summaryrefslogtreecommitdiffstats
path: root/kernel
diff options
context:
space:
mode:
authorDavid S. Miller <davem@davemloft.net>2018-08-07 11:02:05 -0700
committerDavid S. Miller <davem@davemloft.net>2018-08-07 11:02:05 -0700
commit1ba982806ce58baac44c52e6c7812245afad27c8 (patch)
treeb0a8061a496c3d01eed36eadf22cea845a2a995b /kernel
parentc5d99d2b35dadebab2408bb10dcd50364eaaf9f4 (diff)
parent85fc4b16aaf05fc8978d242c556a1711dce15cf8 (diff)
downloadlinux-1ba982806ce58baac44c52e6c7812245afad27c8.tar.bz2
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
Daniel Borkmann says: ==================== pull-request: bpf-next 2018-08-07 The following pull-request contains BPF updates for your *net-next* tree. The main changes are: 1) Add cgroup local storage for BPF programs, which provides a fast accessible memory for storing various per-cgroup data like number of transmitted packets, etc, from Roman. 2) Support bpf_get_socket_cookie() BPF helper in several more program types that have a full socket available, from Andrey. 3) Significantly improve the performance of perf events which are reported from BPF offload. Also convert a couple of BPF AF_XDP samples overto use libbpf, both from Jakub. 4) seg6local LWT provides the End.DT6 action, which allows to decapsulate an outer IPv6 header containing a Segment Routing Header. Adds this action now to the seg6local BPF interface, from Mathieu. 5) Do not mark dst register as unbounded in MOV64 instruction when both src and dst register are the same, from Arthur. 6) Define u_smp_rmb() and u_smp_wmb() to their respective barrier instructions on arm64 for the AF_XDP sample code, from Brian. 7) Convert the tcp_client.py and tcp_server.py BPF selftest scripts over from Python 2 to Python 3, from Jeremy. 8) Enable BTF build flags to the BPF sample code Makefile, from Taeung. 9) Remove an unnecessary rcu_read_lock() in run_lwt_bpf(), from Taehee. 10) Several improvements to the README.rst from the BPF documentation to make it more consistent with RST format, from Tobin. 11) Replace all occurrences of strerror() by calls to strerror_r() in libbpf and fix a FORTIFY_SOURCE build error along with it, from Thomas. 12) Fix a bug in bpftool's get_btf() function to correctly propagate an error via PTR_ERR(), from Yue. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'kernel')
-rw-r--r--kernel/bpf/Makefile1
-rw-r--r--kernel/bpf/cgroup.c157
-rw-r--r--kernel/bpf/core.c77
-rw-r--r--kernel/bpf/helpers.c20
-rw-r--r--kernel/bpf/local_storage.c378
-rw-r--r--kernel/bpf/map_in_map.c3
-rw-r--r--kernel/bpf/syscall.c61
-rw-r--r--kernel/bpf/verifier.c44
8 files changed, 617 insertions, 124 deletions
diff --git a/kernel/bpf/Makefile b/kernel/bpf/Makefile
index f27f5496d6fe..e8906cbad81f 100644
--- a/kernel/bpf/Makefile
+++ b/kernel/bpf/Makefile
@@ -3,6 +3,7 @@ obj-y := core.o
obj-$(CONFIG_BPF_SYSCALL) += syscall.o verifier.o inode.o helpers.o tnum.o
obj-$(CONFIG_BPF_SYSCALL) += hashtab.o arraymap.o percpu_freelist.o bpf_lru_list.o lpm_trie.o map_in_map.o
+obj-$(CONFIG_BPF_SYSCALL) += local_storage.o
obj-$(CONFIG_BPF_SYSCALL) += disasm.o
obj-$(CONFIG_BPF_SYSCALL) += btf.o
ifeq ($(CONFIG_NET),y)
diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c
index badabb0b435c..6a7d931bbc55 100644
--- a/kernel/bpf/cgroup.c
+++ b/kernel/bpf/cgroup.c
@@ -34,6 +34,8 @@ void cgroup_bpf_put(struct cgroup *cgrp)
list_for_each_entry_safe(pl, tmp, progs, node) {
list_del(&pl->node);
bpf_prog_put(pl->prog);
+ bpf_cgroup_storage_unlink(pl->storage);
+ bpf_cgroup_storage_free(pl->storage);
kfree(pl);
static_branch_dec(&cgroup_bpf_enabled_key);
}
@@ -115,15 +117,18 @@ static int compute_effective_progs(struct cgroup *cgrp,
cnt = 0;
p = cgrp;
do {
- if (cnt == 0 || (p->bpf.flags[type] & BPF_F_ALLOW_MULTI))
- list_for_each_entry(pl,
- &p->bpf.progs[type], node) {
- if (!pl->prog)
- continue;
- progs->progs[cnt++] = pl->prog;
- }
- p = cgroup_parent(p);
- } while (p);
+ if (cnt > 0 && !(p->bpf.flags[type] & BPF_F_ALLOW_MULTI))
+ continue;
+
+ list_for_each_entry(pl, &p->bpf.progs[type], node) {
+ if (!pl->prog)
+ continue;
+
+ progs->items[cnt].prog = pl->prog;
+ progs->items[cnt].cgroup_storage = pl->storage;
+ cnt++;
+ }
+ } while ((p = cgroup_parent(p)));
rcu_assign_pointer(*array, progs);
return 0;
@@ -172,6 +177,45 @@ cleanup:
return -ENOMEM;
}
+static int update_effective_progs(struct cgroup *cgrp,
+ enum bpf_attach_type type)
+{
+ struct cgroup_subsys_state *css;
+ int err;
+
+ /* allocate and recompute effective prog arrays */
+ css_for_each_descendant_pre(css, &cgrp->self) {
+ struct cgroup *desc = container_of(css, struct cgroup, self);
+
+ err = compute_effective_progs(desc, type, &desc->bpf.inactive);
+ if (err)
+ goto cleanup;
+ }
+
+ /* all allocations were successful. Activate all prog arrays */
+ css_for_each_descendant_pre(css, &cgrp->self) {
+ struct cgroup *desc = container_of(css, struct cgroup, self);
+
+ activate_effective_progs(desc, type, desc->bpf.inactive);
+ desc->bpf.inactive = NULL;
+ }
+
+ return 0;
+
+cleanup:
+ /* oom while computing effective. Free all computed effective arrays
+ * since they were not activated
+ */
+ css_for_each_descendant_pre(css, &cgrp->self) {
+ struct cgroup *desc = container_of(css, struct cgroup, self);
+
+ bpf_prog_array_free(desc->bpf.inactive);
+ desc->bpf.inactive = NULL;
+ }
+
+ return err;
+}
+
#define BPF_CGROUP_MAX_PROGS 64
/**
@@ -188,7 +232,7 @@ int __cgroup_bpf_attach(struct cgroup *cgrp, struct bpf_prog *prog,
{
struct list_head *progs = &cgrp->bpf.progs[type];
struct bpf_prog *old_prog = NULL;
- struct cgroup_subsys_state *css;
+ struct bpf_cgroup_storage *storage, *old_storage = NULL;
struct bpf_prog_list *pl;
bool pl_was_allocated;
int err;
@@ -210,72 +254,71 @@ int __cgroup_bpf_attach(struct cgroup *cgrp, struct bpf_prog *prog,
if (prog_list_length(progs) >= BPF_CGROUP_MAX_PROGS)
return -E2BIG;
+ storage = bpf_cgroup_storage_alloc(prog);
+ if (IS_ERR(storage))
+ return -ENOMEM;
+
if (flags & BPF_F_ALLOW_MULTI) {
- list_for_each_entry(pl, progs, node)
- if (pl->prog == prog)
+ list_for_each_entry(pl, progs, node) {
+ if (pl->prog == prog) {
/* disallow attaching the same prog twice */
+ bpf_cgroup_storage_free(storage);
return -EINVAL;
+ }
+ }
pl = kmalloc(sizeof(*pl), GFP_KERNEL);
- if (!pl)
+ if (!pl) {
+ bpf_cgroup_storage_free(storage);
return -ENOMEM;
+ }
+
pl_was_allocated = true;
pl->prog = prog;
+ pl->storage = storage;
list_add_tail(&pl->node, progs);
} else {
if (list_empty(progs)) {
pl = kmalloc(sizeof(*pl), GFP_KERNEL);
- if (!pl)
+ if (!pl) {
+ bpf_cgroup_storage_free(storage);
return -ENOMEM;
+ }
pl_was_allocated = true;
list_add_tail(&pl->node, progs);
} else {
pl = list_first_entry(progs, typeof(*pl), node);
old_prog = pl->prog;
+ old_storage = pl->storage;
+ bpf_cgroup_storage_unlink(old_storage);
pl_was_allocated = false;
}
pl->prog = prog;
+ pl->storage = storage;
}
cgrp->bpf.flags[type] = flags;
- /* allocate and recompute effective prog arrays */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- err = compute_effective_progs(desc, type, &desc->bpf.inactive);
- if (err)
- goto cleanup;
- }
-
- /* all allocations were successful. Activate all prog arrays */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- activate_effective_progs(desc, type, desc->bpf.inactive);
- desc->bpf.inactive = NULL;
- }
+ err = update_effective_progs(cgrp, type);
+ if (err)
+ goto cleanup;
static_branch_inc(&cgroup_bpf_enabled_key);
+ if (old_storage)
+ bpf_cgroup_storage_free(old_storage);
if (old_prog) {
bpf_prog_put(old_prog);
static_branch_dec(&cgroup_bpf_enabled_key);
}
+ bpf_cgroup_storage_link(storage, cgrp, type);
return 0;
cleanup:
- /* oom while computing effective. Free all computed effective arrays
- * since they were not activated
- */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- bpf_prog_array_free(desc->bpf.inactive);
- desc->bpf.inactive = NULL;
- }
-
/* and cleanup the prog list */
pl->prog = old_prog;
+ bpf_cgroup_storage_free(pl->storage);
+ pl->storage = old_storage;
+ bpf_cgroup_storage_link(old_storage, cgrp, type);
if (pl_was_allocated) {
list_del(&pl->node);
kfree(pl);
@@ -298,7 +341,6 @@ int __cgroup_bpf_detach(struct cgroup *cgrp, struct bpf_prog *prog,
struct list_head *progs = &cgrp->bpf.progs[type];
u32 flags = cgrp->bpf.flags[type];
struct bpf_prog *old_prog = NULL;
- struct cgroup_subsys_state *css;
struct bpf_prog_list *pl;
int err;
@@ -337,25 +379,14 @@ int __cgroup_bpf_detach(struct cgroup *cgrp, struct bpf_prog *prog,
pl->prog = NULL;
}
- /* allocate and recompute effective prog arrays */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- err = compute_effective_progs(desc, type, &desc->bpf.inactive);
- if (err)
- goto cleanup;
- }
-
- /* all allocations were successful. Activate all prog arrays */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- activate_effective_progs(desc, type, desc->bpf.inactive);
- desc->bpf.inactive = NULL;
- }
+ err = update_effective_progs(cgrp, type);
+ if (err)
+ goto cleanup;
/* now can actually delete it from this cgroup list */
list_del(&pl->node);
+ bpf_cgroup_storage_unlink(pl->storage);
+ bpf_cgroup_storage_free(pl->storage);
kfree(pl);
if (list_empty(progs))
/* last program was detached, reset flags to zero */
@@ -366,16 +397,6 @@ int __cgroup_bpf_detach(struct cgroup *cgrp, struct bpf_prog *prog,
return 0;
cleanup:
- /* oom while computing effective. Free all computed effective arrays
- * since they were not activated
- */
- css_for_each_descendant_pre(css, &cgrp->self) {
- struct cgroup *desc = container_of(css, struct cgroup, self);
-
- bpf_prog_array_free(desc->bpf.inactive);
- desc->bpf.inactive = NULL;
- }
-
/* and restore back old_prog */
pl->prog = old_prog;
return err;
@@ -654,6 +675,8 @@ cgroup_dev_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
return &bpf_map_delete_elem_proto;
case BPF_FUNC_get_current_uid_gid:
return &bpf_get_current_uid_gid_proto;
+ case BPF_FUNC_get_local_storage:
+ return &bpf_get_local_storage_proto;
case BPF_FUNC_trace_printk:
if (capable(CAP_SYS_ADMIN))
return bpf_get_trace_printk_proto();
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 253aa8e79c7b..4d09e610777f 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -1542,7 +1542,8 @@ struct bpf_prog_array *bpf_prog_array_alloc(u32 prog_cnt, gfp_t flags)
{
if (prog_cnt)
return kzalloc(sizeof(struct bpf_prog_array) +
- sizeof(struct bpf_prog *) * (prog_cnt + 1),
+ sizeof(struct bpf_prog_array_item) *
+ (prog_cnt + 1),
flags);
return &empty_prog_array.hdr;
@@ -1556,43 +1557,45 @@ void bpf_prog_array_free(struct bpf_prog_array __rcu *progs)
kfree_rcu(progs, rcu);
}
-int bpf_prog_array_length(struct bpf_prog_array __rcu *progs)
+int bpf_prog_array_length(struct bpf_prog_array __rcu *array)
{
- struct bpf_prog **prog;
+ struct bpf_prog_array_item *item;
u32 cnt = 0;
rcu_read_lock();
- prog = rcu_dereference(progs)->progs;
- for (; *prog; prog++)
- if (*prog != &dummy_bpf_prog.prog)
+ item = rcu_dereference(array)->items;
+ for (; item->prog; item++)
+ if (item->prog != &dummy_bpf_prog.prog)
cnt++;
rcu_read_unlock();
return cnt;
}
-static bool bpf_prog_array_copy_core(struct bpf_prog **prog,
+
+static bool bpf_prog_array_copy_core(struct bpf_prog_array __rcu *array,
u32 *prog_ids,
u32 request_cnt)
{
+ struct bpf_prog_array_item *item;
int i = 0;
- for (; *prog; prog++) {
- if (*prog == &dummy_bpf_prog.prog)
+ item = rcu_dereference(array)->items;
+ for (; item->prog; item++) {
+ if (item->prog == &dummy_bpf_prog.prog)
continue;
- prog_ids[i] = (*prog)->aux->id;
+ prog_ids[i] = item->prog->aux->id;
if (++i == request_cnt) {
- prog++;
+ item++;
break;
}
}
- return !!(*prog);
+ return !!(item->prog);
}
-int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *progs,
+int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *array,
__u32 __user *prog_ids, u32 cnt)
{
- struct bpf_prog **prog;
unsigned long err = 0;
bool nospc;
u32 *ids;
@@ -1611,8 +1614,7 @@ int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *progs,
if (!ids)
return -ENOMEM;
rcu_read_lock();
- prog = rcu_dereference(progs)->progs;
- nospc = bpf_prog_array_copy_core(prog, ids, cnt);
+ nospc = bpf_prog_array_copy_core(array, ids, cnt);
rcu_read_unlock();
err = copy_to_user(prog_ids, ids, cnt * sizeof(u32));
kfree(ids);
@@ -1623,14 +1625,14 @@ int bpf_prog_array_copy_to_user(struct bpf_prog_array __rcu *progs,
return 0;
}
-void bpf_prog_array_delete_safe(struct bpf_prog_array __rcu *progs,
+void bpf_prog_array_delete_safe(struct bpf_prog_array __rcu *array,
struct bpf_prog *old_prog)
{
- struct bpf_prog **prog = progs->progs;
+ struct bpf_prog_array_item *item = array->items;
- for (; *prog; prog++)
- if (*prog == old_prog) {
- WRITE_ONCE(*prog, &dummy_bpf_prog.prog);
+ for (; item->prog; item++)
+ if (item->prog == old_prog) {
+ WRITE_ONCE(item->prog, &dummy_bpf_prog.prog);
break;
}
}
@@ -1641,7 +1643,7 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array,
struct bpf_prog_array **new_array)
{
int new_prog_cnt, carry_prog_cnt = 0;
- struct bpf_prog **existing_prog;
+ struct bpf_prog_array_item *existing;
struct bpf_prog_array *array;
bool found_exclude = false;
int new_prog_idx = 0;
@@ -1650,15 +1652,15 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array,
* the new array.
*/
if (old_array) {
- existing_prog = old_array->progs;
- for (; *existing_prog; existing_prog++) {
- if (*existing_prog == exclude_prog) {
+ existing = old_array->items;
+ for (; existing->prog; existing++) {
+ if (existing->prog == exclude_prog) {
found_exclude = true;
continue;
}
- if (*existing_prog != &dummy_bpf_prog.prog)
+ if (existing->prog != &dummy_bpf_prog.prog)
carry_prog_cnt++;
- if (*existing_prog == include_prog)
+ if (existing->prog == include_prog)
return -EEXIST;
}
}
@@ -1684,15 +1686,17 @@ int bpf_prog_array_copy(struct bpf_prog_array __rcu *old_array,
/* Fill in the new prog array */
if (carry_prog_cnt) {
- existing_prog = old_array->progs;
- for (; *existing_prog; existing_prog++)
- if (*existing_prog != exclude_prog &&
- *existing_prog != &dummy_bpf_prog.prog)
- array->progs[new_prog_idx++] = *existing_prog;
+ existing = old_array->items;
+ for (; existing->prog; existing++)
+ if (existing->prog != exclude_prog &&
+ existing->prog != &dummy_bpf_prog.prog) {
+ array->items[new_prog_idx++].prog =
+ existing->prog;
+ }
}
if (include_prog)
- array->progs[new_prog_idx++] = include_prog;
- array->progs[new_prog_idx] = NULL;
+ array->items[new_prog_idx++].prog = include_prog;
+ array->items[new_prog_idx].prog = NULL;
*new_array = array;
return 0;
}
@@ -1701,7 +1705,6 @@ int bpf_prog_array_copy_info(struct bpf_prog_array __rcu *array,
u32 *prog_ids, u32 request_cnt,
u32 *prog_cnt)
{
- struct bpf_prog **prog;
u32 cnt = 0;
if (array)
@@ -1714,8 +1717,7 @@ int bpf_prog_array_copy_info(struct bpf_prog_array __rcu *array,
return 0;
/* this function is called under trace/bpf_trace.c: bpf_event_mutex */
- prog = rcu_dereference_check(array, 1)->progs;
- return bpf_prog_array_copy_core(prog, prog_ids, request_cnt) ? -ENOSPC
+ return bpf_prog_array_copy_core(array, prog_ids, request_cnt) ? -ENOSPC
: 0;
}
@@ -1793,6 +1795,7 @@ const struct bpf_func_proto bpf_get_current_comm_proto __weak;
const struct bpf_func_proto bpf_sock_map_update_proto __weak;
const struct bpf_func_proto bpf_sock_hash_update_proto __weak;
const struct bpf_func_proto bpf_get_current_cgroup_id_proto __weak;
+const struct bpf_func_proto bpf_get_local_storage_proto __weak;
const struct bpf_func_proto * __weak bpf_get_trace_printk_proto(void)
{
diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index 73065e2d23c2..1991466b8327 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -193,4 +193,24 @@ const struct bpf_func_proto bpf_get_current_cgroup_id_proto = {
.gpl_only = false,
.ret_type = RET_INTEGER,
};
+
+DECLARE_PER_CPU(void*, bpf_cgroup_storage);
+
+BPF_CALL_2(bpf_get_local_storage, struct bpf_map *, map, u64, flags)
+{
+ /* map and flags arguments are not used now,
+ * but provide an ability to extend the API
+ * for other types of local storages.
+ * verifier checks that their values are correct.
+ */
+ return (unsigned long) this_cpu_read(bpf_cgroup_storage);
+}
+
+const struct bpf_func_proto bpf_get_local_storage_proto = {
+ .func = bpf_get_local_storage,
+ .gpl_only = false,
+ .ret_type = RET_PTR_TO_MAP_VALUE,
+ .arg1_type = ARG_CONST_MAP_PTR,
+ .arg2_type = ARG_ANYTHING,
+};
#endif
diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c
new file mode 100644
index 000000000000..fc4e37f68f2a
--- /dev/null
+++ b/kernel/bpf/local_storage.c
@@ -0,0 +1,378 @@
+//SPDX-License-Identifier: GPL-2.0
+#include <linux/bpf-cgroup.h>
+#include <linux/bpf.h>
+#include <linux/bug.h>
+#include <linux/filter.h>
+#include <linux/mm.h>
+#include <linux/rbtree.h>
+#include <linux/slab.h>
+
+DEFINE_PER_CPU(void*, bpf_cgroup_storage);
+
+#ifdef CONFIG_CGROUP_BPF
+
+#define LOCAL_STORAGE_CREATE_FLAG_MASK \
+ (BPF_F_NUMA_NODE | BPF_F_RDONLY | BPF_F_WRONLY)
+
+struct bpf_cgroup_storage_map {
+ struct bpf_map map;
+
+ spinlock_t lock;
+ struct bpf_prog *prog;
+ struct rb_root root;
+ struct list_head list;
+};
+
+static struct bpf_cgroup_storage_map *map_to_storage(struct bpf_map *map)
+{
+ return container_of(map, struct bpf_cgroup_storage_map, map);
+}
+
+static int bpf_cgroup_storage_key_cmp(
+ const struct bpf_cgroup_storage_key *key1,
+ const struct bpf_cgroup_storage_key *key2)
+{
+ if (key1->cgroup_inode_id < key2->cgroup_inode_id)
+ return -1;
+ else if (key1->cgroup_inode_id > key2->cgroup_inode_id)
+ return 1;
+ else if (key1->attach_type < key2->attach_type)
+ return -1;
+ else if (key1->attach_type > key2->attach_type)
+ return 1;
+ return 0;
+}
+
+static struct bpf_cgroup_storage *cgroup_storage_lookup(
+ struct bpf_cgroup_storage_map *map, struct bpf_cgroup_storage_key *key,
+ bool locked)
+{
+ struct rb_root *root = &map->root;
+ struct rb_node *node;
+
+ if (!locked)
+ spin_lock_bh(&map->lock);
+
+ node = root->rb_node;
+ while (node) {
+ struct bpf_cgroup_storage *storage;
+
+ storage = container_of(node, struct bpf_cgroup_storage, node);
+
+ switch (bpf_cgroup_storage_key_cmp(key, &storage->key)) {
+ case -1:
+ node = node->rb_left;
+ break;
+ case 1:
+ node = node->rb_right;
+ break;
+ default:
+ if (!locked)
+ spin_unlock_bh(&map->lock);
+ return storage;
+ }
+ }
+
+ if (!locked)
+ spin_unlock_bh(&map->lock);
+
+ return NULL;
+}
+
+static int cgroup_storage_insert(struct bpf_cgroup_storage_map *map,
+ struct bpf_cgroup_storage *storage)
+{
+ struct rb_root *root = &map->root;
+ struct rb_node **new = &(root->rb_node), *parent = NULL;
+
+ while (*new) {
+ struct bpf_cgroup_storage *this;
+
+ this = container_of(*new, struct bpf_cgroup_storage, node);
+
+ parent = *new;
+ switch (bpf_cgroup_storage_key_cmp(&storage->key, &this->key)) {
+ case -1:
+ new = &((*new)->rb_left);
+ break;
+ case 1:
+ new = &((*new)->rb_right);
+ break;
+ default:
+ return -EEXIST;
+ }
+ }
+
+ rb_link_node(&storage->node, parent, new);
+ rb_insert_color(&storage->node, root);
+
+ return 0;
+}
+
+static void *cgroup_storage_lookup_elem(struct bpf_map *_map, void *_key)
+{
+ struct bpf_cgroup_storage_map *map = map_to_storage(_map);
+ struct bpf_cgroup_storage_key *key = _key;
+ struct bpf_cgroup_storage *storage;
+
+ storage = cgroup_storage_lookup(map, key, false);
+ if (!storage)
+ return NULL;
+
+ return &READ_ONCE(storage->buf)->data[0];
+}
+
+static int cgroup_storage_update_elem(struct bpf_map *map, void *_key,
+ void *value, u64 flags)
+{
+ struct bpf_cgroup_storage_key *key = _key;
+ struct bpf_cgroup_storage *storage;
+ struct bpf_storage_buffer *new;
+
+ if (flags & BPF_NOEXIST)
+ return -EINVAL;
+
+ storage = cgroup_storage_lookup((struct bpf_cgroup_storage_map *)map,
+ key, false);
+ if (!storage)
+ return -ENOENT;
+
+ new = kmalloc_node(sizeof(struct bpf_storage_buffer) +
+ map->value_size, __GFP_ZERO | GFP_USER,
+ map->numa_node);
+ if (!new)
+ return -ENOMEM;
+
+ memcpy(&new->data[0], value, map->value_size);
+
+ new = xchg(&storage->buf, new);
+ kfree_rcu(new, rcu);
+
+ return 0;
+}
+
+static int cgroup_storage_get_next_key(struct bpf_map *_map, void *_key,
+ void *_next_key)
+{
+ struct bpf_cgroup_storage_map *map = map_to_storage(_map);
+ struct bpf_cgroup_storage_key *key = _key;
+ struct bpf_cgroup_storage_key *next = _next_key;
+ struct bpf_cgroup_storage *storage;
+
+ spin_lock_bh(&map->lock);
+
+ if (list_empty(&map->list))
+ goto enoent;
+
+ if (key) {
+ storage = cgroup_storage_lookup(map, key, true);
+ if (!storage)
+ goto enoent;
+
+ storage = list_next_entry(storage, list);
+ if (!storage)
+ goto enoent;
+ } else {
+ storage = list_first_entry(&map->list,
+ struct bpf_cgroup_storage, list);
+ }
+
+ spin_unlock_bh(&map->lock);
+ next->attach_type = storage->key.attach_type;
+ next->cgroup_inode_id = storage->key.cgroup_inode_id;
+ return 0;
+
+enoent:
+ spin_unlock_bh(&map->lock);
+ return -ENOENT;
+}
+
+static struct bpf_map *cgroup_storage_map_alloc(union bpf_attr *attr)
+{
+ int numa_node = bpf_map_attr_numa_node(attr);
+ struct bpf_cgroup_storage_map *map;
+
+ if (attr->key_size != sizeof(struct bpf_cgroup_storage_key))
+ return ERR_PTR(-EINVAL);
+
+ if (attr->value_size > PAGE_SIZE)
+ return ERR_PTR(-E2BIG);
+
+ if (attr->map_flags & ~LOCAL_STORAGE_CREATE_FLAG_MASK)
+ /* reserved bits should not be used */
+ return ERR_PTR(-EINVAL);
+
+ if (attr->max_entries)
+ /* max_entries is not used and enforced to be 0 */
+ return ERR_PTR(-EINVAL);
+
+ map = kmalloc_node(sizeof(struct bpf_cgroup_storage_map),
+ __GFP_ZERO | GFP_USER, numa_node);
+ if (!map)
+ return ERR_PTR(-ENOMEM);
+
+ map->map.pages = round_up(sizeof(struct bpf_cgroup_storage_map),
+ PAGE_SIZE) >> PAGE_SHIFT;
+
+ /* copy mandatory map attributes */
+ bpf_map_init_from_attr(&map->map, attr);
+
+ spin_lock_init(&map->lock);
+ map->root = RB_ROOT;
+ INIT_LIST_HEAD(&map->list);
+
+ return &map->map;
+}
+
+static void cgroup_storage_map_free(struct bpf_map *_map)
+{
+ struct bpf_cgroup_storage_map *map = map_to_storage(_map);
+
+ WARN_ON(!RB_EMPTY_ROOT(&map->root));
+ WARN_ON(!list_empty(&map->list));
+
+ kfree(map);
+}
+
+static int cgroup_storage_delete_elem(struct bpf_map *map, void *key)
+{
+ return -EINVAL;
+}
+
+const struct bpf_map_ops cgroup_storage_map_ops = {
+ .map_alloc = cgroup_storage_map_alloc,
+ .map_free = cgroup_storage_map_free,
+ .map_get_next_key = cgroup_storage_get_next_key,
+ .map_lookup_elem = cgroup_storage_lookup_elem,
+ .map_update_elem = cgroup_storage_update_elem,
+ .map_delete_elem = cgroup_storage_delete_elem,
+};
+
+int bpf_cgroup_storage_assign(struct bpf_prog *prog, struct bpf_map *_map)
+{
+ struct bpf_cgroup_storage_map *map = map_to_storage(_map);
+ int ret = -EBUSY;
+
+ spin_lock_bh(&map->lock);
+
+ if (map->prog && map->prog != prog)
+ goto unlock;
+ if (prog->aux->cgroup_storage && prog->aux->cgroup_storage != _map)
+ goto unlock;
+
+ map->prog = prog;
+ prog->aux->cgroup_storage = _map;
+ ret = 0;
+unlock:
+ spin_unlock_bh(&map->lock);
+
+ return ret;
+}
+
+void bpf_cgroup_storage_release(struct bpf_prog *prog, struct bpf_map *_map)
+{
+ struct bpf_cgroup_storage_map *map = map_to_storage(_map);
+
+ spin_lock_bh(&map->lock);
+ if (map->prog == prog) {
+ WARN_ON(prog->aux->cgroup_storage != _map);
+ map->prog = NULL;
+ prog->aux->cgroup_storage = NULL;
+ }
+ spin_unlock_bh(&map->lock);
+}
+
+struct bpf_cgroup_storage *bpf_cgroup_storage_alloc(struct bpf_prog *prog)
+{
+ struct bpf_cgroup_storage *storage;
+ struct bpf_map *map;
+ u32 pages;
+
+ map = prog->aux->cgroup_storage;
+ if (!map)
+ return NULL;
+
+ pages = round_up(sizeof(struct bpf_cgroup_storage) +
+ sizeof(struct bpf_storage_buffer) +
+ map->value_size, PAGE_SIZE) >> PAGE_SHIFT;
+ if (bpf_map_charge_memlock(map, pages))
+ return ERR_PTR(-EPERM);
+
+ storage = kmalloc_node(sizeof(struct bpf_cgroup_storage),
+ __GFP_ZERO | GFP_USER, map->numa_node);
+ if (!storage) {
+ bpf_map_uncharge_memlock(map, pages);
+ return ERR_PTR(-ENOMEM);
+ }
+
+ storage->buf = kmalloc_node(sizeof(struct bpf_storage_buffer) +
+ map->value_size, __GFP_ZERO | GFP_USER,
+ map->numa_node);
+ if (!storage->buf) {
+ bpf_map_uncharge_memlock(map, pages);
+ kfree(storage);
+ return ERR_PTR(-ENOMEM);
+ }
+
+ storage->map = (struct bpf_cgroup_storage_map *)map;
+
+ return storage;
+}
+
+void bpf_cgroup_storage_free(struct bpf_cgroup_storage *storage)
+{
+ u32 pages;
+ struct bpf_map *map;
+
+ if (!storage)
+ return;
+
+ map = &storage->map->map;
+ pages = round_up(sizeof(struct bpf_cgroup_storage) +
+ sizeof(struct bpf_storage_buffer) +
+ map->value_size, PAGE_SIZE) >> PAGE_SHIFT;
+ bpf_map_uncharge_memlock(map, pages);
+
+ kfree_rcu(storage->buf, rcu);
+ kfree_rcu(storage, rcu);
+}
+
+void bpf_cgroup_storage_link(struct bpf_cgroup_storage *storage,
+ struct cgroup *cgroup,
+ enum bpf_attach_type type)
+{
+ struct bpf_cgroup_storage_map *map;
+
+ if (!storage)
+ return;
+
+ storage->key.attach_type = type;
+ storage->key.cgroup_inode_id = cgroup->kn->id.id;
+
+ map = storage->map;
+
+ spin_lock_bh(&map->lock);
+ WARN_ON(cgroup_storage_insert(map, storage));
+ list_add(&storage->list, &map->list);
+ spin_unlock_bh(&map->lock);
+}
+
+void bpf_cgroup_storage_unlink(struct bpf_cgroup_storage *storage)
+{
+ struct bpf_cgroup_storage_map *map;
+ struct rb_root *root;
+
+ if (!storage)
+ return;
+
+ map = storage->map;
+
+ spin_lock_bh(&map->lock);
+ root = &map->root;
+ rb_erase(&storage->node, root);
+
+ list_del(&storage->list);
+ spin_unlock_bh(&map->lock);
+}
+
+#endif
diff --git a/kernel/bpf/map_in_map.c b/kernel/bpf/map_in_map.c
index 1da574612bea..3bfbf4464416 100644
--- a/kernel/bpf/map_in_map.c
+++ b/kernel/bpf/map_in_map.c
@@ -23,7 +23,8 @@ struct bpf_map *bpf_map_meta_alloc(int inner_map_ufd)
* is a runtime binding. Doing static check alone
* in the verifier is not enough.
*/
- if (inner_map->map_type == BPF_MAP_TYPE_PROG_ARRAY) {
+ if (inner_map->map_type == BPF_MAP_TYPE_PROG_ARRAY ||
+ inner_map->map_type == BPF_MAP_TYPE_CGROUP_STORAGE) {
fdput(f);
return ERR_PTR(-ENOTSUPP);
}
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index a31a1ba0f8ea..5af4e9e2722d 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -181,32 +181,60 @@ int bpf_map_precharge_memlock(u32 pages)
return 0;
}
-static int bpf_map_charge_memlock(struct bpf_map *map)
+static int bpf_charge_memlock(struct user_struct *user, u32 pages)
{
- struct user_struct *user = get_current_user();
- unsigned long memlock_limit;
+ unsigned long memlock_limit = rlimit(RLIMIT_MEMLOCK) >> PAGE_SHIFT;
- memlock_limit = rlimit(RLIMIT_MEMLOCK) >> PAGE_SHIFT;
+ if (atomic_long_add_return(pages, &user->locked_vm) > memlock_limit) {
+ atomic_long_sub(pages, &user->locked_vm);
+ return -EPERM;
+ }
+ return 0;
+}
- atomic_long_add(map->pages, &user->locked_vm);
+static void bpf_uncharge_memlock(struct user_struct *user, u32 pages)
+{
+ atomic_long_sub(pages, &user->locked_vm);
+}
+
+static int bpf_map_init_memlock(struct bpf_map *map)
+{
+ struct user_struct *user = get_current_user();
+ int ret;
- if (atomic_long_read(&user->locked_vm) > memlock_limit) {
- atomic_long_sub(map->pages, &user->locked_vm);
+ ret = bpf_charge_memlock(user, map->pages);
+ if (ret) {
free_uid(user);
- return -EPERM;
+ return ret;
}
map->user = user;
- return 0;
+ return ret;
}
-static void bpf_map_uncharge_memlock(struct bpf_map *map)
+static void bpf_map_release_memlock(struct bpf_map *map)
{
struct user_struct *user = map->user;
-
- atomic_long_sub(map->pages, &user->locked_vm);
+ bpf_uncharge_memlock(user, map->pages);
free_uid(user);
}
+int bpf_map_charge_memlock(struct bpf_map *map, u32 pages)
+{
+ int ret;
+
+ ret = bpf_charge_memlock(map->user, pages);
+ if (ret)
+ return ret;
+ map->pages += pages;
+ return ret;
+}
+
+void bpf_map_uncharge_memlock(struct bpf_map *map, u32 pages)
+{
+ bpf_uncharge_memlock(map->user, pages);
+ map->pages -= pages;
+}
+
static int bpf_map_alloc_id(struct bpf_map *map)
{
int id;
@@ -256,7 +284,7 @@ static void bpf_map_free_deferred(struct work_struct *work)
{
struct bpf_map *map = container_of(work, struct bpf_map, work);
- bpf_map_uncharge_memlock(map);
+ bpf_map_release_memlock(map);
security_bpf_map_free(map);
/* implementation dependent freeing */
map->ops->map_free(map);
@@ -492,7 +520,7 @@ static int map_create(union bpf_attr *attr)
if (err)
goto free_map_nouncharge;
- err = bpf_map_charge_memlock(map);
+ err = bpf_map_init_memlock(map);
if (err)
goto free_map_sec;
@@ -515,7 +543,7 @@ static int map_create(union bpf_attr *attr)
return err;
free_map:
- bpf_map_uncharge_memlock(map);
+ bpf_map_release_memlock(map);
free_map_sec:
security_bpf_map_free(map);
free_map_nouncharge:
@@ -929,6 +957,9 @@ static void free_used_maps(struct bpf_prog_aux *aux)
{
int i;
+ if (aux->cgroup_storage)
+ bpf_cgroup_storage_release(aux->prog, aux->cgroup_storage);
+
for (i = 0; i < aux->used_map_cnt; i++)
bpf_map_put(aux->used_maps[i]);
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 25e47c195874..587468a9c37d 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -2127,6 +2127,10 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env,
func_id != BPF_FUNC_current_task_under_cgroup)
goto error;
break;
+ case BPF_MAP_TYPE_CGROUP_STORAGE:
+ if (func_id != BPF_FUNC_get_local_storage)
+ goto error;
+ break;
/* devmap returns a pointer to a live net_device ifindex that we cannot
* allow to be modified from bpf side. So do not allow lookup elements
* for now.
@@ -2209,6 +2213,10 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env,
if (map->map_type != BPF_MAP_TYPE_SOCKHASH)
goto error;
break;
+ case BPF_FUNC_get_local_storage:
+ if (map->map_type != BPF_MAP_TYPE_CGROUP_STORAGE)
+ goto error;
+ break;
default:
break;
}
@@ -2533,6 +2541,16 @@ static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn
}
regs = cur_regs(env);
+
+ /* check that flags argument in get_local_storage(map, flags) is 0,
+ * this is required because get_local_storage() can't return an error.
+ */
+ if (func_id == BPF_FUNC_get_local_storage &&
+ !register_is_null(&regs[BPF_REG_2])) {
+ verbose(env, "get_local_storage() doesn't support non-zero flags\n");
+ return -EINVAL;
+ }
+
/* reset caller saved regs */
for (i = 0; i < CALLER_SAVED_REGS; i++) {
mark_reg_not_init(env, regs, caller_saved[i]);
@@ -2545,8 +2563,12 @@ static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn
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) {
- regs[BPF_REG_0].type = PTR_TO_MAP_VALUE_OR_NULL;
+ } else if (fn->ret_type == RET_PTR_TO_MAP_VALUE_OR_NULL ||
+ fn->ret_type == RET_PTR_TO_MAP_VALUE) {
+ if (fn->ret_type == RET_PTR_TO_MAP_VALUE)
+ regs[BPF_REG_0].type = PTR_TO_MAP_VALUE;
+ else
+ regs[BPF_REG_0].type = PTR_TO_MAP_VALUE_OR_NULL;
/* There is no offset yet applied, variable or fixed */
mark_reg_known_zero(env, regs, BPF_REG_0);
regs[BPF_REG_0].off = 0;
@@ -3238,8 +3260,8 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
}
}
- /* check dest operand */
- err = check_reg_arg(env, insn->dst_reg, DST_OP);
+ /* check dest operand, mark as required later */
+ err = check_reg_arg(env, insn->dst_reg, DST_OP_NO_MARK);
if (err)
return err;
@@ -3265,6 +3287,8 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
/* case: R = imm
* remember the value we stored into this reg
*/
+ /* clear any state __mark_reg_known doesn't set */
+ mark_reg_unknown(env, regs, insn->dst_reg);
regs[insn->dst_reg].type = SCALAR_VALUE;
if (BPF_CLASS(insn->code) == BPF_ALU64) {
__mark_reg_known(regs + insn->dst_reg,
@@ -5152,6 +5176,14 @@ static int replace_map_fd_with_map_ptr(struct bpf_verifier_env *env)
}
env->used_maps[env->used_map_cnt++] = map;
+ if (map->map_type == BPF_MAP_TYPE_CGROUP_STORAGE &&
+ bpf_cgroup_storage_assign(env->prog, map)) {
+ verbose(env,
+ "only one cgroup storage is allowed\n");
+ fdput(f);
+ return -EBUSY;
+ }
+
fdput(f);
next_insn:
insn++;
@@ -5178,6 +5210,10 @@ static void release_maps(struct bpf_verifier_env *env)
{
int i;
+ if (env->prog->aux->cgroup_storage)
+ bpf_cgroup_storage_release(env->prog,
+ env->prog->aux->cgroup_storage);
+
for (i = 0; i < env->used_map_cnt; i++)
bpf_map_put(env->used_maps[i]);
}