swung0x48/MobileGL
1
0
Fork 0
mirror of https://github.com/MobileGL-Dev/MobileGL synced 2026-11-07 18:13:23 +09:00
Code Issues Packages Projects Releases Wiki Activity
Files
4fe2edb6fa3fac018089f980c514b61793ef5277
MobileGL/docs/Disaggregated/notes/p5b/BRIEF-P5B.md
T
swung0x48 aba329cc53 [Docs] (Disaggregated): P6's plan and contract draft; the working notes move from a local directory into the repository 
Two things land together: the plan and contract draft P6 needs before it starts, and the
working notes that until now existed on exactly one machine.

1. P6's plan and contract draft (neither started; not one line of code)

     docs/Disaggregated/P6-SPAWN-PLAN.md      the package plan
     docs/Disaggregated/P6-CONTRACT-DRAFT.md  the contract draft (moves to
                                              MG_Remote/CONTRACT-P6.md when c6 lands)

   The three findings that mattered, all written into the documents:

   a. P6 is SMALLER than its roadmap cell suggests. The transport primitives are already in
      the tree and already tested: SocketDoorbell including hangup detection and the two
      lost-wakeup fences (Doorbell.cpp:211-360, four cases in FdPassingTest), SCM_RIGHTS fd
      passing, ShmSegment::Adopt, and the SurfaceOp / WindowKind schema already defined in
      protocol.fbs. inproc uses a real second shm mapping rather than an alias, and
      SessionRings.h:22-28 says that was done so today is not the first time it runs. P6 is
      assembly plus a control plane plus a process, not "write a socket transport".

   b. The evidence for "P6 is only a transport swap" has EXPIRED. It was bought by P5c, whose
      audit found 59 direct memory accesses in a tree that already looked clean. P5e changed
      the tree again. So the first package, a6, is a read-only audit that writes no code, and
      every contract row that depends on it is marked PENDING a6 rather than decided in
      advance.

   c. P6 does not get a game running on the phone. ANativeWindow* is meaningless in the child,
      and protocol.fbs:186-196's own comment says Android transfers the window out of band -
      that band is P12. P6 lands pbuffer / surfaceless / offscreen, which is what HeadlessGL,
      the split-equivalent lane and trace replay need, and which is exactly what P6's own exit
      gate measures. A real window reaching the server is a named refusal. This gets its own
      section because "P6 is done" would otherwise read as "the game runs now".

   The contract adds rule G: a role may not name a process-local handle. It extends P5c's
   rule E (a role may not name the other role's memory) from memory to identity - EGLDisplay /
   EGLSurface / EGLContext / ANativeWindow* / fd number / pid / mapped address either cross as
   a token this contract defines or they do not cross, and a descriptor crosses only through
   SCM_RIGHTS, where the kernel renumbers it.

   What P5e left for P6 gets its own section. The latch semantics are already designed
   (ARCHITECTURE.md:504) and are copied unchanged. What is new is that after run-ahead the
   client no longer waits while the peer process can DIE: the client holds records that were
   published and will never be applied, and Android's freezer suspends the child without
   closing anything. So slow and dead must be distinguishable WITHOUT a timeout threshold.
   That is P6's only new design; everything else is carriage. It shares a predicate with E1's
   outstanding re-specification (ID-122), and the two should be designed together.

   Two gates carried over from P5e's lessons: G2 requires integration-spawn's test-name set to
   match integration-split's name for name, because a differing set means the lane is quietly
   running less; and the arm-proof gate requires every entry to record the child pid and
   transport=spawn in its OWN private log - ID-124 established that this kind of false green
   really happens, and a spawn lane that actually ran monolith would pass every other gate.
   S2's shape must be settled before it is written, and the document says why: E1 is the
   cautionary case.

2. The working notes move into the repository (docs/Disaggregated/notes/, 415 .md files)

   ROADMAP / CURRENT_STAGE_PROGRESS / MEASUREMENTS cite ruling numbers, audit line numbers and
   the provenance of gate figures from ~/w7/notes/, a directory that exists on one machine and
   is under no version control. A record cited hundreds of times that lives only locally is not
   a record, so it moves in whole, with its directory structure intact.

   Only .md moves. Scripts, run logs, census and gate result JSON, patches, APKs and nm dumps
   stay where they are: they are evidence DATA rather than documents - large, regenerated per
   run, and mostly carrying absolute paths and device serials that mean something only on the
   machine that produced them. The mapping and this split are written in
   docs/Disaggregated/notes/README.md.

   Two exceptions are recorded there too: P6's BRIEF has been promoted to a first-class
   document and keeps no copy under notes/, so the two cannot drift apart; and recovered/ holds
   material salvaged from failed workflow runs, which may duplicate the like-named phase
   directories and is kept because a few of those were the only complete version at the time.

   CI's citation lint takes docs/Disaggregated/*.md only (non-recursive), and notes/ is
   deliberately outside it: those notes' file:line citations were written against their own
   code heads, and resolving them against today's HEAD would fail wholesale - which is not an
   error, it is their correct state as a historical record. To check one, use the revision it
   declares (--rev).

3. Inconsistencies fixed in passing

   - README.md's status line still said P5e was in progress while the stage table already said
     closed.
   - CURRENT_STAGE_PROGRESS's next-step item 1 still listed the exit gate as pending, though
     (a) and (b) were run in the previous commit; rewritten to closed plus the one item that
     actually remains (E1).
   - ROADMAP's P6 row still had the pre-phase plan in three of its cells; it now carries the
     package order, the scope boundary and the two new gates.

Checked with check_doc_citations.py: 125 citations across docs/Disaggregated/*.md, 0 problems.
2026-09-18 23:31:39 -04:00

7.5 KiB
Raw Blame History

BRIEF-P5B — verb migration under inproc, Minecraft-first

Goal, one sentence: fully separate-thread rendering of REAL workloads as fast as possible — every Minecraft trace, and the four A/B trace cases, rendered by the apply thread under MOBILEGL_TRANSPORT=inproc — by migrating the class-C slots the census measured, in the order it measured them, with the inproc lane's abort count and the traces' first-blocker list as the only gates (ID-66, ID-68).

Read PACKAGE-PREAMBLE.md first, then MobileGL/MG_Remote/CONTRACT-P5B.md (the wire contract, every row of it), then this file.

1. What c0b landed (head p5b/c0b, see p5b-results/c0b-v1.md)

Every package compiles and links on day one against real headers:

  • the wire rows: draw_vbo grew kDrawIsIndirect + the MGPDrawIndirect second tail; five rows appended at opcodes 72..76 (bind_shader_image, patch_parameter, bind_stream_output, set_storage_block_binding, copy_framebuffer_to_texture); MGPCopyRegion grew its two GL names; the clear discriminants have one spelling in MGPipeTypes.h;
  • the codec: encode/decode arms for every row, tail cross-checks, the name blob;
  • WireVerbSink has one method per row (13 new + the widened OnDrawVbo), and ServerVerbSink overrides every one with a Fatal{UnmigratedVerb, "<GL slot>"} stub;
  • the client emit table's class-C list is PARTITIONED per package (MGR_UNMIGRATED_{D1,I1,T2,F1,TAIL}_SLOTS, kEmittedSlots{D1,I1,T2,F1}, ownership static_asserts 19 / 7 / 7 / 11 / 20) so four packages edit four disjoint regions;
  • the stamp rows (FieldOwnership.def), the generated tables, the pinned counts, and a round-trip unit case per row in PipeWireCodecTest.cpp.

2. The four packages

Each package owns its slots, its list and count in EmitTables.cpp, its stub bodies in PipeApplier.cpp, and its cases in PipeWireCodecTest.cpp + the integration lane. CONTRACT-P5B §2 gives every slot's record fields, sink dispatch and refusal names; §8 the two pre-granted edits to c0b's files.

package measured slots (census entries) companions (same rows, owned too) rows model
d1 — indexed / instanced / multi-draw / indirect DrawElements 56 (+ every Minecraft trace), MultiDrawElementsBaseVertex 18, DrawArraysInstancedBaseInstance 4, DrawElementsBaseVertex 4, MultiDrawArrays 2, MultiDrawElements 2, DrawArraysInstanced 2, MultiDrawElementsIndirect 2, MultiDrawArraysIndirectCount 2, DrawElementsInstancedBaseVertex (trace) DrawElementsInstanced, DrawElementsInstancedBaseInstance, DrawElementsInstancedBaseVertexBaseInstance, DrawRangeElements, DrawRangeElementsBaseVertex, MultiDrawArraysIndirect, MultiDrawElementsIndirectCount, DrawArraysIndirect, DrawElementsIndirect draw_vbo (59) Claude, premium
i1 — image / compute / barrier / copy-image / storage block BindImageTexture 138, DispatchCompute 49, CopyImageSubData 30, MemoryBarrier 18, ShaderStorageBlockBinding 4 DispatchComputeIndirect, MemoryBarrierByRegion bind_shader_image (72), launch_grid (60), memory_barrier (61), resource_copy_region (53), set_storage_block_binding (75) Claude
t2 — XFB spans, XFB object bind, patch parameter BeginTransformFeedback 95, PatchParameteri 43, BindTransformFeedback 2 EndTransformFeedback, PauseTransformFeedback, ResumeTransformFeedback, (DeleteTransformFeedback: no row, see contract) begin/end/pause/resume_stream_output (62..65), bind_stream_output (74), patch_parameter (73) Claude
f1 — clears, framebuffer-sourced copies, mips ClearBufferiv 12, ClearBufferuiv 8, ClearBufferfv 6, ClearBufferfi 1, ClearNamedFramebufferfv 2, CopyTexImage2D 3, GenerateMipmap 2 ClearNamedFramebufferfi/iv/uiv, CopyTexSubImage2D clear (57), copy_framebuffer_to_texture (76), generate_mipmap (54) codex astra — additive and mechanical

Per slot the work is the same three edits: the emitter (class C → B in EmitTables.cpp, the pre-verb hook order kept), the ServerVerbSink body (the backend call the contract names), and a lane case that proves the record crossed. Nothing in the backends, unless the contract names the site (i1's mirror skip, f1's optional mip-storage re-derivation) - and then only behind #if MOBILEGL_BUILD_DISAGGREGATED.

3. The per-package gate

  • the inproc lane's abort count for your slots → 0, and each affected scenario reaches its NEXT first blocker, reported by name (Fatal{UnmigratedVerb, "…"}, Fatal{UnmigratedEmulation, "…"}, or a wrong-answer failure - whichever comes first). Run ~/w7/p5b-c0b-census.sh <your tree> <outdir> and diff results.json against c0b's baseline (~/w7/p5b-c0b-census-logs/results.json: 432 passed / 185 skipped / 505 aborted / 27 failed, all 505 aborts UnmigratedVerb, the per-slot table identical to census-classC.md's).
  • d1 additionally: every Minecraft trace reaches its next first blocker under inproc, reported by name per fixture (the census's trace table), via the same per-entry private-log method as p6_census_traces.{sh,py}; a trace that RENDERS is reported with its SSIM.
  • the integrator's quick gate: unit lanes green (c0b's baseline + yours), integration-split inproc 22/22, push integration-gpu 1128/1128 name-for-name, G1 0/0/0/0 .text +0.
  • no reviews between rounds; one codex review at P5b's close.

4. File ownership

file owner in P5b others
MG_Pipe/MGPipeTypes.h, PipeCalls.def, PipeFields.def, FieldOwnership.def, Coverage.def, FillPoints.def, scripts/gen_pipe*.py, MG_Remote/CONTRACT-P5B.md c0b / integrator t2: kCapBackendOwnsXfbCapture (contract §6.5); i1: the GetProgramForDispatch row (§6.9). Anything else: ask.
MG_Remote/Wire/PipeWireCodec.{h,cpp} w1's file; c0b landed the arms a package edits the layout arm of ITS row only, with the integrator
MG_Remote/Server/PipeApplier.{h,cpp} v1's file each package replaces ITS stub bodies (the // ---- d1/i1/t2/f1 ---- blocks) and may add ITS tallies
MG_Remote/Client/EmitTables.{h,cpp} c1's file each package edits ITS MGR_UNMIGRATED_*_SLOTS, ITS kEmittedSlots*, ITS emitters
MG_Remote/Client/SlotCaps.h c1's file t2: the EndTransformFeedback probe → kCapBackendOwnsXfbCapture
MG_Test/Wire/PipeWireCodecTest.cpp, MG_Test/Wire/RemoteClientTest.cpp shared append your cases below c0b's; the partition pins (UnmigratedSlotCount() etc.) move with your counts
MG_IntegrationTest/** t1's file each package adds ITS lane cases / scenario arms
MG_Backend/DirectGLES/**, MG_Backend/DirectVulkan/** untouched by c0b only the sites the contract names, only behind #if MOBILEGL_BUILD_DISAGGREGATED
~/w7/notes/p5b/p5b-results/<slug>-v1.md you

5. The loop that follows

  1. The four packages land, each on its own gate. The integrator merges in landing order, regenerates (gen_pipe.py, gen_pipe_field_ownership.py - never hand-merge generated/), runs the quick gate.
  2. Re-census on the merged head: the lane (p5b-c0b-census.sh) and the traces. The new first-blocker table is the next round's package list, same shape, same gates.
  3. Migrate the new first blockers (wave 3's tail is P9/P10-shaped: readbacks, queries, syncs) and the residual-input debts rsp points at where a verb's pull is what blocks it.
  4. Repeat until the four A/B trace cases render under inproc on the Redmi. That is P5b's exit; P6's spawn transport follows (ID-68).
Reference in New Issue View Git Blame Copy Permalink