swung0x48
aba329cc53
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.
4.5 KiB
Read-only audit brief (Kimi): every apply-thread read of client-owned memory on the DirectGLES inproc draw path
Repository: the current directory (MobileGL, branch feat/disaggregated, head 2fde7034). READ-ONLY:
do not edit any file, do not build, do not run tests, do not touch adb or WSL. Use your file reading
and grep tools only. Write the report file named at the end BEFORE your final message.
Context
MobileGL's split (MOBILEGL_TRANSPORT=inproc) runs the GL frontend on the application's GL thread
("client role") and the backend (MobileGL/MG_Backend/DirectGLES) on an apply thread ("server role",
MobileGL/MG_Remote/Server/ServerLoop.cpp, records applied by MobileGL/MG_Remote/Server/PipeApplier.cpp).
Today every draw/clear/blit/dispatch waits for the apply thread (MobileGL/MG_Remote/Client/ClientSession.cpp,
EmitAndWaitTails, the R-1 barrier). The next phase (P5e) wants the client to publish those records
and continue WITHOUT waiting. That is only safe if, while applying a record, the apply thread reads
nothing that the GL thread may be mutating at the same time.
Client-owned memory = anything reachable from MG_State::pGLContext / the GLContext object (binding
tables, TextureUnit arrays, BindingSlots, VertexArrayObject attributes, ProgramObject internals,
FramebufferObject attachments, BufferObject shadow bytes), the frontend objects those slots hold
(ITextureObject, BufferObject, SamplerObject, ...), the client slot allocator MGPipeSlots()
(MobileGL/MG_Impl/Pipe/SlotAllocator.*, probed via HandleOf/FindByLifetimeId/GetOrCreate(const StatePtr&)
in MobileGL/MG_Backend/DirectGLES/SlotTables.h and HandleOfBuffer in Managers.cpp), and the
BARRIER_PULLED rows of gPipeInputs (MobileGL/MG_Pipe/FieldOwnership.def, storage
MobileGL/MG_Backend/MGPipe/PipeInputs.h, read through MGB_CTX->...). Server-owned memory =
MG_Pipe::MGPipeApplier() state, the twin objects in the DirectGLES registries, the server staged
stores (ServerStaged(), ServerStagedTexture()), records and SEG_STAGE bytes.
Task
Starting from the record sinks in MobileGL/MG_Remote/Server/PipeApplier.cpp (ServerVerbSink::OnDrawVbo,
OnClear, OnBlit..., OnDispatch..., OnGenerateMipmap, OnCopyFramebufferToTexture, and every
MGPipeApply* record apply in MobileGL/MG_Pipe/PipeApply.cpp), follow the calls into
MobileGL/MG_Backend/DirectGLES/DirectGLES.cpp and Managers.cpp (PrepareForDraw, PrepareForCompute,
SyncNeccessaryBuffers, SyncVaoAttributeBuffersByHandle, SyncNeccessaryTextures, SyncCurrentProgram,
BindCurrentProgramWithResources, BindCurrentTextures, SyncCurrentFBO, BindCurrentFBO,
SyncImageTextureBindingsForDraw, SyncBufferBindingPoints, Clear, BlitFramebuffer, GenerateMipmap,
EnsureBufferResource*, IsBufferDrawClean*, and everything they call) and list EVERY site that reads
client-owned memory, with the arm it is on when the transport is active (some sites have a record
arm and a legacy arm; say which runs under MG_Config::Transport != Monolith with all
MOBILEGL_PIPE_PUSH subsystem bits set).
For each site: file:line, the function, what is read (which client object/field/table), the kind
(live dereference / lifetime-id + allocator probe / allocator mint / gPipeInputs BARRIER_PULLED row),
whether it runs on every draw or only on a memo miss / rare path, and which pushed record or applier
state field already carries the same information (look in MobileGL/MG_Pipe/PipeApply.h
MGPipeApplierState, MobileGL/MG_Pipe/MGPipeTypes.h, MobileGL/MG_Pipe/PipeCalls.def) or "none".
Also list: (a) every MGPipeFrontendKeyedRegistryScope / MGPipeReverseAnnouncementScope construction
site and what it wraps; (b) every place the apply thread holds a SharedPtr to a frontend object
across records (memos, sync lists); (c) every MGB_CTX-> accessor used from MG_Backend/DirectGLES
whose row in FieldOwnership.def is BARRIER_PULLED.
Output: a markdown table per family (VAO/vertex, buffers, textures/samplers, framebuffers/attachments,
images, programs/shader state, binding-point tables, XFB, misc), then the three lists, then a short
"what would have to change for the apply thread to read only server-owned memory" per family
(one paragraph each, no code). No process narration. <= 400 lines. Every file:line must resolve
at head 2fde7034.
Write the report to: C:/Users/geekerwan/AppData/Local/Temp/claude/C--Users-geekerwan-AndroidStudioProjects-FoldCraftLauncher-MobileGL-disagg/5007453d-cb19-4ebf-96a3-60ba4fb55b47/scratchpad/p5e/kimi-audit.md