Remove documentation that describes things that no longer exist (#2017)

Three kinds of stale documentation, all of it naming something the tree does not
contain. `doc-consistency.md` §3: when a referent is gone, the reference goes too.

**`problems.md`, deleted.** A personal working audit of `host_build_graph` —
findings, their status, and notes on what to look at next. It was never meant to be
here. It arrived in #1974 through a `git add -A` that swept it up with that
change's 958 files, and #2012 then updated it rather than noticing it. Nothing
references it (`git grep problems.md` is empty), it is absent from `mkdocs.yml`,
and `wheel.packages` is `["simpler_setup", "python/simpler"]` so it never shipped.
Its closed items name the PRs that closed them, each of which carries the
reasoning in its own commit message.

**`last_task_alive`, in hbg's `RUNTIME_LOGIC.md`.** The paragraph gives four
reasons `SchedulerState` holds no per-run content, and the third was "hbg never
advances `last_task_alive`". That was accurate when written — hbg inherited the
field from `tensormap_and_ringbuffer` and did not advance it — but #1837 deleted
the field with the rest of the reclamation paths, for exactly the reason the
sentence gives, and #2004 then removed the ring that held it. The clause now names
the one thing in that list that does not exist. It is replaced by the live fact
from the same runtime: polling reserves no wiring or dependency pool, because
readiness comes from the task table's `completion_flags`.

**`ring 0`, in hbg's `SCALAR_DATA_ACCESS.md`.** Ownership validation was described
as checking that "the task ID ... carries ring 0, the only ring HBG places tasks
on". hbg has had no ring since #2004 (`git grep 'rings\['` over its tree is
empty), and since #2012 the check is on the id *space*, not a ring index. The
bullet now says what the code does and why a `GRAPH_NODE` id fails it.

**`pto_runtime2_init`, in three `scheduler.h` files.** A comment credited it with
reserving and wiring the early-dispatch queues. `git grep pto_runtime2_init`
returned only those three comments — the symbol exists nowhere. The work is done by
`SchedulerState::init_data_from_layout`, which the comment now names. Only three of
the four runtime trees carried the line; `src/a5/runtime/tensormap_and_ringbuffer`
words that member differently.

**`run_from_blob`, in `docs/buffer-abi.md`.** A present-tense table row told the
chip leaf to "hand the POD blob to `run_from_blob`", a name that has never existed
(introduced with the row in #1599). The row now names the three functions that do
the work — `read_args_from_blob`, `ImportRegistry.materialize_args`,
`_submit_chip_run_materialized` — matching `worker.py`'s `submit_frame`.

Deliberately untouched: `tensormap_and_ringbuffer`'s ring documentation.
`current_task_index`, `task_window_mask`, `advance_lock` and `last_task_alive` are
all live there (162 uses across 20 non-doc files); #2004 dropped the ring from
`host_build_graph` only, and said so.

Author: ChaoWao

docs/buffer-abi.md

@@ -250,7 +250,7 @@ A `Tensor` on the wire is materialized differently by each consumer:
 
 | Consumer | What it does |
 | -------- | ------------ |
-| **Chip leaf (L2 runtime)** | Materialize each one to a `ChipTensor` (map-once, keyed by identity), including **strided** views; hand the POD blob to `run_from_blob`. |
+| **Chip leaf (L2 runtime)** | Decode the POD blob with `read_args_from_blob`, then materialize each one to a `ChipTensor` (map-once, keyed by canonical identity) via `ImportRegistry.materialize_args`, including **strided** views, and submit with `_submit_chip_run_materialized`. |
 | **Python sub-worker** (compute) | Map each one into a `MappedArg`; the callable computes with `torch.frombuffer(arg.buffer, ...)`. No `ChipTensor`. |
 | **Nested L4→L3 orch** (forwarding) | **Re-export** each backing to a descriptor `H'` that keeps the source's canonical identity — no pass-through, no map on the forwarding hop. |
 

problems.md

@@ -124,8 +124,9 @@ scalar `rt_orchestration_done` publishes into the runtime header.
 
 **Why the scheduler state is device-written.** `SchedulerState` holds no
 per-run content: `sm_header` and the task-header pointer derive from a pooled SM base,
-queue capacities are compile-time constants, hbg never advances
-`last_task_alive`, and it has no host-side entry point at all. So the host would
+queue capacities are compile-time constants, polling reserves no wiring or
+dependency pool (readiness comes from the task table's `completion_flags`, which
+the task header owns), and it has no host-side entry point at all. So the host would
 only be writing an initialization pattern — 203,392 bytes
 of it, dominated by `AsyncWaitList::entries` — for the device to receive and never
 read. `RuntimeContext` therefore holds a *pointer* to it, wired from

src/a2a3/runtime/host_build_graph/docs/RUNTIME_LOGIC.md

@@ -70,7 +70,9 @@ it derives for itself, from a heap block whose prior contents it never reads.
 
 Before a wait slot is used, the runtime verifies:
 
-- the task ID is valid and carries ring 0, the only ring HBG places tasks on;
+- the task ID is valid and lies in the `RING` id space — a `GRAPH_NODE` id packs
+  its outer task and node index into the low bits, so it indexes no task table
+  slot (see `src/common/host_build_graph/task_id_encoding.h`);
 - the task table slot that ID indexes has a bound task descriptor; and
 - the descriptor's full task ID matches the tensor's owner/producer ID.
 

src/a2a3/runtime/host_build_graph/docs/SCALAR_DATA_ACCESS.md

@@ -689,11 +689,11 @@ struct SchedulerState {
     EarlyDispatchDoorbell early_dispatch_doorbell_table[EARLY_DISPATCH_CORE_MASK_WORDS * 64]{};
 
     // Cross-thread early-dispatch work queues, one ChipReadyQueue MPMC instance per
-    // resource shape (AIC/AIV/MIX) — arena-backed, reserved/wired in pto_runtime2_init
-    // alongside the per-shape ready queues, and indexed the same way. A candidate is
-    // pushed to the queue for its own shape (active_mask.to_shape()) so the drain can
-    // pop per shape and size the pop to that shape's free cores, exactly as normal
-    // dispatch pops ready_queues[shape].
+    // resource shape (AIC/AIV/MIX) — arena-backed, reserved and wired by
+    // SchedulerState::init_data_from_layout alongside the per-shape ready queues, and
+    // indexed the same way. A candidate is pushed to the queue for its own shape
+    // (active_mask.to_shape()) so the drain can pop per shape and size the pop to that
+    // shape's free cores, exactly as normal dispatch pops ready_queues[shape].
     //
     // A consumer's SPMD blocks span cores owned by several AICPU threads, but only a
     // thread RUNNING the consumer's producer discovers it (via the producer's

src/a2a3/runtime/host_build_graph/runtime/scheduler/scheduler.h

@@ -735,11 +735,11 @@ struct SchedulerState {
     EarlyDispatchDoorbell early_dispatch_doorbell_table[EARLY_DISPATCH_CORE_MASK_WORDS * 64]{};
 
     // Cross-thread early-dispatch work queues, one ChipReadyQueue MPMC instance per
-    // resource shape (AIC/AIV/MIX) — arena-backed, reserved/wired in pto_runtime2_init
-    // alongside the per-shape ready queues, and indexed the same way. A candidate is
-    // pushed to the queue for its own shape (active_mask.to_shape()) so the drain can
-    // pop per shape and size the pop to that shape's free cores, exactly as normal
-    // dispatch pops ready_queues[shape].
+    // resource shape (AIC/AIV/MIX) — arena-backed, reserved and wired by
+    // SchedulerState::init_data_from_layout alongside the per-shape ready queues, and
+    // indexed the same way. A candidate is pushed to the queue for its own shape
+    // (active_mask.to_shape()) so the drain can pop per shape and size the pop to that
+    // shape's free cores, exactly as normal dispatch pops ready_queues[shape].
     //
     // A consumer's SPMD blocks span cores owned by several AICPU threads, but only a
     // thread RUNNING the consumer's producer discovers it (via the producer's

src/a2a3/runtime/tensormap_and_ringbuffer/runtime/scheduler/scheduler.h

@@ -124,8 +124,9 @@ scalar `rt_orchestration_done` publishes into the runtime header.
 
 **Why the scheduler state is device-written.** `SchedulerState` holds no
 per-run content: `sm_header` and the task-header pointer derive from a pooled SM base,
-queue capacities are compile-time constants, hbg never advances
-`last_task_alive`, and it has no host-side entry point at all. So the host would
+queue capacities are compile-time constants, polling reserves no wiring or
+dependency pool (readiness comes from the task table's `completion_flags`, which
+the task header owns), and it has no host-side entry point at all. So the host would
 only be writing an initialization pattern — 203,392 bytes
 of it, dominated by `AsyncWaitList::entries` — for the device to receive and never
 read. `RuntimeContext` therefore holds a *pointer* to it, wired from

src/a5/runtime/host_build_graph/docs/RUNTIME_LOGIC.md

@@ -70,7 +70,9 @@ it derives for itself, from a heap block whose prior contents it never reads.
 
 Before a wait slot is used, the runtime verifies:
 
-- the task ID is valid and carries ring 0, the only ring HBG places tasks on;
+- the task ID is valid and lies in the `RING` id space — a `GRAPH_NODE` id packs
+  its outer task and node index into the low bits, so it indexes no task table
+  slot (see `src/common/host_build_graph/task_id_encoding.h`);
 - the task table slot that ID indexes has a bound task descriptor; and
 - the descriptor's full task ID matches the tensor's owner/producer ID.
 

src/a5/runtime/host_build_graph/docs/SCALAR_DATA_ACCESS.md

@@ -689,11 +689,11 @@ struct SchedulerState {
     EarlyDispatchDoorbell early_dispatch_doorbell_table[EARLY_DISPATCH_CORE_MASK_WORDS * 64]{};
 
     // Cross-thread early-dispatch work queues, one ChipReadyQueue MPMC instance per
-    // resource shape (AIC/AIV/MIX) — arena-backed, reserved/wired in pto_runtime2_init
-    // alongside the per-shape ready queues, and indexed the same way. A candidate is
-    // pushed to the queue for its own shape (active_mask.to_shape()) so the drain can
-    // pop per shape and size the pop to that shape's free cores, exactly as normal
-    // dispatch pops ready_queues[shape].
+    // resource shape (AIC/AIV/MIX) — arena-backed, reserved and wired by
+    // SchedulerState::init_data_from_layout alongside the per-shape ready queues, and
+    // indexed the same way. A candidate is pushed to the queue for its own shape
+    // (active_mask.to_shape()) so the drain can pop per shape and size the pop to that
+    // shape's free cores, exactly as normal dispatch pops ready_queues[shape].
     //
     // A consumer's SPMD blocks span cores owned by several AICPU threads, but only a
     // thread RUNNING the consumer's producer discovers it (via the producer's