summaryrefslogtreecommitdiff
path: root/kernel/bpf
diff options
context:
space:
mode:
Diffstat (limited to 'kernel/bpf')
-rw-r--r--kernel/bpf/bpf_inode_storage.c2
-rw-r--r--kernel/bpf/bpf_struct_ops.c2
-rw-r--r--kernel/bpf/core.c4
-rw-r--r--kernel/bpf/preload/bpf_preload_kern.c19
-rw-r--r--kernel/bpf/syscall.c5
-rw-r--r--kernel/bpf/trampoline.c218
-rw-r--r--kernel/bpf/verifier.c37
7 files changed, 217 insertions, 70 deletions
diff --git a/kernel/bpf/bpf_inode_storage.c b/kernel/bpf/bpf_inode_storage.c
index da753721457c..2921ca39a93e 100644
--- a/kernel/bpf/bpf_inode_storage.c
+++ b/kernel/bpf/bpf_inode_storage.c
@@ -109,7 +109,7 @@ static void *bpf_fd_inode_storage_lookup_elem(struct bpf_map *map, void *key)
fd = *(int *)key;
f = fget_raw(fd);
if (!f)
- return NULL;
+ return ERR_PTR(-EBADF);
sdata = inode_storage_lookup(f->f_inode, map, true);
fput(f);
diff --git a/kernel/bpf/bpf_struct_ops.c b/kernel/bpf/bpf_struct_ops.c
index 1a666a975416..70f6fd4fa305 100644
--- a/kernel/bpf/bpf_struct_ops.c
+++ b/kernel/bpf/bpf_struct_ops.c
@@ -430,7 +430,7 @@ static int bpf_struct_ops_map_update_elem(struct bpf_map *map, void *key,
tprogs[BPF_TRAMP_FENTRY].progs[0] = prog;
tprogs[BPF_TRAMP_FENTRY].nr_progs = 1;
- err = arch_prepare_bpf_trampoline(image,
+ err = arch_prepare_bpf_trampoline(NULL, image,
st_map->image + PAGE_SIZE,
&st_ops->func_models[i], 0,
tprogs, NULL);
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 3a283bf97f2f..75244ecb2389 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -827,7 +827,7 @@ static int __init bpf_jit_charge_init(void)
}
pure_initcall(bpf_jit_charge_init);
-static int bpf_jit_charge_modmem(u32 pages)
+int bpf_jit_charge_modmem(u32 pages)
{
if (atomic_long_add_return(pages, &bpf_jit_current) >
(bpf_jit_limit >> PAGE_SHIFT)) {
@@ -840,7 +840,7 @@ static int bpf_jit_charge_modmem(u32 pages)
return 0;
}
-static void bpf_jit_uncharge_modmem(u32 pages)
+void bpf_jit_uncharge_modmem(u32 pages)
{
atomic_long_sub(pages, &bpf_jit_current);
}
diff --git a/kernel/bpf/preload/bpf_preload_kern.c b/kernel/bpf/preload/bpf_preload_kern.c
index 79c5772465f1..53736e52c1df 100644
--- a/kernel/bpf/preload/bpf_preload_kern.c
+++ b/kernel/bpf/preload/bpf_preload_kern.c
@@ -60,9 +60,12 @@ static int finish(void)
&magic, sizeof(magic), &pos);
if (n != sizeof(magic))
return -EPIPE;
+
tgid = umd_ops.info.tgid;
- wait_event(tgid->wait_pidfd, thread_group_exited(tgid));
- umd_ops.info.tgid = NULL;
+ if (tgid) {
+ wait_event(tgid->wait_pidfd, thread_group_exited(tgid));
+ umd_cleanup_helper(&umd_ops.info);
+ }
return 0;
}
@@ -80,10 +83,18 @@ static int __init load_umd(void)
static void __exit fini_umd(void)
{
+ struct pid *tgid;
+
bpf_preload_ops = NULL;
+
/* kill UMD in case it's still there due to earlier error */
- kill_pid(umd_ops.info.tgid, SIGKILL, 1);
- umd_ops.info.tgid = NULL;
+ tgid = umd_ops.info.tgid;
+ if (tgid) {
+ kill_pid(tgid, SIGKILL, 1);
+
+ wait_event(tgid->wait_pidfd, thread_group_exited(tgid));
+ umd_cleanup_helper(&umd_ops.info);
+ }
umd_unload_blob(&umd_ops.info);
}
late_initcall(load_umd);
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index c859bc46d06c..250503482cda 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -854,6 +854,11 @@ static int map_create(union bpf_attr *attr)
err = PTR_ERR(btf);
goto free_map;
}
+ if (btf_is_kernel(btf)) {
+ btf_put(btf);
+ err = -EACCES;
+ goto free_map;
+ }
map->btf = btf;
if (attr->btf_value_type_id) {
diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c
index 7bc3b3209224..1f3a4be4b175 100644
--- a/kernel/bpf/trampoline.c
+++ b/kernel/bpf/trampoline.c
@@ -57,19 +57,10 @@ void bpf_image_ksym_del(struct bpf_ksym *ksym)
PAGE_SIZE, true, ksym->name);
}
-static void bpf_trampoline_ksym_add(struct bpf_trampoline *tr)
-{
- struct bpf_ksym *ksym = &tr->ksym;
-
- snprintf(ksym->name, KSYM_NAME_LEN, "bpf_trampoline_%llu", tr->key);
- bpf_image_ksym_add(tr->image, ksym);
-}
-
static struct bpf_trampoline *bpf_trampoline_lookup(u64 key)
{
struct bpf_trampoline *tr;
struct hlist_head *head;
- void *image;
int i;
mutex_lock(&trampoline_mutex);
@@ -84,14 +75,6 @@ static struct bpf_trampoline *bpf_trampoline_lookup(u64 key)
if (!tr)
goto out;
- /* is_root was checked earlier. No need for bpf_jit_charge_modmem() */
- image = bpf_jit_alloc_exec_page();
- if (!image) {
- kfree(tr);
- tr = NULL;
- goto out;
- }
-
tr->key = key;
INIT_HLIST_NODE(&tr->hlist);
hlist_add_head(&tr->hlist, head);
@@ -99,9 +82,6 @@ static struct bpf_trampoline *bpf_trampoline_lookup(u64 key)
mutex_init(&tr->mutex);
for (i = 0; i < BPF_TRAMP_MAX; i++)
INIT_HLIST_HEAD(&tr->progs_hlist[i]);
- tr->image = image;
- INIT_LIST_HEAD_RCU(&tr->ksym.lnode);
- bpf_trampoline_ksym_add(tr);
out:
mutex_unlock(&trampoline_mutex);
return tr;
@@ -185,10 +165,142 @@ bpf_trampoline_get_progs(const struct bpf_trampoline *tr, int *total)
return tprogs;
}
+static void __bpf_tramp_image_put_deferred(struct work_struct *work)
+{
+ struct bpf_tramp_image *im;
+
+ im = container_of(work, struct bpf_tramp_image, work);
+ bpf_image_ksym_del(&im->ksym);
+ bpf_jit_free_exec(im->image);
+ bpf_jit_uncharge_modmem(1);
+ percpu_ref_exit(&im->pcref);
+ kfree_rcu(im, rcu);
+}
+
+/* callback, fexit step 3 or fentry step 2 */
+static void __bpf_tramp_image_put_rcu(struct rcu_head *rcu)
+{
+ struct bpf_tramp_image *im;
+
+ im = container_of(rcu, struct bpf_tramp_image, rcu);
+ INIT_WORK(&im->work, __bpf_tramp_image_put_deferred);
+ schedule_work(&im->work);
+}
+
+/* callback, fexit step 2. Called after percpu_ref_kill confirms. */
+static void __bpf_tramp_image_release(struct percpu_ref *pcref)
+{
+ struct bpf_tramp_image *im;
+
+ im = container_of(pcref, struct bpf_tramp_image, pcref);
+ call_rcu_tasks(&im->rcu, __bpf_tramp_image_put_rcu);
+}
+
+/* callback, fexit or fentry step 1 */
+static void __bpf_tramp_image_put_rcu_tasks(struct rcu_head *rcu)
+{
+ struct bpf_tramp_image *im;
+
+ im = container_of(rcu, struct bpf_tramp_image, rcu);
+ if (im->ip_after_call)
+ /* the case of fmod_ret/fexit trampoline and CONFIG_PREEMPTION=y */
+ percpu_ref_kill(&im->pcref);
+ else
+ /* the case of fentry trampoline */
+ call_rcu_tasks(&im->rcu, __bpf_tramp_image_put_rcu);
+}
+
+static void bpf_tramp_image_put(struct bpf_tramp_image *im)
+{
+ /* The trampoline image that calls original function is using:
+ * rcu_read_lock_trace to protect sleepable bpf progs
+ * rcu_read_lock to protect normal bpf progs
+ * percpu_ref to protect trampoline itself
+ * rcu tasks to protect trampoline asm not covered by percpu_ref
+ * (which are few asm insns before __bpf_tramp_enter and
+ * after __bpf_tramp_exit)
+ *
+ * The trampoline is unreachable before bpf_tramp_image_put().
+ *
+ * First, patch the trampoline to avoid calling into fexit progs.
+ * The progs will be freed even if the original function is still
+ * executing or sleeping.
+ * In case of CONFIG_PREEMPT=y use call_rcu_tasks() to wait on
+ * first few asm instructions to execute and call into
+ * __bpf_tramp_enter->percpu_ref_get.
+ * Then use percpu_ref_kill to wait for the trampoline and the original
+ * function to finish.
+ * Then use call_rcu_tasks() to make sure few asm insns in
+ * the trampoline epilogue are done as well.
+ *
+ * In !PREEMPT case the task that got interrupted in the first asm
+ * insns won't go through an RCU quiescent state which the
+ * percpu_ref_kill will be waiting for. Hence the first
+ * call_rcu_tasks() is not necessary.
+ */
+ if (im->ip_after_call) {
+ int err = bpf_arch_text_poke(im->ip_after_call, BPF_MOD_JUMP,
+ NULL, im->ip_epilogue);
+ WARN_ON(err);
+ if (IS_ENABLED(CONFIG_PREEMPTION))
+ call_rcu_tasks(&im->rcu, __bpf_tramp_image_put_rcu_tasks);
+ else
+ percpu_ref_kill(&im->pcref);
+ return;
+ }
+
+ /* The trampoline without fexit and fmod_ret progs doesn't call original
+ * function and doesn't use percpu_ref.
+ * Use call_rcu_tasks_trace() to wait for sleepable progs to finish.
+ * Then use call_rcu_tasks() to wait for the rest of trampoline asm
+ * and normal progs.
+ */
+ call_rcu_tasks_trace(&im->rcu, __bpf_tramp_image_put_rcu_tasks);
+}
+
+static struct bpf_tramp_image *bpf_tramp_image_alloc(u64 key, u32 idx)
+{
+ struct bpf_tramp_image *im;
+ struct bpf_ksym *ksym;
+ void *image;
+ int err = -ENOMEM;
+
+ im = kzalloc(sizeof(*im), GFP_KERNEL);
+ if (!im)
+ goto out;
+
+ err = bpf_jit_charge_modmem(1);
+ if (err)
+ goto out_free_im;
+
+ err = -ENOMEM;
+ im->image = image = bpf_jit_alloc_exec_page();
+ if (!image)
+ goto out_uncharge;
+
+ err = percpu_ref_init(&im->pcref, __bpf_tramp_image_release, 0, GFP_KERNEL);
+ if (err)
+ goto out_free_image;
+
+ ksym = &im->ksym;
+ INIT_LIST_HEAD_RCU(&ksym->lnode);
+ snprintf(ksym->name, KSYM_NAME_LEN, "bpf_trampoline_%llu_%u", key, idx);
+ bpf_image_ksym_add(image, ksym);
+ return im;
+
+out_free_image:
+ bpf_jit_free_exec(im->image);
+out_uncharge:
+ bpf_jit_uncharge_modmem(1);
+out_free_im:
+ kfree(im);
+out:
+ return ERR_PTR(err);
+}
+
static int bpf_trampoline_update(struct bpf_trampoline *tr)
{
- void *old_image = tr->image + ((tr->selector + 1) & 1) * PAGE_SIZE/2;
- void *new_image = tr->image + (tr->selector & 1) * PAGE_SIZE/2;
+ struct bpf_tramp_image *im;
struct bpf_tramp_progs *tprogs;
u32 flags = BPF_TRAMP_F_RESTORE_REGS;
int err, total;
@@ -198,41 +310,42 @@ static int bpf_trampoline_update(struct bpf_trampoline *tr)
return PTR_ERR(tprogs);
if (total == 0) {
- err = unregister_fentry(tr, old_image);
+ err = unregister_fentry(tr, tr->cur_image->image);
+ bpf_tramp_image_put(tr->cur_image);
+ tr->cur_image = NULL;
tr->selector = 0;
goto out;
}
+ im = bpf_tramp_image_alloc(tr->key, tr->selector);
+ if (IS_ERR(im)) {
+ err = PTR_ERR(im);
+ goto out;
+ }
+
if (tprogs[BPF_TRAMP_FEXIT].nr_progs ||
tprogs[BPF_TRAMP_MODIFY_RETURN].nr_progs)
flags = BPF_TRAMP_F_CALL_ORIG | BPF_TRAMP_F_SKIP_FRAME;
- /* Though the second half of trampoline page is unused a task could be
- * preempted in the middle of the first half of trampoline and two
- * updates to trampoline would change the code from underneath the
- * preempted task. Hence wait for tasks to voluntarily schedule or go
- * to userspace.
- * The same trampoline can hold both sleepable and non-sleepable progs.
- * synchronize_rcu_tasks_trace() is needed to make sure all sleepable
- * programs finish executing.
- * Wait for these two grace periods together.
- */
- synchronize_rcu_mult(call_rcu_tasks, call_rcu_tasks_trace);
-
- err = arch_prepare_bpf_trampoline(new_image, new_image + PAGE_SIZE / 2,
+ err = arch_prepare_bpf_trampoline(im, im->image, im->image + PAGE_SIZE,
&tr->func.model, flags, tprogs,
tr->func.addr);
if (err < 0)
goto out;
- if (tr->selector)
+ WARN_ON(tr->cur_image && tr->selector == 0);
+ WARN_ON(!tr->cur_image && tr->selector);
+ if (tr->cur_image)
/* progs already running at this address */
- err = modify_fentry(tr, old_image, new_image);
+ err = modify_fentry(tr, tr->cur_image->image, im->image);
else
/* first time registering */
- err = register_fentry(tr, new_image);
+ err = register_fentry(tr, im->image);
if (err)
goto out;
+ if (tr->cur_image)
+ bpf_tramp_image_put(tr->cur_image);
+ tr->cur_image = im;
tr->selector++;
out:
kfree(tprogs);
@@ -364,17 +477,12 @@ void bpf_trampoline_put(struct bpf_trampoline *tr)
goto out;
if (WARN_ON_ONCE(!hlist_empty(&tr->progs_hlist[BPF_TRAMP_FEXIT])))
goto out;
- bpf_image_ksym_del(&tr->ksym);
- /* This code will be executed when all bpf progs (both sleepable and
- * non-sleepable) went through
- * bpf_prog_put()->call_rcu[_tasks_trace]()->bpf_prog_free_deferred().
- * Hence no need for another synchronize_rcu_tasks_trace() here,
- * but synchronize_rcu_tasks() is still needed, since trampoline
- * may not have had any sleepable programs and we need to wait
- * for tasks to get out of trampoline code before freeing it.
+ /* This code will be executed even when the last bpf_tramp_image
+ * is alive. All progs are detached from the trampoline and the
+ * trampoline image is patched with jmp into epilogue to skip
+ * fexit progs. The fentry-only trampoline will be freed via
+ * multiple rcu callbacks.
*/
- synchronize_rcu_tasks();
- bpf_jit_free_exec(tr->image);
hlist_del(&tr->hlist);
kfree(tr);
out:
@@ -478,8 +586,18 @@ void notrace __bpf_prog_exit_sleepable(struct bpf_prog *prog, u64 start)
rcu_read_unlock_trace();
}
+void notrace __bpf_tramp_enter(struct bpf_tramp_image *tr)
+{
+ percpu_ref_get(&tr->pcref);
+}
+
+void notrace __bpf_tramp_exit(struct bpf_tramp_image *tr)
+{
+ percpu_ref_put(&tr->pcref);
+}
+
int __weak
-arch_prepare_bpf_trampoline(void *image, void *image_end,
+arch_prepare_bpf_trampoline(struct bpf_tramp_image *tr, void *image, void *image_end,
const struct btf_func_model *m, u32 flags,
struct bpf_tramp_progs *tprogs,
void *orig_call)
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index f9096b049cd6..999bf36ffeb1 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -6045,10 +6045,14 @@ static int retrieve_ptr_limit(const struct bpf_reg_state *ptr_reg,
{
bool mask_to_left = (opcode == BPF_ADD && off_is_neg) ||
(opcode == BPF_SUB && !off_is_neg);
- u32 off;
+ u32 off, max;
switch (ptr_reg->type) {
case PTR_TO_STACK:
+ /* Offset 0 is out-of-bounds, but acceptable start for the
+ * left direction, see BPF_REG_FP.
+ */
+ max = MAX_BPF_STACK + mask_to_left;
/* Indirect variable offset stack access is prohibited in
* unprivileged mode so it's not handled here.
*/
@@ -6056,8 +6060,8 @@ static int retrieve_ptr_limit(const struct bpf_reg_state *ptr_reg,
if (mask_to_left)
*ptr_limit = MAX_BPF_STACK + off;
else
- *ptr_limit = -off;
- return 0;
+ *ptr_limit = -off - 1;
+ return *ptr_limit >= max ? -ERANGE : 0;
case PTR_TO_MAP_KEY:
/* Currently, this code is not exercised as the only use
* is bpf_for_each_map_elem() helper which requires
@@ -6072,13 +6076,14 @@ static int retrieve_ptr_limit(const struct bpf_reg_state *ptr_reg,
}
return 0;
case PTR_TO_MAP_VALUE:
+ max = ptr_reg->map_ptr->value_size;
if (mask_to_left) {
*ptr_limit = ptr_reg->umax_value + ptr_reg->off;
} else {
off = ptr_reg->smin_value + ptr_reg->off;
- *ptr_limit = ptr_reg->map_ptr->value_size - off;
+ *ptr_limit = ptr_reg->map_ptr->value_size - off - 1;
}
- return 0;
+ return *ptr_limit >= max ? -ERANGE : 0;
default:
return -EINVAL;
}
@@ -6131,6 +6136,7 @@ static int sanitize_ptr_alu(struct bpf_verifier_env *env,
u32 alu_state, alu_limit;
struct bpf_reg_state tmp;
bool ret;
+ int err;
if (can_skip_alu_sanitation(env, insn))
return 0;
@@ -6146,10 +6152,13 @@ static int sanitize_ptr_alu(struct bpf_verifier_env *env,
alu_state |= ptr_is_dst_reg ?
BPF_ALU_SANITIZE_SRC : BPF_ALU_SANITIZE_DST;
- if (retrieve_ptr_limit(ptr_reg, &alu_limit, opcode, off_is_neg))
- return 0;
- if (update_alu_sanitation_state(aux, alu_state, alu_limit))
- return -EACCES;
+ err = retrieve_ptr_limit(ptr_reg, &alu_limit, opcode, off_is_neg);
+ if (err < 0)
+ return err;
+
+ err = update_alu_sanitation_state(aux, alu_state, alu_limit);
+ if (err < 0)
+ return err;
do_sim:
/* Simulate and find potential out-of-bounds access under
* speculative execution from truncation as a result of
@@ -6301,7 +6310,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
case BPF_ADD:
ret = sanitize_ptr_alu(env, insn, ptr_reg, dst_reg, smin_val < 0);
if (ret < 0) {
- verbose(env, "R%d tried to add from different maps or paths\n", dst);
+ verbose(env, "R%d tried to add from different maps, paths, or prohibited types\n", dst);
return ret;
}
/* We can take a fixed offset as long as it doesn't overflow
@@ -6356,7 +6365,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
case BPF_SUB:
ret = sanitize_ptr_alu(env, insn, ptr_reg, dst_reg, smin_val < 0);
if (ret < 0) {
- verbose(env, "R%d tried to sub from different maps or paths\n", dst);
+ verbose(env, "R%d tried to sub from different maps, paths, or prohibited types\n", dst);
return ret;
}
if (dst_reg == off_reg) {
@@ -9276,6 +9285,10 @@ static int check_btf_info(struct bpf_verifier_env *env,
btf = btf_get_by_fd(attr->prog_btf_fd);
if (IS_ERR(btf))
return PTR_ERR(btf);
+ if (btf_is_kernel(btf)) {
+ btf_put(btf);
+ return -EACCES;
+ }
env->prog->aux->btf = btf;
err = check_btf_func(env, attr, uattr);
@@ -11916,7 +11929,7 @@ static int do_misc_fixups(struct bpf_verifier_env *env)
off_reg = issrc ? insn->src_reg : insn->dst_reg;
if (isneg)
*patch++ = BPF_ALU64_IMM(BPF_MUL, off_reg, -1);
- *patch++ = BPF_MOV32_IMM(BPF_REG_AX, aux->alu_limit - 1);
+ *patch++ = BPF_MOV32_IMM(BPF_REG_AX, aux->alu_limit);
*patch++ = BPF_ALU64_REG(BPF_SUB, BPF_REG_AX, off_reg);
*patch++ = BPF_ALU64_REG(BPF_OR, BPF_REG_AX, off_reg);
*patch++ = BPF_ALU64_IMM(BPF_NEG, BPF_REG_AX, 0);
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-11 06:02:09 +0000