Files
apk-reverse/tests/unit/test_hw_watch.py
T
root ab22b225ff Hardware pass: hardware watchpoints, ETE differential tracing, kernel-side event injection
Add the layer under the software tooling, in the repository's shape: what runs
on a host is measured, and every device-bound or kernel-bound piece is a gated,
labelled template.

- scripts/hw_watch.py: arm64 hardware watchpoints/execution breakpoints. WCR/
  BCR encoders with the 8-byte-window and alignment rules (unit-tested to the
  canonical 0x1FFD/0x3F05), a self-contained on-device ptrace helper that
  compiles clean as a static aarch64 ELF (observed: cross gcc 16.2), adb
  orchestration, and hit decoding to module+offset. The stealth ledger is
  stated three times on purpose: the ptrace route leaves TracerPid visible.
- scripts/ete_capture.py: CoreSight ETE capture with the five gates (kernel
  era, coresight sysfs, cs_etm PMU, perf-with-CORESIGHT, permission), the
  perf/OpenCSD build recipe, and the per-thread record wrapper whose
  --control one-instruction run must pass before any real capture.
- scripts/trace_diff.py: differential reduction of two execution traces into
  first-divergence / divergent regions / blocks-unique-to-each-side.
  Producer-agnostic: Stalker BLK logs (raw or run_probe-wrapped) and decoded
  cs_etm address streams alike, host-tested on synthetic flattened-dispatcher
  fixtures.
- scripts/kernel_instrument_scaffold.py: KPM/LKM watchpoint + register-rewrite
  templates plus an honest userspace carrier; every kernel file carries a
  NOT COMPILED AND NOT LOADED banner that verify enforces, and the
  register-rewrite action is fenced as experimental in the template itself.
- references/hardware-watchpoints.md, references/ete-diff-tracing.md,
  references/kernel-event-injection.md: the mechanism, the gates, the traps,
  and the composition recipes (watch the key buffer; diff two runs; flip the
  decision at the trap), with hand-offs instead of duplication.
- Registration: 3 symptom rows (SKILL.md + routing mirror), 3 reference rows,
  4 script rows, coverage/route-inventory/evidence rows, capability-matrix
  entries, known-limitations entry, eval case 4, README trees and coverage
  paragraphs in both languages, and three EXTENSION evidence records stating
  exactly what was measured on this host (encoders, compile, fixtures,
  generate/verify) and what was not (no watchpoint armed, no cs_etm stream,
  nothing kernel-side compiled or loaded anywhere).

Hardware pass 2: SPE/BRBE crash forensics, address symbolization, VM de-virtualization

Calibrated against the two public obfuscator implementations the user supplied
(dpkg123/ollvm-aosp-toolchain: Hikari-family and Pluto-family flattening, the
xvmp IR virtualizer at 63fec41f/9b7cf75e; fetched read-only into
tools/_work/refs, git-ignored). Source facts observed, runtime inferred.

- scripts/trace_addr2sym.py: runtime addresses -> symbol + static offset,
  zero-dependency ELF64 (PT_LOAD vaddr<->file-off conversion, .dynsym/.symtab,
  stripped section+off fallback). Host-verified against a real cross-compiled
  aarch64 .so with readelf as the independent cross-check, plus a hand-built
  minimal ELF covering vaddr!=offset; accepts hw_watch hits, trace mod+off
  keys, or raw address lists. Fixes on the way: --base int coercion, and
  nearest-below symbol semantics for PCs in inter-symbol gaps.
- scripts/vm_handler_slice.py: slice an interpreter trace into handler
  segments (dispatcher detection = histogram top + predecessor diversity; the
  first 4x-margin rule refused the calibrated fixture itself and was replaced),
  signature clustering, open-tail handling, and the guest-opcode trajectory via
  an optional handler->opcode map. Calibrated against xvmp's documented shape:
  vm_exec(code,size,vm_args,num) stubs, boxed arguments observable before
  interpretation, 11 opcodes + 19 arithmetic sub-ops, const-materialization
  dispatch as a distinct segment.
- references/hardware-sampling-and-crash.md: SPE (data flow without a known
  address; PC + data address per sample), BRBE (branch records when a crash
  stack cannot be trusted), the crash loop (tombstone -> branch records ->
  symbolize), and what each adds to the OLLVM/VMP routes. All
  documented-and-gated: no SPE sample or BRBE record was captured here.
- Flattening-recipe corrections from source: Hikari keeps a frame-resident
  switchVar written through an indirect pointer with scramble32-permuted cases;
  Pluto stores switchVar/switchXor rolling-XOR so memory only ever holds noise
  -- hardware-watchpoints.md's state-variable recipe now distinguishes the
  families, and ete-diff-tracing.md carries the family table plus the xvmp
  shape. AntiDebugging injects raw svc inline asm with randomized fields, so
  byte signatures cannot survive it; semantic site-location can.
- Registration: 3 symptom rows (untrustworthy crash stack; runtime addresses
  needing symbols; interpreter loop reading as noise), reference/script rows,
  coverage/evidence/capability-matrix entries, README trees in both languages,
  and EXTENSION-sampling-crash.md recording exactly what was measured (real .so
  symbolization, slicer fixtures, the slicer's self-found dispatcher-rule bug)
  and what was not (no SPE/BRBE capture, no hardened binary ever traced).

Fix all PR #1 review findings by severity; un-entrench the tests that pinned them

Major (device-breaking / wrong results):
- hw_watch: match the arm64 user_hwdebug_state slot ABI (address register
  first, control second -- the swap armed nothing correctly)
- hw_watch: every post-seize error path now clears slots and detaches;
  PTRACE_O_EXITKILL turned an analysis failure into a SIGKILL of the target
- hw_watch: forward genuine target SIGTRAPs (raise/tgkill/int3); only ptrace
  event stops are consumed -- the reference's signal-preservation promise
- hw_watch: run checks push/build/chmod and the helper's exit status; a dead
  capture is RESULT=blocked, never ok-no-hits
- hw_watch: decode offsets from the module base, not each mapping's start
- ete_capture: --control now requires the control record to DECODE
  (perf script event lines), the documented gate, not just a zero exit
- ete_capture: -o before the -- workload separator in --cmd mode; all
  values shell-quoted into su -c; stale remote perf.data removed first;
  perf exit status checked (timeout's 124 is the normal window end)
- kernel scaffold: --set-reg restricted to x0..x30; the rewrite is gated by
  its own breakpoint event so a watchpoint hit cannot run it; LKM pins the
  task (get/put_task_struct) across registration and calls the four-argument
  register_user_hw_breakpoint; --break rows initialize .addr/.len/.bp_type
  (there is no .pc member); --id validated against the module-id grammar
- trace_addr2sym: exact module matching (path or basename, never substring);
  load bias aligned against the PT_LOAD (bias = start - (foff - p_offset)
  - p_vaddr); PCs outside the ELF span are refused instead of landing on the
  last symbol; --json stdout is a pure document (RESULT= in text mode only)
- trace_diff: block keys need 0x or >=8 hex digits (PIDs/counts no longer
  parse); vm_handler_slice: dispatcher needs predecessor diversity, no
  frequency-margin fallback; both JSON outputs pure (check_commands' rule)

Minor/nitpick:
- capability-matrix + evidence-summary: trace reduction status ok->partial
  (no real-target producer ran it; the file's own definition of ok)
- hardware-sampling-and-crash.md and the script examples carry --elf
- kernel-event-injection.md no longer claims the callback's first statement
  is a filter (it prints unfiltered; the filter is a required change)
- Elf64 caches symbol addresses for per-row bisect
- docstring coverage of the touched surface: 16.67% -> 100%

Tests: every json.loads workaround stripping RESULT= is gone; new regression
tests cover the slot ABI, release-on-error, SIGTRAP forwarding, non-first
mapping offsets, -o placement and su -c quoting, the control decode gate,
x31/--id refusals, the guarded rewrite, outside-ELF PCs and bias math,
short-token refusal and the dispatcher margin rule. The minimal-ELF fixture
now packs a full 8-field Elf64_Phdr (its 7-field pack silently shifted
p_memsz onto p_align). 739 passed; flake8/mypy and all five gates clean.

Address the follow-up review on 0bd8dfe (6 findings)

- ete_capture: two-layer quoting for the control-record, decode and cleanup
  commands -- shlex.quote spliced between literal single quotes broke out of
  su -c for any spaced output name; --seconds must be positive (timeout 0
  disables the timer), refused before any adb check
- kernel scaffold: --set-reg now refuses sp/pc as well (the template has no
  rewrite path for them; a generate that silently drops the action is the
  misleading kind of success); both templates arm with a sampling period
  (LKM via hw_breakpoint_init, KPM spells out sample_period=1) -- a zero
  period is a non-sampling event whose callback never fires: armed, silent
- trace_addr2sym: contains_vaddr tests each PT_LOAD's loaded span instead
  of the whole-file extent; a PC in the unmapped gap between segments is
  refused rather than named after the nearest symbol below
- trace_diff: a bare all-digit token is indistinguishable from a date or
  count, so block keys need 0x or a non-decimal hex digit; a date-only log
  stays refused as a capture failure

Tests: two-layer quoting with a spaced control name, --seconds refusal,
sp/pc in the rejection list, sampling-period assertions for both templates,
a two-segment gap fixture, and a date-only parser case. 743 passed; flake8,
mypy and all five gates clean.
2026-10-15 15:33:28 +08:00

252 lines
11 KiB
Python

"""Unit tests for hw_watch.py: encoders, refusals, emit, hit decoding.
What is asserted here is the **host-measured half** of the watchpoint route: the
WCR/BCR bit encodings against the ARM DDI 0487 field map, the alignment rules
that refuse impossible specs locally instead of on the device, the emitted
helper's shape, and the hit-log decoder. Device behaviour (seize/arm/trap) is
deliberately not simulated -- see docs/tool-verification/EXTENSION-hw-watchpoints.md.
"""
import json
import shutil
import subprocess
import pytest
from conftest import load_script, run_script
hw = load_script("hw_watch.py")
# ---------------------------------------------------------------------------
# encoders -- the two canonical values plus the variations
# ---------------------------------------------------------------------------
class TestWatchEncoders:
def test_canonical_8byte_rw(self):
# BAS=0xff, LSC=both(0b11), PAC=EL0(0b10), E=1
"""Canonical 8byte rw."""
assert hw.encode_watch(0x7b2000000000, 8, "rw") == (0x7b2000000000, 0x1FFD)
def test_load_only_and_store_only(self):
"""Load only and store only."""
wvr, wcr_r = hw.encode_watch(0x1000, 8, "r")
wvr, wcr_w = hw.encode_watch(0x1000, 8, "w")
assert wcr_r == 0x1FED # LSC=01 (load)
assert wcr_w == 0x1FF5 # LSC=10 (store)
def test_unaligned_start_shifts_bas(self):
# bytes 4..7 of the window: BAS = 0xf0 << 5 = 0x1E00
"""Unaligned start shifts bas."""
wvr, wcr = hw.encode_watch(0x1004, 4, "w")
assert wvr == 0x1000 # WVR canonicalised to the aligned window
assert wcr & 0x1FE0 == 0x1E00
assert wcr == 0x1E15 # | store | PAC | E
def test_len_1_at_offset_7(self):
"""Len 1 at offset 7."""
wvr, wcr = hw.encode_watch(0x1007, 1, "rw")
assert wcr & 0x1FE0 == (0x80 << 5)
def test_window_crossing_refused(self):
"""Window crossing refused."""
with pytest.raises(hw.PlanError, match="crosses"):
hw.encode_watch(0x1006, 8, "rw")
with pytest.raises(hw.PlanError, match="crosses"):
hw.encode_watch(0x1004, 8, "r")
def test_length_bounds(self):
"""Length bounds."""
with pytest.raises(hw.PlanError):
hw.encode_watch(0x1000, 0, "rw")
with pytest.raises(hw.PlanError):
hw.encode_watch(0x1000, 9, "rw")
def test_access_vocabulary(self):
"""Access vocabulary."""
with pytest.raises(hw.PlanError):
hw.encode_watch(0x1000, 8, "x")
class TestBreakEncoders:
def test_canonical_address_match(self):
# BAS all-ones (0x3F << 8), PMC=EL0 (0b10 << 1), BT=0, E=1
"""Canonical address match."""
assert hw.encode_break(0x7b20001234c0) == (0x7b20001234c0, 0x3F05)
def test_misaligned_pc_refused(self):
"""Misaligned pc refused."""
with pytest.raises(hw.PlanError, match="4-byte aligned"):
hw.encode_break(0x1002)
with pytest.raises(hw.PlanError):
hw.encode_break(0x1001)
def test_kernel_abi_constants_are_the_uapi_values():
"""Pinned because a drifted constant compiles fine and arms nothing."""
assert hw.PTRACE_GETREGSET == 0x4204
assert hw.PTRACE_SETREGSET == 0x4205
assert hw.PTRACE_SEIZE == 0x4206
assert hw.NT_ARM_HW_BREAK == 0x402
assert hw.NT_ARM_HW_WATCH == 0x403
assert hw.TRAP_HWBKPT == 4
assert hw.USER_PT_REGS_BYTES == 272
assert hw.REG_OFFSETS["pc"] == 256 and hw.REG_OFFSETS["sp"] == 248
# ---------------------------------------------------------------------------
# emit -- the helper's shape and buildability
# ---------------------------------------------------------------------------
def test_emit_writes_helper_and_readme(tmp_path):
"""Emit writes helper and readme."""
out = tmp_path / "hww"
r = run_script("hw_watch.py", ["emit", "--out", str(out)])
assert r.returncode == 0, r.output
c = (out / "hwwatch_helper.c").read_text(encoding="utf-8")
readme = (out / "README.md").read_text(encoding="utf-8")
# the ABI surface the encoders promised
for marker in ("NT_ARM_HW_WATCH", "NT_ARM_HW_BREAK", "TRAP_HWBKPT",
"PTRACE_SEIZE", "PTRACE_O_TRACECLONE", "PTRACE_SINGLESTEP",
"PTRACE_LISTEN", "0x3Fu << 8"):
assert marker in c, marker
# the honesty devices
assert "TracerPid" in c
assert "TEMPLATE" in c
assert "aarch64-linux-android24-clang" in readme
def test_emitted_helper_compiles_when_a_cross_toolchain_exists(tmp_path):
"""When gcc targets aarch64 here (it does on this CI image), the emitted
helper must build as a static ELF. Skipped where no compiler is present:
the compile is evidence, not a portability promise."""
cc = shutil.which("gcc")
if cc is None:
pytest.skip("no C compiler on this host")
probe = subprocess.run([cc, "-dumpmachine"], capture_output=True, text=True)
if "aarch64" not in (probe.stdout or ""):
pytest.skip("host gcc does not target aarch64 (%r)" % probe.stdout.strip())
out = tmp_path / "hww"
run_script("hw_watch.py", ["emit", "--out", str(out)])
exe = tmp_path / "hwwatch"
built = subprocess.run([cc, "-static", "-O2", "-o", str(exe),
str(out / "hwwatch_helper.c")],
capture_output=True, text=True)
assert built.returncode == 0, built.stderr
assert exe.stat().st_size > 10000
# ---------------------------------------------------------------------------
# decode -- hits in, module+offset out
# ---------------------------------------------------------------------------
HIT_LOG = """READY pid=1234 watch_slots=4 break_slots=6
ARM tid=1234 watch=1 break=0
ARM tid=1240 watch=1 break=0
HIT tid=1240 kind=watch pc=0x7b20001234c0 addr=0x7b2000feedd0 sp=0x7fc000123000 lr=0x7b2000120000 x0=0x1 x1=0x2 x2=0x3 x3=0x4 x4=0x5 x5=0x6 x6=0x7 x7=0x8
ARMFAIL tid=1250 slot=watch1 capacity=0
HIT tid=1234 kind=break pc=0x7b2000129000 addr=0x7b2000129000 sp=0x7fc000124000 lr=0x0 x0=0x0 x1=0x0 x2=0x0 x3=0x0 x4=0x0 x5=0x0 x6=0x0 x7=0x0
DONE hits=2 reason=seconds
"""
MAPS = """7b2000000000-7b2000200000 r--p 00000000 fd:01 100 /data/app/liba.so
7b2000200000-7b2000400000 r-xp 00200000 fd:01 100 /data/app/liba.so
7b2000400000-7b2000420000 rw-p 00400000 fd:01 100 /data/app/liba.so
"""
def test_parse_hits_tolerates_interleaved_lines(tmp_path):
"""Parse hits tolerates interleaved lines."""
log = tmp_path / "hits.log"
log.write_text(HIT_LOG, encoding="utf-8")
rows = hw.parse_hits(str(log))
assert len(rows) == 2
assert rows[0]["kind"] == "watch" and rows[0]["pc"] == "0x7b20001234c0"
assert rows[1]["kind"] == "break"
def test_decode_resolves_module_offset(tmp_path):
"""Decode resolves module offset."""
log = tmp_path / "hits.log"
log.write_text(HIT_LOG, encoding="utf-8")
maps = tmp_path / "maps.txt"
maps.write_text(MAPS, encoding="utf-8")
r = run_script("hw_watch.py", ["decode", "--hits", str(log), "--maps", str(maps)])
assert r.returncode == 0, r.output
assert "+0x1234c0" in r.output # pc lands in the first r--p range
assert "unresolved_pc=0" in r.output
rj = run_script("hw_watch.py", ["decode", "--hits", str(log), "--maps", str(maps),
"--json"])
doc = json.loads(rj.stdout) # --json stdout is one pure document
assert doc["status"] == "ok" and doc["hits"] == 2
assert doc["rows"][0]["site"].endswith("liba.so+0x1234c0")
def test_decode_offsets_come_from_the_module_base_not_the_mapping(tmp_path):
"""A PC in a library's r-xp mapping (its second mapping) must report the
offset from the MODULE base -- the number IDA/Ghidra expects -- not from
that mapping's own start, which resets the offset at each boundary."""
log = tmp_path / "hits.log"
log.write_text("HIT tid=1 kind=break pc=0x7b2000212345 addr=0x7b2000212345 "
"sp=0x0 lr=0x0 x0=0x0\n", encoding="utf-8")
maps = tmp_path / "maps.txt"
maps.write_text(MAPS, encoding="utf-8")
r = run_script("hw_watch.py", ["decode", "--hits", str(log), "--maps", str(maps)])
assert r.returncode == 0, r.output
assert "liba.so+0x212345" in r.output # not +0x12345 (mapping-relative)
def test_decode_without_maps_leaves_pcs_raw(tmp_path):
"""Decode without maps leaves pcs raw."""
log = tmp_path / "hits.log"
log.write_text(HIT_LOG, encoding="utf-8")
r = run_script("hw_watch.py", ["decode", "--hits", str(log)])
assert r.returncode == 0
assert "unresolved" in r.output and "no --maps" in r.output
def test_plan_json_and_refusal_exit_codes():
"""Plan json and refusal exit codes."""
r = run_script("hw_watch.py", ["plan", "--addr", "0x1000", "--len", "8",
"--access", "rw", "--json"])
assert r.returncode == 0
doc = json.loads(r.stdout)
assert doc["specs"][0]["cr"] == 0x1FFD
r = run_script("hw_watch.py", ["plan", "--addr", "0x1004", "--len", "8"])
assert r.returncode == 2
assert "crosses" in r.output
# ---------------------------------------------------------------------------
# the emitted helper: kernel-ABI regressions pinned in the C source
# ---------------------------------------------------------------------------
def emitted_helper(tmp_path):
"""Emit the on-device helper; return its C source."""
out = tmp_path / "hww"
run_script("hw_watch.py", ["emit", "--out", str(out)])
return (out / "hwwatch_helper.c").read_text(encoding="utf-8")
def test_helper_slot_layout_matches_the_kernel_abi(tmp_path):
"""user_hwdebug_state carries the 64-bit address register FIRST and the
32-bit control register SECOND (arch/arm64 uapi); swapping them arms
every slot with the ctrl bits as the address."""
c = emitted_helper(tmp_path)
assert "struct hwslot { uint64_t wvr; uint32_t wcr; uint32_t pad; };" in c
def test_helper_error_paths_release_seized_threads(tmp_path):
"""PTRACE_O_EXITKILL turns an exit-without-detach into a SIGKILL of the
target: every post-seize failure must route through release_all()."""
c = emitted_helper(tmp_path)
assert "static void release_all(void)" in c
# capacity probe failure and arming failure both detach instead of return
assert c.count("release_all();") >= 3 # cap fail, arm fail, clean exit
# partial seize failure detaches the threads seized so far
assert "EXITKILLed when we exit" in c
def test_helper_forwards_genuine_sigtrap(tmp_path):
"""Only our TRAP_HWBKPT hits and ptrace event stops are consumed; a
target-generated SIGTRAP (raise/tgkill/int3) must be forwarded or the
app's handler never runs."""
c = emitted_helper(tmp_path)
assert "if (event != 0) {" in c # event stops: consumed
assert "genuine" in c and "pt(PTRACE_CONT, tid, 0, sig);" in c