Commit Graph
21 Commits
Author SHA1 Message Date
swung0x48 590d74525d [Feat] (MG_Remote, P11 B2): T0 behind MOBILEGL_IPC_ADOPT_TIER - control revision 5 (LinkTerms.adoptTier ask), kCapAdoptT0 granted at the first native bind after the backend's POST, the AHardwareBuffer Offer on the aux socket matched to its map_persistent record by seq, named T2 fallbacks instead of A1's handshake Fatal, the server allow switch, done-no-bytes CPU reads 2026-09-29 17:40:56 -04:00
swung0x48 bcff9b00cc [Feat] (MG_Remote): fence polls are answered from the server's reports - the apply thread posts kEventFenceSignaled as the GPU finishes each client fence (and submits and re-checks while idle), glClientWaitSync(0) / glGetSynciv answer locally, and only every MOBILEGL_IPC_POLL_ESCALATE-th unanswered poll crosses 2026-09-29 03:59:44 -04:00
swung0x48 362daafeed [Fix] (MG_Remote): the pack-buffer readbacks check their whole shape before the read and refuse it at one latch site per sink, each with a raw-peer control, and a verb counts as statically barriered only when every op that stamps it waits 2026-09-29 02:21:36 -04:00
swung0x48 6485ae1dc0 [Feat] (MG_Remote): a read into a pixel pack buffer crosses as read_pixels_to_buffer / get_texture_image_to_buffer and waits for nothing - the server lands the rows through the buffer-write gate on both backends, the client marks the buffer GPU-written, and a device lost with the write still owed ends in Fatal{ReadbackDeclined, "buffer-writeback"} 2026-09-29 02:02:29 -04:00
BZLZHH b1ef484116 [Perf] (MG_Remote): the residency budget does not make the window usable either - the answer is gone before the client next looks, so a deferral needs RETENTION, not a budget
Steps 1 and 2 of the plan the design review set out, landed and measured on the
device. Step 1 answers the question step 2 depended on; step 2 answers the one the
whole line depended on, and the answer is that the window's shape is wrong.

STEP 1 - CAN AN ANSWER BE LOST IN DELIVERY? NO. ReplyPool now counts what the link
took (PostedReplies) and what it refused (FailedReplies), and a non-OK result is
NAMED instead of swallowed: it checked only BUFFER_TOO_SMALL and let
TRANSPORT_CLOSED vanish with no line and no counter, so "the server never sent it"
and "it was sent and the buffer overwrote it" were the same symptom with opposite
fixes. Measured: posted=4501 failed=0. The twenty-answer gap between the server's
tally and the client's was a sampling difference, not a drop - and the server's
m_answered counts ATTEMPTS (incremented before PostReply), which is why it could
never have settled it.

STEP 2 - DOES A RESIDENCY BUDGET FIX THE WINDOW? NO.
Wire_ResourceCreate now measures how many reply-owning records the client has
emitted since the oldest window entry was pushed (ClientSession::ReplyPostings, a
new clock) and takes the oldest answer back as soon as that approaches the reply
buffer's depth, instead of deferring past the point where the answer can still
exist. On the device:

    P12win call=4501 cfg=2 eff=2 pending=2 taken=63 pushed=65 blocked=4435
           budget=4451 readfail=8872 posts=54377
    kind=frame-op frame=14 wait_replies=9078 by_op=2=9071   wall 27.87 s

budget=4451: it fires on essentially EVERY create, the window stops deferring
(pushed=65), and the drain still reads nothing back. THE REASON IS STRUCTURAL, not
a constant:

  * the window only gets control WHEN A CREATE ARRIVES. Between two creates this
    client emits on the order of a thousand records - at the FIRST failed read the
    reply ring already spanned [1175..1504] while the head was seq 310 - and nothing
    looks at the window in that gap;
  * what displaces a deferred answer is not what this client emits, it is WHAT THE
    SERVER ANSWERS. The server is the one behind (one answer per ~5 ms wakeup while
    the client runs ahead), so the replies arriving during the gap are its LATE
    ANSWERS TO THE EARLIER CREATES - they fill the 8 slots first;
  * so the check fires too late by construction. Tightening it does not fix that; it
    only removes the deferral entirely, which is what pushed=65 is.

WHAT THAT LEAVES. A deferral is readable only if the client KEEPS the answers it
deferred. The buffer is bounded by COUNT and the client cannot bound how many
answers the server will post between two of its creates, so hoping the buffer still
holds one is a race the client loses - which is also why the earlier two diagnoses
(reply-pool churn, then FIFO-with-no-way-out) were both describing symptoms of the
same thing. The shape that can work is retention: the reader keeps the answers for
the seqs still outstanding (a handful), drops them when read, and reports a LOSS
when a read asks for one it no longer has. That is a transport change - StreamLink's
ring becomes a retention map with a cap and a lost counter - and it is the next step.

BOTH STEPS STAY: step 1 is a net gain on its own, and step 2 is the safe direction
(it can only degrade the window to no-deferral, i.e. the pre-window shape) and its
clock is the one the retention design needs. MOBILEGL_IPC_CREATE_WINDOW stays at 1.

172/172 and 13/13 after the probes were removed. The measurements and the revised
conclusion are in notes/p12/CREATE-WINDOW-MEASURED.md section 9.
2026-09-26 02:35:05 +08:00
swung0x48 2bcec27ac2 [MG_Remote, MG_Test] (P7): bound readback scratch by server reply limit 2026-09-23 13:43:44 -04:00
swung0x48 018fab0f5b [WIP] (MG_Remote): implement P6.5 TCP and stream transport 2026-09-22 07:19:18 -04:00
swung0x48 70df57bb23 [MG_Remote] (Disaggregated): carry native query results and complete XFB object deletion 2026-09-20 11:52:13 -04:00
swung0x48 8df7861f07 [MG_Remote, MG_Test] (Disaggregated): P5e ID-111 - the barriered stamp asks the server's own capability
- ApplyOne stamped `kMGPipeP5eClientWaitRuleLanded ? wireSaysBarriered : true`, which reads a
  build constant and never asks which server this is. kCapRunAheadApply is never published on
  DirectVulkan (ID-90), yet draw_vbo / blit / clear / launch_grid are all kWaitNone - so the day
  the constant flips, a Magma server would stamp every draw record UNBARRIERED while its client
  is still lockstep. CountBarrierPull is an unconditional Fatal on an unbarriered pull, no knob
  involved, so Magma dies on its first draw and MagmaP7AllocatorDebtScope's exemption goes with it.
- The stamp is now `MGPipeApplierStampsBarriered(waitRuleLanded, serverPublishesRunAhead, wireSaysBarriered)`,
  a pure function beside the applier, so ID-90 is something a unit case can hold AT THE POST-FLIP
  value of the constant rather than a claim that first becomes testable on integration day. That
  is MGPipeRunAheadCapBitsFor's shape, for MGPipeRunAheadCapBitsFor's reason.
- Both predicates stay computed unconditionally, as arguments rather than short-circuit arms:
  ID-103's reason, extended to the capability probe.
- `ServerPublishesRunAhead` is promoted out of the anonymous namespace as
  `MGPipeServerPublishesRunAhead` (body unchanged) so the red-once can state "this session lacks
  bit 10" in its own words instead of inferring it from the stamp under test (ID-102).
- Red-once: ADrawOnASessionWithoutTheRunAheadBitStaysStampedBarriered drives a real DrawVbo
  through ApplyOne on a SetCapabilityBits(0) session. Deleting the new conjunct makes the
  post-flip assertion false - "RemoteClientControls.inc:143: Failure / Expected: true / exited 124".
2026-09-18 11:43:17 -04:00
swung0x48 a7e55ee62a [MG_Pipe, MG_Remote, MG_Backend, MG_Impl, MG_IntegrationTest] (Disaggregated): P5c ct - the applier_reset and object_death control records
Two appended wire records replace the two cross-role control bypasses:
applier_reset (opcode 77, MGPApplierReset{Uint64 ContextSerial}, kScreen) is emitted by
the GL thread at the FreshlyPrimed make-current edge, before the client-side resets, so
the record's barrier orders the server's MGPipeApplierReset against every verb that
follows; a direct MGPipeApplierReset() from a non-apply thread with a live session is
Fatal{RoleViolation, "g_applier"} (MGPipeApply.cpp carries the layer-2 guard; the
apply-thread and no-session bring-up shapes keep the direct call verbatim).
object_death (opcode 78, reuses MGPipeHandleOnly, kCtxObject) is the framebuffer
family's first wire delete opcode: the client thread resolves the dying object's handle
in its OWN allocator, emits nothing for a handle the server never saw, and EmitAndWaits
otherwise - the wait preserving the death's ordering against in-flight records that
name the handle, the only property the blocking mailbox hop provided. The sink releases
per-kind twins through a handle-keyed ReleaseTwinByHandle (seven kinds, SamplerViewCso's
idempotent second path kept); a null arriving handle is
Fatal{ProtocolCorruption, "ObjectDeath.Handle"}. A documented NoSession fallback arm
delivers a death to a server-only fixture's loop; monolith keeps the mailbox hop and
the FreshlyPrimed direct call character for character.

Evidence: gen_pipe --check/--self-test and gen_pipe_field_ownership --check/--self-test
green; unit 2155/2155; integration-split 111/111 (107 plus four new CtWireScenario
entries, all genuinely run, each with its own log path and an EXPECT_EXIT death case);
red-once - shorting the emission to the direct call turns the four Ct cases red with
Fatal{RoleViolation, "g_applier"} by name, then restores green verbatim.
CONTRACT-P5C sections 1, 5.1-5.2, 7.
2026-09-17 06:08:54 -04:00
swung0x48 f018c0ef05 [Feat] (MG_Remote, P5b sync): migrate fence lifecycle and wait replies to the apply thread 2026-09-16 13:46:49 -04:00
swung0x48 9606466ab6 [Merge] (MG_Remote, P5b): combine d1 draw migration with f1 i1 t2 and r2 2026-09-16 13:30:01 -04:00
swung0x48 5443bced9a [Merge] (P5b): integrate t2 and reconcile all migrated emitter families 2026-09-16 13:27:15 -04:00
swung0x48 dd293203ca [Feat] (MG_Remote, P5b/t2): emit the four stream-output span rows, the XFB object bind and the patch parameter, and give each ServerVerbSink body its backend call 2026-09-16 12:23:57 -04:00
swung0x48 e564939615 [Feat] (MG_Remote, P5b d1): flip the nineteen indexed / instanced / multi-draw / indirect draw slots to class B on draw_vbo - the emitters plan the record from the GL call and the bound element / indirect / parameter buffers, stage a client index array as the kDrawHasUserIndices span, refuse +CLIENT_INDICES / +CLIENT_COMMANDS / +INDEX_OFFSET by name, and the sink dispatches every shape to the backend slot the monolith calls (a base vertex of 0 is DrawElements, never DrawElementsBaseVertex with 0: ANGLE has no glDrawElementsBaseVertex) 2026-09-16 12:10:38 -04:00
swung0x48 734287ac56 [Feat] (MG_Remote, P5b): emit i1 seven image/compute/barrier/copy-image/storage-block slots as class B and give each ServerVerbSink body the backend call the contract names 2026-09-16 12:09:55 -04:00
swung0x48 f567610339 [Merge] (P5b, c0b): merge feat/disaggregated at f5614995 (v1 round 3, ID-62/ID-67) into p5b/c0b 2026-09-16 11:16:09 -04:00
swung0x48 a164113029 [Feat] (MG_Remote, P5b): decode every P5b row to a WireVerbSink method, stub each ServerVerbSink body with a named UnmigratedVerb Fatal, and partition the emit table class-C list per migration package 2026-09-16 11:15:45 -04:00
swung0x48 49fecc1686 [Fix] (DirectGLES/Managers, MG_Backend/Init, MG_Remote/Server): N-7 m-5's freed-base null is transport-guarded like the other two split hunks so monolith transport in a split build changes nothing by construction; N-8 the second consumer-mask check compares MGPipeGetResourceOps() against the table step 2 saw and refuses a replaced table as Fatal{ConsumerMaskLie, resource ops table replaced} (the re-run alone could not see a client-registered table); ServerVerbSink::ReadbackScratchBytes for ID-49's tight-size gate 2026-09-16 10:17:29 -04:00
swung0x48 8dbe230ae3 [Feat] (MG_Remote/Server): PipeApplier - attach the decoder on the apply thread, stamp/apply/clear per record, and ServerVerbSink for the five class-B verbs 2026-09-16 05:45:15 -04:00
swung0x48 a45c0a0e18 [Feat] (MG_Remote): land the P5 contract headers and their named-Fatal stubs so all seven parallel packages compile and link on day one 2026-09-11 13:05:05 -04:00