bpf: move global verifier log into verifier environment
authorJakub Kicinski <jakub.kicinski@netronome.com>
Mon, 9 Oct 2017 17:30:11 +0000 (10:30 -0700)
committerBen Hutchings <ben@decadent.org.uk>
Fri, 22 Dec 2017 14:12:23 +0000 (14:12 +0000)
The biggest piece of global state protected by the verifier lock
is the verifier_log.  Move that log to struct bpf_verifier_env.
struct bpf_verifier_env has to be passed now to all invocations
of verbose().

Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 4.14]

Gbp-Pq: Topic bugfix/all
Gbp-Pq: Name bpf-move-global-verifier-log-into-verifier-environme.patch

include/linux/bpf_verifier.h
kernel/bpf/verifier.c

index 163541ba70d9792ef98528ec3bb5c58c44c8a1b0..5ddb9a626a51e2b8ac06ebdb7eb59ce0c25a7750 100644 (file)
@@ -152,6 +152,8 @@ struct bpf_verifier_env {
        bool allow_ptr_leaks;
        bool seen_direct_write;
        struct bpf_insn_aux_data *insn_aux_data; /* array of per-insn state */
+
+       struct bpf_verifer_log log;
 };
 
 int bpf_analyzer(struct bpf_prog *prog, const struct bpf_ext_analyzer_ops *ops,
index e5ecc3678f991194c762302efe62b27a891b8ab1..470adb8593ce6f57191f96eece5f01c2fa364282 100644 (file)
@@ -153,20 +153,16 @@ struct bpf_call_arg_meta {
        int access_size;
 };
 
-/* verbose verifier prints what it's seeing
- * bpf_check() is called under lock, so no race to access these global vars
- */
-static struct bpf_verifer_log verifier_log;
-
 static DEFINE_MUTEX(bpf_verifier_lock);
 
 /* log_level controls verbosity level of eBPF verifier.
  * verbose() is used to dump the verification trace to the log, so the user
  * can figure out what's wrong with the program
  */
-static __printf(1, 2) void verbose(const char *fmt, ...)
+static __printf(2, 3) void verbose(struct bpf_verifier_env *env,
+                                  const char *fmt, ...)
 {
-       struct bpf_verifer_log *log = &verifier_log;
+       struct bpf_verifer_log *log = &env->log;
        va_list args;
 
        if (!log->level || bpf_verifier_log_full(log))
@@ -207,7 +203,8 @@ static const char *func_id_name(int id)
                return "unknown";
 }
 
-static void print_verifier_state(struct bpf_verifier_state *state)
+static void print_verifier_state(struct bpf_verifier_env *env,
+                                struct bpf_verifier_state *state)
 {
        struct bpf_reg_state *reg;
        enum bpf_reg_type t;
@@ -218,21 +215,21 @@ static void print_verifier_state(struct bpf_verifier_state *state)
                t = reg->type;
                if (t == NOT_INIT)
                        continue;
-               verbose(" R%d=%s", i, reg_type_str[t]);
+               verbose(env, " R%d=%s", i, reg_type_str[t]);
                if ((t == SCALAR_VALUE || t == PTR_TO_STACK) &&
                    tnum_is_const(reg->var_off)) {
                        /* reg->off should be 0 for SCALAR_VALUE */
-                       verbose("%lld", reg->var_off.value + reg->off);
+                       verbose(env, "%lld", reg->var_off.value + reg->off);
                } else {
-                       verbose("(id=%d", reg->id);
+                       verbose(env, "(id=%d", reg->id);
                        if (t != SCALAR_VALUE)
-                               verbose(",off=%d", reg->off);
+                               verbose(env, ",off=%d", reg->off);
                        if (t == PTR_TO_PACKET)
-                               verbose(",r=%d", reg->range);
+                               verbose(env, ",r=%d", reg->range);
                        else if (t == CONST_PTR_TO_MAP ||
                                 t == PTR_TO_MAP_VALUE ||
                                 t == PTR_TO_MAP_VALUE_OR_NULL)
-                               verbose(",ks=%d,vs=%d",
+                               verbose(env, ",ks=%d,vs=%d",
                                        reg->map_ptr->key_size,
                                        reg->map_ptr->value_size);
                        if (tnum_is_const(reg->var_off)) {
@@ -240,38 +237,38 @@ static void print_verifier_state(struct bpf_verifier_state *state)
                                 * could be a pointer whose offset is too big
                                 * for reg->off
                                 */
-                               verbose(",imm=%llx", reg->var_off.value);
+                               verbose(env, ",imm=%llx", reg->var_off.value);
                        } else {
                                if (reg->smin_value != reg->umin_value &&
                                    reg->smin_value != S64_MIN)
-                                       verbose(",smin_value=%lld",
+                                       verbose(env, ",smin_value=%lld",
                                                (long long)reg->smin_value);
                                if (reg->smax_value != reg->umax_value &&
                                    reg->smax_value != S64_MAX)
-                                       verbose(",smax_value=%lld",
+                                       verbose(env, ",smax_value=%lld",
                                                (long long)reg->smax_value);
                                if (reg->umin_value != 0)
-                                       verbose(",umin_value=%llu",
+                                       verbose(env, ",umin_value=%llu",
                                                (unsigned long long)reg->umin_value);
                                if (reg->umax_value != U64_MAX)
-                                       verbose(",umax_value=%llu",
+                                       verbose(env, ",umax_value=%llu",
                                                (unsigned long long)reg->umax_value);
                                if (!tnum_is_unknown(reg->var_off)) {
                                        char tn_buf[48];
 
                                        tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
-                                       verbose(",var_off=%s", tn_buf);
+                                       verbose(env, ",var_off=%s", tn_buf);
                                }
                        }
-                       verbose(")");
+                       verbose(env, ")");
                }
        }
        for (i = 0; i < MAX_BPF_STACK; i += BPF_REG_SIZE) {
                if (state->stack_slot_type[i] == STACK_SPILL)
-                       verbose(" fp%d=%s", -MAX_BPF_STACK + i,
+                       verbose(env, " fp%d=%s", -MAX_BPF_STACK + i,
                                reg_type_str[state->spilled_regs[i / BPF_REG_SIZE].type]);
        }
-       verbose("\n");
+       verbose(env, "\n");
 }
 
 static const char *const bpf_class_string[] = {
@@ -326,21 +323,21 @@ static const char *const bpf_jmp_string[16] = {
        [BPF_EXIT >> 4] = "exit",
 };
 
-static void print_bpf_insn(const struct bpf_verifier_env *env,
+static void print_bpf_insn(struct bpf_verifier_env *env,
                           const struct bpf_insn *insn)
 {
        u8 class = BPF_CLASS(insn->code);
 
        if (class == BPF_ALU || class == BPF_ALU64) {
                if (BPF_SRC(insn->code) == BPF_X)
-                       verbose("(%02x) %sr%d %s %sr%d\n",
+                       verbose(env, "(%02x) %sr%d %s %sr%d\n",
                                insn->code, class == BPF_ALU ? "(u32) " : "",
                                insn->dst_reg,
                                bpf_alu_string[BPF_OP(insn->code) >> 4],
                                class == BPF_ALU ? "(u32) " : "",
                                insn->src_reg);
                else
-                       verbose("(%02x) %sr%d %s %s%d\n",
+                       verbose(env, "(%02x) %sr%d %s %s%d\n",
                                insn->code, class == BPF_ALU ? "(u32) " : "",
                                insn->dst_reg,
                                bpf_alu_string[BPF_OP(insn->code) >> 4],
@@ -348,46 +345,46 @@ static void print_bpf_insn(const struct bpf_verifier_env *env,
                                insn->imm);
        } else if (class == BPF_STX) {
                if (BPF_MODE(insn->code) == BPF_MEM)
-                       verbose("(%02x) *(%s *)(r%d %+d) = r%d\n",
+                       verbose(env, "(%02x) *(%s *)(r%d %+d) = r%d\n",
                                insn->code,
                                bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                                insn->dst_reg,
                                insn->off, insn->src_reg);
                else if (BPF_MODE(insn->code) == BPF_XADD)
-                       verbose("(%02x) lock *(%s *)(r%d %+d) += r%d\n",
+                       verbose(env, "(%02x) lock *(%s *)(r%d %+d) += r%d\n",
                                insn->code,
                                bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                                insn->dst_reg, insn->off,
                                insn->src_reg);
                else
-                       verbose("BUG_%02x\n", insn->code);
+                       verbose(env, "BUG_%02x\n", insn->code);
        } else if (class == BPF_ST) {
                if (BPF_MODE(insn->code) != BPF_MEM) {
-                       verbose("BUG_st_%02x\n", insn->code);
+                       verbose(env, "BUG_st_%02x\n", insn->code);
                        return;
                }
-               verbose("(%02x) *(%s *)(r%d %+d) = %d\n",
+               verbose(env, "(%02x) *(%s *)(r%d %+d) = %d\n",
                        insn->code,
                        bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                        insn->dst_reg,
                        insn->off, insn->imm);
        } else if (class == BPF_LDX) {
                if (BPF_MODE(insn->code) != BPF_MEM) {
-                       verbose("BUG_ldx_%02x\n", insn->code);
+                       verbose(env, "BUG_ldx_%02x\n", insn->code);
                        return;
                }
-               verbose("(%02x) r%d = *(%s *)(r%d %+d)\n",
+               verbose(env, "(%02x) r%d = *(%s *)(r%d %+d)\n",
                        insn->code, insn->dst_reg,
                        bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                        insn->src_reg, insn->off);
        } else if (class == BPF_LD) {
                if (BPF_MODE(insn->code) == BPF_ABS) {
-                       verbose("(%02x) r0 = *(%s *)skb[%d]\n",
+                       verbose(env, "(%02x) r0 = *(%s *)skb[%d]\n",
                                insn->code,
                                bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                                insn->imm);
                } else if (BPF_MODE(insn->code) == BPF_IND) {
-                       verbose("(%02x) r0 = *(%s *)skb[r%d + %d]\n",
+                       verbose(env, "(%02x) r0 = *(%s *)skb[r%d + %d]\n",
                                insn->code,
                                bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
                                insn->src_reg, insn->imm);
@@ -402,36 +399,37 @@ static void print_bpf_insn(const struct bpf_verifier_env *env,
                        if (map_ptr && !env->allow_ptr_leaks)
                                imm = 0;
 
-                       verbose("(%02x) r%d = 0x%llx\n", insn->code,
+                       verbose(env, "(%02x) r%d = 0x%llx\n", insn->code,
                                insn->dst_reg, (unsigned long long)imm);
                } else {
-                       verbose("BUG_ld_%02x\n", insn->code);
+                       verbose(env, "BUG_ld_%02x\n", insn->code);
                        return;
                }
        } else if (class == BPF_JMP) {
                u8 opcode = BPF_OP(insn->code);
 
                if (opcode == BPF_CALL) {
-                       verbose("(%02x) call %s#%d\n", insn->code,
+                       verbose(env, "(%02x) call %s#%d\n", insn->code,
                                func_id_name(insn->imm), insn->imm);
                } else if (insn->code == (BPF_JMP | BPF_JA)) {
-                       verbose("(%02x) goto pc%+d\n",
+                       verbose(env, "(%02x) goto pc%+d\n",
                                insn->code, insn->off);
                } else if (insn->code == (BPF_JMP | BPF_EXIT)) {
-                       verbose("(%02x) exit\n", insn->code);
+                       verbose(env, "(%02x) exit\n", insn->code);
                } else if (BPF_SRC(insn->code) == BPF_X) {
-                       verbose("(%02x) if r%d %s r%d goto pc%+d\n",
+                       verbose(env, "(%02x) if r%d %s r%d goto pc%+d\n",
                                insn->code, insn->dst_reg,
                                bpf_jmp_string[BPF_OP(insn->code) >> 4],
                                insn->src_reg, insn->off);
                } else {
-                       verbose("(%02x) if r%d %s 0x%x goto pc%+d\n",
+                       verbose(env, "(%02x) if r%d %s 0x%x goto pc%+d\n",
                                insn->code, insn->dst_reg,
                                bpf_jmp_string[BPF_OP(insn->code) >> 4],
                                insn->imm, insn->off);
                }
        } else {
-               verbose("(%02x) %s\n", insn->code, bpf_class_string[class]);
+               verbose(env, "(%02x) %s\n",
+                       insn->code, bpf_class_string[class]);
        }
 }
 
@@ -470,7 +468,7 @@ static struct bpf_verifier_state *push_stack(struct bpf_verifier_env *env,
        env->head = elem;
        env->stack_size++;
        if (env->stack_size > BPF_COMPLEXITY_LIMIT_STACK) {
-               verbose("BPF program is too complex\n");
+               verbose(env, "BPF program is too complex\n");
                goto err;
        }
        return &elem->st;
@@ -508,10 +506,11 @@ static void __mark_reg_known_zero(struct bpf_reg_state *reg)
        __mark_reg_known(reg, 0);
 }
 
-static void mark_reg_known_zero(struct bpf_reg_state *regs, u32 regno)
+static void mark_reg_known_zero(struct bpf_verifier_env *env,
+                               struct bpf_reg_state *regs, u32 regno)
 {
        if (WARN_ON(regno >= MAX_BPF_REG)) {
-               verbose("mark_reg_known_zero(regs, %u)\n", regno);
+               verbose(env, "mark_reg_known_zero(regs, %u)\n", regno);
                /* Something bad happened, let's kill all regs */
                for (regno = 0; regno < MAX_BPF_REG; regno++)
                        __mark_reg_not_init(regs + regno);
@@ -596,10 +595,11 @@ static void __mark_reg_unknown(struct bpf_reg_state *reg)
        __mark_reg_unbounded(reg);
 }
 
-static void mark_reg_unknown(struct bpf_reg_state *regs, u32 regno)
+static void mark_reg_unknown(struct bpf_verifier_env *env,
+                            struct bpf_reg_state *regs, u32 regno)
 {
        if (WARN_ON(regno >= MAX_BPF_REG)) {
-               verbose("mark_reg_unknown(regs, %u)\n", regno);
+               verbose(env, "mark_reg_unknown(regs, %u)\n", regno);
                /* Something bad happened, let's kill all regs */
                for (regno = 0; regno < MAX_BPF_REG; regno++)
                        __mark_reg_not_init(regs + regno);
@@ -614,10 +614,11 @@ static void __mark_reg_not_init(struct bpf_reg_state *reg)
        reg->type = NOT_INIT;
 }
 
-static void mark_reg_not_init(struct bpf_reg_state *regs, u32 regno)
+static void mark_reg_not_init(struct bpf_verifier_env *env,
+                             struct bpf_reg_state *regs, u32 regno)
 {
        if (WARN_ON(regno >= MAX_BPF_REG)) {
-               verbose("mark_reg_not_init(regs, %u)\n", regno);
+               verbose(env, "mark_reg_not_init(regs, %u)\n", regno);
                /* Something bad happened, let's kill all regs */
                for (regno = 0; regno < MAX_BPF_REG; regno++)
                        __mark_reg_not_init(regs + regno);
@@ -626,22 +627,23 @@ static void mark_reg_not_init(struct bpf_reg_state *regs, u32 regno)
        __mark_reg_not_init(regs + regno);
 }
 
-static void init_reg_state(struct bpf_reg_state *regs)
+static void init_reg_state(struct bpf_verifier_env *env,
+                          struct bpf_reg_state *regs)
 {
        int i;
 
        for (i = 0; i < MAX_BPF_REG; i++) {
-               mark_reg_not_init(regs, i);
+               mark_reg_not_init(env, regs, i);
                regs[i].live = REG_LIVE_NONE;
        }
 
        /* frame pointer */
        regs[BPF_REG_FP].type = PTR_TO_STACK;
-       mark_reg_known_zero(regs, BPF_REG_FP);
+       mark_reg_known_zero(env, regs, BPF_REG_FP);
 
        /* 1st arg to a function */
        regs[BPF_REG_1].type = PTR_TO_CTX;
-       mark_reg_known_zero(regs, BPF_REG_1);
+       mark_reg_known_zero(env, regs, BPF_REG_1);
 }
 
 enum reg_arg_type {
@@ -675,26 +677,26 @@ static int check_reg_arg(struct bpf_verifier_env *env, u32 regno,
        struct bpf_reg_state *regs = env->cur_state.regs;
 
        if (regno >= MAX_BPF_REG) {
-               verbose("R%d is invalid\n", regno);
+               verbose(env, "R%d is invalid\n", regno);
                return -EINVAL;
        }
 
        if (t == SRC_OP) {
                /* check whether register used as source operand can be read */
                if (regs[regno].type == NOT_INIT) {
-                       verbose("R%d !read_ok\n", regno);
+                       verbose(env, "R%d !read_ok\n", regno);
                        return -EACCES;
                }
                mark_reg_read(&env->cur_state, regno);
        } else {
                /* check whether register used as dest operand can be written to */
                if (regno == BPF_REG_FP) {
-                       verbose("frame pointer is read only\n");
+                       verbose(env, "frame pointer is read only\n");
                        return -EACCES;
                }
                regs[regno].live |= REG_LIVE_WRITTEN;
                if (t == DST_OP)
-                       mark_reg_unknown(regs, regno);
+                       mark_reg_unknown(env, regs, regno);
        }
        return 0;
 }
@@ -718,7 +720,8 @@ static bool is_spillable_regtype(enum bpf_reg_type type)
 /* check_stack_read/write functions track spill/fill of registers,
  * stack boundary and alignment are checked in check_mem_access()
  */
-static int check_stack_write(struct bpf_verifier_state *state, int off,
+static int check_stack_write(struct bpf_verifier_env *env,
+                            struct bpf_verifier_state *state, int off,
                             int size, int value_regno)
 {
        int i, spi = (MAX_BPF_STACK + off) / BPF_REG_SIZE;
@@ -731,7 +734,7 @@ static int check_stack_write(struct bpf_verifier_state *state, int off,
 
                /* register containing pointer is being spilled into stack */
                if (size != BPF_REG_SIZE) {
-                       verbose("invalid size of register spill\n");
+                       verbose(env, "invalid size of register spill\n");
                        return -EACCES;
                }
 
@@ -766,7 +769,8 @@ static void mark_stack_slot_read(const struct bpf_verifier_state *state, int slo
        }
 }
 
-static int check_stack_read(struct bpf_verifier_state *state, int off, int size,
+static int check_stack_read(struct bpf_verifier_env *env,
+                           struct bpf_verifier_state *state, int off, int size,
                            int value_regno)
 {
        u8 *slot_type;
@@ -776,12 +780,12 @@ static int check_stack_read(struct bpf_verifier_state *state, int off, int size,
 
        if (slot_type[0] == STACK_SPILL) {
                if (size != BPF_REG_SIZE) {
-                       verbose("invalid size of register spill\n");
+                       verbose(env, "invalid size of register spill\n");
                        return -EACCES;
                }
                for (i = 1; i < BPF_REG_SIZE; i++) {
                        if (slot_type[i] != STACK_SPILL) {
-                               verbose("corrupted spill memory\n");
+                               verbose(env, "corrupted spill memory\n");
                                return -EACCES;
                        }
                }
@@ -797,14 +801,14 @@ static int check_stack_read(struct bpf_verifier_state *state, int off, int size,
        } else {
                for (i = 0; i < size; i++) {
                        if (slot_type[i] != STACK_MISC) {
-                               verbose("invalid read from stack off %d+%d size %d\n",
+                               verbose(env, "invalid read from stack off %d+%d size %d\n",
                                        off, i, size);
                                return -EACCES;
                        }
                }
                if (value_regno >= 0)
                        /* have read misc data from the stack */
-                       mark_reg_unknown(state->regs, value_regno);
+                       mark_reg_unknown(env, state->regs, value_regno);
                return 0;
        }
 }
@@ -816,7 +820,7 @@ static int __check_map_access(struct bpf_verifier_env *env, u32 regno, int off,
        struct bpf_map *map = env->cur_state.regs[regno].map_ptr;
 
        if (off < 0 || size <= 0 || off + size > map->value_size) {
-               verbose("invalid access to map value, value_size=%d off=%d size=%d\n",
+               verbose(env, "invalid access to map value, value_size=%d off=%d size=%d\n",
                        map->value_size, off, size);
                return -EACCES;
        }
@@ -835,8 +839,8 @@ static int check_map_access(struct bpf_verifier_env *env, u32 regno,
         * need to try adding each of min_value and max_value to off
         * to make sure our theoretical access will be safe.
         */
-       if (verifier_log.level)
-               print_verifier_state(state);
+       if (env->log.level)
+               print_verifier_state(env, state);
        /* The minimum value is only important with signed
         * comparisons where we can't assume the floor of a
         * value is 0.  If we are using signed variables for our
@@ -844,13 +848,14 @@ static int check_map_access(struct bpf_verifier_env *env, u32 regno,
         * will have a set floor within our range.
         */
        if (reg->smin_value < 0) {
-               verbose("R%d min value is negative, either use unsigned index or do a if (index >=0) check.\n",
+               verbose(env, "R%d min value is negative, either use unsigned index or do a if (index >=0) check.\n",
                        regno);
                return -EACCES;
        }
        err = __check_map_access(env, regno, reg->smin_value + off, size);
        if (err) {
-               verbose("R%d min value is outside of the array range\n", regno);
+               verbose(env, "R%d min value is outside of the array range\n",
+                       regno);
                return err;
        }
 
@@ -859,13 +864,14 @@ static int check_map_access(struct bpf_verifier_env *env, u32 regno,
         * If reg->umax_value + off could overflow, treat that as unbounded too.
         */
        if (reg->umax_value >= BPF_MAX_VAR_OFF) {
-               verbose("R%d unbounded memory access, make sure to bounds check any array access into a map\n",
+               verbose(env, "R%d unbounded memory access, make sure to bounds check any array access into a map\n",
                        regno);
                return -EACCES;
        }
        err = __check_map_access(env, regno, reg->umax_value + off, size);
        if (err)
-               verbose("R%d max value is outside of the array range\n", regno);
+               verbose(env, "R%d max value is outside of the array range\n",
+                       regno);
        return err;
 }
 
@@ -904,7 +910,7 @@ static int __check_packet_access(struct bpf_verifier_env *env, u32 regno,
        struct bpf_reg_state *reg = &regs[regno];
 
        if (off < 0 || size <= 0 || (u64)off + size > reg->range) {
-               verbose("invalid access to packet, off=%d size=%d, R%d(id=%d,off=%d,r=%d)\n",
+               verbose(env, "invalid access to packet, off=%d size=%d, R%d(id=%d,off=%d,r=%d)\n",
                        off, size, regno, reg->id, reg->off, reg->range);
                return -EACCES;
        }
@@ -927,13 +933,13 @@ static int check_packet_access(struct bpf_verifier_env *env, u32 regno, int off,
         * detail to prove they're safe.
         */
        if (reg->smin_value < 0) {
-               verbose("R%d min value is negative, either use unsigned index or do a if (index >=0) check.\n",
+               verbose(env, "R%d min value is negative, either use unsigned index or do a if (index >=0) check.\n",
                        regno);
                return -EACCES;
        }
        err = __check_packet_access(env, regno, off, size);
        if (err) {
-               verbose("R%d offset is outside of the packet\n", regno);
+               verbose(env, "R%d offset is outside of the packet\n", regno);
                return err;
        }
        return err;
@@ -969,7 +975,7 @@ static int check_ctx_access(struct bpf_verifier_env *env, int insn_idx, int off,
                return 0;
        }
 
-       verbose("invalid bpf_context access off=%d size=%d\n", off, size);
+       verbose(env, "invalid bpf_context access off=%d size=%d\n", off, size);
        return -EACCES;
 }
 
@@ -987,7 +993,8 @@ static bool is_pointer_value(struct bpf_verifier_env *env, int regno)
        return __is_pointer_value(env->allow_ptr_leaks, &env->cur_state.regs[regno]);
 }
 
-static int check_pkt_ptr_alignment(const struct bpf_reg_state *reg,
+static int check_pkt_ptr_alignment(struct bpf_verifier_env *env,
+                                  const struct bpf_reg_state *reg,
                                   int off, int size, bool strict)
 {
        struct tnum reg_off;
@@ -1012,7 +1019,8 @@ static int check_pkt_ptr_alignment(const struct bpf_reg_state *reg,
                char tn_buf[48];
 
                tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
-               verbose("misaligned packet access off %d+%s+%d+%d size %d\n",
+               verbose(env,
+                       "misaligned packet access off %d+%s+%d+%d size %d\n",
                        ip_align, tn_buf, reg->off, off, size);
                return -EACCES;
        }
@@ -1020,7 +1028,8 @@ static int check_pkt_ptr_alignment(const struct bpf_reg_state *reg,
        return 0;
 }
 
-static int check_generic_ptr_alignment(const struct bpf_reg_state *reg,
+static int check_generic_ptr_alignment(struct bpf_verifier_env *env,
+                                      const struct bpf_reg_state *reg,
                                       const char *pointer_desc,
                                       int off, int size, bool strict)
 {
@@ -1035,7 +1044,7 @@ static int check_generic_ptr_alignment(const struct bpf_reg_state *reg,
                char tn_buf[48];
 
                tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
-               verbose("misaligned %saccess off %s+%d+%d size %d\n",
+               verbose(env, "misaligned %saccess off %s+%d+%d size %d\n",
                        pointer_desc, tn_buf, reg->off, off, size);
                return -EACCES;
        }
@@ -1053,7 +1062,7 @@ static int check_ptr_alignment(struct bpf_verifier_env *env,
        switch (reg->type) {
        case PTR_TO_PACKET:
                /* special case, because of NET_IP_ALIGN */
-               return check_pkt_ptr_alignment(reg, off, size, strict);
+               return check_pkt_ptr_alignment(env, reg, off, size, strict);
        case PTR_TO_MAP_VALUE:
                pointer_desc = "value ";
                break;
@@ -1066,7 +1075,8 @@ static int check_ptr_alignment(struct bpf_verifier_env *env,
        default:
                break;
        }
-       return check_generic_ptr_alignment(reg, pointer_desc, off, size, strict);
+       return check_generic_ptr_alignment(env, reg, pointer_desc, off, size,
+                                          strict);
 }
 
 /* check whether memory at (regno + off) is accessible for t = (read | write)
@@ -1098,27 +1108,27 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
        if (reg->type == PTR_TO_MAP_VALUE) {
                if (t == BPF_WRITE && value_regno >= 0 &&
                    is_pointer_value(env, value_regno)) {
-                       verbose("R%d leaks addr into map\n", value_regno);
+                       verbose(env, "R%d leaks addr into map\n", value_regno);
                        return -EACCES;
                }
 
                err = check_map_access(env, regno, off, size);
                if (!err && t == BPF_READ && value_regno >= 0)
-                       mark_reg_unknown(state->regs, value_regno);
+                       mark_reg_unknown(env, state->regs, value_regno);
 
        } else if (reg->type == PTR_TO_CTX) {
                enum bpf_reg_type reg_type = SCALAR_VALUE;
 
                if (t == BPF_WRITE && value_regno >= 0 &&
                    is_pointer_value(env, value_regno)) {
-                       verbose("R%d leaks addr into ctx\n", value_regno);
+                       verbose(env, "R%d leaks addr into ctx\n", value_regno);
                        return -EACCES;
                }
                /* ctx accesses must be at a fixed offset, so that we can
                 * determine what type of data were returned.
                 */
                if (reg->off) {
-                       verbose("dereference of modified ctx ptr R%d off=%d+%d, ctx+const is allowed, ctx+const+const is not\n",
+                       verbose(env, "dereference of modified ctx ptr R%d off=%d+%d, ctx+const is allowed, ctx+const+const is not\n",
                                regno, reg->off, off - reg->off);
                        return -EACCES;
                }
@@ -1126,7 +1136,8 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
                        char tn_buf[48];
 
                        tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
-                       verbose("variable ctx access var_off=%s off=%d size=%d",
+                       verbose(env,
+                               "variable ctx access var_off=%s off=%d size=%d",
                                tn_buf, off, size);
                        return -EACCES;
                }
@@ -1137,9 +1148,10 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
                         * the offset is zero.
                         */
                        if (reg_type == SCALAR_VALUE)
-                               mark_reg_unknown(state->regs, value_regno);
+                               mark_reg_unknown(env, state->regs, value_regno);
                        else
-                               mark_reg_known_zero(state->regs, value_regno);
+                               mark_reg_known_zero(env, state->regs,
+                                                   value_regno);
                        state->regs[value_regno].id = 0;
                        state->regs[value_regno].off = 0;
                        state->regs[value_regno].range = 0;
@@ -1155,13 +1167,14 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
                        char tn_buf[48];
 
                        tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
-                       verbose("variable stack access var_off=%s off=%d size=%d",
+                       verbose(env, "variable stack access var_off=%s off=%d size=%d",
                                tn_buf, off, size);
                        return -EACCES;
                }
                off += reg->var_off.value;
                if (off >= 0 || off < -MAX_BPF_STACK) {
-                       verbose("invalid stack off=%d size=%d\n", off, size);
+                       verbose(env, "invalid stack off=%d size=%d\n", off,
+                               size);
                        return -EACCES;
                }
 
@@ -1172,28 +1185,31 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
                        if (!env->allow_ptr_leaks &&
                            state->stack_slot_type[MAX_BPF_STACK + off] == STACK_SPILL &&
                            size != BPF_REG_SIZE) {
-                               verbose("attempt to corrupt spilled pointer on stack\n");
+                               verbose(env, "attempt to corrupt spilled pointer on stack\n");
                                return -EACCES;
                        }
-                       err = check_stack_write(state, off, size, value_regno);
+                       err = check_stack_write(env, state, off, size,
+                                               value_regno);
                } else {
-                       err = check_stack_read(state, off, size, value_regno);
+                       err = check_stack_read(env, state, off, size,
+                                              value_regno);
                }
        } else if (reg->type == PTR_TO_PACKET) {
                if (t == BPF_WRITE && !may_access_direct_pkt_data(env, NULL, t)) {
-                       verbose("cannot write into packet\n");
+                       verbose(env, "cannot write into packet\n");
                        return -EACCES;
                }
                if (t == BPF_WRITE && value_regno >= 0 &&
                    is_pointer_value(env, value_regno)) {
-                       verbose("R%d leaks addr into packet\n", value_regno);
+                       verbose(env, "R%d leaks addr into packet\n",
+                               value_regno);
                        return -EACCES;
                }
                err = check_packet_access(env, regno, off, size);
                if (!err && t == BPF_READ && value_regno >= 0)
-                       mark_reg_unknown(state->regs, value_regno);
+                       mark_reg_unknown(env, state->regs, value_regno);
        } else {
-               verbose("R%d invalid mem access '%s'\n",
+               verbose(env, "R%d invalid mem access '%s'\n",
                        regno, reg_type_str[reg->type]);
                return -EACCES;
        }
@@ -1214,7 +1230,7 @@ static int check_xadd(struct bpf_verifier_env *env, int insn_idx, struct bpf_ins
 
        if ((BPF_SIZE(insn->code) != BPF_W && BPF_SIZE(insn->code) != BPF_DW) ||
            insn->imm != 0) {
-               verbose("BPF_XADD uses reserved fields\n");
+               verbose(env, "BPF_XADD uses reserved fields\n");
                return -EINVAL;
        }
 
@@ -1229,7 +1245,7 @@ static int check_xadd(struct bpf_verifier_env *env, int insn_idx, struct bpf_ins
                return err;
 
        if (is_pointer_value(env, insn->src_reg)) {
-               verbose("R%d leaks addr into mem\n", insn->src_reg);
+               verbose(env, "R%d leaks addr into mem\n", insn->src_reg);
                return -EACCES;
        }
 
@@ -1270,7 +1286,7 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
                    register_is_null(regs[regno]))
                        return 0;
 
-               verbose("R%d type=%s expected=%s\n", regno,
+               verbose(env, "R%d type=%s expected=%s\n", regno,
                        reg_type_str[regs[regno].type],
                        reg_type_str[PTR_TO_STACK]);
                return -EACCES;
@@ -1281,13 +1297,13 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
                char tn_buf[48];
 
                tnum_strn(tn_buf, sizeof(tn_buf), regs[regno].var_off);
-               verbose("invalid variable stack read R%d var_off=%s\n",
+               verbose(env, "invalid variable stack read R%d var_off=%s\n",
                        regno, tn_buf);
        }
        off = regs[regno].off + regs[regno].var_off.value;
        if (off >= 0 || off < -MAX_BPF_STACK || off + access_size > 0 ||
            access_size <= 0) {
-               verbose("invalid stack type R%d off=%d access_size=%d\n",
+               verbose(env, "invalid stack type R%d off=%d access_size=%d\n",
                        regno, off, access_size);
                return -EACCES;
        }
@@ -1303,7 +1319,7 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
 
        for (i = 0; i < access_size; i++) {
                if (state->stack_slot_type[MAX_BPF_STACK + off + i] != STACK_MISC) {
-                       verbose("invalid indirect read from stack off %d+%d size %d\n",
+                       verbose(env, "invalid indirect read from stack off %d+%d size %d\n",
                                off, i, access_size);
                        return -EACCES;
                }
@@ -1345,7 +1361,8 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 
        if (arg_type == ARG_ANYTHING) {
                if (is_pointer_value(env, regno)) {
-                       verbose("R%d leaks addr into helper function\n", regno);
+                       verbose(env, "R%d leaks addr into helper function\n",
+                               regno);
                        return -EACCES;
                }
                return 0;
@@ -1353,7 +1370,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 
        if (type == PTR_TO_PACKET &&
            !may_access_direct_pkt_data(env, meta, BPF_READ)) {
-               verbose("helper access to the packet is not allowed\n");
+               verbose(env, "helper access to the packet is not allowed\n");
                return -EACCES;
        }
 
@@ -1389,7 +1406,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                        goto err_type;
                meta->raw_mode = arg_type == ARG_PTR_TO_UNINIT_MEM;
        } else {
-               verbose("unsupported arg_type %d\n", arg_type);
+               verbose(env, "unsupported arg_type %d\n", arg_type);
                return -EFAULT;
        }
 
@@ -1407,7 +1424,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                         * we have to check map_key here. Otherwise it means
                         * that kernel subsystem misconfigured verifier
                         */
-                       verbose("invalid map_ptr to access map->key\n");
+                       verbose(env, "invalid map_ptr to access map->key\n");
                        return -EACCES;
                }
                if (type == PTR_TO_PACKET)
@@ -1423,7 +1440,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                 */
                if (!meta->map_ptr) {
                        /* kernel subsystem misconfigured verifier */
-                       verbose("invalid map_ptr to access map->value\n");
+                       verbose(env, "invalid map_ptr to access map->value\n");
                        return -EACCES;
                }
                if (type == PTR_TO_PACKET)
@@ -1443,7 +1460,8 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                 */
                if (regno == 0) {
                        /* kernel subsystem misconfigured verifier */
-                       verbose("ARG_CONST_SIZE cannot be first argument\n");
+                       verbose(env,
+                               "ARG_CONST_SIZE cannot be first argument\n");
                        return -EACCES;
                }
 
@@ -1460,7 +1478,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                        meta = NULL;
 
                if (reg->smin_value < 0) {
-                       verbose("R%d min value is negative, either use unsigned or 'var &= const'\n",
+                       verbose(env, "R%d min value is negative, either use unsigned or 'var &= const'\n",
                                regno);
                        return -EACCES;
                }
@@ -1474,7 +1492,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
                }
 
                if (reg->umax_value >= BPF_MAX_VAR_SIZ) {
-                       verbose("R%d unbounded memory access, use 'var &= const' or 'if (var < const)'\n",
+                       verbose(env, "R%d unbounded memory access, use 'var &= const' or 'if (var < const)'\n",
                                regno);
                        return -EACCES;
                }
@@ -1485,12 +1503,13 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 
        return err;
 err_type:
-       verbose("R%d type=%s expected=%s\n", regno,
+       verbose(env, "R%d type=%s expected=%s\n", regno,
                reg_type_str[type], reg_type_str[expected_type]);
        return -EACCES;
 }
 
-static int check_map_func_compatibility(struct bpf_map *map, int func_id)
+static int check_map_func_compatibility(struct bpf_verifier_env *env,
+                                       struct bpf_map *map, int func_id)
 {
        if (!map)
                return 0;
@@ -1576,7 +1595,7 @@ static int check_map_func_compatibility(struct bpf_map *map, int func_id)
 
        return 0;
 error:
-       verbose("cannot pass map_type %d into func %s#%d\n",
+       verbose(env, "cannot pass map_type %d into func %s#%d\n",
                map->map_type, func_id_name(func_id), func_id);
        return -EINVAL;
 }
@@ -1611,7 +1630,7 @@ static void clear_all_pkt_pointers(struct bpf_verifier_env *env)
        for (i = 0; i < MAX_BPF_REG; i++)
                if (regs[i].type == PTR_TO_PACKET ||
                    regs[i].type == PTR_TO_PACKET_END)
-                       mark_reg_unknown(regs, i);
+                       mark_reg_unknown(env, regs, i);
 
        for (i = 0; i < MAX_BPF_STACK; i += BPF_REG_SIZE) {
                if (state->stack_slot_type[i] != STACK_SPILL)
@@ -1635,7 +1654,8 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
 
        /* find function prototype */
        if (func_id < 0 || func_id >= __BPF_FUNC_MAX_ID) {
-               verbose("invalid func %s#%d\n", func_id_name(func_id), func_id);
+               verbose(env, "invalid func %s#%d\n", func_id_name(func_id),
+                       func_id);
                return -EINVAL;
        }
 
@@ -1643,13 +1663,14 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
                fn = env->prog->aux->ops->get_func_proto(func_id);
 
        if (!fn) {
-               verbose("unknown func %s#%d\n", func_id_name(func_id), func_id);
+               verbose(env, "unknown func %s#%d\n", func_id_name(func_id),
+                       func_id);
                return -EINVAL;
        }
 
        /* eBPF programs must be GPL compatible to use GPL-ed functions */
        if (!env->prog->gpl_compatible && fn->gpl_only) {
-               verbose("cannot call GPL only function from proprietary program\n");
+               verbose(env, "cannot call GPL only function from proprietary program\n");
                return -EINVAL;
        }
 
@@ -1663,7 +1684,7 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
         */
        err = check_raw_mode(fn);
        if (err) {
-               verbose("kernel subsystem misconfigured func %s#%d\n",
+               verbose(env, "kernel subsystem misconfigured func %s#%d\n",
                        func_id_name(func_id), func_id);
                return err;
        }
@@ -1696,14 +1717,14 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
 
        /* reset caller saved regs */
        for (i = 0; i < CALLER_SAVED_REGS; i++) {
-               mark_reg_not_init(regs, caller_saved[i]);
+               mark_reg_not_init(env, regs, caller_saved[i]);
                check_reg_arg(env, caller_saved[i], DST_OP_NO_MARK);
        }
 
        /* update return register (already marked as written above) */
        if (fn->ret_type == RET_INTEGER) {
                /* sets type to SCALAR_VALUE */
-               mark_reg_unknown(regs, BPF_REG_0);
+               mark_reg_unknown(env, regs, BPF_REG_0);
        } else if (fn->ret_type == RET_VOID) {
                regs[BPF_REG_0].type = NOT_INIT;
        } else if (fn->ret_type == RET_PTR_TO_MAP_VALUE_OR_NULL) {
@@ -1711,14 +1732,15 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
 
                regs[BPF_REG_0].type = PTR_TO_MAP_VALUE_OR_NULL;
                /* There is no offset yet applied, variable or fixed */
-               mark_reg_known_zero(regs, BPF_REG_0);
+               mark_reg_known_zero(env, regs, BPF_REG_0);
                regs[BPF_REG_0].off = 0;
                /* remember map_ptr, so that check_map_access()
                 * can check 'value_size' boundary of memory access
                 * to map element returned from bpf_map_lookup_elem()
                 */
                if (meta.map_ptr == NULL) {
-                       verbose("kernel subsystem misconfigured verifier\n");
+                       verbose(env,
+                               "kernel subsystem misconfigured verifier\n");
                        return -EINVAL;
                }
                regs[BPF_REG_0].map_ptr = meta.map_ptr;
@@ -1729,12 +1751,12 @@ static int check_call(struct bpf_verifier_env *env, int func_id, int insn_idx)
                else if (insn_aux->map_ptr != meta.map_ptr)
                        insn_aux->map_ptr = BPF_MAP_PTR_POISON;
        } else {
-               verbose("unknown return type %d of func %s#%d\n",
+               verbose(env, "unknown return type %d of func %s#%d\n",
                        fn->ret_type, func_id_name(func_id), func_id);
                return -EINVAL;
        }
 
-       err = check_map_func_compatibility(meta.map_ptr, func_id);
+       err = check_map_func_compatibility(env, meta.map_ptr, func_id);
        if (err)
                return err;
 
@@ -1793,39 +1815,42 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
        dst_reg = &regs[dst];
 
        if (WARN_ON_ONCE(known && (smin_val != smax_val))) {
-               print_verifier_state(&env->cur_state);
-               verbose("verifier internal error: known but bad sbounds\n");
+               print_verifier_state(env, &env->cur_state);
+               verbose(env,
+                       "verifier internal error: known but bad sbounds\n");
                return -EINVAL;
        }
        if (WARN_ON_ONCE(known && (umin_val != umax_val))) {
-               print_verifier_state(&env->cur_state);
-               verbose("verifier internal error: known but bad ubounds\n");
+               print_verifier_state(env, &env->cur_state);
+               verbose(env,
+                       "verifier internal error: known but bad ubounds\n");
                return -EINVAL;
        }
 
        if (BPF_CLASS(insn->code) != BPF_ALU64) {
                /* 32-bit ALU ops on pointers produce (meaningless) scalars */
                if (!env->allow_ptr_leaks)
-                       verbose("R%d 32-bit pointer arithmetic prohibited\n",
+                       verbose(env,
+                               "R%d 32-bit pointer arithmetic prohibited\n",
                                dst);
                return -EACCES;
        }
 
        if (ptr_reg->type == PTR_TO_MAP_VALUE_OR_NULL) {
                if (!env->allow_ptr_leaks)
-                       verbose("R%d pointer arithmetic on PTR_TO_MAP_VALUE_OR_NULL prohibited, null-check it first\n",
+                       verbose(env, "R%d pointer arithmetic on PTR_TO_MAP_VALUE_OR_NULL prohibited, null-check it first\n",
                                dst);
                return -EACCES;
        }
        if (ptr_reg->type == CONST_PTR_TO_MAP) {
                if (!env->allow_ptr_leaks)
-                       verbose("R%d pointer arithmetic on CONST_PTR_TO_MAP prohibited\n",
+                       verbose(env, "R%d pointer arithmetic on CONST_PTR_TO_MAP prohibited\n",
                                dst);
                return -EACCES;
        }
        if (ptr_reg->type == PTR_TO_PACKET_END) {
                if (!env->allow_ptr_leaks)
-                       verbose("R%d pointer arithmetic on PTR_TO_PACKET_END prohibited\n",
+                       verbose(env, "R%d pointer arithmetic on PTR_TO_PACKET_END prohibited\n",
                                dst);
                return -EACCES;
        }
@@ -1890,7 +1915,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
                if (dst_reg == off_reg) {
                        /* scalar -= pointer.  Creates an unknown scalar */
                        if (!env->allow_ptr_leaks)
-                               verbose("R%d tried to subtract pointer from scalar\n",
+                               verbose(env, "R%d tried to subtract pointer from scalar\n",
                                        dst);
                        return -EACCES;
                }
@@ -1900,7 +1925,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
                 */
                if (ptr_reg->type == PTR_TO_STACK) {
                        if (!env->allow_ptr_leaks)
-                               verbose("R%d subtraction from stack pointer prohibited\n",
+                               verbose(env, "R%d subtraction from stack pointer prohibited\n",
                                        dst);
                        return -EACCES;
                }
@@ -1955,13 +1980,13 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
                 * ptr &= ~3 which would reduce min_value by 3.)
                 */
                if (!env->allow_ptr_leaks)
-                       verbose("R%d bitwise operator %s on pointer prohibited\n",
+                       verbose(env, "R%d bitwise operator %s on pointer prohibited\n",
                                dst, bpf_alu_string[opcode >> 4]);
                return -EACCES;
        default:
                /* other operators (e.g. MUL,LSH) produce non-pointer results */
                if (!env->allow_ptr_leaks)
-                       verbose("R%d pointer arithmetic with %s operator prohibited\n",
+                       verbose(env, "R%d pointer arithmetic with %s operator prohibited\n",
                                dst, bpf_alu_string[opcode >> 4]);
                return -EACCES;
        }
@@ -2127,7 +2152,7 @@ static int adjust_scalar_min_max_vals(struct bpf_verifier_env *env,
                        /* Shifts greater than 63 are undefined.  This includes
                         * shifts by a negative number.
                         */
-                       mark_reg_unknown(regs, insn->dst_reg);
+                       mark_reg_unknown(env, regs, insn->dst_reg);
                        break;
                }
                /* We lose all sign bit information (except what we can pick
@@ -2155,7 +2180,7 @@ static int adjust_scalar_min_max_vals(struct bpf_verifier_env *env,
                        /* Shifts greater than 63 are undefined.  This includes
                         * shifts by a negative number.
                         */
-                       mark_reg_unknown(regs, insn->dst_reg);
+                       mark_reg_unknown(env, regs, insn->dst_reg);
                        break;
                }
                /* BPF_RSH is an unsigned shift, so make the appropriate casts */
@@ -2183,7 +2208,7 @@ static int adjust_scalar_min_max_vals(struct bpf_verifier_env *env,
                __update_reg_bounds(dst_reg);
                break;
        default:
-               mark_reg_unknown(regs, insn->dst_reg);
+               mark_reg_unknown(env, regs, insn->dst_reg);
                break;
        }
 
@@ -2215,12 +2240,12 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env,
                                 * an arbitrary scalar.
                                 */
                                if (!env->allow_ptr_leaks) {
-                                       verbose("R%d pointer %s pointer prohibited\n",
+                                       verbose(env, "R%d pointer %s pointer prohibited\n",
                                                insn->dst_reg,
                                                bpf_alu_string[opcode >> 4]);
                                        return -EACCES;
                                }
-                               mark_reg_unknown(regs, insn->dst_reg);
+                               mark_reg_unknown(env, regs, insn->dst_reg);
                                return 0;
                        } else {
                                /* scalar += pointer
@@ -2272,13 +2297,13 @@ static int adjust_reg_min_max_vals(struct bpf_verifier_env *env,
 
        /* Got here implies adding two SCALAR_VALUEs */
        if (WARN_ON_ONCE(ptr_reg)) {
-               print_verifier_state(&env->cur_state);
-               verbose("verifier internal error: unexpected ptr_reg\n");
+               print_verifier_state(env, &env->cur_state);
+               verbose(env, "verifier internal error: unexpected ptr_reg\n");
                return -EINVAL;
        }
        if (WARN_ON(!src_reg)) {
-               print_verifier_state(&env->cur_state);
-               verbose("verifier internal error: no src_reg\n");
+               print_verifier_state(env, &env->cur_state);
+               verbose(env, "verifier internal error: no src_reg\n");
                return -EINVAL;
        }
        return adjust_scalar_min_max_vals(env, insn, dst_reg, *src_reg);
@@ -2296,14 +2321,14 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                        if (BPF_SRC(insn->code) != 0 ||
                            insn->src_reg != BPF_REG_0 ||
                            insn->off != 0 || insn->imm != 0) {
-                               verbose("BPF_NEG uses reserved fields\n");
+                               verbose(env, "BPF_NEG uses reserved fields\n");
                                return -EINVAL;
                        }
                } else {
                        if (insn->src_reg != BPF_REG_0 || insn->off != 0 ||
                            (insn->imm != 16 && insn->imm != 32 && insn->imm != 64) ||
                            BPF_CLASS(insn->code) == BPF_ALU64) {
-                               verbose("BPF_END uses reserved fields\n");
+                               verbose(env, "BPF_END uses reserved fields\n");
                                return -EINVAL;
                        }
                }
@@ -2314,7 +2339,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                        return err;
 
                if (is_pointer_value(env, insn->dst_reg)) {
-                       verbose("R%d pointer arithmetic prohibited\n",
+                       verbose(env, "R%d pointer arithmetic prohibited\n",
                                insn->dst_reg);
                        return -EACCES;
                }
@@ -2328,7 +2353,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
 
                if (BPF_SRC(insn->code) == BPF_X) {
                        if (insn->imm != 0 || insn->off != 0) {
-                               verbose("BPF_MOV uses reserved fields\n");
+                               verbose(env, "BPF_MOV uses reserved fields\n");
                                return -EINVAL;
                        }
 
@@ -2338,7 +2363,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                                return err;
                } else {
                        if (insn->src_reg != BPF_REG_0 || insn->off != 0) {
-                               verbose("BPF_MOV uses reserved fields\n");
+                               verbose(env, "BPF_MOV uses reserved fields\n");
                                return -EINVAL;
                        }
                }
@@ -2358,11 +2383,12 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                        } else {
                                /* R1 = (u32) R2 */
                                if (is_pointer_value(env, insn->src_reg)) {
-                                       verbose("R%d partial copy of pointer\n",
+                                       verbose(env,
+                                               "R%d partial copy of pointer\n",
                                                insn->src_reg);
                                        return -EACCES;
                                }
-                               mark_reg_unknown(regs, insn->dst_reg);
+                               mark_reg_unknown(env, regs, insn->dst_reg);
                                /* high 32 bits are known zero. */
                                regs[insn->dst_reg].var_off = tnum_cast(
                                                regs[insn->dst_reg].var_off, 4);
@@ -2377,14 +2403,14 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                }
 
        } else if (opcode > BPF_END) {
-               verbose("invalid BPF_ALU opcode %x\n", opcode);
+               verbose(env, "invalid BPF_ALU opcode %x\n", opcode);
                return -EINVAL;
 
        } else {        /* all other ALU ops: and, sub, xor, add, ... */
 
                if (BPF_SRC(insn->code) == BPF_X) {
                        if (insn->imm != 0 || insn->off != 0) {
-                               verbose("BPF_ALU uses reserved fields\n");
+                               verbose(env, "BPF_ALU uses reserved fields\n");
                                return -EINVAL;
                        }
                        /* check src1 operand */
@@ -2393,7 +2419,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                                return err;
                } else {
                        if (insn->src_reg != BPF_REG_0 || insn->off != 0) {
-                               verbose("BPF_ALU uses reserved fields\n");
+                               verbose(env, "BPF_ALU uses reserved fields\n");
                                return -EINVAL;
                        }
                }
@@ -2405,7 +2431,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
 
                if ((opcode == BPF_MOD || opcode == BPF_DIV) &&
                    BPF_SRC(insn->code) == BPF_K && insn->imm == 0) {
-                       verbose("div by zero\n");
+                       verbose(env, "div by zero\n");
                        return -EINVAL;
                }
 
@@ -2414,7 +2440,7 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
                        int size = BPF_CLASS(insn->code) == BPF_ALU64 ? 64 : 32;
 
                        if (insn->imm < 0 || insn->imm >= size) {
-                               verbose("invalid shift %d\n", insn->imm);
+                               verbose(env, "invalid shift %d\n", insn->imm);
                                return -EINVAL;
                        }
                }
@@ -2775,13 +2801,13 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env,
        int err;
 
        if (opcode > BPF_JSLE) {
-               verbose("invalid BPF_JMP opcode %x\n", opcode);
+               verbose(env, "invalid BPF_JMP opcode %x\n", opcode);
                return -EINVAL;
        }
 
        if (BPF_SRC(insn->code) == BPF_X) {
                if (insn->imm != 0) {
-                       verbose("BPF_JMP uses reserved fields\n");
+                       verbose(env, "BPF_JMP uses reserved fields\n");
                        return -EINVAL;
                }
 
@@ -2791,13 +2817,13 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env,
                        return err;
 
                if (is_pointer_value(env, insn->src_reg)) {
-                       verbose("R%d pointer comparison prohibited\n",
+                       verbose(env, "R%d pointer comparison prohibited\n",
                                insn->src_reg);
                        return -EACCES;
                }
        } else {
                if (insn->src_reg != BPF_REG_0) {
-                       verbose("BPF_JMP uses reserved fields\n");
+                       verbose(env, "BPF_JMP uses reserved fields\n");
                        return -EINVAL;
                }
        }
@@ -2913,11 +2939,12 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env,
                /* pkt_end <= pkt_data' */
                find_good_pkt_pointers(this_branch, &regs[insn->src_reg], true);
        } else if (is_pointer_value(env, insn->dst_reg)) {
-               verbose("R%d pointer comparison prohibited\n", insn->dst_reg);
+               verbose(env, "R%d pointer comparison prohibited\n",
+                       insn->dst_reg);
                return -EACCES;
        }
-       if (verifier_log.level)
-               print_verifier_state(this_branch);
+       if (env->log.level)
+               print_verifier_state(env, this_branch);
        return 0;
 }
 
@@ -2936,11 +2963,11 @@ static int check_ld_imm(struct bpf_verifier_env *env, struct bpf_insn *insn)
        int err;
 
        if (BPF_SIZE(insn->code) != BPF_DW) {
-               verbose("invalid BPF_LD_IMM insn\n");
+               verbose(env, "invalid BPF_LD_IMM insn\n");
                return -EINVAL;
        }
        if (insn->off != 0) {
-               verbose("BPF_LD_IMM64 uses reserved fields\n");
+               verbose(env, "BPF_LD_IMM64 uses reserved fields\n");
                return -EINVAL;
        }
 
@@ -2998,14 +3025,14 @@ static int check_ld_abs(struct bpf_verifier_env *env, struct bpf_insn *insn)
        int i, err;
 
        if (!may_access_skb(env->prog->type)) {
-               verbose("BPF_LD_[ABS|IND] instructions not allowed for this program type\n");
+               verbose(env, "BPF_LD_[ABS|IND] instructions not allowed for this program type\n");
                return -EINVAL;
        }
 
        if (insn->dst_reg != BPF_REG_0 || insn->off != 0 ||
            BPF_SIZE(insn->code) == BPF_DW ||
            (mode == BPF_ABS && insn->src_reg != BPF_REG_0)) {
-               verbose("BPF_LD_[ABS|IND] uses reserved fields\n");
+               verbose(env, "BPF_LD_[ABS|IND] uses reserved fields\n");
                return -EINVAL;
        }
 
@@ -3015,7 +3042,8 @@ static int check_ld_abs(struct bpf_verifier_env *env, struct bpf_insn *insn)
                return err;
 
        if (regs[BPF_REG_6].type != PTR_TO_CTX) {
-               verbose("at the time of BPF_LD_ABS|IND R6 != pointer to skb\n");
+               verbose(env,
+                       "at the time of BPF_LD_ABS|IND R6 != pointer to skb\n");
                return -EINVAL;
        }
 
@@ -3028,7 +3056,7 @@ static int check_ld_abs(struct bpf_verifier_env *env, struct bpf_insn *insn)
 
        /* reset caller saved regs to unreadable */
        for (i = 0; i < CALLER_SAVED_REGS; i++) {
-               mark_reg_not_init(regs, caller_saved[i]);
+               mark_reg_not_init(env, regs, caller_saved[i]);
                check_reg_arg(env, caller_saved[i], DST_OP_NO_MARK);
        }
 
@@ -3036,7 +3064,7 @@ static int check_ld_abs(struct bpf_verifier_env *env, struct bpf_insn *insn)
         * the value fetched from the packet.
         * Already marked as written above.
         */
-       mark_reg_unknown(regs, BPF_REG_0);
+       mark_reg_unknown(env, regs, BPF_REG_0);
        return 0;
 }
 
@@ -3100,7 +3128,7 @@ static int push_insn(int t, int w, int e, struct bpf_verifier_env *env)
                return 0;
 
        if (w < 0 || w >= env->prog->len) {
-               verbose("jump out of range from insn %d to %d\n", t, w);
+               verbose(env, "jump out of range from insn %d to %d\n", t, w);
                return -EINVAL;
        }
 
@@ -3117,13 +3145,13 @@ static int push_insn(int t, int w, int e, struct bpf_verifier_env *env)
                insn_stack[cur_stack++] = w;
                return 1;
        } else if ((insn_state[w] & 0xF0) == DISCOVERED) {
-               verbose("back-edge from insn %d to %d\n", t, w);
+               verbose(env, "back-edge from insn %d to %d\n", t, w);
                return -EINVAL;
        } else if (insn_state[w] == EXPLORED) {
                /* forward- or cross-edge */
                insn_state[t] = DISCOVERED | e;
        } else {
-               verbose("insn state internal bug\n");
+               verbose(env, "insn state internal bug\n");
                return -EFAULT;
        }
        return 0;
@@ -3217,7 +3245,7 @@ peek_stack:
 mark_explored:
        insn_state[t] = EXPLORED;
        if (cur_stack-- <= 0) {
-               verbose("pop stack internal bug\n");
+               verbose(env, "pop stack internal bug\n");
                ret = -EFAULT;
                goto err_free;
        }
@@ -3226,7 +3254,7 @@ mark_explored:
 check_state:
        for (i = 0; i < insn_cnt; i++) {
                if (insn_state[i] != EXPLORED) {
-                       verbose("unreachable insn %d\n", i);
+                       verbose(env, "unreachable insn %d\n", i);
                        ret = -EINVAL;
                        goto err_free;
                }
@@ -3606,7 +3634,7 @@ static int do_check(struct bpf_verifier_env *env)
        int insn_processed = 0;
        bool do_print_state = false;
 
-       init_reg_state(regs);
+       init_reg_state(env, regs);
        state->parent = NULL;
        insn_idx = 0;
        for (;;) {
@@ -3615,7 +3643,7 @@ static int do_check(struct bpf_verifier_env *env)
                int err;
 
                if (insn_idx >= insn_cnt) {
-                       verbose("invalid insn idx %d insn_cnt %d\n",
+                       verbose(env, "invalid insn idx %d insn_cnt %d\n",
                                insn_idx, insn_cnt);
                        return -EFAULT;
                }
@@ -3624,7 +3652,8 @@ static int do_check(struct bpf_verifier_env *env)
                class = BPF_CLASS(insn->code);
 
                if (++insn_processed > BPF_COMPLEXITY_LIMIT_INSNS) {
-                       verbose("BPF program is too large. Processed %d insn\n",
+                       verbose(env,
+                               "BPF program is too large. Processed %d insn\n",
                                insn_processed);
                        return -E2BIG;
                }
@@ -3634,12 +3663,12 @@ static int do_check(struct bpf_verifier_env *env)
                        return err;
                if (err == 1) {
                        /* found equivalent state, can prune the search */
-                       if (verifier_log.level) {
+                       if (env->log.level) {
                                if (do_print_state)
-                                       verbose("\nfrom %d to %d: safe\n",
+                                       verbose(env, "\nfrom %d to %d: safe\n",
                                                prev_insn_idx, insn_idx);
                                else
-                                       verbose("%d: safe\n", insn_idx);
+                                       verbose(env, "%d: safe\n", insn_idx);
                        }
                        goto process_bpf_exit;
                }
@@ -3647,19 +3676,18 @@ static int do_check(struct bpf_verifier_env *env)
                if (need_resched())
                        cond_resched();
 
-               if (verifier_log.level > 1 ||
-                   (verifier_log.level && do_print_state)) {
-                       if (verifier_log.level > 1)
-                               verbose("%d:", insn_idx);
+               if (env->log.level > 1 || (env->log.level && do_print_state)) {
+                       if (env->log.level > 1)
+                               verbose(env, "%d:", insn_idx);
                        else
-                               verbose("\nfrom %d to %d:",
+                               verbose(env, "\nfrom %d to %d:",
                                        prev_insn_idx, insn_idx);
-                       print_verifier_state(&env->cur_state);
+                       print_verifier_state(env, &env->cur_state);
                        do_print_state = false;
                }
 
-               if (verifier_log.level) {
-                       verbose("%d: ", insn_idx);
+               if (env->log.level) {
+                       verbose(env, "%d: ", insn_idx);
                        print_bpf_insn(env, insn);
                }
 
@@ -3716,7 +3744,7 @@ static int do_check(struct bpf_verifier_env *env)
                                 * src_reg == stack|map in some other branch.
                                 * Reject it.
                                 */
-                               verbose("same insn cannot be used with different pointers\n");
+                               verbose(env, "same insn cannot be used with different pointers\n");
                                return -EINVAL;
                        }
 
@@ -3756,14 +3784,14 @@ static int do_check(struct bpf_verifier_env *env)
                        } else if (dst_reg_type != *prev_dst_type &&
                                   (dst_reg_type == PTR_TO_CTX ||
                                    *prev_dst_type == PTR_TO_CTX)) {
-                               verbose("same insn cannot be used with different pointers\n");
+                               verbose(env, "same insn cannot be used with different pointers\n");
                                return -EINVAL;
                        }
 
                } else if (class == BPF_ST) {
                        if (BPF_MODE(insn->code) != BPF_MEM ||
                            insn->src_reg != BPF_REG_0) {
-                               verbose("BPF_ST uses reserved fields\n");
+                               verbose(env, "BPF_ST uses reserved fields\n");
                                return -EINVAL;
                        }
                        /* check src operand */
@@ -3786,7 +3814,7 @@ static int do_check(struct bpf_verifier_env *env)
                                    insn->off != 0 ||
                                    insn->src_reg != BPF_REG_0 ||
                                    insn->dst_reg != BPF_REG_0) {
-                                       verbose("BPF_CALL uses reserved fields\n");
+                                       verbose(env, "BPF_CALL uses reserved fields\n");
                                        return -EINVAL;
                                }
 
@@ -3799,7 +3827,7 @@ static int do_check(struct bpf_verifier_env *env)
                                    insn->imm != 0 ||
                                    insn->src_reg != BPF_REG_0 ||
                                    insn->dst_reg != BPF_REG_0) {
-                                       verbose("BPF_JA uses reserved fields\n");
+                                       verbose(env, "BPF_JA uses reserved fields\n");
                                        return -EINVAL;
                                }
 
@@ -3811,7 +3839,7 @@ static int do_check(struct bpf_verifier_env *env)
                                    insn->imm != 0 ||
                                    insn->src_reg != BPF_REG_0 ||
                                    insn->dst_reg != BPF_REG_0) {
-                                       verbose("BPF_EXIT uses reserved fields\n");
+                                       verbose(env, "BPF_EXIT uses reserved fields\n");
                                        return -EINVAL;
                                }
 
@@ -3826,7 +3854,7 @@ static int do_check(struct bpf_verifier_env *env)
                                        return err;
 
                                if (is_pointer_value(env, BPF_REG_0)) {
-                                       verbose("R0 leaks addr as return value\n");
+                                       verbose(env, "R0 leaks addr as return value\n");
                                        return -EACCES;
                                }
 
@@ -3858,19 +3886,19 @@ process_bpf_exit:
 
                                insn_idx++;
                        } else {
-                               verbose("invalid BPF_LD mode\n");
+                               verbose(env, "invalid BPF_LD mode\n");
                                return -EINVAL;
                        }
                } else {
-                       verbose("unknown insn class %d\n", class);
+                       verbose(env, "unknown insn class %d\n", class);
                        return -EINVAL;
                }
 
                insn_idx++;
        }
 
-       verbose("processed %d insns, stack depth %d\n",
-               insn_processed, env->prog->aux->stack_depth);
+       verbose(env, "processed %d insns, stack depth %d\n", insn_processed,
+               env->prog->aux->stack_depth);
        return 0;
 }
 
@@ -3882,7 +3910,8 @@ static int check_map_prealloc(struct bpf_map *map)
                !(map->map_flags & BPF_F_NO_PREALLOC);
 }
 
-static int check_map_prog_compatibility(struct bpf_map *map,
+static int check_map_prog_compatibility(struct bpf_verifier_env *env,
+                                       struct bpf_map *map,
                                        struct bpf_prog *prog)
 
 {
@@ -3893,12 +3922,12 @@ static int check_map_prog_compatibility(struct bpf_map *map,
         */
        if (prog->type == BPF_PROG_TYPE_PERF_EVENT) {
                if (!check_map_prealloc(map)) {
-                       verbose("perf_event programs can only use preallocated hash map\n");
+                       verbose(env, "perf_event programs can only use preallocated hash map\n");
                        return -EINVAL;
                }
                if (map->inner_map_meta &&
                    !check_map_prealloc(map->inner_map_meta)) {
-                       verbose("perf_event programs can only use preallocated inner hash map\n");
+                       verbose(env, "perf_event programs can only use preallocated inner hash map\n");
                        return -EINVAL;
                }
        }
@@ -3921,14 +3950,14 @@ static int replace_map_fd_with_map_ptr(struct bpf_verifier_env *env)
        for (i = 0; i < insn_cnt; i++, insn++) {
                if (BPF_CLASS(insn->code) == BPF_LDX &&
                    (BPF_MODE(insn->code) != BPF_MEM || insn->imm != 0)) {
-                       verbose("BPF_LDX uses reserved fields\n");
+                       verbose(env, "BPF_LDX uses reserved fields\n");
                        return -EINVAL;
                }
 
                if (BPF_CLASS(insn->code) == BPF_STX &&
                    ((BPF_MODE(insn->code) != BPF_MEM &&
                      BPF_MODE(insn->code) != BPF_XADD) || insn->imm != 0)) {
-                       verbose("BPF_STX uses reserved fields\n");
+                       verbose(env, "BPF_STX uses reserved fields\n");
                        return -EINVAL;
                }
 
@@ -3939,7 +3968,7 @@ static int replace_map_fd_with_map_ptr(struct bpf_verifier_env *env)
                        if (i == insn_cnt - 1 || insn[1].code != 0 ||
                            insn[1].dst_reg != 0 || insn[1].src_reg != 0 ||
                            insn[1].off != 0) {
-                               verbose("invalid bpf_ld_imm64 insn\n");
+                               verbose(env, "invalid bpf_ld_imm64 insn\n");
                                return -EINVAL;
                        }
 
@@ -3948,19 +3977,20 @@ static int replace_map_fd_with_map_ptr(struct bpf_verifier_env *env)
                                goto next_insn;
 
                        if (insn->src_reg != BPF_PSEUDO_MAP_FD) {
-                               verbose("unrecognized bpf_ld_imm64 insn\n");
+                               verbose(env,
+                                       "unrecognized bpf_ld_imm64 insn\n");
                                return -EINVAL;
                        }
 
                        f = fdget(insn->imm);
                        map = __bpf_map_get(f);
                        if (IS_ERR(map)) {
-                               verbose("fd %d is not pointing to valid bpf_map\n",
+                               verbose(env, "fd %d is not pointing to valid bpf_map\n",
                                        insn->imm);
                                return PTR_ERR(map);
                        }
 
-                       err = check_map_prog_compatibility(map, env->prog);
+                       err = check_map_prog_compatibility(env, map, env->prog);
                        if (err) {
                                fdput(f);
                                return err;
@@ -4082,7 +4112,7 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
                cnt = ops->gen_prologue(insn_buf, env->seen_direct_write,
                                        env->prog);
                if (cnt >= ARRAY_SIZE(insn_buf)) {
-                       verbose("bpf verifier is misconfigured\n");
+                       verbose(env, "bpf verifier is misconfigured\n");
                        return -EINVAL;
                } else if (cnt) {
                        new_prog = bpf_patch_insn_data(env, 0, insn_buf, cnt);
@@ -4130,7 +4160,7 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
                        u8 size_code;
 
                        if (type == BPF_WRITE) {
-                               verbose("bpf verifier narrow ctx access misconfigured\n");
+                               verbose(env, "bpf verifier narrow ctx access misconfigured\n");
                                return -EINVAL;
                        }
 
@@ -4149,7 +4179,7 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
                                              &target_size);
                if (cnt == 0 || cnt >= ARRAY_SIZE(insn_buf) ||
                    (ctx_field_size && !target_size)) {
-                       verbose("bpf verifier is misconfigured\n");
+                       verbose(env, "bpf verifier is misconfigured\n");
                        return -EINVAL;
                }
 
@@ -4231,7 +4261,7 @@ static int fixup_bpf_calls(struct bpf_verifier_env *env)
 
                        cnt = map_ptr->ops->map_gen_lookup(map_ptr, insn_buf);
                        if (cnt == 0 || cnt >= ARRAY_SIZE(insn_buf)) {
-                               verbose("bpf verifier is misconfigured\n");
+                               verbose(env, "bpf verifier is misconfigured\n");
                                return -EINVAL;
                        }
 
@@ -4275,7 +4305,8 @@ patch_call_imm:
                 * programs to call them, must be real in-kernel functions
                 */
                if (!fn->func) {
-                       verbose("kernel subsystem misconfigured func %s#%d\n",
+                       verbose(env,
+                               "kernel subsystem misconfigured func %s#%d\n",
                                func_id_name(insn->imm), insn->imm);
                        return -EFAULT;
                }
@@ -4309,8 +4340,8 @@ static void free_states(struct bpf_verifier_env *env)
 
 int bpf_check(struct bpf_prog **prog, union bpf_attr *attr)
 {
-       struct bpf_verifer_log *log = &verifier_log;
        struct bpf_verifier_env *env;
+       struct bpf_verifer_log *log;
        int ret = -EINVAL;
 
        /* 'struct bpf_verifier_env' can be global, but since it's not small,
@@ -4319,6 +4350,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr)
        env = kzalloc(sizeof(struct bpf_verifier_env), GFP_KERNEL);
        if (!env)
                return -ENOMEM;
+       log = &env->log;
 
        env->insn_aux_data = vzalloc(sizeof(struct bpf_insn_aux_data) *
                                     (*prog)->len);
@@ -4337,7 +4369,6 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr)
                log->level = attr->log_level;
                log->ubuf = (char __user *) (unsigned long) attr->log_buf;
                log->len_total = attr->log_size;
-               log->len_used = 0;
 
                ret = -EINVAL;
                /* log attributes have to be sane */
@@ -4349,8 +4380,6 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr)
                log->kbuf = vmalloc(log->len_total);
                if (!log->kbuf)
                        goto err_unlock;
-       } else {
-               log->level = 0;
        }
 
        env->strict_alignment = !!(attr->prog_flags & BPF_F_STRICT_ALIGNMENT);
@@ -4461,8 +4490,6 @@ int bpf_analyzer(struct bpf_prog *prog, const struct bpf_ext_analyzer_ops *ops,
        /* grab the mutex to protect few globals used by verifier */
        mutex_lock(&bpf_verifier_lock);
 
-       verifier_log.level = 0;
-
        env->strict_alignment = false;
        if (!IS_ENABLED(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS))
                env->strict_alignment = true;