linux.git/tools/testing/selftests/bpf/verifier, branch v5.19

bpf, selftests: Add verifier test case for jmp32's jeq/jne

2022-07-01T19:56:27+00:00

Add a test case to trigger the verifier's incorrect conclusion in the
case of jmp32's jeq/jne. Also here, make use of dead code elimination,
so that we can see the verifier bailing out on unfixed kernels.

Before:

  # ./test_verifier 724
  #724/p jeq32/jne32: bounds checking FAIL
  Failed to load prog 'Permission denied'!
  R4 !read_ok
  verification time 8 usec
  stack depth 0
  processed 8 insns (limit 1000000) max_states_per_insn 0 total_states 1 peak_states 1 mark_read 0
  Summary: 0 PASSED, 0 SKIPPED, 1 FAILED

After:

  # ./test_verifier 724
  #724/p jeq32/jne32: bounds checking OK
  Summary: 1 PASSED, 0 SKIPPED, 0 FAILED

Signed-off-by: Daniel Borkmann 
Signed-off-by: Andrii Nakryiko 
Link: https://lore.kernel.org/bpf/20220701124727.11153-4-daniel@iogearbox.net

bpf, selftests: Add verifier test case for imm=0,umin=0,umax=1 scalar

2022-07-01T19:56:27+00:00

Add a test case to trigger the constant scalar issue which leaves the
register in scalar(imm=0,umin=0,umax=1,var_off=(0x0; 0x0)) state. Make
use of dead code elimination, so that we can see the verifier bailing
out on unfixed kernels. For the condition, we use jle given it checks
on umax bound.

Before:

  # ./test_verifier 743
  #743/p jump & dead code elimination FAIL
  Failed to load prog 'Permission denied'!
  R4 !read_ok
  verification time 11 usec
  stack depth 0
  processed 13 insns (limit 1000000) max_states_per_insn 0 total_states 1 peak_states 1 mark_read 1
  Summary: 0 PASSED, 0 SKIPPED, 1 FAILED

After:

  # ./test_verifier 743
  #743/p jump & dead code elimination OK
  Summary: 1 PASSED, 0 SKIPPED, 0 FAILED

Signed-off-by: Daniel Borkmann 
Signed-off-by: Andrii Nakryiko 
Link: https://lore.kernel.org/bpf/20220701124727.11153-3-daniel@iogearbox.net

bpf: Prepare prog_test_struct kfuncs for runtime tests

2022-05-11T23:57:27+00:00

In an effort to actually test the refcounting logic at runtime, add a
refcount_t member to prog_test_ref_kfunc and use it in selftests to
verify and test the whole logic more exhaustively.

The kfunc calls for prog_test_member do not require runtime refcounting,
as they are only used for verifier selftests, not during runtime
execution. Hence, their implementation now has a WARN_ON_ONCE as it is
not meant to be reachable code at runtime. It is strictly used in tests
triggering failure cases in the verifier. bpf_kfunc_call_memb_release is
called from map free path, since prog_test_member is embedded in map
value for some verifier tests, so we skip WARN_ON_ONCE for it.

Signed-off-by: Kumar Kartikeya Dwivedi 
Link: https://lore.kernel.org/r/20220511194654.765705-3-memxor@gmail.com
Signed-off-by: Alexei Starovoitov

selftests/bpf: Add test for strict BTF type check

2022-04-26T03:26:45+00:00

Ensure that the edge case where first member type was matched
successfully even if it didn't match BTF type of register is caught and
rejected by the verifier.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220424214901.2743946-14-memxor@gmail.com

selftests/bpf: Add verifier tests for kptr

2022-04-26T03:26:44+00:00

Reuse bpf_prog_test functions to test the support for PTR_TO_BTF_ID in
BPF map case, including some tests that verify implementation sanity and
corner cases.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220424214901.2743946-13-memxor@gmail.com

bpf: Tag argument to be released in bpf_func_proto

2022-04-26T00:31:35+00:00

Add a new type flag for bpf_arg_type that when set tells verifier that
for a release function, that argument's register will be the one for
which meta.ref_obj_id will be set, and which will then be released
using release_reference. To capture the regno, introduce a new field
release_regno in bpf_call_arg_meta.

This would be required in the next patch, where we may either pass NULL
or a refcounted pointer as an argument to the release function
bpf_kptr_xchg. Just releasing only when meta.ref_obj_id is set is not
enough, as there is a case where the type of argument needed matches,
but the ref_obj_id is set to 0. Hence, we must enforce that whenever
meta.ref_obj_id is zero, the register that is to be released can only
be NULL for a release function.

Since we now indicate whether an argument is to be released in
bpf_func_proto itself, is_release_function helper has lost its utitlity,
hence refactor code to work without it, and just rely on
meta.release_regno to know when to release state for a ref_obj_id.
Still, the restriction of one release argument and only one ref_obj_id
passed to BPF helper or kfunc remains. This may be lifted in the future.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220424214901.2743946-3-memxor@gmail.com

selftests/bpf: Add tests for kfunc register offset checks

2022-03-05T23:29:36+00:00

Include a few verifier selftests that test against the problems being
fixed by previous commits, i.e. release kfunc always require
PTR_TO_BTF_ID fixed and var_off to be 0, and negative offset is not
permitted and returns a helpful error message.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220304224645.3677453-9-memxor@gmail.com

bpf: Disallow negative offset in check_ptr_off_reg

2022-03-05T23:29:35+00:00

check_ptr_off_reg only allows fixed offset to be set for PTR_TO_BTF_ID,
where reg->off < 0 doesn't make sense. This would shift the pointer
backwards, and fails later in btf_struct_ids_match or btf_struct_walk
due to out of bounds access (since offset is interpreted as unsigned).

Improve the verifier by rejecting this case by using a better error
message for BPF helpers and kfunc, by putting a check inside the
check_func_arg_reg_off function.

Also, update existing verifier selftests to work with new error string.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220304224645.3677453-4-memxor@gmail.com

bpf: Small BPF verifier log improvements

2022-03-03T15:54:10+00:00

In particular these include:

  1) Remove output of inv for scalars in print_verifier_state
  2) Replace inv with scalar in verifier error messages
  3) Remove _value suffixes for umin/umax/s32_min/etc (except map_value)
  4) Remove output of id=0
  5) Remove output of ref_obj_id=0

Signed-off-by: Mykola Lysenko 
Signed-off-by: Daniel Borkmann 
Acked-by: Andrii Nakryiko 
Link: https://lore.kernel.org/bpf/20220301222745.1667206-1-mykolal@fb.com

selftests/bpf: Add test for reg2btf_ids out of bounds access

2022-02-22T04:24:15+00:00

This test tries to pass a PTR_TO_BTF_ID_OR_NULL to the release function,
which would trigger a out of bounds access without the fix in commit
45ce4b4f9009 ("bpf: Fix crash due to out of bounds access into reg2btf_ids.")
but after the fix, it should only index using base_type(reg->type),
which should be less than __BPF_REG_TYPE_MAX, and also not permit any
type flags to be set for the reg->type.

Signed-off-by: Kumar Kartikeya Dwivedi 
Signed-off-by: Alexei Starovoitov 
Link: https://lore.kernel.org/bpf/20220220023138.2224652-1-memxor@gmail.com