From b1cd609d9b517f01867c211bd520cc805db3068a Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Tue, 12 Mar 2019 09:27:09 -0700 Subject: [PATCH 01/68] bpf: Add base proto function for cgroup-bpf programs Currently kernel/bpf/cgroup.c contains only one program type and one proto function cgroup_dev_func_proto(). It'd be useful to have base proto function that can be reused for new cgroup-bpf program types coming soon. Introduce cgroup_base_func_proto(). Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- kernel/bpf/cgroup.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index 4e807973aa80..f6cd38746df2 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -701,7 +701,7 @@ int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor, EXPORT_SYMBOL(__cgroup_bpf_check_dev_permission); static const struct bpf_func_proto * -cgroup_dev_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +cgroup_base_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { switch (func_id) { case BPF_FUNC_map_lookup_elem: @@ -725,6 +725,12 @@ cgroup_dev_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) } } +static const struct bpf_func_proto * +cgroup_dev_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +{ + return cgroup_base_func_proto(func_id, prog); +} + static bool cgroup_dev_is_valid_access(int off, int size, enum bpf_access_type type, const struct bpf_prog *prog, From 7b146cebe30cb481b0f70d85779da938da818637 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Wed, 27 Feb 2019 12:59:24 -0800 Subject: [PATCH 02/68] bpf: Sysctl hook Containerized applications may run as root and it may create problems for whole host. Specifically such applications may change a sysctl and affect applications in other containers. Furthermore in existing infrastructure it may not be possible to just completely disable writing to sysctl, instead such a process should be gradual with ability to log what sysctl are being changed by a container, investigate, limit the set of writable sysctl to currently used ones (so that new ones can not be changed) and eventually reduce this set to zero. The patch introduces new program type BPF_PROG_TYPE_CGROUP_SYSCTL and attach type BPF_CGROUP_SYSCTL to solve these problems on cgroup basis. New program type has access to following minimal context: struct bpf_sysctl { __u32 write; }; Where @write indicates whether sysctl is being read (= 0) or written (= 1). Helpers to access sysctl name and value will be introduced separately. BPF_CGROUP_SYSCTL attach point is added to sysctl code right before passing control to ctl_table->proc_handler so that BPF program can either allow or deny access to sysctl. Suggested-by: Roman Gushchin Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- fs/proc/proc_sysctl.c | 5 +++ include/linux/bpf-cgroup.h | 18 ++++++++ include/linux/bpf_types.h | 1 + include/linux/filter.h | 8 ++++ include/uapi/linux/bpf.h | 9 ++++ kernel/bpf/cgroup.c | 92 ++++++++++++++++++++++++++++++++++++++ kernel/bpf/syscall.c | 7 +++ kernel/bpf/verifier.c | 1 + 8 files changed, 141 insertions(+) diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index d65390727541..e01b02150340 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -13,6 +13,7 @@ #include #include #include +#include #include "internal.h" static const struct dentry_operations proc_sys_dentry_operations; @@ -588,6 +589,10 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, if (!table->proc_handler) goto out; + error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write); + if (error) + goto out; + /* careful: calling conventions are nasty here */ res = count; error = table->proc_handler(table, write, buf, &res, ppos); diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index a4c644c1c091..b1c45da20a26 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -17,6 +17,8 @@ struct bpf_map; struct bpf_prog; struct bpf_sock_ops_kern; struct bpf_cgroup_storage; +struct ctl_table; +struct ctl_table_header; #ifdef CONFIG_CGROUP_BPF @@ -109,6 +111,10 @@ int __cgroup_bpf_run_filter_sock_ops(struct sock *sk, int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor, short access, enum bpf_attach_type type); +int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, + struct ctl_table *table, int write, + enum bpf_attach_type type); + static inline enum bpf_cgroup_storage_type cgroup_storage_type( struct bpf_map *map) { @@ -253,6 +259,17 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key, \ __ret; \ }) + + +#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write) \ +({ \ + int __ret = 0; \ + if (cgroup_bpf_enabled) \ + __ret = __cgroup_bpf_run_filter_sysctl(head, table, write, \ + BPF_CGROUP_SYSCTL); \ + __ret; \ +}) + int cgroup_bpf_prog_attach(const union bpf_attr *attr, enum bpf_prog_type ptype, struct bpf_prog *prog); int cgroup_bpf_prog_detach(const union bpf_attr *attr, @@ -321,6 +338,7 @@ static inline int bpf_percpu_cgroup_storage_update(struct bpf_map *map, #define BPF_CGROUP_RUN_PROG_UDP6_SENDMSG_LOCK(sk, uaddr, t_ctx) ({ 0; }) #define BPF_CGROUP_RUN_PROG_SOCK_OPS(sock_ops) ({ 0; }) #define BPF_CGROUP_RUN_PROG_DEVICE_CGROUP(type,major,minor,access) ({ 0; }) +#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write) ({ 0; }) #define for_each_cgroup_storage_type(stype) for (; false; ) diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h index 08bf2f1fe553..d26991a16894 100644 --- a/include/linux/bpf_types.h +++ b/include/linux/bpf_types.h @@ -28,6 +28,7 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT, raw_tracepoint) #endif #ifdef CONFIG_CGROUP_BPF BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_DEVICE, cg_dev) +BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_SYSCTL, cg_sysctl) #endif #ifdef CONFIG_BPF_LIRC_MODE2 BPF_PROG_TYPE(BPF_PROG_TYPE_LIRC_MODE2, lirc_mode2) diff --git a/include/linux/filter.h b/include/linux/filter.h index 6074aa064b54..a17732057880 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -33,6 +33,8 @@ struct bpf_prog_aux; struct xdp_rxq_info; struct xdp_buff; struct sock_reuseport; +struct ctl_table; +struct ctl_table_header; /* ArgX, context and stack frame pointer register positions. Note, * Arg1, Arg2, Arg3, etc are used as argument mappings of function @@ -1177,4 +1179,10 @@ struct bpf_sock_ops_kern { */ }; +struct bpf_sysctl_kern { + struct ctl_table_header *head; + struct ctl_table *table; + int write; +}; + #endif /* __LINUX_FILTER_H__ */ diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 2e96d0b4bf65..cc2a2466d5f3 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -167,6 +167,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_LIRC_MODE2, BPF_PROG_TYPE_SK_REUSEPORT, BPF_PROG_TYPE_FLOW_DISSECTOR, + BPF_PROG_TYPE_CGROUP_SYSCTL, }; enum bpf_attach_type { @@ -188,6 +189,7 @@ enum bpf_attach_type { BPF_CGROUP_UDP6_SENDMSG, BPF_LIRC_MODE2, BPF_FLOW_DISSECTOR, + BPF_CGROUP_SYSCTL, __MAX_BPF_ATTACH_TYPE }; @@ -3308,4 +3310,11 @@ struct bpf_line_info { struct bpf_spin_lock { __u32 val; }; + +struct bpf_sysctl { + __u32 write; /* Sysctl is being read (= 0) or written (= 1). + * Allows 1,2,4-byte read, but no write. + */ +}; + #endif /* _UAPI__LINUX_BPF_H__ */ diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index f6cd38746df2..610491b5f0aa 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -11,7 +11,9 @@ #include #include #include +#include #include +#include #include #include #include @@ -768,3 +770,93 @@ const struct bpf_verifier_ops cg_dev_verifier_ops = { .get_func_proto = cgroup_dev_func_proto, .is_valid_access = cgroup_dev_is_valid_access, }; + +/** + * __cgroup_bpf_run_filter_sysctl - Run a program on sysctl + * + * @head: sysctl table header + * @table: sysctl table + * @write: sysctl is being read (= 0) or written (= 1) + * @type: type of program to be executed + * + * Program is run when sysctl is being accessed, either read or written, and + * can allow or deny such access. + * + * This function will return %-EPERM if an attached program is found and + * returned value != 1 during execution. In all other cases 0 is returned. + */ +int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, + struct ctl_table *table, int write, + enum bpf_attach_type type) +{ + struct bpf_sysctl_kern ctx = { + .head = head, + .table = table, + .write = write, + }; + struct cgroup *cgrp; + int ret; + + rcu_read_lock(); + cgrp = task_dfl_cgroup(current); + ret = BPF_PROG_RUN_ARRAY(cgrp->bpf.effective[type], &ctx, BPF_PROG_RUN); + rcu_read_unlock(); + + return ret == 1 ? 0 : -EPERM; +} +EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl); + +static const struct bpf_func_proto * +sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +{ + return cgroup_base_func_proto(func_id, prog); +} + +static bool sysctl_is_valid_access(int off, int size, enum bpf_access_type type, + const struct bpf_prog *prog, + struct bpf_insn_access_aux *info) +{ + const int size_default = sizeof(__u32); + + if (off < 0 || off + size > sizeof(struct bpf_sysctl) || + off % size || type != BPF_READ) + return false; + + switch (off) { + case offsetof(struct bpf_sysctl, write): + bpf_ctx_record_field_size(info, size_default); + return bpf_ctx_narrow_access_ok(off, size, size_default); + default: + return false; + } +} + +static u32 sysctl_convert_ctx_access(enum bpf_access_type type, + const struct bpf_insn *si, + struct bpf_insn *insn_buf, + struct bpf_prog *prog, u32 *target_size) +{ + struct bpf_insn *insn = insn_buf; + + switch (si->off) { + case offsetof(struct bpf_sysctl, write): + *insn++ = BPF_LDX_MEM( + BPF_SIZE(si->code), si->dst_reg, si->src_reg, + bpf_target_off(struct bpf_sysctl_kern, write, + FIELD_SIZEOF(struct bpf_sysctl_kern, + write), + target_size)); + break; + } + + return insn - insn_buf; +} + +const struct bpf_verifier_ops cg_sysctl_verifier_ops = { + .get_func_proto = sysctl_func_proto, + .is_valid_access = sysctl_is_valid_access, + .convert_ctx_access = sysctl_convert_ctx_access, +}; + +const struct bpf_prog_ops cg_sysctl_prog_ops = { +}; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index d995eedfdd16..92c9b8a32b50 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1888,6 +1888,9 @@ static int bpf_prog_attach(const union bpf_attr *attr) case BPF_FLOW_DISSECTOR: ptype = BPF_PROG_TYPE_FLOW_DISSECTOR; break; + case BPF_CGROUP_SYSCTL: + ptype = BPF_PROG_TYPE_CGROUP_SYSCTL; + break; default: return -EINVAL; } @@ -1966,6 +1969,9 @@ static int bpf_prog_detach(const union bpf_attr *attr) return lirc_prog_detach(attr); case BPF_FLOW_DISSECTOR: return skb_flow_dissector_bpf_prog_detach(attr); + case BPF_CGROUP_SYSCTL: + ptype = BPF_PROG_TYPE_CGROUP_SYSCTL; + break; default: return -EINVAL; } @@ -1999,6 +2005,7 @@ static int bpf_prog_query(const union bpf_attr *attr, case BPF_CGROUP_UDP6_SENDMSG: case BPF_CGROUP_SOCK_OPS: case BPF_CGROUP_DEVICE: + case BPF_CGROUP_SYSCTL: break; case BPF_LIRC_MODE2: return lirc_prog_query(attr, uattr); diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index f25b7c9c20ba..20808e3c95a8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5267,6 +5267,7 @@ static int check_return_code(struct bpf_verifier_env *env) case BPF_PROG_TYPE_CGROUP_SOCK_ADDR: case BPF_PROG_TYPE_SOCK_OPS: case BPF_PROG_TYPE_CGROUP_DEVICE: + case BPF_PROG_TYPE_CGROUP_SYSCTL: break; default: return 0; From 808649fb787d918a48a360a668ee4ee9023f0c11 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Wed, 27 Feb 2019 13:28:48 -0800 Subject: [PATCH 03/68] bpf: Introduce bpf_sysctl_get_name helper Add bpf_sysctl_get_name() helper to copy sysctl name (/proc/sys/ entry) into provided by BPF_PROG_TYPE_CGROUP_SYSCTL program buffer. By default full name (w/o /proc/sys/) is copied, e.g. "net/ipv4/tcp_mem". If BPF_F_SYSCTL_BASE_NAME flag is set, only base name will be copied, e.g. "tcp_mem". Documentation for the new helper is provided in bpf.h UAPI. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- include/uapi/linux/bpf.h | 22 ++++++++++++- kernel/bpf/cgroup.c | 70 +++++++++++++++++++++++++++++++++++++++- 2 files changed, 90 insertions(+), 2 deletions(-) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index cc2a2466d5f3..9c8a2f3ccb9b 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2506,6 +2506,22 @@ union bpf_attr { * Return * 0 if iph and th are a valid SYN cookie ACK, or a negative error * otherwise. + * + * int bpf_sysctl_get_name(struct bpf_sysctl *ctx, char *buf, size_t buf_len, u64 flags) + * Description + * Get name of sysctl in /proc/sys/ and copy it into provided by + * program buffer *buf* of size *buf_len*. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * + * If *flags* is zero, full name (e.g. "net/ipv4/tcp_mem") is + * copied. Use **BPF_F_SYSCTL_BASE_NAME** flag to copy base name + * only (e.g. "tcp_mem"). + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2608,7 +2624,8 @@ union bpf_attr { FN(skb_ecn_set_ce), \ FN(get_listener_sock), \ FN(skc_lookup_tcp), \ - FN(tcp_check_syncookie), + FN(tcp_check_syncookie), \ + FN(sysctl_get_name), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call @@ -2681,6 +2698,9 @@ enum bpf_func_id { BPF_ADJ_ROOM_ENCAP_L2_MASK) \ << BPF_ADJ_ROOM_ENCAP_L2_SHIFT) +/* BPF_FUNC_sysctl_get_name flags. */ +#define BPF_F_SYSCTL_BASE_NAME (1ULL << 0) + /* Mode for BPF_FUNC_skb_adjust_room helper. */ enum bpf_adj_room_mode { BPF_ADJ_ROOM_NET, diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index 610491b5f0aa..a68387043244 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include #include @@ -806,10 +807,77 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, } EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl); +static ssize_t sysctl_cpy_dir(const struct ctl_dir *dir, char **bufp, + size_t *lenp) +{ + ssize_t tmp_ret = 0, ret; + + if (dir->header.parent) { + tmp_ret = sysctl_cpy_dir(dir->header.parent, bufp, lenp); + if (tmp_ret < 0) + return tmp_ret; + } + + ret = strscpy(*bufp, dir->header.ctl_table[0].procname, *lenp); + if (ret < 0) + return ret; + *bufp += ret; + *lenp -= ret; + ret += tmp_ret; + + /* Avoid leading slash. */ + if (!ret) + return ret; + + tmp_ret = strscpy(*bufp, "/", *lenp); + if (tmp_ret < 0) + return tmp_ret; + *bufp += tmp_ret; + *lenp -= tmp_ret; + + return ret + tmp_ret; +} + +BPF_CALL_4(bpf_sysctl_get_name, struct bpf_sysctl_kern *, ctx, char *, buf, + size_t, buf_len, u64, flags) +{ + ssize_t tmp_ret = 0, ret; + + if (!buf) + return -EINVAL; + + if (!(flags & BPF_F_SYSCTL_BASE_NAME)) { + if (!ctx->head) + return -EINVAL; + tmp_ret = sysctl_cpy_dir(ctx->head->parent, &buf, &buf_len); + if (tmp_ret < 0) + return tmp_ret; + } + + ret = strscpy(buf, ctx->table->procname, buf_len); + + return ret < 0 ? ret : tmp_ret + ret; +} + +static const struct bpf_func_proto bpf_sysctl_get_name_proto = { + .func = bpf_sysctl_get_name, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_PTR_TO_MEM, + .arg3_type = ARG_CONST_SIZE, + .arg4_type = ARG_ANYTHING, +}; + static const struct bpf_func_proto * sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { - return cgroup_base_func_proto(func_id, prog); + switch (func_id) { + case BPF_FUNC_sysctl_get_name: + return &bpf_sysctl_get_name_proto; + default: + return cgroup_base_func_proto(func_id, prog); + } } static bool sysctl_is_valid_access(int off, int size, enum bpf_access_type type, From 1d11b3016cec4ed9770b98e82a61708c8f4926e7 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Thu, 28 Feb 2019 19:22:15 -0800 Subject: [PATCH 04/68] bpf: Introduce bpf_sysctl_get_current_value helper Add bpf_sysctl_get_current_value() helper to copy current sysctl value into provided by BPF_PROG_TYPE_CGROUP_SYSCTL program buffer. It provides same string as user space can see by reading corresponding file in /proc/sys/, including new line, etc. Documentation for the new helper is provided in bpf.h UAPI. Since current value is kept in ctl_table->data in a parsed form, ctl_table->proc_handler() with write=0 is called to read that data and convert it to a string. Such a string can later be parsed by a program using helpers that will be introduced separately. Unfortunately it's not trivial to provide API to access parsed data due to variety of data representations (string, intvec, uintvec, ulongvec, custom structures, even NULL, etc). Instead it's assumed that user know how to handle specific sysctl they're interested in and appropriate helpers can be used. Since ctl_table->proc_handler() expects __user buffer, conversion to __user happens for kernel allocated one where the value is stored. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- include/linux/filter.h | 2 ++ include/uapi/linux/bpf.h | 22 +++++++++++++- kernel/bpf/cgroup.c | 65 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 88 insertions(+), 1 deletion(-) diff --git a/include/linux/filter.h b/include/linux/filter.h index a17732057880..f254ff92819f 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -1182,6 +1182,8 @@ struct bpf_sock_ops_kern { struct bpf_sysctl_kern { struct ctl_table_header *head; struct ctl_table *table; + void *cur_val; + size_t cur_len; int write; }; diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 9c8a2f3ccb9b..063543afa359 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2522,6 +2522,25 @@ union bpf_attr { * * **-E2BIG** if the buffer wasn't big enough (*buf* will contain * truncated name in this case). + * + * int bpf_sysctl_get_current_value(struct bpf_sysctl *ctx, char *buf, size_t buf_len) + * Description + * Get current value of sysctl as it is presented in /proc/sys + * (incl. newline, etc), and copy it as a string into provided + * by program buffer *buf* of size *buf_len*. + * + * The whole value is copied, no matter what file position user + * space issued e.g. sys_read at. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). + * + * **-EINVAL** if current value was unavailable, e.g. because + * sysctl is uninitialized and read returns -EIO for it. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2625,7 +2644,8 @@ union bpf_attr { FN(get_listener_sock), \ FN(skc_lookup_tcp), \ FN(tcp_check_syncookie), \ - FN(sysctl_get_name), + FN(sysctl_get_name), \ + FN(sysctl_get_current_value), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index a68387043244..c6b2cf29a54b 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -794,15 +794,37 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, .head = head, .table = table, .write = write, + .cur_val = NULL, + .cur_len = PAGE_SIZE, }; struct cgroup *cgrp; int ret; + ctx.cur_val = kmalloc_track_caller(ctx.cur_len, GFP_KERNEL); + if (ctx.cur_val) { + mm_segment_t old_fs; + loff_t pos = 0; + + old_fs = get_fs(); + set_fs(KERNEL_DS); + if (table->proc_handler(table, 0, (void __user *)ctx.cur_val, + &ctx.cur_len, &pos)) { + /* Let BPF program decide how to proceed. */ + ctx.cur_len = 0; + } + set_fs(old_fs); + } else { + /* Let BPF program decide how to proceed. */ + ctx.cur_len = 0; + } + rcu_read_lock(); cgrp = task_dfl_cgroup(current); ret = BPF_PROG_RUN_ARRAY(cgrp->bpf.effective[type], &ctx, BPF_PROG_RUN); rcu_read_unlock(); + kfree(ctx.cur_val); + return ret == 1 ? 0 : -EPERM; } EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl); @@ -869,12 +891,55 @@ static const struct bpf_func_proto bpf_sysctl_get_name_proto = { .arg4_type = ARG_ANYTHING, }; +static int copy_sysctl_value(char *dst, size_t dst_len, char *src, + size_t src_len) +{ + if (!dst) + return -EINVAL; + + if (!dst_len) + return -E2BIG; + + if (!src || !src_len) { + memset(dst, 0, dst_len); + return -EINVAL; + } + + memcpy(dst, src, min(dst_len, src_len)); + + if (dst_len > src_len) { + memset(dst + src_len, '\0', dst_len - src_len); + return src_len; + } + + dst[dst_len - 1] = '\0'; + + return -E2BIG; +} + +BPF_CALL_3(bpf_sysctl_get_current_value, struct bpf_sysctl_kern *, ctx, + char *, buf, size_t, buf_len) +{ + return copy_sysctl_value(buf, buf_len, ctx->cur_val, ctx->cur_len); +} + +static const struct bpf_func_proto bpf_sysctl_get_current_value_proto = { + .func = bpf_sysctl_get_current_value, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_PTR_TO_UNINIT_MEM, + .arg3_type = ARG_CONST_SIZE, +}; + static const struct bpf_func_proto * sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { switch (func_id) { case BPF_FUNC_sysctl_get_name: return &bpf_sysctl_get_name_proto; + case BPF_FUNC_sysctl_get_current_value: + return &bpf_sysctl_get_current_value_proto; default: return cgroup_base_func_proto(func_id, prog); } From 4e63acdff864654cee0ac5aaeda3913798ee78f6 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Thu, 7 Mar 2019 18:38:43 -0800 Subject: [PATCH 05/68] bpf: Introduce bpf_sysctl_{get,set}_new_value helpers Add helpers to work with new value being written to sysctl by user space. bpf_sysctl_get_new_value() copies value being written to sysctl into provided buffer. bpf_sysctl_set_new_value() overrides new value being written by user space with a one from provided buffer. Buffer should contain string representation of the value, similar to what can be seen in /proc/sys/. Both helpers can be used only on sysctl write. File position matters and can be managed by an interface that will be introduced separately. E.g. if user space calls sys_write to a file in /proc/sys/ at file position = X, where X > 0, then the value set by bpf_sysctl_set_new_value() will be written starting from X. If program wants to override whole value with specified buffer, file position has to be set to zero. Documentation for the new helpers is provided in bpf.h UAPI. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- fs/proc/proc_sysctl.c | 22 ++++++++--- include/linux/bpf-cgroup.h | 8 ++-- include/linux/filter.h | 3 ++ include/uapi/linux/bpf.h | 38 +++++++++++++++++- kernel/bpf/cgroup.c | 81 +++++++++++++++++++++++++++++++++++++- 5 files changed, 142 insertions(+), 10 deletions(-) diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index e01b02150340..023101c6f0d7 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -570,8 +570,8 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, struct inode *inode = file_inode(filp); struct ctl_table_header *head = grab_header(inode); struct ctl_table *table = PROC_I(inode)->sysctl_entry; + void *new_buf = NULL; ssize_t error; - size_t res; if (IS_ERR(head)) return PTR_ERR(head); @@ -589,15 +589,27 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, if (!table->proc_handler) goto out; - error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write); + error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, &count, + &new_buf); if (error) goto out; /* careful: calling conventions are nasty here */ - res = count; - error = table->proc_handler(table, write, buf, &res, ppos); + if (new_buf) { + mm_segment_t old_fs; + + old_fs = get_fs(); + set_fs(KERNEL_DS); + error = table->proc_handler(table, write, (void __user *)new_buf, + &count, ppos); + set_fs(old_fs); + kfree(new_buf); + } else { + error = table->proc_handler(table, write, buf, &count, ppos); + } + if (!error) - error = res; + error = count; out: sysctl_head_finish(head); diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index b1c45da20a26..1e97271f9a10 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -113,7 +113,8 @@ int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor, int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, struct ctl_table *table, int write, - enum bpf_attach_type type); + void __user *buf, size_t *pcount, + void **new_buf, enum bpf_attach_type type); static inline enum bpf_cgroup_storage_type cgroup_storage_type( struct bpf_map *map) @@ -261,11 +262,12 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key, }) -#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write) \ +#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, count, nbuf) \ ({ \ int __ret = 0; \ if (cgroup_bpf_enabled) \ __ret = __cgroup_bpf_run_filter_sysctl(head, table, write, \ + buf, count, nbuf, \ BPF_CGROUP_SYSCTL); \ __ret; \ }) @@ -338,7 +340,7 @@ static inline int bpf_percpu_cgroup_storage_update(struct bpf_map *map, #define BPF_CGROUP_RUN_PROG_UDP6_SENDMSG_LOCK(sk, uaddr, t_ctx) ({ 0; }) #define BPF_CGROUP_RUN_PROG_SOCK_OPS(sock_ops) ({ 0; }) #define BPF_CGROUP_RUN_PROG_DEVICE_CGROUP(type,major,minor,access) ({ 0; }) -#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write) ({ 0; }) +#define BPF_CGROUP_RUN_PROG_SYSCTL(head,table,write,buf,count,nbuf) ({ 0; }) #define for_each_cgroup_storage_type(stype) for (; false; ) diff --git a/include/linux/filter.h b/include/linux/filter.h index f254ff92819f..a23653f9460c 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -1184,6 +1184,9 @@ struct bpf_sysctl_kern { struct ctl_table *table; void *cur_val; size_t cur_len; + void *new_val; + size_t new_len; + int new_updated; int write; }; diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 063543afa359..547b8258d731 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2541,6 +2541,40 @@ union bpf_attr { * * **-EINVAL** if current value was unavailable, e.g. because * sysctl is uninitialized and read returns -EIO for it. + * + * int bpf_sysctl_get_new_value(struct bpf_sysctl *ctx, char *buf, size_t buf_len) + * Description + * Get new value being written by user space to sysctl (before + * the actual write happens) and copy it as a string into + * provided by program buffer *buf* of size *buf_len*. + * + * User space may write new value at file position > 0. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). + * + * **-EINVAL** if sysctl is being read. + * + * int bpf_sysctl_set_new_value(struct bpf_sysctl *ctx, const char *buf, size_t buf_len) + * Description + * Override new value being written by user space to sysctl with + * value provided by program in buffer *buf* of size *buf_len*. + * + * *buf* should contain a string in same form as provided by user + * space on sysctl write. + * + * User space may write new value at file position > 0. To override + * the whole sysctl value file position should be set to zero. + * Return + * 0 on success. + * + * **-E2BIG** if the *buf_len* is too big. + * + * **-EINVAL** if sysctl is being read. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2645,7 +2679,9 @@ union bpf_attr { FN(skc_lookup_tcp), \ FN(tcp_check_syncookie), \ FN(sysctl_get_name), \ - FN(sysctl_get_current_value), + FN(sysctl_get_current_value), \ + FN(sysctl_get_new_value), \ + FN(sysctl_set_new_value), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index c6b2cf29a54b..ba4e21986760 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -778,6 +778,13 @@ const struct bpf_verifier_ops cg_dev_verifier_ops = { * @head: sysctl table header * @table: sysctl table * @write: sysctl is being read (= 0) or written (= 1) + * @buf: pointer to buffer passed by user space + * @pcount: value-result argument: value is size of buffer pointed to by @buf, + * result is size of @new_buf if program set new value, initial value + * otherwise + * @new_buf: pointer to pointer to new buffer that will be allocated if program + * overrides new value provided by user space on sysctl write + * NOTE: it's caller responsibility to free *new_buf if it was set * @type: type of program to be executed * * Program is run when sysctl is being accessed, either read or written, and @@ -788,7 +795,8 @@ const struct bpf_verifier_ops cg_dev_verifier_ops = { */ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, struct ctl_table *table, int write, - enum bpf_attach_type type) + void __user *buf, size_t *pcount, + void **new_buf, enum bpf_attach_type type) { struct bpf_sysctl_kern ctx = { .head = head, @@ -796,6 +804,9 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, .write = write, .cur_val = NULL, .cur_len = PAGE_SIZE, + .new_val = NULL, + .new_len = 0, + .new_updated = 0, }; struct cgroup *cgrp; int ret; @@ -818,6 +829,18 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, ctx.cur_len = 0; } + if (write && buf && *pcount) { + /* BPF program should be able to override new value with a + * buffer bigger than provided by user. + */ + ctx.new_val = kmalloc_track_caller(PAGE_SIZE, GFP_KERNEL); + ctx.new_len = min(PAGE_SIZE, *pcount); + if (!ctx.new_val || + copy_from_user(ctx.new_val, buf, ctx.new_len)) + /* Let BPF program decide how to proceed. */ + ctx.new_len = 0; + } + rcu_read_lock(); cgrp = task_dfl_cgroup(current); ret = BPF_PROG_RUN_ARRAY(cgrp->bpf.effective[type], &ctx, BPF_PROG_RUN); @@ -825,6 +848,13 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, kfree(ctx.cur_val); + if (ret == 1 && ctx.new_updated) { + *new_buf = ctx.new_val; + *pcount = ctx.new_len; + } else { + kfree(ctx.new_val); + } + return ret == 1 ? 0 : -EPERM; } EXPORT_SYMBOL(__cgroup_bpf_run_filter_sysctl); @@ -932,6 +962,51 @@ static const struct bpf_func_proto bpf_sysctl_get_current_value_proto = { .arg3_type = ARG_CONST_SIZE, }; +BPF_CALL_3(bpf_sysctl_get_new_value, struct bpf_sysctl_kern *, ctx, char *, buf, + size_t, buf_len) +{ + if (!ctx->write) { + if (buf && buf_len) + memset(buf, '\0', buf_len); + return -EINVAL; + } + return copy_sysctl_value(buf, buf_len, ctx->new_val, ctx->new_len); +} + +static const struct bpf_func_proto bpf_sysctl_get_new_value_proto = { + .func = bpf_sysctl_get_new_value, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_PTR_TO_UNINIT_MEM, + .arg3_type = ARG_CONST_SIZE, +}; + +BPF_CALL_3(bpf_sysctl_set_new_value, struct bpf_sysctl_kern *, ctx, + const char *, buf, size_t, buf_len) +{ + if (!ctx->write || !ctx->new_val || !ctx->new_len || !buf || !buf_len) + return -EINVAL; + + if (buf_len > PAGE_SIZE - 1) + return -E2BIG; + + memcpy(ctx->new_val, buf, buf_len); + ctx->new_len = buf_len; + ctx->new_updated = 1; + + return 0; +} + +static const struct bpf_func_proto bpf_sysctl_set_new_value_proto = { + .func = bpf_sysctl_set_new_value, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_PTR_TO_MEM, + .arg3_type = ARG_CONST_SIZE, +}; + static const struct bpf_func_proto * sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { @@ -940,6 +1015,10 @@ sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_sysctl_get_name_proto; case BPF_FUNC_sysctl_get_current_value: return &bpf_sysctl_get_current_value_proto; + case BPF_FUNC_sysctl_get_new_value: + return &bpf_sysctl_get_new_value_proto; + case BPF_FUNC_sysctl_set_new_value: + return &bpf_sysctl_set_new_value_proto; default: return cgroup_base_func_proto(func_id, prog); } From e1550bfe0de47e30484ba91de1e50a91ec1c31f5 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Thu, 7 Mar 2019 18:50:52 -0800 Subject: [PATCH 06/68] bpf: Add file_pos field to bpf_sysctl ctx Add file_pos field to bpf_sysctl context to read and write sysctl file position at which sysctl is being accessed (read or written). The field can be used to e.g. override whole sysctl value on write to sysctl even when sys_write is called by user space with file_pos > 0. Or BPF program may reject such accesses. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- fs/proc/proc_sysctl.c | 2 +- include/linux/bpf-cgroup.h | 9 ++++--- include/linux/filter.h | 3 +++ include/uapi/linux/bpf.h | 3 +++ kernel/bpf/cgroup.c | 54 +++++++++++++++++++++++++++++++++++--- 5 files changed, 63 insertions(+), 8 deletions(-) diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 023101c6f0d7..2d61e5e8c863 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -590,7 +590,7 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, goto out; error = BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, &count, - &new_buf); + ppos, &new_buf); if (error) goto out; diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index 1e97271f9a10..cb3c6b3b89c8 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -114,7 +114,8 @@ int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor, int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, struct ctl_table *table, int write, void __user *buf, size_t *pcount, - void **new_buf, enum bpf_attach_type type); + loff_t *ppos, void **new_buf, + enum bpf_attach_type type); static inline enum bpf_cgroup_storage_type cgroup_storage_type( struct bpf_map *map) @@ -262,12 +263,12 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key, }) -#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, count, nbuf) \ +#define BPF_CGROUP_RUN_PROG_SYSCTL(head, table, write, buf, count, pos, nbuf) \ ({ \ int __ret = 0; \ if (cgroup_bpf_enabled) \ __ret = __cgroup_bpf_run_filter_sysctl(head, table, write, \ - buf, count, nbuf, \ + buf, count, pos, nbuf, \ BPF_CGROUP_SYSCTL); \ __ret; \ }) @@ -340,7 +341,7 @@ static inline int bpf_percpu_cgroup_storage_update(struct bpf_map *map, #define BPF_CGROUP_RUN_PROG_UDP6_SENDMSG_LOCK(sk, uaddr, t_ctx) ({ 0; }) #define BPF_CGROUP_RUN_PROG_SOCK_OPS(sock_ops) ({ 0; }) #define BPF_CGROUP_RUN_PROG_DEVICE_CGROUP(type,major,minor,access) ({ 0; }) -#define BPF_CGROUP_RUN_PROG_SYSCTL(head,table,write,buf,count,nbuf) ({ 0; }) +#define BPF_CGROUP_RUN_PROG_SYSCTL(head,table,write,buf,count,pos,nbuf) ({ 0; }) #define for_each_cgroup_storage_type(stype) for (; false; ) diff --git a/include/linux/filter.h b/include/linux/filter.h index a23653f9460c..fb0edad75971 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -1188,6 +1188,9 @@ struct bpf_sysctl_kern { size_t new_len; int new_updated; int write; + loff_t *ppos; + /* Temporary "register" for indirect stores to ppos. */ + u64 tmp_reg; }; #endif /* __LINUX_FILTER_H__ */ diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 547b8258d731..89976de909af 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -3391,6 +3391,9 @@ struct bpf_sysctl { __u32 write; /* Sysctl is being read (= 0) or written (= 1). * Allows 1,2,4-byte read, but no write. */ + __u32 file_pos; /* Sysctl file position to read from, write to. + * Allows 1,2,4-byte read an 4-byte write. + */ }; #endif /* _UAPI__LINUX_BPF_H__ */ diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index ba4e21986760..b2adf22139b3 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -782,6 +782,9 @@ const struct bpf_verifier_ops cg_dev_verifier_ops = { * @pcount: value-result argument: value is size of buffer pointed to by @buf, * result is size of @new_buf if program set new value, initial value * otherwise + * @ppos: value-result argument: value is position at which read from or write + * to sysctl is happening, result is new position if program overrode it, + * initial value otherwise * @new_buf: pointer to pointer to new buffer that will be allocated if program * overrides new value provided by user space on sysctl write * NOTE: it's caller responsibility to free *new_buf if it was set @@ -796,12 +799,14 @@ const struct bpf_verifier_ops cg_dev_verifier_ops = { int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, struct ctl_table *table, int write, void __user *buf, size_t *pcount, - void **new_buf, enum bpf_attach_type type) + loff_t *ppos, void **new_buf, + enum bpf_attach_type type) { struct bpf_sysctl_kern ctx = { .head = head, .table = table, .write = write, + .ppos = ppos, .cur_val = NULL, .cur_len = PAGE_SIZE, .new_val = NULL, @@ -1030,14 +1035,22 @@ static bool sysctl_is_valid_access(int off, int size, enum bpf_access_type type, { const int size_default = sizeof(__u32); - if (off < 0 || off + size > sizeof(struct bpf_sysctl) || - off % size || type != BPF_READ) + if (off < 0 || off + size > sizeof(struct bpf_sysctl) || off % size) return false; switch (off) { case offsetof(struct bpf_sysctl, write): + if (type != BPF_READ) + return false; bpf_ctx_record_field_size(info, size_default); return bpf_ctx_narrow_access_ok(off, size, size_default); + case offsetof(struct bpf_sysctl, file_pos): + if (type == BPF_READ) { + bpf_ctx_record_field_size(info, size_default); + return bpf_ctx_narrow_access_ok(off, size, size_default); + } else { + return size == size_default; + } default: return false; } @@ -1059,6 +1072,41 @@ static u32 sysctl_convert_ctx_access(enum bpf_access_type type, write), target_size)); break; + case offsetof(struct bpf_sysctl, file_pos): + /* ppos is a pointer so it should be accessed via indirect + * loads and stores. Also for stores additional temporary + * register is used since neither src_reg nor dst_reg can be + * overridden. + */ + if (type == BPF_WRITE) { + int treg = BPF_REG_9; + + if (si->src_reg == treg || si->dst_reg == treg) + --treg; + if (si->src_reg == treg || si->dst_reg == treg) + --treg; + *insn++ = BPF_STX_MEM( + BPF_DW, si->dst_reg, treg, + offsetof(struct bpf_sysctl_kern, tmp_reg)); + *insn++ = BPF_LDX_MEM( + BPF_FIELD_SIZEOF(struct bpf_sysctl_kern, ppos), + treg, si->dst_reg, + offsetof(struct bpf_sysctl_kern, ppos)); + *insn++ = BPF_STX_MEM( + BPF_SIZEOF(u32), treg, si->src_reg, 0); + *insn++ = BPF_LDX_MEM( + BPF_DW, treg, si->dst_reg, + offsetof(struct bpf_sysctl_kern, tmp_reg)); + } else { + *insn++ = BPF_LDX_MEM( + BPF_FIELD_SIZEOF(struct bpf_sysctl_kern, ppos), + si->dst_reg, si->src_reg, + offsetof(struct bpf_sysctl_kern, ppos)); + *insn++ = BPF_LDX_MEM( + BPF_SIZE(si->code), si->dst_reg, si->dst_reg, 0); + } + *target_size = sizeof(u32); + break; } return insn - insn_buf; From 196398d4c0ac3de6f016f7774c952a069e116e71 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Thu, 7 Mar 2019 19:09:29 -0800 Subject: [PATCH 07/68] bpf: Sync bpf.h to tools/ Sync BPF_PROG_TYPE_CGROUP_SYSCTL related bpf UAPI changes to tools/. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/include/uapi/linux/bpf.h | 90 +++++++++++++++++++++++++++++++++- 1 file changed, 89 insertions(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 2e96d0b4bf65..89976de909af 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -167,6 +167,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_LIRC_MODE2, BPF_PROG_TYPE_SK_REUSEPORT, BPF_PROG_TYPE_FLOW_DISSECTOR, + BPF_PROG_TYPE_CGROUP_SYSCTL, }; enum bpf_attach_type { @@ -188,6 +189,7 @@ enum bpf_attach_type { BPF_CGROUP_UDP6_SENDMSG, BPF_LIRC_MODE2, BPF_FLOW_DISSECTOR, + BPF_CGROUP_SYSCTL, __MAX_BPF_ATTACH_TYPE }; @@ -2504,6 +2506,75 @@ union bpf_attr { * Return * 0 if iph and th are a valid SYN cookie ACK, or a negative error * otherwise. + * + * int bpf_sysctl_get_name(struct bpf_sysctl *ctx, char *buf, size_t buf_len, u64 flags) + * Description + * Get name of sysctl in /proc/sys/ and copy it into provided by + * program buffer *buf* of size *buf_len*. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * + * If *flags* is zero, full name (e.g. "net/ipv4/tcp_mem") is + * copied. Use **BPF_F_SYSCTL_BASE_NAME** flag to copy base name + * only (e.g. "tcp_mem"). + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). + * + * int bpf_sysctl_get_current_value(struct bpf_sysctl *ctx, char *buf, size_t buf_len) + * Description + * Get current value of sysctl as it is presented in /proc/sys + * (incl. newline, etc), and copy it as a string into provided + * by program buffer *buf* of size *buf_len*. + * + * The whole value is copied, no matter what file position user + * space issued e.g. sys_read at. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). + * + * **-EINVAL** if current value was unavailable, e.g. because + * sysctl is uninitialized and read returns -EIO for it. + * + * int bpf_sysctl_get_new_value(struct bpf_sysctl *ctx, char *buf, size_t buf_len) + * Description + * Get new value being written by user space to sysctl (before + * the actual write happens) and copy it as a string into + * provided by program buffer *buf* of size *buf_len*. + * + * User space may write new value at file position > 0. + * + * The buffer is always NUL terminated, unless it's zero-sized. + * Return + * Number of character copied (not including the trailing NUL). + * + * **-E2BIG** if the buffer wasn't big enough (*buf* will contain + * truncated name in this case). + * + * **-EINVAL** if sysctl is being read. + * + * int bpf_sysctl_set_new_value(struct bpf_sysctl *ctx, const char *buf, size_t buf_len) + * Description + * Override new value being written by user space to sysctl with + * value provided by program in buffer *buf* of size *buf_len*. + * + * *buf* should contain a string in same form as provided by user + * space on sysctl write. + * + * User space may write new value at file position > 0. To override + * the whole sysctl value file position should be set to zero. + * Return + * 0 on success. + * + * **-E2BIG** if the *buf_len* is too big. + * + * **-EINVAL** if sysctl is being read. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2606,7 +2677,11 @@ union bpf_attr { FN(skb_ecn_set_ce), \ FN(get_listener_sock), \ FN(skc_lookup_tcp), \ - FN(tcp_check_syncookie), + FN(tcp_check_syncookie), \ + FN(sysctl_get_name), \ + FN(sysctl_get_current_value), \ + FN(sysctl_get_new_value), \ + FN(sysctl_set_new_value), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call @@ -2679,6 +2754,9 @@ enum bpf_func_id { BPF_ADJ_ROOM_ENCAP_L2_MASK) \ << BPF_ADJ_ROOM_ENCAP_L2_SHIFT) +/* BPF_FUNC_sysctl_get_name flags. */ +#define BPF_F_SYSCTL_BASE_NAME (1ULL << 0) + /* Mode for BPF_FUNC_skb_adjust_room helper. */ enum bpf_adj_room_mode { BPF_ADJ_ROOM_NET, @@ -3308,4 +3386,14 @@ struct bpf_line_info { struct bpf_spin_lock { __u32 val; }; + +struct bpf_sysctl { + __u32 write; /* Sysctl is being read (= 0) or written (= 1). + * Allows 1,2,4-byte read, but no write. + */ + __u32 file_pos; /* Sysctl file position to read from, write to. + * Allows 1,2,4-byte read an 4-byte write. + */ +}; + #endif /* _UAPI__LINUX_BPF_H__ */ From 063cc9f06ee6cac12dbc68a504bc4ff56bde790d Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 09:15:26 -0800 Subject: [PATCH 08/68] libbpf: Support sysctl hook Support BPF_PROG_TYPE_CGROUP_SYSCTL program in libbpf: identifying program and attach types by section name, probe. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/libbpf.c | 3 +++ tools/lib/bpf/libbpf_probes.c | 1 + 2 files changed, 4 insertions(+) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 67484cf32b2e..e5b77ad97795 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -2064,6 +2064,7 @@ static bool bpf_prog_type__needs_kver(enum bpf_prog_type type) case BPF_PROG_TYPE_TRACEPOINT: case BPF_PROG_TYPE_RAW_TRACEPOINT: case BPF_PROG_TYPE_PERF_EVENT: + case BPF_PROG_TYPE_CGROUP_SYSCTL: return false; case BPF_PROG_TYPE_KPROBE: default: @@ -3004,6 +3005,8 @@ static const struct { BPF_CGROUP_UDP4_SENDMSG), BPF_EAPROG_SEC("cgroup/sendmsg6", BPF_PROG_TYPE_CGROUP_SOCK_ADDR, BPF_CGROUP_UDP6_SENDMSG), + BPF_EAPROG_SEC("cgroup/sysctl", BPF_PROG_TYPE_CGROUP_SYSCTL, + BPF_CGROUP_SYSCTL), }; #undef BPF_PROG_SEC_IMPL diff --git a/tools/lib/bpf/libbpf_probes.c b/tools/lib/bpf/libbpf_probes.c index 8c3a1c04dcb2..0f25541632e3 100644 --- a/tools/lib/bpf/libbpf_probes.c +++ b/tools/lib/bpf/libbpf_probes.c @@ -97,6 +97,7 @@ probe_load(enum bpf_prog_type prog_type, const struct bpf_insn *insns, case BPF_PROG_TYPE_LIRC_MODE2: case BPF_PROG_TYPE_SK_REUSEPORT: case BPF_PROG_TYPE_FLOW_DISSECTOR: + case BPF_PROG_TYPE_CGROUP_SYSCTL: default: break; } From 7007af63da3bf6764c9208029a3585756260b55f Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 09:17:45 -0800 Subject: [PATCH 09/68] selftests/bpf: Test sysctl section name Add unit test to verify that program and attach types are properly identified for "cgroup/sysctl" section name. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_section_names.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/tools/testing/selftests/bpf/test_section_names.c b/tools/testing/selftests/bpf/test_section_names.c index 7c4f41572b1c..bebd4fbca1f4 100644 --- a/tools/testing/selftests/bpf/test_section_names.c +++ b/tools/testing/selftests/bpf/test_section_names.c @@ -119,6 +119,11 @@ static struct sec_name_test tests[] = { {0, BPF_PROG_TYPE_CGROUP_SOCK_ADDR, BPF_CGROUP_UDP6_SENDMSG}, {0, BPF_CGROUP_UDP6_SENDMSG}, }, + { + "cgroup/sysctl", + {0, BPF_PROG_TYPE_CGROUP_SYSCTL, BPF_CGROUP_SYSCTL}, + {0, BPF_CGROUP_SYSCTL}, + }, }; static int test_prog_type_by_name(const struct sec_name_test *test) From 1f5fa9ab6e2eb16ba81284280a498fe9c543a2d8 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 15:08:21 -0800 Subject: [PATCH 10/68] selftests/bpf: Test BPF_CGROUP_SYSCTL Add unit test for BPF_PROG_TYPE_CGROUP_SYSCTL program type. Test that program can allow/deny access. Test both valid and invalid accesses to ctx->write. Example of output: # ./test_sysctl Test case: sysctl wrong attach_type .. [PASS] Test case: sysctl:read allow all .. [PASS] Test case: sysctl:read deny all .. [PASS] Test case: ctx:write sysctl:read read ok .. [PASS] Test case: ctx:write sysctl:write read ok .. [PASS] Test case: ctx:write sysctl:read write reject .. [PASS] Summary: 6 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/Makefile | 3 +- tools/testing/selftests/bpf/test_sysctl.c | 291 ++++++++++++++++++++++ 2 files changed, 293 insertions(+), 1 deletion(-) create mode 100644 tools/testing/selftests/bpf/test_sysctl.c diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 078283d073b0..f9d83ba7843e 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -23,7 +23,7 @@ TEST_GEN_PROGS = test_verifier test_tag test_maps test_lru_map test_lpm_map test test_align test_verifier_log test_dev_cgroup test_tcpbpf_user \ test_sock test_btf test_sockmap test_lirc_mode2_user get_cgroup_id_user \ test_socket_cookie test_cgroup_storage test_select_reuseport test_section_names \ - test_netcnt test_tcpnotify_user test_sock_fields + test_netcnt test_tcpnotify_user test_sock_fields test_sysctl BPF_OBJ_FILES = $(patsubst %.c,%.o, $(notdir $(wildcard progs/*.c))) TEST_GEN_FILES = $(BPF_OBJ_FILES) @@ -93,6 +93,7 @@ $(OUTPUT)/get_cgroup_id_user: cgroup_helpers.c $(OUTPUT)/test_cgroup_storage: cgroup_helpers.c $(OUTPUT)/test_netcnt: cgroup_helpers.c $(OUTPUT)/test_sock_fields: cgroup_helpers.c +$(OUTPUT)/test_sysctl: cgroup_helpers.c .PHONY: force diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c new file mode 100644 index 000000000000..6d0ab09789ad --- /dev/null +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -0,0 +1,291 @@ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2019 Facebook + +#include +#include +#include +#include +#include +#include + +#include + +#include + +#include "bpf_rlimit.h" +#include "bpf_util.h" +#include "cgroup_helpers.h" + +#define CG_PATH "/foo" +#define MAX_INSNS 512 + +char bpf_log_buf[BPF_LOG_BUF_SIZE]; + +struct sysctl_test { + const char *descr; + struct bpf_insn insns[MAX_INSNS]; + enum bpf_attach_type attach_type; + const char *sysctl; + int open_flags; + const char *newval; + enum { + LOAD_REJECT, + ATTACH_REJECT, + OP_EPERM, + SUCCESS, + } result; +}; + +static struct sysctl_test tests[] = { + { + .descr = "sysctl wrong attach_type", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = 0, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = ATTACH_REJECT, + }, + { + .descr = "sysctl:read allow all", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl:read deny all", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = OP_EPERM, + }, + { + .descr = "ctx:write sysctl:read read ok", + .insns = { + /* If (write) */ + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_1, + offsetof(struct bpf_sysctl, write)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 1, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "ctx:write sysctl:write read ok", + .insns = { + /* If (write) */ + BPF_LDX_MEM(BPF_B, BPF_REG_7, BPF_REG_1, + offsetof(struct bpf_sysctl, write)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 1, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/domainname", + .open_flags = O_WRONLY, + .newval = "(none)", /* same as default, should fail anyway */ + .result = OP_EPERM, + }, + { + .descr = "ctx:write sysctl:read write reject", + .insns = { + /* write = X */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_W, BPF_REG_1, BPF_REG_0, + offsetof(struct bpf_sysctl, write)), + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = LOAD_REJECT, + }, +}; + +static size_t probe_prog_length(const struct bpf_insn *fp) +{ + size_t len; + + for (len = MAX_INSNS - 1; len > 0; --len) + if (fp[len].code != 0 || fp[len].imm != 0) + break; + return len + 1; +} + +static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) +{ + struct bpf_insn *prog = test->insns; + struct bpf_load_program_attr attr; + int ret; + + memset(&attr, 0, sizeof(struct bpf_load_program_attr)); + attr.prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL; + attr.insns = prog; + attr.insns_cnt = probe_prog_length(attr.insns); + attr.license = "GPL"; + + ret = bpf_load_program_xattr(&attr, bpf_log_buf, BPF_LOG_BUF_SIZE); + if (ret < 0 && test->result != LOAD_REJECT) { + log_err(">>> Loading program error.\n" + ">>> Verifier output:\n%s\n-------\n", bpf_log_buf); + } + + return ret; +} + +static int access_sysctl(const char *sysctl_path, + const struct sysctl_test *test) +{ + int err = 0; + int fd; + + fd = open(sysctl_path, test->open_flags | O_CLOEXEC); + if (fd < 0) + return fd; + + if (test->open_flags == O_RDONLY) { + char buf[128]; + + if (read(fd, buf, sizeof(buf)) == -1) + goto err; + } else if (test->open_flags == O_WRONLY) { + if (!test->newval) { + log_err("New value for sysctl is not set"); + goto err; + } + if (write(fd, test->newval, strlen(test->newval)) == -1) + goto err; + } else { + log_err("Unexpected sysctl access: neither read nor write"); + goto err; + } + + goto out; +err: + err = -1; +out: + close(fd); + return err; +} + +static int run_test_case(int cgfd, struct sysctl_test *test) +{ + enum bpf_attach_type atype = test->attach_type; + char sysctl_path[128]; + int progfd = -1; + int err = 0; + + printf("Test case: %s .. ", test->descr); + + snprintf(sysctl_path, sizeof(sysctl_path), "/proc/sys/%s", + test->sysctl); + + progfd = load_sysctl_prog(test, sysctl_path); + if (progfd < 0) { + if (test->result == LOAD_REJECT) + goto out; + else + goto err; + } + + if (bpf_prog_attach(progfd, cgfd, atype, BPF_F_ALLOW_OVERRIDE) == -1) { + if (test->result == ATTACH_REJECT) + goto out; + else + goto err; + } + + if (access_sysctl(sysctl_path, test) == -1) { + if (test->result == OP_EPERM && errno == EPERM) + goto out; + else + goto err; + } + + if (test->result != SUCCESS) { + log_err("Unexpected failure"); + goto err; + } + + goto out; +err: + err = -1; +out: + /* Detaching w/o checking return code: best effort attempt. */ + if (progfd != -1) + bpf_prog_detach(cgfd, atype); + close(progfd); + printf("[%s]\n", err ? "FAIL" : "PASS"); + return err; +} + +static int run_tests(int cgfd) +{ + int passes = 0; + int fails = 0; + int i; + + for (i = 0; i < ARRAY_SIZE(tests); ++i) { + if (run_test_case(cgfd, &tests[i])) + ++fails; + else + ++passes; + } + printf("Summary: %d PASSED, %d FAILED\n", passes, fails); + return fails ? -1 : 0; +} + +int main(int argc, char **argv) +{ + int cgfd = -1; + int err = 0; + + if (setup_cgroup_environment()) + goto err; + + cgfd = create_and_get_cgroup(CG_PATH); + if (cgfd < 0) + goto err; + + if (join_cgroup(CG_PATH)) + goto err; + + if (run_tests(cgfd)) + goto err; + + goto out; +err: + err = -1; +out: + close(cgfd); + cleanup_cgroup_environment(); + return err; +} From 6041c67f28d8b297ef53ab47abdcd319626c765e Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 15:13:43 -0800 Subject: [PATCH 11/68] selftests/bpf: Test bpf_sysctl_get_name helper Test w/ and w/o BPF_F_SYSCTL_BASE_NAME, buffers with enough space and too small buffers to get E2BIG and truncated result, etc. # ./test_sysctl ... Test case: sysctl_get_name sysctl_value:base ok .. [PASS] Test case: sysctl_get_name sysctl_value:base E2BIG truncated .. [PASS] Test case: sysctl_get_name sysctl:full ok .. [PASS] Test case: sysctl_get_name sysctl:full E2BIG truncated .. [PASS] Test case: sysctl_get_name sysctl:full E2BIG truncated small .. [PASS] Summary: 11 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_sysctl.c | 222 ++++++++++++++++++++++ 1 file changed, 222 insertions(+) diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 6d0ab09789ad..2e7ef04503a8 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -128,6 +128,228 @@ static struct sysctl_test tests[] = { .open_flags = O_RDONLY, .result = LOAD_REJECT, }, + { + .descr = "sysctl_get_name sysctl_value:base ok", + .insns = { + /* sysctl_get_name arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_name arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* sysctl_get_name arg4 (flags) */ + BPF_MOV64_IMM(BPF_REG_4, BPF_F_SYSCTL_BASE_NAME), + + /* sysctl_get_name(ctx, buf, buf_len, flags) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_name), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, sizeof("tcp_mem") - 1, 6), + /* buf == "tcp_mem\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x006d656d5f706374ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_name sysctl_value:base E2BIG truncated", + .insns = { + /* sysctl_get_name arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_name arg3 (buf_len) too small */ + BPF_MOV64_IMM(BPF_REG_3, 7), + + /* sysctl_get_name arg4 (flags) */ + BPF_MOV64_IMM(BPF_REG_4, BPF_F_SYSCTL_BASE_NAME), + + /* sysctl_get_name(ctx, buf, buf_len, flags) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_name), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -E2BIG, 6), + + /* buf[0:7] == "tcp_me\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x00656d5f706374ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_name sysctl:full ok", + .insns = { + /* sysctl_get_name arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -24), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 16), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_name arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 17), + + /* sysctl_get_name arg4 (flags) */ + BPF_MOV64_IMM(BPF_REG_4, 0), + + /* sysctl_get_name(ctx, buf, buf_len, flags) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_name), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 16, 14), + + /* buf[0:8] == "net/ipv4" && */ + BPF_LD_IMM64(BPF_REG_8, 0x347670692f74656eULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 10), + + /* buf[8:16] == "/tcp_mem" && */ + BPF_LD_IMM64(BPF_REG_8, 0x6d656d5f7063742fULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 8), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 6), + + /* buf[16:24] == "\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x0ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 16), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_name sysctl:full E2BIG truncated", + .insns = { + /* sysctl_get_name arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -16), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 8), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_name arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 16), + + /* sysctl_get_name arg4 (flags) */ + BPF_MOV64_IMM(BPF_REG_4, 0), + + /* sysctl_get_name(ctx, buf, buf_len, flags) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_name), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -E2BIG, 10), + + /* buf[0:8] == "net/ipv4" && */ + BPF_LD_IMM64(BPF_REG_8, 0x347670692f74656eULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 6), + + /* buf[8:16] == "/tcp_me\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x00656d5f7063742fULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 8), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_name sysctl:full E2BIG truncated small", + .insns = { + /* sysctl_get_name arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_name arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 7), + + /* sysctl_get_name arg4 (flags) */ + BPF_MOV64_IMM(BPF_REG_4, 0), + + /* sysctl_get_name(ctx, buf, buf_len, flags) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_name), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -E2BIG, 6), + + /* buf[0:8] == "net/ip\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x000070692f74656eULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, }; static size_t probe_prog_length(const struct bpf_insn *fp) From 11ff34f74e32f5a4ec1f71621508789cd60775b3 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 15:17:51 -0800 Subject: [PATCH 12/68] selftests/bpf: Test sysctl_get_current_value helper Test sysctl_get_current_value on sysctl read and write, buffers with enough space and too small buffers to get E2BIG and truncated result, etc. # ./test_sysctl ... Test case: sysctl_get_current_value sysctl:read ok, gt .. [PASS] Test case: sysctl_get_current_value sysctl:read ok, eq .. [PASS] Test case: sysctl_get_current_value sysctl:read E2BIG truncated .. [PASS] Test case: sysctl_get_current_value sysctl:read EINVAL .. [PASS] Test case: sysctl_get_current_value sysctl:write ok .. [PASS] Summary: 16 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_sysctl.c | 228 ++++++++++++++++++++++ 1 file changed, 228 insertions(+) diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 2e7ef04503a8..9fb6560fa775 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -18,11 +18,13 @@ #define CG_PATH "/foo" #define MAX_INSNS 512 +#define FIXUP_SYSCTL_VALUE 0 char bpf_log_buf[BPF_LOG_BUF_SIZE]; struct sysctl_test { const char *descr; + size_t fixup_value_insn; struct bpf_insn insns[MAX_INSNS]; enum bpf_attach_type attach_type; const char *sysctl; @@ -350,6 +352,190 @@ static struct sysctl_test tests[] = { .open_flags = O_RDONLY, .result = SUCCESS, }, + { + .descr = "sysctl_get_current_value sysctl:read ok, gt", + .insns = { + /* sysctl_get_current_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_current_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* sysctl_get_current_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_current_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 6, 6), + + /* buf[0:6] == "Linux\n\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x000a78756e694cULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_current_value sysctl:read ok, eq", + .insns = { + /* sysctl_get_current_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_B, BPF_REG_7, BPF_REG_0, 7), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_current_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 7), + + /* sysctl_get_current_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_current_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 6, 6), + + /* buf[0:6] == "Linux\n\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x000a78756e694cULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_current_value sysctl:read E2BIG truncated", + .insns = { + /* sysctl_get_current_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_H, BPF_REG_7, BPF_REG_0, 6), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_current_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 6), + + /* sysctl_get_current_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_current_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -E2BIG, 6), + + /* buf[0:6] == "Linux\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x000078756e694cULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_current_value sysctl:read EINVAL", + .insns = { + /* sysctl_get_current_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_current_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* sysctl_get_current_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_current_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 4), + + /* buf[0:8] is NUL-filled) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 0, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv6/conf/lo/stable_secret", /* -EIO */ + .open_flags = O_RDONLY, + .result = OP_EPERM, + }, + { + .descr = "sysctl_get_current_value sysctl:write ok", + .fixup_value_insn = 6, + .insns = { + /* sysctl_get_current_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_current_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* sysctl_get_current_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_current_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 4, 6), + + /* buf[0:4] == expected) */ + BPF_LD_IMM64(BPF_REG_8, FIXUP_SYSCTL_VALUE), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_WRONLY, + .newval = "600", /* same as default, should fail anyway */ + .result = OP_EPERM, + }, }; static size_t probe_prog_length(const struct bpf_insn *fp) @@ -362,6 +548,27 @@ static size_t probe_prog_length(const struct bpf_insn *fp) return len + 1; } +static int fixup_sysctl_value(const char *buf, size_t buf_len, + struct bpf_insn *prog, size_t insn_num) +{ + uint32_t value_num = 0; + uint8_t c, i; + + if (buf_len > sizeof(value_num)) { + log_err("Value is too big (%zd) to use in fixup", buf_len); + return -1; + } + + for (i = 0; i < buf_len; ++i) { + c = buf[i]; + value_num |= (c << i * 8); + } + + prog[insn_num].imm = value_num; + + return 0; +} + static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) { struct bpf_insn *prog = test->insns; @@ -374,6 +581,27 @@ static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) attr.insns_cnt = probe_prog_length(attr.insns); attr.license = "GPL"; + if (test->fixup_value_insn) { + char buf[128]; + ssize_t len; + int fd; + + fd = open(sysctl_path, O_RDONLY | O_CLOEXEC); + if (fd < 0) { + log_err("open(%s) failed", sysctl_path); + return -1; + } + len = read(fd, buf, sizeof(buf)); + if (len == -1) { + log_err("read(%s) failed", sysctl_path); + close(fd); + return -1; + } + close(fd); + if (fixup_sysctl_value(buf, len, prog, test->fixup_value_insn)) + return -1; + } + ret = bpf_load_program_xattr(&attr, bpf_log_buf, BPF_LOG_BUF_SIZE); if (ret < 0 && test->result != LOAD_REJECT) { log_err(">>> Loading program error.\n" From 786047dd08de3334722af036e2827ce3b34205cc Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 15:22:03 -0800 Subject: [PATCH 13/68] selftests/bpf: Test bpf_sysctl_{get,set}_new_value helpers Test that new value provided by user space on sysctl write can be read by bpf_sysctl_get_new_value and overridden by bpf_sysctl_set_new_value. # ./test_sysctl ... Test case: sysctl_get_new_value sysctl:read EINVAL .. [PASS] Test case: sysctl_get_new_value sysctl:write ok .. [PASS] Test case: sysctl_get_new_value sysctl:write ok long .. [PASS] Test case: sysctl_get_new_value sysctl:write E2BIG .. [PASS] Test case: sysctl_set_new_value sysctl:read EINVAL .. [PASS] Test case: sysctl_set_new_value sysctl:write ok .. [PASS] Summary: 22 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_sysctl.c | 222 ++++++++++++++++++++++ 1 file changed, 222 insertions(+) diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 9fb6560fa775..95437b72404f 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -536,6 +536,228 @@ static struct sysctl_test tests[] = { .newval = "600", /* same as default, should fail anyway */ .result = OP_EPERM, }, + { + .descr = "sysctl_get_new_value sysctl:read EINVAL", + .insns = { + /* sysctl_get_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* sysctl_get_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_new_value), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_get_new_value sysctl:write ok", + .insns = { + /* sysctl_get_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 4), + + /* sysctl_get_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_new_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 3, 4), + + /* buf[0:4] == "606\0") */ + BPF_LDX_MEM(BPF_W, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 0x00363036, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_WRONLY, + .newval = "606", + .result = OP_EPERM, + }, + { + .descr = "sysctl_get_new_value sysctl:write ok long", + .insns = { + /* sysctl_get_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -24), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 24), + + /* sysctl_get_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_new_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 23, 14), + + /* buf[0:8] == "3000000 " && */ + BPF_LD_IMM64(BPF_REG_8, 0x2030303030303033ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 10), + + /* buf[8:16] == "4000000 " && */ + BPF_LD_IMM64(BPF_REG_8, 0x2030303030303034ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 8), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 6), + + /* buf[16:24] == "6000000\0") */ + BPF_LD_IMM64(BPF_REG_8, 0x0030303030303036ULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 16), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_WRONLY, + .newval = "3000000 4000000 6000000", + .result = OP_EPERM, + }, + { + .descr = "sysctl_get_new_value sysctl:write E2BIG", + .insns = { + /* sysctl_get_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_B, BPF_REG_7, BPF_REG_0, 3), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_get_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 3), + + /* sysctl_get_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_get_new_value), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -E2BIG, 4), + + /* buf[0:3] == "60\0") */ + BPF_LDX_MEM(BPF_W, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 0x003036, 2), + + /* return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_JMP_A(1), + + /* else return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_WRONLY, + .newval = "606", + .result = OP_EPERM, + }, + { + .descr = "sysctl_set_new_value sysctl:read EINVAL", + .insns = { + /* sysctl_set_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_set_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 3), + + /* sysctl_set_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_set_new_value), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "sysctl_set_new_value sysctl:write ok", + .fixup_value_insn = 2, + .insns = { + /* sysctl_set_new_value arg2 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, FIXUP_SYSCTL_VALUE), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_2, BPF_REG_7), + + /* sysctl_set_new_value arg3 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_3, 3), + + /* sysctl_set_new_value(ctx, buf, buf_len) */ + BPF_EMIT_CALL(BPF_FUNC_sysctl_set_new_value), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 0, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_WRONLY, + .newval = "606", + .result = SUCCESS, + }, }; static size_t probe_prog_length(const struct bpf_insn *fp) From 9a1027e52535db1a0adf7831afdfce745dc0a061 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 8 Mar 2019 15:25:02 -0800 Subject: [PATCH 14/68] selftests/bpf: Test file_pos field in bpf_sysctl ctx Test access to file_pos field of bpf_sysctl context, both read (incl. narrow read) and write. # ./test_sysctl ... Test case: ctx:file_pos sysctl:read read ok .. [PASS] Test case: ctx:file_pos sysctl:read read ok narrow .. [PASS] Test case: ctx:file_pos sysctl:read write ok .. [PASS] ... Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_sysctl.c | 64 +++++++++++++++++++++++ 1 file changed, 64 insertions(+) diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 95437b72404f..43008aae32d3 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -30,6 +30,7 @@ struct sysctl_test { const char *sysctl; int open_flags; const char *newval; + const char *oldval; enum { LOAD_REJECT, ATTACH_REJECT, @@ -130,6 +131,64 @@ static struct sysctl_test tests[] = { .open_flags = O_RDONLY, .result = LOAD_REJECT, }, + { + .descr = "ctx:file_pos sysctl:read read ok", + .insns = { + /* If (file_pos == X) */ + BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_1, + offsetof(struct bpf_sysctl, file_pos)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 0, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "ctx:file_pos sysctl:read read ok narrow", + .insns = { + /* If (file_pos == X) */ + BPF_LDX_MEM(BPF_B, BPF_REG_7, BPF_REG_1, + offsetof(struct bpf_sysctl, file_pos)), + BPF_JMP_IMM(BPF_JNE, BPF_REG_7, 0, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + .descr = "ctx:file_pos sysctl:read write ok", + .insns = { + /* file_pos = X */ + BPF_MOV64_IMM(BPF_REG_0, 2), + BPF_STX_MEM(BPF_W, BPF_REG_1, BPF_REG_0, + offsetof(struct bpf_sysctl, file_pos)), + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "kernel/ostype", + .open_flags = O_RDONLY, + .oldval = "nux\n", + .result = SUCCESS, + }, { .descr = "sysctl_get_name sysctl_value:base ok", .insns = { @@ -848,6 +907,11 @@ static int access_sysctl(const char *sysctl_path, if (read(fd, buf, sizeof(buf)) == -1) goto err; + if (test->oldval && + strncmp(buf, test->oldval, strlen(test->oldval))) { + log_err("Read value %s != %s", buf, test->oldval); + goto err; + } } else if (test->open_flags == O_WRONLY) { if (!test->newval) { log_err("New value for sysctl is not set"); From 57c3bb725a3dd97d960d7e1cd0845d88de53217f Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Mon, 18 Mar 2019 16:57:10 -0700 Subject: [PATCH 15/68] bpf: Introduce ARG_PTR_TO_{INT,LONG} arg types Currently the way to pass result from BPF helper to BPF program is to provide memory area defined by pointer and size: func(void *, size_t). It works great for generic use-case, but for simple types, such as int, it's overkill and consumes two arguments when it could use just one. Introduce new argument types ARG_PTR_TO_INT and ARG_PTR_TO_LONG to be able to pass result from helper to program via pointer to int and long correspondingly: func(int *) or func(long *). New argument types are similar to ARG_PTR_TO_MEM with the following differences: * they don't require corresponding ARG_CONST_SIZE argument, predefined access sizes are used instead (32bit for int, 64bit for long); * it's possible to use more than one such an argument in a helper; * provided pointers have to be aligned. It's easy to introduce similar ARG_PTR_TO_CHAR and ARG_PTR_TO_SHORT argument types. It's not done due to lack of use-case though. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- include/linux/bpf.h | 2 ++ kernel/bpf/verifier.c | 29 +++++++++++++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index e4d4c1771ab0..fd06ada941ad 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -202,6 +202,8 @@ enum bpf_arg_type { ARG_ANYTHING, /* any (initialized) argument is ok */ ARG_PTR_TO_SPIN_LOCK, /* pointer to bpf_spin_lock */ ARG_PTR_TO_SOCK_COMMON, /* pointer to sock_common */ + ARG_PTR_TO_INT, /* pointer to int */ + ARG_PTR_TO_LONG, /* pointer to long */ }; /* type of values returned from helper functions */ diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 20808e3c95a8..15ab6fa817ce 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -2462,6 +2462,22 @@ static bool arg_type_is_mem_size(enum bpf_arg_type type) type == ARG_CONST_SIZE_OR_ZERO; } +static bool arg_type_is_int_ptr(enum bpf_arg_type type) +{ + return type == ARG_PTR_TO_INT || + type == ARG_PTR_TO_LONG; +} + +static int int_ptr_type_to_size(enum bpf_arg_type type) +{ + if (type == ARG_PTR_TO_INT) + return sizeof(u32); + else if (type == ARG_PTR_TO_LONG) + return sizeof(u64); + + return -EINVAL; +} + static int check_func_arg(struct bpf_verifier_env *env, u32 regno, enum bpf_arg_type arg_type, struct bpf_call_arg_meta *meta) @@ -2554,6 +2570,12 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno, type != expected_type) goto err_type; meta->raw_mode = arg_type == ARG_PTR_TO_UNINIT_MEM; + } else if (arg_type_is_int_ptr(arg_type)) { + expected_type = PTR_TO_STACK; + if (!type_is_pkt_pointer(type) && + type != PTR_TO_MAP_VALUE && + type != expected_type) + goto err_type; } else { verbose(env, "unsupported arg_type %d\n", arg_type); return -EFAULT; @@ -2635,6 +2657,13 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno, err = check_helper_mem_access(env, regno - 1, reg->umax_value, zero_size_allowed, meta); + } else if (arg_type_is_int_ptr(arg_type)) { + int size = int_ptr_type_to_size(arg_type); + + err = check_helper_mem_access(env, regno, size, false, meta); + if (err) + return err; + err = check_ptr_alignment(env, reg, 0, size, true); } return err; From d7a4cb9b6705a89937d12c8158a35a3145dc967a Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Mon, 18 Mar 2019 17:55:26 -0700 Subject: [PATCH 16/68] bpf: Introduce bpf_strtol and bpf_strtoul helpers Add bpf_strtol and bpf_strtoul to convert a string to long and unsigned long correspondingly. It's similar to user space strtol(3) and strtoul(3) with a few changes to the API: * instead of NUL-terminated C string the helpers expect buffer and buffer length; * resulting long or unsigned long is returned in a separate result-argument; * return value is used to indicate success or failure, on success number of consumed bytes is returned that can be used to identify position to read next if the buffer is expected to contain multiple integers; * instead of *base* argument, *flags* is used that provides base in 5 LSB, other bits are reserved for future use; * number of supported bases is limited. Documentation for the new helpers is provided in bpf.h UAPI. The helpers are made available to BPF_PROG_TYPE_CGROUP_SYSCTL programs to be able to convert string input to e.g. "ulongvec" output. E.g. "net/ipv4/tcp_mem" consists of three ulong integers. They can be parsed by calling to bpf_strtoul three times. Implementation notes: Implementation includes "../../lib/kstrtox.h" to reuse integer parsing functions. It's done exactly same way as fs/proc/base.c already does. Unfortunately existing kstrtoX function can't be used directly since they fail if any invalid character is present right after integer in the string. Existing simple_strtoX functions can't be used either since they're obsolete and don't handle overflow properly. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- include/linux/bpf.h | 2 + include/uapi/linux/bpf.h | 51 ++++++++++++++- kernel/bpf/cgroup.c | 4 ++ kernel/bpf/helpers.c | 131 +++++++++++++++++++++++++++++++++++++++ 4 files changed, 187 insertions(+), 1 deletion(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index fd06ada941ad..f15432d90728 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -989,6 +989,8 @@ extern const struct bpf_func_proto bpf_sk_redirect_map_proto; extern const struct bpf_func_proto bpf_spin_lock_proto; extern const struct bpf_func_proto bpf_spin_unlock_proto; extern const struct bpf_func_proto bpf_get_local_storage_proto; +extern const struct bpf_func_proto bpf_strtol_proto; +extern const struct bpf_func_proto bpf_strtoul_proto; /* Shared helpers among cBPF and eBPF. */ void bpf_user_rnd_init_once(void); diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 89976de909af..c26be24fd5e2 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2575,6 +2575,53 @@ union bpf_attr { * **-E2BIG** if the *buf_len* is too big. * * **-EINVAL** if sysctl is being read. + * + * int bpf_strtol(const char *buf, size_t buf_len, u64 flags, long *res) + * Description + * Convert the initial part of the string from buffer *buf* of + * size *buf_len* to a long integer according to the given base + * and save the result in *res*. + * + * The string may begin with an arbitrary amount of white space + * (as determined by isspace(3)) followed by a single optional '-' + * sign. + * + * Five least significant bits of *flags* encode base, other bits + * are currently unused. + * + * Base must be either 8, 10, 16 or 0 to detect it automatically + * similar to user space strtol(3). + * Return + * Number of characters consumed on success. Must be positive but + * no more than buf_len. + * + * **-EINVAL** if no valid digits were found or unsupported base + * was provided. + * + * **-ERANGE** if resulting value was out of range. + * + * int bpf_strtoul(const char *buf, size_t buf_len, u64 flags, unsigned long *res) + * Description + * Convert the initial part of the string from buffer *buf* of + * size *buf_len* to an unsigned long integer according to the + * given base and save the result in *res*. + * + * The string may begin with an arbitrary amount of white space + * (as determined by isspace(3)). + * + * Five least significant bits of *flags* encode base, other bits + * are currently unused. + * + * Base must be either 8, 10, 16 or 0 to detect it automatically + * similar to user space strtoul(3). + * Return + * Number of characters consumed on success. Must be positive but + * no more than buf_len. + * + * **-EINVAL** if no valid digits were found or unsupported base + * was provided. + * + * **-ERANGE** if resulting value was out of range. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2681,7 +2728,9 @@ union bpf_attr { FN(sysctl_get_name), \ FN(sysctl_get_current_value), \ FN(sysctl_get_new_value), \ - FN(sysctl_set_new_value), + FN(sysctl_set_new_value), \ + FN(strtol), \ + FN(strtoul), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index b2adf22139b3..789d4ab2336e 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -1016,6 +1016,10 @@ static const struct bpf_func_proto * sysctl_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { switch (func_id) { + case BPF_FUNC_strtol: + return &bpf_strtol_proto; + case BPF_FUNC_strtoul: + return &bpf_strtoul_proto; case BPF_FUNC_sysctl_get_name: return &bpf_sysctl_get_name_proto; case BPF_FUNC_sysctl_get_current_value: diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c index a411fc17d265..4266ffde07ca 100644 --- a/kernel/bpf/helpers.c +++ b/kernel/bpf/helpers.c @@ -18,6 +18,9 @@ #include #include #include +#include + +#include "../../lib/kstrtox.h" /* If kernel subsystem is allowing eBPF programs to call this function, * inside its own verifier_ops->get_func_proto() callback it should return @@ -363,4 +366,132 @@ const struct bpf_func_proto bpf_get_local_storage_proto = { .arg2_type = ARG_ANYTHING, }; #endif + +#define BPF_STRTOX_BASE_MASK 0x1F + +static int __bpf_strtoull(const char *buf, size_t buf_len, u64 flags, + unsigned long long *res, bool *is_negative) +{ + unsigned int base = flags & BPF_STRTOX_BASE_MASK; + const char *cur_buf = buf; + size_t cur_len = buf_len; + unsigned int consumed; + size_t val_len; + char str[64]; + + if (!buf || !buf_len || !res || !is_negative) + return -EINVAL; + + if (base != 0 && base != 8 && base != 10 && base != 16) + return -EINVAL; + + if (flags & ~BPF_STRTOX_BASE_MASK) + return -EINVAL; + + while (cur_buf < buf + buf_len && isspace(*cur_buf)) + ++cur_buf; + + *is_negative = (cur_buf < buf + buf_len && *cur_buf == '-'); + if (*is_negative) + ++cur_buf; + + consumed = cur_buf - buf; + cur_len -= consumed; + if (!cur_len) + return -EINVAL; + + cur_len = min(cur_len, sizeof(str) - 1); + memcpy(str, cur_buf, cur_len); + str[cur_len] = '\0'; + cur_buf = str; + + cur_buf = _parse_integer_fixup_radix(cur_buf, &base); + val_len = _parse_integer(cur_buf, base, res); + + if (val_len & KSTRTOX_OVERFLOW) + return -ERANGE; + + if (val_len == 0) + return -EINVAL; + + cur_buf += val_len; + consumed += cur_buf - str; + + return consumed; +} + +static int __bpf_strtoll(const char *buf, size_t buf_len, u64 flags, + long long *res) +{ + unsigned long long _res; + bool is_negative; + int err; + + err = __bpf_strtoull(buf, buf_len, flags, &_res, &is_negative); + if (err < 0) + return err; + if (is_negative) { + if ((long long)-_res > 0) + return -ERANGE; + *res = -_res; + } else { + if ((long long)_res < 0) + return -ERANGE; + *res = _res; + } + return err; +} + +BPF_CALL_4(bpf_strtol, const char *, buf, size_t, buf_len, u64, flags, + long *, res) +{ + long long _res; + int err; + + err = __bpf_strtoll(buf, buf_len, flags, &_res); + if (err < 0) + return err; + if (_res != (long)_res) + return -ERANGE; + *res = _res; + return err; +} + +const struct bpf_func_proto bpf_strtol_proto = { + .func = bpf_strtol, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_MEM, + .arg2_type = ARG_CONST_SIZE, + .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_PTR_TO_LONG, +}; + +BPF_CALL_4(bpf_strtoul, const char *, buf, size_t, buf_len, u64, flags, + unsigned long *, res) +{ + unsigned long long _res; + bool is_negative; + int err; + + err = __bpf_strtoull(buf, buf_len, flags, &_res, &is_negative); + if (err < 0) + return err; + if (is_negative) + return -EINVAL; + if (_res != (unsigned long)_res) + return -ERANGE; + *res = _res; + return err; +} + +const struct bpf_func_proto bpf_strtoul_proto = { + .func = bpf_strtoul, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_MEM, + .arg2_type = ARG_CONST_SIZE, + .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_PTR_TO_LONG, +}; #endif From b457e5534c9988bc6bcf30a22fa7cd765b14ecbd Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Mon, 18 Mar 2019 18:07:01 -0700 Subject: [PATCH 17/68] bpf: Sync bpf.h to tools/ Sync bpf_strtoX related bpf UAPI changes to tools/. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/include/uapi/linux/bpf.h | 51 +++++++++++++++++++++++++++++++++- 1 file changed, 50 insertions(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 89976de909af..c26be24fd5e2 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -2575,6 +2575,53 @@ union bpf_attr { * **-E2BIG** if the *buf_len* is too big. * * **-EINVAL** if sysctl is being read. + * + * int bpf_strtol(const char *buf, size_t buf_len, u64 flags, long *res) + * Description + * Convert the initial part of the string from buffer *buf* of + * size *buf_len* to a long integer according to the given base + * and save the result in *res*. + * + * The string may begin with an arbitrary amount of white space + * (as determined by isspace(3)) followed by a single optional '-' + * sign. + * + * Five least significant bits of *flags* encode base, other bits + * are currently unused. + * + * Base must be either 8, 10, 16 or 0 to detect it automatically + * similar to user space strtol(3). + * Return + * Number of characters consumed on success. Must be positive but + * no more than buf_len. + * + * **-EINVAL** if no valid digits were found or unsupported base + * was provided. + * + * **-ERANGE** if resulting value was out of range. + * + * int bpf_strtoul(const char *buf, size_t buf_len, u64 flags, unsigned long *res) + * Description + * Convert the initial part of the string from buffer *buf* of + * size *buf_len* to an unsigned long integer according to the + * given base and save the result in *res*. + * + * The string may begin with an arbitrary amount of white space + * (as determined by isspace(3)). + * + * Five least significant bits of *flags* encode base, other bits + * are currently unused. + * + * Base must be either 8, 10, 16 or 0 to detect it automatically + * similar to user space strtoul(3). + * Return + * Number of characters consumed on success. Must be positive but + * no more than buf_len. + * + * **-EINVAL** if no valid digits were found or unsupported base + * was provided. + * + * **-ERANGE** if resulting value was out of range. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -2681,7 +2728,9 @@ union bpf_attr { FN(sysctl_get_name), \ FN(sysctl_get_current_value), \ FN(sysctl_get_new_value), \ - FN(sysctl_set_new_value), + FN(sysctl_set_new_value), \ + FN(strtol), \ + FN(strtoul), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call From 99f57973ac5b8e38ee5b5e5f518cd5bf95e3b1e3 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sat, 23 Mar 2019 15:47:05 -0700 Subject: [PATCH 18/68] selftests/bpf: Add sysctl and strtoX helpers to bpf_helpers.h Add bpf_sysctl_* and bpf_strtoX helpers to bpf_helpers.h. Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/bpf_helpers.h | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/tools/testing/selftests/bpf/bpf_helpers.h b/tools/testing/selftests/bpf/bpf_helpers.h index e85d62cb53d0..59e221586cf7 100644 --- a/tools/testing/selftests/bpf/bpf_helpers.h +++ b/tools/testing/selftests/bpf/bpf_helpers.h @@ -192,6 +192,25 @@ static int (*bpf_skb_ecn_set_ce)(void *ctx) = static int (*bpf_tcp_check_syncookie)(struct bpf_sock *sk, void *ip, int ip_len, void *tcp, int tcp_len) = (void *) BPF_FUNC_tcp_check_syncookie; +static int (*bpf_sysctl_get_name)(void *ctx, char *buf, + unsigned long long buf_len, + unsigned long long flags) = + (void *) BPF_FUNC_sysctl_get_name; +static int (*bpf_sysctl_get_current_value)(void *ctx, char *buf, + unsigned long long buf_len) = + (void *) BPF_FUNC_sysctl_get_current_value; +static int (*bpf_sysctl_get_new_value)(void *ctx, char *buf, + unsigned long long buf_len) = + (void *) BPF_FUNC_sysctl_get_new_value; +static int (*bpf_sysctl_set_new_value)(void *ctx, const char *buf, + unsigned long long buf_len) = + (void *) BPF_FUNC_sysctl_set_new_value; +static int (*bpf_strtol)(const char *buf, unsigned long long buf_len, + unsigned long long flags, long *res) = + (void *) BPF_FUNC_strtol; +static int (*bpf_strtoul)(const char *buf, unsigned long long buf_len, + unsigned long long flags, unsigned long *res) = + (void *) BPF_FUNC_strtoul; /* llvm builtin functions that eBPF C program may use to * emit BPF_LD_ABS and BPF_LD_IND instructions From c2d5f12e4c6cf7e94cfcb98389db51557cf05f9a Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Mon, 18 Mar 2019 18:17:03 -0700 Subject: [PATCH 19/68] selftests/bpf: Test ARG_PTR_TO_LONG arg type Test that verifier handles new argument types properly, including uninitialized or partially initialized value, misaligned stack access, etc. Example of output: #456/p ARG_PTR_TO_LONG uninitialized OK #457/p ARG_PTR_TO_LONG half-uninitialized OK #458/p ARG_PTR_TO_LONG misaligned OK #459/p ARG_PTR_TO_LONG size < sizeof(long) OK #460/p ARG_PTR_TO_LONG initialized OK Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- .../testing/selftests/bpf/verifier/int_ptr.c | 160 ++++++++++++++++++ 1 file changed, 160 insertions(+) create mode 100644 tools/testing/selftests/bpf/verifier/int_ptr.c diff --git a/tools/testing/selftests/bpf/verifier/int_ptr.c b/tools/testing/selftests/bpf/verifier/int_ptr.c new file mode 100644 index 000000000000..ca3b4729df66 --- /dev/null +++ b/tools/testing/selftests/bpf/verifier/int_ptr.c @@ -0,0 +1,160 @@ +{ + "ARG_PTR_TO_LONG uninitialized", + .insns = { + /* bpf_strtoul arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* bpf_strtoul arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* bpf_strtoul arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* bpf_strtoul arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + /* bpf_strtoul() */ + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .result = REJECT, + .prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL, + .errstr = "invalid indirect read from stack off -16+0 size 8", +}, +{ + "ARG_PTR_TO_LONG half-uninitialized", + .insns = { + /* bpf_strtoul arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* bpf_strtoul arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* bpf_strtoul arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* bpf_strtoul arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_W, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + /* bpf_strtoul() */ + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .result = REJECT, + .prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL, + .errstr = "invalid indirect read from stack off -16+4 size 8", +}, +{ + "ARG_PTR_TO_LONG misaligned", + .insns = { + /* bpf_strtoul arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* bpf_strtoul arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* bpf_strtoul arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* bpf_strtoul arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -12), + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_STX_MEM(BPF_W, BPF_REG_7, BPF_REG_0, 0), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 4), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + /* bpf_strtoul() */ + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .result = REJECT, + .prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL, + .errstr = "misaligned stack access off (0x0; 0x0)+-20+0 size 8", +}, +{ + "ARG_PTR_TO_LONG size < sizeof(long)", + .insns = { + /* bpf_strtoul arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -16), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* bpf_strtoul arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* bpf_strtoul arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* bpf_strtoul arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, 12), + BPF_STX_MEM(BPF_W, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + /* bpf_strtoul() */ + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .result = REJECT, + .prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL, + .errstr = "invalid stack type R4 off=-4 access_size=8", +}, +{ + "ARG_PTR_TO_LONG initialized", + .insns = { + /* bpf_strtoul arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* bpf_strtoul arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* bpf_strtoul arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* bpf_strtoul arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + /* bpf_strtoul() */ + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .result = ACCEPT, + .prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL, +}, From 8549ddc832d6f36be47279d201e95cc8ade6faa9 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Mon, 18 Mar 2019 18:21:18 -0700 Subject: [PATCH 20/68] selftests/bpf: Test bpf_strtol and bpf_strtoul helpers Test that bpf_strtol and bpf_strtoul helpers can be used to convert provided buffer to long or unsigned long correspondingly and return both correct result and number of consumed bytes, or proper errno. Example of output: # ./test_sysctl .. Test case: bpf_strtoul one number string .. [PASS] Test case: bpf_strtoul multi number string .. [PASS] Test case: bpf_strtoul buf_len = 0, reject .. [PASS] Test case: bpf_strtoul supported base, ok .. [PASS] Test case: bpf_strtoul unsupported base, EINVAL .. [PASS] Test case: bpf_strtoul buf with spaces only, EINVAL .. [PASS] Test case: bpf_strtoul negative number, EINVAL .. [PASS] Test case: bpf_strtol negative number, ok .. [PASS] Test case: bpf_strtol hex number, ok .. [PASS] Test case: bpf_strtol max long .. [PASS] Test case: bpf_strtol overflow, ERANGE .. [PASS] Summary: 36 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_sysctl.c | 485 ++++++++++++++++++++++ 1 file changed, 485 insertions(+) diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 43008aae32d3..885675480af9 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -817,6 +817,491 @@ static struct sysctl_test tests[] = { .newval = "606", .result = SUCCESS, }, + { + "bpf_strtoul one number string", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 3, 4), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 600, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtoul multi number string", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + /* "600 602\0" */ + BPF_LD_IMM64(BPF_REG_0, 0x0032303620303036ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 8), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 3, 18), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 600, 16), + + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_ALU64_REG(BPF_ADD, BPF_REG_7, BPF_REG_0), + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 8), + BPF_ALU64_REG(BPF_SUB, BPF_REG_2, BPF_REG_0), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -16), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 4, 4), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 602, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtoul buf_len = 0, reject", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 0), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = LOAD_REJECT, + }, + { + "bpf_strtoul supported base, ok", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00373730), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 8), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 3, 4), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 63, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtoul unsupported base, EINVAL", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00303036), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 3), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtoul buf with spaces only, EINVAL", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x090a0c0d), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtoul negative number, EINVAL", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00362d0a), /* " -6\0" */ + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtoul), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -EINVAL, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtol negative number, ok", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x00362d0a), /* " -6\0" */ + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 10), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtol), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 3, 4), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, -6, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtol hex number, ok", + .insns = { + /* arg1 (buf) */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_MOV64_IMM(BPF_REG_0, 0x65667830), /* "0xfe" */ + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 4), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtol), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 4, 4), + /* res == expected) */ + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 254, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtol max long", + .insns = { + /* arg1 (buf) 9223372036854775807 */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -24), + BPF_LD_IMM64(BPF_REG_0, 0x3032373333323239ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_LD_IMM64(BPF_REG_0, 0x3537373435383633ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 8), + BPF_LD_IMM64(BPF_REG_0, 0x0000000000373038ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 16), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 19), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtol), + + /* if (ret == expected && */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 19, 6), + /* res == expected) */ + BPF_LD_IMM64(BPF_REG_8, 0x7fffffffffffffffULL), + BPF_LDX_MEM(BPF_DW, BPF_REG_9, BPF_REG_7, 0), + BPF_JMP_REG(BPF_JNE, BPF_REG_8, BPF_REG_9, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, + { + "bpf_strtol overflow, ERANGE", + .insns = { + /* arg1 (buf) 9223372036854775808 */ + BPF_MOV64_REG(BPF_REG_7, BPF_REG_10), + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -24), + BPF_LD_IMM64(BPF_REG_0, 0x3032373333323239ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_LD_IMM64(BPF_REG_0, 0x3537373435383633ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 8), + BPF_LD_IMM64(BPF_REG_0, 0x0000000000383038ULL), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 16), + + BPF_MOV64_REG(BPF_REG_1, BPF_REG_7), + + /* arg2 (buf_len) */ + BPF_MOV64_IMM(BPF_REG_2, 19), + + /* arg3 (flags) */ + BPF_MOV64_IMM(BPF_REG_3, 0), + + /* arg4 (res) */ + BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, -8), + BPF_STX_MEM(BPF_DW, BPF_REG_7, BPF_REG_0, 0), + BPF_MOV64_REG(BPF_REG_4, BPF_REG_7), + + BPF_EMIT_CALL(BPF_FUNC_strtol), + + /* if (ret == expected) */ + BPF_JMP_IMM(BPF_JNE, BPF_REG_0, -ERANGE, 2), + + /* return ALLOW; */ + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_JMP_A(1), + + /* else return DENY; */ + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, }; static size_t probe_prog_length(const struct bpf_insn *fp) From 7568f4cbbeae687e4c545516275479f50c15a7cc Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sat, 23 Mar 2019 15:51:00 -0700 Subject: [PATCH 21/68] selftests/bpf: C based test for sysctl and strtoX Add C based test for a few bpf_sysctl_* helpers and bpf_strtoul. Make sure that sysctl can be identified by name and that multiple integers can be parsed from sysctl value with bpf_strtoul. net/ipv4/tcp_mem is chosen as a testing sysctl, it contains 3 unsigned longs, they all are parsed and compared (val[0] < val[1] < val[2]). Example of output: # ./test_sysctl ... Test case: C prog: deny all writes .. [PASS] Test case: C prog: deny access by name .. [PASS] Test case: C prog: read tcp_mem .. [PASS] Summary: 39 PASSED, 0 FAILED Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/progs/test_sysctl_prog.c | 70 +++++++++++++++++++ tools/testing/selftests/bpf/test_sysctl.c | 57 ++++++++++++++- 2 files changed, 126 insertions(+), 1 deletion(-) create mode 100644 tools/testing/selftests/bpf/progs/test_sysctl_prog.c diff --git a/tools/testing/selftests/bpf/progs/test_sysctl_prog.c b/tools/testing/selftests/bpf/progs/test_sysctl_prog.c new file mode 100644 index 000000000000..a295cad805d7 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_sysctl_prog.c @@ -0,0 +1,70 @@ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2019 Facebook + +#include +#include + +#include +#include + +#include "bpf_helpers.h" +#include "bpf_util.h" + +/* Max supported length of a string with unsigned long in base 10 (pow2 - 1). */ +#define MAX_ULONG_STR_LEN 0xF + +/* Max supported length of sysctl value string (pow2). */ +#define MAX_VALUE_STR_LEN 0x40 + +static __always_inline int is_tcp_mem(struct bpf_sysctl *ctx) +{ + char tcp_mem_name[] = "net/ipv4/tcp_mem"; + unsigned char i; + char name[64]; + int ret; + + memset(name, 0, sizeof(name)); + ret = bpf_sysctl_get_name(ctx, name, sizeof(name), 0); + if (ret < 0 || ret != sizeof(tcp_mem_name) - 1) + return 0; + +#pragma clang loop unroll(full) + for (i = 0; i < sizeof(tcp_mem_name); ++i) + if (name[i] != tcp_mem_name[i]) + return 0; + + return 1; +} + +SEC("cgroup/sysctl") +int sysctl_tcp_mem(struct bpf_sysctl *ctx) +{ + unsigned long tcp_mem[3] = {0, 0, 0}; + char value[MAX_VALUE_STR_LEN]; + unsigned char i, off = 0; + int ret; + + if (ctx->write) + return 0; + + if (!is_tcp_mem(ctx)) + return 0; + + ret = bpf_sysctl_get_current_value(ctx, value, MAX_VALUE_STR_LEN); + if (ret < 0 || ret >= MAX_VALUE_STR_LEN) + return 0; + +#pragma clang loop unroll(full) + for (i = 0; i < ARRAY_SIZE(tcp_mem); ++i) { + ret = bpf_strtoul(value + off, MAX_ULONG_STR_LEN, 0, + tcp_mem + i); + if (ret <= 0 || ret > MAX_ULONG_STR_LEN) + return 0; + off += ret & MAX_ULONG_STR_LEN; + } + + + return tcp_mem[0] < tcp_mem[1] && tcp_mem[1] < tcp_mem[2]; +} + +char _license[] SEC("license") = "GPL"; diff --git a/tools/testing/selftests/bpf/test_sysctl.c b/tools/testing/selftests/bpf/test_sysctl.c index 885675480af9..a3bebd7c68dd 100644 --- a/tools/testing/selftests/bpf/test_sysctl.c +++ b/tools/testing/selftests/bpf/test_sysctl.c @@ -11,6 +11,7 @@ #include #include +#include #include "bpf_rlimit.h" #include "bpf_util.h" @@ -26,6 +27,7 @@ struct sysctl_test { const char *descr; size_t fixup_value_insn; struct bpf_insn insns[MAX_INSNS]; + const char *prog_file; enum bpf_attach_type attach_type; const char *sysctl; int open_flags; @@ -1302,6 +1304,31 @@ static struct sysctl_test tests[] = { .open_flags = O_RDONLY, .result = SUCCESS, }, + { + "C prog: deny all writes", + .prog_file = "./test_sysctl_prog.o", + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_WRONLY, + .newval = "123 456 789", + .result = OP_EPERM, + }, + { + "C prog: deny access by name", + .prog_file = "./test_sysctl_prog.o", + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/route/mtu_expires", + .open_flags = O_RDONLY, + .result = OP_EPERM, + }, + { + "C prog: read tcp_mem", + .prog_file = "./test_sysctl_prog.o", + .attach_type = BPF_CGROUP_SYSCTL, + .sysctl = "net/ipv4/tcp_mem", + .open_flags = O_RDONLY, + .result = SUCCESS, + }, }; static size_t probe_prog_length(const struct bpf_insn *fp) @@ -1335,7 +1362,8 @@ static int fixup_sysctl_value(const char *buf, size_t buf_len, return 0; } -static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) +static int load_sysctl_prog_insns(struct sysctl_test *test, + const char *sysctl_path) { struct bpf_insn *prog = test->insns; struct bpf_load_program_attr attr; @@ -1377,6 +1405,33 @@ static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) return ret; } +static int load_sysctl_prog_file(struct sysctl_test *test) +{ + struct bpf_prog_load_attr attr; + struct bpf_object *obj; + int prog_fd; + + memset(&attr, 0, sizeof(struct bpf_prog_load_attr)); + attr.file = test->prog_file; + attr.prog_type = BPF_PROG_TYPE_CGROUP_SYSCTL; + + if (bpf_prog_load_xattr(&attr, &obj, &prog_fd)) { + if (test->result != LOAD_REJECT) + log_err(">>> Loading program (%s) error.\n", + test->prog_file); + return -1; + } + + return prog_fd; +} + +static int load_sysctl_prog(struct sysctl_test *test, const char *sysctl_path) +{ + return test->prog_file + ? load_sysctl_prog_file(test) + : load_sysctl_prog_insns(test, sysctl_path); +} + static int access_sysctl(const char *sysctl_path, const struct sysctl_test *test) { From 51356ac89b5a15e5207e8740d5f4f8b71cb7332f Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Fri, 12 Apr 2019 16:01:01 -0700 Subject: [PATCH 22/68] bpf: Fix distinct pointer types warning for ARCH=i386 Fix a new warning reported by kbuild for make ARCH=i386: In file included from kernel/bpf/cgroup.c:11:0: kernel/bpf/cgroup.c: In function '__cgroup_bpf_run_filter_sysctl': include/linux/kernel.h:827:29: warning: comparison of distinct pointer types lacks a cast (!!(sizeof((typeof(x) *)1 == (typeof(y) *)1))) ^ include/linux/kernel.h:841:4: note: in expansion of macro '__typecheck' (__typecheck(x, y) && __no_side_effects(x, y)) ^~~~~~~~~~~ include/linux/kernel.h:851:24: note: in expansion of macro '__safe_cmp' __builtin_choose_expr(__safe_cmp(x, y), \ ^~~~~~~~~~ include/linux/kernel.h:860:19: note: in expansion of macro '__careful_cmp' #define min(x, y) __careful_cmp(x, y, <) ^~~~~~~~~~~~~ >> kernel/bpf/cgroup.c:837:17: note: in expansion of macro 'min' ctx.new_len = min(PAGE_SIZE, *pcount); ^~~ Fixes: 4e63acdff864 ("bpf: Introduce bpf_sysctl_{get,set}_new_value helpers") Signed-off-by: Andrey Ignatov Signed-off-by: Alexei Starovoitov --- kernel/bpf/cgroup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index 789d4ab2336e..e58a6c247f56 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -839,7 +839,7 @@ int __cgroup_bpf_run_filter_sysctl(struct ctl_table_header *head, * buffer bigger than provided by user. */ ctx.new_val = kmalloc_track_caller(PAGE_SIZE, GFP_KERNEL); - ctx.new_len = min(PAGE_SIZE, *pcount); + ctx.new_len = min_t(size_t, PAGE_SIZE, *pcount); if (!ctx.new_val || copy_from_user(ctx.new_val, buf, ctx.new_len)) /* Let BPF program decide how to proceed. */ From 1b04aee7e2182454a663950e68084fa5ada9625a Mon Sep 17 00:00:00 2001 From: Jiong Wang Date: Fri, 12 Apr 2019 22:59:34 +0100 Subject: [PATCH 23/68] bpf: refactor propagate_liveness to eliminate duplicated for loop Propagation for register and stack slot are finished in separate for loop, while they are perfect to be put into a single loop. This could also let them share some common variables in later patches. Signed-off-by: Jiong Wang Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 15ab6fa817ce..da285df492fd 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6254,10 +6254,8 @@ static int propagate_liveness(struct bpf_verifier_env *env, return err; } } - } - /* ... and stack slots */ - for (frame = 0; frame <= vstate->curframe; frame++) { + /* Propagate stack slots. */ state = vstate->frame[frame]; parent = vparent->frame[frame]; for (i = 0; i < state->allocated_stack / BPF_REG_SIZE && From 3f8cafa4131f67d47c8de326c7dcd561cc65fb38 Mon Sep 17 00:00:00 2001 From: Jiong Wang Date: Fri, 12 Apr 2019 22:59:35 +0100 Subject: [PATCH 24/68] bpf: refactor propagate_liveness to eliminate code redundance Access to reg states were not factored out, the consequence is long code for dereferencing them which made the indentation not good for reading. This patch factor out these code so the core code in the loop could be easier to follow. Reviewed-by: Jakub Kicinski Signed-off-by: Jiong Wang Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 34 ++++++++++++++++++++-------------- 1 file changed, 20 insertions(+), 14 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index da285df492fd..6fd36a8ba1a0 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6232,8 +6232,9 @@ static int propagate_liveness(struct bpf_verifier_env *env, const struct bpf_verifier_state *vstate, struct bpf_verifier_state *vparent) { - int i, frame, err = 0; + struct bpf_reg_state *state_reg, *parent_reg; struct bpf_func_state *state, *parent; + int i, frame, err = 0; if (vparent->curframe != vstate->curframe) { WARN(1, "propagate_live: parent frame %d current frame %d\n", @@ -6243,28 +6244,33 @@ static int propagate_liveness(struct bpf_verifier_env *env, /* Propagate read liveness of registers... */ BUILD_BUG_ON(BPF_REG_FP + 1 != MAX_BPF_REG); for (frame = 0; frame <= vstate->curframe; frame++) { + parent = vparent->frame[frame]; + state = vstate->frame[frame]; + parent_reg = parent->regs; + state_reg = state->regs; /* We don't need to worry about FP liveness, it's read-only */ for (i = frame < vstate->curframe ? BPF_REG_6 : 0; i < BPF_REG_FP; i++) { - if (vparent->frame[frame]->regs[i].live & REG_LIVE_READ) + if (parent_reg[i].live & REG_LIVE_READ) continue; - if (vstate->frame[frame]->regs[i].live & REG_LIVE_READ) { - err = mark_reg_read(env, &vstate->frame[frame]->regs[i], - &vparent->frame[frame]->regs[i]); - if (err) - return err; - } + if (!(state_reg[i].live & REG_LIVE_READ)) + continue; + err = mark_reg_read(env, &state_reg[i], &parent_reg[i]); + if (err) + return err; } /* Propagate stack slots. */ - state = vstate->frame[frame]; - parent = vparent->frame[frame]; for (i = 0; i < state->allocated_stack / BPF_REG_SIZE && i < parent->allocated_stack / BPF_REG_SIZE; i++) { - if (parent->stack[i].spilled_ptr.live & REG_LIVE_READ) + parent_reg = &parent->stack[i].spilled_ptr; + state_reg = &state->stack[i].spilled_ptr; + if (parent_reg->live & REG_LIVE_READ) continue; - if (state->stack[i].spilled_ptr.live & REG_LIVE_READ) - mark_reg_read(env, &state->stack[i].spilled_ptr, - &parent->stack[i].spilled_ptr); + if (!(state_reg->live & REG_LIVE_READ)) + continue; + err = mark_reg_read(env, state_reg, parent_reg); + if (err) + return err; } } return err; From 55e7f3b5ac94a8c9f2e35961a45e9aa526a9e41d Mon Sep 17 00:00:00 2001 From: Jiong Wang Date: Fri, 12 Apr 2019 22:59:36 +0100 Subject: [PATCH 25/68] bpf: factor out reg and stack slot propagation into "propagate_liveness_reg" After code refactor in previous patches, the propagation logic inside the for loop in "propagate_liveness" becomes clear that they are good enough to be factored out into a common function "propagate_liveness_reg". Reviewed-by: Jakub Kicinski Signed-off-by: Jiong Wang Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 30 ++++++++++++++++++++---------- 1 file changed, 20 insertions(+), 10 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6fd36a8ba1a0..3fdb301c4f8c 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -6221,6 +6221,22 @@ static bool states_equal(struct bpf_verifier_env *env, return true; } +static int propagate_liveness_reg(struct bpf_verifier_env *env, + struct bpf_reg_state *reg, + struct bpf_reg_state *parent_reg) +{ + int err; + + if (parent_reg->live & REG_LIVE_READ || !(reg->live & REG_LIVE_READ)) + return 0; + + err = mark_reg_read(env, reg, parent_reg); + if (err) + return err; + + return 0; +} + /* A write screens off any subsequent reads; but write marks come from the * straight-line code between a state and its parent. When we arrive at an * equivalent state (jump target or such) we didn't arrive by the straight-line @@ -6250,11 +6266,8 @@ static int propagate_liveness(struct bpf_verifier_env *env, state_reg = state->regs; /* We don't need to worry about FP liveness, it's read-only */ for (i = frame < vstate->curframe ? BPF_REG_6 : 0; i < BPF_REG_FP; i++) { - if (parent_reg[i].live & REG_LIVE_READ) - continue; - if (!(state_reg[i].live & REG_LIVE_READ)) - continue; - err = mark_reg_read(env, &state_reg[i], &parent_reg[i]); + err = propagate_liveness_reg(env, &state_reg[i], + &parent_reg[i]); if (err) return err; } @@ -6264,11 +6277,8 @@ static int propagate_liveness(struct bpf_verifier_env *env, i < parent->allocated_stack / BPF_REG_SIZE; i++) { parent_reg = &parent->stack[i].spilled_ptr; state_reg = &state->stack[i].spilled_ptr; - if (parent_reg->live & REG_LIVE_READ) - continue; - if (!(state_reg->live & REG_LIVE_READ)) - continue; - err = mark_reg_read(env, state_reg, parent_reg); + err = propagate_liveness_reg(env, state_reg, + parent_reg); if (err) return err; } From c342dc109aa5a4f0bb36335cb441aaafc98b98ef Mon Sep 17 00:00:00 2001 From: Jiong Wang Date: Fri, 12 Apr 2019 22:59:37 +0100 Subject: [PATCH 26/68] bpf: refactor "check_reg_arg" to eliminate code redundancy There are a few "regs[regno]" here are there across "check_reg_arg", this patch factor it out into a simple "reg" pointer. The intention is to simplify code indentation and make the later patches in this set look cleaner. Reviewed-by: Jakub Kicinski Signed-off-by: Jiong Wang Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 3fdb301c4f8c..c7220153c5b1 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1177,30 +1177,32 @@ static int check_reg_arg(struct bpf_verifier_env *env, u32 regno, { struct bpf_verifier_state *vstate = env->cur_state; struct bpf_func_state *state = vstate->frame[vstate->curframe]; - struct bpf_reg_state *regs = state->regs; + struct bpf_reg_state *reg, *regs = state->regs; if (regno >= MAX_BPF_REG) { verbose(env, "R%d is invalid\n", regno); return -EINVAL; } + reg = ®s[regno]; if (t == SRC_OP) { /* check whether register used as source operand can be read */ - if (regs[regno].type == NOT_INIT) { + if (reg->type == NOT_INIT) { verbose(env, "R%d !read_ok\n", regno); return -EACCES; } /* We don't need to worry about FP liveness because it's read-only */ - if (regno != BPF_REG_FP) - return mark_reg_read(env, ®s[regno], - regs[regno].parent); + if (regno == BPF_REG_FP) + return 0; + + return mark_reg_read(env, reg, reg->parent); } else { /* check whether register used as dest operand can be written to */ if (regno == BPF_REG_FP) { verbose(env, "frame pointer is read only\n"); return -EACCES; } - regs[regno].live |= REG_LIVE_WRITTEN; + reg->live |= REG_LIVE_WRITTEN; if (t == DST_OP) mark_reg_unknown(env, regs, regno); } From 68e5ab1fc8bd9c17ed94ac172ffe0b9b7e85a59a Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Mon, 15 Apr 2019 14:58:39 -0700 Subject: [PATCH 27/68] kbuild: handle old pahole more gracefully when generating BTF When CONFIG_DEBUG_INFO_BTF is enabled but available version of pahole is too old to support BTF generation, build script is supposed to emit warning and proceed with the build. Due to using exit instead of return from BASH function, existing handling code prematurely exits exit code 0, not completing some of the build steps. This patch fixes issue by correctly returning just from gen_btf() function only. Fixes: e83b9f55448a ("kbuild: add ability to generate BTF type info for vmlinux") Cc: Masahiro Yamada Cc: Arnaldo Carvalho de Melo Cc: Daniel Borkmann Cc: Alexei Starovoitov Cc: Yonghong Song Cc: Martin KaFai Lau Signed-off-by: Andrii Nakryiko Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- scripts/link-vmlinux.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/scripts/link-vmlinux.sh b/scripts/link-vmlinux.sh index dd2b31ccca6a..6a148d0d51bf 100755 --- a/scripts/link-vmlinux.sh +++ b/scripts/link-vmlinux.sh @@ -99,7 +99,7 @@ gen_btf() pahole_ver=$(${PAHOLE} --version | sed -E 's/v([0-9]+)\.([0-9]+)/\1\2/') if [ "${pahole_ver}" -lt "113" ]; then info "BTF" "${1}: pahole version $(${PAHOLE} --version) is too old, need at least v1.13" - exit 0 + return 0 fi info "BTF" ${1} From 189cf5a4a7d5e0349f3079f05a8ccf4e4a2f0c3e Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Mon, 15 Apr 2019 16:48:07 -0700 Subject: [PATCH 28/68] btf: add support for VAR and DATASEC in btf_dedup() This patch adds support for VAR and DATASEC in btf_dedup(). VAR/DATASEC are never deduplicated, but they need to be processed anyway as types they refer to might need to be remapped due to deduplication and compaction. Cc: Daniel Borkmann Cc: Yonghong Song Cc: Alexei Starovoitov Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann --- tools/lib/bpf/btf.c | 29 +++++++++++++++++++++++++++-- 1 file changed, 27 insertions(+), 2 deletions(-) diff --git a/tools/lib/bpf/btf.c b/tools/lib/bpf/btf.c index 701e7e28ada3..75eaf10b9e1a 100644 --- a/tools/lib/bpf/btf.c +++ b/tools/lib/bpf/btf.c @@ -1354,8 +1354,16 @@ static struct btf_dedup *btf_dedup_new(struct btf *btf, struct btf_ext *btf_ext, } /* special BTF "void" type is made canonical immediately */ d->map[0] = 0; - for (i = 1; i <= btf->nr_types; i++) - d->map[i] = BTF_UNPROCESSED_ID; + for (i = 1; i <= btf->nr_types; i++) { + struct btf_type *t = d->btf->types[i]; + __u16 kind = BTF_INFO_KIND(t->info); + + /* VAR and DATASEC are never deduped and are self-canonical */ + if (kind == BTF_KIND_VAR || kind == BTF_KIND_DATASEC) + d->map[i] = i; + else + d->map[i] = BTF_UNPROCESSED_ID; + } d->hypot_map = malloc(sizeof(__u32) * (1 + btf->nr_types)); if (!d->hypot_map) { @@ -1946,6 +1954,8 @@ static int btf_dedup_prim_type(struct btf_dedup *d, __u32 type_id) case BTF_KIND_UNION: case BTF_KIND_FUNC: case BTF_KIND_FUNC_PROTO: + case BTF_KIND_VAR: + case BTF_KIND_DATASEC: return 0; case BTF_KIND_INT: @@ -2699,6 +2709,7 @@ static int btf_dedup_remap_type(struct btf_dedup *d, __u32 type_id) case BTF_KIND_PTR: case BTF_KIND_TYPEDEF: case BTF_KIND_FUNC: + case BTF_KIND_VAR: r = btf_dedup_remap_type_id(d, t->type); if (r < 0) return r; @@ -2753,6 +2764,20 @@ static int btf_dedup_remap_type(struct btf_dedup *d, __u32 type_id) break; } + case BTF_KIND_DATASEC: { + struct btf_var_secinfo *var = (struct btf_var_secinfo *)(t + 1); + __u16 vlen = BTF_INFO_VLEN(t->info); + + for (i = 0; i < vlen; i++) { + r = btf_dedup_remap_type_id(d, var->type); + if (r < 0) + return r; + var->type = r; + var++; + } + break; + } + default: return -EINVAL; } From efb2ddc4ce5dba9b6c5ec106528d18a645424f3f Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Mon, 15 Apr 2019 16:48:08 -0700 Subject: [PATCH 29/68] selftests/btf: add VAR and DATASEC case for dedup tests Add test case verifying that dedup happens (INTs are deduped in this case) and VAR/DATASEC types are not deduped, but have their referenced type IDs adjusted correctly. Cc: Daniel Borkmann Cc: Yonghong Song Cc: Alexei Starovoitov Signed-off-by: Andrii Nakryiko Signed-off-by: Daniel Borkmann --- tools/testing/selftests/bpf/test_btf.c | 49 ++++++++++++++++++++++++++ 1 file changed, 49 insertions(+) diff --git a/tools/testing/selftests/bpf/test_btf.c b/tools/testing/selftests/bpf/test_btf.c index 44cd3378d216..f8eb7987b794 100644 --- a/tools/testing/selftests/bpf/test_btf.c +++ b/tools/testing/selftests/bpf/test_btf.c @@ -6642,6 +6642,51 @@ const struct btf_dedup_test dedup_tests[] = { .dont_resolve_fwds = false, }, }, +{ + .descr = "dedup: datasec and vars pass-through", + .input = { + .raw_types = { + /* int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + /* static int t */ + BTF_VAR_ENC(NAME_NTH(2), 1, 0), /* [2] */ + /* .bss section */ /* [3] */ + BTF_TYPE_ENC(NAME_NTH(1), BTF_INFO_ENC(BTF_KIND_DATASEC, 0, 1), 4), + BTF_VAR_SECINFO_ENC(2, 0, 4), + /* int, referenced from [5] */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [4] */ + /* another static int t */ + BTF_VAR_ENC(NAME_NTH(2), 4, 0), /* [5] */ + /* another .bss section */ /* [6] */ + BTF_TYPE_ENC(NAME_NTH(1), BTF_INFO_ENC(BTF_KIND_DATASEC, 0, 1), 4), + BTF_VAR_SECINFO_ENC(5, 0, 4), + BTF_END_RAW, + }, + BTF_STR_SEC("\0.bss\0t"), + }, + .expect = { + .raw_types = { + /* int */ + BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4), /* [1] */ + /* static int t */ + BTF_VAR_ENC(NAME_NTH(2), 1, 0), /* [2] */ + /* .bss section */ /* [3] */ + BTF_TYPE_ENC(NAME_NTH(1), BTF_INFO_ENC(BTF_KIND_DATASEC, 0, 1), 4), + BTF_VAR_SECINFO_ENC(2, 0, 4), + /* another static int t */ + BTF_VAR_ENC(NAME_NTH(2), 1, 0), /* [4] */ + /* another .bss section */ /* [5] */ + BTF_TYPE_ENC(NAME_NTH(1), BTF_INFO_ENC(BTF_KIND_DATASEC, 0, 1), 4), + BTF_VAR_SECINFO_ENC(4, 0, 4), + BTF_END_RAW, + }, + BTF_STR_SEC("\0.bss\0t"), + }, + .opts = { + .dont_resolve_fwds = false, + .dedup_table_size = 1 + }, +}, }; @@ -6671,6 +6716,10 @@ static int btf_type_size(const struct btf_type *t) return base_size + vlen * sizeof(struct btf_member); case BTF_KIND_FUNC_PROTO: return base_size + vlen * sizeof(struct btf_param); + case BTF_KIND_VAR: + return base_size + sizeof(struct btf_var); + case BTF_KIND_DATASEC: + return base_size + vlen * sizeof(struct btf_var_secinfo); default: fprintf(stderr, "Unsupported BTF_KIND:%u\n", kind); return -EINVAL; From bcbccad694b7ef0de9a993ecd918231c10a1496a Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Thu, 11 Apr 2019 15:53:16 -0700 Subject: [PATCH 30/68] selftests/bpf: bring back (void *) cast to set_ipv4_csum in test_tc_tunnel It was removed in commit 166b5a7f2ca3 ("selftests_bpf: extend test_tc_tunnel for UDP encap") without any explanation. Otherwise I see: progs/test_tc_tunnel.c:160:17: warning: taking address of packed member 'ip' of class or structure 'v4hdr' may result in an unaligned pointer value [-Waddress-of-packed-member] set_ipv4_csum(&h_outer.ip); ^~~~~~~~~~ 1 warning generated. Cc: Alan Maguire Cc: Willem de Bruijn Fixes: 166b5a7f2ca3 ("selftests_bpf: extend test_tc_tunnel for UDP encap") Signed-off-by: Stanislav Fomichev Acked-by: Song Liu Reviewed-by: Alan Maguire Signed-off-by: Daniel Borkmann --- tools/testing/selftests/bpf/progs/test_tc_tunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/progs/test_tc_tunnel.c b/tools/testing/selftests/bpf/progs/test_tc_tunnel.c index bcb00d737e95..ab56a6a72b7a 100644 --- a/tools/testing/selftests/bpf/progs/test_tc_tunnel.c +++ b/tools/testing/selftests/bpf/progs/test_tc_tunnel.c @@ -157,7 +157,7 @@ static __always_inline int encap_ipv4(struct __sk_buff *skb, __u8 encap_proto, bpf_ntohs(h_outer.ip.tot_len)); h_outer.ip.protocol = encap_proto; - set_ipv4_csum(&h_outer.ip); + set_ipv4_csum((void *)&h_outer.ip); /* store new outer network header */ if (bpf_skb_store_bytes(skb, ETH_HLEN, &h_outer, olen, From bfb35c27c65fce60a12e188135ae1344d1b89e24 Mon Sep 17 00:00:00 2001 From: Alan Maguire Date: Fri, 12 Apr 2019 12:27:34 +0100 Subject: [PATCH 31/68] bpf: fix whitespace for ENCAP_L2 defines in bpf.h replace tab after #define with space in line with rest of definitions Signed-off-by: Alan Maguire Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- include/uapi/linux/bpf.h | 6 +++--- tools/include/uapi/linux/bpf.h | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index c26be24fd5e2..704bb69514a2 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2792,14 +2792,14 @@ enum bpf_func_id { /* BPF_FUNC_skb_adjust_room flags. */ #define BPF_F_ADJ_ROOM_FIXED_GSO (1ULL << 0) -#define BPF_ADJ_ROOM_ENCAP_L2_MASK 0xff -#define BPF_ADJ_ROOM_ENCAP_L2_SHIFT 56 +#define BPF_ADJ_ROOM_ENCAP_L2_MASK 0xff +#define BPF_ADJ_ROOM_ENCAP_L2_SHIFT 56 #define BPF_F_ADJ_ROOM_ENCAP_L3_IPV4 (1ULL << 1) #define BPF_F_ADJ_ROOM_ENCAP_L3_IPV6 (1ULL << 2) #define BPF_F_ADJ_ROOM_ENCAP_L4_GRE (1ULL << 3) #define BPF_F_ADJ_ROOM_ENCAP_L4_UDP (1ULL << 4) -#define BPF_F_ADJ_ROOM_ENCAP_L2(len) (((__u64)len & \ +#define BPF_F_ADJ_ROOM_ENCAP_L2(len) (((__u64)len & \ BPF_ADJ_ROOM_ENCAP_L2_MASK) \ << BPF_ADJ_ROOM_ENCAP_L2_SHIFT) diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index c26be24fd5e2..704bb69514a2 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -2792,14 +2792,14 @@ enum bpf_func_id { /* BPF_FUNC_skb_adjust_room flags. */ #define BPF_F_ADJ_ROOM_FIXED_GSO (1ULL << 0) -#define BPF_ADJ_ROOM_ENCAP_L2_MASK 0xff -#define BPF_ADJ_ROOM_ENCAP_L2_SHIFT 56 +#define BPF_ADJ_ROOM_ENCAP_L2_MASK 0xff +#define BPF_ADJ_ROOM_ENCAP_L2_SHIFT 56 #define BPF_F_ADJ_ROOM_ENCAP_L3_IPV4 (1ULL << 1) #define BPF_F_ADJ_ROOM_ENCAP_L3_IPV6 (1ULL << 2) #define BPF_F_ADJ_ROOM_ENCAP_L4_GRE (1ULL << 3) #define BPF_F_ADJ_ROOM_ENCAP_L4_UDP (1ULL << 4) -#define BPF_F_ADJ_ROOM_ENCAP_L2(len) (((__u64)len & \ +#define BPF_F_ADJ_ROOM_ENCAP_L2(len) (((__u64)len & \ BPF_ADJ_ROOM_ENCAP_L2_MASK) \ << BPF_ADJ_ROOM_ENCAP_L2_SHIFT) From 43537b8e2dc515e037e855504db3f6c7cf73c79f Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Fri, 12 Apr 2019 09:30:48 -0400 Subject: [PATCH 32/68] bpf: reserve flags in bpf_skb_net_shrink The ENCAP flags in bpf_skb_adjust_room are ignored on decap with bpf_skb_net_shrink. Reserve these bits for future use. Fixes: 868d523535c2d ("bpf: add bpf_skb_adjust_room encap flags") Signed-off-by: Willem de Bruijn Reviewed-by: Alan Maguire Signed-off-by: Daniel Borkmann --- net/core/filter.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 95a27fdf9a40..bd1f51907b83 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3069,6 +3069,9 @@ static int bpf_skb_net_shrink(struct sk_buff *skb, u32 off, u32 len_diff, { int ret; + if (flags & ~BPF_F_ADJ_ROOM_FIXED_GSO) + return -EINVAL; + if (skb_is_gso(skb) && !skb_is_gso_tcp(skb)) { /* udp gso_size delineates datagrams, only allow if fixed */ if (!(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4) || From 031ebc1aac3dc26c5ef9f5236ad35366bd57de61 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 12 Apr 2019 14:29:34 +0100 Subject: [PATCH 33/68] tools: bpftool: remove blank line after btf_id when listing programs Commit 569b0c77735d ("tools/bpftool: show btf id in program information") made bpftool print an empty line after each program entry when listing the BPF programs loaded on the system (plain output). This is especially confusing when some programs have an associated BTF id, and others don't. Let's remove the blank line. Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/prog.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index 81067803189e..4b7a307b9fc9 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -323,7 +323,7 @@ static void print_prog_plain(struct bpf_prog_info *info, int fd) } if (info->btf_id) - printf("\n\tbtf_id %d\n", info->btf_id); + printf("\n\tbtf_id %d", info->btf_id); printf("\n"); } From 39c9f10639a3e8ead677492a7ab41cee5fa481a7 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 12 Apr 2019 14:29:35 +0100 Subject: [PATCH 34/68] tools: bpftool: reset errno for "bpftool cgroup tree" When trying to dump the tree of all cgroups under a given root node, bpftool attempts to query programs of all available attach types. Some of those attach types do not support queries, therefore several of the calls are actually expected to fail. Those calls set errno to EINVAL, which has no consequence for dumping the rest of the tree. It does have consequences however if errno is inspected at a later time. For example, bpftool batch mode relies on errno to determine whether a command has succeeded, and whether it should carry on with the next command. Setting errno to EINVAL when everything worked as expected would therefore make such command fail: # echo 'cgroup tree \n net show' | \ bpftool batch file - To improve this, reset errno when its value is EINVAL after attempting to show programs for all existing attach types in do_show_tree_fn(). Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/cgroup.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tools/bpf/bpftool/cgroup.c b/tools/bpf/bpftool/cgroup.c index 4b5c8da2a7c0..a81b34343eb8 100644 --- a/tools/bpf/bpftool/cgroup.c +++ b/tools/bpf/bpftool/cgroup.c @@ -248,6 +248,13 @@ static int do_show_tree_fn(const char *fpath, const struct stat *sb, for (type = 0; type < __MAX_BPF_ATTACH_TYPE; type++) show_attached_bpf_progs(cgroup_fd, type, ftw->level); + if (errno == EINVAL) + /* Last attach type does not support query. + * Do not report an error for this, especially because batch + * mode would stop processing commands. + */ + errno = 0; + if (json_output) { jsonw_end_array(json_wtr); jsonw_end_object(json_wtr); From 9a487883bd6bffbfb99e5f58bbfab290a42cd1a6 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 12 Apr 2019 14:29:36 +0100 Subject: [PATCH 35/68] tools: bpftool: fix man page documentation for "pinmaps" keyword The "pinmaps" keyword is present in the man page, in the verbose description of the "bpftool prog load" command. However, it is missing from the summary of available commands at the beginning of the file. Add it there as well. Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/Documentation/bpftool-prog.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index 9386bd6e0396..49c33939239f 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -25,7 +25,7 @@ PROG COMMANDS | **bpftool** **prog dump xlated** *PROG* [{**file** *FILE* | **opcodes** | **visual** | **linum**}] | **bpftool** **prog dump jited** *PROG* [{**file** *FILE* | **opcodes** | **linum**}] | **bpftool** **prog pin** *PROG* *FILE* -| **bpftool** **prog { load | loadall }** *OBJ* *PATH* [**type** *TYPE*] [**map** {**idx** *IDX* | **name** *NAME*} *MAP*] [**dev** *NAME*] +| **bpftool** **prog { load | loadall }** *OBJ* *PATH* [**type** *TYPE*] [**map** {**idx** *IDX* | **name** *NAME*} *MAP*] [**dev** *NAME*] [**pinmaps** *MAP_DIR*] | **bpftool** **prog attach** *PROG* *ATTACH_TYPE* [*MAP*] | **bpftool** **prog detach** *PROG* *ATTACH_TYPE* [*MAP*] | **bpftool** **prog tracelog** From 88b3eed805e9edf2646d7ba3e701b616225572e1 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 12 Apr 2019 14:29:37 +0100 Subject: [PATCH 36/68] tools: bpftool: fix short option name for printing version in man pages Manual pages would tell that option "-v" (lower case) would print the version number for bpftool. This is wrong: the short name of the option is "-V" (upper case). Fix the documentation accordingly. Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/Documentation/bpftool-cgroup.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-feature.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-map.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-net.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-perf.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool-prog.rst | 2 +- tools/bpf/bpftool/Documentation/bpftool.rst | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index 9bb9ace54ba8..5e3b7d9d7599 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -99,7 +99,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool-feature.rst b/tools/bpf/bpftool/Documentation/bpftool-feature.rst index 82de03dd8f52..10177343b7ce 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-feature.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-feature.rst @@ -63,7 +63,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool-map.rst b/tools/bpf/bpftool/Documentation/bpftool-map.rst index 5c984ffc9f01..55ecf80ca03e 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-map.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-map.rst @@ -135,7 +135,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool-net.rst b/tools/bpf/bpftool/Documentation/bpftool-net.rst index 779dab3650ee..6b692b428373 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-net.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-net.rst @@ -55,7 +55,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool-perf.rst b/tools/bpf/bpftool/Documentation/bpftool-perf.rst index bca5590a80d0..86154740fabb 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-perf.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-perf.rst @@ -43,7 +43,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index 49c33939239f..f2beb79a77e7 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -144,7 +144,7 @@ OPTIONS -h, --help Print short generic help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json diff --git a/tools/bpf/bpftool/Documentation/bpftool.rst b/tools/bpf/bpftool/Documentation/bpftool.rst index 4f2188845dd8..deb2ca911ddf 100644 --- a/tools/bpf/bpftool/Documentation/bpftool.rst +++ b/tools/bpf/bpftool/Documentation/bpftool.rst @@ -49,7 +49,7 @@ OPTIONS -h, --help Print short help message (similar to **bpftool help**). - -v, --version + -V, --version Print version number (similar to **bpftool version**). -j, --json From 25df480def17eb792860d86b8f90fda00035f0f9 Mon Sep 17 00:00:00 2001 From: Quentin Monnet Date: Fri, 12 Apr 2019 14:29:38 +0100 Subject: [PATCH 37/68] tools: bpftool: add a note on program statistics in man page Linux kernel now supports statistics for BPF programs, and bpftool is able to dump them. However, these statistics are not enabled by default, and administrators may not know how to access them. Add a paragraph in bpftool documentation, under the description of the "bpftool prog show" command, to explain that such statistics are available and that their collection is controlled via a dedicated sysctl knob. Signed-off-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/Documentation/bpftool-prog.rst | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index f2beb79a77e7..bb9bb00c0c2c 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -56,6 +56,14 @@ DESCRIPTION Output will start with program ID followed by program type and zero or more named attributes (depending on kernel version). + Since Linux 5.1 the kernel can collect statistics on BPF + programs (such as the total time spent running the program, + and the number of times it was run). If available, bpftool + shows such statistics. However, the kernel does not collect + them by defaults, as it slightly impacts performance on each + program run. Activation or deactivation of the feature is + performed via the **kernel.bpf_stats_enabled** sysctl knob. + **bpftool prog dump xlated** *PROG* [{ **file** *FILE* | **opcodes** | **visual** | **linum** }] Dump eBPF instructions of the program from the kernel. By default, eBPF will be disassembled and printed to standard From 0478c3bf812451315da6eeb79f8cf151db094767 Mon Sep 17 00:00:00 2001 From: Benjamin Poirier Date: Mon, 15 Apr 2019 16:15:35 +0900 Subject: [PATCH 38/68] bpftool: Use print_entry_error() in case of ENOENT when dumping Commit bf598a8f0f77 ("bpftool: Improve handling of ENOENT on map dumps") used print_entry_plain() in case of ENOENT. However, that commit introduces dead code. Per-cpu maps are zero-filled. When reading them, it's all or nothing. There will never be a case where some cpus have an entry and others don't. The truth is that ENOENT is an error case. Use print_entry_error() to output the desired message. That function's "value" parameter is also renamed to indicate that we never use it for an actual map value. The output format is unchanged. Signed-off-by: Benjamin Poirier Reviewed-by: Quentin Monnet Reviewed-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/map.c | 33 ++++++++++++++------------------- 1 file changed, 14 insertions(+), 19 deletions(-) diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index e96903078991..df958af56b6c 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -261,20 +261,20 @@ static void print_entry_json(struct bpf_map_info *info, unsigned char *key, } static void print_entry_error(struct bpf_map_info *info, unsigned char *key, - const char *value) + const char *error_msg) { - int value_size = strlen(value); + int msg_size = strlen(error_msg); bool single_line, break_names; - break_names = info->key_size > 16 || value_size > 16; - single_line = info->key_size + value_size <= 24 && !break_names; + break_names = info->key_size > 16 || msg_size > 16; + single_line = info->key_size + msg_size <= 24 && !break_names; printf("key:%c", break_names ? '\n' : ' '); fprint_hex(stdout, key, info->key_size, " "); printf(single_line ? " " : "\n"); - printf("value:%c%s", break_names ? '\n' : ' ', value); + printf("value:%c%s", break_names ? '\n' : ' ', error_msg); printf("\n"); } @@ -298,11 +298,7 @@ static void print_entry_plain(struct bpf_map_info *info, unsigned char *key, if (info->value_size) { printf("value:%c", break_names ? '\n' : ' '); - if (value) - fprint_hex(stdout, value, info->value_size, - " "); - else - printf(""); + fprint_hex(stdout, value, info->value_size, " "); } printf("\n"); @@ -321,11 +317,8 @@ static void print_entry_plain(struct bpf_map_info *info, unsigned char *key, for (i = 0; i < n; i++) { printf("value (CPU %02d):%c", i, info->value_size > 16 ? '\n' : ' '); - if (value) - fprint_hex(stdout, value + i * step, - info->value_size, " "); - else - printf(""); + fprint_hex(stdout, value + i * step, + info->value_size, " "); printf("\n"); } } @@ -722,11 +715,13 @@ static int dump_map_elem(int fd, void *key, void *value, jsonw_string_field(json_wtr, "error", strerror(lookup_errno)); jsonw_end_object(json_wtr); } else { + const char *msg = NULL; + if (errno == ENOENT) - print_entry_plain(map_info, key, NULL); - else - print_entry_error(map_info, key, - strerror(lookup_errno)); + msg = ""; + + print_entry_error(map_info, key, + msg ? : strerror(lookup_errno)); } return 0; From 3da6e7e408b9c229e5e5cc1ddab7445c7561afc3 Mon Sep 17 00:00:00 2001 From: Benjamin Poirier Date: Mon, 15 Apr 2019 16:15:36 +0900 Subject: [PATCH 39/68] bpftool: Improve handling of ENOSPC on reuseport_array map dumps avoids outputting a series of value: No space left on device The value itself is not wrong but bpf_fd_reuseport_array_lookup_elem() can only return it if the map was created with value_size = 8. There's nothing bpftool can do about it. Instead of repeating this error for every key in the map, print an explanatory warning and a specialized error. example before: key: 00 00 00 00 value: No space left on device key: 01 00 00 00 value: No space left on device key: 02 00 00 00 value: No space left on device Found 0 elements example after: Warning: cannot read values from reuseport_sockarray map with value_size != 8 key: 00 00 00 00 value: key: 01 00 00 00 value: key: 02 00 00 00 value: Found 0 elements Signed-off-by: Benjamin Poirier Reviewed-by: Quentin Monnet Reviewed-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- tools/bpf/bpftool/map.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index df958af56b6c..44b192e87708 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -719,6 +719,9 @@ static int dump_map_elem(int fd, void *key, void *value, if (errno == ENOENT) msg = ""; + else if (lookup_errno == ENOSPC && + map_info->type == BPF_MAP_TYPE_REUSEPORT_SOCKARRAY) + msg = ""; print_entry_error(map_info, key, msg ? : strerror(lookup_errno)); @@ -775,6 +778,10 @@ static int do_dump(int argc, char **argv) } } + if (info.type == BPF_MAP_TYPE_REUSEPORT_SOCKARRAY && + info.value_size != 8) + p_info("Warning: cannot read values from %s map with value_size != 8", + map_type_name[info.type]); while (true) { err = bpf_map_get_next_key(fd, prev_key, key); if (err) { From 08de198c95438fcbfad7bc06121176794ec92c6e Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Fri, 12 Apr 2019 14:41:32 -0700 Subject: [PATCH 40/68] selftests/bpf: two scale tests Add two tests to check that sequence of 1024 jumps is verifiable. Signed-off-by: Alexei Starovoitov Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- tools/testing/selftests/bpf/test_verifier.c | 70 ++++++++++++++++++++ tools/testing/selftests/bpf/verifier/scale.c | 18 +++++ 2 files changed, 88 insertions(+) create mode 100644 tools/testing/selftests/bpf/verifier/scale.c diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index e2ebcaddbe78..6cb6a1074fd1 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -208,6 +208,76 @@ static void bpf_fill_rand_ld_dw(struct bpf_test *self) self->retval = (uint32_t)res; } +/* test the sequence of 1k jumps */ +static void bpf_fill_scale1(struct bpf_test *self) +{ + struct bpf_insn *insn = self->fill_insns; + int i = 0, k = 0; + + insn[i++] = BPF_MOV64_REG(BPF_REG_6, BPF_REG_1); + /* test to check that the sequence of 1024 jumps is acceptable */ + while (k++ < 1024) { + insn[i++] = BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, + BPF_FUNC_get_prandom_u32); + insn[i++] = BPF_JMP_IMM(BPF_JGT, BPF_REG_0, bpf_semi_rand_get(), 2); + insn[i++] = BPF_MOV64_REG(BPF_REG_1, BPF_REG_10); + insn[i++] = BPF_STX_MEM(BPF_DW, BPF_REG_1, BPF_REG_6, + -8 * (k % 64 + 1)); + } + /* every jump adds 1024 steps to insn_processed, so to stay exactly + * within 1m limit add MAX_TEST_INSNS - 1025 MOVs and 1 EXIT + */ + while (i < MAX_TEST_INSNS - 1025) + insn[i++] = BPF_ALU32_IMM(BPF_MOV, BPF_REG_0, 42); + insn[i] = BPF_EXIT_INSN(); + self->prog_len = i + 1; + self->retval = 42; +} + +/* test the sequence of 1k jumps in inner most function (function depth 8)*/ +static void bpf_fill_scale2(struct bpf_test *self) +{ + struct bpf_insn *insn = self->fill_insns; + int i = 0, k = 0; + +#define FUNC_NEST 7 + for (k = 0; k < FUNC_NEST; k++) { + insn[i++] = BPF_CALL_REL(1); + insn[i++] = BPF_EXIT_INSN(); + } + insn[i++] = BPF_MOV64_REG(BPF_REG_6, BPF_REG_1); + /* test to check that the sequence of 1024 jumps is acceptable */ + while (k++ < 1024) { + insn[i++] = BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, + BPF_FUNC_get_prandom_u32); + insn[i++] = BPF_JMP_IMM(BPF_JGT, BPF_REG_0, bpf_semi_rand_get(), 2); + insn[i++] = BPF_MOV64_REG(BPF_REG_1, BPF_REG_10); + insn[i++] = BPF_STX_MEM(BPF_DW, BPF_REG_1, BPF_REG_6, + -8 * (k % (64 - 4 * FUNC_NEST) + 1)); + } + /* every jump adds 1024 steps to insn_processed, so to stay exactly + * within 1m limit add MAX_TEST_INSNS - 1025 MOVs and 1 EXIT + */ + while (i < MAX_TEST_INSNS - 1025) + insn[i++] = BPF_ALU32_IMM(BPF_MOV, BPF_REG_0, 42); + insn[i] = BPF_EXIT_INSN(); + self->prog_len = i + 1; + self->retval = 42; +} + +static void bpf_fill_scale(struct bpf_test *self) +{ + switch (self->retval) { + case 1: + return bpf_fill_scale1(self); + case 2: + return bpf_fill_scale2(self); + default: + self->prog_len = 0; + break; + } +} + /* BPF_SK_LOOKUP contains 13 instructions, if you need to fix up maps */ #define BPF_SK_LOOKUP(func) \ /* struct bpf_sock_tuple tuple = {} */ \ diff --git a/tools/testing/selftests/bpf/verifier/scale.c b/tools/testing/selftests/bpf/verifier/scale.c new file mode 100644 index 000000000000..7f868d4802e0 --- /dev/null +++ b/tools/testing/selftests/bpf/verifier/scale.c @@ -0,0 +1,18 @@ +{ + "scale: scale test 1", + .insns = { }, + .data = { }, + .fill_helper = bpf_fill_scale, + .prog_type = BPF_PROG_TYPE_SCHED_CLS, + .result = ACCEPT, + .retval = 1, +}, +{ + "scale: scale test 2", + .insns = { }, + .data = { }, + .fill_helper = bpf_fill_scale, + .prog_type = BPF_PROG_TYPE_SCHED_CLS, + .result = ACCEPT, + .retval = 2, +}, From a5cb33464e53482ecc645b4b7703f5b6d151095f Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Fri, 12 Apr 2019 16:43:10 -0700 Subject: [PATCH 41/68] selftests/bpf: make flow dissector tests more extensible Rewrite selftest to iterate over an array with input packet and expected flow_keys. This should make it easier to extend this test with additional cases without too much boilerplate. Signed-off-by: Stanislav Fomichev Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- .../selftests/bpf/prog_tests/flow_dissector.c | 197 +++++++++++------- 1 file changed, 116 insertions(+), 81 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c index fc818bc1d729..1a73937826b0 100644 --- a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c +++ b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c @@ -2,7 +2,7 @@ #include #define CHECK_FLOW_KEYS(desc, got, expected) \ - CHECK(memcmp(&got, &expected, sizeof(got)) != 0, \ + CHECK_ATTR(memcmp(&got, &expected, sizeof(got)) != 0, \ desc, \ "nhoff=%u/%u " \ "thoff=%u/%u " \ @@ -10,6 +10,7 @@ "is_frag=%u/%u " \ "is_first_frag=%u/%u " \ "is_encap=%u/%u " \ + "ip_proto=0x%x/0x%x " \ "n_proto=0x%x/0x%x " \ "sport=%u/%u " \ "dport=%u/%u\n", \ @@ -19,53 +20,32 @@ got.is_frag, expected.is_frag, \ got.is_first_frag, expected.is_first_frag, \ got.is_encap, expected.is_encap, \ + got.ip_proto, expected.ip_proto, \ got.n_proto, expected.n_proto, \ got.sport, expected.sport, \ got.dport, expected.dport) -static struct bpf_flow_keys pkt_v4_flow_keys = { - .nhoff = 0, - .thoff = sizeof(struct iphdr), - .addr_proto = ETH_P_IP, - .ip_proto = IPPROTO_TCP, - .n_proto = __bpf_constant_htons(ETH_P_IP), -}; +struct ipv4_pkt { + struct ethhdr eth; + struct iphdr iph; + struct tcphdr tcp; +} __packed; -static struct bpf_flow_keys pkt_v6_flow_keys = { - .nhoff = 0, - .thoff = sizeof(struct ipv6hdr), - .addr_proto = ETH_P_IPV6, - .ip_proto = IPPROTO_TCP, - .n_proto = __bpf_constant_htons(ETH_P_IPV6), -}; - -#define VLAN_HLEN 4 - -static struct { +struct svlan_ipv4_pkt { struct ethhdr eth; __u16 vlan_tci; __u16 vlan_proto; struct iphdr iph; struct tcphdr tcp; -} __packed pkt_vlan_v4 = { - .eth.h_proto = __bpf_constant_htons(ETH_P_8021Q), - .vlan_proto = __bpf_constant_htons(ETH_P_IP), - .iph.ihl = 5, - .iph.protocol = IPPROTO_TCP, - .iph.tot_len = __bpf_constant_htons(MAGIC_BYTES), - .tcp.urg_ptr = 123, - .tcp.doff = 5, -}; +} __packed; -static struct bpf_flow_keys pkt_vlan_v4_flow_keys = { - .nhoff = VLAN_HLEN, - .thoff = VLAN_HLEN + sizeof(struct iphdr), - .addr_proto = ETH_P_IP, - .ip_proto = IPPROTO_TCP, - .n_proto = __bpf_constant_htons(ETH_P_IP), -}; +struct ipv6_pkt { + struct ethhdr eth; + struct ipv6hdr iph; + struct tcphdr tcp; +} __packed; -static struct { +struct dvlan_ipv6_pkt { struct ethhdr eth; __u16 vlan_tci; __u16 vlan_proto; @@ -73,31 +53,97 @@ static struct { __u16 vlan_proto2; struct ipv6hdr iph; struct tcphdr tcp; -} __packed pkt_vlan_v6 = { - .eth.h_proto = __bpf_constant_htons(ETH_P_8021AD), - .vlan_proto = __bpf_constant_htons(ETH_P_8021Q), - .vlan_proto2 = __bpf_constant_htons(ETH_P_IPV6), - .iph.nexthdr = IPPROTO_TCP, - .iph.payload_len = __bpf_constant_htons(MAGIC_BYTES), - .tcp.urg_ptr = 123, - .tcp.doff = 5, +} __packed; + +struct test { + const char *name; + union { + struct ipv4_pkt ipv4; + struct svlan_ipv4_pkt svlan_ipv4; + struct ipv6_pkt ipv6; + struct dvlan_ipv6_pkt dvlan_ipv6; + } pkt; + struct bpf_flow_keys keys; }; -static struct bpf_flow_keys pkt_vlan_v6_flow_keys = { - .nhoff = VLAN_HLEN * 2, - .thoff = VLAN_HLEN * 2 + sizeof(struct ipv6hdr), - .addr_proto = ETH_P_IPV6, - .ip_proto = IPPROTO_TCP, - .n_proto = __bpf_constant_htons(ETH_P_IPV6), +#define VLAN_HLEN 4 + +struct test tests[] = { + { + .name = "ipv4", + .pkt.ipv4 = { + .eth.h_proto = __bpf_constant_htons(ETH_P_IP), + .iph.ihl = 5, + .iph.protocol = IPPROTO_TCP, + .iph.tot_len = __bpf_constant_htons(MAGIC_BYTES), + .tcp.doff = 5, + }, + .keys = { + .nhoff = 0, + .thoff = sizeof(struct iphdr), + .addr_proto = ETH_P_IP, + .ip_proto = IPPROTO_TCP, + .n_proto = __bpf_constant_htons(ETH_P_IP), + }, + }, + { + .name = "ipv6", + .pkt.ipv6 = { + .eth.h_proto = __bpf_constant_htons(ETH_P_IPV6), + .iph.nexthdr = IPPROTO_TCP, + .iph.payload_len = __bpf_constant_htons(MAGIC_BYTES), + .tcp.doff = 5, + }, + .keys = { + .nhoff = 0, + .thoff = sizeof(struct ipv6hdr), + .addr_proto = ETH_P_IPV6, + .ip_proto = IPPROTO_TCP, + .n_proto = __bpf_constant_htons(ETH_P_IPV6), + }, + }, + { + .name = "802.1q-ipv4", + .pkt.svlan_ipv4 = { + .eth.h_proto = __bpf_constant_htons(ETH_P_8021Q), + .vlan_proto = __bpf_constant_htons(ETH_P_IP), + .iph.ihl = 5, + .iph.protocol = IPPROTO_TCP, + .iph.tot_len = __bpf_constant_htons(MAGIC_BYTES), + .tcp.doff = 5, + }, + .keys = { + .nhoff = VLAN_HLEN, + .thoff = VLAN_HLEN + sizeof(struct iphdr), + .addr_proto = ETH_P_IP, + .ip_proto = IPPROTO_TCP, + .n_proto = __bpf_constant_htons(ETH_P_IP), + }, + }, + { + .name = "802.1ad-ipv6", + .pkt.dvlan_ipv6 = { + .eth.h_proto = __bpf_constant_htons(ETH_P_8021AD), + .vlan_proto = __bpf_constant_htons(ETH_P_8021Q), + .vlan_proto2 = __bpf_constant_htons(ETH_P_IPV6), + .iph.nexthdr = IPPROTO_TCP, + .iph.payload_len = __bpf_constant_htons(MAGIC_BYTES), + .tcp.doff = 5, + }, + .keys = { + .nhoff = VLAN_HLEN * 2, + .thoff = VLAN_HLEN * 2 + sizeof(struct ipv6hdr), + .addr_proto = ETH_P_IPV6, + .ip_proto = IPPROTO_TCP, + .n_proto = __bpf_constant_htons(ETH_P_IPV6), + }, + }, }; void test_flow_dissector(void) { - struct bpf_flow_keys flow_keys; struct bpf_object *obj; - __u32 duration, retval; int err, prog_fd; - __u32 size; err = bpf_flow_load(&obj, "./bpf_flow.o", "flow_dissector", "jmp_table", &prog_fd); @@ -106,35 +152,24 @@ void test_flow_dissector(void) return; } - err = bpf_prog_test_run(prog_fd, 10, &pkt_v4, sizeof(pkt_v4), - &flow_keys, &size, &retval, &duration); - CHECK(size != sizeof(flow_keys) || err || retval != 1, "ipv4", - "err %d errno %d retval %d duration %d size %u/%lu\n", - err, errno, retval, duration, size, sizeof(flow_keys)); - CHECK_FLOW_KEYS("ipv4_flow_keys", flow_keys, pkt_v4_flow_keys); + for (int i = 0; i < ARRAY_SIZE(tests); i++) { + struct bpf_flow_keys flow_keys; + struct bpf_prog_test_run_attr tattr = { + .prog_fd = prog_fd, + .data_in = &tests[i].pkt, + .data_size_in = sizeof(tests[i].pkt), + .data_out = &flow_keys, + }; - err = bpf_prog_test_run(prog_fd, 10, &pkt_v6, sizeof(pkt_v6), - &flow_keys, &size, &retval, &duration); - CHECK(size != sizeof(flow_keys) || err || retval != 1, "ipv6", - "err %d errno %d retval %d duration %d size %u/%lu\n", - err, errno, retval, duration, size, sizeof(flow_keys)); - CHECK_FLOW_KEYS("ipv6_flow_keys", flow_keys, pkt_v6_flow_keys); - - err = bpf_prog_test_run(prog_fd, 10, &pkt_vlan_v4, sizeof(pkt_vlan_v4), - &flow_keys, &size, &retval, &duration); - CHECK(size != sizeof(flow_keys) || err || retval != 1, "vlan_ipv4", - "err %d errno %d retval %d duration %d size %u/%lu\n", - err, errno, retval, duration, size, sizeof(flow_keys)); - CHECK_FLOW_KEYS("vlan_ipv4_flow_keys", flow_keys, - pkt_vlan_v4_flow_keys); - - err = bpf_prog_test_run(prog_fd, 10, &pkt_vlan_v6, sizeof(pkt_vlan_v6), - &flow_keys, &size, &retval, &duration); - CHECK(size != sizeof(flow_keys) || err || retval != 1, "vlan_ipv6", - "err %d errno %d retval %d duration %d size %u/%lu\n", - err, errno, retval, duration, size, sizeof(flow_keys)); - CHECK_FLOW_KEYS("vlan_ipv6_flow_keys", flow_keys, - pkt_vlan_v6_flow_keys); + err = bpf_prog_test_run_xattr(&tattr); + CHECK_ATTR(tattr.data_size_out != sizeof(flow_keys) || + err || tattr.retval != 1, + tests[i].name, + "err %d errno %d retval %d duration %d size %u/%lu\n", + err, errno, tattr.retval, tattr.duration, + tattr.data_size_out, sizeof(flow_keys)); + CHECK_FLOW_KEYS(tests[i].name, flow_keys, tests[i].keys); + } bpf_object__close(obj); } From 02a8c817a31606b6b37c2b755f6569903f44241e Mon Sep 17 00:00:00 2001 From: Alban Crequy Date: Sun, 14 Apr 2019 18:58:46 +0200 Subject: [PATCH 42/68] bpf: add map helper functions push, pop, peek in more BPF programs commit f1a2e44a3aec ("bpf: add queue and stack maps") introduced new BPF helper functions: - BPF_FUNC_map_push_elem - BPF_FUNC_map_pop_elem - BPF_FUNC_map_peek_elem but they were made available only for network BPF programs. This patch makes them available for tracepoint, cgroup and lirc programs. Signed-off-by: Alban Crequy Cc: Mauricio Vasquez B Acked-by: Song Liu Signed-off-by: Daniel Borkmann --- drivers/media/rc/bpf-lirc.c | 6 ++++++ kernel/bpf/cgroup.c | 6 ++++++ kernel/trace/bpf_trace.c | 6 ++++++ 3 files changed, 18 insertions(+) diff --git a/drivers/media/rc/bpf-lirc.c b/drivers/media/rc/bpf-lirc.c index 390a722e6211..ee657003c1a1 100644 --- a/drivers/media/rc/bpf-lirc.c +++ b/drivers/media/rc/bpf-lirc.c @@ -97,6 +97,12 @@ lirc_mode2_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_map_update_elem_proto; case BPF_FUNC_map_delete_elem: return &bpf_map_delete_elem_proto; + case BPF_FUNC_map_push_elem: + return &bpf_map_push_elem_proto; + case BPF_FUNC_map_pop_elem: + return &bpf_map_pop_elem_proto; + case BPF_FUNC_map_peek_elem: + return &bpf_map_peek_elem_proto; case BPF_FUNC_ktime_get_ns: return &bpf_ktime_get_ns_proto; case BPF_FUNC_tail_call: diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index e58a6c247f56..fcde0f7b2585 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -713,6 +713,12 @@ cgroup_base_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_map_update_elem_proto; case BPF_FUNC_map_delete_elem: return &bpf_map_delete_elem_proto; + case BPF_FUNC_map_push_elem: + return &bpf_map_push_elem_proto; + case BPF_FUNC_map_pop_elem: + return &bpf_map_pop_elem_proto; + case BPF_FUNC_map_peek_elem: + return &bpf_map_peek_elem_proto; case BPF_FUNC_get_current_uid_gid: return &bpf_get_current_uid_gid_proto; case BPF_FUNC_get_local_storage: diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index d64c00afceb5..91800be0c8eb 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -569,6 +569,12 @@ tracing_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_map_update_elem_proto; case BPF_FUNC_map_delete_elem: return &bpf_map_delete_elem_proto; + case BPF_FUNC_map_push_elem: + return &bpf_map_push_elem_proto; + case BPF_FUNC_map_pop_elem: + return &bpf_map_pop_elem_proto; + case BPF_FUNC_map_peek_elem: + return &bpf_map_peek_elem_proto; case BPF_FUNC_probe_read: return &bpf_probe_read_proto; case BPF_FUNC_ktime_get_ns: From 809041e765055ba311f3a0b8751db1c65739ad51 Mon Sep 17 00:00:00 2001 From: Peter Oskolkov Date: Wed, 3 Apr 2019 08:43:38 -0700 Subject: [PATCH 43/68] selftests: bpf: add VRF test cases to lwt_ip_encap test. This patch adds tests validating that VRF and BPF-LWT encap work together well, as requested by David Ahern. Signed-off-by: Peter Oskolkov Acked-by: Martin KaFai Lau Signed-off-by: Alexei Starovoitov --- .../selftests/bpf/test_lwt_ip_encap.sh | 134 +++++++++++------- 1 file changed, 86 insertions(+), 48 deletions(-) diff --git a/tools/testing/selftests/bpf/test_lwt_ip_encap.sh b/tools/testing/selftests/bpf/test_lwt_ip_encap.sh index d4d3391cc13a..acf7a74f97cd 100755 --- a/tools/testing/selftests/bpf/test_lwt_ip_encap.sh +++ b/tools/testing/selftests/bpf/test_lwt_ip_encap.sh @@ -129,6 +129,24 @@ setup() ip link set veth7 netns ${NS2} ip link set veth8 netns ${NS3} + if [ ! -z "${VRF}" ] ; then + ip -netns ${NS1} link add red type vrf table 1001 + ip -netns ${NS1} link set red up + ip -netns ${NS1} route add table 1001 unreachable default metric 8192 + ip -netns ${NS1} -6 route add table 1001 unreachable default metric 8192 + ip -netns ${NS1} link set veth1 vrf red + ip -netns ${NS1} link set veth5 vrf red + + ip -netns ${NS2} link add red type vrf table 1001 + ip -netns ${NS2} link set red up + ip -netns ${NS2} route add table 1001 unreachable default metric 8192 + ip -netns ${NS2} -6 route add table 1001 unreachable default metric 8192 + ip -netns ${NS2} link set veth2 vrf red + ip -netns ${NS2} link set veth3 vrf red + ip -netns ${NS2} link set veth6 vrf red + ip -netns ${NS2} link set veth7 vrf red + fi + # configure addesses: the top route (1-2-3-4) ip -netns ${NS1} addr add ${IPv4_1}/24 dev veth1 ip -netns ${NS2} addr add ${IPv4_2}/24 dev veth2 @@ -163,29 +181,29 @@ setup() # NS1 # top route - ip -netns ${NS1} route add ${IPv4_2}/32 dev veth1 - ip -netns ${NS1} route add default dev veth1 via ${IPv4_2} # go top by default - ip -netns ${NS1} -6 route add ${IPv6_2}/128 dev veth1 - ip -netns ${NS1} -6 route add default dev veth1 via ${IPv6_2} # go top by default + ip -netns ${NS1} route add ${IPv4_2}/32 dev veth1 ${VRF} + ip -netns ${NS1} route add default dev veth1 via ${IPv4_2} ${VRF} # go top by default + ip -netns ${NS1} -6 route add ${IPv6_2}/128 dev veth1 ${VRF} + ip -netns ${NS1} -6 route add default dev veth1 via ${IPv6_2} ${VRF} # go top by default # bottom route - ip -netns ${NS1} route add ${IPv4_6}/32 dev veth5 - ip -netns ${NS1} route add ${IPv4_7}/32 dev veth5 via ${IPv4_6} - ip -netns ${NS1} route add ${IPv4_8}/32 dev veth5 via ${IPv4_6} - ip -netns ${NS1} -6 route add ${IPv6_6}/128 dev veth5 - ip -netns ${NS1} -6 route add ${IPv6_7}/128 dev veth5 via ${IPv6_6} - ip -netns ${NS1} -6 route add ${IPv6_8}/128 dev veth5 via ${IPv6_6} + ip -netns ${NS1} route add ${IPv4_6}/32 dev veth5 ${VRF} + ip -netns ${NS1} route add ${IPv4_7}/32 dev veth5 via ${IPv4_6} ${VRF} + ip -netns ${NS1} route add ${IPv4_8}/32 dev veth5 via ${IPv4_6} ${VRF} + ip -netns ${NS1} -6 route add ${IPv6_6}/128 dev veth5 ${VRF} + ip -netns ${NS1} -6 route add ${IPv6_7}/128 dev veth5 via ${IPv6_6} ${VRF} + ip -netns ${NS1} -6 route add ${IPv6_8}/128 dev veth5 via ${IPv6_6} ${VRF} # NS2 # top route - ip -netns ${NS2} route add ${IPv4_1}/32 dev veth2 - ip -netns ${NS2} route add ${IPv4_4}/32 dev veth3 - ip -netns ${NS2} -6 route add ${IPv6_1}/128 dev veth2 - ip -netns ${NS2} -6 route add ${IPv6_4}/128 dev veth3 + ip -netns ${NS2} route add ${IPv4_1}/32 dev veth2 ${VRF} + ip -netns ${NS2} route add ${IPv4_4}/32 dev veth3 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_1}/128 dev veth2 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_4}/128 dev veth3 ${VRF} # bottom route - ip -netns ${NS2} route add ${IPv4_5}/32 dev veth6 - ip -netns ${NS2} route add ${IPv4_8}/32 dev veth7 - ip -netns ${NS2} -6 route add ${IPv6_5}/128 dev veth6 - ip -netns ${NS2} -6 route add ${IPv6_8}/128 dev veth7 + ip -netns ${NS2} route add ${IPv4_5}/32 dev veth6 ${VRF} + ip -netns ${NS2} route add ${IPv4_8}/32 dev veth7 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_5}/128 dev veth6 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_8}/128 dev veth7 ${VRF} # NS3 # top route @@ -207,16 +225,16 @@ setup() ip -netns ${NS3} tunnel add gre_dev mode gre remote ${IPv4_1} local ${IPv4_GRE} ttl 255 ip -netns ${NS3} link set gre_dev up ip -netns ${NS3} addr add ${IPv4_GRE} dev gre_dev - ip -netns ${NS1} route add ${IPv4_GRE}/32 dev veth5 via ${IPv4_6} - ip -netns ${NS2} route add ${IPv4_GRE}/32 dev veth7 via ${IPv4_8} + ip -netns ${NS1} route add ${IPv4_GRE}/32 dev veth5 via ${IPv4_6} ${VRF} + ip -netns ${NS2} route add ${IPv4_GRE}/32 dev veth7 via ${IPv4_8} ${VRF} # configure IPv6 GRE device in NS3, and a route to it via the "bottom" route ip -netns ${NS3} -6 tunnel add name gre6_dev mode ip6gre remote ${IPv6_1} local ${IPv6_GRE} ttl 255 ip -netns ${NS3} link set gre6_dev up ip -netns ${NS3} -6 addr add ${IPv6_GRE} nodad dev gre6_dev - ip -netns ${NS1} -6 route add ${IPv6_GRE}/128 dev veth5 via ${IPv6_6} - ip -netns ${NS2} -6 route add ${IPv6_GRE}/128 dev veth7 via ${IPv6_8} + ip -netns ${NS1} -6 route add ${IPv6_GRE}/128 dev veth5 via ${IPv6_6} ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_GRE}/128 dev veth7 via ${IPv6_8} ${VRF} # rp_filter gets confused by what these tests are doing, so disable it ip netns exec ${NS1} sysctl -wq net.ipv4.conf.all.rp_filter=0 @@ -244,18 +262,18 @@ trap cleanup EXIT remove_routes_to_gredev() { - ip -netns ${NS1} route del ${IPv4_GRE} dev veth5 - ip -netns ${NS2} route del ${IPv4_GRE} dev veth7 - ip -netns ${NS1} -6 route del ${IPv6_GRE}/128 dev veth5 - ip -netns ${NS2} -6 route del ${IPv6_GRE}/128 dev veth7 + ip -netns ${NS1} route del ${IPv4_GRE} dev veth5 ${VRF} + ip -netns ${NS2} route del ${IPv4_GRE} dev veth7 ${VRF} + ip -netns ${NS1} -6 route del ${IPv6_GRE}/128 dev veth5 ${VRF} + ip -netns ${NS2} -6 route del ${IPv6_GRE}/128 dev veth7 ${VRF} } add_unreachable_routes_to_gredev() { - ip -netns ${NS1} route add unreachable ${IPv4_GRE}/32 - ip -netns ${NS2} route add unreachable ${IPv4_GRE}/32 - ip -netns ${NS1} -6 route add unreachable ${IPv6_GRE}/128 - ip -netns ${NS2} -6 route add unreachable ${IPv6_GRE}/128 + ip -netns ${NS1} route add unreachable ${IPv4_GRE}/32 ${VRF} + ip -netns ${NS2} route add unreachable ${IPv4_GRE}/32 ${VRF} + ip -netns ${NS1} -6 route add unreachable ${IPv6_GRE}/128 ${VRF} + ip -netns ${NS2} -6 route add unreachable ${IPv6_GRE}/128 ${VRF} } test_ping() @@ -265,10 +283,10 @@ test_ping() local RET=0 if [ "${PROTO}" == "IPv4" ] ; then - ip netns exec ${NS1} ping -c 1 -W 1 -I ${IPv4_SRC} ${IPv4_DST} 2>&1 > /dev/null + ip netns exec ${NS1} ping -c 1 -W 1 -I veth1 ${IPv4_DST} 2>&1 > /dev/null RET=$? elif [ "${PROTO}" == "IPv6" ] ; then - ip netns exec ${NS1} ping6 -c 1 -W 6 -I ${IPv6_SRC} ${IPv6_DST} 2>&1 > /dev/null + ip netns exec ${NS1} ping6 -c 1 -W 6 -I veth1 ${IPv6_DST} 2>&1 > /dev/null RET=$? else echo " test_ping: unknown PROTO: ${PROTO}" @@ -328,7 +346,7 @@ test_gso() test_egress() { local readonly ENCAP=$1 - echo "starting egress ${ENCAP} encap test" + echo "starting egress ${ENCAP} encap test ${VRF}" setup # by default, pings work @@ -336,26 +354,35 @@ test_egress() test_ping IPv6 0 # remove NS2->DST routes, ping fails - ip -netns ${NS2} route del ${IPv4_DST}/32 dev veth3 - ip -netns ${NS2} -6 route del ${IPv6_DST}/128 dev veth3 + ip -netns ${NS2} route del ${IPv4_DST}/32 dev veth3 ${VRF} + ip -netns ${NS2} -6 route del ${IPv6_DST}/128 dev veth3 ${VRF} test_ping IPv4 1 test_ping IPv6 1 # install replacement routes (LWT/eBPF), pings succeed if [ "${ENCAP}" == "IPv4" ] ; then - ip -netns ${NS1} route add ${IPv4_DST} encap bpf xmit obj test_lwt_ip_encap.o sec encap_gre dev veth1 - ip -netns ${NS1} -6 route add ${IPv6_DST} encap bpf xmit obj test_lwt_ip_encap.o sec encap_gre dev veth1 + ip -netns ${NS1} route add ${IPv4_DST} encap bpf xmit obj \ + test_lwt_ip_encap.o sec encap_gre dev veth1 ${VRF} + ip -netns ${NS1} -6 route add ${IPv6_DST} encap bpf xmit obj \ + test_lwt_ip_encap.o sec encap_gre dev veth1 ${VRF} elif [ "${ENCAP}" == "IPv6" ] ; then - ip -netns ${NS1} route add ${IPv4_DST} encap bpf xmit obj test_lwt_ip_encap.o sec encap_gre6 dev veth1 - ip -netns ${NS1} -6 route add ${IPv6_DST} encap bpf xmit obj test_lwt_ip_encap.o sec encap_gre6 dev veth1 + ip -netns ${NS1} route add ${IPv4_DST} encap bpf xmit obj \ + test_lwt_ip_encap.o sec encap_gre6 dev veth1 ${VRF} + ip -netns ${NS1} -6 route add ${IPv6_DST} encap bpf xmit obj \ + test_lwt_ip_encap.o sec encap_gre6 dev veth1 ${VRF} else echo " unknown encap ${ENCAP}" TEST_STATUS=1 fi test_ping IPv4 0 test_ping IPv6 0 - test_gso IPv4 - test_gso IPv6 + + # skip GSO tests with VRF: VRF routing needs properly assigned + # source IP/device, which is easy to do with ping and hard with dd/nc. + if [ -z "${VRF}" ] ; then + test_gso IPv4 + test_gso IPv6 + fi # a negative test: remove routes to GRE devices: ping fails remove_routes_to_gredev @@ -374,7 +401,7 @@ test_egress() test_ingress() { local readonly ENCAP=$1 - echo "starting ingress ${ENCAP} encap test" + echo "starting ingress ${ENCAP} encap test ${VRF}" setup # need to wait a bit for IPv6 to autoconf, otherwise @@ -385,18 +412,22 @@ test_ingress() test_ping IPv6 0 # remove NS2->DST routes, pings fail - ip -netns ${NS2} route del ${IPv4_DST}/32 dev veth3 - ip -netns ${NS2} -6 route del ${IPv6_DST}/128 dev veth3 + ip -netns ${NS2} route del ${IPv4_DST}/32 dev veth3 ${VRF} + ip -netns ${NS2} -6 route del ${IPv6_DST}/128 dev veth3 ${VRF} test_ping IPv4 1 test_ping IPv6 1 # install replacement routes (LWT/eBPF), pings succeed if [ "${ENCAP}" == "IPv4" ] ; then - ip -netns ${NS2} route add ${IPv4_DST} encap bpf in obj test_lwt_ip_encap.o sec encap_gre dev veth2 - ip -netns ${NS2} -6 route add ${IPv6_DST} encap bpf in obj test_lwt_ip_encap.o sec encap_gre dev veth2 + ip -netns ${NS2} route add ${IPv4_DST} encap bpf in obj \ + test_lwt_ip_encap.o sec encap_gre dev veth2 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_DST} encap bpf in obj \ + test_lwt_ip_encap.o sec encap_gre dev veth2 ${VRF} elif [ "${ENCAP}" == "IPv6" ] ; then - ip -netns ${NS2} route add ${IPv4_DST} encap bpf in obj test_lwt_ip_encap.o sec encap_gre6 dev veth2 - ip -netns ${NS2} -6 route add ${IPv6_DST} encap bpf in obj test_lwt_ip_encap.o sec encap_gre6 dev veth2 + ip -netns ${NS2} route add ${IPv4_DST} encap bpf in obj \ + test_lwt_ip_encap.o sec encap_gre6 dev veth2 ${VRF} + ip -netns ${NS2} -6 route add ${IPv6_DST} encap bpf in obj \ + test_lwt_ip_encap.o sec encap_gre6 dev veth2 ${VRF} else echo "FAIL: unknown encap ${ENCAP}" TEST_STATUS=1 @@ -418,6 +449,13 @@ test_ingress() process_test_results } +VRF="" +test_egress IPv4 +test_egress IPv6 +test_ingress IPv4 +test_ingress IPv6 + +VRF="vrf red" test_egress IPv4 test_egress IPv6 test_ingress IPv4 From 725721a6506eea53bfde81a34e91a06d6162c216 Mon Sep 17 00:00:00 2001 From: Viet Hoang Tran Date: Mon, 15 Apr 2019 09:54:55 +0000 Subject: [PATCH 44/68] bpf: allow clearing all sock_ops callback flags The helper function bpf_sock_ops_cb_flags_set() can be used to both set and clear the sock_ops callback flags. However, its current behavior is not consistent. BPF program may clear a flag if more than one were set, or replace a flag with another one, but cannot clear all flags. This patch also updates the documentation to clarify the ability to clear flags of this helper function. Signed-off-by: Hoang Tran Acked-by: Martin KaFai Lau Signed-off-by: Alexei Starovoitov --- include/uapi/linux/bpf.h | 9 ++++++++- net/core/filter.c | 3 +-- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 704bb69514a2..eaf2d3284248 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -1737,12 +1737,19 @@ union bpf_attr { * error if an eBPF program tries to set a callback that is not * supported in the current kernel. * - * The supported callback values that *argval* can combine are: + * *argval* is a flag array which can combine these flags: * * * **BPF_SOCK_OPS_RTO_CB_FLAG** (retransmission time out) * * **BPF_SOCK_OPS_RETRANS_CB_FLAG** (retransmission) * * **BPF_SOCK_OPS_STATE_CB_FLAG** (TCP state change) * + * Therefore, this function can be used to clear a callback flag by + * setting the appropriate bit to zero. e.g. to disable the RTO + * callback: + * + * **bpf_sock_ops_cb_flags_set(bpf_sock,** + * **bpf_sock->bpf_sock_ops_cb_flags & ~BPF_SOCK_OPS_RTO_CB_FLAG)** + * * Here are some examples of where one could call such eBPF * program: * diff --git a/net/core/filter.c b/net/core/filter.c index bd1f51907b83..1833926a63fc 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -4437,8 +4437,7 @@ BPF_CALL_2(bpf_sock_ops_cb_flags_set, struct bpf_sock_ops_kern *, bpf_sock, if (!IS_ENABLED(CONFIG_INET) || !sk_fullsock(sk)) return -EINVAL; - if (val) - tcp_sk(sk)->bpf_sock_ops_cb_flags = val; + tcp_sk(sk)->bpf_sock_ops_cb_flags = val; return argval & (~BPF_SOCK_OPS_ALL_CB_FLAGS); } From 0d306c31b2f77391dacdeaad4470c577f2aecc4f Mon Sep 17 00:00:00 2001 From: Prashant Bhole Date: Tue, 16 Apr 2019 18:13:01 +0900 Subject: [PATCH 45/68] bpf: use BPF_CAST_CALL for casting bpf call verifier.c uses BPF_CAST_CALL for casting bpf call except at one place in jit_subprogs(). Let's use the macro for consistency. Signed-off-by: Prashant Bhole Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index c7220153c5b1..db301e9b5295 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -7647,9 +7647,8 @@ static int jit_subprogs(struct bpf_verifier_env *env) insn->src_reg != BPF_PSEUDO_CALL) continue; subprog = insn->off; - insn->imm = (u64 (*)(u64, u64, u64, u64, u64)) - func[subprog]->bpf_func - - __bpf_call_base; + insn->imm = BPF_CAST_CALL(func[subprog]->bpf_func) - + __bpf_call_base; } /* we use the aux data to keep a list of the start addresses From e1d1dc4653ecdea55cb0e96844f88da62c65cd4f Mon Sep 17 00:00:00 2001 From: Andrii Nakryiko Date: Tue, 16 Apr 2019 11:47:17 -0700 Subject: [PATCH 46/68] libbpf: fix printf formatter for ptrdiff_t argument Using %ld for printing out value of ptrdiff_t type is not portable between 32-bit and 64-bit archs. This is causing compilation errors for libbpf on 32-bit platform (discovered as part of an effort to integrate libbpf into systemd ([0])). Proper formatter is %td, which is used in this patch. v2->v1: - add Reported-by - provide more context on how this issue was discovered [0] https://github.com/systemd/systemd/pull/12151 Reported-by: Evgeny Vereshchagin Cc: Daniel Borkmann Cc: Alexei Starovoitov Cc: Yonghong Song Signed-off-by: Andrii Nakryiko Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/libbpf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index e5b77ad97795..d817bf20f3d6 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -817,7 +817,7 @@ bpf_object__init_internal_map(struct bpf_object *obj, struct bpf_map *map, memcpy(*data_buff, data->d_buf, data->d_size); } - pr_debug("map %ld is \"%s\"\n", map - obj->maps, map->name); + pr_debug("map %td is \"%s\"\n", map - obj->maps, map->name); return 0; } From f25377ee4fb1118650a08b403234aa6f57ce25a9 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Tue, 16 Apr 2019 13:13:47 -0700 Subject: [PATCH 47/68] bpftool: Support sysctl hook Add support for recently added BPF_PROG_TYPE_CGROUP_SYSCTL program type and BPF_CGROUP_SYSCTL attach type. Example of bpftool output with sysctl program from selftests: # bpftool p load ./test_sysctl_prog.o /mnt/bpf/sysctl_prog type cgroup/sysctl # bpftool p l 9: cgroup_sysctl name sysctl_tcp_mem tag 0dd05f81a8d0d52e gpl loaded_at 2019-04-16T12:57:27-0700 uid 0 xlated 1008B jited 623B memlock 4096B # bpftool c a /mnt/cgroup2/bla sysctl id 9 # bpftool c t CgroupPath ID AttachType AttachFlags Name /mnt/cgroup2/bla 9 sysctl sysctl_tcp_mem # bpftool c d /mnt/cgroup2/bla sysctl id 9 # bpftool c t CgroupPath ID AttachType AttachFlags Name Signed-off-by: Andrey Ignatov Acked-by: Song Liu Acked-by: Jakub Kicinski Signed-off-by: Alexei Starovoitov --- tools/bpf/bpftool/Documentation/bpftool-cgroup.rst | 5 +++-- tools/bpf/bpftool/Documentation/bpftool-prog.rst | 3 ++- tools/bpf/bpftool/bash-completion/bpftool | 7 ++++--- tools/bpf/bpftool/cgroup.c | 3 ++- tools/bpf/bpftool/main.h | 1 + tools/bpf/bpftool/prog.c | 2 +- 6 files changed, 13 insertions(+), 8 deletions(-) diff --git a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst index 5e3b7d9d7599..89b6b10e2183 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-cgroup.rst @@ -29,7 +29,7 @@ CGROUP COMMANDS | *PROG* := { **id** *PROG_ID* | **pinned** *FILE* | **tag** *PROG_TAG* } | *ATTACH_TYPE* := { **ingress** | **egress** | **sock_create** | **sock_ops** | **device** | | **bind4** | **bind6** | **post_bind4** | **post_bind6** | **connect4** | **connect6** | -| **sendmsg4** | **sendmsg6** } +| **sendmsg4** | **sendmsg6** | **sysctl** } | *ATTACH_FLAGS* := { **multi** | **override** } DESCRIPTION @@ -85,7 +85,8 @@ DESCRIPTION **sendmsg4** call to sendto(2), sendmsg(2), sendmmsg(2) for an unconnected udp4 socket (since 4.18); **sendmsg6** call to sendto(2), sendmsg(2), sendmmsg(2) for an - unconnected udp6 socket (since 4.18). + unconnected udp6 socket (since 4.18); + **sysctl** sysctl access (since 5.2). **bpftool cgroup detach** *CGROUP* *ATTACH_TYPE* *PROG* Detach *PROG* from the cgroup *CGROUP* and attach type diff --git a/tools/bpf/bpftool/Documentation/bpftool-prog.rst b/tools/bpf/bpftool/Documentation/bpftool-prog.rst index bb9bb00c0c2c..2f183ffd8351 100644 --- a/tools/bpf/bpftool/Documentation/bpftool-prog.rst +++ b/tools/bpf/bpftool/Documentation/bpftool-prog.rst @@ -39,7 +39,8 @@ PROG COMMANDS | **cgroup/sock** | **cgroup/dev** | **lwt_in** | **lwt_out** | **lwt_xmit** | | **lwt_seg6local** | **sockops** | **sk_skb** | **sk_msg** | **lirc_mode2** | | **cgroup/bind4** | **cgroup/bind6** | **cgroup/post_bind4** | **cgroup/post_bind6** | -| **cgroup/connect4** | **cgroup/connect6** | **cgroup/sendmsg4** | **cgroup/sendmsg6** +| **cgroup/connect4** | **cgroup/connect6** | **cgroup/sendmsg4** | **cgroup/sendmsg6** | +| **cgroup/sysctl** | } | *ATTACH_TYPE* := { | **msg_verdict** | **stream_verdict** | **stream_parser** | **flow_dissector** diff --git a/tools/bpf/bpftool/bash-completion/bpftool b/tools/bpf/bpftool/bash-completion/bpftool index b803827d01e8..9f3ffe1e26ab 100644 --- a/tools/bpf/bpftool/bash-completion/bpftool +++ b/tools/bpf/bpftool/bash-completion/bpftool @@ -370,7 +370,8 @@ _bpftool() lirc_mode2 cgroup/bind4 cgroup/bind6 \ cgroup/connect4 cgroup/connect6 \ cgroup/sendmsg4 cgroup/sendmsg6 \ - cgroup/post_bind4 cgroup/post_bind6" -- \ + cgroup/post_bind4 cgroup/post_bind6 \ + cgroup/sysctl" -- \ "$cur" ) ) return 0 ;; @@ -619,7 +620,7 @@ _bpftool() attach|detach) local ATTACH_TYPES='ingress egress sock_create sock_ops \ device bind4 bind6 post_bind4 post_bind6 connect4 \ - connect6 sendmsg4 sendmsg6' + connect6 sendmsg4 sendmsg6 sysctl' local ATTACH_FLAGS='multi override' local PROG_TYPE='id pinned tag' case $prev in @@ -629,7 +630,7 @@ _bpftool() ;; ingress|egress|sock_create|sock_ops|device|bind4|bind6|\ post_bind4|post_bind6|connect4|connect6|sendmsg4|\ - sendmsg6) + sendmsg6|sysctl) COMPREPLY=( $( compgen -W "$PROG_TYPE" -- \ "$cur" ) ) return 0 diff --git a/tools/bpf/bpftool/cgroup.c b/tools/bpf/bpftool/cgroup.c index a81b34343eb8..7e22f115c8c1 100644 --- a/tools/bpf/bpftool/cgroup.c +++ b/tools/bpf/bpftool/cgroup.c @@ -25,7 +25,7 @@ " ATTACH_TYPE := { ingress | egress | sock_create |\n" \ " sock_ops | device | bind4 | bind6 |\n" \ " post_bind4 | post_bind6 | connect4 |\n" \ - " connect6 | sendmsg4 | sendmsg6 }" + " connect6 | sendmsg4 | sendmsg6 | sysctl }" static const char * const attach_type_strings[] = { [BPF_CGROUP_INET_INGRESS] = "ingress", @@ -41,6 +41,7 @@ static const char * const attach_type_strings[] = { [BPF_CGROUP_INET6_POST_BIND] = "post_bind6", [BPF_CGROUP_UDP4_SENDMSG] = "sendmsg4", [BPF_CGROUP_UDP6_SENDMSG] = "sendmsg6", + [BPF_CGROUP_SYSCTL] = "sysctl", [__MAX_BPF_ATTACH_TYPE] = NULL, }; diff --git a/tools/bpf/bpftool/main.h b/tools/bpf/bpftool/main.h index d7dd84d3c660..1ccc46169a19 100644 --- a/tools/bpf/bpftool/main.h +++ b/tools/bpf/bpftool/main.h @@ -73,6 +73,7 @@ static const char * const prog_type_name[] = { [BPF_PROG_TYPE_LIRC_MODE2] = "lirc_mode2", [BPF_PROG_TYPE_SK_REUSEPORT] = "sk_reuseport", [BPF_PROG_TYPE_FLOW_DISSECTOR] = "flow_dissector", + [BPF_PROG_TYPE_CGROUP_SYSCTL] = "cgroup_sysctl", }; extern const char * const map_type_name[]; diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c index 4b7a307b9fc9..fc495b27f0fc 100644 --- a/tools/bpf/bpftool/prog.c +++ b/tools/bpf/bpftool/prog.c @@ -1060,7 +1060,7 @@ static int do_help(int argc, char **argv) " tracepoint | raw_tracepoint | xdp | perf_event | cgroup/skb |\n" " cgroup/sock | cgroup/dev | lwt_in | lwt_out | lwt_xmit |\n" " lwt_seg6local | sockops | sk_skb | sk_msg | lirc_mode2 |\n" - " sk_reuseport | flow_dissector |\n" + " sk_reuseport | flow_dissector | cgroup/sysctl |\n" " cgroup/bind4 | cgroup/bind6 | cgroup/post_bind4 |\n" " cgroup/post_bind6 | cgroup/connect4 | cgroup/connect6 |\n" " cgroup/sendmsg4 | cgroup/sendmsg6 }\n" From d459b59ee0f5ed2fa96e77bbd919e095b1aeceb0 Mon Sep 17 00:00:00 2001 From: Prashant Bhole Date: Wed, 17 Apr 2019 09:22:58 +0900 Subject: [PATCH 48/68] tools/bpftool: re-organize newline printing for map listing Let's move the final newline printing in show_map_close_plain() at the end of the function because it looks correct and consistent with prog.c. Also let's do related changes for the line which prints pinned file name. Signed-off-by: Prashant Bhole Reviewed-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- tools/bpf/bpftool/map.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index 44b192e87708..846d2e5c64b4 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -597,15 +597,16 @@ static int show_map_close_plain(int fd, struct bpf_map_info *info) } close(fd); - printf("\n"); if (!hash_empty(map_table.table)) { struct pinned_obj *obj; hash_for_each_possible(map_table.table, obj, hash, info->id) { if (obj->id == info->id) - printf("\tpinned %s\n", obj->path); + printf("\n\tpinned %s", obj->path); } } + + printf("\n"); return 0; } From d1b7725dfea3e6c1aff2ee94baf3658aba450fbe Mon Sep 17 00:00:00 2001 From: Prashant Bhole Date: Wed, 17 Apr 2019 09:22:59 +0900 Subject: [PATCH 49/68] tools/bpftool: show btf_id in map listing Let's print btf id of map similar to the way we are printing it for programs. Sample output: user@test# bpftool map -f 61: lpm_trie flags 0x1 key 20B value 8B max_entries 1 memlock 4096B 133: array name test_btf_id flags 0x0 key 4B value 4B max_entries 4 memlock 4096B pinned /sys/fs/bpf/test100 btf_id 174 170: array name test_btf_id flags 0x0 key 4B value 4B max_entries 4 memlock 4096B btf_id 240 Signed-off-by: Prashant Bhole Reviewed-by: Quentin Monnet Reviewed-by: Jakub Kicinski Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- tools/bpf/bpftool/map.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/tools/bpf/bpftool/map.c b/tools/bpf/bpftool/map.c index 846d2e5c64b4..10b6c9d3e525 100644 --- a/tools/bpf/bpftool/map.c +++ b/tools/bpf/bpftool/map.c @@ -531,6 +531,9 @@ static int show_map_close_json(int fd, struct bpf_map_info *info) } close(fd); + if (info->btf_id) + jsonw_int_field(json_wtr, "btf_id", info->btf_id); + if (!hash_empty(map_table.table)) { struct pinned_obj *obj; @@ -606,6 +609,9 @@ static int show_map_close_plain(int fd, struct bpf_map_info *info) } } + if (info->btf_id) + printf("\n\tbtf_id %d", info->btf_id); + printf("\n"); return 0; } From f63666de2ba9c1c3ac0ec57fc5d3032514ec80f1 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 16 Apr 2019 14:58:08 +0200 Subject: [PATCH 50/68] xsk: fix XDP socket ring buffer memory ordering The ring buffer code of XDP sockets is missing a memory barrier on the consumer side between the load of the data and the write that signals that it is ok for the producer to put new data into the buffer. On architectures that does not guarantee that stores are not reordered with older loads, the producer might put data into the ring before the consumer had the chance to read it. As IA does guarantee this ordering, it would only need a compiler barrier here, but there are no primitives in Linux for this specific case (hinder writes to be ordered before older reads) so I had to add a smp_mb() here which will translate into a run-time synch operation on IA. Added a longish comment in the code explaining what each barrier in the ring implementation accomplishes and what would happen if we removed one of them. Signed-off-by: Magnus Karlsson Signed-off-by: Alexei Starovoitov --- net/xdp/xsk_queue.h | 56 +++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 52 insertions(+), 4 deletions(-) diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h index 610c0bdc0c2b..88b9ae24658d 100644 --- a/net/xdp/xsk_queue.h +++ b/net/xdp/xsk_queue.h @@ -43,6 +43,48 @@ struct xsk_queue { u64 invalid_descs; }; +/* The structure of the shared state of the rings are the same as the + * ring buffer in kernel/events/ring_buffer.c. For the Rx and completion + * ring, the kernel is the producer and user space is the consumer. For + * the Tx and fill rings, the kernel is the consumer and user space is + * the producer. + * + * producer consumer + * + * if (LOAD ->consumer) { LOAD ->producer + * (A) smp_rmb() (C) + * STORE $data LOAD $data + * smp_wmb() (B) smp_mb() (D) + * STORE ->producer STORE ->consumer + * } + * + * (A) pairs with (D), and (B) pairs with (C). + * + * Starting with (B), it protects the data from being written after + * the producer pointer. If this barrier was missing, the consumer + * could observe the producer pointer being set and thus load the data + * before the producer has written the new data. The consumer would in + * this case load the old data. + * + * (C) protects the consumer from speculatively loading the data before + * the producer pointer actually has been read. If we do not have this + * barrier, some architectures could load old data as speculative loads + * are not discarded as the CPU does not know there is a dependency + * between ->producer and data. + * + * (A) is a control dependency that separates the load of ->consumer + * from the stores of $data. In case ->consumer indicates there is no + * room in the buffer to store $data we do not. So no barrier is needed. + * + * (D) protects the load of the data to be observed to happen after the + * store of the consumer pointer. If we did not have this memory + * barrier, the producer could observe the consumer pointer being set + * and overwrite the data with a new value before the consumer got the + * chance to read the old value. The consumer would thus miss reading + * the old entry and very likely read the new entry twice, once right + * now and again after circling through the ring. + */ + /* Common functions operating for both RXTX and umem queues */ static inline u64 xskq_nb_invalid_descs(struct xsk_queue *q) @@ -106,6 +148,7 @@ static inline u64 *xskq_validate_addr(struct xsk_queue *q, u64 *addr) static inline u64 *xskq_peek_addr(struct xsk_queue *q, u64 *addr) { if (q->cons_tail == q->cons_head) { + smp_mb(); /* D, matches A */ WRITE_ONCE(q->ring->consumer, q->cons_tail); q->cons_head = q->cons_tail + xskq_nb_avail(q, RX_BATCH_SIZE); @@ -128,10 +171,11 @@ static inline int xskq_produce_addr(struct xsk_queue *q, u64 addr) if (xskq_nb_free(q, q->prod_tail, 1) == 0) return -ENOSPC; + /* A, matches D */ ring->desc[q->prod_tail++ & q->ring_mask] = addr; /* Order producer and data */ - smp_wmb(); + smp_wmb(); /* B, matches C */ WRITE_ONCE(q->ring->producer, q->prod_tail); return 0; @@ -144,6 +188,7 @@ static inline int xskq_produce_addr_lazy(struct xsk_queue *q, u64 addr) if (xskq_nb_free(q, q->prod_head, LAZY_UPDATE_THRESHOLD) == 0) return -ENOSPC; + /* A, matches D */ ring->desc[q->prod_head++ & q->ring_mask] = addr; return 0; } @@ -152,7 +197,7 @@ static inline void xskq_produce_flush_addr_n(struct xsk_queue *q, u32 nb_entries) { /* Order producer and data */ - smp_wmb(); + smp_wmb(); /* B, matches C */ q->prod_tail += nb_entries; WRITE_ONCE(q->ring->producer, q->prod_tail); @@ -163,6 +208,7 @@ static inline int xskq_reserve_addr(struct xsk_queue *q) if (xskq_nb_free(q, q->prod_head, 1) == 0) return -ENOSPC; + /* A, matches D */ q->prod_head++; return 0; } @@ -204,11 +250,12 @@ static inline struct xdp_desc *xskq_peek_desc(struct xsk_queue *q, struct xdp_desc *desc) { if (q->cons_tail == q->cons_head) { + smp_mb(); /* D, matches A */ WRITE_ONCE(q->ring->consumer, q->cons_tail); q->cons_head = q->cons_tail + xskq_nb_avail(q, RX_BATCH_SIZE); /* Order consumer and data */ - smp_rmb(); + smp_rmb(); /* C, matches B */ } return xskq_validate_desc(q, desc); @@ -228,6 +275,7 @@ static inline int xskq_produce_batch_desc(struct xsk_queue *q, if (xskq_nb_free(q, q->prod_head, 1) == 0) return -ENOSPC; + /* A, matches D */ idx = (q->prod_head++) & q->ring_mask; ring->desc[idx].addr = addr; ring->desc[idx].len = len; @@ -238,7 +286,7 @@ static inline int xskq_produce_batch_desc(struct xsk_queue *q, static inline void xskq_produce_flush_desc(struct xsk_queue *q) { /* Order producer and data */ - smp_wmb(); + smp_wmb(); /* B, matches C */ q->prod_tail = q->prod_head, WRITE_ONCE(q->ring->producer, q->prod_tail); From d5e63fdd443378531fd1a67a15a720a799635d93 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 16 Apr 2019 14:58:09 +0200 Subject: [PATCH 51/68] libbpf: fix XDP socket ring buffer memory ordering The ring buffer code of XDP sockets is missing a memory barrier on the consumer side between the load of the data and the write that signals that it is ok for the producer to put new data into the buffer. On architectures that does not guarantee that stores are not reordered with older loads, the producer might put data into the ring before the consumer had the chance to read it. As IA does guarantee this ordering, it would only need a compiler barrier here, but there are no primitives in barrier.h for this specific case (hinder writes to be ordered before older reads) so I had to add a smp_mb() here which will translate into a run-time synch operation on IA. Signed-off-by: Magnus Karlsson Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/xsk.h | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/tools/lib/bpf/xsk.h b/tools/lib/bpf/xsk.h index a497f00e2962..1b35c40dff73 100644 --- a/tools/lib/bpf/xsk.h +++ b/tools/lib/bpf/xsk.h @@ -36,6 +36,10 @@ struct name { \ DEFINE_XSK_RING(xsk_ring_prod); DEFINE_XSK_RING(xsk_ring_cons); +/* For a detailed explanation on the memory barriers associated with the + * ring, please take a look at net/xdp/xsk_queue.h. + */ + struct xsk_umem; struct xsk_socket; @@ -116,8 +120,8 @@ static inline size_t xsk_ring_prod__reserve(struct xsk_ring_prod *prod, static inline void xsk_ring_prod__submit(struct xsk_ring_prod *prod, size_t nb) { - /* Make sure everything has been written to the ring before signalling - * this to the kernel. + /* Make sure everything has been written to the ring before indicating + * this to the kernel by writing the producer pointer. */ smp_wmb(); @@ -144,6 +148,11 @@ static inline size_t xsk_ring_cons__peek(struct xsk_ring_cons *cons, static inline void xsk_ring_cons__release(struct xsk_ring_cons *cons, size_t nb) { + /* Make sure data has been read before indicating we are done + * with the entries by updating the consumer pointer. + */ + smp_mb(); + *cons->consumer += nb; } From a06d729646e87afc5ce06e539aecf762cc26c6e3 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 16 Apr 2019 14:58:10 +0200 Subject: [PATCH 52/68] libbpf: remove likely/unlikely in xsk.h This patch removes the use of likely and unlikely in xsk.h since they create a dependency on Linux headers as reported by several users. There have also been reports that the use of these decreases performance as the compiler puts the code on two different cache lines instead of on a single one. All in all, I think we are better off without them. Fixes: 1cad07884239 ("libbpf: add support for using AF_XDP sockets") Signed-off-by: Magnus Karlsson Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/xsk.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/lib/bpf/xsk.h b/tools/lib/bpf/xsk.h index 1b35c40dff73..f264f24f06ac 100644 --- a/tools/lib/bpf/xsk.h +++ b/tools/lib/bpf/xsk.h @@ -109,7 +109,7 @@ static inline __u32 xsk_cons_nb_avail(struct xsk_ring_cons *r, __u32 nb) static inline size_t xsk_ring_prod__reserve(struct xsk_ring_prod *prod, size_t nb, __u32 *idx) { - if (unlikely(xsk_prod_nb_free(prod, nb) < nb)) + if (xsk_prod_nb_free(prod, nb) < nb) return 0; *idx = prod->cached_prod; @@ -133,7 +133,7 @@ static inline size_t xsk_ring_cons__peek(struct xsk_ring_cons *cons, { size_t entries = xsk_cons_nb_avail(cons, nb); - if (likely(entries > 0)) { + if (entries > 0) { /* Make sure we do not speculatively read the data before * we have received the packet buffers from the ring. */ From b7e3a28019c92ffe1f55de278c5641de33b6259a Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 16 Apr 2019 14:58:11 +0200 Subject: [PATCH 53/68] libbpf: remove dependency on barrier.h in xsk.h The use of smp_rmb() and smp_wmb() creates a Linux header dependency on barrier.h that is unnecessary in most parts. This patch implements the two small defines that are needed from barrier.h. As a bonus, the new implementations are faster than the default ones as they default to sfence and lfence for x86, while we only need a compiler barrier in our case. Just as it is when the same ring access code is compiled in the kernel. Fixes: 1cad07884239 ("libbpf: add support for using AF_XDP sockets") Signed-off-by: Magnus Karlsson Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/libbpf_util.h | 25 +++++++++++++++++++++++++ tools/lib/bpf/xsk.h | 7 ++++--- 2 files changed, 29 insertions(+), 3 deletions(-) diff --git a/tools/lib/bpf/libbpf_util.h b/tools/lib/bpf/libbpf_util.h index 81ecda0cb9c9..71fbb2898b87 100644 --- a/tools/lib/bpf/libbpf_util.h +++ b/tools/lib/bpf/libbpf_util.h @@ -23,6 +23,31 @@ do { \ #define pr_info(fmt, ...) __pr(LIBBPF_INFO, fmt, ##__VA_ARGS__) #define pr_debug(fmt, ...) __pr(LIBBPF_DEBUG, fmt, ##__VA_ARGS__) +/* Use these barrier functions instead of smp_[rw]mb() when they are + * used in a libbpf header file. That way they can be built into the + * application that uses libbpf. + */ +#if defined(__i386__) || defined(__x86_64__) +# define libbpf_smp_rmb() asm volatile("" : : : "memory") +# define libbpf_smp_wmb() asm volatile("" : : : "memory") +# define libbpf_smp_mb() \ + asm volatile("lock; addl $0,-4(%%rsp)" : : : "memory", "cc") +#elif defined(__aarch64__) +# define libbpf_smp_rmb() asm volatile("dmb ishld" : : : "memory") +# define libbpf_smp_wmb() asm volatile("dmb ishst" : : : "memory") +# define libbpf_smp_mb() asm volatile("dmb ish" : : : "memory") +#elif defined(__arm__) +/* These are only valid for armv7 and above */ +# define libbpf_smp_rmb() asm volatile("dmb ish" : : : "memory") +# define libbpf_smp_wmb() asm volatile("dmb ishst" : : : "memory") +# define libbpf_smp_mb() asm volatile("dmb ish" : : : "memory") +#else +# warning Architecture missing native barrier functions in libbpf_util.h. +# define libbpf_smp_rmb() __sync_synchronize() +# define libbpf_smp_wmb() __sync_synchronize() +# define libbpf_smp_mb() __sync_synchronize() +#endif + #ifdef __cplusplus } /* extern "C" */ #endif diff --git a/tools/lib/bpf/xsk.h b/tools/lib/bpf/xsk.h index f264f24f06ac..2377c7a7f1b1 100644 --- a/tools/lib/bpf/xsk.h +++ b/tools/lib/bpf/xsk.h @@ -16,6 +16,7 @@ #include #include "libbpf.h" +#include "libbpf_util.h" #ifdef __cplusplus extern "C" { @@ -123,7 +124,7 @@ static inline void xsk_ring_prod__submit(struct xsk_ring_prod *prod, size_t nb) /* Make sure everything has been written to the ring before indicating * this to the kernel by writing the producer pointer. */ - smp_wmb(); + libbpf_smp_wmb(); *prod->producer += nb; } @@ -137,7 +138,7 @@ static inline size_t xsk_ring_cons__peek(struct xsk_ring_cons *cons, /* Make sure we do not speculatively read the data before * we have received the packet buffers from the ring. */ - smp_rmb(); + libbpf_smp_rmb(); *idx = cons->cached_cons; cons->cached_cons += entries; @@ -151,7 +152,7 @@ static inline void xsk_ring_cons__release(struct xsk_ring_cons *cons, size_t nb) /* Make sure data has been read before indicating we are done * with the entries by updating the consumer pointer. */ - smp_mb(); + libbpf_smp_mb(); *cons->consumer += nb; } From 2c5935f1b2b642cee8e1562396ec8a7781fc4c6d Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Tue, 16 Apr 2019 14:58:12 +0200 Subject: [PATCH 54/68] libbpf: optimize barrier for XDP socket rings The full memory barrier in the XDP socket rings on the consumer side between the load of the data and the store of the consumer ring is there to protect the store from being executed before the load of the data. If this was allowed to happen, the producer might overwrite the data field with a new entry before the consumer got the chance to read it. On x86, stores are guaranteed not to be reordered with older loads, so it does not need a full memory barrier here. A compile time barrier would be enough. This patch introdcues a new primitive in libbpf_util.h that implements a new barrier type (libbpf_smp_rwmb) hindering stores to be reordered with older loads. It is then used in the XDP socket ring access code in libbpf to improve performance. Signed-off-by: Magnus Karlsson Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/libbpf_util.h | 5 +++++ tools/lib/bpf/xsk.h | 2 +- 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/lib/bpf/libbpf_util.h b/tools/lib/bpf/libbpf_util.h index 71fbb2898b87..172b707e007b 100644 --- a/tools/lib/bpf/libbpf_util.h +++ b/tools/lib/bpf/libbpf_util.h @@ -32,20 +32,25 @@ do { \ # define libbpf_smp_wmb() asm volatile("" : : : "memory") # define libbpf_smp_mb() \ asm volatile("lock; addl $0,-4(%%rsp)" : : : "memory", "cc") +/* Hinders stores to be observed before older loads. */ +# define libbpf_smp_rwmb() asm volatile("" : : : "memory") #elif defined(__aarch64__) # define libbpf_smp_rmb() asm volatile("dmb ishld" : : : "memory") # define libbpf_smp_wmb() asm volatile("dmb ishst" : : : "memory") # define libbpf_smp_mb() asm volatile("dmb ish" : : : "memory") +# define libbpf_smp_rwmb() libbpf_smp_mb() #elif defined(__arm__) /* These are only valid for armv7 and above */ # define libbpf_smp_rmb() asm volatile("dmb ish" : : : "memory") # define libbpf_smp_wmb() asm volatile("dmb ishst" : : : "memory") # define libbpf_smp_mb() asm volatile("dmb ish" : : : "memory") +# define libbpf_smp_rwmb() libbpf_smp_mb() #else # warning Architecture missing native barrier functions in libbpf_util.h. # define libbpf_smp_rmb() __sync_synchronize() # define libbpf_smp_wmb() __sync_synchronize() # define libbpf_smp_mb() __sync_synchronize() +# define libbpf_smp_rwmb() __sync_synchronize() #endif #ifdef __cplusplus diff --git a/tools/lib/bpf/xsk.h b/tools/lib/bpf/xsk.h index 2377c7a7f1b1..82ea71a0f3ec 100644 --- a/tools/lib/bpf/xsk.h +++ b/tools/lib/bpf/xsk.h @@ -152,7 +152,7 @@ static inline void xsk_ring_cons__release(struct xsk_ring_cons *cons, size_t nb) /* Make sure data has been read before indicating we are done * with the entries by updating the consumer pointer. */ - libbpf_smp_mb(); + libbpf_smp_rwmb(); *cons->consumer += nb; } From 77361825bb01ecadf3ac8622e2e4dbc28806e858 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 12 Apr 2019 17:07:32 +0200 Subject: [PATCH 55/68] bpf: cpumap use ptr_ring_consume_batched Move ptr_ring dequeue outside loop, that allocate SKBs and calls network stack, as these operations that can take some time. The ptr_ring is a communication channel between CPUs, where we want to reduce/limit any cacheline bouncing. Do a concentrated bulk dequeue via ptr_ring_consume_batched, to shorten the period and times the remote cacheline in ptr_ring is read Batch size 8 is both to (1) limit BH-disable period, and (2) consume one cacheline on 64-bit archs. After reducing the BH-disable section further then we can consider changing this, while still thinking about L1 cacheline size being active. Signed-off-by: Jesper Dangaard Brouer Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- kernel/bpf/cpumap.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-) diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index 3c18260403dd..430103e182a0 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -240,6 +240,8 @@ static void put_cpu_map_entry(struct bpf_cpu_map_entry *rcpu) } } +#define CPUMAP_BATCH 8 + static int cpu_map_kthread_run(void *data) { struct bpf_cpu_map_entry *rcpu = data; @@ -252,8 +254,9 @@ static int cpu_map_kthread_run(void *data) * kthread_stop signal until queue is empty. */ while (!kthread_should_stop() || !__ptr_ring_empty(rcpu->queue)) { - unsigned int processed = 0, drops = 0, sched = 0; - struct xdp_frame *xdpf; + unsigned int drops = 0, sched = 0; + void *frames[CPUMAP_BATCH]; + int i, n; /* Release CPU reschedule checks */ if (__ptr_ring_empty(rcpu->queue)) { @@ -269,14 +272,16 @@ static int cpu_map_kthread_run(void *data) sched = cond_resched(); } - /* Process packets in rcpu->queue */ - local_bh_disable(); /* * The bpf_cpu_map_entry is single consumer, with this * kthread CPU pinned. Lockless access to ptr_ring * consume side valid as no-resize allowed of queue. */ - while ((xdpf = __ptr_ring_consume(rcpu->queue))) { + n = ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH); + + local_bh_disable(); + for (i = 0; i < n; i++) { + struct xdp_frame *xdpf = frames[i]; struct sk_buff *skb; int ret; @@ -290,13 +295,9 @@ static int cpu_map_kthread_run(void *data) ret = netif_receive_skb_core(skb); if (ret == NET_RX_DROP) drops++; - - /* Limit BH-disable period */ - if (++processed == 8) - break; } /* Feedback loop via tracepoint */ - trace_xdp_cpumap_kthread(rcpu->map_id, processed, drops, sched); + trace_xdp_cpumap_kthread(rcpu->map_id, n, drops, sched); local_bh_enable(); /* resched point, may call do_softirq() */ } From ba0509b6881efd0c8b26c36490cba87d8fb324c0 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 12 Apr 2019 17:07:37 +0200 Subject: [PATCH 56/68] net: core: introduce build_skb_around The function build_skb() also have the responsibility to allocate and clear the SKB structure. Introduce a new function build_skb_around(), that moves the responsibility of allocation and clearing to the caller. This allows caller to use kmem_cache (slab/slub) bulk allocation API. Next patch use this function combined with kmem_cache_alloc_bulk. Signed-off-by: Jesper Dangaard Brouer Acked-by: Song Liu Acked-by: Eric Dumazet Signed-off-by: Alexei Starovoitov --- include/linux/skbuff.h | 2 ++ net/core/skbuff.c | 71 +++++++++++++++++++++++++++++++----------- 2 files changed, 54 insertions(+), 19 deletions(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index a06275a618f0..e81f2b0e8a83 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1042,6 +1042,8 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t priority, int flags, int node); struct sk_buff *__build_skb(void *data, unsigned int frag_size); struct sk_buff *build_skb(void *data, unsigned int frag_size); +struct sk_buff *build_skb_around(struct sk_buff *skb, + void *data, unsigned int frag_size); /** * alloc_skb - allocate a network buffer diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 9901f5322852..087622298d77 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -258,6 +258,33 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, } EXPORT_SYMBOL(__alloc_skb); +/* Caller must provide SKB that is memset cleared */ +static struct sk_buff *__build_skb_around(struct sk_buff *skb, + void *data, unsigned int frag_size) +{ + struct skb_shared_info *shinfo; + unsigned int size = frag_size ? : ksize(data); + + size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info)); + + /* Assumes caller memset cleared SKB */ + skb->truesize = SKB_TRUESIZE(size); + refcount_set(&skb->users, 1); + skb->head = data; + skb->data = data; + skb_reset_tail_pointer(skb); + skb->end = skb->tail + size; + skb->mac_header = (typeof(skb->mac_header))~0U; + skb->transport_header = (typeof(skb->transport_header))~0U; + + /* make sure we initialize shinfo sequentially */ + shinfo = skb_shinfo(skb); + memset(shinfo, 0, offsetof(struct skb_shared_info, dataref)); + atomic_set(&shinfo->dataref, 1); + + return skb; +} + /** * __build_skb - build a network buffer * @data: data buffer provided by caller @@ -279,32 +306,15 @@ EXPORT_SYMBOL(__alloc_skb); */ struct sk_buff *__build_skb(void *data, unsigned int frag_size) { - struct skb_shared_info *shinfo; struct sk_buff *skb; - unsigned int size = frag_size ? : ksize(data); skb = kmem_cache_alloc(skbuff_head_cache, GFP_ATOMIC); - if (!skb) + if (unlikely(!skb)) return NULL; - size -= SKB_DATA_ALIGN(sizeof(struct skb_shared_info)); - memset(skb, 0, offsetof(struct sk_buff, tail)); - skb->truesize = SKB_TRUESIZE(size); - refcount_set(&skb->users, 1); - skb->head = data; - skb->data = data; - skb_reset_tail_pointer(skb); - skb->end = skb->tail + size; - skb->mac_header = (typeof(skb->mac_header))~0U; - skb->transport_header = (typeof(skb->transport_header))~0U; - /* make sure we initialize shinfo sequentially */ - shinfo = skb_shinfo(skb); - memset(shinfo, 0, offsetof(struct skb_shared_info, dataref)); - atomic_set(&shinfo->dataref, 1); - - return skb; + return __build_skb_around(skb, data, frag_size); } /* build_skb() is wrapper over __build_skb(), that specifically @@ -325,6 +335,29 @@ struct sk_buff *build_skb(void *data, unsigned int frag_size) } EXPORT_SYMBOL(build_skb); +/** + * build_skb_around - build a network buffer around provided skb + * @skb: sk_buff provide by caller, must be memset cleared + * @data: data buffer provided by caller + * @frag_size: size of data, or 0 if head was kmalloced + */ +struct sk_buff *build_skb_around(struct sk_buff *skb, + void *data, unsigned int frag_size) +{ + if (unlikely(!skb)) + return NULL; + + skb = __build_skb_around(skb, data, frag_size); + + if (skb && frag_size) { + skb->head_frag = 1; + if (page_is_pfmemalloc(virt_to_head_page(data))) + skb->pfmemalloc = 1; + } + return skb; +} +EXPORT_SYMBOL(build_skb_around); + #define NAPI_SKB_CACHE_SIZE 64 struct napi_alloc_cache { From 8f0504a97e1ba6b70e1c8b5a88255c280f263287 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 12 Apr 2019 17:07:43 +0200 Subject: [PATCH 57/68] bpf: cpumap do bulk allocation of SKBs As cpumap now batch consume xdp_frame's from the ptr_ring, it knows how many SKBs it need to allocate. Thus, lets bulk allocate these SKBs via kmem_cache_alloc_bulk() API, and use the previously introduced function build_skb_around(). Notice that the flag __GFP_ZERO asks the slab/slub allocator to clear the memory for us. This does clear a larger area than needed, but my micro benchmarks on Intel CPUs show that this is slightly faster due to being a cacheline aligned area is cleared for the SKBs. (For SLUB allocator, there is a future optimization potential, because SKBs will with high probability originate from same page. If we can find/identify continuous memory areas then the Intel CPU memset rep stos will have a real performance gain.) Signed-off-by: Jesper Dangaard Brouer Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- kernel/bpf/cpumap.c | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-) diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index 430103e182a0..732d6ced3987 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -160,12 +160,12 @@ static void cpu_map_kthread_stop(struct work_struct *work) } static struct sk_buff *cpu_map_build_skb(struct bpf_cpu_map_entry *rcpu, - struct xdp_frame *xdpf) + struct xdp_frame *xdpf, + struct sk_buff *skb) { unsigned int hard_start_headroom; unsigned int frame_size; void *pkt_data_start; - struct sk_buff *skb; /* Part of headroom was reserved to xdpf */ hard_start_headroom = sizeof(struct xdp_frame) + xdpf->headroom; @@ -191,8 +191,8 @@ static struct sk_buff *cpu_map_build_skb(struct bpf_cpu_map_entry *rcpu, SKB_DATA_ALIGN(sizeof(struct skb_shared_info)); pkt_data_start = xdpf->data - hard_start_headroom; - skb = build_skb(pkt_data_start, frame_size); - if (!skb) + skb = build_skb_around(skb, pkt_data_start, frame_size); + if (unlikely(!skb)) return NULL; skb_reserve(skb, hard_start_headroom); @@ -256,7 +256,9 @@ static int cpu_map_kthread_run(void *data) while (!kthread_should_stop() || !__ptr_ring_empty(rcpu->queue)) { unsigned int drops = 0, sched = 0; void *frames[CPUMAP_BATCH]; - int i, n; + void *skbs[CPUMAP_BATCH]; + gfp_t gfp = __GFP_ZERO | GFP_ATOMIC; + int i, n, m; /* Release CPU reschedule checks */ if (__ptr_ring_empty(rcpu->queue)) { @@ -278,14 +280,20 @@ static int cpu_map_kthread_run(void *data) * consume side valid as no-resize allowed of queue. */ n = ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH); + m = kmem_cache_alloc_bulk(skbuff_head_cache, gfp, n, skbs); + if (unlikely(m == 0)) { + for (i = 0; i < n; i++) + skbs[i] = NULL; /* effect: xdp_return_frame */ + drops = n; + } local_bh_disable(); for (i = 0; i < n; i++) { struct xdp_frame *xdpf = frames[i]; - struct sk_buff *skb; + struct sk_buff *skb = skbs[i]; int ret; - skb = cpu_map_build_skb(rcpu, xdpf); + skb = cpu_map_build_skb(rcpu, xdpf, skb); if (!skb) { xdp_return_frame(xdpf); continue; From 86d231459d6dc9094e70c35c3517f4ef860b2f1e Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Fri, 12 Apr 2019 17:07:48 +0200 Subject: [PATCH 58/68] bpf: cpumap memory prefetchw optimizations for struct page A lot of the performance gain comes from this patch. While analysing performance overhead it was found that the largest CPU stalls were caused when touching the struct page area. It is first read with a READ_ONCE from build_skb_around via page_is_pfmemalloc(), and when freed written by page_frag_free() call. Measurements show that the prefetchw (W) variant operation is needed to achieve the performance gain. We believe this optimization it two fold, first the W-variant saves one step in the cache-coherency protocol, and second it helps us to avoid the non-temporal prefetch HW optimizations and bring this into all cache-levels. It might be worth investigating if prefetch into L2 will have the same benefit. Signed-off-by: Jesper Dangaard Brouer Acked-by: Ilias Apalodimas Acked-by: Song Liu Signed-off-by: Alexei Starovoitov --- kernel/bpf/cpumap.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c index 732d6ced3987..cf727d77c6c6 100644 --- a/kernel/bpf/cpumap.c +++ b/kernel/bpf/cpumap.c @@ -280,6 +280,18 @@ static int cpu_map_kthread_run(void *data) * consume side valid as no-resize allowed of queue. */ n = ptr_ring_consume_batched(rcpu->queue, frames, CPUMAP_BATCH); + + for (i = 0; i < n; i++) { + void *f = frames[i]; + struct page *page = virt_to_page(f); + + /* Bring struct page memory area to curr CPU. Read by + * build_skb_around via page_is_pfmemalloc(), and when + * freed written by page_frag_free call. + */ + prefetchw(page); + } + m = kmem_cache_alloc_bulk(skbuff_head_cache, gfp, n, skbs); if (unlikely(m == 0)) { for (i = 0; i < n; i++) From ba02de1aa04e392e15ef503c6dd5166915d9d4de Mon Sep 17 00:00:00 2001 From: Yonghong Song Date: Wed, 17 Apr 2019 22:23:30 -0700 Subject: [PATCH 59/68] selftests/bpf: fix a compilation error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit I hit the following compilation error with gcc 4.8.5. prog_tests/flow_dissector.c: In function ‘test_flow_dissector’: prog_tests/flow_dissector.c:155:2: error: ‘for’ loop initial declarations are only allowed in C99 mode for (int i = 0; i < ARRAY_SIZE(tests); i++) { ^ prog_tests/flow_dissector.c:155:2: note: use option -std=c99 or -std=gnu99 to compile your code Let us fix the issue by avoiding this particular c99 feature. Fixes: a5cb33464e53 ("selftests/bpf: make flow dissector tests more extensible") Signed-off-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/prog_tests/flow_dissector.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c index 1a73937826b0..126319f9a97c 100644 --- a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c +++ b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c @@ -143,7 +143,7 @@ struct test tests[] = { void test_flow_dissector(void) { struct bpf_object *obj; - int err, prog_fd; + int i, err, prog_fd; err = bpf_flow_load(&obj, "./bpf_flow.o", "flow_dissector", "jmp_table", &prog_fd); @@ -152,7 +152,7 @@ void test_flow_dissector(void) return; } - for (int i = 0; i < ARRAY_SIZE(tests); i++) { + for (i = 0; i < ARRAY_SIZE(tests); i++) { struct bpf_flow_keys flow_keys; struct bpf_prog_test_run_attr tattr = { .prog_fd = prog_fd, From da70314917862d4da4a8d7601cd47339df8b3c23 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Wed, 17 Apr 2019 22:28:57 -0700 Subject: [PATCH 60/68] bpf: Document BPF_PROG_TYPE_CGROUP_SYSCTL Add documentation for BPF_PROG_TYPE_CGROUP_SYSCTL, including general info, attach type, context, return code, helpers, example and usage considerations. A separate file prog_cgroup_sysctl.rst is added to Documentation/bpf/. In the future more program types can be documented in their own prog_.rst files. Another way to place program type specific documentation would be to group program types somehow (e.g. cgroup.rst for all cgroup-bpf programs), but it may not scale well since some program types may belong to different groups, e.g. BPF_PROG_TYPE_CGROUP_SKB can be documented together with either cgroup-bpf programs or programs that access skb. The new file is added to the index and verified by `make htmldocs` / sanity-check by lynx. Signed-off-by: Andrey Ignatov Acked-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- Documentation/bpf/index.rst | 9 ++ Documentation/bpf/prog_cgroup_sysctl.rst | 125 +++++++++++++++++++++++ 2 files changed, 134 insertions(+) create mode 100644 Documentation/bpf/prog_cgroup_sysctl.rst diff --git a/Documentation/bpf/index.rst b/Documentation/bpf/index.rst index 4e77932959cc..dadcaa9a9f5f 100644 --- a/Documentation/bpf/index.rst +++ b/Documentation/bpf/index.rst @@ -36,6 +36,15 @@ Two sets of Questions and Answers (Q&A) are maintained. bpf_devel_QA +Program types +============= + +.. toctree:: + :maxdepth: 1 + + prog_cgroup_sysctl + + .. Links: .. _Documentation/networking/filter.txt: ../networking/filter.txt .. _man-pages: https://www.kernel.org/doc/man-pages/ diff --git a/Documentation/bpf/prog_cgroup_sysctl.rst b/Documentation/bpf/prog_cgroup_sysctl.rst new file mode 100644 index 000000000000..677d6c637cf3 --- /dev/null +++ b/Documentation/bpf/prog_cgroup_sysctl.rst @@ -0,0 +1,125 @@ +.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) + +=========================== +BPF_PROG_TYPE_CGROUP_SYSCTL +=========================== + +This document describes ``BPF_PROG_TYPE_CGROUP_SYSCTL`` program type that +provides cgroup-bpf hook for sysctl. + +The hook has to be attached to a cgroup and will be called every time a +process inside that cgroup tries to read from or write to sysctl knob in proc. + +1. Attach type +************** + +``BPF_CGROUP_SYSCTL`` attach type has to be used to attach +``BPF_PROG_TYPE_CGROUP_SYSCTL`` program to a cgroup. + +2. Context +********** + +``BPF_PROG_TYPE_CGROUP_SYSCTL`` provides access to the following context from +BPF program:: + + struct bpf_sysctl { + __u32 write; + __u32 file_pos; + }; + +* ``write`` indicates whether sysctl value is being read (``0``) or written + (``1``). This field is read-only. + +* ``file_pos`` indicates file position sysctl is being accessed at, read + or written. This field is read-write. Writing to the field sets the starting + position in sysctl proc file ``read(2)`` will be reading from or ``write(2)`` + will be writing to. Writing zero to the field can be used e.g. to override + whole sysctl value by ``bpf_sysctl_set_new_value()`` on ``write(2)`` even + when it's called by user space on ``file_pos > 0``. Writing non-zero + value to the field can be used to access part of sysctl value starting from + specified ``file_pos``. Not all sysctl support access with ``file_pos != + 0``, e.g. writes to numeric sysctl entries must always be at file position + ``0``. See also ``kernel.sysctl_writes_strict`` sysctl. + +See `linux/bpf.h`_ for more details on how context field can be accessed. + +3. Return code +************** + +``BPF_PROG_TYPE_CGROUP_SYSCTL`` program must return one of the following +return codes: + +* ``0`` means "reject access to sysctl"; +* ``1`` means "proceed with access". + +If program returns ``0`` user space will get ``-1`` from ``read(2)`` or +``write(2)`` and ``errno`` will be set to ``EPERM``. + +4. Helpers +********** + +Since sysctl knob is represented by a name and a value, sysctl specific BPF +helpers focus on providing access to these properties: + +* ``bpf_sysctl_get_name()`` to get sysctl name as it is visible in + ``/proc/sys`` into provided by BPF program buffer; + +* ``bpf_sysctl_get_current_value()`` to get string value currently held by + sysctl into provided by BPF program buffer. This helper is available on both + ``read(2)`` from and ``write(2)`` to sysctl; + +* ``bpf_sysctl_get_new_value()`` to get new string value currently being + written to sysctl before actual write happens. This helper can be used only + on ``ctx->write == 1``; + +* ``bpf_sysctl_set_new_value()`` to override new string value currently being + written to sysctl before actual write happens. Sysctl value will be + overridden starting from the current ``ctx->file_pos``. If the whole value + has to be overridden BPF program can set ``file_pos`` to zero before calling + to the helper. This helper can be used only on ``ctx->write == 1``. New + string value set by the helper is treated and verified by kernel same way as + an equivalent string passed by user space. + +BPF program sees sysctl value same way as user space does in proc filesystem, +i.e. as a string. Since many sysctl values represent an integer or a vector +of integers, the following helpers can be used to get numeric value from the +string: + +* ``bpf_strtol()`` to convert initial part of the string to long integer + similar to user space `strtol(3)`_; +* ``bpf_strtoul()`` to convert initial part of the string to unsigned long + integer similar to user space `strtoul(3)`_; + +See `linux/bpf.h`_ for more details on helpers described here. + +5. Examples +*********** + +See `test_sysctl_prog.c`_ for an example of BPF program in C that access +sysctl name and value, parses string value to get vector of integers and uses +the result to make decision whether to allow or deny access to sysctl. + +6. Notes +******** + +``BPF_PROG_TYPE_CGROUP_SYSCTL`` is intended to be used in **trusted** root +environment, for example to monitor sysctl usage or catch unreasonable values +an application, running as root in a separate cgroup, is trying to set. + +Since `task_dfl_cgroup(current)` is called at `sys_read` / `sys_write` time it +may return results different from that at `sys_open` time, i.e. process that +opened sysctl file in proc filesystem may differ from process that is trying +to read from / write to it and two such processes may run in different +cgroups, what means ``BPF_PROG_TYPE_CGROUP_SYSCTL`` should not be used as a +security mechanism to limit sysctl usage. + +As with any cgroup-bpf program additional care should be taken if an +application running as root in a cgroup should not be allowed to +detach/replace BPF program attached by administrator. + +.. Links +.. _linux/bpf.h: ../../include/uapi/linux/bpf.h +.. _strtol(3): http://man7.org/linux/man-pages/man3/strtol.3p.html +.. _strtoul(3): http://man7.org/linux/man-pages/man3/strtoul.3p.html +.. _test_sysctl_prog.c: + ../../tools/testing/selftests/bpf/progs/test_sysctl_prog.c From 79b1b30e4c209c2ff1ddd6559b41dba1dfc8bcd8 Mon Sep 17 00:00:00 2001 From: Magnus Karlsson Date: Thu, 18 Apr 2019 09:21:10 +0200 Subject: [PATCH 61/68] libbpf: remove compile time warning from libbpf_util.h Having a helpful compile time warning in libbpf_util.h is not a good idea since all warnings are treated as errors. Change this into a comment in the code instead. Fixes: b7e3a28019c9 ("libbpf: remove dependency on barrier.h in xsk.h") Signed-off-by: Magnus Karlsson Acked-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/libbpf_util.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/lib/bpf/libbpf_util.h b/tools/lib/bpf/libbpf_util.h index 172b707e007b..da94c4cb2e4d 100644 --- a/tools/lib/bpf/libbpf_util.h +++ b/tools/lib/bpf/libbpf_util.h @@ -46,7 +46,7 @@ do { \ # define libbpf_smp_mb() asm volatile("dmb ish" : : : "memory") # define libbpf_smp_rwmb() libbpf_smp_mb() #else -# warning Architecture missing native barrier functions in libbpf_util.h. +/* Architecture missing native barrier functions. */ # define libbpf_smp_rmb() __sync_synchronize() # define libbpf_smp_wmb() __sync_synchronize() # define libbpf_smp_mb() __sync_synchronize() From 5de35e3ae9d01796a4909fd0dd8c3660d9316a77 Mon Sep 17 00:00:00 2001 From: Wang YanQing Date: Fri, 19 Apr 2019 02:05:13 +0800 Subject: [PATCH 62/68] selftests/bpf: fix compile errors due to unsync linux/in6.h and netinet/in.h I meet below compile errors: " In file included from test_tcpnotify_kern.c:12: /usr/include/netinet/in.h:101:5: error: expected identifier IPPROTO_HOPOPTS = 0, /* IPv6 Hop-by-Hop options. */ ^ /usr/include/linux/in6.h:131:26: note: expanded from macro 'IPPROTO_HOPOPTS' ^ In file included from test_tcpnotify_kern.c:12: /usr/include/netinet/in.h:103:5: error: expected identifier IPPROTO_ROUTING = 43, /* IPv6 routing header. */ ^ /usr/include/linux/in6.h:132:26: note: expanded from macro 'IPPROTO_ROUTING' ^ In file included from test_tcpnotify_kern.c:12: /usr/include/netinet/in.h:105:5: error: expected identifier IPPROTO_FRAGMENT = 44, /* IPv6 fragmentation header. */ ^ /usr/include/linux/in6.h:133:26: note: expanded from macro 'IPPROTO_FRAGMENT' " The same compile errors are reported for test_tcpbpf_kern.c too. My environment: lsb_release -a: No LSB modules are available. Distributor ID: Ubuntu Description: Ubuntu 16.04.6 LTS Release: 16.04 Codename: xenial dpkg -l | grep libc-dev: ii libc-dev-bin 2.23-0ubuntu11 amd64 GNU C Library: Development binaries ii linux-libc-dev:amd64 4.4.0-145.171 amd64 Linux Kernel Headers for development. The reason is linux/in6.h and netinet/in.h aren't synchronous about how to handle the same definitions, IPPROTO_HOPOPTS, etc. This patch fixes the compile errors by moving to before the . Signed-off-by: Wang YanQing Acked-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c | 2 +- tools/testing/selftests/bpf/progs/test_tcpnotify_kern.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c index 74f73b33a7b0..c7c3240e0dd4 100644 --- a/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c +++ b/tools/testing/selftests/bpf/progs/test_tcpbpf_kern.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 #include #include +#include #include #include #include @@ -9,7 +10,6 @@ #include #include #include -#include #include "bpf_helpers.h" #include "bpf_endian.h" #include "test_tcpbpf.h" diff --git a/tools/testing/selftests/bpf/progs/test_tcpnotify_kern.c b/tools/testing/selftests/bpf/progs/test_tcpnotify_kern.c index edbca203ce2d..ec6db6e64c41 100644 --- a/tools/testing/selftests/bpf/progs/test_tcpnotify_kern.c +++ b/tools/testing/selftests/bpf/progs/test_tcpnotify_kern.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 #include #include +#include #include #include #include @@ -9,7 +10,6 @@ #include #include #include -#include #include "bpf_helpers.h" #include "bpf_endian.h" #include "test_tcpnotify.h" From 849f257f61ff7dde49d59c62802e5913ff7a7cbb Mon Sep 17 00:00:00 2001 From: Martin KaFai Lau Date: Thu, 18 Apr 2019 14:33:30 -0700 Subject: [PATCH 63/68] bpf: Increase MAX_NR_MAPS to 17 in test_verifier.c map_fds[16] is the last one index-ed by fixup_map_array_small. Hence, the MAX_NR_MAPS should be 17 instead. Fixes: fb2abb73e575 ("bpf, selftest: test {rd, wr}only flags and direct value access") Signed-off-by: Martin KaFai Lau Acked-by: Yonghong Song Signed-off-by: Alexei Starovoitov --- tools/testing/selftests/bpf/test_verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index 6cb6a1074fd1..ed9e894afef3 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -52,7 +52,7 @@ #define MAX_INSNS BPF_MAXINSNS #define MAX_TEST_INSNS 1000000 #define MAX_FIXUPS 8 -#define MAX_NR_MAPS 16 +#define MAX_NR_MAPS 17 #define MAX_TEST_RUNS 8 #define POINTER_VALUE 0xcafe4all #define TEST_DATA_LEN 64 From 80695946737dff4cfc1ecdefd4ebf300f132d8ee Mon Sep 17 00:00:00 2001 From: Stanislav Fomichev Date: Thu, 18 Apr 2019 16:47:52 -0700 Subject: [PATCH 64/68] bpf: move BPF_PROG_TYPE_FLOW_DISSECTOR documentation to a new common place In commit da7031491786 ("bpf: Document BPF_PROG_TYPE_CGROUP_SYSCTL") Andrey proposes to put per-prog type docs under Documentation/bpf/ Let's move flow dissector documentation there as well. Signed-off-by: Stanislav Fomichev Signed-off-by: Alexei Starovoitov --- Documentation/bpf/index.rst | 1 + .../bpf_flow_dissector.rst => bpf/prog_flow_dissector.rst} | 6 +++--- Documentation/networking/index.rst | 1 - 3 files changed, 4 insertions(+), 4 deletions(-) rename Documentation/{networking/bpf_flow_dissector.rst => bpf/prog_flow_dissector.rst} (98%) diff --git a/Documentation/bpf/index.rst b/Documentation/bpf/index.rst index dadcaa9a9f5f..d3fe4cac0c90 100644 --- a/Documentation/bpf/index.rst +++ b/Documentation/bpf/index.rst @@ -43,6 +43,7 @@ Program types :maxdepth: 1 prog_cgroup_sysctl + prog_flow_dissector .. Links: diff --git a/Documentation/networking/bpf_flow_dissector.rst b/Documentation/bpf/prog_flow_dissector.rst similarity index 98% rename from Documentation/networking/bpf_flow_dissector.rst rename to Documentation/bpf/prog_flow_dissector.rst index b375ae2ec2c4..ed343abe541e 100644 --- a/Documentation/networking/bpf_flow_dissector.rst +++ b/Documentation/bpf/prog_flow_dissector.rst @@ -1,8 +1,8 @@ .. SPDX-License-Identifier: GPL-2.0 -================== -BPF Flow Dissector -================== +============================ +BPF_PROG_TYPE_FLOW_DISSECTOR +============================ Overview ======== diff --git a/Documentation/networking/index.rst b/Documentation/networking/index.rst index 984e68f9e026..5449149be496 100644 --- a/Documentation/networking/index.rst +++ b/Documentation/networking/index.rst @@ -9,7 +9,6 @@ Contents: netdev-FAQ af_xdp batman-adv - bpf_flow_dissector can can_ucan_protocol device_drivers/freescale/dpaa2/index From 4519efa6f8ea343e43ade21b0189b0b295439202 Mon Sep 17 00:00:00 2001 From: "McCabe, Robert J" Date: Fri, 19 Apr 2019 18:37:20 -0500 Subject: [PATCH 65/68] libbpf: fix BPF_LOG_BUF_SIZE off-by-one error The BPF_PROG_LOAD condition for kernel version <= 5.1 is log->len_total > UINT_MAX >> 8 /* (16 * 1024 * 1024) - 1 */ Signed-off-by: McCabe, Robert J Signed-off-by: Alexei Starovoitov --- tools/lib/bpf/bpf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h index bc30783d1403..ff2506fe6bd2 100644 --- a/tools/lib/bpf/bpf.h +++ b/tools/lib/bpf/bpf.h @@ -92,7 +92,7 @@ struct bpf_load_program_attr { #define MAPS_RELAX_COMPAT 0x01 /* Recommend log buffer size */ -#define BPF_LOG_BUF_SIZE (16 * 1024 * 1024) /* verifier maximum in kernels <= 5.1 */ +#define BPF_LOG_BUF_SIZE (UINT32_MAX >> 8) /* verifier maximum in kernels <= 5.1 */ LIBBPF_API int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr, char *log_buf, size_t log_buf_sz); From 3b8802446d27522cd6d32178ba975cc492611f31 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Wed, 17 Apr 2019 18:27:01 -0700 Subject: [PATCH 66/68] bpf: document the verifier limits Document the verifier limits. Signed-off-by: Alexei Starovoitov Acked-by: Yonghong Song Signed-off-by: Daniel Borkmann --- Documentation/bpf/bpf_design_QA.rst | 29 +++++++++++++++++++++++++++-- 1 file changed, 27 insertions(+), 2 deletions(-) diff --git a/Documentation/bpf/bpf_design_QA.rst b/Documentation/bpf/bpf_design_QA.rst index 10453c627135..cb402c59eca5 100644 --- a/Documentation/bpf/bpf_design_QA.rst +++ b/Documentation/bpf/bpf_design_QA.rst @@ -85,8 +85,33 @@ Q: Can loops be supported in a safe way? A: It's not clear yet. BPF developers are trying to find a way to -support bounded loops where the verifier can guarantee that -the program terminates in less than 4096 instructions. +support bounded loops. + +Q: What are the verifier limits? +-------------------------------- +A: The only limit known to the user space is BPF_MAXINSNS (4096). +It's the maximum number of instructions that the unprivileged bpf +program can have. The verifier has various internal limits. +Like the maximum number of instructions that can be explored during +program analysis. Currently, that limit is set to 1 million. +Which essentially means that the largest program can consist +of 1 million NOP instructions. There is a limit to the maximum number +of subsequent branches, a limit to the number of nested bpf-to-bpf +calls, a limit to the number of the verifier states per instruction, +a limit to the number of maps used by the program. +All these limits can be hit with a sufficiently complex program. +There are also non-numerical limits that can cause the program +to be rejected. The verifier used to recognize only pointer + constant +expressions. Now it can recognize pointer + bounded_register. +bpf_lookup_map_elem(key) had a requirement that 'key' must be +a pointer to the stack. Now, 'key' can be a pointer to map value. +The verifier is steadily getting 'smarter'. The limits are +being removed. The only way to know that the program is going to +be accepted by the verifier is to try to load it. +The bpf development process guarantees that the future kernel +versions will accept all bpf programs that were accepted by +the earlier versions. + Instruction level questions --------------------------- From 7df737e991069d75eec1ded1c8b37e81b8c54df9 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Fri, 19 Apr 2019 07:44:54 -0700 Subject: [PATCH 67/68] bpf: remove global variables Move three global variables protected by bpf_verifier_lock into 'struct bpf_verifier_env' to allow parallel verification. Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- include/linux/bpf_verifier.h | 5 +++++ kernel/bpf/verifier.c | 25 +++++++++++++------------ 2 files changed, 18 insertions(+), 12 deletions(-) diff --git a/include/linux/bpf_verifier.h b/include/linux/bpf_verifier.h index b3ab61fe1932..1305ccbd8fe6 100644 --- a/include/linux/bpf_verifier.h +++ b/include/linux/bpf_verifier.h @@ -295,6 +295,11 @@ struct bpf_verifier_env { const struct bpf_line_info *prev_linfo; struct bpf_verifier_log log; struct bpf_subprog_info subprog_info[BPF_MAX_SUBPROGS + 1]; + struct { + int *insn_state; + int *insn_stack; + int cur_stack; + } cfg; u32 subprog_cnt; /* number of instructions analyzed by the verifier */ u32 insn_processed; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index db301e9b5295..5f0eb5bd5589 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -5369,10 +5369,6 @@ enum { #define STATE_LIST_MARK ((struct bpf_verifier_state_list *) -1L) -static int *insn_stack; /* stack of insns to process */ -static int cur_stack; /* current stack index */ -static int *insn_state; - /* t, w, e - match pseudo-code above: * t - index of current instruction * w - next instruction @@ -5380,6 +5376,9 @@ static int *insn_state; */ static int push_insn(int t, int w, int e, struct bpf_verifier_env *env) { + int *insn_stack = env->cfg.insn_stack; + int *insn_state = env->cfg.insn_state; + if (e == FALLTHROUGH && insn_state[t] >= (DISCOVERED | FALLTHROUGH)) return 0; @@ -5400,9 +5399,9 @@ static int push_insn(int t, int w, int e, struct bpf_verifier_env *env) /* tree-edge */ insn_state[t] = DISCOVERED | e; insn_state[w] = DISCOVERED; - if (cur_stack >= env->prog->len) + if (env->cfg.cur_stack >= env->prog->len) return -E2BIG; - insn_stack[cur_stack++] = w; + insn_stack[env->cfg.cur_stack++] = w; return 1; } else if ((insn_state[w] & 0xF0) == DISCOVERED) { verbose_linfo(env, t, "%d: ", t); @@ -5426,14 +5425,15 @@ static int check_cfg(struct bpf_verifier_env *env) { struct bpf_insn *insns = env->prog->insnsi; int insn_cnt = env->prog->len; + int *insn_stack, *insn_state; int ret = 0; int i, t; - insn_state = kvcalloc(insn_cnt, sizeof(int), GFP_KERNEL); + insn_state = env->cfg.insn_state = kvcalloc(insn_cnt, sizeof(int), GFP_KERNEL); if (!insn_state) return -ENOMEM; - insn_stack = kvcalloc(insn_cnt, sizeof(int), GFP_KERNEL); + insn_stack = env->cfg.insn_stack = kvcalloc(insn_cnt, sizeof(int), GFP_KERNEL); if (!insn_stack) { kvfree(insn_state); return -ENOMEM; @@ -5441,12 +5441,12 @@ static int check_cfg(struct bpf_verifier_env *env) insn_state[0] = DISCOVERED; /* mark 1st insn as discovered */ insn_stack[0] = 0; /* 0 is the first instruction */ - cur_stack = 1; + env->cfg.cur_stack = 1; peek_stack: - if (cur_stack == 0) + if (env->cfg.cur_stack == 0) goto check_state; - t = insn_stack[cur_stack - 1]; + t = insn_stack[env->cfg.cur_stack - 1]; if (BPF_CLASS(insns[t].code) == BPF_JMP || BPF_CLASS(insns[t].code) == BPF_JMP32) { @@ -5515,7 +5515,7 @@ static int check_cfg(struct bpf_verifier_env *env) mark_explored: insn_state[t] = EXPLORED; - if (cur_stack-- <= 0) { + if (env->cfg.cur_stack-- <= 0) { verbose(env, "pop stack internal bug\n"); ret = -EFAULT; goto err_free; @@ -5535,6 +5535,7 @@ static int check_cfg(struct bpf_verifier_env *env) err_free: kvfree(insn_state); kvfree(insn_stack); + env->cfg.insn_state = env->cfg.insn_stack = NULL; return ret; } From 45a73c17bfb92c3ceebedc80a750ef2c2931c26b Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Fri, 19 Apr 2019 07:44:55 -0700 Subject: [PATCH 68/68] bpf: drop bpf_verifier_lock Drop bpf_verifier_lock for root to avoid being DoS-ed by unprivileged. The BPF verifier is now fully parallel. All unpriv users are still serialized by bpf_verifier_lock to avoid exhausting kernel memory by running N parallel verifications. Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- kernel/bpf/verifier.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 5f0eb5bd5589..423f242a5efb 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -8132,9 +8132,11 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, env->insn_aux_data[i].orig_idx = i; env->prog = *prog; env->ops = bpf_verifier_ops[env->prog->type]; + is_priv = capable(CAP_SYS_ADMIN); /* grab the mutex to protect few globals used by verifier */ - mutex_lock(&bpf_verifier_lock); + if (!is_priv) + mutex_lock(&bpf_verifier_lock); if (attr->log_level || attr->log_buf || attr->log_size) { /* user requested verbose verifier output @@ -8157,7 +8159,6 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, if (attr->prog_flags & BPF_F_ANY_ALIGNMENT) env->strict_alignment = false; - is_priv = capable(CAP_SYS_ADMIN); env->allow_ptr_leaks = is_priv; ret = replace_map_fd_with_map_ptr(env); @@ -8270,7 +8271,8 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr, release_maps(env); *prog = env->prog; err_unlock: - mutex_unlock(&bpf_verifier_lock); + if (!is_priv) + mutex_unlock(&bpf_verifier_lock); vfree(env->insn_aux_data); err_free_env: kfree(env);