Luiz Capitulino reported the test_verifier test failed:
"precise: ST insn causing spi > allocated_stack".
And it was introduced by the following upstream commit:
ecdf985d7615 ("bpf: track immediate values written to stack by BPF_ST instruction")
Eduard's investigation [4] shows that test failure is not a bug, but a
difference in BPF verifier behavior between upstream, where commits
[1,2,3] by Andrii are present, and 5.10, where these commits are absent.
Backporting strategy is consistent with Eduard in kernel version 6.1 [5],
but with some conflicts in patch #1, #4 and #6 due to the bpf of 5.10
doesn't support more features.
Commits of Andrii:
[1] be2ef8161572 ("bpf: allow precision tracking for programs with subprogs")
[2] f63181b6ae79 ("bpf: stop setting precise in current state")
[3] 7a830b53c17b ("bpf: aggressively forget precise markings during state checkpointing")
Links:
[4] https://lore.kernel.org/stable/c9b10a8a551edafdfec855fbd35757c6238ad258.camel@xxxxxxxxx/
[5] https://lore.kernel.org/all/20230724124223.1176479-2-eddyz87@xxxxxxxxx/
Andrii Nakryiko (4):
bpf: allow precision tracking for programs with subprogs
bpf: stop setting precise in current state
bpf: aggressively forget precise markings during state checkpointing
selftests/bpf: make test_align selftest more robust
Ilya Leoshkevich (1):
selftests/bpf: Fix sk_assign on s390x
Yonghong Song (1):
selftests/bpf: Workaround verification failure for
fexit_bpf2bpf/func_replace_return_code
kernel/bpf/verifier.c | 175 ++++++++++++++++--
.../testing/selftests/bpf/prog_tests/align.c | 36 ++--
.../selftests/bpf/prog_tests/sk_assign.c | 25 ++-
.../selftests/bpf/progs/connect4_prog.c | 2 +-
.../selftests/bpf/progs/test_sk_assign.c | 11 ++
.../bpf/progs/test_sk_assign_libbpf.c | 3 +
6 files changed, 219 insertions(+), 33 deletions(-)
create mode 100644 tools/testing/selftests/bpf/progs/test_sk_assign_libbpf.c
--
2.25.1
