linux/kernel/bpf
David Vernet 51302c951c bpf: Teach verifier that trusted PTR_TO_BTF_ID pointers are non-NULL
In reg_type_not_null(), we currently assume that a pointer may be NULL
if it has the PTR_MAYBE_NULL modifier, or if it doesn't belong to one of
several base type of pointers that are never NULL-able. For example,
PTR_TO_CTX, PTR_TO_MAP_VALUE, etc.

It turns out that in some cases, PTR_TO_BTF_ID can never be NULL as
well, though we currently don't specify it. For example, if you had the
following program:

SEC("tc")
long example_refcnt_fail(void *ctx)
{
	struct bpf_cpumask *mask1, *mask2;

	mask1 = bpf_cpumask_create();
	mask2 = bpf_cpumask_create();

        if (!mask1 || !mask2)
		goto error_release;

	bpf_cpumask_test_cpu(0, (const struct cpumask *)mask1);
	bpf_cpumask_test_cpu(0, (const struct cpumask *)mask2);

error_release:
	if (mask1)
		bpf_cpumask_release(mask1);
	if (mask2)
		bpf_cpumask_release(mask2);
	return ret;
}

The verifier will incorrectly fail to load the program, thinking
(unintuitively) that we have a possibly-unreleased reference if the mask
is NULL, because we (correctly) don't issue a bpf_cpumask_release() on
the NULL path.

The reason the verifier gets confused is due to the fact that we don't
explicitly tell the verifier that trusted PTR_TO_BTF_ID pointers can
never be NULL. Basically, if we successfully get past the if check
(meaning both pointers go from ptr_or_null_bpf_cpumask to
ptr_bpf_cpumask), the verifier will correctly assume that the references
need to be dropped on any possible branch that leads to program exit.
However, it will _incorrectly_ think that the ptr == NULL branch is
possible, and will erroneously detect it as a branch on which we failed
to drop the reference.

The solution is of course to teach the verifier that trusted
PTR_TO_BTF_ID pointers can never be NULL, so that it doesn't incorrectly
think it's possible for the reference to be present on the ptr == NULL
branch.

A follow-on patch will add a selftest that verifies this behavior.

Signed-off-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20230602150112.1494194-1-void@manifault.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2023-06-05 14:36:57 -07:00
..
preload bpf: Replace all non-returning strlcpy with strscpy 2023-05-31 13:04:20 +02:00
arraymap.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
bloom_filter.c bpf: compute hashes in bloom filter similar to hashmap 2023-04-02 08:44:49 -07:00
bpf_cgrp_storage.c bpf: Teach verifier that certain helpers accept NULL pointer. 2023-04-04 16:57:16 -07:00
bpf_inode_storage.c Networking changes for 6.4. 2023-04-26 16:07:23 -07:00
bpf_iter.c bpf: implement numbers iterator 2023-03-08 16:19:51 -08:00
bpf_local_storage.c bpf: Handle NULL in bpf_local_storage_free. 2023-04-12 10:27:50 -07:00
bpf_lru_list.c bpf: Address KCSAN report on bpf_lru_list 2023-05-12 12:01:03 -07:00
bpf_lru_list.h bpf: Address KCSAN report on bpf_lru_list 2023-05-12 12:01:03 -07:00
bpf_lsm.c bpf: Fix the kernel crash caused by bpf_setsockopt(). 2023-01-26 23:26:40 -08:00
bpf_struct_ops_types.h
bpf_struct_ops.c bpf: Check IS_ERR for the bpf_map_get() return value 2023-03-24 12:40:47 -07:00
bpf_task_storage.c bpf: Teach verifier that certain helpers accept NULL pointer. 2023-04-04 16:57:16 -07:00
btf.c bpf: Silence a warning in btf_type_id_size() 2023-05-30 14:34:46 -07:00
cgroup_iter.c cgroup: bpf: use cgroup_lock()/cgroup_unlock() wrappers 2023-03-17 12:07:13 -10:00
cgroup.c bpf-next-for-netdev 2023-05-16 19:50:05 -07:00
core.c bpf: Support 64-bit pointers to kfuncs 2023-04-13 21:36:41 -07:00
cpumap.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
cpumask.c bpf: Treat KF_RELEASE kfuncs as KF_TRUSTED_ARGS 2023-03-25 16:56:22 -07:00
devmap.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
disasm.c
disasm.h
dispatcher.c bpf: Synchronize dispatcher update with bpf_dispatcher_xdp_func 2022-12-14 12:02:14 -08:00
hashtab.c bpf: fix a memory leak in the LRU and LRU_PERCPU hash maps 2023-05-22 10:26:39 -07:00
helpers.c bpf: Make bpf_refcount_acquire fallible for non-owning refs 2023-06-05 13:17:20 -07:00
inode.c bpf: Support O_PATH FDs in BPF_OBJ_PIN and BPF_OBJ_GET commands 2023-05-23 23:31:42 +02:00
Kconfig
link_iter.c
local_storage.c cgroup changes for v6.4-rc1 2023-04-29 10:05:22 -07:00
log.c bpf: drop unnecessary user-triggerable WARN_ONCE in verifierl log 2023-05-16 22:34:50 -07:00
lpm_trie.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
Makefile bpf: Split off basic BPF verifier log into separate file 2023-04-11 18:05:42 +02:00
map_in_map.c bpf: Remove btf_field_offs, use btf_record's fields instead 2023-04-15 17:36:49 -07:00
map_in_map.h
map_iter.c
memalloc.c bpf: Add a few bpf mem allocator functions 2023-03-25 19:52:51 -07:00
mmap_unlock_work.h
net_namespace.c
offload.c bpf: netdev: init the offload table earlier 2023-05-15 07:07:41 -07:00
percpu_freelist.c bpf: Initialize same number of free nodes for each pcpu_freelist 2022-11-11 12:05:14 -08:00
percpu_freelist.h
prog_iter.c
queue_stack_maps.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
reuseport_array.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
ringbuf.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
stackmap.c bpf: return long from bpf_map_ops funcs 2023-03-22 15:11:30 -07:00
syscall.c bpf: Fix bad unlock balance on freeze_mutex 2023-05-26 12:16:12 +02:00
sysfs_btf.c
task_iter.c bpf: keep a reference to the mm, in case the task is dead. 2022-12-28 14:11:48 -08:00
tnum.c
trampoline.c bpf: Fix memleak due to fentry attach failure 2023-05-15 23:41:59 +02:00
verifier.c bpf: Teach verifier that trusted PTR_TO_BTF_ID pointers are non-NULL 2023-06-05 14:36:57 -07:00