From e542f2c4cd16d49392abf3349341d58153d3c603 Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Tue, 14 Dec 2021 11:59:03 -0800 Subject: libbpf: Auto-bump RLIMIT_MEMLOCK if kernel needs it for BPF The need to increase RLIMIT_MEMLOCK to do anything useful with BPF is one of the first extremely frustrating gotchas that all new BPF users go through and in some cases have to learn it a very hard way. Luckily, starting with upstream Linux kernel version 5.11, BPF subsystem dropped the dependency on memlock and uses memcg-based memory accounting instead. Unfortunately, detecting memcg-based BPF memory accounting is far from trivial (as can be evidenced by this patch), so in practice most BPF applications still do unconditional RLIMIT_MEMLOCK increase. As we move towards libbpf 1.0, it would be good to allow users to forget about RLIMIT_MEMLOCK vs memcg and let libbpf do the sensible adjustment automatically. This patch paves the way forward in this matter. Libbpf will do feature detection of memcg-based accounting, and if detected, will do nothing. But if the kernel is too old, just like BCC, libbpf will automatically increase RLIMIT_MEMLOCK on behalf of user application ([0]). As this is technically a breaking change, during the transition period applications have to opt into libbpf 1.0 mode by setting LIBBPF_STRICT_AUTO_RLIMIT_MEMLOCK bit when calling libbpf_set_strict_mode(). Libbpf allows to control the exact amount of set RLIMIT_MEMLOCK limit with libbpf_set_memlock_rlim_max() API. Passing 0 will make libbpf do nothing with RLIMIT_MEMLOCK. libbpf_set_memlock_rlim_max() has to be called before the first bpf_prog_load(), bpf_btf_load(), or bpf_object__load() call, otherwise it has no effect and will return -EBUSY. [0] Closes: https://github.com/libbpf/libbpf/issues/369 Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20211214195904.1785155-2-andrii@kernel.org --- tools/lib/bpf/libbpf_internal.h | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) (limited to 'tools/lib/bpf/libbpf_internal.h') diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h index 5e8166a2f3d8..5dbe4f463880 100644 --- a/tools/lib/bpf/libbpf_internal.h +++ b/tools/lib/bpf/libbpf_internal.h @@ -291,6 +291,45 @@ static inline bool libbpf_validate_opts(const char *opts, (opts)->sz - __off); \ }) +enum kern_feature_id { + /* v4.14: kernel support for program & map names. */ + FEAT_PROG_NAME, + /* v5.2: kernel support for global data sections. */ + FEAT_GLOBAL_DATA, + /* BTF support */ + FEAT_BTF, + /* BTF_KIND_FUNC and BTF_KIND_FUNC_PROTO support */ + FEAT_BTF_FUNC, + /* BTF_KIND_VAR and BTF_KIND_DATASEC support */ + FEAT_BTF_DATASEC, + /* BTF_FUNC_GLOBAL is supported */ + FEAT_BTF_GLOBAL_FUNC, + /* BPF_F_MMAPABLE is supported for arrays */ + FEAT_ARRAY_MMAP, + /* kernel support for expected_attach_type in BPF_PROG_LOAD */ + FEAT_EXP_ATTACH_TYPE, + /* bpf_probe_read_{kernel,user}[_str] helpers */ + FEAT_PROBE_READ_KERN, + /* BPF_PROG_BIND_MAP is supported */ + FEAT_PROG_BIND_MAP, + /* Kernel support for module BTFs */ + FEAT_MODULE_BTF, + /* BTF_KIND_FLOAT support */ + FEAT_BTF_FLOAT, + /* BPF perf link support */ + FEAT_PERF_LINK, + /* BTF_KIND_DECL_TAG support */ + FEAT_BTF_DECL_TAG, + /* BTF_KIND_TYPE_TAG support */ + FEAT_BTF_TYPE_TAG, + /* memcg-based accounting for BPF maps and progs */ + FEAT_MEMCG_ACCOUNT, + __FEAT_CNT, +}; + +int probe_memcg_account(void); +bool kernel_supports(const struct bpf_object *obj, enum kern_feature_id feat_id); +int bump_rlimit_memlock(void); int parse_cpu_mask_str(const char *s, bool **mask, int *mask_sz); int parse_cpu_mask_file(const char *fcpu, bool **mask, int *mask_sz); -- cgit v1.2.3