[MG_IntegrationTest, docs] (P7): the whole integration binary runs under Magma split, and it is one red

Tier 2 replays the split-arm registrations - ~100 scenarios, the set Espryt's disaggregation work
has carried since P5 - and it measured green on Magma across all three transports. That is a real
result but it is NOT exit gate 1. The gate says the INTEGRATION BINARY is green under Magma split,
and the set that names is the one the monolith DirectVulkan. arm runs: the whole suite minus
MGL_ITEST_MONOLITH_MECHANISM_FILTER. 510 cases against tier 2's 102, and the readback matrices,
the same-texture copy shapes and the odd-offset uniform ranges are in it and nowhere else in the
disaggregated registry.

integration-magma-full-{split,spawn,tcp}, in no gating step, private log per entry through a list
of its own. THE RESULT: 453/0/57 on inproc, 453/0/57 on spawn, 454/1/57 on tcp - and not one of
the 1530 entries reaches any of the tree's fifteen @P7 named refusals. The single red is
IterationRPProgram203Scenario.FixedCompleteInputProducesFixedCompleteGoldenOutput on the TCP arm:
one texel of 262656, texel 0, which is the case's own exposure value
(IterationRPProgram203Scenario.cpp:371, actual half bits (0x357a, 0x4b65) against golden
(0x3a74, 0x4821)). Deterministic 5/5 on tcp; 5/5 GREEN on inproc, on spawn and on the monolith
control. spawn and tcp differ only in MOBILEGL_IPC_DATA, so the suspicion is the STREAM data path
and not "two processes"; the GLES tcp lane is green in the same run, so it is not the fixture.
Not fixed here - it is what the census is for.

TWO ENVIRONMENT DIFFERENCES FROM THE GATED TIERS, both measured rather than assumed:

  * NO MGITEST_SPLIT_LANE=1. That marker arms ScenarioFixture's destructor assertion - a case in
    an armed split lane must have moved the client encoder's record ordinal - which is right for
    a curated lane and wrong for a whole-binary census. With it set, the first run produced 13
    reds, all `EmitSeq was 20 at SetUp and is 20 now`: AdvertisedLimitsScenario (8),
    SpirvShaderBinaryScenario (4) and ProgramPipelineScenario.AReservedName - a limits query is
    answered from the client's CapsMirror and a shader-binary error surface never reaches a verb.
    Those 13 would red on a DirectGLES census too. What it gives up is stated in the comment: this
    tier does not prove its records crossed, and the tcp topology proof is inert in it.
  * DualBlockScenario.* filtered out. Its green REQUIRES MOBILEGL_IPC_ROLE_SPLIT_STATE=1 and a
    census must not set that knob; without it the case is red by name, which is its documented
    negative-control behaviour. The monolith arm skips it for the same class of reason.

CtWireScenario gets MOBILEGL_TEST_SERVER_COUNTERS=1 in this tier - a switch that makes a counter
readable, not an oracle that asserts - because without it the spawn arm's ApplierReset case reds
on "server control-record telemetry missing", which says nothing about Magma. The strict-arming,
pmap and persistent-map-arm markers stay out.

The inproc-only exclusion in spawn_lane_parity.py is now PER TIER: in the gated tier the two
scenarios are registered on the inproc arm alone, in the census they are on every arm and simply
skip, so excluding them there would manufacture a spurious "extra".

Measured after the change: every pre-existing lane unchanged and green (magma-split 73,
spawn 52, tcp 54, all-* 102/102/104, GLES 186/102/104, buffers 19, runahead 14, caches 7), the
dual-block census 257 green / 2 skipped / 0 red with an empty ratchet, both new CI steps green,
parity RC=0 across all three tiers, G14 4155 -> 6095 with removed=0.

The note now carries the grouped red list, the 57 named skips by reason, the four pieces of
evidence that the tcp red is a defect and not plumbing, and six findings - the first being that
wave 2's clusters cannot be scheduled from this lane's red list, because on lavapipe there is
essentially none.
This commit is contained in:
2026-09-22 14:24:19 -04:00
parent 78e10ccf80
commit 1de8f02ebe
4 changed files with 348 additions and 164 deletions
@@ -29,6 +29,29 @@ foreach(listName IN ITEMS @MGL_SPLIT_ARM_TEST_LISTS@)
endif()
endforeach()
endforeach()
# P7 package L tier 3: the whole-suite Magma census. A LIST OF ITS OWN, and the log path is the
# ONLY property it gets - the loop above also arms the strict-arming, pmap and server-counter
# lane markers by entry name, and a census is here to be READ, not to assert three unrelated
# oracles it never opted into. The path itself is not optional: these entries are
# DirectVulkan.{Split,Spawn,Tcp}. names, so split_log_paths.py's ownership gate requires exactly
# one absolute private path each, and on the spawn and tcp arms two processes append to it.
#
# ONE EXCEPTION, AND IT IS TELEMETRY RATHER THAN AN ORACLE: CtWireScenario reads the server's
# own control-record tallies, and without MOBILEGL_TEST_SERVER_COUNTERS=1 they are not published
# at all - the case then reds on `server control-record telemetry missing`, which says nothing
# about Magma. Measured: that is exactly what the spawn arm's ApplierReset case did on this
# tier's first run. A switch that makes a counter readable is not the same kind of thing as the
# strict-arming, pmap or persistent-map-arm markers above, which make a case ASSERT something it
# never opted into, and those stay out of the census.
foreach(listName IN ITEMS @MGL_MAGMA_FULL_TEST_LISTS@)
foreach(entry IN LISTS ${listName})
set_tests_properties("${entry}" PROPERTIES ENVIRONMENT
"MOBILEGL_LOG_FILE_PATH=@CMAKE_CURRENT_BINARY_DIR@/split-logs/${entry}.log")
if(entry MATCHES "CtWireScenario\\.")
set_tests_properties("${entry}" PROPERTIES ENVIRONMENT "MOBILEGL_TEST_SERVER_COUNTERS=1")
endif()
endforeach()
endforeach()
foreach(entry IN LISTS MGL_SPLIT_PA_STRICT_DRAW_TESTS MGL_SPLIT_PA_STRICT_DISPATCH_TESTS)
set_tests_properties("${entry}" PROPERTIES ENVIRONMENT
"MOBILEGL_LOG_FILE_PATH=@CMAKE_CURRENT_BINARY_DIR@/split-logs/${entry}.log")