From 4e10df9a60d96ced321dd2af71da558c6b750078 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Mon, 20 Jul 2015 20:34:18 -0700 Subject: bpf: introduce bpf_skb_vlan_push/pop() helpers Allow eBPF programs attached to TC qdiscs call skb_vlan_push/pop via helper functions. These functions may change skb->data/hlen which are cached by some JITs to improve performance of ld_abs/ld_ind instructions. Therefore JITs need to recognize bpf_skb_vlan_push/pop() calls, re-compute header len and re-cache skb->data/hlen back into cpu registers. Note, skb->data/hlen are not directly accessible from the programs, so any changes to skb->data done either by these helpers or by other TC actions are safe. eBPF JIT supported by three architectures: - arm64 JIT is using bpf_load_pointer() without caching, so it's ok as-is. - x64 JIT re-caches skb->data/hlen unconditionally after vlan_push/pop calls (experiments showed that conditional re-caching is slower). - s390 JIT falls back to interpreter for now when bpf_skb_vlan_push() is present in the program (re-caching is tbd). These helpers allow more scalable handling of vlan from the programs. Instead of creating thousands of vlan netdevs on top of eth0 and attaching TC+ingress+bpf to all of them, the program can be attached to eth0 directly and manipulate vlans as necessary. Signed-off-by: Alexei Starovoitov Signed-off-by: David S. Miller --- include/linux/bpf.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/linux/bpf.h') diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 4383476a0d48..139d6d2e123f 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -192,5 +192,7 @@ extern const struct bpf_func_proto bpf_ktime_get_ns_proto; extern const struct bpf_func_proto bpf_get_current_pid_tgid_proto; extern const struct bpf_func_proto bpf_get_current_uid_gid_proto; extern const struct bpf_func_proto bpf_get_current_comm_proto; +extern const struct bpf_func_proto bpf_skb_vlan_push_proto; +extern const struct bpf_func_proto bpf_skb_vlan_pop_proto; #endif /* _LINUX_BPF_H */ -- cgit v1.2.3-70-g09d2 From 2a36f0b92eb638dd023870574eb471b1c56be9ad Mon Sep 17 00:00:00 2001 From: Wang Nan Date: Thu, 6 Aug 2015 07:02:33 +0000 Subject: bpf: Make the bpf_prog_array_map more generic All the map backends are of generic nature. In order to avoid adding much special code into the eBPF core, rewrite part of the bpf_prog_array map code and make it more generic. So the new perf_event_array map type can reuse most of code with bpf_prog_array map and add fewer lines of special code. Signed-off-by: Wang Nan Signed-off-by: Kaixu Xia Signed-off-by: David S. Miller --- arch/x86/net/bpf_jit_comp.c | 6 ++-- include/linux/bpf.h | 8 +++-- kernel/bpf/arraymap.c | 80 +++++++++++++++++++++++++++------------------ kernel/bpf/core.c | 2 +- kernel/bpf/syscall.c | 2 +- 5 files changed, 60 insertions(+), 38 deletions(-) (limited to 'include/linux/bpf.h') diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c index ec5214f39aa8..70efcd0940f9 100644 --- a/arch/x86/net/bpf_jit_comp.c +++ b/arch/x86/net/bpf_jit_comp.c @@ -246,7 +246,7 @@ static void emit_prologue(u8 **pprog) * goto out; * if (++tail_call_cnt > MAX_TAIL_CALL_CNT) * goto out; - * prog = array->prog[index]; + * prog = array->ptrs[index]; * if (prog == NULL) * goto out; * goto *(prog->bpf_func + prologue_size); @@ -284,9 +284,9 @@ static void emit_bpf_tail_call(u8 **pprog) EMIT3(0x83, 0xC0, 0x01); /* add eax, 1 */ EMIT2_off32(0x89, 0x85, -STACKSIZE + 36); /* mov dword ptr [rbp - 516], eax */ - /* prog = array->prog[index]; */ + /* prog = array->ptrs[index]; */ EMIT4_off32(0x48, 0x8D, 0x84, 0xD6, /* lea rax, [rsi + rdx * 8 + offsetof(...)] */ - offsetof(struct bpf_array, prog)); + offsetof(struct bpf_array, ptrs)); EMIT3(0x48, 0x8B, 0x00); /* mov rax, qword ptr [rax] */ /* if (prog == NULL) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 139d6d2e123f..d495211d63d1 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -24,6 +24,10 @@ struct bpf_map_ops { void *(*map_lookup_elem)(struct bpf_map *map, void *key); int (*map_update_elem)(struct bpf_map *map, void *key, void *value, u64 flags); int (*map_delete_elem)(struct bpf_map *map, void *key); + + /* funcs called by prog_array and perf_event_array map */ + void *(*map_fd_get_ptr) (struct bpf_map *map, int fd); + void (*map_fd_put_ptr) (void *ptr); }; struct bpf_map { @@ -142,13 +146,13 @@ struct bpf_array { bool owner_jited; union { char value[0] __aligned(8); - struct bpf_prog *prog[0] __aligned(8); + void *ptrs[0] __aligned(8); }; }; #define MAX_TAIL_CALL_CNT 32 u64 bpf_tail_call(u64 ctx, u64 r2, u64 index, u64 r4, u64 r5); -void bpf_prog_array_map_clear(struct bpf_map *map); +void bpf_fd_array_map_clear(struct bpf_map *map); bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp); const struct bpf_func_proto *bpf_get_trace_printk_proto(void); diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index cb31229a6fa4..45df6572ecfd 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -150,15 +150,15 @@ static int __init register_array_map(void) } late_initcall(register_array_map); -static struct bpf_map *prog_array_map_alloc(union bpf_attr *attr) +static struct bpf_map *fd_array_map_alloc(union bpf_attr *attr) { - /* only bpf_prog file descriptors can be stored in prog_array map */ + /* only file descriptors can be stored in this type of map */ if (attr->value_size != sizeof(u32)) return ERR_PTR(-EINVAL); return array_map_alloc(attr); } -static void prog_array_map_free(struct bpf_map *map) +static void fd_array_map_free(struct bpf_map *map) { struct bpf_array *array = container_of(map, struct bpf_array, map); int i; @@ -167,21 +167,21 @@ static void prog_array_map_free(struct bpf_map *map) /* make sure it's empty */ for (i = 0; i < array->map.max_entries; i++) - BUG_ON(array->prog[i] != NULL); + BUG_ON(array->ptrs[i] != NULL); kvfree(array); } -static void *prog_array_map_lookup_elem(struct bpf_map *map, void *key) +static void *fd_array_map_lookup_elem(struct bpf_map *map, void *key) { return NULL; } /* only called from syscall */ -static int prog_array_map_update_elem(struct bpf_map *map, void *key, - void *value, u64 map_flags) +static int fd_array_map_update_elem(struct bpf_map *map, void *key, + void *value, u64 map_flags) { struct bpf_array *array = container_of(map, struct bpf_array, map); - struct bpf_prog *prog, *old_prog; + void *new_ptr, *old_ptr; u32 index = *(u32 *)key, ufd; if (map_flags != BPF_ANY) @@ -191,57 +191,75 @@ static int prog_array_map_update_elem(struct bpf_map *map, void *key, return -E2BIG; ufd = *(u32 *)value; - prog = bpf_prog_get(ufd); - if (IS_ERR(prog)) - return PTR_ERR(prog); - - if (!bpf_prog_array_compatible(array, prog)) { - bpf_prog_put(prog); - return -EINVAL; - } + new_ptr = map->ops->map_fd_get_ptr(map, ufd); + if (IS_ERR(new_ptr)) + return PTR_ERR(new_ptr); - old_prog = xchg(array->prog + index, prog); - if (old_prog) - bpf_prog_put_rcu(old_prog); + old_ptr = xchg(array->ptrs + index, new_ptr); + if (old_ptr) + map->ops->map_fd_put_ptr(old_ptr); return 0; } -static int prog_array_map_delete_elem(struct bpf_map *map, void *key) +static int fd_array_map_delete_elem(struct bpf_map *map, void *key) { struct bpf_array *array = container_of(map, struct bpf_array, map); - struct bpf_prog *old_prog; + void *old_ptr; u32 index = *(u32 *)key; if (index >= array->map.max_entries) return -E2BIG; - old_prog = xchg(array->prog + index, NULL); - if (old_prog) { - bpf_prog_put_rcu(old_prog); + old_ptr = xchg(array->ptrs + index, NULL); + if (old_ptr) { + map->ops->map_fd_put_ptr(old_ptr); return 0; } else { return -ENOENT; } } +static void *prog_fd_array_get_ptr(struct bpf_map *map, int fd) +{ + struct bpf_array *array = container_of(map, struct bpf_array, map); + struct bpf_prog *prog = bpf_prog_get(fd); + if (IS_ERR(prog)) + return prog; + + if (!bpf_prog_array_compatible(array, prog)) { + bpf_prog_put(prog); + return ERR_PTR(-EINVAL); + } + return prog; +} + +static void prog_fd_array_put_ptr(void *ptr) +{ + struct bpf_prog *prog = ptr; + + bpf_prog_put_rcu(prog); +} + /* decrement refcnt of all bpf_progs that are stored in this map */ -void bpf_prog_array_map_clear(struct bpf_map *map) +void bpf_fd_array_map_clear(struct bpf_map *map) { struct bpf_array *array = container_of(map, struct bpf_array, map); int i; for (i = 0; i < array->map.max_entries; i++) - prog_array_map_delete_elem(map, &i); + fd_array_map_delete_elem(map, &i); } static const struct bpf_map_ops prog_array_ops = { - .map_alloc = prog_array_map_alloc, - .map_free = prog_array_map_free, + .map_alloc = fd_array_map_alloc, + .map_free = fd_array_map_free, .map_get_next_key = array_map_get_next_key, - .map_lookup_elem = prog_array_map_lookup_elem, - .map_update_elem = prog_array_map_update_elem, - .map_delete_elem = prog_array_map_delete_elem, + .map_lookup_elem = fd_array_map_lookup_elem, + .map_update_elem = fd_array_map_update_elem, + .map_delete_elem = fd_array_map_delete_elem, + .map_fd_get_ptr = prog_fd_array_get_ptr, + .map_fd_put_ptr = prog_fd_array_put_ptr, }; static struct bpf_map_type_list prog_array_type __read_mostly = { diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index fafa74161445..67c380cfa9ca 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -450,7 +450,7 @@ select_insn: tail_call_cnt++; - prog = READ_ONCE(array->prog[index]); + prog = READ_ONCE(array->ptrs[index]); if (unlikely(!prog)) goto out; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index a1b14d197a4f..dc9b464fefa9 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -72,7 +72,7 @@ static int bpf_map_release(struct inode *inode, struct file *filp) /* prog_array stores refcnt-ed bpf_prog pointers * release them all when user space closes prog_array_fd */ - bpf_prog_array_map_clear(map); + bpf_fd_array_map_clear(map); bpf_map_put(map); return 0; -- cgit v1.2.3-70-g09d2 From ea317b267e9d03a8241893aa176fba7661d07579 Mon Sep 17 00:00:00 2001 From: Kaixu Xia Date: Thu, 6 Aug 2015 07:02:34 +0000 Subject: bpf: Add new bpf map type to store the pointer to struct perf_event Introduce a new bpf map type 'BPF_MAP_TYPE_PERF_EVENT_ARRAY'. This map only stores the pointer to struct perf_event. The user space event FDs from perf_event_open() syscall are converted to the pointer to struct perf_event and stored in map. Signed-off-by: Kaixu Xia Signed-off-by: David S. Miller --- include/linux/bpf.h | 1 + include/uapi/linux/bpf.h | 1 + kernel/bpf/arraymap.c | 57 ++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 59 insertions(+) (limited to 'include/linux/bpf.h') diff --git a/include/linux/bpf.h b/include/linux/bpf.h index d495211d63d1..4fc1f4070789 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -10,6 +10,7 @@ #include #include #include +#include struct bpf_map; diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 2ce13c109b00..a1814e8e53a7 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -114,6 +114,7 @@ enum bpf_map_type { BPF_MAP_TYPE_HASH, BPF_MAP_TYPE_ARRAY, BPF_MAP_TYPE_PROG_ARRAY, + BPF_MAP_TYPE_PERF_EVENT_ARRAY, }; enum bpf_prog_type { diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 45df6572ecfd..29ace107f236 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -273,3 +273,60 @@ static int __init register_prog_array_map(void) return 0; } late_initcall(register_prog_array_map); + +static void perf_event_array_map_free(struct bpf_map *map) +{ + bpf_fd_array_map_clear(map); + fd_array_map_free(map); +} + +static void *perf_event_fd_array_get_ptr(struct bpf_map *map, int fd) +{ + struct perf_event *event; + const struct perf_event_attr *attr; + + event = perf_event_get(fd); + if (IS_ERR(event)) + return event; + + attr = perf_event_attrs(event); + if (IS_ERR(attr)) + return (void *)attr; + + if (attr->type != PERF_TYPE_RAW && + attr->type != PERF_TYPE_HARDWARE) { + perf_event_release_kernel(event); + return ERR_PTR(-EINVAL); + } + return event; +} + +static void perf_event_fd_array_put_ptr(void *ptr) +{ + struct perf_event *event = ptr; + + perf_event_release_kernel(event); +} + +static const struct bpf_map_ops perf_event_array_ops = { + .map_alloc = fd_array_map_alloc, + .map_free = perf_event_array_map_free, + .map_get_next_key = array_map_get_next_key, + .map_lookup_elem = fd_array_map_lookup_elem, + .map_update_elem = fd_array_map_update_elem, + .map_delete_elem = fd_array_map_delete_elem, + .map_fd_get_ptr = perf_event_fd_array_get_ptr, + .map_fd_put_ptr = perf_event_fd_array_put_ptr, +}; + +static struct bpf_map_type_list perf_event_array_type __read_mostly = { + .ops = &perf_event_array_ops, + .type = BPF_MAP_TYPE_PERF_EVENT_ARRAY, +}; + +static int __init register_perf_event_array_map(void) +{ + bpf_register_map_type(&perf_event_array_type); + return 0; +} +late_initcall(register_perf_event_array_map); -- cgit v1.2.3-70-g09d2 From 35578d7984003097af2b1e34502bc943d40c1804 Mon Sep 17 00:00:00 2001 From: Kaixu Xia Date: Thu, 6 Aug 2015 07:02:35 +0000 Subject: bpf: Implement function bpf_perf_event_read() that get the selected hardware PMU conuter According to the perf_event_map_fd and index, the function bpf_perf_event_read() can convert the corresponding map value to the pointer to struct perf_event and return the Hardware PMU counter value. Signed-off-by: Kaixu Xia Signed-off-by: David S. Miller --- include/linux/bpf.h | 1 + include/uapi/linux/bpf.h | 1 + kernel/bpf/verifier.c | 48 +++++++++++++++++++++++++++++++++--------------- kernel/trace/bpf_trace.c | 31 +++++++++++++++++++++++++++++++ 4 files changed, 66 insertions(+), 15 deletions(-) (limited to 'include/linux/bpf.h') diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 4fc1f4070789..f57d7fed9ec3 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -190,6 +190,7 @@ extern const struct bpf_func_proto bpf_map_lookup_elem_proto; extern const struct bpf_func_proto bpf_map_update_elem_proto; extern const struct bpf_func_proto bpf_map_delete_elem_proto; +extern const struct bpf_func_proto bpf_perf_event_read_proto; extern const struct bpf_func_proto bpf_get_prandom_u32_proto; extern const struct bpf_func_proto bpf_get_smp_processor_id_proto; extern const struct bpf_func_proto bpf_tail_call_proto; diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index a1814e8e53a7..92a48e2d5461 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -271,6 +271,7 @@ enum bpf_func_id { */ BPF_FUNC_skb_get_tunnel_key, BPF_FUNC_skb_set_tunnel_key, + BPF_FUNC_perf_event_read, /* u64 bpf_perf_event_read(&map, index) */ __BPF_FUNC_MAX_ID, }; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index cd307df98cb3..48e1c7192560 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -238,6 +238,14 @@ static const char * const reg_type_str[] = { [CONST_IMM] = "imm", }; +static const struct { + int map_type; + int func_id; +} func_limit[] = { + {BPF_MAP_TYPE_PROG_ARRAY, BPF_FUNC_tail_call}, + {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_read}, +}; + static void print_verifier_state(struct verifier_env *env) { enum bpf_reg_type t; @@ -837,6 +845,28 @@ static int check_func_arg(struct verifier_env *env, u32 regno, return err; } +static int check_map_func_compatibility(struct bpf_map *map, int func_id) +{ + bool bool_map, bool_func; + int i; + + if (!map) + return 0; + + for (i = 0; i <= ARRAY_SIZE(func_limit); i++) { + bool_map = (map->map_type == func_limit[i].map_type); + bool_func = (func_id == func_limit[i].func_id); + /* only when map & func pair match it can continue. + * don't allow any other map type to be passed into + * the special func; + */ + if (bool_map != bool_func) + return -EINVAL; + } + + return 0; +} + static int check_call(struct verifier_env *env, int func_id) { struct verifier_state *state = &env->cur_state; @@ -912,21 +942,9 @@ static int check_call(struct verifier_env *env, int func_id) return -EINVAL; } - if (map && map->map_type == BPF_MAP_TYPE_PROG_ARRAY && - func_id != BPF_FUNC_tail_call) - /* prog_array map type needs extra care: - * only allow to pass it into bpf_tail_call() for now. - * bpf_map_delete_elem() can be allowed in the future, - * while bpf_map_update_elem() must only be done via syscall - */ - return -EINVAL; - - if (func_id == BPF_FUNC_tail_call && - map->map_type != BPF_MAP_TYPE_PROG_ARRAY) - /* don't allow any other map type to be passed into - * bpf_tail_call() - */ - return -EINVAL; + err = check_map_func_compatibility(map, func_id); + if (err) + return err; return 0; } diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 88a041adee90..ef9936df1b04 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -158,6 +158,35 @@ const struct bpf_func_proto *bpf_get_trace_printk_proto(void) return &bpf_trace_printk_proto; } +static u64 bpf_perf_event_read(u64 r1, u64 index, u64 r3, u64 r4, u64 r5) +{ + struct bpf_map *map = (struct bpf_map *) (unsigned long) r1; + struct bpf_array *array = container_of(map, struct bpf_array, map); + struct perf_event *event; + + if (unlikely(index >= array->map.max_entries)) + return -E2BIG; + + event = (struct perf_event *)array->ptrs[index]; + if (!event) + return -ENOENT; + + /* + * we don't know if the function is run successfully by the + * return value. It can be judged in other places, such as + * eBPF programs. + */ + return perf_event_read_local(event); +} + +const struct bpf_func_proto bpf_perf_event_read_proto = { + .func = bpf_perf_event_read, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_CONST_MAP_PTR, + .arg2_type = ARG_ANYTHING, +}; + static const struct bpf_func_proto *kprobe_prog_func_proto(enum bpf_func_id func_id) { switch (func_id) { @@ -183,6 +212,8 @@ static const struct bpf_func_proto *kprobe_prog_func_proto(enum bpf_func_id func return bpf_get_trace_printk_proto(); case BPF_FUNC_get_smp_processor_id: return &bpf_get_smp_processor_id_proto; + case BPF_FUNC_perf_event_read: + return &bpf_perf_event_read_proto; default: return NULL; } -- cgit v1.2.3-70-g09d2