summaryrefslogtreecommitdiff
path: root/kernel/bpf
diff options
context:
space:
mode:
authorDavid S. Miller <davem@davemloft.net>2021-02-17 00:14:06 +0300
committerDavid S. Miller <davem@davemloft.net>2021-02-17 00:14:06 +0300
commitb8af417e4d93caeefb89bbfbd56ec95dedd8dab5 (patch)
tree1c8d22e1aec330238830a43cc8aee0cf768ae1c7 /kernel/bpf
parent9ec5eea5b6acfae7279203097eeec5d02d01d9b7 (diff)
parent45159b27637b0fef6d5ddb86fc7c46b13c77960f (diff)
downloadlinux-b8af417e4d93caeefb89bbfbd56ec95dedd8dab5.tar.xz
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
Daniel Borkmann says: ==================== pull-request: bpf-next 2021-02-16 The following pull-request contains BPF updates for your *net-next* tree. There's a small merge conflict between 7eeba1706eba ("tcp: Add receive timestamp support for receive zerocopy.") from net-next tree and 9cacf81f8161 ("bpf: Remove extra lock_sock for TCP_ZEROCOPY_RECEIVE") from bpf-next tree. Resolve as follows: [...] lock_sock(sk); err = tcp_zerocopy_receive(sk, &zc, &tss); err = BPF_CGROUP_RUN_PROG_GETSOCKOPT_KERN(sk, level, optname, &zc, &len, err); release_sock(sk); [...] We've added 116 non-merge commits during the last 27 day(s) which contain a total of 156 files changed, 5662 insertions(+), 1489 deletions(-). The main changes are: 1) Adds support of pointers to types with known size among global function args to overcome the limit on max # of allowed args, from Dmitrii Banshchikov. 2) Add bpf_iter for task_vma which can be used to generate information similar to /proc/pid/maps, from Song Liu. 3) Enable bpf_{g,s}etsockopt() from all sock_addr related program hooks. Allow rewriting bind user ports from BPF side below the ip_unprivileged_port_start range, both from Stanislav Fomichev. 4) Prevent recursion on fentry/fexit & sleepable programs and allow map-in-map as well as per-cpu maps for the latter, from Alexei Starovoitov. 5) Add selftest script to run BPF CI locally. Also enable BPF ringbuffer for sleepable programs, both from KP Singh. 6) Extend verifier to enable variable offset read/write access to the BPF program stack, from Andrei Matei. 7) Improve tc & XDP MTU handling and add a new bpf_check_mtu() helper to query device MTU from programs, from Jesper Dangaard Brouer. 8) Allow bpf_get_socket_cookie() helper also be called from [sleepable] BPF tracing programs, from Florent Revest. 9) Extend x86 JIT to pad JMPs with NOPs for helping image to converge when otherwise too many passes are required, from Gary Lin. 10) Verifier fixes on atomics with BPF_FETCH as well as function-by-function verification both related to zero-extension handling, from Ilya Leoshkevich. 11) Better kernel build integration of resolve_btfids tool, from Jiri Olsa. 12) Batch of AF_XDP selftest cleanups and small performance improvement for libbpf's xsk map redirect for newer kernels, from Björn Töpel. 13) Follow-up BPF doc and verifier improvements around atomics with BPF_FETCH, from Brendan Jackman. 14) Permit zero-sized data sections e.g. if ELF .rodata section contains read-only data from local variables, from Yonghong Song. 15) veth driver skb bulk-allocation for ndo_xdp_xmit, from Lorenzo Bianconi. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'kernel/bpf')
-rw-r--r--kernel/bpf/bpf_iter.c2
-rw-r--r--kernel/bpf/bpf_lru_list.c7
-rw-r--r--kernel/bpf/btf.c76
-rw-r--r--kernel/bpf/cgroup.c120
-rw-r--r--kernel/bpf/core.c18
-rw-r--r--kernel/bpf/cpumap.c46
-rw-r--r--kernel/bpf/devmap.c4
-rw-r--r--kernel/bpf/disasm.c2
-rw-r--r--kernel/bpf/hashtab.c4
-rw-r--r--kernel/bpf/helpers.c12
-rw-r--r--kernel/bpf/syscall.c16
-rw-r--r--kernel/bpf/task_iter.c267
-rw-r--r--kernel/bpf/trampoline.c77
-rw-r--r--kernel/bpf/verifier.c877
14 files changed, 1187 insertions, 341 deletions
diff --git a/kernel/bpf/bpf_iter.c b/kernel/bpf/bpf_iter.c
index 5454161407f1..a0d9eade9c80 100644
--- a/kernel/bpf/bpf_iter.c
+++ b/kernel/bpf/bpf_iter.c
@@ -287,7 +287,7 @@ int bpf_iter_reg_target(const struct bpf_iter_reg *reg_info)
{
struct bpf_iter_target_info *tinfo;
- tinfo = kmalloc(sizeof(*tinfo), GFP_KERNEL);
+ tinfo = kzalloc(sizeof(*tinfo), GFP_KERNEL);
if (!tinfo)
return -ENOMEM;
diff --git a/kernel/bpf/bpf_lru_list.c b/kernel/bpf/bpf_lru_list.c
index 1b6b9349cb85..d99e89f113c4 100644
--- a/kernel/bpf/bpf_lru_list.c
+++ b/kernel/bpf/bpf_lru_list.c
@@ -502,13 +502,14 @@ struct bpf_lru_node *bpf_lru_pop_free(struct bpf_lru *lru, u32 hash)
static void bpf_common_lru_push_free(struct bpf_lru *lru,
struct bpf_lru_node *node)
{
+ u8 node_type = READ_ONCE(node->type);
unsigned long flags;
- if (WARN_ON_ONCE(node->type == BPF_LRU_LIST_T_FREE) ||
- WARN_ON_ONCE(node->type == BPF_LRU_LOCAL_LIST_T_FREE))
+ if (WARN_ON_ONCE(node_type == BPF_LRU_LIST_T_FREE) ||
+ WARN_ON_ONCE(node_type == BPF_LRU_LOCAL_LIST_T_FREE))
return;
- if (node->type == BPF_LRU_LOCAL_LIST_T_PENDING) {
+ if (node_type == BPF_LRU_LOCAL_LIST_T_PENDING) {
struct bpf_lru_locallist *loc_l;
loc_l = per_cpu_ptr(lru->common_lru.local_list, node->cpu);
diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
index 8962f988514f..2efeb5f4b343 100644
--- a/kernel/bpf/btf.c
+++ b/kernel/bpf/btf.c
@@ -3540,11 +3540,6 @@ static s32 btf_datasec_check_meta(struct btf_verifier_env *env,
return -EINVAL;
}
- if (!btf_type_vlen(t)) {
- btf_verifier_log_type(env, t, "vlen == 0");
- return -EINVAL;
- }
-
if (!t->size) {
btf_verifier_log_type(env, t, "size == 0");
return -EINVAL;
@@ -5296,15 +5291,16 @@ int btf_check_type_match(struct bpf_verifier_log *log, const struct bpf_prog *pr
* Only PTR_TO_CTX and SCALAR_VALUE states are recognized.
*/
int btf_check_func_arg_match(struct bpf_verifier_env *env, int subprog,
- struct bpf_reg_state *reg)
+ struct bpf_reg_state *regs)
{
struct bpf_verifier_log *log = &env->log;
struct bpf_prog *prog = env->prog;
struct btf *btf = prog->aux->btf;
const struct btf_param *args;
- const struct btf_type *t;
- u32 i, nargs, btf_id;
+ const struct btf_type *t, *ref_t;
+ u32 i, nargs, btf_id, type_size;
const char *tname;
+ bool is_global;
if (!prog->aux->func_info)
return -EINVAL;
@@ -5338,38 +5334,57 @@ int btf_check_func_arg_match(struct bpf_verifier_env *env, int subprog,
bpf_log(log, "Function %s has %d > 5 args\n", tname, nargs);
goto out;
}
+
+ is_global = prog->aux->func_info_aux[subprog].linkage == BTF_FUNC_GLOBAL;
/* check that BTF function arguments match actual types that the
* verifier sees.
*/
for (i = 0; i < nargs; i++) {
+ struct bpf_reg_state *reg = &regs[i + 1];
+
t = btf_type_by_id(btf, args[i].type);
while (btf_type_is_modifier(t))
t = btf_type_by_id(btf, t->type);
if (btf_type_is_int(t) || btf_type_is_enum(t)) {
- if (reg[i + 1].type == SCALAR_VALUE)
+ if (reg->type == SCALAR_VALUE)
continue;
bpf_log(log, "R%d is not a scalar\n", i + 1);
goto out;
}
if (btf_type_is_ptr(t)) {
- if (reg[i + 1].type == SCALAR_VALUE) {
- bpf_log(log, "R%d is not a pointer\n", i + 1);
- goto out;
- }
/* If function expects ctx type in BTF check that caller
* is passing PTR_TO_CTX.
*/
if (btf_get_prog_ctx_type(log, btf, t, prog->type, i)) {
- if (reg[i + 1].type != PTR_TO_CTX) {
+ if (reg->type != PTR_TO_CTX) {
bpf_log(log,
"arg#%d expected pointer to ctx, but got %s\n",
i, btf_kind_str[BTF_INFO_KIND(t->info)]);
goto out;
}
- if (check_ctx_reg(env, &reg[i + 1], i + 1))
+ if (check_ctx_reg(env, reg, i + 1))
goto out;
continue;
}
+
+ if (!is_global)
+ goto out;
+
+ t = btf_type_skip_modifiers(btf, t->type, NULL);
+
+ ref_t = btf_resolve_size(btf, t, &type_size);
+ if (IS_ERR(ref_t)) {
+ bpf_log(log,
+ "arg#%d reference type('%s %s') size cannot be determined: %ld\n",
+ i, btf_type_str(t), btf_name_by_offset(btf, t->name_off),
+ PTR_ERR(ref_t));
+ goto out;
+ }
+
+ if (check_mem_reg(env, reg, i + 1, type_size))
+ goto out;
+
+ continue;
}
bpf_log(log, "Unrecognized arg#%d type %s\n",
i, btf_kind_str[BTF_INFO_KIND(t->info)]);
@@ -5393,14 +5408,14 @@ out:
* (either PTR_TO_CTX or SCALAR_VALUE).
*/
int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog,
- struct bpf_reg_state *reg)
+ struct bpf_reg_state *regs)
{
struct bpf_verifier_log *log = &env->log;
struct bpf_prog *prog = env->prog;
enum bpf_prog_type prog_type = prog->type;
struct btf *btf = prog->aux->btf;
const struct btf_param *args;
- const struct btf_type *t;
+ const struct btf_type *t, *ref_t;
u32 i, nargs, btf_id;
const char *tname;
@@ -5464,16 +5479,35 @@ int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog,
* Only PTR_TO_CTX and SCALAR are supported atm.
*/
for (i = 0; i < nargs; i++) {
+ struct bpf_reg_state *reg = &regs[i + 1];
+
t = btf_type_by_id(btf, args[i].type);
while (btf_type_is_modifier(t))
t = btf_type_by_id(btf, t->type);
if (btf_type_is_int(t) || btf_type_is_enum(t)) {
- reg[i + 1].type = SCALAR_VALUE;
+ reg->type = SCALAR_VALUE;
continue;
}
- if (btf_type_is_ptr(t) &&
- btf_get_prog_ctx_type(log, btf, t, prog_type, i)) {
- reg[i + 1].type = PTR_TO_CTX;
+ if (btf_type_is_ptr(t)) {
+ if (btf_get_prog_ctx_type(log, btf, t, prog_type, i)) {
+ reg->type = PTR_TO_CTX;
+ continue;
+ }
+
+ t = btf_type_skip_modifiers(btf, t->type, NULL);
+
+ ref_t = btf_resolve_size(btf, t, &reg->mem_size);
+ if (IS_ERR(ref_t)) {
+ bpf_log(log,
+ "arg#%d reference type('%s %s') size cannot be determined: %ld\n",
+ i, btf_type_str(t), btf_name_by_offset(btf, t->name_off),
+ PTR_ERR(ref_t));
+ return -EINVAL;
+ }
+
+ reg->type = PTR_TO_MEM_OR_NULL;
+ reg->id = ++env->id_gen;
+
continue;
}
bpf_log(log, "Arg#%d type %s in %s() is not supported yet.\n",
diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index 6aa9e10c6335..b567ca46555c 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -19,7 +19,7 @@
#include "../cgroup/cgroup-internal.h"
-DEFINE_STATIC_KEY_FALSE(cgroup_bpf_enabled_key);
+DEFINE_STATIC_KEY_ARRAY_FALSE(cgroup_bpf_enabled_key, MAX_BPF_ATTACH_TYPE);
EXPORT_SYMBOL(cgroup_bpf_enabled_key);
void cgroup_bpf_offline(struct cgroup *cgrp)
@@ -128,7 +128,7 @@ static void cgroup_bpf_release(struct work_struct *work)
if (pl->link)
bpf_cgroup_link_auto_detach(pl->link);
kfree(pl);
- static_branch_dec(&cgroup_bpf_enabled_key);
+ static_branch_dec(&cgroup_bpf_enabled_key[type]);
}
old_array = rcu_dereference_protected(
cgrp->bpf.effective[type],
@@ -499,7 +499,7 @@ int __cgroup_bpf_attach(struct cgroup *cgrp,
if (old_prog)
bpf_prog_put(old_prog);
else
- static_branch_inc(&cgroup_bpf_enabled_key);
+ static_branch_inc(&cgroup_bpf_enabled_key[type]);
bpf_cgroup_storages_link(new_storage, cgrp, type);
return 0;
@@ -698,7 +698,7 @@ int __cgroup_bpf_detach(struct cgroup *cgrp, struct bpf_prog *prog,
cgrp->bpf.flags[type] = 0;
if (old_prog)
bpf_prog_put(old_prog);
- static_branch_dec(&cgroup_bpf_enabled_key);
+ static_branch_dec(&cgroup_bpf_enabled_key[type]);
return 0;
cleanup:
@@ -1055,6 +1055,8 @@ EXPORT_SYMBOL(__cgroup_bpf_run_filter_sk);
* @uaddr: sockaddr struct provided by user
* @type: The type of program to be exectuted
* @t_ctx: Pointer to attach type specific context
+ * @flags: Pointer to u32 which contains higher bits of BPF program
+ * return value (OR'ed together).
*
* socket is expected to be of type INET or INET6.
*
@@ -1064,7 +1066,8 @@ EXPORT_SYMBOL(__cgroup_bpf_run_filter_sk);
int __cgroup_bpf_run_filter_sock_addr(struct sock *sk,
struct sockaddr *uaddr,
enum bpf_attach_type type,
- void *t_ctx)
+ void *t_ctx,
+ u32 *flags)
{
struct bpf_sock_addr_kern ctx = {
.sk = sk,
@@ -1087,7 +1090,8 @@ int __cgroup_bpf_run_filter_sock_addr(struct sock *sk,
}
cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
- ret = BPF_PROG_RUN_ARRAY(cgrp->bpf.effective[type], &ctx, BPF_PROG_RUN);
+ ret = BPF_PROG_RUN_ARRAY_FLAGS(cgrp->bpf.effective[type], &ctx,
+ BPF_PROG_RUN, flags);
return ret == 1 ? 0 : -EPERM;
}
@@ -1298,7 +1302,8 @@ static bool __cgroup_bpf_prog_array_is_empty(struct cgroup *cgrp,
return empty;
}
-static int sockopt_alloc_buf(struct bpf_sockopt_kern *ctx, int max_optlen)
+static int sockopt_alloc_buf(struct bpf_sockopt_kern *ctx, int max_optlen,
+ struct bpf_sockopt_buf *buf)
{
if (unlikely(max_optlen < 0))
return -EINVAL;
@@ -1310,6 +1315,15 @@ static int sockopt_alloc_buf(struct bpf_sockopt_kern *ctx, int max_optlen)
max_optlen = PAGE_SIZE;
}
+ if (max_optlen <= sizeof(buf->data)) {
+ /* When the optval fits into BPF_SOCKOPT_KERN_BUF_SIZE
+ * bytes avoid the cost of kzalloc.
+ */
+ ctx->optval = buf->data;
+ ctx->optval_end = ctx->optval + max_optlen;
+ return max_optlen;
+ }
+
ctx->optval = kzalloc(max_optlen, GFP_USER);
if (!ctx->optval)
return -ENOMEM;
@@ -1319,16 +1333,26 @@ static int sockopt_alloc_buf(struct bpf_sockopt_kern *ctx, int max_optlen)
return max_optlen;
}
-static void sockopt_free_buf(struct bpf_sockopt_kern *ctx)
+static void sockopt_free_buf(struct bpf_sockopt_kern *ctx,
+ struct bpf_sockopt_buf *buf)
{
+ if (ctx->optval == buf->data)
+ return;
kfree(ctx->optval);
}
+static bool sockopt_buf_allocated(struct bpf_sockopt_kern *ctx,
+ struct bpf_sockopt_buf *buf)
+{
+ return ctx->optval != buf->data;
+}
+
int __cgroup_bpf_run_filter_setsockopt(struct sock *sk, int *level,
int *optname, char __user *optval,
int *optlen, char **kernel_optval)
{
struct cgroup *cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
+ struct bpf_sockopt_buf buf = {};
struct bpf_sockopt_kern ctx = {
.sk = sk,
.level = *level,
@@ -1340,8 +1364,7 @@ int __cgroup_bpf_run_filter_setsockopt(struct sock *sk, int *level,
* attached to the hook so we don't waste time allocating
* memory and locking the socket.
*/
- if (!cgroup_bpf_enabled ||
- __cgroup_bpf_prog_array_is_empty(cgrp, BPF_CGROUP_SETSOCKOPT))
+ if (__cgroup_bpf_prog_array_is_empty(cgrp, BPF_CGROUP_SETSOCKOPT))
return 0;
/* Allocate a bit more than the initial user buffer for
@@ -1350,7 +1373,7 @@ int __cgroup_bpf_run_filter_setsockopt(struct sock *sk, int *level,
*/
max_optlen = max_t(int, 16, *optlen);
- max_optlen = sockopt_alloc_buf(&ctx, max_optlen);
+ max_optlen = sockopt_alloc_buf(&ctx, max_optlen, &buf);
if (max_optlen < 0)
return max_optlen;
@@ -1390,14 +1413,31 @@ int __cgroup_bpf_run_filter_setsockopt(struct sock *sk, int *level,
*/
if (ctx.optlen != 0) {
*optlen = ctx.optlen;
- *kernel_optval = ctx.optval;
+ /* We've used bpf_sockopt_kern->buf as an intermediary
+ * storage, but the BPF program indicates that we need
+ * to pass this data to the kernel setsockopt handler.
+ * No way to export on-stack buf, have to allocate a
+ * new buffer.
+ */
+ if (!sockopt_buf_allocated(&ctx, &buf)) {
+ void *p = kmalloc(ctx.optlen, GFP_USER);
+
+ if (!p) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ memcpy(p, ctx.optval, ctx.optlen);
+ *kernel_optval = p;
+ } else {
+ *kernel_optval = ctx.optval;
+ }
/* export and don't free sockopt buf */
return 0;
}
}
out:
- sockopt_free_buf(&ctx);
+ sockopt_free_buf(&ctx, &buf);
return ret;
}
@@ -1407,6 +1447,7 @@ int __cgroup_bpf_run_filter_getsockopt(struct sock *sk, int level,
int retval)
{
struct cgroup *cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
+ struct bpf_sockopt_buf buf = {};
struct bpf_sockopt_kern ctx = {
.sk = sk,
.level = level,
@@ -1419,13 +1460,12 @@ int __cgroup_bpf_run_filter_getsockopt(struct sock *sk, int level,
* attached to the hook so we don't waste time allocating
* memory and locking the socket.
*/
- if (!cgroup_bpf_enabled ||
- __cgroup_bpf_prog_array_is_empty(cgrp, BPF_CGROUP_GETSOCKOPT))
+ if (__cgroup_bpf_prog_array_is_empty(cgrp, BPF_CGROUP_GETSOCKOPT))
return retval;
ctx.optlen = max_optlen;
- max_optlen = sockopt_alloc_buf(&ctx, max_optlen);
+ max_optlen = sockopt_alloc_buf(&ctx, max_optlen, &buf);
if (max_optlen < 0)
return max_optlen;
@@ -1488,9 +1528,55 @@ int __cgroup_bpf_run_filter_getsockopt(struct sock *sk, int level,
ret = ctx.retval;
out:
- sockopt_free_buf(&ctx);
+ sockopt_free_buf(&ctx, &buf);
return ret;
}
+
+int __cgroup_bpf_run_filter_getsockopt_kern(struct sock *sk, int level,
+ int optname, void *optval,
+ int *optlen, int retval)
+{
+ struct cgroup *cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data);
+ struct bpf_sockopt_kern ctx = {
+ .sk = sk,
+ .level = level,
+ .optname = optname,
+ .retval = retval,
+ .optlen = *optlen,
+ .optval = optval,
+ .optval_end = optval + *optlen,
+ };
+ int ret;
+
+ /* Note that __cgroup_bpf_run_filter_getsockopt doesn't copy
+ * user data back into BPF buffer when reval != 0. This is
+ * done as an optimization to avoid extra copy, assuming
+ * kernel won't populate the data in case of an error.
+ * Here we always pass the data and memset() should
+ * be called if that data shouldn't be "exported".
+ */
+
+ ret = BPF_PROG_RUN_ARRAY(cgrp->bpf.effective[BPF_CGROUP_GETSOCKOPT],
+ &ctx, BPF_PROG_RUN);
+ if (!ret)
+ return -EPERM;
+
+ if (ctx.optlen > *optlen)
+ return -EFAULT;
+
+ /* BPF programs only allowed to set retval to 0, not some
+ * arbitrary value.
+ */
+ if (ctx.retval != 0 && ctx.retval != retval)
+ return -EFAULT;
+
+ /* BPF programs can shrink the buffer, export the modifications.
+ */
+ if (ctx.optlen != 0)
+ *optlen = ctx.optlen;
+
+ return ctx.retval;
+}
#endif
static ssize_t sysctl_cpy_dir(const struct ctl_dir *dir, char **bufp,
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 5bbd4884ff7a..0ae015ad1e05 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -91,6 +91,12 @@ struct bpf_prog *bpf_prog_alloc_no_stats(unsigned int size, gfp_t gfp_extra_flag
vfree(fp);
return NULL;
}
+ fp->active = alloc_percpu_gfp(int, GFP_KERNEL_ACCOUNT | gfp_extra_flags);
+ if (!fp->active) {
+ vfree(fp);
+ kfree(aux);
+ return NULL;
+ }
fp->pages = size / PAGE_SIZE;
fp->aux = aux;
@@ -114,8 +120,9 @@ struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags)
if (!prog)
return NULL;
- prog->aux->stats = alloc_percpu_gfp(struct bpf_prog_stats, gfp_flags);
- if (!prog->aux->stats) {
+ prog->stats = alloc_percpu_gfp(struct bpf_prog_stats, gfp_flags);
+ if (!prog->stats) {
+ free_percpu(prog->active);
kfree(prog->aux);
vfree(prog);
return NULL;
@@ -124,7 +131,7 @@ struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags)
for_each_possible_cpu(cpu) {
struct bpf_prog_stats *pstats;
- pstats = per_cpu_ptr(prog->aux->stats, cpu);
+ pstats = per_cpu_ptr(prog->stats, cpu);
u64_stats_init(&pstats->syncp);
}
return prog;
@@ -238,6 +245,8 @@ struct bpf_prog *bpf_prog_realloc(struct bpf_prog *fp_old, unsigned int size,
* reallocated structure.
*/
fp_old->aux = NULL;
+ fp_old->stats = NULL;
+ fp_old->active = NULL;
__bpf_prog_free(fp_old);
}
@@ -249,10 +258,11 @@ void __bpf_prog_free(struct bpf_prog *fp)
if (fp->aux) {
mutex_destroy(&fp->aux->used_maps_mutex);
mutex_destroy(&fp->aux->dst_mutex);
- free_percpu(fp->aux->stats);
kfree(fp->aux->poke_tab);
kfree(fp->aux);
}
+ free_percpu(fp->stats);
+ free_percpu(fp->active);
vfree(fp);
}
diff --git a/kernel/bpf/cpumap.c b/kernel/bpf/cpumap.c
index 747313698178..5d1469de6921 100644
--- a/kernel/bpf/cpumap.c
+++ b/kernel/bpf/cpumap.c
@@ -141,49 +141,6 @@ static void cpu_map_kthread_stop(struct work_struct *work)
kthread_stop(rcpu->kthread);
}
-static struct sk_buff *cpu_map_build_skb(struct xdp_frame *xdpf,
- struct sk_buff *skb)
-{
- unsigned int hard_start_headroom;
- unsigned int frame_size;
- void *pkt_data_start;
-
- /* Part of headroom was reserved to xdpf */
- hard_start_headroom = sizeof(struct xdp_frame) + xdpf->headroom;
-
- /* Memory size backing xdp_frame data already have reserved
- * room for build_skb to place skb_shared_info in tailroom.
- */
- frame_size = xdpf->frame_sz;
-
- pkt_data_start = xdpf->data - hard_start_headroom;
- skb = build_skb_around(skb, pkt_data_start, frame_size);
- if (unlikely(!skb))
- return NULL;
-
- skb_reserve(skb, hard_start_headroom);
- __skb_put(skb, xdpf->len);
- if (xdpf->metasize)
- skb_metadata_set(skb, xdpf->metasize);
-
- /* Essential SKB info: protocol and skb->dev */
- skb->protocol = eth_type_trans(skb, xdpf->dev_rx);
-
- /* Optional SKB info, currently missing:
- * - HW checksum info (skb->ip_summed)
- * - HW RX hash (skb_set_hash)
- * - RX ring dev queue index (skb_record_rx_queue)
- */
-
- /* Until page_pool get SKB return path, release DMA here */
- xdp_release_frame(xdpf);
-
- /* Allow SKB to reuse area used by xdp_frame */
- xdp_scrub_frame(xdpf);
-
- return skb;
-}
-
static void __cpu_map_ring_cleanup(struct ptr_ring *ring)
{
/* The tear-down procedure should have made sure that queue is
@@ -350,7 +307,8 @@ static int cpu_map_kthread_run(void *data)
struct sk_buff *skb = skbs[i];
int ret;
- skb = cpu_map_build_skb(xdpf, skb);
+ skb = __xdp_build_skb_from_frame(xdpf, skb,
+ xdpf->dev_rx);
if (!skb) {
xdp_return_frame(xdpf);
continue;
diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
index f6e9c68afdd4..85d9d1b72a33 100644
--- a/kernel/bpf/devmap.c
+++ b/kernel/bpf/devmap.c
@@ -802,9 +802,7 @@ static int dev_map_notification(struct notifier_block *notifier,
break;
/* will be freed in free_netdev() */
- netdev->xdp_bulkq =
- __alloc_percpu_gfp(sizeof(struct xdp_dev_bulk_queue),
- sizeof(void *), GFP_ATOMIC);
+ netdev->xdp_bulkq = alloc_percpu(struct xdp_dev_bulk_queue);
if (!netdev->xdp_bulkq)
return NOTIFY_BAD;
diff --git a/kernel/bpf/disasm.c b/kernel/bpf/disasm.c
index 19ff8fed7f4b..3acc7e0b6916 100644
--- a/kernel/bpf/disasm.c
+++ b/kernel/bpf/disasm.c
@@ -161,7 +161,7 @@ void print_bpf_insn(const struct bpf_insn_cbs *cbs,
insn->dst_reg,
insn->off, insn->src_reg);
else if (BPF_MODE(insn->code) == BPF_ATOMIC &&
- (insn->imm == BPF_ADD || insn->imm == BPF_ADD ||
+ (insn->imm == BPF_ADD || insn->imm == BPF_AND ||
insn->imm == BPF_OR || insn->imm == BPF_XOR)) {
verbose(cbs->private_data, "(%02x) lock *(%s *)(r%d %+d) %s r%d\n",
insn->code,
diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index c1ac7f964bc9..d63912e73ad9 100644
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -1148,7 +1148,7 @@ static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key,
/* unknown flags */
return -EINVAL;
- WARN_ON_ONCE(!rcu_read_lock_held());
+ WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
key_size = map->key_size;
@@ -1202,7 +1202,7 @@ static int __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
/* unknown flags */
return -EINVAL;
- WARN_ON_ONCE(!rcu_read_lock_held());
+ WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
key_size = map->key_size;
diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index 41ca280b1dc1..308427fe03a3 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -720,14 +720,6 @@ bpf_base_func_proto(enum bpf_func_id func_id)
return &bpf_spin_lock_proto;
case BPF_FUNC_spin_unlock:
return &bpf_spin_unlock_proto;
- case BPF_FUNC_trace_printk:
- if (!perfmon_capable())
- return NULL;
- return bpf_get_trace_printk_proto();
- case BPF_FUNC_snprintf_btf:
- if (!perfmon_capable())
- return NULL;
- return &bpf_snprintf_btf_proto;
case BPF_FUNC_jiffies64:
return &bpf_jiffies64_proto;
case BPF_FUNC_per_cpu_ptr:
@@ -742,6 +734,8 @@ bpf_base_func_proto(enum bpf_func_id func_id)
return NULL;
switch (func_id) {
+ case BPF_FUNC_trace_printk:
+ return bpf_get_trace_printk_proto();
case BPF_FUNC_get_current_task:
return &bpf_get_current_task_proto;
case BPF_FUNC_probe_read_user:
@@ -752,6 +746,8 @@ bpf_base_func_proto(enum bpf_func_id func_id)
return &bpf_probe_read_user_str_proto;
case BPF_FUNC_probe_read_kernel_str:
return &bpf_probe_read_kernel_str_proto;
+ case BPF_FUNC_snprintf_btf:
+ return &bpf_snprintf_btf_proto;
default:
return NULL;
}
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index e5999d86c76e..c859bc46d06c 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -1731,25 +1731,28 @@ static int bpf_prog_release(struct inode *inode, struct file *filp)
static void bpf_prog_get_stats(const struct bpf_prog *prog,
struct bpf_prog_stats *stats)
{
- u64 nsecs = 0, cnt = 0;
+ u64 nsecs = 0, cnt = 0, misses = 0;
int cpu;
for_each_possible_cpu(cpu) {
const struct bpf_prog_stats *st;
unsigned int start;
- u64 tnsecs, tcnt;
+ u64 tnsecs, tcnt, tmisses;
- st = per_cpu_ptr(prog->aux->stats, cpu);
+ st = per_cpu_ptr(prog->stats, cpu);
do {
start = u64_stats_fetch_begin_irq(&st->syncp);
tnsecs = st->nsecs;
tcnt = st->cnt;
+ tmisses = st->misses;
} while (u64_stats_fetch_retry_irq(&st->syncp, start));
nsecs += tnsecs;
cnt += tcnt;
+ misses += tmisses;
}
stats->nsecs = nsecs;
stats->cnt = cnt;
+ stats->misses = misses;
}
#ifdef CONFIG_PROC_FS
@@ -1768,14 +1771,16 @@ static void bpf_prog_show_fdinfo(struct seq_file *m, struct file *filp)
"memlock:\t%llu\n"
"prog_id:\t%u\n"
"run_time_ns:\t%llu\n"
- "run_cnt:\t%llu\n",
+ "run_cnt:\t%llu\n"
+ "recursion_misses:\t%llu\n",
prog->type,
prog->jited,
prog_tag,
prog->pages * 1ULL << PAGE_SHIFT,
prog->aux->id,
stats.nsecs,
- stats.cnt);
+ stats.cnt,
+ stats.misses);
}
#endif
@@ -3438,6 +3443,7 @@ static int bpf_prog_get_info_by_fd(struct file *file,
bpf_prog_get_stats(prog, &stats);
info.run_time_ns = stats.nsecs;
info.run_cnt = stats.cnt;
+ info.recursion_misses = stats.misses;
if (!bpf_capable()) {
info.jited_prog_len = 0;
diff --git a/kernel/bpf/task_iter.c b/kernel/bpf/task_iter.c
index 175b7b42bfc4..b68cb5d6d6eb 100644
--- a/kernel/bpf/task_iter.c
+++ b/kernel/bpf/task_iter.c
@@ -286,9 +286,248 @@ static const struct seq_operations task_file_seq_ops = {
.show = task_file_seq_show,
};
+struct bpf_iter_seq_task_vma_info {
+ /* The first field must be struct bpf_iter_seq_task_common.
+ * this is assumed by {init, fini}_seq_pidns() callback functions.
+ */
+ struct bpf_iter_seq_task_common common;
+ struct task_struct *task;
+ struct vm_area_struct *vma;
+ u32 tid;
+ unsigned long prev_vm_start;
+ unsigned long prev_vm_end;
+};
+
+enum bpf_task_vma_iter_find_op {
+ task_vma_iter_first_vma, /* use mm->mmap */
+ task_vma_iter_next_vma, /* use curr_vma->vm_next */
+ task_vma_iter_find_vma, /* use find_vma() to find next vma */
+};
+
+static struct vm_area_struct *
+task_vma_seq_get_next(struct bpf_iter_seq_task_vma_info *info)
+{
+ struct pid_namespace *ns = info->common.ns;
+ enum bpf_task_vma_iter_find_op op;
+ struct vm_area_struct *curr_vma;
+ struct task_struct *curr_task;
+ u32 curr_tid = info->tid;
+
+ /* If this function returns a non-NULL vma, it holds a reference to
+ * the task_struct, and holds read lock on vma->mm->mmap_lock.
+ * If this function returns NULL, it does not hold any reference or
+ * lock.
+ */
+ if (info->task) {
+ curr_task = info->task;
+ curr_vma = info->vma;
+ /* In case of lock contention, drop mmap_lock to unblock
+ * the writer.
+ *
+ * After relock, call find(mm, prev_vm_end - 1) to find
+ * new vma to process.
+ *
+ * +------+------+-----------+
+ * | VMA1 | VMA2 | VMA3 |
+ * +------+------+-----------+
+ * | | | |
+ * 4k 8k 16k 400k
+ *
+ * For example, curr_vma == VMA2. Before unlock, we set
+ *
+ * prev_vm_start = 8k
+ * prev_vm_end = 16k
+ *
+ * There are a few cases:
+ *
+ * 1) VMA2 is freed, but VMA3 exists.
+ *
+ * find_vma() will return VMA3, just process VMA3.
+ *
+ * 2) VMA2 still exists.
+ *
+ * find_vma() will return VMA2, process VMA2->next.
+ *
+ * 3) no more vma in this mm.
+ *
+ * Process the next task.
+ *
+ * 4) find_vma() returns a different vma, VMA2'.
+ *
+ * 4.1) If VMA2 covers same range as VMA2', skip VMA2',
+ * because we already covered the range;
+ * 4.2) VMA2 and VMA2' covers different ranges, process
+ * VMA2'.
+ */
+ if (mmap_lock_is_contended(curr_task->mm)) {
+ info->prev_vm_start = curr_vma->vm_start;
+ info->prev_vm_end = curr_vma->vm_end;
+ op = task_vma_iter_find_vma;
+ mmap_read_unlock(curr_task->mm);
+ if (mmap_read_lock_killable(curr_task->mm))
+ goto finish;
+ } else {
+ op = task_vma_iter_next_vma;
+ }
+ } else {
+again:
+ curr_task = task_seq_get_next(ns, &curr_tid, true);
+ if (!curr_task) {
+ info->tid = curr_tid + 1;
+ goto finish;
+ }
+
+ if (curr_tid != info->tid) {
+ info->tid = curr_tid;
+ /* new task, process the first vma */
+ op = task_vma_iter_first_vma;
+ } else {
+ /* Found the same tid, which means the user space
+ * finished data in previous buffer and read more.
+ * We dropped mmap_lock before returning to user
+ * space, so it is necessary to use find_vma() to
+ * find the next vma to process.
+ */
+ op = task_vma_iter_find_vma;
+ }
+
+ if (!curr_task->mm)
+ goto next_task;
+
+ if (mmap_read_lock_killable(curr_task->mm))
+ goto finish;
+ }
+
+ switch (op) {
+ case task_vma_iter_first_vma:
+ curr_vma = curr_task->mm->mmap;
+ break;
+ case task_vma_iter_next_vma:
+ curr_vma = curr_vma->vm_next;
+ break;
+ case task_vma_iter_find_vma:
+ /* We dropped mmap_lock so it is necessary to use find_vma
+ * to find the next vma. This is similar to the mechanism
+ * in show_smaps_rollup().
+ */
+ curr_vma = find_vma(curr_task->mm, info->prev_vm_end - 1);
+ /* case 1) and 4.2) above just use curr_vma */
+
+ /* check for case 2) or case 4.1) above */
+ if (curr_vma &&
+ curr_vma->vm_start == info->prev_vm_start &&
+ curr_vma->vm_end == info->prev_vm_end)
+ curr_vma = curr_vma->vm_next;
+ break;
+ }
+ if (!curr_vma) {
+ /* case 3) above, or case 2) 4.1) with vma->next == NULL */
+ mmap_read_unlock(curr_task->mm);
+ goto next_task;
+ }
+ info->task = curr_task;
+ info->vma = curr_vma;
+ return curr_vma;
+
+next_task:
+ put_task_struct(curr_task);
+ info->task = NULL;
+ curr_tid++;
+ goto again;
+
+finish:
+ if (curr_task)
+ put_task_struct(curr_task);
+ info->task = NULL;
+ info->vma = NULL;
+ return NULL;
+}
+
+static void *task_vma_seq_start(struct seq_file *seq, loff_t *pos)
+{
+ struct bpf_iter_seq_task_vma_info *info = seq->private;
+ struct vm_area_struct *vma;
+
+ vma = task_vma_seq_get_next(info);
+ if (vma && *pos == 0)
+ ++*pos;
+
+ return vma;
+}
+
+static void *task_vma_seq_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+ struct bpf_iter_seq_task_vma_info *info = seq->private;
+
+ ++*pos;
+ return task_vma_seq_get_next(info);
+}
+
+struct bpf_iter__task_vma {
+ __bpf_md_ptr(struct bpf_iter_meta *, meta);
+ __bpf_md_ptr(struct task_struct *, task);
+ __bpf_md_ptr(struct vm_area_struct *, vma);
+};
+
+DEFINE_BPF_ITER_FUNC(task_vma, struct bpf_iter_meta *meta,
+ struct task_struct *task, struct vm_area_struct *vma)
+
+static int __task_vma_seq_show(struct seq_file *seq, bool in_stop)
+{
+ struct bpf_iter_seq_task_vma_info *info = seq->private;
+ struct bpf_iter__task_vma ctx;
+ struct bpf_iter_meta meta;
+ struct bpf_prog *prog;
+
+ meta.seq = seq;
+ prog = bpf_iter_get_info(&meta, in_stop);
+ if (!prog)
+ return 0;
+
+ ctx.meta = &meta;
+ ctx.task = info->task;
+ ctx.vma = info->vma;
+ return bpf_iter_run_prog(prog, &ctx);
+}
+
+static int task_vma_seq_show(struct seq_file *seq, void *v)
+{
+ return __task_vma_seq_show(seq, false);
+}
+
+static void task_vma_seq_stop(struct seq_file *seq, void *v)
+{
+ struct bpf_iter_seq_task_vma_info *info = seq->private;
+
+ if (!v) {
+ (void)__task_vma_seq_show(seq, true);
+ } else {
+ /* info->vma has not been seen by the BPF program. If the
+ * user space reads more, task_vma_seq_get_next should
+ * return this vma again. Set prev_vm_start to ~0UL,
+ * so that we don't skip the vma returned by the next
+ * find_vma() (case task_vma_iter_find_vma in
+ * task_vma_seq_get_next()).
+ */
+ info->prev_vm_start = ~0UL;
+ info->prev_vm_end = info->vma->vm_end;
+ mmap_read_unlock(info->task->mm);
+ put_task_struct(info->task);
+ info->task = NULL;
+ }
+}
+
+static const struct seq_operations task_vma_seq_ops = {
+ .start = task_vma_seq_start,
+ .next = task_vma_seq_next,
+ .stop = task_vma_seq_stop,
+ .show = task_vma_seq_show,
+};
+
BTF_ID_LIST(btf_task_file_ids)
BTF_ID(struct, task_struct)
BTF_ID(struct, file)
+BTF_ID(struct, vm_area_struct)
static const struct bpf_iter_seq_info task_seq_info = {
.seq_ops = &task_seq_ops,
@@ -328,6 +567,26 @@ static struct bpf_iter_reg task_file_reg_info = {
.seq_info = &task_file_seq_info,
};
+static const struct bpf_iter_seq_info task_vma_seq_info = {
+ .seq_ops = &task_vma_seq_ops,
+ .init_seq_private = init_seq_pidns,
+ .fini_seq_private = fini_seq_pidns,
+ .seq_priv_size = sizeof(struct bpf_iter_seq_task_vma_info),
+};
+
+static struct bpf_iter_reg task_vma_reg_info = {
+ .target = "task_vma",
+ .feature = BPF_ITER_RESCHED,
+ .ctx_arg_info_size = 2,
+ .ctx_arg_info = {
+ { offsetof(struct bpf_iter__task_vma, task),
+ PTR_TO_BTF_ID_OR_NULL },
+ { offsetof(struct bpf_iter__task_vma, vma),
+ PTR_TO_BTF_ID_OR_NULL },
+ },
+ .seq_info = &task_vma_seq_info,
+};
+
static int __init task_iter_init(void)
{
int ret;
@@ -339,6 +598,12 @@ static int __init task_iter_init(void)
task_file_reg_info.ctx_arg_info[0].btf_id = btf_task_file_ids[0];
task_file_reg_info.ctx_arg_info[1].btf_id = btf_task_file_ids[1];
- return bpf_iter_reg_target(&task_file_reg_info);
+ ret = bpf_iter_reg_target(&task_file_reg_info);
+ if (ret)
+ return ret;
+
+ task_vma_reg_info.ctx_arg_info[0].btf_id = btf_task_file_ids[0];
+ task_vma_reg_info.ctx_arg_info[1].btf_id = btf_task_file_ids[2];
+ return bpf_iter_reg_target(&task_vma_reg_info);
}
late_initcall(task_iter_init);
diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c
index 35c5887d82ff..7bc3b3209224 100644
--- a/kernel/bpf/trampoline.c
+++ b/kernel/bpf/trampoline.c
@@ -381,55 +381,100 @@ out:
mutex_unlock(&trampoline_mutex);
}
+#define NO_START_TIME 1
+static u64 notrace bpf_prog_start_time(void)
+{
+ u64 start = NO_START_TIME;
+
+ if (static_branch_unlikely(&bpf_stats_enabled_key)) {
+ start = sched_clock();
+ if (unlikely(!start))
+ start = NO_START_TIME;
+ }
+ return start;
+}
+
+static void notrace inc_misses_counter(struct bpf_prog *prog)
+{
+ struct bpf_prog_stats *stats;
+
+ stats = this_cpu_ptr(prog->stats);
+ u64_stats_update_begin(&stats->syncp);
+ stats->misses++;
+ u64_stats_update_end(&stats->syncp);
+}
+
/* The logic is similar to BPF_PROG_RUN, but with an explicit
* rcu_read_lock() and migrate_disable() which are required
* for the trampoline. The macro is split into
- * call _bpf_prog_enter
+ * call __bpf_prog_enter
* call prog->bpf_func
* call __bpf_prog_exit
+ *
+ * __bpf_prog_enter returns:
+ * 0 - skip execution of the bpf prog
+ * 1 - execute bpf prog
+ * [2..MAX_U64] - excute bpf prog and record execution time.
+ * This is start time.
*/
-u64 notrace __bpf_prog_enter(void)
+u64 notrace __bpf_prog_enter(struct bpf_prog *prog)
__acquires(RCU)
{
- u64 start = 0;
-
rcu_read_lock();
migrate_disable();
- if (static_branch_unlikely(&bpf_stats_enabled_key))
- start = sched_clock();
- return start;
+ if (unlikely(__this_cpu_inc_return(*(prog->active)) != 1)) {
+ inc_misses_counter(prog);
+ return 0;
+ }
+ return bpf_prog_start_time();
}
-void notrace __bpf_prog_exit(struct bpf_prog *prog, u64 start)
- __releases(RCU)
+static void notrace update_prog_stats(struct bpf_prog *prog,
+ u64 start)
{
struct bpf_prog_stats *stats;
if (static_branch_unlikely(&bpf_stats_enabled_key) &&
- /* static_key could be enabled in __bpf_prog_enter
- * and disabled in __bpf_prog_exit.
+ /* static_key could be enabled in __bpf_prog_enter*
+ * and disabled in __bpf_prog_exit*.
* And vice versa.
- * Hence check that 'start' is not zero.
+ * Hence check that 'start' is valid.
*/
- start) {
- stats = this_cpu_ptr(prog->aux->stats);
+ start > NO_START_TIME) {
+ stats = this_cpu_ptr(prog->stats);
u64_stats_update_begin(&stats->syncp);
stats->cnt++;
stats->nsecs += sched_clock() - start;
u64_stats_update_end(&stats->syncp);
}
+}
+
+void notrace __bpf_prog_exit(struct bpf_prog *prog, u64 start)
+ __releases(RCU)
+{
+ update_prog_stats(prog, start);
+ __this_cpu_dec(*(prog->active));
migrate_enable();
rcu_read_unlock();
}
-void notrace __bpf_prog_enter_sleepable(void)
+u64 notrace __bpf_prog_enter_sleepable(struct bpf_prog *prog)
{
rcu_read_lock_trace();
+ migrate_disable();
might_fault();
+ if (unlikely(__this_cpu_inc_return(*(prog->active)) != 1)) {
+ inc_misses_counter(prog);
+ return 0;
+ }
+ return bpf_prog_start_time();
}
-void notrace __bpf_prog_exit_sleepable(void)
+void notrace __bpf_prog_exit_sleepable(struct bpf_prog *prog, u64 start)
{
+ update_prog_stats(prog, start);
+ __this_cpu_dec(*(prog->active));
+ migrate_enable();
rcu_read_unlock_trace();
}
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 1cffd4e84725..36d1e7339ede 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -228,6 +228,12 @@ static void bpf_map_key_store(struct bpf_insn_aux_data *aux, u64 state)
(poisoned ? BPF_MAP_KEY_POISON : 0ULL);
}
+static bool bpf_pseudo_call(const struct bpf_insn *insn)
+{
+ return insn->code == (BPF_JMP | BPF_CALL) &&
+ insn->src_reg == BPF_PSEUDO_CALL;
+}
+
struct bpf_call_arg_meta {
struct bpf_map *map_ptr;
bool raw_mode;
@@ -1073,6 +1079,51 @@ static void mark_reg_known_zero(struct bpf_verifier_env *env,
__mark_reg_known_zero(regs + regno);
}
+static void mark_ptr_not_null_reg(struct bpf_reg_state *reg)
+{
+ switch (reg->type) {
+ case PTR_TO_MAP_VALUE_OR_NULL: {
+ const struct bpf_map *map = reg->map_ptr;
+
+ if (map->inner_map_meta) {
+ reg->type = CONST_PTR_TO_MAP;
+ reg->map_ptr = map->inner_map_meta;
+ } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) {
+ reg->type = PTR_TO_XDP_SOCK;
+ } else if (map->map_type == BPF_MAP_TYPE_SOCKMAP ||
+ map->map_type == BPF_MAP_TYPE_SOCKHASH) {
+ reg->type = PTR_TO_SOCKET;
+ } else {
+ reg->type = PTR_TO_MAP_VALUE;
+ }
+ break;
+ }
+ case PTR_TO_SOCKET_OR_NULL:
+ reg->type = PTR_TO_SOCKET;
+ break;
+ case PTR_TO_SOCK_COMMON_OR_NULL:
+ reg->type = PTR_TO_SOCK_COMMON;
+ break;
+ case PTR_TO_TCP_SOCK_OR_NULL:
+ reg->type = PTR_TO_TCP_SOCK;
+ break;
+ case PTR_TO_BTF_ID_OR_NULL:
+ reg->type = PTR_TO_BTF_ID;
+ break;
+ case PTR_TO_MEM_OR_NULL:
+ reg->type = PTR_TO_MEM;
+ break;
+ case PTR_TO_RDONLY_BUF_OR_NULL:
+ reg->type = PTR_TO_RDONLY_BUF;
+ break;
+ case PTR_TO_RDWR_BUF_OR_NULL:
+ reg->type = PTR_TO_RDWR_BUF;
+ break;
+ default:
+ WARN_ON("unknown nullable register type");
+ }
+}
+
static bool reg_is_pkt_pointer(const struct bpf_reg_state *reg)
{
return type_is_pkt_pointer(reg->type);
@@ -1486,9 +1537,7 @@ static int check_subprogs(struct bpf_verifier_env *env)
/* determine subprog starts. The end is one before the next starts */
for (i = 0; i < insn_cnt; i++) {
- if (insn[i].code != (BPF_JMP | BPF_CALL))
- continue;
- if (insn[i].src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn + i))
continue;
if (!env->bpf_capable) {
verbose(env,
@@ -2271,12 +2320,14 @@ static void save_register_state(struct bpf_func_state *state,
state->stack[spi].slot_type[i] = STACK_SPILL;
}
-/* check_stack_read/write functions track spill/fill of registers,
+/* check_stack_{read,write}_fixed_off functions track spill/fill of registers,
* stack boundary and alignment are checked in check_mem_access()
*/
-static int check_stack_write(struct bpf_verifier_env *env,
- struct bpf_func_state *state, /* func where register points to */
- int off, int size, int value_regno, int insn_idx)
+static int check_stack_write_fixed_off(struct bpf_verifier_env *env,
+ /* stack frame we're writing to */
+ struct bpf_func_state *state,
+ int off, int size, int value_regno,
+ int insn_idx)
{
struct bpf_func_state *cur; /* state of the current function */
int i, slot = -off - 1, spi = slot / BPF_REG_SIZE, err;
@@ -2402,9 +2453,175 @@ static int check_stack_write(struct bpf_verifier_env *env,
return 0;
}
-static int check_stack_read(struct bpf_verifier_env *env,
- struct bpf_func_state *reg_state /* func where register points to */,
- int off, int size, int value_regno)
+/* Write the stack: 'stack[ptr_regno + off] = value_regno'. 'ptr_regno' is
+ * known to contain a variable offset.
+ * This function checks whether the write is permitted and conservatively
+ * tracks the effects of the write, considering that each stack slot in the
+ * dynamic range is potentially written to.
+ *
+ * 'off' includes 'regno->off'.
+ * 'value_regno' can be -1, meaning that an unknown value is being written to
+ * the stack.
+ *
+ * Spilled pointers in range are not marked as written because we don't know
+ * what's going to be actually written. This means that read propagation for
+ * future reads cannot be terminated by this write.
+ *
+ * For privileged programs, uninitialized stack slots are considered
+ * initialized by this write (even though we don't know exactly what offsets
+ * are going to be written to). The idea is that we don't want the verifier to
+ * reject future reads that access slots written to through variable offsets.
+ */
+static int check_stack_write_var_off(struct bpf_verifier_env *env,
+ /* func where register points to */
+ struct bpf_func_state *state,
+ int ptr_regno, int off, int size,
+ int value_regno, int insn_idx)
+{
+ struct bpf_func_state *cur; /* state of the current function */
+ int min_off, max_off;
+ int i, err;
+ struct bpf_reg_state *ptr_reg = NULL, *value_reg = NULL;
+ bool writing_zero = false;
+ /* set if the fact that we're writing a zero is used to let any
+ * stack slots remain STACK_ZERO
+ */
+ bool zero_used = false;
+
+ cur = env->cur_state->frame[env->cur_state->curframe];
+ ptr_reg = &cur->regs[ptr_regno];
+ min_off = ptr_reg->smin_value + off;
+ max_off = ptr_reg->smax_value + off + size;
+ if (value_regno >= 0)
+ value_reg = &cur->regs[value_regno];
+ if (value_reg && register_is_null(value_reg))
+ writing_zero = true;
+
+ err = realloc_func_state(state, round_up(-min_off, BPF_REG_SIZE),
+ state->acquired_refs, true);
+ if (err)
+ return err;
+
+
+ /* Variable offset writes destroy any spilled pointers in range. */
+ for (i = min_off; i < max_off; i++) {
+ u8 new_type, *stype;
+ int slot, spi;
+
+ slot = -i - 1;
+ spi = slot / BPF_REG_SIZE;
+ stype = &state->stack[spi].slot_type[slot % BPF_REG_SIZE];
+
+ if (!env->allow_ptr_leaks
+ && *stype != NOT_INIT
+ && *stype != SCALAR_VALUE) {
+ /* Reject the write if there's are spilled pointers in
+ * range. If we didn't reject here, the ptr status
+ * would be erased below (even though not all slots are
+ * actually overwritten), possibly opening the door to
+ * leaks.
+ */
+ verbose(env, "spilled ptr in range of var-offset stack write; insn %d, ptr off: %d",
+ insn_idx, i);
+ return -EINVAL;
+ }
+
+ /* Erase all spilled pointers. */
+ state->stack[spi].spilled_ptr.type = NOT_INIT;
+
+ /* Update the slot type. */
+ new_type = STACK_MISC;
+ if (writing_zero && *stype == STACK_ZERO) {
+ new_type = STACK_ZERO;
+ zero_used = true;
+ }
+ /* If the slot is STACK_INVALID, we check whether it's OK to
+ * pretend that it will be initialized by this write. The slot
+ * might not actually be written to, and so if we mark it as
+ * initialized future reads might leak uninitialized memory.
+ * For privileged programs, we will accept such reads to slots
+ * that may or may not be written because, if we're reject
+ * them, the error would be too confusing.
+ */
+ if (*stype == STACK_INVALID && !env->allow_uninit_stack) {
+ verbose(env, "uninit stack in range of var-offset write prohibited for !root; insn %d, off: %d",
+ insn_idx, i);
+ return -EINVAL;
+ }
+ *stype = new_type;
+ }
+ if (zero_used) {
+ /* backtracking doesn't work for STACK_ZERO yet. */
+ err = mark_chain_precision(env, value_regno);
+ if (err)
+ return err;
+ }
+ return 0;
+}
+
+/* When register 'dst_regno' is assigned some values from stack[min_off,
+ * max_off), we set the register's type according to the types of the
+ * respective stack slots. If all the stack values are known to be zeros, then
+ * so is the destination reg. Otherwise, the register is considered to be
+ * SCALAR. This function does not deal with register filling; the caller must
+ * ensure that all spilled registers in the stack range have been marked as
+ * read.
+ */
+static void mark_reg_stack_read(struct bpf_verifier_env *env,
+ /* func where src register points to */
+ struct bpf_func_state *ptr_state,
+ int min_off, int max_off, int dst_regno)
+{
+ struct bpf_verifier_state *vstate = env->cur_state;
+ struct bpf_func_state *state = vstate->frame[vstate->curframe];
+ int i, slot, spi;
+ u8 *stype;
+ int zeros = 0;
+
+ for (i = min_off; i < max_off; i++) {
+ slot = -i - 1;
+ spi = slot / BPF_REG_SIZE;
+ stype = ptr_state->stack[spi].slot_type;
+ if (stype[slot % BPF_REG_SIZE] != STACK_ZERO)
+ break;
+ zeros++;
+ }
+ if (zeros == max_off - min_off) {
+ /* any access_size read into register is zero extended,
+ * so the whole register == const_zero
+ */
+ __mark_reg_const_zero(&state->regs[dst_regno]);
+ /* backtracking doesn't support STACK_ZERO yet,
+ * so mark it precise here, so that later
+ * backtracking can stop here.
+ * Backtracking may not need this if this register
+ * doesn't participate in pointer adjustment.
+ * Forward propagation of precise flag is not
+ * necessary either. This mark is only to stop
+ * backtracking. Any register that contributed
+ * to const 0 was marked precise before spill.
+ */
+ state->regs[dst_regno].precise = true;
+ } else {
+ /* have read misc data from the stack */
+ mark_reg_unknown(env, state->regs, dst_regno);
+ }
+ state->regs[dst_regno].live |= REG_LIVE_WRITTEN;
+}
+
+/* Read the stack at 'off' and put the results into the register indicated by
+ * 'dst_regno'. It handles reg filling if the addressed stack slot is a
+ * spilled reg.
+ *
+ * 'dst_regno' can be -1, meaning that the read value is not going to a
+ * register.
+ *
+ * The access is assumed to be within the current stack bounds.
+ */
+static int check_stack_read_fixed_off(struct bpf_verifier_env *env,
+ /* func where src register points to */
+ struct bpf_func_state *reg_state,
+ int off, int size, int dst_regno)
{
struct bpf_verifier_state *vstate = env->cur_state;
struct bpf_func_state *state = vstate->frame[vstate->curframe];
@@ -2412,11 +2629,6 @@ static int check_stack_read(struct bpf_verifier_env *env,
struct bpf_reg_state *reg;
u8 *stype;
- if (reg_state->allocated_stack <= slot) {
- verbose(env, "invalid read from stack off %d+0 size %d\n",
- off, size);
- return -EACCES;
- }
stype = reg_state->stack[spi].slot_type;
reg = &reg_state->stack[spi].spilled_ptr;
@@ -2427,9 +2639,9 @@ static int check_stack_read(struct bpf_verifier_env *env,
verbose(env, "invalid size of register fill\n");
return -EACCES;
}
- if (value_regno >= 0) {
- mark_reg_unknown(env, state->regs, value_regno);
- state->regs[value_regno].live |= REG_LIVE_WRITTEN;
+ if (dst_regno >= 0) {
+ mark_reg_unknown(env, state->regs, dst_regno);
+ state->regs[dst_regno].live |= REG_LIVE_WRITTEN;
}
mark_reg_read(env, reg, reg->parent, REG_LIVE_READ64);
return 0;
@@ -2441,16 +2653,16 @@ static int check_stack_read(struct bpf_verifier_env *env,
}
}
- if (value_regno >= 0) {
+ if (dst_regno >= 0) {
/* restore register state from stack */
- state->regs[value_regno] = *reg;
+ state->regs[dst_regno] = *reg;
/* mark reg as written since spilled pointer state likely
* has its liveness marks cleared by is_state_visited()
* which resets stack/reg liveness for state transitions
*/
- state->regs[value_regno].live |= REG_LIVE_WRITTEN;
+ state->regs[dst_regno].live |= REG_LIVE_WRITTEN;
} else if (__is_pointer_value(env->allow_ptr_leaks, reg)) {
- /* If value_regno==-1, the caller is asking us whether
+ /* If dst_regno==-1, the caller is asking us whether
* it is acceptable to use this value as a SCALAR_VALUE
* (e.g. for XADD).
* We must not allow unprivileged callers to do that
@@ -2462,70 +2674,167 @@ static int check_stack_read(struct bpf_verifier_env *env,
}
mark_reg_read(env, reg, reg->parent, REG_LIVE_READ64);
} else {
- int zeros = 0;
+ u8 type;
for (i = 0; i < size; i++) {
- if (stype[(slot - i) % BPF_REG_SIZE] == STACK_MISC)
+ type = stype[(slot - i) % BPF_REG_SIZE];
+ if (type == STACK_MISC)
continue;
- if (stype[(slot - i) % BPF_REG_SIZE] == STACK_ZERO) {
- zeros++;
+ if (type == STACK_ZERO)
continue;
- }
verbose(env, "invalid read from stack off %d+%d size %d\n",
off, i, size);
return -EACCES;
}
mark_reg_read(env, reg, reg->parent, REG_LIVE_READ64);
- if (value_regno >= 0) {
- if (zeros == size) {
- /* any size read into register is zero extended,
- * so the whole register == const_zero
- */
- __mark_reg_const_zero(&state->regs[value_regno]);
- /* backtracking doesn't support STACK_ZERO yet,
- * so mark it precise here, so that later
- * backtracking can stop here.
- * Backtracking may not need this if this register
- * doesn't participate in pointer adjustment.
- * Forward propagation of precise flag is not
- * necessary either. This mark is only to stop
- * backtracking. Any register that contributed
- * to const 0 was marked precise before spill.
- */
- state->regs[value_regno].precise = true;
- } else {
- /* have read misc data from the stack */
- mark_reg_unknown(env, state->regs, value_regno);
- }
- state->regs[value_regno].live |= REG_LIVE_WRITTEN;
- }
+ if (dst_regno >= 0)
+ mark_reg_stack_read(env, reg_state, off, off + size, dst_regno);
}
return 0;
}
-static int check_stack_access(struct bpf_verifier_env *env,
- const struct bpf_reg_state *reg,
- int off, int size)
+enum stack_access_src {
+ ACCESS_DIRECT = 1, /* the access is performed by an instruction */
+ ACCESS_HELPER = 2, /* the access is performed by a helper */
+};
+
+static int check_stack_range_initialized(struct bpf_verifier_env *env,
+ int regno, int off, int access_size,
+ bool zero_size_allowed,
+ enum stack_access_src type,
+ struct bpf_call_arg_meta *meta);
+
+static struct bpf_reg_state *reg_state(struct bpf_verifier_env *env, int regno)
+{
+ return cur_regs(env) + regno;
+}
+
+/* Read the stack at 'ptr_regno + off' and put the result into the register
+ * 'dst_regno'.
+ * 'off' includes the pointer register's fixed offset(i.e. 'ptr_regno.off'),
+ * but not its variable offset.
+ * 'size' is assumed to be <= reg size and the access is assumed to be aligned.
+ *
+ * As opposed to check_stack_read_fixed_off, this function doesn't deal with
+ * filling registers (i.e. reads of spilled register cannot be detected when
+ * the offset is not fixed). We conservatively mark 'dst_regno' as containing
+ * SCALAR_VALUE. That's why we assert that the 'ptr_regno' has a variable
+ * offset; for a fixed offset check_stack_read_fixed_off should be used
+ * instead.
+ */
+static int check_stack_read_var_off(struct bpf_verifier_env *env,
+ int ptr_regno, int off, int size, int dst_regno)
{
- /* Stack accesses must be at a fixed offset, so that we
- * can determine what type of data were returned. See
- * check_stack_read().
+ /* The state of the source register. */
+ struct bpf_reg_state *reg = reg_state(env, ptr_regno);
+ struct bpf_func_state *ptr_state = func(env, reg);
+ int err;
+ int min_off, max_off;
+
+ /* Note that we pass a NULL meta, so raw access will not be permitted.
*/
- if (!tnum_is_const(reg->var_off)) {
+ err = check_stack_range_initialized(env, ptr_regno, off, size,
+ false, ACCESS_DIRECT, NULL);
+ if (err)
+ return err;
+
+ min_off = reg->smin_value + off;
+ max_off = reg->smax_value + off;
+ mark_reg_stack_read(env, ptr_state, min_off, max_off + size, dst_regno);
+ return 0;
+}
+
+/* check_stack_read dispatches to check_stack_read_fixed_off or
+ * check_stack_read_var_off.
+ *
+ * The caller must ensure that the offset falls within the allocated stack
+ * bounds.
+ *
+ * 'dst_regno' is a register which will receive the value from the stack. It
+ * can be -1, meaning that the read value is not going to a register.
+ */
+static int check_stack_read(struct bpf_verifier_env *env,
+ int ptr_regno, int off, int size,
+ int dst_regno)
+{
+ struct bpf_reg_state *reg = reg_state(env, ptr_regno);
+ struct bpf_func_state *state = func(env, reg);
+ int err;
+ /* Some accesses are only permitted with a static offset. */
+ bool var_off = !tnum_is_const(reg->var_off);
+
+ /* The offset is required to be static when reads don't go to a
+ * register, in order to not leak pointers (see
+ * check_stack_read_fixed_off).
+ */
+ if (dst_regno < 0 && var_off) {
char tn_buf[48];
tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
- verbose(env, "variable stack access var_off=%s off=%d size=%d\n",
+ verbose(env, "variable offset stack pointer cannot be passed into helper function; var_off=%s off=%d size=%d\n",
tn_buf, off, size);
return -EACCES;
}
+ /* Variable offset is prohibited for unprivileged mode for simplicity
+ * since it requires corresponding support in Spectre masking for stack
+ * ALU. See also retrieve_ptr_limit().
+ */
+ if (!env->bypass_spec_v1 && var_off) {
+ char tn_buf[48];
- if (off >= 0 || off < -MAX_BPF_STACK) {
- verbose(env, "invalid stack off=%d size=%d\n", off, size);
+ tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
+ verbose(env, "R%d variable offset stack access prohibited for !root, var_off=%s\n",
+ ptr_regno, tn_buf);
return -EACCES;
}
- return 0;
+ if (!var_off) {
+ off += reg->var_off.value;
+ err = check_stack_read_fixed_off(env, state, off, size,
+ dst_regno);
+ } else {
+ /* Variable offset stack reads need more conservative handling
+ * than fixed offset ones. Note that dst_regno >= 0 on this
+ * branch.
+ */
+ err = check_stack_read_var_off(env, ptr_regno, off, size,
+ dst_regno);
+ }
+ return err;
+}
+
+
+/* check_stack_write dispatches to check_stack_write_fixed_off or
+ * check_stack_write_var_off.
+ *
+ * 'ptr_regno' is the register used as a pointer into the stack.
+ * 'off' includes 'ptr_regno->off', but not its variable offset (if any).
+ * 'value_regno' is the register whose value we're writing to the stack. It can
+ * be -1, meaning that we're not writing from a register.
+ *
+ * The caller must ensure that the offset falls within the maximum stack size.
+ */
+static int check_stack_write(struct bpf_verifier_env *env,
+ int ptr_regno, int off, int size,
+ int value_regno, int insn_idx)
+{
+ struct bpf_reg_state *reg = reg_state(env, ptr_regno);
+ struct bpf_func_state *state = func(env, reg);
+ int err;
+
+ if (tnum_is_const(reg->var_off)) {
+ off += reg->var_off.value;
+ err = check_stack_write_fixed_off(env, state, off, size,
+ value_regno, insn_idx);
+ } else {
+ /* Variable offset stack reads need more conservative handling
+ * than fixed offset ones.
+ */
+ err = check_stack_write_var_off(env, state,
+ ptr_regno, off, size,
+ value_regno, insn_idx);
+ }
+ return err;
}
static int check_map_access_type(struct bpf_verifier_env *env, u32 regno,
@@ -2858,11 +3167,6 @@ static int check_sock_access(struct bpf_verifier_env *env, int insn_idx,
return -EACCES;
}
-static struct bpf_reg_state *reg_state(struct bpf_verifier_env *env, int regno)
-{
- return cur_regs(env) + regno;
-}
-
static bool is_pointer_value(struct bpf_verifier_env *env, int regno)
{
return __is_pointer_value(env->allow_ptr_leaks, reg_state(env, regno));
@@ -2981,8 +3285,8 @@ static int check_ptr_alignment(struct bpf_verifier_env *env,
break;
case PTR_TO_STACK:
pointer_desc = "stack ";
- /* The stack spill tracking logic in check_stack_write()
- * and check_stack_read() relies on stack accesses being
+ /* The stack spill tracking logic in check_stack_write_fixed_off()
+ * and check_stack_read_fixed_off() relies on stack accesses being
* aligned.
*/
strict = true;
@@ -3074,9 +3378,7 @@ process_func:
continue_func:
subprog_end = subprog[idx + 1].start;
for (; i < subprog_end; i++) {
- if (insn[i].code != (BPF_JMP | BPF_CALL))
- continue;
- if (insn[i].src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn + i))
continue;
/* remember insn and function to return to */
ret_insn[frame] = i + 1;
@@ -3400,6 +3702,91 @@ static int check_ptr_to_map_access(struct bpf_verifier_env *env,
return 0;
}
+/* Check that the stack access at the given offset is within bounds. The
+ * maximum valid offset is -1.
+ *
+ * The minimum valid offset is -MAX_BPF_STACK for writes, and
+ * -state->allocated_stack for reads.
+ */
+static int check_stack_slot_within_bounds(int off,
+ struct bpf_func_state *state,
+ enum bpf_access_type t)
+{
+ int min_valid_off;
+
+ if (t == BPF_WRITE)
+ min_valid_off = -MAX_BPF_STACK;
+ else
+ min_valid_off = -state->allocated_stack;
+
+ if (off < min_valid_off || off > -1)
+ return -EACCES;
+ return 0;
+}
+
+/* Check that the stack access at 'regno + off' falls within the maximum stack
+ * bounds.
+ *
+ * 'off' includes `regno->offset`, but not its dynamic part (if any).
+ */
+static int check_stack_access_within_bounds(
+ struct bpf_verifier_env *env,
+ int regno, int off, int access_size,
+ enum stack_access_src src, enum bpf_access_type type)
+{
+ struct bpf_reg_state *regs = cur_regs(env);
+ struct bpf_reg_state *reg = regs + regno;
+ struct bpf_func_state *state = func(env, reg);
+ int min_off, max_off;
+ int err;
+ char *err_extra;
+
+ if (src == ACCESS_HELPER)
+ /* We don't know if helpers are reading or writing (or both). */
+ err_extra = " indirect access to";
+ else if (type == BPF_READ)
+ err_extra = " read from";
+ else
+ err_extra = " write to";
+
+ if (tnum_is_const(reg->var_off)) {
+ min_off = reg->var_off.value + off;
+ if (access_size > 0)
+ max_off = min_off + access_size - 1;
+ else
+ max_off = min_off;
+ } else {
+ if (reg->smax_value >= BPF_MAX_VAR_OFF ||
+ reg->smin_value <= -BPF_MAX_VAR_OFF) {
+ verbose(env, "invalid unbounded variable-offset%s stack R%d\n",
+ err_extra, regno);
+ return -EACCES;
+ }
+ min_off = reg->smin_value + off;
+ if (access_size > 0)
+ max_off = reg->smax_value + off + access_size - 1;
+ else
+ max_off = min_off;
+ }
+
+ err = check_stack_slot_within_bounds(min_off, state, type);
+ if (!err)
+ err = check_stack_slot_within_bounds(max_off, state, type);
+
+ if (err) {
+ if (tnum_is_const(reg->var_off)) {
+ verbose(env, "invalid%s stack R%d off=%d size=%d\n",
+ err_extra, regno, off, access_size);
+ } else {
+ char tn_buf[48];
+
+ tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
+ verbose(env, "invalid variable-offset%s stack R%d var_off=%s size=%d\n",
+ err_extra, regno, tn_buf, access_size);
+ }
+ }
+ return err;
+}
/* check whether memory at (regno + off) is accessible for t = (read | write)
* if t==write, value_regno is a register which value is stored into memory
@@ -3515,8 +3902,8 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
}
} else if (reg->type == PTR_TO_STACK) {
- off += reg->var_off.value;
- err = check_stack_access(env, reg, off, size);
+ /* Basic bounds checks. */
+ err = check_stack_access_within_bounds(env, regno, off, size, ACCESS_DIRECT, t);
if (err)
return err;
@@ -3525,12 +3912,12 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
if (err)
return err;
- if (t == BPF_WRITE)
- err = check_stack_write(env, state, off, size,
- value_regno, insn_idx);
- else
- err = check_stack_read(env, state, off, size,
+ if (t == BPF_READ)
+ err = check_stack_read(env, regno, off, size,
value_regno);
+ else
+ err = check_stack_write(env, regno, off, size,
+ value_regno, insn_idx);
} else if (reg_is_pkt_pointer(reg)) {
if (t == BPF_WRITE && !may_access_direct_pkt_data(env, NULL, t)) {
verbose(env, "cannot write into packet\n");
@@ -3665,9 +4052,26 @@ static int check_atomic(struct bpf_verifier_env *env, int insn_idx, struct bpf_i
return -EACCES;
}
+ if (insn->imm & BPF_FETCH) {
+ if (insn->imm == BPF_CMPXCHG)
+ load_reg = BPF_REG_0;
+ else
+ load_reg = insn->src_reg;
+
+ /* check and record load of old value */
+ err = check_reg_arg(env, load_reg, DST_OP);
+ if (err)
+ return err;
+ } else {
+ /* This instruction accesses a memory location but doesn't
+ * actually load it into a register.
+ */
+ load_reg = -1;
+ }
+
/* check whether we can read the memory */
err = check_mem_access(env, insn_idx, insn->dst_reg, insn->off,
- BPF_SIZE(insn->code), BPF_READ, -1, true);
+ BPF_SIZE(insn->code), BPF_READ, load_reg, true);
if (err)
return err;
@@ -3677,65 +4081,56 @@ static int check_atomic(struct bpf_verifier_env *env, int insn_idx, struct bpf_i
if (err)
return err;
- if (!(insn->imm & BPF_FETCH))
- return 0;
-
- if (insn->imm == BPF_CMPXCHG)
- load_reg = BPF_REG_0;
- else
- load_reg = insn->src_reg;
-
- /* check and record load of old value */
- err = check_reg_arg(env, load_reg, DST_OP);
- if (err)
- return err;
-
return 0;
}
-static int __check_stack_boundary(struct bpf_verifier_env *env, u32 regno,
- int off, int access_size,
- bool zero_size_allowed)
+/* When register 'regno' is used to read the stack (either directly or through
+ * a helper function) make sure that it's within stack boundary and, depending
+ * on the access type, that all elements of the stack are initialized.
+ *
+ * 'off' includes 'regno->off', but not its dynamic part (if any).
+ *
+ * All registers that have been spilled on the stack in the slots within the
+ * read offsets are marked as read.
+ */
+static int check_stack_range_initialized(
+ struct bpf_verifier_env *env, int regno, int off,
+ int access_size, bool zero_size_allowed,
+ enum stack_access_src type, struct bpf_call_arg_meta *meta)
{
struct bpf_reg_state *reg = reg_state(env, regno);
+ struct bpf_func_state *state = func(env, reg);
+ int err, min_off, max_off, i, j, slot, spi;
+ char *err_extra = type == ACCESS_HELPER ? " indirect" : "";
+ enum bpf_access_type bounds_check_type;
+ /* Some accesses can write anything into the stack, others are
+ * read-only.
+ */
+ bool clobber = false;
- if (off >= 0 || off < -MAX_BPF_STACK || off + access_size > 0 ||
- access_size < 0 || (access_size == 0 && !zero_size_allowed)) {
- if (tnum_is_const(reg->var_off)) {
- verbose(env, "invalid stack type R%d off=%d access_size=%d\n",
- regno, off, access_size);
- } else {
- char tn_buf[48];
-
- tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
- verbose(env, "invalid stack type R%d var_off=%s access_size=%d\n",
- regno, tn_buf, access_size);
- }
+ if (access_size == 0 && !zero_size_allowed) {
+ verbose(env, "invalid zero-sized read\n");
return -EACCES;
}
- return 0;
-}
-/* when register 'regno' is passed into function that will read 'access_size'
- * bytes from that pointer, make sure that it's within stack boundary
- * and all elements of stack are initialized.
- * Unlike most pointer bounds-checking functions, this one doesn't take an
- * 'off' argument, so it has to add in reg->off itself.
- */
-static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
- int access_size, bool zero_size_allowed,
- struct bpf_call_arg_meta *meta)
-{
- struct bpf_reg_state *reg = reg_state(env, regno);
- struct bpf_func_state *state = func(env, reg);
- int err, min_off, max_off, i, j, slot, spi;
+ if (type == ACCESS_HELPER) {
+ /* The bounds checks for writes are more permissive than for
+ * reads. However, if raw_mode is not set, we'll do extra
+ * checks below.
+ */
+ bounds_check_type = BPF_WRITE;
+ clobber = true;
+ } else {
+ bounds_check_type = BPF_READ;
+ }
+ err = check_stack_access_within_bounds(env, regno, off, access_size,
+ type, bounds_check_type);
+ if (err)
+ return err;
+
if (tnum_is_const(reg->var_off)) {
- min_off = max_off = reg->var_off.value + reg->off;
- err = __check_stack_boundary(env, regno, min_off, access_size,
- zero_size_allowed);
- if (err)
- return err;
+ min_off = max_off = reg->var_off.value + off;
} else {
/* Variable offset is prohibited for unprivileged mode for
* simplicity since it requires corresponding support in
@@ -3746,8 +4141,8 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
char tn_buf[48];
tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
- verbose(env, "R%d indirect variable offset stack access prohibited for !root, var_off=%s\n",
- regno, tn_buf);
+ verbose(env, "R%d%s variable offset stack access prohibited for !root, var_off=%s\n",
+ regno, err_extra, tn_buf);
return -EACCES;
}
/* Only initialized buffer on stack is allowed to be accessed
@@ -3759,28 +4154,8 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
if (meta && meta->raw_mode)
meta = NULL;
- if (reg->smax_value >= BPF_MAX_VAR_OFF ||
- reg->smax_value <= -BPF_MAX_VAR_OFF) {
- verbose(env, "R%d unbounded indirect variable offset stack access\n",
- regno);
- return -EACCES;
- }
- min_off = reg->smin_value + reg->off;
- max_off = reg->smax_value + reg->off;
- err = __check_stack_boundary(env, regno, min_off, access_size,
- zero_size_allowed);
- if (err) {
- verbose(env, "R%d min value is outside of stack bound\n",
- regno);
- return err;
- }
- err = __check_stack_boundary(env, regno, max_off, access_size,
- zero_size_allowed);
- if (err) {
- verbose(env, "R%d max value is outside of stack bound\n",
- regno);
- return err;
- }
+ min_off = reg->smin_value + off;
+ max_off = reg->smax_value + off;
}
if (meta && meta->raw_mode) {
@@ -3800,8 +4175,10 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
if (*stype == STACK_MISC)
goto mark;
if (*stype == STACK_ZERO) {
- /* helper can write anything into the stack */
- *stype = STACK_MISC;
+ if (clobber) {
+ /* helper can write anything into the stack */
+ *stype = STACK_MISC;
+ }
goto mark;
}
@@ -3812,22 +4189,24 @@ static int check_stack_boundary(struct bpf_verifier_env *env, int regno,
if (state->stack[spi].slot_type[0] == STACK_SPILL &&
(state->stack[spi].spilled_ptr.type == SCALAR_VALUE ||
env->allow_ptr_leaks)) {
- __mark_reg_unknown(env, &state->stack[spi].spilled_ptr);
- for (j = 0; j < BPF_REG_SIZE; j++)
- state->stack[spi].slot_type[j] = STACK_MISC;
+ if (clobber) {
+ __mark_reg_unknown(env, &state->stack[spi].spilled_ptr);
+ for (j = 0; j < BPF_REG_SIZE; j++)
+ state->stack[spi].slot_type[j] = STACK_MISC;
+ }
goto mark;
}
err:
if (tnum_is_const(reg->var_off)) {
- verbose(env, "invalid indirect read from stack off %d+%d size %d\n",
- min_off, i - min_off, access_size);
+ verbose(env, "invalid%s read from stack R%d off %d+%d size %d\n",
+ err_extra, regno, min_off, i - min_off, access_size);
} else {
char tn_buf[48];
tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
- verbose(env, "invalid indirect read from stack var_off %s+%d size %d\n",
- tn_buf, i - min_off, access_size);
+ verbose(env, "invalid%s read from stack R%d var_off %s+%d size %d\n",
+ err_extra, regno, tn_buf, i - min_off, access_size);
}
return -EACCES;
mark:
@@ -3876,8 +4255,10 @@ static int check_helper_mem_access(struct bpf_verifier_env *env, int regno,
"rdwr",
&env->prog->aux->max_rdwr_access);
case PTR_TO_STACK:
- return check_stack_boundary(env, regno, access_size,
- zero_size_allowed, meta);
+ return check_stack_range_initialized(
+ env,
+ regno, reg->off, access_size,
+ zero_size_allowed, ACCESS_HELPER, meta);
default: /* scalar_value or invalid ptr */
/* Allow zero-byte read from NULL, regardless of pointer type */
if (zero_size_allowed && access_size == 0 &&
@@ -3891,6 +4272,29 @@ static int check_helper_mem_access(struct bpf_verifier_env *env, int regno,
}
}
+int check_mem_reg(struct bpf_verifier_env *env, struct bpf_reg_state *reg,
+ u32 regno, u32 mem_size)
+{
+ if (register_is_null(reg))
+ return 0;
+
+ if (reg_type_may_be_null(reg->type)) {
+ /* Assuming that the register contains a value check if the memory
+ * access is safe. Temporarily save and restore the register's state as
+ * the conversion shouldn't be visible to a caller.
+ */
+ const struct bpf_reg_state saved_reg = *reg;
+ int rv;
+
+ mark_ptr_not_null_reg(reg);
+ rv = check_helper_mem_access(env, regno, mem_size, true, NULL);
+ *reg = saved_reg;
+ return rv;
+ }
+
+ return check_helper_mem_access(env, regno, mem_size, true, NULL);
+}
+
/* Implementation details:
* bpf_map_lookup returns PTR_TO_MAP_VALUE_OR_NULL
* Two bpf_map_lookups (even with the same key) will have different reg->id.
@@ -4875,8 +5279,9 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn,
subprog);
clear_caller_saved_regs(env, caller->regs);
- /* All global functions return SCALAR_VALUE */
+ /* All global functions return a 64-bit SCALAR_VALUE */
mark_reg_unknown(env, caller->regs, BPF_REG_0);
+ caller->regs[BPF_REG_0].subreg_def = DEF_NOT_SUBREG;
/* continue with next insn after call */
return 0;
@@ -5541,6 +5946,41 @@ do_sim:
return !ret ? -EFAULT : 0;
}
+/* check that stack access falls within stack limits and that 'reg' doesn't
+ * have a variable offset.
+ *
+ * Variable offset is prohibited for unprivileged mode for simplicity since it
+ * requires corresponding support in Spectre masking for stack ALU. See also
+ * retrieve_ptr_limit().
+ *
+ *
+ * 'off' includes 'reg->off'.
+ */
+static int check_stack_access_for_ptr_arithmetic(
+ struct bpf_verifier_env *env,
+ int regno,
+ const struct bpf_reg_state *reg,
+ int off)
+{
+ if (!tnum_is_const(reg->var_off)) {
+ char tn_buf[48];
+
+ tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off);
+ verbose(env, "R%d variable stack access prohibited for !root, var_off=%s off=%d\n",
+ regno, tn_buf, off);
+ return -EACCES;
+ }
+
+ if (off >= 0 || off < -MAX_BPF_STACK) {
+ verbose(env, "R%d stack pointer arithmetic goes out of range, "
+ "prohibited for !root; off=%d\n", regno, off);
+ return -EACCES;
+ }
+
+ return 0;
+}
+
+
/* Handles arithmetic on a pointer and a scalar: computes new min/max and var_off.
* Caller should also handle BPF_MOV case separately.
* If we return -EACCES, caller may want to try again treating pointer as a
@@ -5784,10 +6224,9 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
"prohibited for !root\n", dst);
return -EACCES;
} else if (dst_reg->type == PTR_TO_STACK &&
- check_stack_access(env, dst_reg, dst_reg->off +
- dst_reg->var_off.value, 1)) {
- verbose(env, "R%d stack pointer arithmetic goes out of range, "
- "prohibited for !root\n", dst);
+ check_stack_access_for_ptr_arithmetic(
+ env, dst, dst_reg, dst_reg->off +
+ dst_reg->var_off.value)) {
return -EACCES;
}
}
@@ -6266,7 +6705,7 @@ static void scalar32_min_max_rsh(struct bpf_reg_state *dst_reg,
* 3) the signed bounds cross zero, so they tell us nothing
* about the result
* If the value in dst_reg is known nonnegative, then again the
- * unsigned bounts capture the signed bounds.
+ * unsigned bounds capture the signed bounds.
* Thus, in all cases it suffices to blow away our signed bounds
* and rely on inferring new ones from the unsigned bounds and
* var_off of the result.
@@ -6297,7 +6736,7 @@ static void scalar_min_max_rsh(struct bpf_reg_state *dst_reg,
* 3) the signed bounds cross zero, so they tell us nothing
* about the result
* If the value in dst_reg is known nonnegative, then again the
- * unsigned bounts capture the signed bounds.
+ * unsigned bounds capture the signed bounds.
* Thus, in all cases it suffices to blow away our signed bounds
* and rely on inferring new ones from the unsigned bounds and
* var_off of the result.
@@ -7367,43 +7806,19 @@ static void mark_ptr_or_null_reg(struct bpf_func_state *state,
}
if (is_null) {
reg->type = SCALAR_VALUE;
- } else if (reg->type == PTR_TO_MAP_VALUE_OR_NULL) {
- const struct bpf_map *map = reg->map_ptr;
-
- if (map->inner_map_meta) {
- reg->type = CONST_PTR_TO_MAP;
- reg->map_ptr = map->inner_map_meta;
- } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) {
- reg->type = PTR_TO_XDP_SOCK;
- } else if (map->map_type == BPF_MAP_TYPE_SOCKMAP ||
- map->map_type == BPF_MAP_TYPE_SOCKHASH) {
- reg->type = PTR_TO_SOCKET;
- } else {
- reg->type = PTR_TO_MAP_VALUE;
- }
- } else if (reg->type == PTR_TO_SOCKET_OR_NULL) {
- reg->type = PTR_TO_SOCKET;
- } else if (reg->type == PTR_TO_SOCK_COMMON_OR_NULL) {
- reg->type = PTR_TO_SOCK_COMMON;
- } else if (reg->type == PTR_TO_TCP_SOCK_OR_NULL) {
- reg->type = PTR_TO_TCP_SOCK;
- } else if (reg->type == PTR_TO_BTF_ID_OR_NULL) {
- reg->type = PTR_TO_BTF_ID;
- } else if (reg->type == PTR_TO_MEM_OR_NULL) {
- reg->type = PTR_TO_MEM;
- } else if (reg->type == PTR_TO_RDONLY_BUF_OR_NULL) {
- reg->type = PTR_TO_RDONLY_BUF;
- } else if (reg->type == PTR_TO_RDWR_BUF_OR_NULL) {
- reg->type = PTR_TO_RDWR_BUF;
- }
- if (is_null) {
/* We don't need id and ref_obj_id from this point
* onwards anymore, thus we should better reset it,
* so that state pruning has chances to take effect.
*/
reg->id = 0;
reg->ref_obj_id = 0;
- } else if (!reg_may_point_to_spin_lock(reg)) {
+
+ return;
+ }
+
+ mark_ptr_not_null_reg(reg);
+
+ if (!reg_may_point_to_spin_lock(reg)) {
/* For not-NULL ptr, reg->ref_obj_id will be reset
* in release_reg_references().
*
@@ -7986,6 +8401,9 @@ static int check_return_code(struct bpf_verifier_env *env)
env->prog->expected_attach_type == BPF_CGROUP_INET4_GETSOCKNAME ||
env->prog->expected_attach_type == BPF_CGROUP_INET6_GETSOCKNAME)
range = tnum_range(1, 1);
+ if (env->prog->expected_attach_type == BPF_CGROUP_INET4_BIND ||
+ env->prog->expected_attach_type == BPF_CGROUP_INET6_BIND)
+ range = tnum_range(0, 3);
break;
case BPF_PROG_TYPE_CGROUP_SKB:
if (env->prog->expected_attach_type == BPF_CGROUP_INET_EGRESS) {
@@ -10015,15 +10433,22 @@ static int check_map_prog_compatibility(struct bpf_verifier_env *env,
case BPF_MAP_TYPE_HASH:
case BPF_MAP_TYPE_LRU_HASH:
case BPF_MAP_TYPE_ARRAY:
+ case BPF_MAP_TYPE_PERCPU_HASH:
+ case BPF_MAP_TYPE_PERCPU_ARRAY:
+ case BPF_MAP_TYPE_LRU_PERCPU_HASH:
+ case BPF_MAP_TYPE_ARRAY_OF_MAPS:
+ case BPF_MAP_TYPE_HASH_OF_MAPS:
if (!is_preallocated_map(map)) {
verbose(env,
- "Sleepable programs can only use preallocated hash maps\n");
+ "Sleepable programs can only use preallocated maps\n");
return -EINVAL;
}
break;
+ case BPF_MAP_TYPE_RINGBUF:
+ break;
default:
verbose(env,
- "Sleepable programs can only use array and hash maps\n");
+ "Sleepable programs can only use array, hash, and ringbuf maps\n");
return -EINVAL;
}
@@ -10581,6 +11006,7 @@ static int opt_subreg_zext_lo32_rnd_hi32(struct bpf_verifier_env *env,
for (i = 0; i < len; i++) {
int adj_idx = i + delta;
struct bpf_insn insn;
+ u8 load_reg;
insn = insns[adj_idx];
if (!aux[adj_idx].zext_dst) {
@@ -10623,9 +11049,27 @@ static int opt_subreg_zext_lo32_rnd_hi32(struct bpf_verifier_env *env,
if (!bpf_jit_needs_zext())
continue;
+ /* zext_dst means that we want to zero-extend whatever register
+ * the insn defines, which is dst_reg most of the time, with
+ * the notable exception of BPF_STX + BPF_ATOMIC + BPF_FETCH.
+ */
+ if (BPF_CLASS(insn.code) == BPF_STX &&
+ BPF_MODE(insn.code) == BPF_ATOMIC) {
+ /* BPF_STX + BPF_ATOMIC insns without BPF_FETCH do not
+ * define any registers, therefore zext_dst cannot be
+ * set.
+ */
+ if (WARN_ON(!(insn.imm & BPF_FETCH)))
+ return -EINVAL;
+ load_reg = insn.imm == BPF_CMPXCHG ? BPF_REG_0
+ : insn.src_reg;
+ } else {
+ load_reg = insn.dst_reg;
+ }
+
zext_patch[0] = insn;
- zext_patch[1].dst_reg = insn.dst_reg;
- zext_patch[1].src_reg = insn.dst_reg;
+ zext_patch[1].dst_reg = load_reg;
+ zext_patch[1].src_reg = load_reg;
patch = zext_patch;
patch_len = 2;
apply_patch_buffer:
@@ -10841,8 +11285,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
return 0;
for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) {
- if (insn->code != (BPF_JMP | BPF_CALL) ||
- insn->src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn))
continue;
/* Upon error here we cannot fall back to interpreter but
* need a hard reject of the program. Thus -EFAULT is
@@ -10883,7 +11326,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
/* BPF_PROG_RUN doesn't call subprogs directly,
* hence main prog stats include the runtime of subprogs.
* subprogs don't have IDs and not reachable via prog_get_next_id
- * func[i]->aux->stats will never be accessed and stays NULL
+ * func[i]->stats will never be accessed and stays NULL
*/
func[i] = bpf_prog_alloc_no_stats(bpf_prog_size(len), GFP_USER);
if (!func[i])
@@ -10971,8 +11414,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
for (i = 0; i < env->subprog_cnt; i++) {
insn = func[i]->insnsi;
for (j = 0; j < func[i]->len; j++, insn++) {
- if (insn->code != (BPF_JMP | BPF_CALL) ||
- insn->src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn))
continue;
subprog = insn->off;
insn->imm = BPF_CAST_CALL(func[subprog]->bpf_func) -
@@ -11017,8 +11459,7 @@ static int jit_subprogs(struct bpf_verifier_env *env)
* later look the same as if they were interpreted only.
*/
for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) {
- if (insn->code != (BPF_JMP | BPF_CALL) ||
- insn->src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn))
continue;
insn->off = env->insn_aux_data[i].call_imm;
subprog = find_subprog(env, i + insn->off + 1);
@@ -11047,8 +11488,7 @@ out_undo_insn:
/* cleanup main prog to be interpreted */
prog->jit_requested = 0;
for (i = 0, insn = prog->insnsi; i < prog->len; i++, insn++) {
- if (insn->code != (BPF_JMP | BPF_CALL) ||
- insn->src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn))
continue;
insn->off = 0;
insn->imm = env->insn_aux_data[i].call_imm;
@@ -11083,8 +11523,7 @@ static int fixup_call_args(struct bpf_verifier_env *env)
return -EINVAL;
}
for (i = 0; i < prog->len; i++, insn++) {
- if (insn->code != (BPF_JMP | BPF_CALL) ||
- insn->src_reg != BPF_PSEUDO_CALL)
+ if (!bpf_pseudo_call(insn))
continue;
depth = get_callee_stack_depth(env, insn, i);
if (depth < 0)
@@ -11547,6 +11986,13 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog)
mark_reg_known_zero(env, regs, i);
else if (regs[i].type == SCALAR_VALUE)
mark_reg_unknown(env, regs, i);
+ else if (regs[i].type == PTR_TO_MEM_OR_NULL) {
+ const u32 mem_size = regs[i].mem_size;
+
+ mark_reg_known_zero(env, regs, i);
+ regs[i].mem_size = mem_size;
+ regs[i].id = ++env->id_gen;
+ }
}
} else {
/* 1st arg to a function */
@@ -12125,6 +12571,7 @@ int bpf_check(struct bpf_prog **prog, union bpf_attr *attr,
env->strict_alignment = false;
env->allow_ptr_leaks = bpf_allow_ptr_leaks();
+ env->allow_uninit_stack = bpf_allow_uninit_stack();
env->allow_ptr_to_map_access = bpf_allow_ptr_to_map_access();
env->bypass_spec_v1 = bpf_bypass_spec_v1();
env->bypass_spec_v4 = bpf_bypass_spec_v4();