1 # SPDX-License-Identifier: GPL-2.0-only
3 # BPF interpreter that, for example, classic socket filters depend on.
7 # Used by archs to tell that they support BPF JIT compiler plus which
8 # flavour. Only one of the two can be selected for a specific arch since
9 # eBPF JIT supersedes the cBPF JIT.
11 # Classic BPF JIT (cBPF)
15 # Extended BPF JIT (eBPF)
19 # Used by archs to tell that they want the BPF JIT compiler enabled by
20 # default for kernels that were compiled with BPF JIT support.
21 config ARCH_WANT_DEFAULT_BPF_JIT
27 bool "Enable bpf() system call"
30 select TASKS_RCU if PREEMPTION
31 select TASKS_TRACE_RCU
33 select NET_SOCK_MSG if NET
34 select PAGE_POOL if NET
37 Enable the bpf() system call that allows to manipulate BPF programs
38 and maps via file descriptors.
41 bool "Enable BPF Just In Time compiler"
43 depends on HAVE_CBPF_JIT || HAVE_EBPF_JIT
46 BPF programs are normally handled by a BPF interpreter. This option
47 allows the kernel to generate native code when a program is loaded
48 into the kernel. This will significantly speed-up processing of BPF
51 Note, an admin should enable this feature changing:
52 /proc/sys/net/core/bpf_jit_enable
53 /proc/sys/net/core/bpf_jit_harden (optional)
54 /proc/sys/net/core/bpf_jit_kallsyms (optional)
56 config BPF_JIT_ALWAYS_ON
57 bool "Permanently enable BPF JIT and remove BPF interpreter"
58 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
60 Enables BPF JIT and removes BPF interpreter to avoid speculative
61 execution of BPF instructions by the interpreter.
63 When CONFIG_BPF_JIT_ALWAYS_ON is enabled, /proc/sys/net/core/bpf_jit_enable
64 is permanently set to 1 and setting any other value than that will
67 config BPF_JIT_DEFAULT_ON
68 def_bool ARCH_WANT_DEFAULT_BPF_JIT || BPF_JIT_ALWAYS_ON
69 depends on HAVE_EBPF_JIT && BPF_JIT
71 config BPF_UNPRIV_DEFAULT_OFF
72 bool "Disable unprivileged BPF by default"
74 depends on BPF_SYSCALL
76 Disables unprivileged BPF by default by setting the corresponding
77 /proc/sys/kernel/unprivileged_bpf_disabled knob to 2. An admin can
78 still reenable it by setting it to 0 later on, or permanently
79 disable it by setting it to 1 (from which no other transition to
80 0 is possible anymore).
82 Unprivileged BPF could be used to exploit certain potential
83 speculative execution side-channel vulnerabilities on unmitigated
86 If you are unsure how to answer this question, answer Y.
88 source "kernel/bpf/preload/Kconfig"
91 bool "Enable BPF LSM Instrumentation"
93 depends on BPF_SYSCALL
97 Enables instrumentation of the security hooks with BPF programs for
98 implementing dynamic MAC and Audit Policies.
100 If you are unsure how to answer this question, answer N.
102 endmenu # "BPF subsystem"