalds/linux/commit/96b3b6c9091d23289721350e32c63cc8749686be)
`BPF_F_LOCK` flag for lookup / update | 5.1 | [`96049f3afd50`](https://github.com/torvalds/linux/commit/96049f3afd50fe8db69fa0068cdca822e747b1e4)
Restrict access from BPF side | 5.2 | [`591fe9888d78`](https://github.com/torvalds/linux/commit/591fe9888d7809d9ee5c828020b6c6ae27c37229)
`FREEZE` | 5.2 | [`87df15de441b`](https://github.com/torvalds/linux/commit/87df15de441bd4add7876ef584da8cabdd9a042a)
mmap() support for array maps | 5.5 | [`fc9702273e2e`](https://github.com/torvalds/linux/commit/fc9702273e2edb90400a34b3be76f7b08fa3344b)
`LOOKUP_BATCH` | 5.6 | [`cb4d03ab499d`](https://github.com/torvalds/linux/commit/cb4d03ab499d4c040f4ab6fd4389d2b49f42b5a5)
`UPDATE_BATCH`, `DELETE_BATCH` | 5.6 | [`aa2e93b8e58e`](https://github.com/torvalds/linux/commit/aa2e93b8e58e18442edfb2427446732415bc215e)
`LOOKUP_AND_DELETE_BATCH` | 5.6 | [`057996380a42`](https://github.com/torvalds/linux/commit/057996380a42bb64ccc04383cfa9c0ace4ea11f0)
`LOOKUP_AND_DELETE_ELEM` support for hash maps | 5.14 | [`3e87f192b405`](https://github.com/torvalds/linux/commit/3e87f192b405960c0fe83e0925bd0dadf4f8cf43)

## XDP

An approximate list of drivers or components supporting XDP programs for your
kernel can be retrieved with:

    git grep -l XDP_SETUP_PROG drivers/

Feature / Driver | Kernel version | Commit
-----------------|----------------|-------
XDP core architecture | 4.8 | [`6a773a15a1e8`](https://github.com/torvalds/linux/commit/6a773a15a1e8874e5eccd2f29190c31085912c95)
Action: drop | 4.8 | [`6a773a15a1e8`](https://github.com/torvalds/linux/commit/6a773a15a1e8874e5eccd2f29190c31085912c95)
Action: pass on to stack | 4.8 | [`6a773a15a1e8`](https://github.com/torvalds/linux/commit/6a773a15a1e8874e5eccd2f29190c31085912c95)
Action: direct forwarding (on same port) | 4.8 | [`6ce96ca348a9`](https://github.com/torvalds/linux/commit/6ce96ca348a9e949f8c43f4d3e98db367d93cffd)
Direct packet data write | 4.8 | [`4acf6c0b84c9`](https://github.com/torvalds/linux/commit/4acf6c0b84c91243c705303cd9ff16421914150d)
Mellanox `mlx4` driver | 4.8 | [`47a38e155037`](https://github.com/torvalds/linux/commit/47a38e155037f417c5740e24ccae6482aedf4b68)
Mellanox `mlx5` driver | 4.9 | [`86994156c736`](https://github.com/torvalds/linux/commit/86994156c736978d113e7927455d4eeeb2128b9f)
Netronome `nfp` driver | 4.10 | [`ecd63a0217d5`](https://github.com/torvalds/linux/commit/ecd63a0217d5f1e8a92f7516f5586d1177b95de2)
QLogic (Cavium) `qed*` drivers | 4.10 | [`496e05170958`](https://github.com/torvalds/linux/commit/496e051709588f832d7a6a420f44f8642b308a87)
`virtio_net` driver | 4.10 | [`f600b6905015`](https://github.com/torvalds/linux/commit/f600b690501550b94e83e07295d9c8b9c4c39f4e)
Broadcom `bnxt_en` driver | 4.11 | [`c6d30e8391b8`](https://github.com/torvalds/linux/commit/c6d30e8391b85e00eb544e6cf047ee0160ee9938)
Intel `ixgbe*` drivers | 4.12 | [`924708081629`](https://github.com/torvalds/linux/commit/9247080816297de4e31abb684939c0e53e3a8a67)
Cavium `thunderx` driver | 4.12 | [`05c773f52b96`](https://github.com/torvalds/linux/commit/05c773f52b96ef3fbc7d9bfa21caadc6247ef7a8)
Generic XDP | 4.12 | [`b5cdae3291f7`](https://github.com/torvalds/linux/commit/b5cdae3291f7be7a34e75affe4c0ec1f7f328b64)
Intel `i40e` driver | 4.13 | [`0c8493d90b6b`](https://github.com/torvalds/linux/commit/0c8493d90b6bb0f5c4fe9217db8f7203f24c0f28)
Action: redirect | 4.14 | [`6453073987ba`](https://github.com/torvalds/linux/commit/6453073987ba392510ab6c8b657844a9312c67f7)
Support for tap | 4.14 | [`761876c857cb`](https://github.com/torvalds/linux/commit/761876c857cb2ef8489fbee01907151da902af91)
Support for veth | 4.14 | [`d445516966dc`](https://github.com/torvalds/linux/commit/d445516966dcb2924741b13b27738b54df2af01a)
Intel `ixgbevf` driver | 4.17 | [`c7aec59657b6`](https://github.com/torvalds/linux/commit/c7aec59657b60f3a29fc7d3274ebefd698879301)
Freescale `dpaa2` driver | 5.0 | [`7e273a8ebdd3`](https://github.com/torvalds/linux/commit/7e273a8ebdd3b83f94eb8b49fc8ee61464f47cc2)
Socionext `netsec` driver | 5.3 | [`ba2b232108d3`](https://github.com/torvalds/linux/commit/ba2b232108d3c2951bab02930a00f23b0cffd5af)
TI `cpsw` driver | 5.3 | [`9ed4050c0d75`](https://github.com/torvalds/linux/commit/9ed4050c0d75768066a07cf66eef4f8dc9d79b52)
Intel `ice` driver |5.5| [`efc2214b6047`](https://github.com/torvalds/linux/commit/efc2214b6047b6f5b4ca53151eba62521b9452d6)
Solarflare `sfc` driver | 5.5 | [`eb9a36be7f3e`](https://github.com/torvalds/linux/commit/eb9a36be7f3ec414700af9a616f035eda1f1e63e)
Marvell `mvneta` driver | 5.5 | [`0db51da7a8e9`](https://github.com/torvalds/linux/commit/0db51da7a8e99f0803ec3a8e25c1a66234a219cb)
Microsoft `hv_netvsc` driver | 5.6 | [`351e1581395f`](https://github.com/torvalds/linux/commit/351e1581395fcc7fb952bbd7dda01238f69968fd)
Amazon `ena` driver | 5.6 | [`838c93dc5449`](https://github.com/torvalds/linux/commit/838c93dc5449e5d6378bae117b0a65a122cf7361)
`xen-netfront` driver | 5.9 | [`6c5aa6fc4def`](https://github.com/torvalds/linux/commit/6c5aa6fc4defc2a0977a2c59e4710d50fa1e834c)
Marvell `mvpp2` driver | 5.9 | [`07dd0a7aae7f`](https://github.com/torvalds/linux/commit/07dd0a7aae7f72af7cec18909581c2bb570edddc)
Intel `igb` driver | 5.10 | [`9cbc948b5a20`](https://github.com/torvalds/linux/commit/9cbc948b5a20c9c054d9631099c0426c16da546b)
Freescale `dpaa` driver | 5.11 | [`86c0c196cbe4`](https://github.com/torvalds/linux/commit/86c0c196cbe48f844721783d9162e46bc35c0c5a)
Intel `igc` driver | 5.13 | [`26575105d6ed`](https://github.com/torvalds/linux/commit/26575105d6ed8e2a8e43bd008fc7d98b75b90d5c)
Freescale `enetc` driver | 5.13 | [`d1b15102dd16`](https://github.com/torvalds/linux/commit/d1b15102dd16adc17fd5e4db8a485e6459f98906)
STMicro `stmmac` driver | 5.13 | [`5fabb01207a2`](https://github.com/torvalds/linux/commit/5fabb01207a2d3439a6abe1d08640de9c942945f)
`bonding` driver | 5.15 | [`9e2ee5c7e7c3`](https://github.com/torvalds/linux/commit/9e2ee5c7e7c35d195e2aa0692a7241d47a433d1e)
Marvell `otx2` | 5.16 | [`06059a1a9a4a`](https://github.com/torvalds/linux/commit/06059a1a9a4a58f139352c65b02989ea6077091a)
Microsoft `mana` driver | 5.17 | [`ed5356b53f07`](https://github.com/torvalds/linux/commit/ed5356b53f070dea5dff5a01b740561cb8222199)
Fungible `fun` driver | 5.18 | [`db37bc177dae`](https://github.com/torvalds/linux/commit/db37bc177dae89cef6fc37bdbe6b223929f70245)
Aquantia `atlantic` driver | 5.19 | [`0d14657f4083`](https://github.com/torvalds/linux/commit/0d14657f40830243266f972766f1e4d00436e648)
Mediatek `mtk` driver | 6.0 | [`7c26c20da5d4`](https://github.com/torvalds/linux/commit/7c26c20da5d420cde55618263be4aa2f6de53056)
Freescale `fec_enet` driver | 6.2 | [`6d6b39f180b8`](https://github.com/torvalds/linux/commit/6d6b39f180b83dfe1e938382b68dd1e6cb51363c)
Microchip `lan966x` driver | 6.2 | [`6a2159be7604`](https://github.com/torvalds/linux/commit/6a2159be7604f5cdd7f574f4e0922f61e63c3f16)
Engleder `tsnep` driver | 6.3 | [`d24bc0bcbbff`](https://github.com/torvalds/linux/commit/d24bc0bcbbfff74c00d92d0630ef99e8d91ef37a)
Google `gve` driver | 6.4 | [`2e80aeae9f80`](https://github.com/torvalds/linux/commit/2e80aeae9f807ac7e967dea7633abea5829c6531)
VMware `vmxnet3` driver | 6.6 | [`54f00cce1178`](https://github.com/torvalds/linux/commit/54f00cce11786742bd11e5e68c3bf85e6dc048c9)

## Helpers

The list of helpers supported in your kernel can be found in file
[`include/uapi/linux/bpf.h`](https://github.com/torvalds/linux/blob/master/include/uapi/linux/bpf.h):

    git grep '	FN(' include/uapi/linux/bpf.h

Alphabetical order

Helper | Kernel version | License | Commit |
-------|----------------|---------|--------|
`BPF_FUNC_bind()` | 4.17 |  | [`d74bad4e74ee`](https://github.com/torvalds/linux/commit/d74bad4e74ee373787a9ae24197c17b7cdc428d5) |
`BPF_FUNC_bprm_opts_set()` | 5.11 |  | [`3f6719c7b62f`](https://github.com/torvalds/linux/commit/3f6719c7b62f0327c9091e26d0da10e65668229e)
`BPF_FUNC_btf_find_by_name_kind()` | 5.14 |  | [`3d78417b60fb`](https://github.com/torvalds/linux/commit/3d78417b60fba249cc555468cb72d96f5cde2964)
`BPF_FUNC_cgrp_storage_delete()` | 6.2 | | [`c4bcfb38a95e`](https://github.com/torvalds/linux/commit/c4bcfb38a95edb1021a53f2d0356a78120ecfbe4)
`BPF_FUNC_cgrp_storage_get()` | 6.2 | | [`c4bcfb38a95e`](https://github.com/torvalds/linux/commit/c4bcfb38a95edb1021a53f2d0356a78120ecfbe4)
`BPF_FUNC_check_mtu()` | 5.12 |  | [`34b2021cc616`](https://github.com/torvalds/linux/commit/34b2021cc61642d61c3cf943d9e71925b827941b)
`BPF_FUNC_clone_redirect()` | 4.2 |  | [`3896d655f4d4`](https://github.com/torvalds/linux/commit/3896d655f4d491c67d669a15f275a39f713410f8)
`BPF_FUNC_copy_from_user()` | 5.10 |  | [`07be4c4a3e7a`](https://github.com/torvalds/linux/commit/07be4c4a3e7a0db148e44b16c5190e753d1c8569)
`BPF_FUNC_copy_from_user_task()` | 5.18 | GPL | [`376040e47334`](https://github.com/torvalds/linux/commit/376040e47334c6dc6a939a32197acceb00fe4acf)
`BPF_FUNC_csum_diff()` | 4.6 |  | [`7d672345ed29`](https://github.com/torvalds/linux/commit/7d672345ed295b1356a5d9f7111da1d1d7d65867)
`BPF_FUNC_csum_level()` | 5.7 |  | [`7cdec54f9713`](https://github.com/torvalds/linux/commit/7cdec54f9713256bb170873a1fc5c75c9127c9d2)
`BPF_FUNC_csum_update()` | 4.9 |  | [`36bbef52c7eb`](https://github.com/torvalds/linux/commit/36bbef52c7eb646ed6247055a2acd3851e317857)
`BPF_FUNC_current_task_under_cgroup()` | 4.9 |  | [`60d20f9195b2`](https://github.com/torvalds/linux/commit/60d20f9195b260bdf0ac10c275ae9f6016f9c069)
`BPF_FUNC_d_path()` | 5.10 |  | [`6e22ab9da793`](https://github.com/torvalds/linux/commit/6e22ab9da79343532cd3cde39df25e5a5478c692)
`BPF_FUNC_dynptr_data()` | 5.19 |  | [`34d4ef5775f7`](https://github.com/torvalds/linux/commit/34d4ef5775f776ec4b0d53a02d588bf3195cada6)
`BPF_FUNC_dynptr_from_mem()` | 5.19 |  | [`263ae152e962`](https://github.com/torvalds/linux/commit/263ae152e96253f40c2c276faad8629e096b3bad)
`BPF_FUNC_dynptr_read()` | 5.19 |  | [`13bbbfbea759`](https://github.com/torvalds/linux/commit/13bbbfbea7598ea9f8d9c3d73bf053bb57f9c4b2)
`BPF_FUNC_dynptr_write()` | 5.19 |  | [`13bbbfbea759`](https://github.com/torvalds/linux/commit/13bbbfbea7598ea9f8d9c3d73bf053bb57f9c4b2)
`BPF_FUNC_fib_lookup()` | 4.18 | GPL | [`87f5fc7e48dd`](https://github.com/torvalds/linux/commit/87f5fc7e48dd3175b30dd03b41564e1a8e136323)
`BPF_FUNC_find_vma()` | 5.17 | | [`7c7e3d31e785`](https://github.com/torvalds/linux/commit/7c7e3d31e7856a8260a254f8c71db416f7f9f5a1)
`BPF_FUNC_for_each_map_elem()` | 5.13 | | [`69c087ba6225`](https://github.com/torvalds/linux/commit/69c087ba6225b574afb6e505b72cb75242a3d844)
`BPF_FUNC_get_attach_cookie()` | 5.15 |  | [`7adfc6c9b315`](https://github.com/torvalds/linux/commit/7adfc6c9b315e174cf8743b21b7b691c8766791b)
`BPF_FUNC_get_branch_snapshot()` | 5.16 | GPL | [`856c02dbce4f`](https://github.com/torvalds/linux/commit/856c02dbce4f8d6a5644083db22c11750aa11481)
`BPF_FUNC_get_current_ancestor_cgroup_id()` | 5.6 |  | [`b4490c5c4e02`](https://github.com/torvalds/linux/commit/b4490c5c4e023f09b7d27c9a9d3e7ad7d09ea6bf)
`BPF_FUNC_get_cgroup_classid()` | 4.3 |  | [`8d20aabe1c76`](https://github.com/torvalds/linux/commit/8d20aabe1c76cccac544d9fcc3ad7823d9e98a2d)
`BPF_FUNC_get_current_cgroup_id()` | 4.18 |  | [`bf6fa2c893c5`](https://github.com/torvalds/linux/commit/bf6fa2c893c5237b48569a13fa3c673041430b6c)
`BPF_FUNC_get_current_comm()` | 4.2 |  | [`ffeedafbf023`](https://github.com/torvalds/linux/commit/ffeedafbf0236f03aeb2e8db273b3e5ae5f5bc89)
`BPF_FUNC_get_current_pid_tgid()` | 4.2 |  | [`ffeedafbf023`](https://github.com/torvalds/linux/commit/ffeedafbf0236f03aeb2e8db273b3e5ae5f5bc89)
`BPF_FUNC_get_current_task()` | 4.8 | GPL | [`606274c5abd8`](https://github.com/torvalds/linux/commit/606274c5abd8e245add01bc7145a8cbb92b69ba8)
`BPF_FUNC_get_current_task_btf()` | 5.11 | GPL | [`3ca1032ab7ab`](https://github.com/torvalds/linux/commit/3ca1032ab7ab010eccb107aa515598788f7d93bb)
`BPF_FUNC_get_current_uid_gid()` | 4.2 |  | [`ffeedafbf023`](https://github.com/torvalds/linux/commit/ffeedafbf0236f03aeb2e8db273b3e5ae5f5bc89)
`BPF_FUNC_get_func_arg()` | 5.17 |  | [`f92c1e183604`](https://github.com/torvalds/linux/commit/f92c1e183604c20ce00eb889315fdaa8f2d9e509)
`BPF_FUNC_get_func_arg_cnt()` | 5.17 |  | [`f92c1e183604`](https://github.com/torvalds/linux/commit/f92c1e183604c20ce00eb889315fdaa8f2d9e509)
`BPF_FUNC_get_func_ip()` | 5.15 |  | [`5d8b583d04ae`](https://github.com/torvalds/linux/commit/5d8b583d04aedb3bd5f6d227a334c210c7d735f9)
`BPF_FUNC_get_func_ret()` | 5.17 |  | [`f92c1e183604`](https://github.com/torvalds/linux/commit/f92c1e183604c20ce00eb889315fdaa8f2d9e509)
`BPF_FUNC_get_retval()` | 5.18 |  | [`b44123b4a3dc`](https://github.com/torvalds/linux/commit/b44123b4a3dcad4664d3a0f72c011ffd4c9c4d93)
`BPF_FUNC_get_hash_recalc()` | 4.8 |  | [`13c5c240f789`](https://github.com/torvalds/linux/commit/13c5c240f789bbd2bcacb14a23771491485ae61f)
`BPF_FUNC_get_listener_sock()` | 5.1 |  | [`dbafd7ddd623`](https://github.com/torvalds/linux/commit/dbafd7ddd62369b2f3926ab847cbf8fc40e800b7)
`BPF_FUNC_get_local_storage()` | 4.19 |  | [`cd3394317653`](https://github.com/torvalds/linux/commit/cd3394317653837e2eb5c5d0904a8996102af9fc)
`BPF_FUNC_get_netns_cookie()` | 5.7 |  | [`f318903c0bf4`](https://github.com/torvalds/linux/commit/f318903c0bf42448b4c884732df2bbb0ef7a2284)
`BPF_FUNC_get_ns_current_pid_tgid()` | 5.7 |  | [`b4490c5c4e02`](https://github.com/torvalds/linux/commit/b4490c5c4e023f09b7d27c9a9d3e7ad7d09ea6bf)
`BPF_FUNC_get_numa_node_id()` | 4.10 |  | [`2d0e30c30f84`](https://github.com/torvalds/linux/commit/2d0e30c30f84d08dc16f0f2af41f1b8a85f0755e)
`BPF_FUNC_get_prandom_u32()` | 4.1 |  | [`03e69b508b6f`](https://github.com/torvalds/linux/commit/03e69b508b6f7c51743055c9f61d1dfeadf4b635)
`BPF_FUNC_get_route_realm()` | 4.4 |  | [`c46646d0484f`](https://github.com/torvalds/linux/commit/c46646d0484f5d08e2bede9b45034ba5b8b489cc)
`BPF_FUNC_get_smp_processor_id()` | 4.1 |  | [`c04167ce2ca0`](https://github.com/torvalds/linux/commit/c04167ce2ca0ecaeaafef006cb0d65cf01b68e42)
`BPF_FUNC_get_socket_cookie()` | 4.12 |  | [`91b8270f2a4d`](https://github.com/torvalds/linux/commit/91b8270f2a4d1d9b268de90451cdca63a70052d6)
`BPF_FUNC_get_socket_uid()` | 4.12 |  | [`6acc5c291068`](https://github.com/torvalds/linux/commit/6acc5c2910689fc6ee181bf63085c5efff6a42bd)
`BPF_FUNC_get_stack()` | 4.18 | GPL | [`de2ff05f48af`](https://github.com/torvalds/linux/commit/de2ff05f48afcde816ff4edb217417f62f624ab5)
`BPF_FUNC_get_stackid()` | 4.6 | GPL | [`d5a3b1f69186`](https://github.com/torvalds/linux/commit/d5a3b1f691865be576c2bffa708549b8cdccda19)
`BPF_FUNC_get_task_stack()` | 5.9 | | [`fa28dcb82a38`](https://github.com/torvalds/linux/commit/fa28dcb82a38f8e3993b0fae9106b1a80b59e4f0)
`BPF_FUNC_getsockopt()` | 4.15 |  | [`cd86d1fd2102`](https://github.com/torvalds/linux/commit/cd86d1fd21025fdd6daf23d1288da405e7ad0ec6)
`BPF_FUNC_ima_file_hash()` | 5.18 |  | [`174b16946e39`](https://github.com/torvalds/linux/commit/174b16946e39ebd369097e0f773536c91a8c1a4c)
`BPF_FUNC_ima_inode_hash()` | 5.11 |  | [`27672f0d280a`](https://github.com/torvalds/linux/commit/27672f0d280a3f286a410a8db2004f46ace72a17)
`BPF_FUNC_inode_storage_delete()` | 5.10 |  | [`8ea636848aca`](https://github.com/torvalds/linux/commit/8ea636848aca35b9f97c5b5dee30225cf2dd0fe6)
`BPF_FUNC_inode_storage_get()` | 5.10 |  | [`8ea636848aca`](https://github.com/torvalds/linux/commit/8ea636848aca35b9f97c5b5dee30225cf2dd0fe6)
`BPF_FUNC_jiffies64()` | 5.5 |  | [`5576b991e9c1`](https://github.com/torvalds/linux/commit/5576b991e9c1a11d2cc21c4b94fc75ec27603896)
`BPF_FUNC_kallsyms_lookup_name()` | 5.16 | | [`d6aef08a872b`](https://github.com/torvalds/linux/commit/d6aef08a872b9e23eecc92d0e92393473b13c497)
`BPF_FUNC_kptr_xchg()` | 5.19 | | [`c0a5a21c25f3`](https://github.com/torvalds/linux/commit/c0a5a21c25f37c9fd7b36072f9968cdff1e4aa13)
`BPF_FUNC_ktime_get_boot_ns()` | 5.8 | | [`71d19214776e`](https://github.com/torvalds/linux/commit/71d19214776e61b33da48f7c1b46e522c7f78221)
`BPF_FUNC_ktime_get_coarse_ns()` | 5.11 | | [`d05512618056`](https://github.com/torvalds/linux/commit/d055126180564a57fe533728a4e93d0cb53d49b3)
`BPF_FUNC_ktime_get_ns()` | 4.1 | | [`d9847d310ab4`](https://github.com/torvalds/linux/commit/d9847d310ab4003725e6ed1822682e24bd406908)
`BPF_FUNC_ktime_get_tai_ns()` | 6.1 |  | [`c8996c98f703`](https://github.com/torvalds/linux/commit/c8996c98f703b09afe77a1d247dae691c9849dc1)
`BPF_FUNC_l3_csum_replace()` | 4.1 |  | [`91bc4822c3d6`](https://github.com/torvalds/linux/commit/91bc4822c3d61b9bb7ef66d3b77948a4f9177954)
`BPF_FUNC_l4_csum_replace()` | 4.1 |  | [`91bc4822c3d6`](https://github.com/torvalds/linux/commit/91bc4822c3d61b9bb7ef66d3b77948a4f9177954)
`BPF_FUNC_load_hdr_opt()` | 5.10 |  | [`0813a841566f`](https://github.com/torvalds/linux/commit/0813a841566f0962a5551be7749b43c45f0022a0)
`BPF_FUNC_loop()` | 5.17 |  | [`e6f2dd0f8067`](https://github.com/torvalds/linux/commit/e6f2dd0f80674e9d5960337b3e9c2a242441b326)
`BPF_FUNC_lwt_push_encap()` | 4.18 |  | [`fe94cc290f53`](https://github.com/torvalds/linux/commit/fe94cc290f535709d3c5ebd1e472dfd0aec7ee79)
`BPF_FUNC_lwt_seg6_action()` | 4.18 |  | [`fe94cc290f53`](https://github.com/torvalds/linux/commit/fe94cc290f535709d3c5ebd1e472dfd0aec7ee79)
`BPF_FUNC_lwt_seg6_adjust_srh()` | 4.18 |  | [`fe94cc290f53`](https://github.com/torvalds/linux/commit/fe94cc290f535709d3c5ebd1e472dfd0aec7ee79)
`BPF_FUNC_lwt_seg6_store_bytes()` | 4.18 |  | [`fe94cc290f53`](https://github.com/torvalds/linux/commit/fe94cc290f535709d3c5ebd1e472dfd0aec7ee79)
`BPF_FUNC_map_delete_elem()` | 3.19 |  | [`d0003ec01c66`](https://github.com/torvalds/linux/commit/d0003ec01c667b731c139e23de3306a8b328ccf5)
`BPF_FUNC_map_lookup_elem()` | 3.19 |  | [`d0003ec01c66`](https://github.com/torvalds/linux/commit/d0003ec01c667b731c139e23de3306a8b328ccf5)
`BPF_FUNC_map_lookup_percpu_elem()` | 5.19 |  | [`07343110b293`](https://github.com/torvalds/linux/commit/07343110b293456d30393e89b86c4dee1ac051c8)
`BPF_FUNC_map_peek_elem()` | 4.20 |  | [`f1a2e44a3aec`](https://github.com/torvalds/linux/commit/f1a2e44a3aeccb3ff18d3ccc0b0203e70b95bd92)
`BPF_FUNC_map_pop_elem()` | 4.20 |  | [`f1a2e44a3aec`](https://github.com/torvalds/linux/commit/f1a2e44a3aeccb3ff18d3ccc0b0203e70b95bd92)
`BPF_FUNC_map_push_elem()` | 4.20 |  | [`f1a2e44a3aec`](https://github.com/torvalds/linux/commit/f1a2e44a3aeccb3ff18d3ccc0b0203e70b95bd92)
`BPF_FUNC_map_update_elem()` | 3.19 |  | [`d0003ec01c66`](https://github.com/torvalds/linux/commit/d0003ec01c667b731c139e23de3306a8b328ccf5)
`BPF_FUNC_msg_apply_bytes()` | 4.17 |  | [`2a100317c9eb`](https://github.com/torvalds/linux/commit/2a100317c9ebc204a166f16294884fbf9da074ce)
`BPF_FUNC_msg_cork_bytes()` | 4.17 |  | [`91843d540a13`](https://github.com/torvalds/linux/commit/91843d540a139eb8070bcff8aa10089164436deb)
`BPF_FUNC_msg_pop_data()` | 5.0 |  | [`7246d8ed4dcc`](https://github.com/torvalds/linux/commit/7246d8ed4dcce23f7509949a77be15fa9f0e3d28)
`BPF_FUNC_msg_pull_data()` | 4.17 |  | [`015632bb30da`](https://github.com/torvalds/linux/commit/015632bb30daaaee64e1bcac07570860e0bf3092)
`BPF_FUNC_msg_push_data()` | 4.20 |  | [`6fff607e2f14`](https://github.com/torvalds/linux/commit/6fff607e2f14bd7c63c06c464a6f93b8efbabe28)
`BPF_FUNC_msg_redirect_hash()` | 4.18 |  | [`81110384441a`](https://github.com/torvalds/linux/commit/81110384441a59cff47430f20f049e69b98c17f4)
`BPF_FUNC_msg_redirect_map()` | 4.17 |  | [`4f738adba30a`](https://github.com/torvalds/linux/commit/4f738adba30a7cfc006f605707e7aee847ffefa0)
`BPF_FUNC_per_cpu_ptr()` | 5.10 |  | [`eaa6bcb71ef6`](https://github.com/torvalds/linux/commit/eaa6bcb71ef6ed3dc18fc525ee7e293b06b4882b) |
`BPF_FUNC_perf_event_output()` | 4.4 | GPL | [`a43eec304259`](https://github.com/torvalds/linux/commit/a43eec304259a6c637f4014a6d4767159b6a3aa3)
`BPF_FUNC_perf_event_read()` | 4.3 | GPL | [`35578d798400`](https://github.com/torvalds/linux/commit/35578d7984003097af2b1e34502bc943d40c1804)
`BPF_FUNC_perf_event_read_value()` | 4.15 | GPL | [`908432ca84fc`](https://github.com/torvalds/linux/commit/908432ca84fc229e906ba164219e9ad0fe56f755)
`BPF_FUNC_perf_prog_read_value()` | 4.15 | GPL | [`4bebdc7a85aa`](https://github.com/torvalds/linux/commit/4bebdc7a85aa400c0222b5329861e4ad9252f1e5)
`BPF_FUNC_probe_read()` | 4.1 | GPL | [`2541517c32be`](https://github.com/torvalds/linux/commit/2541517c32be2531e0da59dfd7efc1ce844644f5)
`BPF_FUNC_probe_read_kernel()` | 5.5 | GPL | [`6ae08ae3dea2`](https://github.com/torvalds/linux/commit/6ae08ae3dea2cfa03dd3665a3c8475c2d429ef47)
`BPF_FUNC_probe_read_kernel_str()` | 5.5 | GPL | [`6ae08ae3dea2`](https://github.com/torvalds/linux/commit/6ae08ae3dea2cfa03dd3665a3c8475c2d429ef47)
`BPF_FUNC_probe_read_user()` | 5.5 | GPL | [`6ae08ae3dea2`](https://github.com/torvalds/linux/commit/6ae08ae3dea2cfa03dd3665a3c8475c2d429ef47)
`BPF_FUNC_probe_read_user_str()` | 5.5 | GPL | [`6ae08ae3dea2`](https://github.com/torvalds/linux/commit/6ae08ae3dea2cfa03dd3665a3c8475c2d429ef47)
`BPF_FUNC_probe_read_str()` | 4.11 | GPL | [`a5e8c07059d0`](https://github.com/torvalds/linux/commit/a5e8c07059d0f0b31737408711d44794928ac218)
`BPF_FUNC_probe_write_user()` | 4.8 | GPL | [`96ae52279594`](https://github.com/torvalds/linux/commit/96ae52279594470622ff0585621a13e96b700600)
`BPF_FUNC_rc_keydown()` | 4.18 | GPL | [`f4364dcfc86d`](https://github.com/torvalds/linux/commit/f4364dcfc86df7c1ca47b256eaf6b6d0cdd0d936)
`BPF_FUNC_rc_pointer_rel()` | 5.0 | GPL | [`01d3240a04f4`](https://github.com/torvalds/linux/commit/01d3240a04f4c09392e13c77b54d4423ebce2d72)
`BPF_FUNC_rc_repeat()` | 4.18 | GPL | [`f4364dcfc86d`](https://github.com/torvalds/linux/commit/f4364dcfc86df7c1ca47b256eaf6b6d0cdd0d936)
`BPF_FUNC_read_branch_records()` | 5.6 | GPL | [`fff7b64355ea`](https://github.com/torvalds/linux/commit/fff7b64355eac6e29b50229ad1512315bc04b44e)
`BPF_FUNC_redirect()` | 4.4 |  | [`27b29f63058d`](https://github.com/torvalds/linux/commit/27b29f63058d26c6c1742f1993338280d5a41dc6)
`BPF_FUNC_redirect_map()` | 4.14 |  | [`97f91a7cf04f`](https://github.com/torvalds/linux/commit/97f91a7cf04ff605845c20948b8a80e54cbd3376)
`BPF_FUNC_redirect_neigh()` | 5.10 |  | [`b4ab31414970`](https://github.com/torvalds/linux/commit/b4ab31414970a7a03a5d55d75083f2c101a30592)
`BPF_FUNC_redirect_peer()` | 5.10 |  | [`9aa1206e8f48`](https://github.com/torvalds/linux/commit/9aa1206e8f48222f35a0c809f33b2f4aaa1e2661)
`BPF_FUNC_reserve_hdr_opt()` | 5.10 |  | [`0813a841566f`](https://github.com/torvalds/linux/commit/0813a841566f0962a5551be7749b43c45f0022a0)
`BPF_FUNC_ringbuf_discard()` | 5.8 |  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab)
`BPF_FUNC_ringbuf_discard_dynptr()` | 5.19 |  | [`bc34dee65a65`](https://github.com/torvalds/linux/commit/bc34dee65a65e9c920c420005b8a43f2a721a458)
`BPF_FUNC_ringbuf_output()` | 5.8 |  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab)
`BPF_FUNC_ringbuf_query()` | 5.8 |  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab)
`BPF_FUNC_ringbuf_reserve()` | 5.8 |  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab)
`BPF_FUNC_ringbuf_reserve_dynptr()` | 5.19 |  | [`bc34dee65a65`](https://github.com/torvalds/linux/commit/bc34dee65a65e9c920c420005b8a43f2a721a458)
`BPF_FUNC_ringbuf_submit()` | 5.8 |  | [`457f44363a88`](https://github.com/torvalds/linux/commit/457f44363a8894135c85b7a9afd2bd8196db24ab)
`BPF_FUNC_ringbuf_submit_dynptr()` | 5.19 |  | [`bc34dee65a65`](https://github.com/torvalds/linux/commit/bc34dee65a65e9c920c420005b8a43f2a721a458)
`BPF_FUNC_send_signal()` | 5.3 |  | [`8b401f9ed244`](https://github.com/torvalds/linux/commit/8b401f9ed2441ad9e219953927a842d24ed051fc)
`BPF_FUNC_send_signal_thread()` | 5.5 |  | [`8482941f0906`](https://github.com/torvalds/linux/commit/8482941f09067da42f9c3362e15bfb3f3c19d610)
`BPF_FUNC_seq_printf()` | 5.7 | GPL | [`492e639f0c22`](https://github.com/torvalds/linux/commit/492e639f0c222784e2e0f121966375f641c61b15)
`BPF_FUNC_seq_printf_btf()` | 5.10 | | [`eb411377aed9`](https://github.com/torvalds/linux/commit/eb411377aed9e27835e77ee0710ee8f4649958f3)
`BPF_FUNC_seq_write()` | 5.7 | GPL | [`492e639f0c22`](https://github.com/torvalds/linux/commit/492e639f0c222784e2e0f121966375f641c61b15)
`BPF_FUNC_set_hash()` | 4.13 |  | [`ded092cd73c2`](https://github.com/torvalds/linux/commit/ded092cd73c2c56a394b936f86897f29b2e131c0)
`BPF_FUNC_set_hash_invalid()` | 4.9 |  | [`7a4b28c6cc9f`](https://github.com/torvalds/linux/commit/7a4b28c6cc9ffac50f791b99cc7e46106436e5d8)
`BPF_FUNC_set_retval()` | 5.18 |  | [`b44123b4a3dc`](https://github.com/torvalds/linux/commit/b44123b4a3dcad4664d3a0f72c011ffd4c9c4d93)
`BPF_FUNC_setsockopt()` | 4.13 |  | [`8c4b4c7e9ff0`](https://github.com/torvalds/linux/commit/8c4b4c7e9ff0447995750d9329949fa082520269)
`BPF_FUNC_sk_ancestor_cgroup_id()` | 5.7 |  | [`f307fa2cb4c9`](https://github.com/torvalds/linux/commit/f307fa2cb4c935f7f1ff0aeb880c7b44fb9a642b)
`BPF_FUNC_sk_assign()` | 5.6 |  | [`cf7fbe660f2d`](https://github.com/torvalds/linux/commit/cf7fbe660f2dbd738ab58aea8e9b0ca6ad232449)
`BPF_FUNC_sk_cgroup_id()` | 5.7 |  | [`f307fa2cb4c9`](https://github.com/torvalds/linux/commit/f307fa2cb4c935f7f1ff0aeb880c7b44fb9a642b)
`BPF_FUNC_sk_fullsock()` | 5.1 |  | [`46f8bc92758c`](https://github.com/torvalds/linux/commit/46f8bc92758c6259bcf945e9216098661c1587cd)
`BPF_FUNC_sk_lookup_tcp()` | 4.20 |  | [`6acc9b432e67`](https://github.com/torvalds/linux/commit/6acc9b432e6714d72d7d77ec7c27f6f8358d0c71)
`BPF_FUNC_sk_lookup_udp()` | 4.20 |  | [`6acc9b432e67`](https://github.com/torvalds/linux/commit/6acc9b432e6714d72d7d77ec7c27f6f8358d0c71)
`BPF_FUNC_sk_redirect_hash()` | 4.18 |  | [`81110384441a`](https://github.com/torvalds/linux/commit/81110384441a59cff47430f20f049e69b98c17f4)
`BPF_FUNC_sk_redirect_map()` | 4.14 |  | [`174a79ff9515`](https://github.com/torvalds/linux/commit/174a79ff9515f400b9a6115643dafd62a635b7e6)
`BPF_FUNC_sk_release()` | 4.20 |  | [`6acc9b432e67`](https://github.com/torvalds/linux/commit/6acc9b432e6714d72d7d77ec7c27f6f8358d0c71)
`BPF_FUNC_sk_select_reuseport()` | 4.19 |  | [`2dbb9b9e6df6`](https://github.com/torvalds/linux/commit/2dbb9b9e6df67d444fbe425c7f6014858d337adf)
`BPF_FUNC_sk_storage_delete()` | 5.2 |  | [`6ac99e8f23d4`](https://github.com/torvalds/linux/commit/6ac99e8f23d4b10258406ca0dd7bffca5f31da9d)
`BPF_FUNC_sk_storage_get()` | 5.2 |  | [`6ac99e8f23d4`](https://github.com/torvalds/linux/commit/6ac99e8f23d4b10258406ca0dd7bffca5f31da9d)
`BPF_FUNC_skb_adjust_room()` | 4.13 |  | [`2be7e212d541`](https://github.com/torvalds/linux/commit/2be7e212d5419a400d051c84ca9fdd083e5aacac)
`BPF_FUNC_skb_ancestor_cgroup_id()` | 4.19 |  | [`7723628101aa`](https://github.com/torvalds/linux/commit/7723628101aaeb1d723786747529b4ea65c5b5c5)
`BPF_FUNC_skb_change_head()` | 4.10 |  | [`3a0af8fd61f9`](https://github.com/torvalds/linux/commit/3a0af8fd61f90920f6fa04e4f1e9a6a73c1b4fd2)
`BPF_FUNC_skb_change_proto()` | 4.8 |  | [`6578171a7ff0`](https://github.com/torvalds/linux/commit/6578171a7ff0c31dc73258f93da7407510abf085)
`BPF_FUNC_skb_change_tail()` | 4.9 |  | [`5293efe62df8`](https://github.com/torvalds/linux/commit/5293efe62df81908f2e90c9820c7edcc8e61f5e9)
`BPF_FUNC_skb_change_type()` | 4.8 |  | [`d2485c4242a8`](https://github.com/torvalds/linux/commit/d2485c4242a826fdf493fd3a27b8b792965b9b9e)
`BPF_FUNC_skb_cgroup_classid()` | 5.10 |  | [`b426ce83baa7`](https://github.com/torvalds/linux/commit/b426ce83baa7dff947fb354118d3133f2953aac8)
`BPF_FUNC_skb_cgroup_id()` | 4.18 |  | [`cb20b08ead40`](https://github.com/torvalds/linux/commit/cb20b08ead401fd17627a36f035c0bf5bfee5567)
`BPF_FUNC_skb_ecn_set_ce()` | 5.1 |  | [`f7c917ba11a6`](https://github.com/torvalds/linux/commit/f7c917ba11a67632a8452ea99fe132f626a7a2cc)
`BPF_FUNC_skb_get_tunnel_key()` | 4.3 |  | [`d3aa45ce6b94`](https://github.com/torvalds/linux/commit/d3aa45ce6b94c65b83971257317867db13e5f492)
`BPF_FUNC_skb_get_tunnel_opt()` | 4.6 |  | [`14ca0751c96f`](https://github.com/torvalds/linux/commit/14ca0751c96f8d3d0f52e8ed3b3236f8b34d3460)
`BPF_FUNC_skb_get_xfrm_state()` | 4.18 |  | [`12bed760a78d`](https://github.com/torvalds/linux/commit/12bed760a78da6e12ac8252fec64d019a9eac523)
`BPF_FUNC_skb_load_bytes()` | 4.5 |  | [`05c74e5e53f6`](https://github.com/torvalds/linux/commit/05c74e5e53f6cb07502c3e6a820f33e2777b6605)
`BPF_FUNC_skb_load_bytes_relative()` | 4.18 |  | [`4e1ec56cdc59`](https://github.com/torvalds/linux/commit/4e1ec56cdc59746943b2acfab3c171b930187bbe)
`BPF_FUNC_skb_output()` | 5.5 |  | [`a7658e1a4164`](https://github.com/torvalds/linux/commit/a7658e1a4164ce2b9eb4a11aadbba38586e93bd6)
`BPF_FUNC_skb_pull_data()` | 4.9 |  | [`36bbef52c7eb`](https://github.com/torvalds/linux/commit/36bbef52c7eb646ed6247055a2acd3851e317857)
`BPF_FUNC_skb_set_tstamp()` | 5.18 |  | [`9bb984f28d5b`](https://github.com/torvalds/linux/commit/9bb984f28d5bcb917d35d930fcfb89f90f9449fd)
`BPF_FUNC_skb_set_tunnel_key()` | 4.3 |  | [`d3aa45ce6b94`](https://github.com/torvalds/linux/commit/d3aa45ce6b94c65b83971257317867db13e5f492)
`BPF_FUNC_skb_set_tunnel_opt()` | 4.6 |  | [`14ca0751c96f`](https://github.com/torvalds/linux/commit/14ca0751c96f8d3d0f52e8ed3b3236f8b34d3460)
`BPF_FUNC_skb_store_bytes()` | 4.1 |  | [`91bc4822c3d6`](https://github.com/torvalds/linux/commit/91bc4822c3d61b9bb7ef66d3b77948a4f9177954)
`BPF_FUNC_skb_under_cgroup()` | 4.8 |  | [`4a482f34afcc`](https://github.com/torvalds/linux/commit/4a482f34afcc162d8456f449b137ec2a95be60d8)
`BPF_FUNC_skb_vlan_pop()` | 4.3 |  | [`4e10df9a60d9`](https://github.com/torvalds/linux/commit/4e10df9a60d96ced321dd2af71da558c6b750078)
`BPF_FUNC_skb_vlan_push()` | 4.3 |  | [`4e10df9a60d9`](https://github.com/torvalds/linux/commit/4e10df9a60d96ced321dd2af71da558c6b750078)
`BPF_FUNC_skc_lookup_tcp()` | 5.2 |  | [`edbf8c01de5a`](https://github.com/torvalds/linux/commit/edbf8c01de5a104a71ed6df2bf6421ceb2836a8e)
`BPF_FUNC_skc_to_mctcp_sock()` | 5.19 |  | [`3bc253c2e652`](https://github.com/torvalds/linux/commit/3bc253c2e652cf5f12cd8c00d80d8ec55d67d1a7)
`BPF_FUNC_skc_to_tcp_sock()` | 5.9 |  | [`478cfbdf5f13`](https://github.com/torvalds/linux/commit/478cfbdf5f13dfe09cfd0b1cbac821f5e27f6108)
`BPF_FUNC_skc_to_tcp_request_sock()` | 5.9 |  | [`478cfbdf5f13`](https://github.com/torvalds/linux/commit/478cfbdf5f13dfe09cfd0b1cbac821f5e27f6108)
`BPF_FUNC_skc_to_tcp_timewait_sock()` | 5.9 |  | [`478cfbdf5f13`](https://github.com/torvalds/linux/commit/478cfbdf5f13dfe09cfd0b1cbac821f5e27f6108)
`BPF_FUNC_skc_to_tcp6_sock()` | 5.9 |  | [`af7ec1383361`](https://github.com/torvalds/linux/commit/af7ec13833619e17f03aa73a785a2f871da6d66b)
`BPF_FUNC_skc_to_udp6_sock()` | 5.9 |  | [`0d4fad3e57df`](https://github.com/torvalds/linux/commit/0d4fad3e57df2bf61e8ffc8d12a34b1caf9b8835)
`BPF_FUNC_skc_to_unix_sock()` | 5.16 |  | [`9eeb3aa33ae0`](https://github.com/torvalds/linux/commit/9eeb3aa33ae005526f672b394c1791578463513f)
`BPF_FUNC_snprintf()` | 5.13 | | [`7b15523a989b`](https://github.com/torvalds/linux/commit/7b15523a989b63927c2bb08e9b5b0bbc10b58bef)
`BPF_FUNC_snprintf_btf()` | 5.10 | | [`c4d0bfb45068`](https://github.com/torvalds/linux/commit/c4d0bfb45068d853a478b9067a95969b1886a30f)
`BPF_FUNC_sock_from_file()` | 5.11 |  | [`4f19cab76136`](https://github.com/torvalds/linux/commit/4f19cab76136e800a3f04d8c9aa4d8e770e3d3d8)
`BPF_FUNC_sock_hash_update()` | 4.18 |  | [`81110384441a`](https://github.com/torvalds/linux/commit/81110384441a59cff47430f20f049e69b98c17f4)
`BPF_FUNC_sock_map_update()` | 4.14 |  | [`174a79ff9515`](https://github.com/torvalds/linux/commit/174a79ff9515f400b9a6115643dafd62a635b7e6)
`BPF_FUNC_spin_lock()` | 5.1 |  | [`d83525ca62cf`](https://github.com/torvalds/linux/commit/d83525ca62cf8ebe3271d14c36fb900c294274a2)
`BPF_FUNC_spin_unlock()` | 5.1 |  | [`d83525ca62cf`](https://github.com/torvalds/linux/commit/d83525ca62cf8ebe3271d14c36fb900c294274a2)
`BPF_FUNC_store_hdr_opt()` | 5.10 |  | [`0813a841566f`](https://github.com/torvalds/linux/commit/0813a841566f0962a5551be7749b43c45f0022a0)
`BPF_FUNC_strncmp()` | 5.17 |  | [`c5fb19937455`](https://github.com/torvalds/linux/commit/c5fb19937455095573a19ddcbff32e993ed10e35)
`BPF_FUNC_strtol()` | 5.2 |  | [`d7a4cb9b6705`](https://github.com/torvalds/linux/commit/d7a4cb9b6705a89937d12c8158a35a3145dc967a)
`BPF_FUNC_strtoul()` | 5.2 |  | [`d7a4cb9b6705`](https://github.com/torvalds/linux/commit/d7a4cb9b6705a89937d12c8158a35a3145dc967a)
`BPF_FUNC_sys_bpf()` | 5.14 |  | [`79a7f8bdb159`](https://github.com/torvalds/linux/commit/79a7f8bdb159d9914b58740f3d31d602a6e4aca8)
`BPF_FUNC_sys_close()` | 5.14 |  | [`3abea089246f`](https://github.com/torvalds/linux/commit/3abea089246f76c1517b054ddb5946f3f1dbd2c0)
`BPF_FUNC_sysctl_get_current_value()` | 5.2 |  | [`1d11b3016cec`](https://github.com/torvalds/linux/commit/1d11b3016cec4ed9770b98e82a61708c8f4926e7)
`BPF_FUNC_sysctl_get_name()` | 5.2 |  | [`808649fb787d`](https://github.com/torvalds/linux/commit/808649fb787d918a48a360a668ee4ee9023f0c11)
`BPF_FUNC_sysctl_get_new_value()` | 5.2 |  | [`4e63acdff864`](https://github.com/torvalds/linux/commit/4e63acdff864654cee0ac5aaeda3913798ee78f6)
`BPF_FUNC_sysctl_set_new_value()` | 5.2 |  | [`4e63acdff864`](https://github.com/torvalds/linux/commit/4e63acdff864654cee0ac5aaeda3913798ee78f6)
`BPF_FUNC_tail_call()` | 4.2 |  | [`04fd61ab36ec`](https://github.com/torvalds/linux/commit/04fd61ab36ec065e194ab5e74ae34a5240d992bb)
`BPF_FUNC_task_pt_regs()` | 5.15 | GPL | [`dd6e10fbd9f`](https://github.com/torvalds/linux/commit/dd6e10fbd9fb86a571d925602c8a24bb4d09a2a7)
`BPF_FUNC_task_storage_delete()` | 5.11 |  | [`4cf1bc1f1045`](https://github.com/torvalds/linux/commit/4cf1bc1f10452065a29d576fc5693fc4fab5b919)
`BPF_FUNC_task_storage_get()` | 5.11 |  | [`4cf1bc1f1045`](https://github.com/torvalds/linux/commit/4cf1bc1f10452065a29d576fc5693fc4fab5b919)
`BPF_FUNC_tcp_check_syncookie()` | 5.2 |  | [`399040847084`](https://github.com/torvalds/linux/commit/399040847084a69f345e0a52fd62f04654e0fce3)
`BPF_FUNC_tcp_gen_syncookie()` | 5.3 |  | [`70d66244317e`](https://github.com/torvalds/linux/commit/70d66244317e958092e9c971b08dd5b7fd29d9cb#diff-05da4bf36c7fbcd176254e1615d98b28)
`BPF_FUNC_tcp_raw_check_syncookie_ipv4()` | 6.0 |  | [`33bf9885040c`](https://github.com/torvalds/linux/commit/33bf9885040c399cf6a95bd33216644126728e14)
`BPF_FUNC_tcp_raw_check_syncookie_ipv6()` | 6.0 |  | [`33bf9885040c`](https://github.com/torvalds/linux/commit/33bf9885040c399cf6a95bd33216644126728e14)
`BPF_FUNC_tcp_raw_gen_syncookie_ipv4()` | 6.0 |  | [`33bf9885040c`](https://github.com/torvalds/linux/commit/33bf9885040c399cf6a95bd33216644126728e14)
`BPF_FUNC_tcp_raw_gen_syncookie_ipv6()` | 6.0 |  | [`33bf9885040c`](https://github.com/torvalds/linux/commit/33bf9885040c399cf6a95bd33216644126728e14)
`BPF_FUNC_tcp_send_ack()` | 5.5 | | [`206057fe020a`](https://github.com/torvalds/linux/commit/206057fe020ac5c037d5e2dd6562a9bd216ec765)
`BPF_FUNC_tcp_sock()` | 5.1 |  | [`655a51e536c0`](https://github.com/torvalds/linux/commit/655a51e536c09d15ffa3603b1b6fce2b45b85a1f)
`BPF_FUNC_this_cpu_ptr()` | 5.10 |  | [`63d9b80dcf2c`](https://github.com/torvalds/linux/commit/63d9b80dcf2c67bc5ade61cbbaa09d7af21f43f1) |
`BPF_FUNC_timer_init()` | 5.15 |  | [`b00628b1c7d5`](https://github.com/torvalds/linux/commit/b00628b1c7d595ae5b544e059c27b1f5828314b4)
`BPF_FUNC_timer_set_callback()` | 5.15 |  | [`b00628b1c7d5`](https://github.com/torvalds/linux/commit/b00628b1c7d595ae5b544e059c27b1f5828314b4)
`BPF_FUNC_timer_start()` | 5.15 |  | [`b00628b1c7d5`](https://github.com/torvalds/linux/commit/b00628b1c7d595ae5b544e059c27b1f5828314b4)
`BPF_FUNC_timer_cancel()` | 5.15 |  | [`b00628b1c7d5`](https://github.com/torvalds/linux/commit/b00628b1c7d595ae5b544e059c27b1f5828314b4)
`BPF_FUNC_trace_printk()` | 4.1 | GPL | [`9c959c863f82`](https://github.com/torvalds/linux/commit/9c959c863f8217a2ff3d7c296e8223654d240569)
`BPF_FUNC_trace_vprintk()` | 5.16 | GPL | [`10aceb629e19`](https://github.com/torvalds/linux/commit/10aceb629e198429c849d5e995c3bb1ba7a9aaa3)
`BPF_FUNC_user_ringbuf_drain()` | 6.1 | | [`205715673844`](https://github.com/torvalds/linux/commit/20571567384428dfc9fe5cf9f2e942e1df13c2dd)
`BPF_FUNC_xdp_adjust_head()` | 4.10 |  | [`17bedab27231`](https://github.com/torvalds/linux/commit/17bedab2723145d17b14084430743549e6943d03)
`BPF_FUNC_xdp_adjust_meta()` | 4.15 |  | [`de8f3a83b0a0`](https://github.com/torvalds/linux/commit/de8f3a83b0a0fddb2cf56e7a718127e9619ea3da)
`BPF_FUNC_xdp_adjust_tail()` | 4.18 |  | [`b32cc5b9a346`](https://github.com/torvalds/linux/commit/b32cc5b9a346319c171e3ad905e0cddda032b5eb)
`BPF_FUNC_xdp_get_buff_len()` | 5.18 |  | [`0165cc817075`](https://github.com/torvalds/linux/commit/0165cc817075cf701e4289838f1d925ff1911b3e)
`BPF_FUNC_xdp_load_bytes()` | 5.18 |  | [`3f364222d032`](https://github.com/torvalds/linux/commit/3f364222d032eea6b245780e845ad213dab28cdd)
`BPF_FUNC_xdp_store_bytes()` | 5.18 |  | [`3f364222d032`](https://github.com/torvalds/linux/commit/3f364222d032eea6b245780e845ad213dab28cdd)
`BPF_FUNC_xdp_output()` | 5.6 | GPL | [`d831ee84bfc9`](https://github.com/torvalds/linux/commit/d831ee84bfc9173eecf30dbbc2553ae81b996c60)
`BPF_FUNC_override_return()` | 4.16 | GPL | [`9802d86585db`](https://github.com/torvalds/linux/commit/9802d86585db91655c7d1929a4f6bbe0952ea88e)
`BPF_FUNC_sock_ops_cb_flags_set()` | 4.16 |  | [`b13d88072172`](https://github.com/torvalds/linux/commit/b13d880721729384757f235166068c315326f4a1)

Note: GPL-only BPF helpers require a GPL-compatible license. The current licenses considered GPL-compatible by the kernel are:

* GPL
* GPL v2
* GPL and additional rights
* Dual BSD/GPL
* Dual MIT/GPL
* Dual MPL/GPL

Check the list of GPL-compatible licenses in your [kernel source code](https://github.com/torvalds/linux/blob/master/include/linux/license.h).

## Program Types
The list of program types and supported helper functions can be retrieved with:

    git grep -W 'func_proto(enum bpf_func_id func_id' kernel/ net/ drivers/

|Program Type| Helper Functions|
|------------|-----------------|
|`BPF_PROG_TYPE_SOCKET_FILTER`|`BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_skb_load_bytes_relative()`  `BPF_FUNC_get_socket_cookie()`  `BPF_FUNC_get_socket_uid()`  `BPF_FUNC_perf_event_output()`  `Base functions`|
|`BPF_PROG_TYPE_KPROBE`|`BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_stackid()`  `BPF_FUNC_get_stack()`  `BPF_FUNC_perf_event_read_value()`  `BPF_FUNC_override_return()`  `Tracing functions`|
|`BPF_PROG_TYPE_SCHED_CLS`  `BPF_PROG_TYPE_SCHED_ACT`|`BPF_FUNC_skb_store_bytes()`  `BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_skb_load_bytes_relative()`  `BPF_FUNC_skb_pull_data()`  `BPF_FUNC_csum_diff()`  `BPF_FUNC_csum_update()`  `BPF_FUNC_l3_csum_replace()`  `BPF_FUNC_l4_csum_replace()`  `BPF_FUNC_clone_redirect()`  `BPF_FUNC_get_cgroup_classid()`  `BPF_FUNC_skb_vlan_push()`  `BPF_FUNC_skb_vlan_pop()`  `BPF_FUNC_skb_change_proto()`  `BPF_FUNC_skb_change_type()`  `BPF_FUNC_skb_adjust_room()`  `BPF_FUNC_skb_change_tail()`  `BPF_FUNC_skb_get_tunnel_key()`  `BPF_FUNC_skb_set_tunnel_key()`  `BPF_FUNC_skb_get_tunnel_opt()`  `BPF_FUNC_skb_set_tunnel_opt()`  `BPF_FUNC_redirect()`  `BPF_FUNC_get_route_realm()`  `BPF_FUNC_get_hash_recalc()`  `BPF_FUNC_set_hash_invalid()`  `BPF_FUNC_set_hash()`  `BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_smp_processor_id()`  `BPF_FUNC_skb_under_cgroup()`  `BPF_FUNC_get_socket_cookie()`  `BPF_FUNC_get_socket_uid()`  `BPF_FUNC_fib_lookup()`  `BPF_FUNC_skb_get_xfrm_state()`  `BPF_FUNC_skb_cgroup_id()`  `Base functions`|
|`BPF_PROG_TYPE_TRACEPOINT`|`BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_stackid()`  `BPF_FUNC_get_stack()`  `BPF_FUNC_d_path()`  `Tracing functions`|
|`BPF_PROG_TYPE_XDP`| `BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_smp_processor_id()`  `BPF_FUNC_csum_diff()`  `BPF_FUNC_xdp_adjust_head()`  `BPF_FUNC_xdp_adjust_meta()`  `BPF_FUNC_redirect()`  `BPF_FUNC_redirect_map()`  `BPF_FUNC_xdp_adjust_tail()`  `BPF_FUNC_fib_lookup()`  `Base functions`|
|`BPF_PROG_TYPE_PERF_EVENT`| `BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_stackid()`  `BPF_FUNC_get_stack()`  `BPF_FUNC_perf_prog_read_value()`  `Tracing functions`|
|`BPF_PROG_TYPE_CGROUP_SKB`|`BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_skb_load_bytes_relative()`  `BPF_FUNC_get_socket_cookie()`  `BPF_FUNC_get_socket_uid()`  `Base functions`|
|`BPF_PROG_TYPE_CGROUP_SOCK`|`BPF_FUNC_get_current_uid_gid()`  `Base functions`|
|`BPF_PROG_TYPE_LWT_IN`|`BPF_FUNC_lwt_push_encap()`  `LWT functions`  `Base functions`|
|`BPF_PROG_TYPE_LWT_OUT`| `LWT functions`  `Base functions`|
|`BPF_PROG_TYPE_LWT_XMIT`| `BPF_FUNC_skb_get_tunnel_key()`  `BPF_FUNC_skb_set_tunnel_key()`  `BPF_FUNC_skb_get_tunnel_opt()`  `BPF_FUNC_skb_set_tunnel_opt()`  `BPF_FUNC_redirect()`  `BPF_FUNC_clone_redirect()`  `BPF_FUNC_skb_change_tail()`  `BPF_FUNC_skb_change_head()`  `BPF_FUNC_skb_store_bytes()`  `BPF_FUNC_csum_update()`  `BPF_FUNC_l3_csum_replace()`  `BPF_FUNC_l4_csum_replace()`  `BPF_FUNC_set_hash_invalid()`  `LWT functions`|
|`BPF_PROG_TYPE_SOCK_OPS`|`BPF_FUNC_setsockopt()`  `BPF_FUNC_getsockopt()`  `BPF_FUNC_sock_ops_cb_flags_set()`  `BPF_FUNC_sock_map_update()`  `BPF_FUNC_sock_hash_update()`  `BPF_FUNC_get_socket_cookie()`  `Base functions`|
|`BPF_PROG_TYPE_SK_SKB`|`BPF_FUNC_skb_store_bytes()`  `BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_skb_pull_data()`  `BPF_FUNC_skb_change_tail()`  `BPF_FUNC_skb_change_head()`  `BPF_FUNC_get_socket_cookie()`  `BPF_FUNC_get_socket_uid()`  `BPF_FUNC_sk_redirect_map()`  `BPF_FUNC_sk_redirect_hash()`  `BPF_FUNC_sk_lookup_tcp()`  `BPF_FUNC_sk_lookup_udp()`  `BPF_FUNC_sk_release()`  `Base functions`|
|`BPF_PROG_TYPE_CGROUP_DEVICE`|`BPF_FUNC_map_lookup_elem()`  `BPF_FUNC_map_update_elem()`  `BPF_FUNC_map_delete_elem()`  `BPF_FUNC_get_current_uid_gid()`  `BPF_FUNC_trace_printk()`|
|`BPF_PROG_TYPE_SK_MSG`|`BPF_FUNC_msg_redirect_map()`  `BPF_FUNC_msg_redirect_hash()`  `BPF_FUNC_msg_apply_bytes()`  `BPF_FUNC_msg_cork_bytes()`  `BPF_FUNC_msg_pull_data()`  `BPF_FUNC_msg_push_data()`  `BPF_FUNC_msg_pop_data()`  `Base functions`|
|`BPF_PROG_TYPE_RAW_TRACEPOINT`|`BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_stackid()`  `BPF_FUNC_get_stack()`  `BPF_FUNC_skb_output()`  `Tracing functions`|
|`BPF_PROG_TYPE_CGROUP_SOCK_ADDR`|`BPF_FUNC_get_current_uid_gid()`  `BPF_FUNC_bind()`  `BPF_FUNC_get_socket_cookie()`  `Base functions`|
|`BPF_PROG_TYPE_LWT_SEG6LOCAL`|`BPF_FUNC_lwt_seg6_store_bytes()`  `BPF_FUNC_lwt_seg6_action()`  `BPF_FUNC_lwt_seg6_adjust_srh()`  `LWT functions`|
|`BPF_PROG_TYPE_LIRC_MODE2`|`BPF_FUNC_rc_repeat()`  `BPF_FUNC_rc_keydown()`  `BPF_FUNC_rc_pointer_rel()`  `BPF_FUNC_map_lookup_elem()`  `BPF_FUNC_map_update_elem()`  `BPF_FUNC_map_delete_elem()`  `BPF_FUNC_ktime_get_ns()`  `BPF_FUNC_tail_call()`  `BPF_FUNC_get_prandom_u32()`  `BPF_FUNC_trace_printk()`|
|`BPF_PROG_TYPE_SK_REUSEPORT`|`BPF_FUNC_sk_select_reuseport()`  `BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_load_bytes_relative()`  `Base functions`|
|`BPF_PROG_TYPE_FLOW_DISSECTOR`|`BPF_FUNC_skb_load_bytes()`  `Base functions`|

|Function Group| Functions|
|------------------|-------|
|`Base functions`| `BPF_FUNC_map_lookup_elem()`  `BPF_FUNC_map_update_elem()`  `BPF_FUNC_map_delete_elem()`  `BPF_FUNC_map_peek_elem()`  `BPF_FUNC_map_pop_elem()`  `BPF_FUNC_map_push_elem()`  `BPF_FUNC_get_prandom_u32()`  `BPF_FUNC_get_smp_processor_id()`  `BPF_FUNC_get_numa_node_id()`  `BPF_FUNC_tail_call()`  `BPF_FUNC_ktime_get_boot_ns()`  `BPF_FUNC_ktime_get_ns()`  `BPF_FUNC_trace_printk()`  `BPF_FUNC_spin_lock()`  `BPF_FUNC_spin_unlock()` |
|`Tracing functions`|`BPF_FUNC_map_lookup_elem()`  `BPF_FUNC_map_update_elem()`  `BPF_FUNC_map_delete_elem()`  `BPF_FUNC_probe_read()`  `BPF_FUNC_ktime_get_boot_ns()`  `BPF_FUNC_ktime_get_ns()`  `BPF_FUNC_tail_call()`  `BPF_FUNC_get_current_pid_tgid()`  `BPF_FUNC_get_current_task()`  `BPF_FUNC_get_current_uid_gid()`  `BPF_FUNC_get_current_comm()`  `BPF_FUNC_trace_printk()`  `BPF_FUNC_get_smp_processor_id()`  `BPF_FUNC_get_numa_node_id()`  `BPF_FUNC_perf_event_read()`  `BPF_FUNC_probe_write_user()`  `BPF_FUNC_current_task_under_cgroup()`  `BPF_FUNC_get_prandom_u32()`  `BPF_FUNC_probe_read_str()`  `BPF_FUNC_get_current_cgroup_id()`  `BPF_FUNC_send_signal()`  `BPF_FUNC_probe_read_kernel()`  `BPF_FUNC_probe_read_kernel_str()`  `BPF_FUNC_probe_read_user()`  `BPF_FUNC_probe_read_user_str()`  `BPF_FUNC_send_signal_thread()`  `BPF_FUNC_get_ns_current_pid_tgid()`  `BPF_FUNC_xdp_output()`  `BPF_FUNC_get_task_stack()`|
|`LWT functions`|  `BPF_FUNC_skb_load_bytes()`  `BPF_FUNC_skb_pull_data()`  `BPF_FUNC_csum_diff()`  `BPF_FUNC_get_cgroup_classid()`  `BPF_FUNC_get_route_realm()`  `BPF_FUNC_get_hash_recalc()`  `BPF_FUNC_perf_event_output()`  `BPF_FUNC_get_smp_processor_id()`  `BPF_FUNC_skb_under_cgroup()`|

---

## File: docs/reference_guide.md

# bcc Reference Guide

Intended for search (Ctrl-F) and reference. For tutorials, start with [tutorial.md](tutorial.md).

This guide is incomplete. If something feels missing, check the bcc and kernel source. And if you confirm we're missing something, please send a pull request to fix it, and help out everyone.

## Contents

- [BPF C](#bpf-c)
    - [Events & Arguments](#events--arguments)
        - [1. kprobes](#1-kprobes)
        - [2. kretprobes](#2-kretprobes)
        - [3. Tracepoints](#3-tracepoints)
        - [4. uprobes](#4-uprobes)
        - [5. uretprobes](#5-uretprobes)
        - [6. USDT probes](#6-usdt-probes)
        - [7. Raw Tracepoints](#7-raw-tracepoints)
        - [8. system call tracepoints](#8-system-call-tracepoints)
        - [9. kfuncs](#9-kfuncs)
        - [10. kretfuncs](#10-kretfuncs)
        - [11. lsm probes](#11-lsm-probes)
        - [12. bpf iterators](#12-bpf-iterators)
    - [Data](#data)
        - [1. bpf_probe_read_kernel()](#1-bpf_probe_read_kernel)
        - [2. bpf_probe_read_kernel_str()](#2-bpf_probe_read_kernel_str)
        - [3. bpf_ktime_get_ns()](#3-bpf_ktime_get_ns)
        - [4. bpf_get_current_pid_tgid()](#4-bpf_get_current_pid_tgid)
        - [5. bpf_get_current_uid_gid()](#5-bpf_get_current_uid_gid)
        - [6. bpf_get_current_comm()](#6-bpf_get_current_comm)
        - [7. bpf_get_current_task()](#7-bpf_get_current_task)
        - [8. bpf_log2l()](#8-bpf_log2l)
        - [9. bpf_get_prandom_u32()](#9-bpf_get_prandom_u32)
        - [10. bpf_probe_read_user()](#10-bpf_probe_read_user)
        - [11. bpf_probe_read_user_str()](#11-bpf_probe_read_user_str)
        - [12. bpf_get_ns_current_pid_tgid()](#12-bpf_get_ns_current_pid_tgid)
    - [Debugging](#debugging)
        - [1. bpf_override_return()](#1-bpf_override_return)
    - [Output](#output)
        - [1. bpf_trace_printk()](#1-bpf_trace_printk)
        - [2. BPF_PERF_OUTPUT](#2-bpf_perf_output)
        - [3. perf_submit()](#3-perf_submit)
        - [4. perf_submit_skb()](#4-perf_submit_skb)
        - [5. BPF_RINGBUF_OUTPUT](#5-bpf_ringbuf_output)
        - [6. ringbuf_output()](#6-ringbuf_output)
        - [7. ringbuf_reserve()](#7-ringbuf_reserve)
        - [8. ringbuf_submit()](#8-ringbuf_submit)
        - [9. ringbuf_discard()](#9-ringbuf_discard)
    - [Maps](#maps)
        - [1. BPF_TABLE](#1-bpf_table)
        - [2. BPF_HASH](#2-bpf_hash)
        - [3. BPF_ARRAY](#3-bpf_array)
        - [4. BPF_HISTOGRAM](#4-bpf_histogram)
        - [5. BPF_STACK_TRACE](#5-bpf_stack_trace)
        - [6. BPF_PERF_ARRAY](#6-bpf_perf_array)
        - [7. BPF_PERCPU_HASH](#7-bpf_percpu_hash)
        - [8. BPF_PERCPU_ARRAY](#8-bpf_percpu_array)
        - [9. BPF_LPM_TRIE](#9-bpf_lpm_trie)
        - [10. BPF_PROG_ARRAY](#10-bpf_prog_array)
        - [11. BPF_DEVMAP](#11-bpf_devmap)
        - [12. BPF_CPUMAP](#12-bpf_cpumap)
        - [13. BPF_XSKMAP](#13-bpf_xskmap)
        - [14. BPF_ARRAY_OF_MAPS](#14-bpf_array_of_maps)
        - [15. BPF_HASH_OF_MAPS](#15-bpf_hash_of_maps)
        - [16. BPF_STACK](#16-bpf_stack)
        - [17. BPF_QUEUE](#17-bpf_queue)
        - [18. BPF_SOCKHASH](#18-bpf_sockhash)
        - [19. map.lookup()](#19-maplookup)
        - [20. map.lookup_or_try_init()](#20-maplookup_or_try_init)
        - [21. map.delete()](#21-mapdelete)
        - [22. map.update()](#22-mapupdate)
        - [23. map.insert()](#23-mapinsert)
        - [24. map.increment()](#24-mapincrement)
        - [25. map.get_stackid()](#25-mapget_stackid)
        - [26. map.perf_read()](#26-mapperf_read)
        - [27. map.call()](#27-mapcall)
        - [28. map.redirect_map()](#28-mapredirect_map)
        - [29. map.push()](#29-mappush)
        - [30. map.pop()](#30-mappop)
        - [31. map.peek()](#31-mappeek)
        - [32. map.sock_hash_update()](#32-mapsock_hash_update)
        - [33. map.msg_redirect_hash()](#33-mapmsg_redirect_hash)
        - [34. map.sk_redirect_hash()](#34-mapsk_redirect_hash)
    - [Licensing](#licensing)
    - [Rewriter](#rewriter)

- [bcc Python](#bcc-python)
    - [Initialization](#initialization)
        - [1. BPF](#1-bpf)
        - [2. USDT](#2-usdt)
    - [Events](#events)
        - [1. attach_kprobe()](#1-attach_kprobe)
        - [2. attach_kretprobe()](#2-attach_kretprobe)
        - [3. attach_tracepoint()](#3-attach_tracepoint)
        - [4. attach_uprobe()](#4-attach_uprobe)
        - [5. attach_uretprobe()](#5-attach_uretprobe)
        - [6. USDT.enable_probe()](#6-usdtenable_probe)
        - [7. attach_raw_tracepoint()](#7-attach_raw_tracepoint)
        - [8. attach_raw_socket()](#8-attach_raw_socket)
        - [9. attach_xdp()](#9-attach_xdp)
        - [10. attach_func()](#10-attach_func)
        - [11. detach_func()](#11-detach_func)
        - [12. detach_kprobe()](#12-detach_kprobe)
        - [13. detach_kretprobe()](#13-detach_kretprobe)
    - [Debug Output](#debug-output)
        - [1. trace_print()](#1-trace_print)
        - [2. trace_fields()](#2-trace_fields)
    - [Output APIs](#output-apis)
        - [1. perf_buffer_poll()](#1-perf_buffer_poll)
        - [2. ring_buffer_poll()](#2-ring_buffer_poll)
        - [3. ring_buffer_consume()](#3-ring_buffer_consume)
    - [Map APIs](#map-apis)
        - [1. get_table()](#1-get_table)
        - [2. open_perf_buffer()](#2-open_perf_buffer)
        - [3. items()](#3-items)
        - [4. values()](#4-values)
        - [5. clear()](#5-clear)
        - [6. items_lookup_and_delete_batch()](#6-items_lookup_and_delete_batch)
        - [7. items_lookup_batch()](#7-items_lookup_batch)
        - [8. items_delete_batch()](#8-items_delete_batch)
        - [9. items_update_batch()](#9-items_update_batch)
        - [10. print_log2_hist()](#10-print_log2_hist)
        - [11. print_linear_hist()](#11-print_linear_hist)
        - [12. open_ring_buffer()](#12-open_ring_buffer)
        - [13. push()](#13-push)
        - [14. pop()](#14-pop)
        - [15. peek()](#15-peek)
    - [Helpers](#helpers)
        - [1. ksym()](#1-ksym)
        - [2. ksymname()](#2-ksymname)
        - [3. sym()](#3-sym)
        - [4. num_open_kprobes()](#4-num_open_kprobes)
        - [5. get_syscall_fnname()](#5-get_syscall_fnname)

- [BPF Errors](#bpf-errors)
    - [1. Invalid mem access](#1-invalid-mem-access)
    - [2. Cannot call GPL only function from proprietary program](#2-cannot-call-gpl-only-function-from-proprietary-program)

- [Environment Variables](#Environment-Variables)
    - [1. kernel source directory](#1-kernel-source-directory)
    - [2. kernel version overriding](#2-kernel-version-overriding)

# BPF C

This section describes the C part of a bcc program.

## Events & Arguments

### 1. kprobes

Syntax: kprobe__*kernel_function_name*

```kprobe__``` is a special prefix that creates a kprobe (dynamic tracing of a kernel function call) for the kernel function name provided as the remainder. You can also use kprobes by declaring a normal C function, then using the Python ```BPF.attach_kprobe()``` (covered later) to associate it with a kernel function.

Arguments are specified on the function declaration: kprobe__*kernel_function_name*(struct pt_regs *ctx [, *argument1* ...])

For example:

```C
int kprobe__tcp_v4_connect(struct pt_regs *ctx, struct sock *sk) {
    [...]
}
```

This instruments the tcp_v4_connect() kernel function using a kprobe, with the following arguments:

- ```struct pt_regs *ctx```: Registers and BPF context.
- ```struct sock *sk```: First argument to tcp_v4_connect().

The first argument is always ```struct pt_regs *```, the remainder are the arguments to the function (they don't need to be specified, if you don't intend to use them).

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/examples/tracing/tcpv4connect.py#L28) ([output](https://github.com/iovisor/bcc/blob/5bd0eb21fd148927b078deb8ac29fff2fb044b66/examples/tracing/tcpv4connect_example.txt#L8)),
[code](https://github.com/iovisor/bcc/commit/310ab53710cfd46095c1f6b3e44f1dbc8d1a41d8#diff-8cd1822359ffee26e7469f991ce0ef00R26) ([output](https://github.com/iovisor/bcc/blob/3b9679a3bd9b922c736f6061dc65cb56de7e0250/examples/tracing/bitehist_example.txt#L6))


### 2. kretprobes

Syntax: kretprobe__*kernel_function_name*

```kretprobe__``` is a special prefix that creates a kretprobe (dynamic tracing of a kernel function return) for the kernel function name provided as the remainder. You can also use kretprobes by declaring a normal C function, then using the Python ```BPF.attach_kretprobe()``` (covered later) to associate it with a kernel function.

Return value is available as ```PT_REGS_RC(ctx)```, given a function declaration of: kretprobe__*kernel_function_name*(struct pt_regs *ctx)

For example:

```C
int kretprobe__tcp_v4_connect(struct pt_regs *ctx)
{
    int ret = PT_REGS_RC(ctx);
    [...]
}
```

This instruments the return of the tcp_v4_connect() kernel function using a kretprobe, and stores the return value in ```ret```.

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/examples/tracing/tcpv4connect.py#L38) ([output](https://github.com/iovisor/bcc/blob/5bd0eb21fd148927b078deb8ac29fff2fb044b66/examples/tracing/tcpv4connect_example.txt#L8))

### 3. Tracepoints

Syntax: TRACEPOINT_PROBE(*category*, *event*)

This is a macro that instruments the tracepoint defined by *category*:*event*.

The tracepoint name is `<category>:<event>`.
The probe function name is `tracepoint__<category>__<event>`.

Arguments are available in an ```args``` struct, which are the tracepoint arguments. One way to list these is to cat the relevant format file under /sys/kernel/debug/tracing/events/*category*/*event*/format.

The ```args``` struct can be used in place of ``ctx`` in each functions requiring a context as an argument. This includes notably [perf_submit()](#3-perf_submit).

For example:

```C
TRACEPOINT_PROBE(random, urandom_read) {
    // args is from /sys/kernel/debug/tracing/events/random/urandom_read/format
    bpf_trace_printk("%d\\n", args->got_bits);
    return 0;
}
```

This instruments the tracepoint `random:urandom_read tracepoint`, and prints the tracepoint argument ```got_bits```.
When using Python API, this probe is automatically attached to the right tracepoint target.
For C++, this tracepoint probe can be attached by specifying the tracepoint target and function name explicitly:
`BPF::attach_tracepoint("random:urandom_read", "tracepoint__random__urandom_read")`
Note the name of the probe function defined above is `tracepoint__random__urandom_read`.

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/a4159da8c4ea8a05a3c6e402451f530d6e5a8b41/examples/tracing/urandomread.py#L19) ([output](https://github.com/iovisor/bcc/commit/e422f5e50ecefb96579b6391a2ada7f6367b83c4#diff-41e5ecfae4a3b38de5f4e0887ed160e5R10)),
[search /examples](https://github.com/iovisor/bcc/search?q=TRACEPOINT_PROBE+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=TRACEPOINT_PROBE+path%3Atools&type=Code)

### 4. uprobes

These are instrumented by declaring a normal function in C, then associating it as a uprobe probe in Python via ```BPF.attach_uprobe()``` (covered later).

Arguments can be examined using ```PT_REGS_PARM``` macros.

For example:

```C
int count(struct pt_regs *ctx) {
    char buf[64];
    bpf_probe_read_user(&buf, sizeof(buf), (void *)PT_REGS_PARM1(ctx));
    bpf_trace_printk("%s %d", buf, PT_REGS_PARM2(ctx));
    return(0);
}
```

This reads the first argument as a string, and then prints it with the second argument as an integer.

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/examples/tracing/strlen_count.py#L26)

### 5. uretprobes

These are instrumented by declaring a normal function in C, then associating it as a uretprobe probe in Python via ```BPF.attach_uretprobe()``` (covered later).

Return value is available as ```PT_REGS_RC(ctx)```, given a function declaration of: *function_name*(struct pt_regs *ctx)

For example:

```C
BPF_HISTOGRAM(dist);
int count(struct pt_regs *ctx) {
    dist.increment(PT_REGS_RC(ctx));
    return 0;
}
```

This increments the bucket in the ```dist``` histogram that is indexed by the return value.

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/examples/tracing/strlen_hist.py#L39) ([output](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/examples/tracing/strlen_hist.py#L15)),
[code](https://github.com/iovisor/bcc/blob/4afa96a71c5dbfc4c507c3355e20baa6c184a3a8/tools/bashreadline.py) ([output](https://github.com/iovisor/bcc/commit/aa87997d21e5c1a6a20e2c96dd25eb92adc8e85d#diff-2fd162f9e594206f789246ce97d62cf0R7))

### 6. USDT probes

These are User Statically-Defined Tracing (USDT) probes, which may be placed in some applications or libraries to provide a user-level equivalent of tracepoints. The primary BPF method provided for USDT support method is ```enable_probe()```. USDT probes are instrumented by declaring a normal function in C, then associating it as a USDT probe in Python via ```USDT.enable_probe()```.

Arguments can be read via: bpf_usdt_readarg(*index*, ctx, &addr)

For example:

```C
int do_trace(struct pt_regs *ctx) {
    uint64_t addr;
    char path[128];
    bpf_usdt_readarg(6, ctx, &addr);
    bpf_probe_read_user(&path, sizeof(path), (void *)addr);
    bpf_trace_printk("path:%s\\n", path);
    return 0;
};
```

This reads the sixth USDT argument, and then pulls it in as a string to ```path```.

When initializing USDTs via the third argument of ```BPF::init``` in the C API, if any USDT fails to ```init```, entire ```BPF::init``` will fail. If you're OK with some USDTs failing to ```init```, use ```BPF::init_usdt``` before calling ```BPF::init```.

Examples in situ:
[code](https://github.com/iovisor/bcc/commit/4f88a9401357d7b75e917abd994aa6ea97dda4d3#diff-04a7cad583be5646080970344c48c1f4R24),
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_usdt_readarg+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_usdt_readarg+path%3Atools&type=Code)

### 7. Raw Tracepoints

Syntax: RAW_TRACEPOINT_PROBE(*event*)

This is a macro that instruments the raw tracepoint defined by *event*.

The argument is a pointer to struct ```bpf_raw_tracepoint_args```, which is defined in [bpf.h](https://github.com/iovisor/bcc/blob/master/src/cc/compat/linux/virtual_bpf.h).  The struct field ```args``` contains all parameters of the raw tracepoint where you can found at linux tree [include/trace/events](https://github.com/torvalds/linux/tree/master/include/trace/events)
directory.

For example:
```C
RAW_TRACEPOINT_PROBE(sched_switch)
{
    // TP_PROTO(bool preempt, struct task_struct *prev, struct task_struct *next)
    struct task_struct *prev = (struct task_struct *)ctx->args[1];
    struct task_struct *next= (struct task_struct *)ctx->args[2];
    s32 prev_tgid, next_tgid;

    bpf_probe_read_kernel(&prev_tgid, sizeof(prev->tgid), &prev->tgid);
    bpf_probe_read_kernel(&next_tgid, sizeof(next->tgid), &next->tgid);
    bpf_trace_printk("%d -> %d\\n", prev_tgid, next_tgid);
}
```

This instruments the sched:sched_switch tracepoint, and prints the prev and next tgid.

Examples in situ:
[search /tools](https://github.com/iovisor/bcc/search?q=RAW_TRACEPOINT_PROBE+path%3Atools&type=Code)

### 8. system call tracepoints

Syntax: ```syscall__SYSCALLNAME```

```syscall__``` is a special prefix that creates a kprobe for the system call name provided as the remainder. You can use it by declaring a normal C function, then using the Python ```BPF.get_syscall_fnname(SYSCALLNAME)``` and ```BPF.attach_kprobe()``` to associate it.

Arguments are specified on the function declaration: ```syscall__SYSCALLNAME(struct pt_regs *ctx, [, argument1 ...])```.

For example:
```C
int syscall__execve(struct pt_regs *ctx,
    const char __user *filename,
    const char __user *const __user *__argv,
    const char __user *const __user *__envp)
{
    [...]
}
```

This instruments the execve system call.

The first argument is always ```struct pt_regs *```, the remainder are the arguments to the function (they don't need to be specified, if you don't intend to use them).

Corresponding Python code:
```Python
b = BPF(text=bpf_text)
execve_fnname = b.get_syscall_fnname("execve")
b.attach_kprobe(event=execve_fnname, fn_name="syscall__execve")
```

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/552658edda09298afdccc8a4b5e17311a2d8a771/tools/execsnoop.py#L101) ([output](https://github.com/iovisor/bcc/blob/552658edda09298afdccc8a4b5e17311a2d8a771/tools/execsnoop_example.txt#L8))

### 9. kfuncs

Syntax: KFUNC_PROBE(*function*, typeof(arg1) arg1, typeof(arg2) arge ...)
        MODULE_KFUNC_PROBE(*module*, *function*, typeof(arg1) arg1, typeof(arg2) arge ...)

This is a macro that instruments the kernel function via trampoline
*before* the function is executed. It's defined by *function* name and
the function arguments defined as *argX*.

For example:
```C
KFUNC_PROBE(do_sys_open, int dfd, const char *filename, int flags, int mode)
{
    ...
```

This instruments the do_sys_open kernel function and make its arguments
accessible as standard argument values.

Examples in situ:
[search /tools](https://github.com/iovisor/bcc/search?q=KFUNC_PROBE+path%3Atools&type=Code)

### 10. kretfuncs

Syntax: KRETFUNC_PROBE(*event*, typeof(arg1) arg1, typeof(arg2) arge ..., int ret)
        MODULE_KRETFUNC_PROBE(*module*, *function*, typeof(arg1) arg1, typeof(arg2) arge ...)

This is a macro that instruments the kernel function via trampoline
*after* the function is executed. It's defined by *function* name and
the function arguments defined as *argX*.

The last argument of the probe is the return value of the instrumented function.

For example:
```C
KRETFUNC_PROBE(do_sys_open, int dfd, const char *filename, int flags, int mode, int ret)
{
    ...
```

This instruments the do_sys_open kernel function and make its arguments
accessible as standard argument values together with its return value.

Examples in situ:
[search /tools](https://github.com/iovisor/bcc/search?q=KRETFUNC_PROBE+path%3Atools&type=Code)


### 11. LSM Probes

Syntax: LSM_PROBE(*hook*, typeof(arg1) arg1, typeof(arg2) arg2 ...)

This is a macro that instruments an LSM hook as a BPF program. It can be
used to audit security events and implement MAC security policies in BPF.
It is defined by specifying the hook name followed by its arguments.

Hook names can be found in
[include/linux/security.h](https://github.com/torvalds/linux/blob/v5.15/include/linux/security.h#L260)
by taking functions like `security_hookname` and taking just the `hookname` part.
For example, `security_bpf` would simply become `bpf`.

Unlike other BPF program types, the return value specified in an LSM probe
matters. A return value of 0 allows the hook to succeed, whereas
any non-zero return value will cause the hook to fail and deny the
security operation.

The following example instruments a hook that denies all future BPF operations:
```C
LSM_PROBE(bpf, int cmd, union bpf_attr *attr, unsigned int size)
{
    return -EPERM;
}
```

This instruments the `security_bpf` hook and causes it to return `-EPERM`.
Changing `return -EPERM` to `return 0` would cause the BPF program
to allow the operation instead.

LSM probes require at least a 5.7+ kernel with the following configuation options set:
- `CONFIG_BPF_LSM=y`
- `CONFIG_LSM` comma separated string must contain "bpf" (for example,
  `CONFIG_LSM="lockdown,yama,bpf"`)

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=LSM_PROBE+path%3Atests&type=Code)

### 12. BPF ITERATORS

Syntax: BPF_ITER(target)

This is a macro to define a program signature for a bpf iterator program. The argument *target* specifies what to iterate for the program.

Currently, kernel does not have interface to discover what targets are supported. A good place to find what is supported is in [tools/testing/selftests/bpf/prog_test/bpf_iter.c](https://github.com/torvalds/linux/blob/master/tools/testing/selftests/bpf/prog_tests/bpf_iter.c) and some sample bpf iter programs are in [tools/testing/selftests/bpf/progs](https://github.com/torvalds/linux/tree/master/tools/testing/selftests/bpf/progs) with file name prefix *bpf_iter*.

The following example defines a program for target *task*, which traverses all tasks in the kernel.
```C
BPF_ITER(task)
{
  struct seq_file *seq = ctx->meta->seq;
  struct task_struct *task = ctx->task;

  if (task == (void *)0)
    return 0;

  ... task->pid, task->tgid, task->comm, ...
  return 0;
}
```

BPF iterators are introduced in 5.8 kernel for task, task_file, bpf_map, netlink_sock and ipv6_route . In 5.9, support is added to tcp/udp sockets and bpf map element (hashmap, arraymap and sk_local_storage_map) traversal.

## Data

### 1. bpf_probe_read_kernel()

Syntax: ```int bpf_probe_read_kernel(void *dst, int size, const void *src)```

Return: 0 on success

This copies size bytes from kernel address space to the BPF stack, so that BPF can later operate on it. For safety, all kernel memory reads must pass through bpf_probe_read_kernel(). This happens automatically in some cases, such as dereferencing kernel variables, as bcc will rewrite the BPF program to include the necessary bpf_probe_read_kernel().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_probe_read_kernel+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_probe_read_kernel+path%3Atools&type=Code)

### 2. bpf_probe_read_kernel_str()

Syntax: ```int bpf_probe_read_kernel_str(void *dst, int size, const void *src)```

Return:
  - \> 0 length of the string including the trailing NULL on success
  - \< 0 error

This copies a `NULL` terminated string from kernel address space to the BPF stack, so that BPF can later operate on it. In case the string length is smaller than size, the target is not padded with further `NULL` bytes. In case the string length is larger than size, just `size - 1` bytes are copied and the last byte is set to `NULL`.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_probe_read_kernel_str+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_probe_read_kernel_str+path%3Atools&type=Code)

### 3. bpf_ktime_get_ns()

Syntax: ```u64 bpf_ktime_get_ns(void)```

Return: u64 number of nanoseconds. Starts at system boot time but stops during suspend.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_ktime_get_ns+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_ktime_get_ns+path%3Atools&type=Code)

### 4. bpf_get_current_pid_tgid()

Syntax: ```u64 bpf_get_current_pid_tgid(void)```

Return: ```current->tgid << 32 | current->pid```

Returns the process ID in the lower 32 bits (kernel's view of the PID, which in user space is usually presented as the thread ID), and the thread group ID in the upper 32 bits (what user space often thinks of as the PID). By directly setting this to a u32, we discard the upper 32 bits.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_current_pid_tgid+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_current_pid_tgid+path%3Atools&type=Code)

### 5. bpf_get_current_uid_gid()

Syntax: ```u64 bpf_get_current_uid_gid(void)```

Return: ```current_gid << 32 | current_uid```

Returns the user ID and group IDs.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_current_uid_gid+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_current_uid_gid+path%3Atools&type=Code)

### 6. bpf_get_current_comm()

Syntax: ```bpf_get_current_comm(char *buf, int size_of_buf)```

Return: 0 on success

Populates the first argument address with the current process name. It should be a pointer to a char array of at least size TASK_COMM_LEN, which is defined in linux/sched.h. For example:

```C
#include <linux/sched.h>

int do_trace(struct pt_regs *ctx) {
    char comm[TASK_COMM_LEN];
    bpf_get_current_comm(&comm, sizeof(comm));
[...]
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_current_comm+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_current_comm+path%3Atools&type=Code)

### 7. bpf_get_current_task()

Syntax: ```bpf_get_current_task()```

Return: current task as a pointer to struct task_struct.

Returns a pointer to the current task's task_struct object. This helper can be used to compute the on-CPU time for a process, identify kernel threads, get the current CPU's run queue, or retrieve many other pieces of information.

With Linux 4.13, due to issues with field randomization, you may need two #define directives before the includes:
```C
#define randomized_struct_fields_start  struct {
#define randomized_struct_fields_end    };
#include <linux/sched.h>

int do_trace(void *ctx) {
    struct task_struct *t = (struct task_struct *)bpf_get_current_task();
[...]
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_current_task+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_current_task+path%3Atools&type=Code)

### 8. bpf_log2l()

Syntax: ```unsigned int bpf_log2l(unsigned long v)```

Returns the log-2 of the provided value. This is often used to create indexes for histograms, to construct power-of-2 histograms.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_log2l+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_log2l+path%3Atools&type=Code)

### 9. bpf_get_prandom_u32()

Syntax: ```u32 bpf_get_prandom_u32()```

Returns a pseudo-random u32.

Example in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_prandom_u32+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_prandom_u32+path%3Atools&type=Code)

### 10. bpf_probe_read_user()

Syntax: ```int bpf_probe_read_user(void *dst, int size, const void *src)```

Return: 0 on success

This attempts to safely read size bytes from user address space to the BPF stack, so that BPF can later operate on it. For safety, all user address space memory reads must pass through bpf_probe_read_user().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_probe_read_user+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_probe_read_user+path%3Atools&type=Code)

### 11. bpf_probe_read_user_str()

Syntax: ```int bpf_probe_read_user_str(void *dst, int size, const void *src)```

Return:
  - \> 0 length of the string including the trailing NULL on success
  - \< 0 error

This copies a `NULL` terminated string from user address space to the BPF stack, so that BPF can later operate on it. In case the string length is smaller than size, the target is not padded with further `NULL` bytes. In case the string length is larger than size, just `size - 1` bytes are copied and the last byte is set to `NULL`.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_probe_read_user_str+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_probe_read_user_str+path%3Atools&type=Code)


### 12. bpf_get_ns_current_pid_tgid()

Syntax: ```u32 bpf_get_ns_current_pid_tgid(u64 dev, u64 ino, struct bpf_pidns_info* nsdata, u32 size)```

Values for *pid* and *tgid* as seen from the current *namespace* will be returned in *nsdata*.

Return 0 on success, or one of the following in case of failure:

- **-EINVAL** if dev and inum supplied don't match dev_t and inode number with nsfs of current task, or if dev conversion to dev_t lost high bits.

- **-ENOENT** if pidns does not exists for the current task.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_get_ns_current_pid_tgid+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_get_ns_current_pid_tgid+path%3Atools&type=Code)


## Debugging

### 1. bpf_override_return()

Syntax: ```int bpf_override_return(struct pt_regs *, unsigned long rc)```

Return: 0 on success

When used in a program attached to a function entry kprobe, causes the
execution of the function to be skipped, immediately returning `rc` instead.
This is used for targeted error injection.

bpf_override_return will only work when the kprobed function is whitelisted to
allow error injections. Whitelisting entails tagging a function with
`ALLOW_ERROR_INJECTION()` in the kernel source tree; see `io_ctl_init` for
an example. If the kprobed function is not whitelisted, the bpf program will
fail to attach with ` ioctl(PERF_EVENT_IOC_SET_BPF): Invalid argument`


```C
int kprobe__io_ctl_init(void *ctx) {
	bpf_override_return(ctx, -ENOMEM);
	return 0;
}
```

## Output

### 1. bpf_trace_printk()

Syntax: ```int bpf_trace_printk(const char *fmt, ...)```

Return: 0 on success

A simple kernel facility for printf() to the common trace_pipe (/sys/kernel/debug/tracing/trace_pipe). This is ok for some quick examples, but has limitations: 3 args max, 1 %s only, and trace_pipe is globally shared, so concurrent programs will have clashing output. A better interface is via BPF_PERF_OUTPUT(). Note that calling this helper is made simpler than the original kernel version, which has ```fmt_size``` as the second parameter.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=bpf_trace_printk+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=bpf_trace_printk+path%3Atools&type=Code)

### 2. BPF_PERF_OUTPUT

Syntax: ```BPF_PERF_OUTPUT(name)```

Creates a BPF table for pushing out custom event data to user space via a perf ring buffer. This is the preferred method for pushing per-event data to user space.

For example:

```C
struct data_t {
    u32 pid;
    u64 ts;
    char comm[TASK_COMM_LEN];
};
BPF_PERF_OUTPUT(events);

int hello(struct pt_regs *ctx) {
    struct data_t data = {};

    data.pid = bpf_get_current_pid_tgid();
    data.ts = bpf_ktime_get_ns();
    bpf_get_current_comm(&data.comm, sizeof(data.comm));

    events.perf_submit(ctx, &data, sizeof(data));

    return 0;
}
```

The output table is named ```events```, and data is pushed to it via ```events.perf_submit()```.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_PERF_OUTPUT+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_PERF_OUTPUT+path%3Atools&type=Code)

### 3. perf_submit()

Syntax: ```int perf_submit((void *)ctx, (void *)data, u32 data_size)```

Return: 0 on success

A method of a BPF_PERF_OUTPUT table, for submitting custom event data to user space. See the BPF_PERF_OUTPUT entry. (This ultimately calls bpf_perf_event_output().)

The ```ctx``` parameter is provided in [kprobes](#1-kprobes) or [kretprobes](#2-kretprobes). For ```SCHED_CLS``` or ```SOCKET_FILTER``` programs, the ```struct __sk_buff *skb``` must be used instead.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=perf_submit+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=perf_submit+path%3Atools&type=Code)

### 4. perf_submit_skb()

Syntax: ```int perf_submit_skb((void *)ctx, u32 packet_size, (void *)data, u32 data_size)```

Return: 0 on success

A method of a BPF_PERF_OUTPUT table available in networking program types, for submitting custom event data to user space, along with the first ```packet_size``` bytes of the packet buffer. See the BPF_PERF_OUTPUT entry. (This ultimately calls bpf_perf_event_output().)

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=perf_submit_skb+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=perf_submit_skb+path%3Atools&type=Code)

### 5. BPF_RINGBUF_OUTPUT

Syntax: ```BPF_RINGBUF_OUTPUT(name, page_cnt)```

Creates a BPF table for pushing out custom event data to user space via a ringbuf ring buffer.
```BPF_RINGBUF_OUTPUT``` has several advantages over ```BPF_PERF_OUTPUT```, summarized as follows:

- Buffer is shared across all CPUs, meaning no per-CPU allocation
- Supports two APIs for BPF programs
    - ```map.ringbuf_output()``` works like ```map.perf_submit()``` (covered in [ringbuf_output](#6-ringbuf_output))
    - ```map.ringbuf_reserve()```/```map.ringbuf_submit()```/```map.ringbuf_discard()```
      split the process of reserving buffer space and submitting events into two steps
      (covered in [ringbuf_reserve](#7-ringbuf_reserve), [ringbuf_submit](#8-ringbuf_submit), [ringbuf_discard](#9-ringbuf_discard))
- BPF APIs do not require access to a CPU ctx argument
- Superior performance and latency in userspace thanks to a shared ring buffer manager
- Supports two ways of consuming data in userspace

Starting in Linux 5.8, this should be the preferred method for pushing per-event data to user space.

Example of both APIs:

```C
struct data_t {
    u32 pid;
    u64 ts;
    char comm[TASK_COMM_LEN];
};

// Creates a ringbuf called events with 8 pages of space, shared across all CPUs
BPF_RINGBUF_OUTPUT(events, 8);

int first_api_example(struct pt_regs *ctx) {
    struct data_t data = {};

    data.pid = bpf_get_current_pid_tgid();
    data.ts = bpf_ktime_get_ns();
    bpf_get_current_comm(&data.comm, sizeof(data.comm));

    events.ringbuf_output(&data, sizeof(data), 0 /* flags */);

    return 0;
}

int second_api_example(struct pt_regs *ctx) {
    struct data_t *data = events.ringbuf_reserve(sizeof(struct data_t));
    if (!data) { // Failed to reserve space
        return 1;
    }

    data->pid = bpf_get_current_pid_tgid();
    data->ts = bpf_ktime_get_ns();
    bpf_get_current_comm(&data->comm, sizeof(data->comm));

    events.ringbuf_submit(data, 0 /* flags */);

    return 0;
}
```

The output table is named ```events```. Data is allocated via ```events.ringbuf_reserve()``` and pushed to it via ```events.ringbuf_submit()```.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_RINGBUF_OUTPUT+path%3Aexamples&type=Code),

### 6. ringbuf_output()

Syntax: ```int ringbuf_output((void *)data, u64 data_size, u64 flags)```

Return: 0 on success

Flags:
 - ```BPF_RB_NO_WAKEUP```: Do not sent notification of new data availability
 - ```BPF_RB_FORCE_WAKEUP```: Send notification of new data availability unconditionally

A method of the BPF_RINGBUF_OUTPUT table, for submitting custom event data to user space. This method works like ```perf_submit()```,
although it does not require a ctx argument.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=ringbuf_output+path%3Aexamples&type=Code),

### 7. ringbuf_reserve()

Syntax: ```void* ringbuf_reserve(u64 data_size)```

Return: Pointer to data struct on success, NULL on failure

A method of the BPF_RINGBUF_OUTPUT table, for reserving space in the ring buffer and simultaenously
allocating a data struct for output. Must be used with one of ```ringbuf_submit``` or ```ringbuf_discard```.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=ringbuf_reserve+path%3Aexamples&type=Code),

### 8. ringbuf_submit()

Syntax: ```void ringbuf_submit((void *)data, u64 flags)```

Return: Nothing, always succeeds

Flags:
 - ```BPF_RB_NO_WAKEUP```: Do not sent notification of new data availability
 - ```BPF_RB_FORCE_WAKEUP```: Send notification of new data availability unconditionally

A method of the BPF_RINGBUF_OUTPUT table, for submitting custom event data to user space. Must be preceded by a call to
```ringbuf_reserve()``` to reserve space for the data.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=ringbuf_submit+path%3Aexamples&type=Code),

### 9. ringbuf_discard()

Syntax: ```void ringbuf_discard((void *)data, u64 flags)```

Return: Nothing, always succeeds

Flags:
 - ```BPF_RB_NO_WAKEUP```: Do not sent notification of new data availability
 - ```BPF_RB_FORCE_WAKEUP```: Send notification of new data availability unconditionally

A method of the BPF_RINGBUF_OUTPUT table, for discarding custom event data; userspace
ignores the data associated with the discarded event. Must be preceded by a call to
```ringbuf_reserve()``` to reserve space for the data.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=ringbuf_submit+path%3Aexamples&type=Code),

### 10. ringbuf_query()

Syntax: ```u64 ringbuf_query(u64 flags)```

Return: Requested value, or 0, if flags are not recognized

Flags:
 - ```BPF_RB_AVAIL_DATA```: Amount of data not yet consumed
 - ```BPF_RB_RING_SIZE```: The size of ring buffer
 - ```BPF_RB_CONS_POS```: Consumer position
 - ```BPF_RB_PROD_POS```: Producer(s) position

A method of the BPF_RINGBUF_OUTPUT table, for getting various properties of ring buffer. Returned values are momentarily snapshots of ring buffer state and could be off by the time helper returns, so this should be used only for debugging/reporting reasons or for implementing various heuristics, that take into account highly-changeable nature of some of those characteristics.

Examples in situ: 
[search /examples](https://github.com/iovisor/bcc/search?q=ringbuf_query+path%3Aexamples&type=Code),

## Maps

Maps are BPF data stores, and are the basis for higher level object types including tables, hashes, and histograms.

### 1. BPF_TABLE

Syntax: ```BPF_TABLE(_table_type, _key_type, _leaf_type, _name, _max_entries)```

Creates a map named ```_name```. Most of the time this will be used via higher-level macros, like BPF_HASH, BPF_ARRAY, BPF_HISTOGRAM, etc.

`BPF_F_TABLE` is a variant that takes a flag in the last parameter. `BPF_TABLE(...)` is actually a wrapper to `BPF_F_TABLE(..., 0 /* flag */)`.

Methods (covered later): map.lookup(), map.lookup_or_try_init(), map.delete(), map.update(), map.insert(), map.increment().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_TABLE+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_TABLE+path%3Atools&type=Code)

#### Pinned Maps

Syntax: ```BPF_TABLE_PINNED(_table_type, _key_type, _leaf_type, _name, _max_entries, "/sys/fs/bpf/xyz")```

Create a new map if it doesn't exist and pin it to the bpffs as a FILE, otherwise use the map that was pinned to the bpffs. The type information is not enforced and the actual map type depends on the map that got pinned to the location.

For example:

```C
BPF_TABLE_PINNED("hash", u64, u64, ids, 1024, "/sys/fs/bpf/ids");
```

### 2. BPF_HASH

Syntax: ```BPF_HASH(name [, key_type [, leaf_type [, size]]])```

Creates a hash map (associative array) named ```name```, with optional parameters.

Defaults: ```BPF_HASH(name, key_type=u64, leaf_type=u64, size=10240)```

For example:

```C
BPF_HASH(start, struct request *);
```

This creates a hash named ```start``` where the key is a ```struct request *```, and the value defaults to u64. This hash is used by the disksnoop.py example for saving timestamps for each I/O request, where the key is the pointer to struct request, and the value is the timestamp.

This is a wrapper macro for `BPF_TABLE("hash", ...)`.

Methods (covered later): map.lookup(), map.lookup_or_try_init(), map.delete(), map.update(), map.insert(), map.increment().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_HASH+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_HASH+path%3Atools&type=Code)

### 3. BPF_ARRAY

Syntax: ```BPF_ARRAY(name [, leaf_type [, size]])```

Creates an int-indexed array which is optimized for fastest lookup and update, named ```name```, with optional parameters.

Defaults: ```BPF_ARRAY(name, leaf_type=u64, size=10240)```

For example:

```C
BPF_ARRAY(counts, u64, 32);
```

This creates an array named ```counts``` where with 32 buckets and 64-bit integer values. This array is used by the funccount.py example for saving call count of each function.

This is a wrapper macro for `BPF_TABLE("array", ...)`.

Methods (covered later): map.lookup(), map.update(), map.increment(). Note that all array elements are pre-allocated with zero values and can not be deleted.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_ARRAY+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_ARRAY+path%3Atools&type=Code)

### 4. BPF_HISTOGRAM

Syntax: ```BPF_HISTOGRAM(name [, key_type [, size ]])```

Creates a histogram map named ```name```, with optional parameters.

Defaults: ```BPF_HISTOGRAM(name, key_type=int, size=64)```

For example:

```C
BPF_HISTOGRAM(dist);
```

This creates a histogram named ```dist```, which defaults to 64 buckets indexed by keys of type int.

This is a wrapper macro for `BPF_TABLE("histgram", ...)`.

Methods (covered later): map.increment().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_HISTOGRAM+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_HISTOGRAM+path%3Atools&type=Code)

### 5. BPF_STACK_TRACE

Syntax: ```BPF_STACK_TRACE(name, max_entries)```

Creates stack trace map named ```name```, with a maximum entry count provided. These maps are used to store stack traces.

For example:

```C
BPF_STACK_TRACE(stack_traces, 1024);
```

This creates stack trace map named ```stack_traces```, with a maximum number of stack trace entries of 1024.

This is a wrapper macro for `BPF_TABLE("stacktrace", ...)`.

Methods (covered later): map.get_stackid().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_STACK_TRACE+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_STACK_TRACE+path%3Atools&type=Code)

### 6. BPF_PERF_ARRAY

Syntax: ```BPF_PERF_ARRAY(name, max_entries)```

Creates perf array named ```name```, with a maximum entry count provided, which must be equal to the number of system cpus. These maps are used to fetch hardware performance counters.

For example:

```C
text="""
BPF_PERF_ARRAY(cpu_cycles, NUM_CPUS);
"""
b = bcc.BPF(text=text, cflags=["-DNUM_CPUS=%d" % multiprocessing.cpu_count()])
b["cpu_cycles"].open_perf_event(b["cpu_cycles"].HW_CPU_CYCLES)
```

This creates a perf array named ```cpu_cycles```, with number of entries equal to the number of cpus/cores. The array is configured so that later calling map.perf_read() will return a hardware-calculated counter of the number of cycles elapsed from some point in the past. Only one type of hardware counter may be configured per table at a time.

Methods (covered later): map.perf_read().

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=BPF_PERF_ARRAY+path%3Atests&type=Code)

### 7. BPF_PERCPU_HASH

Syntax: ```BPF_PERCPU_HASH(name [, key_type [, leaf_type [, size]]])```

Creates NUM_CPU int-indexed hash maps (associative arrays) named ```name```, with optional parameters. Each CPU will have a separate copy of this array. The copies are not kept synchronized in any way.

Note that due to limits defined in the kernel (in linux/mm/percpu.c), the ```leaf_type``` cannot have a size of more than 32KB.
In other words, ```BPF_PERCPU_HASH``` elements cannot be larger than 32KB in size.


Defaults: ```BPF_PERCPU_HASH(name, key_type=u64, leaf_type=u64, size=10240)```

For example:

```C
BPF_PERCPU_HASH(start, struct request *);
```

This creates NUM_CPU hashes named ```start``` where the key is a ```struct request *```, and the value defaults to u64.

This is a wrapper macro for `BPF_TABLE("percpu_hash", ...)`.

Methods (covered later): map.lookup(), map.lookup_or_try_init(), map.delete(), map.update(), map.insert(), map.increment().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_PERCPU_HASH+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_PERCPU_HASH+path%3Atools&type=Code)


### 8. BPF_PERCPU_ARRAY

Syntax: ```BPF_PERCPU_ARRAY(name [, leaf_type [, size]])```

Creates NUM_CPU int-indexed arrays which are optimized for fastest lookup and update, named ```name```, with optional parameters. Each CPU will have a separate copy of this array. The copies are not kept synchronized in any way.

Note that due to limits defined in the kernel (in linux/mm/percpu.c), the ```leaf_type``` cannot have a size of more than 32KB.
In other words, ```BPF_PERCPU_ARRAY``` elements cannot be larger than 32KB in size.


Defaults: ```BPF_PERCPU_ARRAY(name, leaf_type=u64, size=10240)```

For example:

```C
BPF_PERCPU_ARRAY(counts, u64, 32);
```

This creates NUM_CPU arrays named ```counts``` where with 32 buckets and 64-bit integer values.

This is a wrapper macro for `BPF_TABLE("percpu_array", ...)`.

Methods (covered later): map.lookup(), map.update(), map.increment(). Note that all array elements are pre-allocated with zero values and can not be deleted.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_PERCPU_ARRAY+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_PERCPU_ARRAY+path%3Atools&type=Code)

### 9. BPF_LPM_TRIE

Syntax: `BPF_LPM_TRIE(name [, key_type [, leaf_type [, size]]])`

Creates a longest prefix match trie map named `name`, with optional parameters.

Defaults: `BPF_LPM_TRIE(name, key_type=u64, leaf_type=u64, size=10240)`

For example:

```c
BPF_LPM_TRIE(trie, struct key_v6);
```

This creates an LPM trie map named `trie` where the key is a `struct key_v6`, and the value defaults to u64.

This is a wrapper macro to `BPF_F_TABLE("lpm_trie", ..., BPF_F_NO_PREALLOC)`.

Methods (covered later): map.lookup(), map.lookup_or_try_init(), map.delete(), map.update(), map.insert(), map.increment().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_LPM_TRIE+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF_LPM_TRIE+path%3Atools&type=Code)

### 10. BPF_PROG_ARRAY

Syntax: ```BPF_PROG_ARRAY(name, size)```

This creates a program array named ```name``` with ```size``` entries. Each entry of the array is either a file descriptor to a bpf program or ```NULL```. The array acts as a jump table so that bpf programs can "tail-call" other bpf programs.

This is a wrapper macro for `BPF_TABLE("prog", ...)`.

Methods (covered later): map.call().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_PROG_ARRAY+path%3Aexamples&type=Code),
[search /tests](https://github.com/iovisor/bcc/search?q=BPF_PROG_ARRAY+path%3Atests&type=Code),
[assign fd](https://github.com/iovisor/bcc/blob/master/examples/networking/tunnel_monitor/monitor.py#L24-L26)

### 11. BPF_DEVMAP

Syntax: ```BPF_DEVMAP(name, size)```

This creates a device map named ```name``` with ```size``` entries. Each entry of the map is an `ifindex` to a network interface. This map is only used in XDP.

For example:
```C
BPF_DEVMAP(devmap, 10);
```

Methods (covered later): map.redirect_map().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_DEVMAP+path%3Aexamples&type=Code),

### 12. BPF_CPUMAP

Syntax: ```BPF_CPUMAP(name, size)```

This creates a cpu map named ```name``` with ```size``` entries. The index of the map represents the CPU id and each entry is the size of the ring buffer allocated for the CPU. This map is only used in XDP.

For example:
```C
BPF_CPUMAP(cpumap, 16);
```

Methods (covered later): map.redirect_map().

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_CPUMAP+path%3Aexamples&type=Code),

### 13. BPF_XSKMAP

Syntax: ```BPF_XSKMAP(name, size [, "/sys/fs/bpf/xyz"])```

This creates a xsk map named ```name``` with ```size``` entries and pin it to the bpffs as a FILE. Each entry represents one NIC's queue id. This map is only used in XDP to redirect packet to an AF_XDP socket. If the AF_XDP socket is binded to a queue which is different than the current packet's queue id, the packet will be dropped. For kernel v5.3 and latter, `lookup` method is available and can be used to check whether and AF_XDP socket is available for the current packet's queue id. More details at [AF_XDP](https://www.kernel.org/doc/html/latest/networking/af_xdp.html).

For example:
```C
BPF_XSKMAP(xsks_map, 8);
```

Methods (covered later): map.redirect_map(). map.lookup()

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF_XSKMAP+path%3Aexamples&type=Code),

### 14. BPF_ARRAY_OF_MAPS

Syntax: ```BPF_ARRAY_OF_MAPS(name, inner_map_name, size)```

This creates an array map with a map-in-map type (BPF_MAP_TYPE_HASH_OF_MAPS) map named ```name``` with ```size``` entries. The inner map meta data is provided by map ```inner_map_name``` and can be most of array or hash maps except ```BPF_MAP_TYPE_PROG_ARRAY```, ```BPF_MAP_TYPE_CGROUP_STORAGE``` and ```BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE```.

For example:
```C
BPF_TABLE("hash", int, int, ex1, 1024);
BPF_TABLE("hash", int, int, ex2, 1024);
BPF_ARRAY_OF_MAPS(maps_array, "ex1", 10);
```

### 15. BPF_HASH_OF_MAPS

Syntax: ```BPF_HASH_OF_MAPS(name, key_type, inner_map_name, size)```

This creates a hash map with a map-in-map type (BPF_MAP_TYPE_HASH_OF_MAPS) map named ```name``` with ```size``` entries. The inner map meta data is provided by map ```inner_map_name``` and can be most of array or hash maps except ```BPF_MAP_TYPE_PROG_ARRAY```, ```BPF_MAP_TYPE_CGROUP_STORAGE``` and ```BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE```.

For example:
```C
BPF_ARRAY(ex1, int, 1024);
BPF_ARRAY(ex2, int, 1024);
BPF_HASH_OF_MAPS(maps_hash, struct custom_key, "ex1", 10);
```

### 16. BPF_STACK

Syntax: ```BPF_STACK(name, leaf_type, max_entries[, flags])```

Creates a stack named ```name``` with value type ```leaf_type``` and max entries ```max_entries```.
Stack and Queue maps are only available from Linux 4.20+.

For example:

```C
BPF_STACK(stack, struct event, 10240);
```

This creates a stack named ```stack``` where the value type is ```struct event```, that holds up to 10240 entries.

Methods (covered later): map.push(), map.pop(), map.peek().

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=BPF_STACK+path%3Atests&type=Code),

### 17. BPF_QUEUE

Syntax: ```BPF_QUEUE(name, leaf_type, max_entries[, flags])```

Creates a queue named ```name``` with value type ```leaf_type``` and max entries ```max_entries```.
Stack and Queue maps are only available from Linux 4.20+.

For example:

```C
BPF_QUEUE(queue, struct event, 10240);
```

This creates a queue named ```queue``` where the value type is ```struct event```, that holds up to 10240 entries.

Methods (covered later): map.push(), map.pop(), map.peek().

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=BPF_QUEUE+path%3Atests&type=Code),

### 18. BPF_SOCKHASH

Syntax: ```BPF_SOCKHASH(name[, key_type [, max_entries)```

Creates a hash named ```name```, with optional parameters. sockhash is only available from Linux 4.18+.

Default: ```BPF_SOCKHASH(name, key_type=u32, max_entries=10240)```

For example:

```C
struct sock_key {
  u32 remote_ip4;
  u32 local_ip4;
  u32 remote_port;
  u32 local_port;
};
BPF_HASH(skh, struct sock_key, 65535);
```

This creates a hash named ```skh``` where the key is a ```struct sock_key```.

A sockhash is a BPF map type that holds references to sock structs. Then with a new sk/msg redirect bpf helper BPF programs can use the map to redirect skbs/msgs between sockets (```map.sk_redirect_hash()/map.msg_redirect_hash()```).

The difference between ```BPF_SOCKHASH``` and ```BPF_SOCKMAP``` is that ```BPF_SOCKMAP``` is implemented based on an array, and enforces keys to be four bytes. While ```BPF_SOCKHASH``` is implemented based on hash table, and the type of key can be specified freely.

Methods (covered later): map.sock_hash_update(), map.msg_redirect_hash(), map.sk_redirect_hash().

[search /tests](https://github.com/iovisor/bcc/search?q=BPF_SOCKHASH+path%3Atests&type=Code)

### 19. map.lookup()

Syntax: ```*val map.lookup(&key)```

Lookup the key in the map, and return a pointer to its value if it exists, else NULL. We pass the key in as an address to a pointer.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=lookup+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=lookup+path%3Atools&type=Code)

### 20. map.lookup_or_try_init()

Syntax: ```*val map.lookup_or_try_init(&key, &zero)```

Lookup the key in the map, and return a pointer to its value if it exists, else initialize the key's value to the second argument. This is often used to initialize values to zero. If the key cannot be inserted (e.g. the map is full) then NULL is returned.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=lookup_or_try_init+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=lookup_or_try_init+path%3Atools&type=Code)

Note: The old map.lookup_or_init() may cause return from the function, so lookup_or_try_init() is recommended as it
does not have this side effect.

### 21. map.delete()

Syntax: ```map.delete(&key)```

Delete the key from the hash.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=delete+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=delete+path%3Atools&type=Code)

### 22. map.update()

Syntax: ```map.update(&key, &val)```

Associate the value in the second argument to the key, overwriting any previous value.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=update+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=update+path%3Atools&type=Code)

### 23. map.insert()

Syntax: ```map.insert(&key, &val)```

Associate the value in the second argument to the key, only if there was no previous value.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=insert+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=insert+path%3Atools&type=Code)

### 24. map.increment()

Syntax: ```map.increment(key[, increment_amount])```

Increments the key's value by `increment_amount`, which defaults to 1. Used for histograms.

```map.increment()``` are not atomic. In the concurrency case. If you want more accurate results, use ```map.atomic_increment()``` instead of ```map.increment()```. The overhead of ```map.increment()``` and ```map.atomic_increment()``` is similar.

Note. When using ```map.atomic_increment()``` to operate on a BPF map of type ```BPF_MAP_TYPE_HASH```, ```map.atomic_increment()``` does not guarantee the atomicity of the operation when the specified key does not exist.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=increment+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=increment+path%3Atools&type=Code)

### 25. map.get_stackid()

Syntax: ```int map.get_stackid(void *ctx, u64 flags)```

This walks the stack found via the struct pt_regs in ```ctx```, saves it in the stack trace map, and returns a unique ID for the stack trace.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=get_stackid+path%3Aexamples&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=get_stackid+path%3Atools&type=Code)

### 26. map.perf_read()

Syntax: ```u64 map.perf_read(u32 cpu)```

This returns the hardware performance counter as configured in [5. BPF_PERF_ARRAY](#5-bpf_perf_array)

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=perf_read+path%3Atests&type=Code)

### 27. map.call()

Syntax: ```void map.call(void *ctx, int index)```

This invokes ```bpf_tail_call()``` to tail-call the bpf program which the ```index``` entry in [BPF_PROG_ARRAY](#10-bpf_prog_array) points to. A tail-call is different from the normal call. It reuses the current stack frame after jumping to another bpf program and never goes back. If the ```index``` entry is empty, it won't jump anywhere and the program execution continues as normal.

For example:

```C
BPF_PROG_ARRAY(prog_array, 10);

int tail_call(void *ctx) {
    bpf_trace_printk("Tail-call\n");
    return 0;
}

int do_tail_call(void *ctx) {
    bpf_trace_printk("Original program\n");
    prog_array.call(ctx, 2);
    return 0;
}
```

```Python
b = BPF(src_file="example.c")
tail_fn = b.load_func("tail_call", BPF.KPROBE)
prog_array = b.get_table("prog_array")
prog_array[c_int(2)] = c_int(tail_fn.fd)
b.attach_kprobe(event="some_kprobe_event", fn_name="do_tail_call")
```

This assigns ```tail_call()``` to ```prog_array[2]```. In the end of ```do_tail_call()```, ```prog_array.call(ctx, 2)``` tail-calls ```tail_call()``` and executes it.

**NOTE:** To prevent infinite loop, the maximum number of tail-calls is 32 ([```MAX_TAIL_CALL_CNT```](https://github.com/torvalds/linux/search?l=C&q=MAX_TAIL_CALL_CNT+path%3Ainclude%2Flinux&type=Code)).

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?l=C&q=call+path%3Aexamples&type=Code),
[search /tests](https://github.com/iovisor/bcc/search?l=C&q=call+path%3Atests&type=Code)

### 28. map.redirect_map()

Syntax: ```int map.redirect_map(int index, int flags)```

This redirects the incoming packets based on the ```index``` entry. If the map is [BPF_DEVMAP](#11-bpf_devmap), the packet will be sent to the transmit queue of the network interface that the entry points to. If the map is [BPF_CPUMAP](#12-bpf_cpumap), the packet will be sent to the ring buffer of the ```index``` CPU and be processed by the CPU later. If the map is [BPF_XSKMAP](#13-bpf_xskmap), the packet will be sent to the AF_XDP socket attached to the queue.

If the packet is redirected successfully, the function will return XDP_REDIRECT. Otherwise, it will return XDP_ABORTED to discard the packet.

For example:
```C
BPF_DEVMAP(devmap, 1);

int redirect_example(struct xdp_md *ctx) {
    return devmap.redirect_map(0, 0);
}
int xdp_dummy(struct xdp_md *ctx) {
    return XDP_PASS;
}
```

```Python
ip = pyroute2.IPRoute()
idx = ip.link_lookup(ifname="eth1")[0]

b = bcc.BPF(src_file="example.c")

devmap = b.get_table("devmap")
devmap[c_uint32(0)] = c_int(idx)

in_fn = b.load_func("redirect_example", BPF.XDP)
out_fn = b.load_func("xdp_dummy", BPF.XDP)
b.attach_xdp("eth0", in_fn, 0)
b.attach_xdp("eth1", out_fn, 0)
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?l=C&q=redirect_map+path%3Aexamples&type=Code),

### 29. map.push()

Syntax: ```int map.push(&val, int flags)```

Push an element onto a Stack or Queue table.
Passing BPF_EXIST as a flag causes the Queue or Stack to discard the oldest element if it is full.
Returns 0 on success, negative error on failure.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=push+path%3Atests&type=Code),

### 30. map.pop()

Syntax: ```int map.pop(&val)```

Pop an element from a Stack or Queue table. ```*val``` is populated with the result.
Unlike peeking, popping removes the element.
Returns 0 on success, negative error on failure.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=pop+path%3Atests&type=Code),

### 31. map.peek()

Syntax: ```int map.peek(&val)```

Peek an element at the head of a Stack or Queue table. ```*val``` is populated with the result.
Unlike popping, peeking does not remove the element.
Returns 0 on success, negative error on failure.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=peek+path%3Atests&type=Code),

### 32. map.sock_hash_update()

Syntax: ```int map.sock_hash_update(struct bpf_sock_ops *skops, &key, int flags)```

Add an entry to, or update a sockhash map referencing sockets. The skops is used as a new value for the entry associated to key. flags is one of:

```
BPF_NOEXIST: The entry for key must not exist in the map.
BPF_EXIST: The entry for key must already exist in the map.
BPF_ANY: No condition on the existence of the entry for key.
```

If the map has eBPF programs (parser and verdict), those will be inherited by the socket being added. If the socket is already attached to eBPF programs, this results in an error.

Return 0 on success, or a negative error in case of failure.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=sock_hash_update+path%3Atests&type=Code),

### 33. map.msg_redirect_hash()

Syntax: ```int map.msg_redirect_hash(struct sk_msg_buff *msg, void *key, u64 flags)```

This helper is used in programs implementing policies at the socket level. If the message msg is allowed to pass (i.e. if the verdict eBPF program returns SK_PASS), redirect it to the socket referenced by map (of type BPF_MAP_TYPE_SOCKHASH) using hash key. Both ingress and egress interfaces can be used for redirection. The BPF_F_INGRESS value in flags is used to make the distinction (ingress path is selected if the flag is present, egress path otherwise). This is the only flag supported for now.

Return SK_PASS on success, or SK_DROP on error.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=msg_redirect_hash+path%3Atests&type=Code),

### 34. map.sk_redirect_hash()

Syntax: ```int map.sk_redirect_hash(struct sk_buff *skb, void *key, u64 flags)```

This helper is used in programs implementing policies at the skb socket level. If the sk_buff skb is allowed to pass (i.e. if the verdict eBPF program returns SK_PASS), redirect it to the socket referenced by map (of  type  BPF_MAP_TYPE_SOCKHASH) using hash key. Both ingress and egress interfaces can be used for redirection. The BPF_F_INGRESS value in flags is used to make the distinction (ingress path is selected if the flag is present, egress otherwise). This is the only flag supported for now.

Return SK_PASS on success, or SK_DROP on error.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=sk_redirect_hash+path%3Atests&type=Code),

## Licensing

Depending on which [BPF helpers](kernel-versions.md#helpers) are used, a GPL-compatible license is required.

The special BCC macro `BPF_LICENSE` specifies the license of the BPF program. You can set the license as a comment in your source code, but the kernel has a special interface to specify it programmatically. If you need to use GPL-only helpers, it is recommended to specify the macro in your C code so that the kernel can understand it:

```C
// SPDX-License-Identifier: GPL-2.0+
#define BPF_LICENSE GPL
```

Otherwise, the kernel may reject loading your program (see the [error description](#2-cannot-call-gpl-only-function-from-proprietary-program) below). Note that it supports multiple words and quotes are not necessary:

```C
// SPDX-License-Identifier: GPL-2.0+ OR BSD-2-Clause
#define BPF_LICENSE Dual BSD/GPL
```

Check the [BPF helpers reference](kernel-versions.md#helpers) to see which helpers are GPL-only and what the kernel understands as GPL-compatible.

**If the macro is not specified, BCC will automatically define the license of the program as GPL.**

## Rewriter

One of jobs for rewriter is to turn implicit memory accesses to explicit ones using kernel helpers. Recent kernel introduced a config option ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE which will be set for architectures whose user address space and kernel address are disjoint. x86 and arm has this config option set while s390 does not. If ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE is not set, the bpf old helper `bpf_probe_read()` will not be available. Some existing users may have implicit memory accesses to access user memory, so using `bpf_probe_read_kernel()` will cause their application to fail. Therefore, for non-s390, the rewriter will use `bpf_probe_read()` for these implicit memory accesses. For s390, `bpf_probe_read_kernel()` is used as default and users should use `bpf_probe_read_user()` explicitly when accessing user memories.

# bcc Python

## Initialization

Constructors.

### 1. BPF

Syntax: ```BPF({text=BPF_program | src_file=filename} [, usdt_contexts=[USDT_object, ...]] [, cflags=[arg1, ...]] [, debug=int])```

Creates a BPF object. This is the main object for defining a BPF program, and interacting with its output.

Exactly one of `text` or `src_file` must be supplied (not both).

The `cflags` specifies additional arguments to be passed to the compiler, for example `-DMACRO_NAME=value` or `-I/include/path`.  The arguments are passed as an array, with each element being an additional argument.  Note that strings are not split on whitespace, so each argument must be a different element of the array, e.g. `["-include", "header.h"]`.

The `debug` flags control debug output, and can be or'ed together:
- `DEBUG_LLVM_IR = 0x1` compiled LLVM IR
- `DEBUG_BPF = 0x2` loaded BPF bytecode and register state on branches
- `DEBUG_PREPROCESSOR = 0x4` pre-processor result
- `DEBUG_SOURCE = 0x8` ASM instructions embedded with source
- `DEBUG_BPF_REGISTER_STATE = 0x10` register state on all instructions in addition to DEBUG_BPF
- `DEBUG_BTF = 0x20` print the messages from the `libbpf` library.

Examples:

```Python
# define entire BPF program in one line:
BPF(text='int do_trace(void *ctx) { bpf_trace_printk("hit!\\n"); return 0; }');

# define program as a variable:
prog = """
int hello(void *ctx) {
    bpf_trace_printk("Hello, World!\\n");
    return 0;
}
"""
b = BPF(text=prog)

# source a file:
b = BPF(src_file = "vfsreadlat.c")

# include a USDT object:
u = USDT(pid=int(pid))
[...]
b = BPF(text=bpf_text, usdt_contexts=[u])

# add include paths:
u = BPF(text=prog, cflags=["-I/path/to/include"])
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=BPF+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=BPF+path%3Atools+language%3Apython&type=Code)

### 2. USDT

Syntax: ```USDT({pid=pid | path=path})```

Creates an object to instrument User Statically-Defined Tracing (USDT) probes. Its primary method is ```enable_probe()```.

Arguments:

- pid: attach to this process ID.
- path: instrument USDT probes from this binary path.

Examples:

```Python
# include a USDT object:
u = USDT(pid=int(pid))
[...]
b = BPF(text=bpf_text, usdt_contexts=[u])
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=USDT+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=USDT+path%3Atools+language%3Apython&type=Code)

## Events

### 1. attach_kprobe()

Syntax: ```BPF.attach_kprobe(event="event", fn_name="name")```

Instruments the kernel function ```event()``` using kernel dynamic tracing of the function entry, and attaches our C defined function ```name()``` to be called when the kernel function is called.

For example:

```Python
b.attach_kprobe(event="sys_clone", fn_name="do_trace")
```

This will instrument the kernel ```sys_clone()``` function, which will then run our BPF defined ```do_trace()``` function each time it is called.

You can call attach_kprobe() more than once, and attach your BPF function to multiple kernel functions.
You can also call attach_kprobe() more than once to attach multiple BPF functions to the same kernel function.

See the previous kprobes section for how to instrument arguments from BPF.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_kprobe+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_kprobe+path%3Atools+language%3Apython&type=Code)

### 2. attach_kretprobe()

Syntax: ```BPF.attach_kretprobe(event="event", fn_name="name" [, maxactive=int])```

Instruments the return of the kernel function ```event()``` using kernel dynamic tracing of the function return, and attaches our C defined function ```name()``` to be called when the kernel function returns.

For example:

```Python
b.attach_kretprobe(event="vfs_read", fn_name="do_return")
```

This will instrument the kernel ```vfs_read()``` function, which will then run our BPF defined ```do_return()``` function each time it is called.

You can call attach_kretprobe() more than once, and attach your BPF function to multiple kernel function returns.
You can also call attach_kretprobe() more than once to attach multiple BPF functions to the same kernel function return.

When a kretprobe is installed on a kernel function, there is a limit on how many parallel calls it can catch. You can change that limit with ```maxactive```. See the kprobes documentation for its default value.

See the previous kretprobes section for how to instrument the return value from BPF.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_kretprobe+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_kretprobe+path%3Atools+language%3Apython&type=Code)

### 3. attach_tracepoint()

Syntax: ```BPF.attach_tracepoint(tp="tracepoint", fn_name="name")```

Instruments the kernel tracepoint described by ```tracepoint```, and when hit, runs the BPF function ```name()```.

This is an explicit way to instrument tracepoints. The ```TRACEPOINT_PROBE``` syntax, covered in the earlier tracepoints section, is an alternate method with the advantage of auto-declaring an ```args``` struct containing the tracepoint arguments. With ```attach_tracepoint()```, the tracepoint arguments need to be declared in the BPF program.

For example:

```Python
# define BPF program
bpf_text = """
#include 

struct urandom_read_args {
    // from /sys/kernel/debug/tracing/events/random/urandom_read/format
    u64 __unused__;
    u32 got_bits;
    u32 pool_left;
    u32 input_left;
};

int printarg(struct urandom_read_args *args) {
    bpf_trace_printk("%d\\n", args->got_bits);
    return 0;
};
"""

# load BPF program
b = BPF(text=bpf_text)
b.attach_tracepoint("random:urandom_read", "printarg")
```

Notice how the first argument to ```printarg()``` is now our defined struct.

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/a4159da8c4ea8a05a3c6e402451f530d6e5a8b41/examples/tracing/urandomread-explicit.py#L41),
[search /examples](https://github.com/iovisor/bcc/search?q=attach_tracepoint+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_tracepoint+path%3Atools+language%3Apython&type=Code)

### 4. attach_uprobe()

Syntax: ```BPF.attach_uprobe(name="location", sym="symbol", fn_name="name" [, sym_off=int])```, ```BPF.attach_uprobe(name="location", sym_re="regex", fn_name="name")```, ```BPF.attach_uprobe(name="location", addr=int, fn_name="name"), BPF.attach_uprobe(name="location", sym="symbol", fn_name="name", [, pid=int])```


Instruments the user-level function ```symbol()``` from either the library or binary named by ```location``` using user-level dynamic tracing of the function entry, and attach our C defined function ```name()``` to be called whenever the user-level function is called. If ```sym_off``` is given, the function is attached to the offset within the symbol.

The real address ```addr``` may be supplied in place of ```sym```, in which case ```sym``` must be set to its default value. If the file is a non-PIE executable, ```addr``` must be a virtual address, otherwise it must be an offset relative to the file load address.

Instead of a symbol name, a regular expression can be provided in ```sym_re```. The uprobe will then attach to symbols that match the provided regular expression.

Uprobes can be attached to a specific process by passing `pid` to `attach_uprobe`.
By default `pid` is set to -1, indicating the `uprobe` will be attached to all processes.
For libraries, the uprobe will attach to the version of the library used by the process if `pid` was given.
For how `pid` is used, see examples in [funcinterval](https://github.com/iovisor/bcc/blob/78423e1667db202012bbb032c567589175a2796c/tools/funcinterval.py#L155-L156).

Libraries can be given in the name argument without the lib prefix, with a versioned SONAME (c.so.6), or with the full path (/usr/lib/...). Binaries can be given only with the full path (/bin/sh).

For example:

```Python
b.attach_uprobe(name="c", sym="strlen", fn_name="count")
```

This will instrument ```strlen()``` function from libc, and call our BPF function ```count()``` when it is called. Note how the "lib" in "libc" is not necessary to specify.

Other examples:

```Python
b.attach_uprobe(name="c", sym="getaddrinfo", fn_name="do_entry")
b.attach_uprobe(name="/usr/bin/python", sym="main", fn_name="do_main")
```

You can call attach_uprobe() more than once, and attach your BPF function to multiple user-level functions.

See the previous uprobes section for how to instrument arguments from BPF.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_uprobe+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_uprobe+path%3Atools+language%3Apython&type=Code)

### 5. attach_uretprobe()

Syntax: ```BPF.attach_uretprobe(name="location", sym="symbol", fn_name="name")```

Instruments the return of the user-level function ```symbol()``` from either the library or binary named by ```location``` using user-level dynamic tracing of the function return, and attach our C defined function ```name()``` to be called whenever the user-level function returns.

For example:

```Python
b.attach_uretprobe(name="c", sym="strlen", fn_name="count")
```

This will instrument ```strlen()``` function from libc, and call our BPF function ```count()``` when it returns.

Other examples:

```Python
b.attach_uretprobe(name="c", sym="getaddrinfo", fn_name="do_return")
b.attach_uretprobe(name="/usr/bin/python", sym="main", fn_name="do_main")
```

You can call attach_uretprobe() more than once, and attach your BPF function to multiple user-level functions.
`BPF.attach_uretprobe` can also be used for a specific process.

See the previous uretprobes section for how to instrument the return value from BPF.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_uretprobe+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_uretprobe+path%3Atools+language%3Apython&type=Code)

### 6. USDT.enable_probe()

Syntax: ```USDT.enable_probe(probe=probe, fn_name=name)```

Attaches a BPF C function ```name``` to the USDT probe ```probe```.

Example:

```Python
# enable USDT probe from given PID
u = USDT(pid=int(pid))
u.enable_probe(probe="http__server__request", fn_name="do_trace")
```

To check if your binary has USDT probes, and what they are, you can run ```readelf -n binary``` and check the stap debug section.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=enable_probe+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=enable_probe+path%3Atools+language%3Apython&type=Code)

### 7. attach_raw_tracepoint()

Syntax: ```BPF.attach_raw_tracepoint(tp="tracepoint", fn_name="name")```

Instruments the kernel raw tracepoint described by ```tracepoint``` (```event``` only, no ```category```), and when hit, runs the BPF function ```name()```.

This is an explicit way to instrument tracepoints. The ```RAW_TRACEPOINT_PROBE``` syntax, covered in the earlier raw tracepoints section, is an alternate method.

For example:

```Python
b.attach_raw_tracepoint("sched_switch", "do_trace")
```

Examples in situ:
[search /tools](https://github.com/iovisor/bcc/search?q=attach_raw_tracepoint+path%3Atools+language%3Apython&type=Code)

### 8. attach_raw_socket()

Syntax: ```BPF.attach_raw_socket(fn, dev)```

Attaches a BPF function to the specified network interface.

The ```fn``` must be the type of ```BPF.function``` and the bpf_prog type needs to be ```BPF_PROG_TYPE_SOCKET_FILTER```  (```fn=BPF.load_func(func_name, BPF.SOCKET_FILTER)```)

```fn.sock``` is a non-blocking raw socket that was created and bound to ```dev```.

All network packets processed by ```dev``` are copied to the ```recv-q``` of ```fn.sock``` after being processed by bpf_prog. Try to recv packet form ```fn.sock``` with rev/recvfrom/recvmsg. Note that if the ```recv-q``` is not read in time after the ```recv-q``` is full, the copied packets will be discarded.

We can use this feature to capture network packets just like ```tcpdump```.

We can use ```ss --bpf --packet -p``` to observe ```fn.sock```.

Example:

```Python
BPF.attach_raw_socket(bpf_func, ifname)
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_raw_socket+path%3Aexamples+language%3Apython&type=Code)
### 9. attach_xdp()
Syntax: ```BPF.attach_xdp(dev="device", fn=b.load_func("fn_name",BPF.XDP), flags)```

Instruments the network driver described by ```dev``` , and then receives the packet, run the BPF function ```fn_name()``` with flags.

Here is a list of optional flags.

```Python
# from xdp_flags uapi/linux/if_link.h
XDP_FLAGS_UPDATE_IF_NOEXIST = (1 << 0)
XDP_FLAGS_SKB_MODE = (1 << 1)
XDP_FLAGS_DRV_MODE = (1 << 2)
XDP_FLAGS_HW_MODE = (1 << 3)
XDP_FLAGS_REPLACE = (1 << 4)
```

You can use flags like this ```BPF.attach_xdp(dev="device", fn=b.load_func("fn_name",BPF.XDP), flags=BPF.XDP_FLAGS_UPDATE_IF_NOEXIST)```

The default value of flags is 0. This means if there is no xdp program with `device`, the fn will run with that device. If there is an xdp program running with device, the old program will be replaced with new fn program.

Currently, bcc does not support XDP_FLAGS_REPLACE flag. The following are the descriptions of other flags.

#### 1. XDP_FLAGS_UPDATE_IF_NOEXIST
If an XDP program is already attached to the specified driver, attaching the XDP program again will fail.

#### 2. XDP_FLAGS_SKB_MODE
Driver doesn’t have support for XDP, but the kernel fakes it.
XDP program works, but there’s no real performance benefit because packets are handed to kernel stack anyways which then emulates XDP – this is usually supported with generic network drivers used in home computers, laptops, and virtualized HW.

#### 3. XDP_FLAGS_DRV_MODE
A driver has XDP support and can hand then to XDP without kernel stack interaction – Few drivers can support it and those are usually for enterprise HW.

#### 4. XDP_FLAGS_HW_MODE
XDP can be loaded and executed directly on the NIC – just a handful of NICs can do that.


For example:

```Python
b.attach_xdp(dev="ens1", fn=b.load_func("do_xdp", BPF.XDP))
```

This will instrument the network device ```ens1``` , which will then run our BPF defined ```do_xdp()``` function each time it receives packets.

Don't forget to call ```b.remove_xdp("ens1")``` at the end!

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=attach_xdp+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=attach_xdp+path%3Atools+language%3Apython&type=Code)

### 10. attach_func()

Syntax: ```BPF.attach_func(fn, attachable_fd, attach_type [, flags])```

Attaches a BPF function of the specified type to a particular ```attachable_fd```. if the ```attach_type``` is ```BPF_FLOW_DISSECTOR```, the function is expected to attach to current net namespace and ```attachable_fd``` must be 0.

For example:

```Python
b.attach_func(fn, cgroup_fd, BPFAttachType.CGROUP_SOCK_OPS)
b.attach_func(fn, map_fd, BPFAttachType.SK_MSG_VERDICT)
```

Note. When attached to "global" hooks (xdp, tc, lwt, cgroup). If the "BPF function" is no longer needed after the program terminates, be sure to call `detach_func` when the program exits.

Examples in situ:

[search /examples](https://github.com/iovisor/bcc/search?q=attach_func+path%3Aexamples+language%3Apython&type=Code),

### 11. detach_func()

Syntax: ```BPF.detach_func(fn, attachable_fd, attach_type)```

Detaches a BPF function of the specified type.

For example:

```Python
b.detach_func(fn, cgroup_fd, BPFAttachType.CGROUP_SOCK_OPS)
b.detach_func(fn, map_fd, BPFAttachType.SK_MSG_VERDICT)
```

Examples in situ:

[search /examples](https://github.com/iovisor/bcc/search?q=detach_func+path%3Aexamples+language%3Apython&type=Code),

### 12. detach_kprobe()

Syntax: ```BPF.detach_kprobe(event="event", fn_name="name")```

Detach a kprobe handler function of the specified event.

For example:

```Python
b.detach_kprobe(event="__page_cache_alloc", fn_name="trace_func_entry")
```

### 13. detach_kretprobe()

Syntax: ```BPF.detach_kretprobe(event="event", fn_name="name")```

Detach a kretprobe handler function of the specified event.

For example:

```Python
b.detach_kretprobe(event="__page_cache_alloc", fn_name="trace_func_return")
```

## Debug Output

### 1. trace_print()

Syntax: ```BPF.trace_print(fmt="fields")```

This method continually reads the globally shared /sys/kernel/debug/tracing/trace_pipe file and prints its contents. This file can be written to via BPF and the bpf_trace_printk() function, however, that method has limitations, including a lack of concurrent tracing support. The BPF_PERF_OUTPUT mechanism, covered earlier, is preferred.

Arguments:

- ```fmt```: optional, and can contain a field formatting string. It defaults to ```None```.

Examples:

```Python
# print trace_pipe output as-is:
b.trace_print()

# print PID and message:
b.trace_print(fmt="{1} {5}")
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=trace_print+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=trace_print+path%3Atools+language%3Apython&type=Code)

### 2. trace_fields()

Syntax: ```BPF.trace_fields(nonblocking=False)```

This method reads one line from the globally shared /sys/kernel/debug/tracing/trace_pipe file and returns it as fields. This file can be written to via BPF and the bpf_trace_printk() function, however, that method has limitations, including a lack of concurrent tracing support. The BPF_PERF_OUTPUT mechanism, covered earlier, is preferred.

Arguments:

- ```nonblocking```: optional, defaults to ```False```. When set to ```True```, the program will not block waiting for input.

Examples:

```Python
while 1:
    try:
        (task, pid, cpu, flags, ts, msg) = b.trace_fields()
    except ValueError:
        continue
    [...]
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=trace_fields+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=trace_fields+path%3Atools+language%3Apython&type=Code)

## Output APIs

Normal output from a BPF program is either:

- per-event: using PERF_EVENT_OUTPUT, open_perf_buffer(), and perf_buffer_poll().
- map summary: using items(), or print_log2_hist(), covered in the Maps section.

### 1. perf_buffer_poll()

Syntax: ```BPF.perf_buffer_poll(timeout=T)```

This polls from all open perf ring buffers, calling the callback function that was provided when calling open_perf_buffer for each entry.

The timeout parameter is optional and measured in milliseconds. In its absence, polling continues indefinitely.

Example:

```Python
# loop with callback to print_event
b["events"].open_perf_buffer(print_event)
while 1:
    try:
        b.perf_buffer_poll()
    except KeyboardInterrupt:
        exit();
```

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/v0.9.0/examples/tracing/hello_perf_output.py#L55),
[search /examples](https://github.com/iovisor/bcc/search?q=perf_buffer_poll+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=perf_buffer_poll+path%3Atools+language%3Apython&type=Code)

### 2. ring_buffer_poll()

Syntax: ```BPF.ring_buffer_poll(timeout=T)```

This polls from all open ringbuf ring buffers, calling the callback function that was provided when calling open_ring_buffer for each entry.

The timeout parameter is optional and measured in milliseconds. In its absence, polling continues until
there is no more data or the callback returns a negative value.

Example:

```Python
# loop with callback to print_event
b["events"].open_ring_buffer(print_event)
while 1:
    try:
        b.ring_buffer_poll(30)
    except KeyboardInterrupt:
        exit();
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=ring_buffer_poll+path%3Aexamples+language%3Apython&type=Code),

### 3. ring_buffer_consume()

Syntax: ```BPF.ring_buffer_consume()```

This consumes from all open ringbuf ring buffers, calling the callback function that was provided when calling open_ring_buffer for each entry.

Unlike ```ring_buffer_poll```, this method **does not poll for data** before attempting to consume.
This reduces latency at the expense of higher CPU consumption. If you are unsure which to use,
use ```ring_buffer_poll```.

Example:

```Python
# loop with callback to print_event
b["events"].open_ring_buffer(print_event)
while 1:
    try:
        b.ring_buffer_consume()
    except KeyboardInterrupt:
        exit();
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=ring_buffer_consume+path%3Aexamples+language%3Apython&type=Code),

## Map APIs

Maps are BPF data stores, and are used in bcc to implement a table, and then higher level objects on top of tables, including hashes and histograms.

### 1. get_table()

Syntax: ```BPF.get_table(name)```

Returns a table object. This is no longer used, as tables can now be read as items from BPF. Eg: ```BPF[name]```.

Examples:

```Python
counts = b.get_table("counts")

counts = b["counts"]
```

These are equivalent.

### 2. open_perf_buffer()

Syntax: ```table.open_perf_buffer(callback, page_cnt=N, lost_cb=None)```

This operates on a table as defined in BPF as BPF_PERF_OUTPUT(), and associates the callback Python function ```callback``` to be called when data is available in the perf ring buffer. This is part of the recommended mechanism for transferring per-event data from kernel to user space. The size of the perf ring buffer can be specified via the ```page_cnt``` parameter, which must be a power of two number of pages and defaults to 8. If the callback is not processing data fast enough, some submitted data may be lost. ```lost_cb``` will be called to log / monitor the lost count. If ```lost_cb``` is the default ```None``` value, it will just print a line of message to ```stderr```.

Example:

```Python
# process event
def print_event(cpu, data, size):
    event = ct.cast(data, ct.POINTER(Data)).contents
    [...]

# loop with callback to print_event
b["events"].open_perf_buffer(print_event)
while 1:
    try:
        b.perf_buffer_poll()
    except KeyboardInterrupt:
        exit()
```

Note that the data structure transferred will need to be declared in C in the BPF program. For example:

```C
// define output data structure in C
struct data_t {
    u32 pid;
    u64 ts;
    char comm[TASK_COMM_LEN];
};
BPF_PERF_OUTPUT(events);
[...]
```

In Python, you can either let bcc generate the data structure from C declaration automatically (recommended):

```Python
def print_event(cpu, data, size):
    event = b["events"].event(data)
[...]
```

or define it manually:

```Python
# define output data structure in Python
TASK_COMM_LEN = 16    # linux/sched.h
class Data(ct.Structure):
    _fields_ = [("pid", ct.c_ulonglong),
                ("ts", ct.c_ulonglong),
                ("comm", ct.c_char * TASK_COMM_LEN)]

def print_event(cpu, data, size):
    event = ct.cast(data, ct.POINTER(Data)).contents
[...]
```

Examples in situ:
[code](https://github.com/iovisor/bcc/blob/v0.9.0/examples/tracing/hello_perf_output.py#L52),
[search /examples](https://github.com/iovisor/bcc/search?q=open_perf_buffer+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=open_perf_buffer+path%3Atools+language%3Apython&type=Code)

### 3. items()

Syntax: ```table.items()```

Returns an array of the keys in a table. This can be used with BPF_HASH maps to fetch, and iterate, over the keys.

Example:

```Python
# print output
print("%10s %s" % ("COUNT", "STRING"))
counts = b.get_table("counts")
for k, v in sorted(counts.items(), key=lambda counts: counts[1].value):
    print("%10d \"%s\"" % (v.value, k.c.encode('string-escape')))
```

This example also uses the ```sorted()``` method to sort by value.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=items+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=items+path%3Atools+language%3Apython&type=Code)

### 4. values()

Syntax: ```table.values()```

Returns an array of the values in a table.

### 5. clear()

Syntax: ```table.clear()```

Clears the table: deletes all entries.

Example:

```Python
# print map summary every second:
while True:
    time.sleep(1)
    print("%-8s\n" % time.strftime("%H:%M:%S"), end="")
    dist.print_log2_hist(sym + " return:")
    dist.clear()
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=clear+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=clear+path%3Atools+language%3Apython&type=Code)

### 6. items_lookup_and_delete_batch()

Syntax: ```table.items_lookup_and_delete_batch()```

Returns an array of the keys in a table with a single call to BPF syscall. This can be used with BPF_HASH maps to fetch, and iterate, over the keys. It also clears the table: deletes all entries.
You should rather use table.items_lookup_and_delete_batch() than table.items() followed by table.clear(). It requires kernel v5.6.

Example:

```Python
# print call rate per second:
print("%9s-%9s-%8s-%9s" % ("PID", "COMM", "fname", "counter"))
while True:
    for k, v in sorted(b['map'].items_lookup_and_delete_batch(), key=lambda kv: (kv[0]).pid):
        print("%9s-%9s-%8s-%9d" % (k.pid, k.comm, k.fname, v.counter))
    sleep(1)
```

### 7. items_lookup_batch()

Syntax: ```table.items_lookup_batch()```

Returns an array of the keys in a table with a single call to BPF syscall. This can be used with BPF_HASH maps to fetch, and iterate, over the keys.
You should rather use table.items_lookup_batch() than table.items(). It requires kernel v5.6.

Example:

```Python
# print current value of map:
print("%9s-%9s-%8s-%9s" % ("PID", "COMM", "fname", "counter"))
while True:
    for k, v in sorted(b['map'].items_lookup_batch(), key=lambda kv: (kv[0]).pid):
        print("%9s-%9s-%8s-%9d" % (k.pid, k.comm, k.fname, v.counter))
```

### 8. items_delete_batch()

Syntax: ```table.items_delete_batch(keys)```

It clears all entries of a BPF_HASH map when keys is None. It is more efficient than table.clear() since it generates only one system call. You can delete a subset of a map by giving an array of keys as parameter. Those keys and their associated values will be deleted. It requires kernel v5.6.

Arguments:

- keys is optional and by default is None.



### 9. items_update_batch()

Syntax: ```table.items_update_batch(keys, values)```

Update all the provided keys with new values. The two arguments must be the same length and within the map limits (between 1 and the maximum entries). It requires kernel v5.6.

Arguments:

- keys is the list of keys to be updated
- values is the list containing the new values.


### 10. print_log2_hist()

Syntax: ```table.print_log2_hist(val_type="value", section_header="Bucket ptr", section_print_fn=None)```

Prints a table as a log2 histogram in ASCII. The table must be stored as log2, which can be done using the BPF function ```bpf_log2l()```.

Arguments:

- val_type: optional, column header.
- section_header: if the histogram has a secondary key, multiple tables will print and section_header can be used as a header description for each.
- section_print_fn: if section_print_fn is not None, it will be passed the bucket value.

Example:

```Python
b = BPF(text="""
BPF_HISTOGRAM(dist);

int kprobe__blk_account_io_done(struct pt_regs *ctx, struct request *req)
{
	dist.increment(bpf_log2l(req->__data_len / 1024));
	return 0;
}
""")
[...]

b["dist"].print_log2_hist("kbytes")
```

Output:

```
     kbytes          : count     distribution
       0 -> 1        : 3        |                                      |
       2 -> 3        : 0        |                                      |
       4 -> 7        : 211      |**********                            |
       8 -> 15       : 0        |                                      |
      16 -> 31       : 0        |                                      |
      32 -> 63       : 0        |                                      |
      64 -> 127      : 1        |                                      |
     128 -> 255      : 800      |**************************************|
```

This output shows a multi-modal distribution, with the largest mode of 128->255 kbytes and a count of 800.

This is an efficient way to summarize data, as the summarization is performed in-kernel, and only the count column is passed to user space.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=print_log2_hist+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=print_log2_hist+path%3Atools+language%3Apython&type=Code)

### 11. print_linear_hist()

Syntax: ```table.print_linear_hist(val_type="value", section_header="Bucket ptr", section_print_fn=None)```

Prints a table as a linear histogram in ASCII. This is intended to visualize small integer ranges, eg, 0 to 100.

Arguments:

- val_type: optional, column header.
- section_header: if the histogram has a secondary key, multiple tables will print and section_header can be used as a header description for each.
- section_print_fn: if section_print_fn is not None, it will be passed the bucket value.

Example:

```Python
b = BPF(text="""
BPF_HISTOGRAM(dist);

int kprobe__blk_account_io_done(struct pt_regs *ctx, struct request *req)
{
	dist.increment(req->__data_len / 1024);
	return 0;
}
""")
[...]

b["dist"].print_linear_hist("kbytes")
```

Output:

```
     kbytes        : count     distribution
        0          : 3        |******                                  |
        1          : 0        |                                        |
        2          : 0        |                                        |
        3          : 0        |                                        |
        4          : 19       |****************************************|
        5          : 0        |                                        |
        6          : 0        |                                        |
        7          : 0        |                                        |
        8          : 4        |********                                |
        9          : 0        |                                        |
        10         : 0        |                                        |
        11         : 0        |                                        |
        12         : 0        |                                        |
        13         : 0        |                                        |
        14         : 0        |                                        |
        15         : 0        |                                        |
        16         : 2        |****                                    |
[...]
```

This is an efficient way to summarize data, as the summarization is performed in-kernel, and only the values in the count column are passed to user space.

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=print_linear_hist+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=print_linear_hist+path%3Atools+language%3Apython&type=Code)

### 12. open_ring_buffer()

Syntax: ```table.open_ring_buffer(callback, ctx=None)```

This operates on a table as defined in BPF as BPF_RINGBUF_OUTPUT(), and associates the callback Python function ```callback``` to be called when data is available in the ringbuf ring buffer. This is part of the new (Linux 5.8+) recommended mechanism for transferring per-event data from kernel to user space. Unlike perf buffers, ringbuf sizes are specified within the BPF program, as part of the ```BPF_RINGBUF_OUTPUT``` macro. If the callback is not processing data fast enough, some submitted data may be lost. In this case, the events should be polled more frequently and/or the size of the ring buffer should be increased.

Example:

```Python
# process event
def print_event(ctx, data, size):
    event = ct.cast(data, ct.POINTER(Data)).contents
    [...]

# loop with callback to print_event
b["events"].open_ring_buffer(print_event)
while 1:
    try:
        b.ring_buffer_poll()
    except KeyboardInterrupt:
        exit()
```

Note that the data structure transferred will need to be declared in C in the BPF program. For example:

```C
// define output data structure in C
struct data_t {
    u32 pid;
    u64 ts;
    char comm[TASK_COMM_LEN];
};
BPF_RINGBUF_OUTPUT(events, 8);
[...]
```

In Python, you can either let bcc generate the data structure from C declaration automatically (recommended):

```Python
def print_event(ctx, data, size):
    event = b["events"].event(data)
[...]
```

or define it manually:

```Python
# define output data structure in Python
TASK_COMM_LEN = 16    # linux/sched.h
class Data(ct.Structure):
    _fields_ = [("pid", ct.c_ulonglong),
                ("ts", ct.c_ulonglong),
                ("comm", ct.c_char * TASK_COMM_LEN)]

def print_event(ctx, data, size):
    event = ct.cast(data, ct.POINTER(Data)).contents
[...]
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=open_ring_buffer+path%3Aexamples+language%3Apython&type=Code),

### 13. push()

Syntax: ```table.push(leaf, flags=0)```

Push an element onto a Stack or Queue table. Raises an exception if the operation does not succeed.
Passing QueueStack.BPF_EXIST as a flag causes the Queue or Stack to discard the oldest element if it is full.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=push+path%3Atests+language%3Apython&type=Code),

### 14. pop()

Syntax: ```leaf = table.pop()```

Pop an element from a Stack or Queue table. Unlike ```peek()```, ```pop()```
removes the element from the table before returning it.
Raises a KeyError exception if the operation does not succeed.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=pop+path%3Atests+language%3Apython&type=Code),

### 15. peek()

Syntax: ```leaf = table.peek()```

Peek the element at the head of a Stack or Queue table. Unlike ```pop()```, ```peek()```
does not remove the element from the table. Raises an exception if the operation does not succeed.

Examples in situ:
[search /tests](https://github.com/iovisor/bcc/search?q=peek+path%3Atests+language%3Apython&type=Code),

## Helpers

Some helper methods provided by bcc. Note that since we're in Python, we can import any Python library and their methods, including, for example, the libraries: argparse, collections, ctypes, datetime, re, socket, struct, subprocess, sys, and time.

### 1. ksym()

Syntax: ```BPF.ksym(addr)```

Translate a kernel memory address into a kernel function name, which is returned.

Example:

```Python
print("kernel function: " + b.ksym(addr))
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=ksym+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=ksym+path%3Atools+language%3Apython&type=Code)

### 2. ksymname()

Syntax: ```BPF.ksymname(name)```

Translate a kernel name into an address. This is the reverse of ksym. Returns -1 when the function name is unknown.

Example:

```Python
print("kernel address: %x" % b.ksymname("vfs_read"))
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=ksymname+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=ksymname+path%3Atools+language%3Apython&type=Code)

### 3. sym()

Syntax: ```BPF.sym(addr, pid, show_module=False, show_offset=False)```

Translate a memory address into a function name for a pid, which is returned. A pid of less than zero will access the kernel symbol cache. The `show_module` and `show_offset` parameters control whether the module in which the symbol lies should be displayed, and whether the instruction offset from the beginning of the symbol should be displayed. These extra parameters default to `False`.

Example:

```Python
print("function: " + b.sym(addr, pid))
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=sym+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=sym+path%3Atools+language%3Apython&type=Code)

### 4. num_open_kprobes()

Syntax: ```BPF.num_open_kprobes()```

Returns the number of open k[ret]probes. Can be useful for scenarios where event_re is used while attaching and detaching probes. Excludes perf_events readers.

Example:

```Python
b.attach_kprobe(event_re=pattern, fn_name="trace_count")
matched = b.num_open_kprobes()
if matched == 0:
    print("0 functions matched by \"%s\". Exiting." % args.pattern)
    exit()
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=num_open_kprobes+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=num_open_kprobes+path%3Atools+language%3Apython&type=Code)

### 5. get_syscall_fnname()

Syntax: ```BPF.get_syscall_fnname(name : str)```

Return the corresponding kernel function name of the syscall. This helper function will try different prefixes and use the right one to concatenate with the syscall name. Note that the return value may vary in different versions of linux kernel and sometimes it will causing trouble. (see [#2590](https://github.com/iovisor/bcc/issues/2590))

Example:

```Python
print("The function name of %s in kernel is %s" % ("clone", b.get_syscall_fnname("clone")))
# sys_clone or __x64_sys_clone or ...
```

Examples in situ:
[search /examples](https://github.com/iovisor/bcc/search?q=get_syscall_fnname+path%3Aexamples+language%3Apython&type=Code),
[search /tools](https://github.com/iovisor/bcc/search?q=get_syscall_fnname+path%3Atools+language%3Apython&type=Code)

# BPF Errors

See the "Understanding eBPF verifier messages" section in the kernel source under Documentation/networking/filter.txt.

## 1. Invalid mem access

This can be due to trying to read memory directly, instead of operating on memory on the BPF stack. All kernel memory reads must be passed via bpf_probe_read_kernel() to copy kernel memory into the BPF stack, which can be automatic by the bcc rewriter in some cases of simple dereferencing. bpf_probe_read_kernel() does all the required checks.

Example:

```
bpf: Permission denied
0: (bf) r6 = r1
1: (79) r7 = *(u64 *)(r6 +80)
2: (85) call 14
3: (bf) r8 = r0
[...]
23: (69) r1 = *(u16 *)(r7 +16)
R7 invalid mem access 'inv'

Traceback (most recent call last):
  File "./tcpaccept", line 179, in <module>
    b = BPF(text=bpf_text)
  File "/usr/lib/python2.7/dist-packages/bcc/__init__.py", line 172, in __init__
    self._trace_autoload()
  File "/usr/lib/python2.7/dist-packages/bcc/__init__.py", line 612, in _trace_autoload
    fn = self.load_func(func_name, BPF.KPROBE)
  File "/usr/lib/python2.7/dist-packages/bcc/__init__.py", line 212, in load_func
    raise Exception("Failed to load BPF program %s" % func_name)
Exception: Failed to load BPF program kretprobe__inet_csk_accept
```

## 2. Cannot call GPL only function from proprietary program

This error happens when a GPL-only helper is called from a non-GPL BPF program. To fix this error, do not use GPL-only helpers from a proprietary BPF program, or relicense the BPF program under a GPL-compatible license. Check which [BPF helpers](https://github.com/iovisor/bcc/blob/master/docs/kernel-versions.md#helpers) are GPL-only, and what licenses are considered GPL-compatible.

Example calling `bpf_get_stackid()`, a GPL-only BPF helper, from a proprietary program (`#define BPF_LICENSE Proprietary`):

```
bpf: Failed to load program: Invalid argument
[...]
8: (85) call bpf_get_stackid#27
cannot call GPL only function from proprietary program
```

# Environment Variables

## 1. Kernel source directory

eBPF program compilation needs kernel sources or kernel headers with headers
compiled. In case your kernel sources are at a non-standard location where BCC
cannot find then, it's possible to provide BCC the absolute path of the location
by setting `BCC_KERNEL_SOURCE` to it.

## 2. Kernel version overriding

By default, BCC stores the `LINUX_VERSION_CODE` in the generated eBPF object
which is then passed along to the kernel when the eBPF program is loaded.
Sometimes this is quite inconvenient especially when the kernel is slightly
updated such as an LTS kernel release. Its extremely unlikely the slight
mismatch would cause any issues with the loaded eBPF program. By setting
`BCC_LINUX_VERSION_CODE` to the version of the kernel that's running, the check
for verifying the kernel version can be bypassed. This is needed for programs
that use kprobes. This needs to be encoded in the format: `(VERSION * 65536) +
(PATCHLEVEL * 256) + SUBLEVEL`. For example, if the running kernel is `4.9.10`,
then can set `export BCC_LINUX_VERSION_CODE=264458` to override the kernel
version check successfully.

---

## File: docs/special_filtering.md

# Special Filtering

Some tools have special filtering capabitilies, the main use case is to trace
processes running in containers, but those mechanisms are generic and could
be used in other cases as well.

## Filtering by cgroups

Some tools have an option to filter by cgroup by referencing a pinned BPF hash
map managed externally.

Examples of commands:

```
# ./opensnoop --cgroupmap /sys/fs/bpf/test01
# ./execsnoop --cgroupmap /sys/fs/bpf/test01
# ./tcpconnect --cgroupmap /sys/fs/bpf/test01
# ./tcpaccept --cgroupmap /sys/fs/bpf/test01
# ./tcptracer --cgroupmap /sys/fs/bpf/test01
```

The commands above will only display results from processes that belong to one
of the cgroups whose id, returned by `bpf_get_current_cgroup_id()`, is in the
pinned BPF hash map.

The BPF hash map can be created by:

```
# bpftool map create /sys/fs/bpf/test01 type hash key 8 value 8 entries 128 \
        name cgroupset flags 0
```

To get a shell in a new cgroup, you can use:

```
# systemd-run --pty --unit test bash
```

The shell will be running in the cgroup
`/sys/fs/cgroup/unified/system.slice/test.service`.

The cgroup id can be discovered using the `name_to_handle_at()` system call. In
the examples/cgroupid, you will find an example of program to get the cgroup
id.

```
# cd examples/cgroupid
# make
# ./cgroupid hex /sys/fs/cgroup/unified/system.slice/test.service
```

or, using Docker:

```
# cd examples/cgroupid
# docker build -t cgroupid .
# docker run --rm --privileged -v /sys/fs/cgroup:/sys/fs/cgroup \
	cgroupid cgroupid hex /sys/fs/cgroup/unified/system.slice/test.service
```

This prints the cgroup id as a hexadecimal string in the host endianness such
as `77 16 00 00 01 00 00 00`.

```
# FILE=/sys/fs/bpf/test01
# CGROUPID_HEX="77 16 00 00 01 00 00 00"
# bpftool map update pinned $FILE key hex $CGROUPID_HEX value hex 00 00 00 00 00 00 00 00 any
```

Now that the shell started by systemd-run has its cgroup id in the BPF hash
map, bcc tools will display results from this shell. Cgroups can be added and
removed from the BPF hash map without restarting the bcc tool.

This feature is useful for integrating bcc tools in external projects.

## Filtering by mount namespace

The BPF hash map can be created by:

```
# bpftool map create /sys/fs/bpf/mnt_ns_set type hash key 8 value 4 entries 128 \
        name mnt_ns_set flags 0
```

Execute the `execsnoop` tool filtering only the mount namespaces
in `/sys/fs/bpf/mnt_ns_set`:

```
# tools/execsnoop.py --mntnsmap /sys/fs/bpf/mnt_ns_set
```

Start a terminal in a new mount namespace:

```
# unshare -m bash
```

Update the hash map with the mount namespace ID of the terminal above:

```
FILE=/sys/fs/bpf/mnt_ns_set
if [ $(printf '\1' | od -dAn) -eq 1 ]; then
 HOST_ENDIAN_CMD=tac
else
  HOST_ENDIAN_CMD=cat
fi

NS_ID_HEX="$(printf '%016x' $(stat -Lc '%i' /proc/self/ns/mnt) | sed 's/.\{2\}/&\n/g' | $HOST_ENDIAN_CMD)"
bpftool map update pinned $FILE key hex $NS_ID_HEX value hex 00 00 00 00 any
```

Execute a command in this terminal:

```
# ping kinvolk.io
```

You'll see how on the `execsnoop` terminal you started above the call is logged:

```
# tools/execsnoop.py --mntnsmap /sys/fs/bpf/mnt_ns_set
[sudo] password for mvb:
PCOMM            PID    PPID   RET ARGS
ping             8096   7970     0 /bin/ping kinvolk.io
```

---

## File: docs/tutorial.md

# bcc Tutorial

This tutorial covers how to use [bcc](https://github.com/iovisor/bcc) tools to quickly solve performance, troubleshooting, and networking issues. If you want to develop new bcc tools, see [tutorial_bcc_python_developer.md](tutorial_bcc_python_developer.md) for that tutorial.

It is assumed for this tutorial that bcc is already installed, and you can run tools like execsnoop successfully. See [INSTALL.md](../INSTALL.md). This uses enhancements added to the Linux 4.x series.

## Observability

Some quick wins.

### 0. Before bcc

Before using bcc, you should start with the Linux basics. One reference is the [Linux Performance Analysis in 60,000 Milliseconds](https://netflixtechblog.com/linux-performance-analysis-in-60-000-milliseconds-accc10403c55) post, which covers these commands:

1. uptime
1. dmesg | tail
1. vmstat 1
1. mpstat -P ALL 1
1. pidstat 1
1. iostat -xz 1
1. free -m
1. sar -n DEV 1
1. sar -n TCP,ETCP 1
1. top

### 1. General Performance

Here is a generic checklist for performance investigations with bcc, first as a list, then in detail:

1. execsnoop
1. opensnoop
1. ext4slower (or btrfs\*, xfs\*, zfs\*)
1. biolatency
1. biosnoop
1. cachestat
1. tcpconnect
1. tcpaccept
1. tcpretrans
1. runqlat
1. profile

These tools may be installed on your system under /usr/share/bcc/tools, or you can run them from the bcc github repo under /tools where they have a .py extension. Browse the 50+ tools available for more analysis options.

#### 1.1 execsnoop

```
# ./execsnoop
PCOMM            PID    RET ARGS
supervise        9660     0 ./run
supervise        9661     0 ./run
mkdir            9662     0 /bin/mkdir -p ./main
run              9663     0 ./run
[...]
```

execsnoop prints one line of output for each new process. Check for short-lived processes. These can consume CPU resources, but not show up in most monitoring tools that periodically take snapshots of which processes are running.

It works by tracing exec(), not the fork(), so it will catch many types of new processes but not all (eg, it won't see an application launching working processes, that doesn't exec() anything else).

More [examples](../tools/execsnoop_example.txt).

#### 1.2. opensnoop

```
# ./opensnoop
PID    COMM               FD ERR PATH
1565   redis-server        5   0 /proc/1565/stat
1565   redis-server        5   0 /proc/1565/stat
1565   redis-server        5   0 /proc/1565/stat
1603   snmpd               9   0 /proc/net/dev
1603   snmpd              11   0 /proc/net/if_inet6
1603   snmpd              -1   2 /sys/class/net/eth0/device/vendor
1603   snmpd              11   0 /proc/sys/net/ipv4/neigh/eth0/retrans_time_ms
1603   snmpd              11   0 /proc/sys/net/ipv6/neigh/eth0/retrans_time_ms
1603   snmpd              11   0 /proc/sys/net/ipv6/conf/eth0/forwarding
[...]
```

opensnoop prints one line of output for each open() syscall, including details.

Files that are opened can tell you a lot about how applications work: identifying their data files, config files, and log files. Sometimes applications can misbehave, and perform poorly, when they are constantly attempting to read files that do not exist. opensnoop gives you a quick look.

More [examples](../tools/opensnoop_example.txt).

#### 1.3. ext4slower (or btrfs\*, xfs\*, zfs\*)

```
# ./ext4slower
Tracing ext4 operations slower than 10 ms
TIME     COMM           PID    T BYTES   OFF_KB   LAT(ms) FILENAME
06:35:01 cron           16464  R 1249    0          16.05 common-auth
06:35:01 cron           16463  R 1249    0          16.04 common-auth
06:35:01 cron           16465  R 1249    0          16.03 common-auth
06:35:01 cron           16465  R 4096    0          10.62 login.defs
06:35:01 cron           16464  R 4096    0          10.61 login.defs
```

ext4slower traces the ext4 file system and times common operations, and then only prints those that exceed a threshold.

This is great for identifying or exonerating one type of performance issue: show individually slow disk i/O via the file system. Disks process I/O asynchronously, and it can be difficult to associate latency at that layer with the latency applications experience. Tracing higher up in the kernel stack, at the VFS -> file system interface, will more closely match what an application suffers. Use this tool to identify if file system latency exceeds a given threshold.

Similar tools exist in bcc for other file systems: btrfsslower, xfsslower, and zfsslower. There is also fileslower, which works at the VFS layer and traces everything (although at some higher overhead).

More [examples](../tools/ext4slower_example.txt).

#### 1.4. biolatency

```
# ./biolatency
Tracing block device I/O... Hit Ctrl-C to end.
^C
     usecs           : count     distribution
       0 -> 1        : 0        |                                      |
       2 -> 3        : 0        |                                      |
       4 -> 7        : 0        |                                      |
       8 -> 15       : 0        |                                      |
      16 -> 31       : 0        |                                      |
      32 -> 63       : 0        |                                      |
      64 -> 127      : 1        |                                      |
     128 -> 255      : 12       |********                              |
     256 -> 511      : 15       |**********                            |
     512 -> 1023     : 43       |*******************************       |
    1024 -> 2047     : 52       |**************************************|
    2048 -> 4095     : 47       |**********************************    |
    4096 -> 8191     : 52       |**************************************|
    8192 -> 16383    : 36       |**************************            |
   16384 -> 32767    : 15       |**********                            |
   32768 -> 65535    : 2        |*                                     |
   65536 -> 131071   : 2        |*                                     |
```

biolatency traces disk I/O latency (time from device issue to completion), and when the tool ends (Ctrl-C, or a given interval), it prints a histogram summary of the latency.

This is great for understanding disk I/O latency beyond the average times given by tools like iostat. I/O latency outliers will be visible at the end of the distribution, as well as multi-mode distributions.

More [examples](../tools/biolatency_example.txt).

#### 1.5. biosnoop

```
# ./biosnoop
TIME(s)        COMM           PID    DISK    T  SECTOR    BYTES   LAT(ms)
0.000004001    supervise      1950   xvda1   W  13092560  4096       0.74
0.000178002    supervise      1950   xvda1   W  13092432  4096       0.61
0.001469001    supervise      1956   xvda1   W  13092440  4096       1.24
0.001588002    supervise      1956   xvda1   W  13115128  4096       1.09
1.022346001    supervise      1950   xvda1   W  13115272  4096       0.98
1.022568002    supervise      1950   xvda1   W  13188496  4096       0.93
[...]
```

biosnoop prints a line of output for each disk I/O, with details including latency (time from device issue to completion).

This allows you to examine disk I/O in more detail, and look for time-ordered patterns (eg, reads queueing behind writes). Note that the output will be verbose if your system performs disk I/O at a high rate.

More [examples](../tools/biosnoop_example.txt).

#### 1.6. cachestat

```
# ./cachestat
    HITS   MISSES  DIRTIES  READ_HIT% WRITE_HIT%   BUFFERS_MB  CACHED_MB
    1074       44       13      94.9%       2.9%            1        223
    2195      170        8      92.5%       6.8%            1        143
     182       53       56      53.6%       1.3%            1        143
   62480    40960    20480      40.6%      19.8%            1        223
       7        2        5      22.2%      22.2%            1        223
     348        0        0     100.0%       0.0%            1        223
[...]
```

cachestat prints a one line summary every second (or every custom interval) showing statistics from the file system cache.

Use this to identify a low cache hit ratio, and a high rate of misses: which gives one lead for performance tuning.

More [examples](../tools/cachestat_example.txt).

#### 1.7. tcpconnect

```
# ./tcpconnect
PID    COMM         IP SADDR            DADDR            DPORT
1479   telnet       4  127.0.0.1        127.0.0.1        23
1469   curl         4  10.201.219.236   54.245.105.25    80
1469   curl         4  10.201.219.236   54.67.101.145    80
1991   telnet       6  ::1              ::1              23
2015   ssh          6  fe80::2000:bff:fe82:3ac fe80::2000:bff:fe82:3ac 22
[...]
```

tcpconnect prints one line of output for every active TCP connection (eg, via connect()), with details including source and destination addresses.

Look for unexpected connections that may point to inefficiencies in application configuration, or an intruder.

More [examples](../tools/tcpconnect_example.txt).

#### 1.8. tcpaccept

```
# ./tcpaccept
PID    COMM         IP RADDR            LADDR            LPORT
907    sshd         4  192.168.56.1     192.168.56.102   22
907    sshd         4  127.0.0.1        127.0.0.1        22
5389   perl         6  1234:ab12:2040:5020:2299:0:5:0 1234:ab12:2040:5020:2299:0:5:0 7001
[...]
```

tcpaccept prints one line of output for every passive TCP connection (eg, via accept()), with details including source and destination addresses.

Look for unexpected connections that may point to inefficiencies in application configuration, or an intruder.

More [examples](../tools/tcpaccept_example.txt).

#### 1.9. tcpretrans

```
# ./tcpretrans
TIME     PID    IP LADDR:LPORT          T> RADDR:RPORT          STATE
01:55:05 0      4  10.153.223.157:22    R> 69.53.245.40:34619   ESTABLISHED
01:55:05 0      4  10.153.223.157:22    R> 69.53.245.40:34619   ESTABLISHED
01:55:17 0      4  10.153.223.157:22    R> 69.53.245.40:22957   ESTABLISHED
[...]
```

tcprerans prints one line of output for every TCP retransmit packet, with details including source and destination addresses, and kernel state of the TCP connection.

TCP retransmissions cause latency and throughput issues. For ESTABLISHED retransmits, look for patterns with networks. For SYN_SENT, this may point to target kernel CPU saturation and kernel packet drops.

More [examples](../tools/tcpretrans_example.txt).

#### 1.10. runqlat

```
# ./runqlat
Tracing run queue latency... Hit Ctrl-C to end.
^C
     usecs               : count     distribution
         0 -> 1          : 233      |***********                             |
         2 -> 3          : 742      |************************************    |
         4 -> 7          : 203      |**********                              |
         8 -> 15         : 173      |********                                |
        16 -> 31         : 24       |*                                       |
        32 -> 63         : 0        |                                        |
        64 -> 127        : 30       |*                                       |
       128 -> 255        : 6        |                                        |
       256 -> 511        : 3        |                                        |
       512 -> 1023       : 5        |                                        |
      1024 -> 2047       : 27       |*                                       |
      2048 -> 4095       : 30       |*                                       |
      4096 -> 8191       : 20       |                                        |
      8192 -> 16383      : 29       |*                                       |
     16384 -> 32767      : 809      |****************************************|
     32768 -> 65535      : 64       |***                                     |
```

runqlat times how long threads were waiting on the CPU run queues, and prints this as a histogram.

This can help quantify time lost waiting for a turn on CPU, during periods of CPU saturation.

More [examples](../tools/runqlat_example.txt).

#### 1.11. profile

```
# ./profile
Sampling at 49 Hertz of all threads by user + kernel stack... Hit Ctrl-C to end.
^C
    00007f31d76c3251 [unknown]
    47a2c1e752bf47f7 [unknown]
    -                sign-file (8877)
        1

    ffffffff813d0af8 __clear_user
    ffffffff813d5277 iov_iter_zero
    ffffffff814ec5f2 read_iter_zero
    ffffffff8120be9d __vfs_read
    ffffffff8120c385 vfs_read
    ffffffff8120d786 sys_read
    ffffffff817cc076 entry_SYSCALL_64_fastpath
    00007fc5652ad9b0 read
    -                dd (25036)
        4

    0000000000400542 func_a
    0000000000400598 main
    00007f12a133e830 __libc_start_main
    083e258d4c544155 [unknown]
    -                func_ab (13549)
        5

[...]

    ffffffff8105eb66 native_safe_halt
    ffffffff8103659e default_idle
    ffffffff81036d1f arch_cpu_idle
    ffffffff810bba5a default_idle_call
    ffffffff810bbd07 cpu_startup_entry
    ffffffff8104df55 start_secondary
    -                swapper/1 (0)
        75
```

profile is a CPU profiler, which takes samples of stack traces at timed intervals, and prints a summary of unique stack traces and a count of their occurrence.

Use this tool to understand the code paths that are consuming CPU resources.

More [examples](../tools/profile_example.txt).

### 2. Observability with Generic Tools

In addition to the above tools for performance tuning, below is a checklist for bcc generic tools, first as a list, and in detail:

1. trace
1. argdist
1. funccount

These generic tools may be useful to provide visibility to solve your specific problems.

#### 2.1. trace

##### Example 1

Suppose you want to track file ownership change. There are three syscalls, `chown`, `fchown` and `lchown` which users can use to change file ownership. The corresponding syscall entry is `SyS_[f|l]chown`.  The following command can be used to print out syscall parameters and the calling process user id. You can use `id` command to find the uid of a particular user.

```
$ trace.py \
  'p::SyS_chown "file = %s, to_uid = %d, to_gid = %d, from_uid = %d", arg1, arg2, arg3, $uid' \
  'p::SyS_fchown "fd = %d, to_uid = %d, to_gid = %d, from_uid = %d", arg1, arg2, arg3, $uid' \
  'p::SyS_lchown "file = %s, to_uid = %d, to_gid = %d, from_uid = %d", arg1, arg2, arg3, $uid'
PID    TID    COMM         FUNC             -
1269255 1269255 python3.6    SyS_lchown       file = /tmp/dotsync-usisgezu/tmp, to_uid = 128203, to_gid = 100, from_uid = 128203
1269441 1269441 zstd         SyS_chown        file = /tmp/dotsync-vic7ygj0/dotsync-package.zst, to_uid = 128203, to_gid = 100, from_uid = 128203
1269255 1269255 python3.6    SyS_lchown       file = /tmp/dotsync-a40zd7ev/tmp, to_uid = 128203, to_gid = 100, from_uid = 128203
1269442 1269442 zstd         SyS_chown        file = /tmp/dotsync-gzp413o_/dotsync-package.zst, to_uid = 128203, to_gid = 100, from_uid = 128203
1269255 1269255 python3.6    SyS_lchown       file = /tmp/dotsync-whx4fivm/tmp/.bash_profile, to_uid = 128203, to_gid = 100, from_uid = 128203
```

##### Example 2

Suppose you want to count nonvoluntary context switches (`nvcsw`) in your bpf based performance monitoring tools and you do not know what is the proper method. `/proc//status` already tells you the number (`nonvoluntary_ctxt_switches`) for a pid and you can use `trace.py` to do a quick experiment to verify your method. With kernel source code, the `nvcsw` is counted at file `linux/kernel/sched/core.c` function `__schedule` and under condition
```
!(!preempt && prev->state) // i.e., preempt || !prev->state
```

The `__schedule` function is marked as `notrace`, and the best place to evaluate the above condition seems in `sched/sched_switch` tracepoint called inside function `__schedule` and defined in `linux/include/trace/events/sched.h`. `trace.py` already has `args` being the pointer to the tracepoint `TP_STRUCT__entry`.  The above condition in function `__schedule` can be represented as
```
args->prev_state == TASK_STATE_MAX || args->prev_state == 0
```

The below command can be used to count the involuntary context switches (per process or per pid) and compare to `/proc//status` or `/proc//task/<task_id>/status` for correctness, as in typical cases, involuntary context switches are not very common.
```
$ trace.py -p 1134138 't:sched:sched_switch (args->prev_state == TASK_STATE_MAX || args->prev_state == 0)'
PID    TID    COMM         FUNC
1134138 1134140 contention_test sched_switch
1134138 1134142 contention_test sched_switch
...
$ trace.py -L 1134140 't:sched:sched_switch (args->prev_state == TASK_STATE_MAX || args->prev_state == 0)'
PID    TID    COMM         FUNC
1134138 1134140 contention_test sched_switch
1134138 1134140 contention_test sched_switch
...
```

##### Example 3

This example is related to issue [1231](https://github.com/iovisor/bcc/issues/1231) and [1516](https://github.com/iovisor/bcc/issues/1516) where uprobe does not work at all in certain cases. First, you can do a `strace` as below

```
$ strace trace.py 'r:bash:readline "%s", retval'
...
perf_event_open(0x7ffd968212f0, -1, 0, -1, 0x8 /* PERF_FLAG_??? */) = -1 EIO (Input/output error)
...
```

The `perf_event_open` syscall returns `-EIO`. Digging into kernel uprobe related codes in `/kernel/trace` and `/kernel/events` directories to search `EIO`, the function `uprobe_register` is the most suspicious. Let us find whether this function is called or not and what is the return value if it is called. In one terminal using the following command to print out the return value of uprobe_register,
```
$ trace.py 'r::uprobe_register "ret = %d", retval'
```
In another terminal run the same bash uretprobe tracing example, and you should get
```
$ trace.py 'r::uprobe_register "ret = %d", retval'
PID    TID    COMM         FUNC             -
1041401 1041401 python2.7    uprobe_register  ret = -5
```

The `-5` error code is EIO. This confirms that the following code in function `uprobe_register` is the most suspicious culprit.
```
 if (!inode->i_mapping->a_ops->readpage && !shmem_mapping(inode->i_mapping))
        return -EIO;
```
The `shmem_mapping` function is defined as
```
bool shmem_mapping(struct address_space *mapping)
{
        return mapping->a_ops == &shmem_aops;
}
```

To confirm the theory, find what is `inode->i_mapping->a_ops` with the following command
```
$ trace.py -I 'linux/fs.h' 'p::uprobe_register(struct inode *inode) "a_ops = %llx", inode->i_mapping->a_ops'
PID    TID    COMM         FUNC             -
814288 814288 python2.7    uprobe_register  a_ops = ffffffff81a2adc0
^C$ grep ffffffff81a2adc0 /proc/kallsyms
ffffffff81a2adc0 R empty_aops
```

The kernel symbol `empty_aops` does not have `readpage` defined and hence the above suspicious condition is true. Further examining the kernel source code shows that `overlayfs` does not provide its own `a_ops` while some other file systems (e.g., ext4) define their own `a_ops` (e.g., `ext4_da_aops`), and `ext4_da_aops` defines `readpage`. Hence, uprobe works fine on ext4 while not on overlayfs.

More [examples](../tools/trace_example.txt).

#### 2.2. argdist

More [examples](../tools/argdist_example.txt).

#### 2.3. funccount

More [examples](../tools/funccount_example.txt).

## Networking

To do.