[Merge] (MGPipe, P5): land j0 rounds 2-3 - the split-lane private logs wired into CI and the skipped-selection control failure (ID-62)

This commit is contained in:
2026-09-16 10:15:58 -04:00
9 changed files with 156 additions and 36 deletions
+5 -1
View File
@@ -19,7 +19,7 @@
set -u
HERE="$(cd "$(dirname "$0")" && pwd)"
WORK="$(mktemp -d)"
WORK="$(mktemp -d)" || exit 1
trap 'rm -rf "${WORK}"' EXIT
STUB_DIR="${WORK}/stub"
@@ -99,6 +99,10 @@ expect FAILED "a pull library passed the split retrace" -- run_retrace retr
echo
echo "smoke test: ${passes} passed, ${failures} failed"
# Keep the private-file cases on the entry point used by CI and the local gate.
if ! bash "${HERE}/testdata/split_private_log_smoke.sh"; then
failures=$((failures + 1))
fi
if [ "${failures}" -gt 0 ]; then
echo "CONTROL_SMOKE_TEST_FAILED"
exit 1
+44 -12
View File
@@ -5,17 +5,19 @@
# X = revert the message check in each control, i.e. put the controls back in the state ID-46
# finding 8 found them in: a non-zero ctest exit is accepted whatever the failure was.
#
# The smoke test must then FAIL, and it must fail on the two cases that exist for this defect -
# The smoke test must then FAIL on the cases that exist for this defect -
# "unrelated failure with a non-empty selection" and "red without the transport-resolution
# message" - and not merely somewhere. A smoke test that goes red for any other reason when the
# evidence check is removed would not be pinning the evidence check.
# message", plus the missing private-file Fatal - and not merely somewhere. A smoke test that goes
# red for any other reason when the evidence check is removed would not be pinning the evidence check.
set -u
HERE="$(cd "$(dirname "$0")" && pwd)"
WORK="$(mktemp -d)"
trap 'cp "${WORK}/split.orig" "${HERE}/split_negative_controls.sh"; cp "${WORK}/retrace.orig" "${HERE}/retrace_pull_library_control.sh"; rm -rf "${WORK}"' EXIT
WORK="$(mktemp -d)" || exit 1
HELPER="${HERE}/../../MobileGL/MG_IntegrationTest/Harness/split_log_paths.py"
trap 'cp "${WORK}/split.orig" "${HERE}/split_negative_controls.sh"; cp "${WORK}/retrace.orig" "${HERE}/retrace_pull_library_control.sh"; cp "${WORK}/helper.orig" "${HELPER}"; rm -rf "${WORK}"' EXIT
cp "${HERE}/split_negative_controls.sh" "${WORK}/split.orig"
cp "${HERE}/retrace_pull_library_control.sh" "${WORK}/retrace.orig"
cp "${HELPER}" "${WORK}/helper.orig"
echo "=== baseline: the smoke test must be GREEN before anything is perturbed"
if ! bash "${HERE}/control_smoke_test.sh" > "${WORK}/before.log" 2>&1; then
@@ -27,27 +29,35 @@ tail -1 "${WORK}/before.log"
echo
echo "=== perturbation: remove the evidence check from both controls"
python3 - "${HERE}/split_negative_controls.sh" "${HERE}/retrace_pull_library_control.sh" <<'PY'
python3 - "${HERE}/split_negative_controls.sh" "${HERE}/retrace_pull_library_control.sh" <<'PY' || exit 1
import sys
import re
split, retrace = sys.argv[1], sys.argv[2]
for path, needle in ((split, 'grep -qE "${evidence}"'), (retrace, 'grep -qF "${EVIDENCE}"')):
text = open(path).read()
if path == split:
text, count = re.subn(
r'(?m)^ python3 "\$\{log_helper\}" evidence[^\n]*\n[^\n]*\|\| exit 1\n',
' : "private-file evidence check removed by red-check"\n', text)
if count != 1:
raise SystemExit(f"expected exactly one private-file evidence check in {path}, found {count}")
out, hit = [], 0
for line in text.splitlines(keepends=True):
if needle in line and line.lstrip().startswith('if ! '):
if needle in line and line.lstrip().startswith(('if ! ', 'elif ! ')):
indent = line[:len(line) - len(line.lstrip())]
out.append(f"{indent}if false; then\n")
keyword = 'elif' if line.lstrip().startswith('elif ') else 'if'
out.append(f"{indent}{keyword} false; then\n")
hit += 1
else:
out.append(line)
if hit != 1:
raise SystemExit(f"expected exactly one evidence check in {path}, found {hit}")
open(path, 'w').write(''.join(out))
print("both evidence checks reverted to 'any non-zero ctest exit is accepted'")
print("E1 private-file, E3(a) and retrace evidence checks reverted to 'any non-zero ctest exit is accepted'")
PY
echo
echo "=== the smoke test on the reverted controls (it MUST be red, on those two cases)"
echo "=== the smoke test on the reverted controls (it MUST be red on the pinned cases)"
bash "${HERE}/control_smoke_test.sh" > "${WORK}/after.log" 2>&1
rc=$?
cat "${WORK}/after.log"
@@ -61,11 +71,33 @@ fi
missed=0
grep -q "NOT OK unrelated failure with a non-empty selection" "${WORK}/after.log" || missed=1
grep -q "NOT OK red without the transport-resolution message" "${WORK}/after.log" || missed=1
grep -qF "NOT OK missing-fatal: control must report FAILED for its own reason" "${WORK}/after.log" || missed=1
if [ "${missed}" -ne 0 ]; then
echo
echo "RED-CHECK FAILED: the smoke test went red, but not on the two cases the evidence check exists for."
echo "RED-CHECK FAILED: the smoke test went red, but not on the unrelated split, unrelated retrace and missing private-Fatal cases."
exit 1
fi
echo
echo "P5_T1_CONTROL_SMOKE_REDCHECK_OK - removing the evidence check reds exactly the two cases that pin it"
cp "${WORK}/split.orig" "${HERE}/split_negative_controls.sh"
cp "${WORK}/retrace.orig" "${HERE}/retrace_pull_library_control.sh"
echo "=== ID-62 perturbation: remove only the skip check"
python3 - "${HELPER}" <<'PY' || exit 1
from pathlib import Path
import sys
path = Path(sys.argv[1])
text = path.read_text()
needle = ' if skipped:\n'
if text.count(needle) != 1:
raise SystemExit('expected exactly one skip check')
path.write_text(text.replace(needle, ' if False: # skip check removed by red-check\n'))
PY
bash "${HERE}/control_smoke_test.sh" > "${WORK}/skips.log" 2>&1
rc=$?
cat "${WORK}/skips.log"
if [ "${rc}" -eq 0 ] || ! grep -qFx 'NOT OK skipped-selection: control must report FAILED for its own reason' "${WORK}/skips.log"; then
echo 'RED-CHECK FAILED: removing the skip check must red the skipped-selection message assertion'
exit 1
fi
echo 'ID-62 red-once: removing only the skip check made skipped-selection red (smoke rc=1)'
echo "P5_T1_CONTROL_SMOKE_REDCHECK_OK - evidence and skipped-selection checks each made their smoke cases red"
+14 -5
View File
@@ -42,8 +42,7 @@
set -u
if [ "${1:-}" = "--self-test" ]; then
bash "$(dirname "$0")/control_smoke_test.sh" &&
bash "$(dirname "$0")/testdata/split_private_log_smoke.sh"
bash "$(dirname "$0")/control_smoke_test.sh"
exit $?
fi
@@ -65,7 +64,8 @@ python3 "${log_helper}" check "${CTEST}" "$PWD" || exit 1
# runtime fact inside each test process (MG_Config::Transport, ClientSession::Active(),
# ImplementedVerbCount(), read by Harness/SplitRuntimePeek), so the only honest way to ask it from a
# shell is to look at what the entries DID.
"${CTEST}" -L integration-split -j 4 --no-tests=error --output-junit "${junit}"
# The metadata ownership test shares the label but cannot arm runtime controls.
"${CTEST}" -L integration-split -R '^DirectGLES\.Split\.' -j 4 --no-tests=error --output-junit "${junit}"
baseline_rc=$?
if [ ! -f "${junit}" ]; then
@@ -115,10 +115,17 @@ run_control() {
python3 "${log_helper}" reset "${manifest}" "${filter}" || exit 1
out="${CONTROL_TMPDIR}/control-output.txt"
env "$@" "${CTEST}" --output-on-failure -L integration-split -R "${filter}" --no-tests=error > "${out}" 2>&1
result="${CONTROL_TMPDIR}/control.xml"
rm -f "${result}"
env "$@" "${CTEST}" --output-on-failure -L integration-split -R "${filter}" --no-tests=error --output-junit "${result}" > "${out}" 2>&1
control_rc=$?
cat "${out}"
# Inspect JUnit before exit status or private Fatal: a skipped pre-flight can carry both.
label='E3(a)'
[ "${evidence}" != private-barrier-fatal ] || label=E1
python3 "${log_helper}" results "${manifest}" "${filter}" "${result}" "${label}" || exit 1
if [ "${control_rc}" -eq 0 ]; then
echo "::error::${name} left ${matched} split entries GREEN, so the knob it turns is not load-bearing and the gate it controls proves nothing."
exit 1
@@ -136,7 +143,9 @@ run_control() {
echo "${name} turned ${matched} selected entries red, and the red carries the scenario's own diagnostic, as it must"
}
# E1: c1 ClientSession::Post emits MGLOG_F Fatal{BarrierViolation, "<slot>"}.
# E1: c1 ClientSession::EmitAndWait emits MGLOG_F Fatal{BarrierViolation, "<slot>"}.
# This asserts an observed overlap with the applier, which is timing-dependent.
# Without that Fatal in a fresh selected file E1 fails; pixel/status fallbacks do not count.
# Keep the ID-53-approved SmallRing selection as well as the default lane.
run_control "negative control E1 (MOBILEGL_IPC_VERB_BARRIER=0)" \
'DirectGLES\.Split\.(SmallRing\.)?(Triangle|ClearThenReadPixels)' \
+14 -2
View File
@@ -7,7 +7,7 @@ trap 'rm -rf "${WORK}"' EXIT
cp "${HERE}/testdata/stub_ctest.sh" "${WORK}/ctest"
chmod +x "${WORK}/ctest"
passes=0
for mode in missing-fatal stdout-fatal stale-fatal evidence e3-unrelated; do
for mode in missing-fatal stdout-fatal stale-fatal evidence e3-unrelated skipped-selection notrun-selection missing-selection partial-fatal wrong-fatal; do
mkdir -p "${WORK}/${mode}"
rc=0
STUB_MODE="${mode}" CTEST="${WORK}/ctest" CONTROL_TMPDIR="${WORK}/${mode}" \
@@ -16,10 +16,22 @@ for mode in missing-fatal stdout-fatal stale-fatal evidence e3-unrelated; do
[ "${rc}" = 0 ] && grep -q "negative control E3(a).*scenario's own diagnostic" "${WORK}/${mode}.out" || {
cat "${WORK}/${mode}.out"; echo "NOT OK private-file Fatal must PASS"; exit 1;
}
grep -qFx "private-log evidence: DirectGLES.Split.ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels: ${WORK}/${mode}/entry.log" "${WORK}/${mode}.out" || {
cat "${WORK}/${mode}.out"; echo "NOT OK private-file evidence line must name the selected entry and path"; exit 1;
}
else
message='E1 FAILED: selected private logs lack expected Fatal'
[ "${mode}" != e3-unrelated ] || message='FAILED: red lacks its persistent-map push diagnostic'
[ "${rc}" != 0 ] && grep -q "${message}" "${WORK}/${mode}.out" || {
case "${mode}" in
skipped-selection) message='SplitLogPaths FAILED: E1 control: the knob killed the pre-flight, not the entry - 1 selected entries skipped' ;;
notrun-selection|missing-selection) message='SplitLogPaths FAILED: E1 control: 1 selected entries did not run' ;;
esac
if [[ "${mode}" = *-selection ]]; then
match=(-qFx)
else
match=(-qF)
fi
[ "${rc}" != 0 ] && grep "${match[@]}" "${message}" "${WORK}/${mode}.out" || {
cat "${WORK}/${mode}.out"; echo "NOT OK ${mode}: control must report FAILED for its own reason"; exit 1;
}
fi
+38 -4
View File
@@ -15,6 +15,11 @@
# stdout-fatal Fatal exists only on stdout, never in the private file
# stale-fatal Fatal exists before reset, never from this control run
# e3-unrelated E1 has its private Fatal; E3(a) fails for an unrelated reason
# skipped-selection knob kills pre-flight; private Fatal exists, entry skips, ctest exits 0
# notrun-selection selected entry was not run, despite a private Fatal
# missing-selection selected entry is absent from the result XML
# partial-fatal two entries fail, but only the first has its expected private Fatal
# wrong-fatal entry fails with a different Fatal in its private file
# green baseline green; the control's own run PASSES (the knob is not load-bearing)
# red-baseline the baseline itself has a failed entry
# all-skipped the baseline is entirely skipped (the disarmed lane, a legitimate exit 0)
@@ -44,10 +49,31 @@ emit_listing() {
return
fi
echo " Test #1: DirectGLES.Split.ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels"
echo "Total Tests: 1"
if [ "${mode}" = partial-fatal ]; then
echo " Test #2: DirectGLES.Split.TriangleScenario.SecondEntry"
echo "Total Tests: 2"
else
echo "Total Tests: 1"
fi
}
write_junit() {
if [ "${MOBILEGL_IPC_VERB_BARRIER:-1}" = 0 ] || [ "${MOBILEGL_IPC_PERSISTENT_BLOCK_KB:-64}" = 0 ]; then
entry=DirectGLES.Split.ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels
[ "${MOBILEGL_IPC_PERSISTENT_BLOCK_KB:-64}" != 0 ] || entry=DirectGLES.Split.PersistentCoherentMapScenario.TwoWritesThroughTheCoherentPointerEachReachTheirOwnDraw
body="<testcase name=\"${entry}\" status=\"fail\"><failure message=\"control red\"/></testcase>"
if [ "${mode}" = partial-fatal ] && [ "${MOBILEGL_IPC_VERB_BARRIER:-1}" = 0 ]; then
body="${body}<testcase name=\"DirectGLES.Split.TriangleScenario.SecondEntry\" status=\"fail\"><failure/></testcase>"
fi
case "${mode}" in
skipped-selection) body="<testcase name=\"${entry}\" status=\"notrun\"><skipped/></testcase>" ;;
notrun-selection) body="<testcase name=\"${entry}\" status=\"notrun\"/>" ;;
missing-selection) body='' ;;
green) body="<testcase name=\"${entry}\" status=\"run\"/>" ;;
esac
printf '%s\n' "<testsuite>${body}</testsuite>" > "$1"
return
fi
case "${mode}" in
red-baseline)
body='<testcase name="DirectGLES.Split.TriangleScenario.AVboBackedTriangleReachesReadPixels" status="failed"><failure message="already red"/></testcase>'
@@ -65,7 +91,7 @@ write_junit() {
# Model the library file sink separately from ctest stdout (ID-53).
log="${CONTROL_TMPDIR}/entry.log"
if [ "${json_requested:-0}" = 1 ]; then
python3 -c 'import json, os; p=os.environ["CONTROL_TMPDIR"]; print(json.dumps({"tests": [{"name": "DirectGLES.Split."+n, "properties": [{"name": "LABELS", "value": ["integration-split"]}, {"name": "ENVIRONMENT", "value": ["MOBILEGL_LOG_FILE_PATH="+p+"/"+f]}]} for n,f in [("ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels", "entry.log"), ("PersistentCoherentMapScenario.TwoWritesThroughTheCoherentPointerEachReachTheirOwnDraw", "pmap.log")]]}))'
python3 -c 'import json, os; p=os.environ["CONTROL_TMPDIR"]; entries=[("ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels", "entry.log"), ("PersistentCoherentMapScenario.TwoWritesThroughTheCoherentPointerEachReachTheirOwnDraw", "pmap.log")]; entries += [("TriangleScenario.SecondEntry", "second.log")] if os.environ["STUB_MODE"] == "partial-fatal" else []; print(json.dumps({"tests": [{"name": "DirectGLES.Split."+n, "properties": [{"name": "LABELS", "value": ["integration-split"]}, {"name": "ENVIRONMENT", "value": ["MOBILEGL_LOG_FILE_PATH="+p+"/"+f]}]} for n,f in entries]}))'
if [ "${mode}" = stale-fatal ]; then
echo 'Fatal{BarrierViolation, "DrawVbo"}' > "${log}"
fi
@@ -79,22 +105,30 @@ fi
if [ -n "${junit}" ]; then
write_junit "${junit}"
if [ "${MOBILEGL_IPC_VERB_BARRIER:-1}" != 0 ] && [ "${MOBILEGL_IPC_PERSISTENT_BLOCK_KB:-64}" != 0 ]; then
case "${mode}" in
red-baseline) echo "1/1 Test #1: ... ***Failed"; exit 8 ;;
*) echo "100% tests passed, 0 tests failed out of 1"; exit 0 ;;
esac
fi
fi
# The control's own run.
if [ "${MOBILEGL_IPC_VERB_BARRIER:-1}" = 0 ]; then
case "${mode}" in
evidence|e3-unrelated) echo 'Fatal{BarrierViolation, "DrawVbo"}' > "${log}" ;;
evidence|e3-unrelated|skipped-selection|notrun-selection|missing-selection|partial-fatal) echo 'Fatal{BarrierViolation, "DrawVbo"}' > "${log}" ;;
wrong-fatal) echo 'Fatal{ReplyMissing, "DrawVbo"}' > "${log}" ;;
missing-fatal) echo "library setup only; no fatal" > "${log}" ;;
stdout-fatal) echo 'Fatal{BarrierViolation, "DrawVbo"}' ;;
esac
fi
case "${mode}" in
unrelated|missing-fatal|stdout-fatal|stale-fatal|e3-unrelated)
skipped-selection|notrun-selection|missing-selection)
echo '1/1 Test #1: selected entry ... ***Skipped'
echo '100% tests passed, 0 tests failed out of 1'
exit 0
;;
unrelated|missing-fatal|stdout-fatal|stale-fatal|e3-unrelated|partial-fatal|wrong-fatal)
echo "1/1 Test #1: DirectGLES.Split.ClearThenReadPixelsScenario.ClearWithNoDrawIsVisibleToDefaultFramebufferReadPixels ...***Failed"
echo "UNRELATED_CONTROL_FAILURE: the harness aborted in setup before the knob was read"
if [ "${MOBILEGL_IPC_PERSISTENT_BLOCK_KB:-64}" = 0 ]; then