mirror of
https://github.com/morluto/rea.git
synced 2026-11-05 14:26:07 +00:00
JANG BYUNGKUNandgensokyo-dev
3ce311757d
* feat(il2cpp): read generic containers on demand and render method spec names Generic container indices and records are demand readers on the same metadata snapshot, so links that do not ask for generic instances allocate nothing. Type rendering keeps one graph walk with its workspace, depth and cycle guards; the C#-like names and the Il2CppDumper GetMethodSpecName names are two styles over it. * feat(il2cpp): link generic method instances with admitted rows and kept issues generic_instances links every genericMethodTable row through methodSpecs, genericInsts and genericMethodPointers, choosing the one CodeRegistration layout whose counts bound every row. The table is reserved against the report budget before any row is resolved; rows and issues are admitted before retention, and a contradicting row stays unknown with an address-backed issue instead of failing the link. Views page the rows by definition, state, address and name, unrequested and historical collections stay unknown, and Ghidra seeds fold instances after the definitions. * feat(il2cpp): accept --generic-instances on the CLI and describe it in tool contracts * docs(il2cpp): describe generic method instances and their limits * fix(il2cpp): keep issued generic rows unknown and name their row A method spec that contradicted the metadata still reported its pointer slot as code, while the schema and provider limitation promised unknown. Such rows now keep no address, every issue carries its genericMethodTable row, the domain validator requires unknown exactly on issued rows, and shared specs name one other row instead of all of them. The row-size lower bound moves to the metadata admission owner, the hex helper is shared, and budget wording says checked rather than reserved. * test(il2cpp): reject ambiguous generic layouts and pin budget, yield and limit paths * fix(il2cpp): keep unmapped generic slots per row and reject undeclared arguments Layout selection no longer requires every genericMethodPointers slot to be mapped; each row reads its own slot, so one unmapped slot leaves only that row unknown with an issue. Method specs whose class or method argument count differs from the owner's declared generic container are issued and unnamed. Fixtures now declare the containers their generic rows fill. * fix(il2cpp): page generic method issues instead of retaining them in summaries Summaries keep generic_methods without its issues and count them in counts.generic_method_issues; the generic_method_issues collection pages them by rows, indices and fields, as code_gen_modules already does. Docs also name the method_indices filter the request schema accepts. * fix(il2cpp): read split generic arguments per slot and method containers on demand Generic argument arrays may span adjacent loader ranges: admit them with hasPointerArrayCoverage and read each slot through readPointer, instead of requiring the whole array in one mapping. Method generic container indices are read on demand from the method record, like type containers, so no link path allocates a per-method column. An owner whose container index is out of range now leaves the method spec unnamed with an issue instead of naming it unchecked. * fix(il2cpp): report generic entries past the address space at their row A damaged methodSpecs or genericMethodPointers pointer could make a spec or slot address exceed 64 bits. The issue kept that address, which the address schema rejects, so inspection failed instead of leaving the row unknown. Check the entry range first and attach the issue to the genericMethodTable row (or its slot field). * chore(il2cpp): remove unrelated formatting changes * fix: restore adb float pattern and mobile fixture from main The merge commit's lint-staged run rewrote these two main files: oxlint --fix dropped the escape in [+\-], which is invalid under the RegExp v flag. * fix(il2cpp): preserve unreadable generic method identities * docs(il2cpp): record generic-instance workflow verification * docs(il2cpp): qualify final integration checks --------- Co-authored-by: gensokyo-dev <342580529+gensokyo-dev@users.noreply.github.com>
139 lines
9.1 KiB
Markdown
139 lines
9.1 KiB
Markdown
# IL2CPP generic-instance verification, 2026-10-24
|
||
|
||
## Scope and pinned implementation
|
||
|
||
PR #5785 adds opt-in static generic-method instance linking to the existing
|
||
IL2CPP reader and exposes retained instances and issues to CLI/MCP views and
|
||
existing Ghidra seed generation. No target is executed. The feature supports
|
||
metadata 27/29/31 with a static CodeRegistration and the existing 64-bit,
|
||
little-endian native image owner. Generic decoding for metadata 24/39 and builds
|
||
without static registration remains explicitly unsupported; this does not remove
|
||
ordinary metadata decoding for those versions.
|
||
|
||
Executed source pin: `11b84c9566027da423f2c7c3752ae2e87f0fdbb7`.
|
||
It includes main through `d87e73f2db50be72d33a4b3c7199af35518b441c` via
|
||
`2718a3c775fb1a80e84066acc0f973e4ffd0e000`; `219ff0b9290e3a6c7c99d40182832b9bbcf4e7e7`
|
||
removes two unrelated formatting differences. This report and the query-choice
|
||
paragraph are documentation-only changes after the measured source pin.
|
||
|
||
## Correctness and review ownership
|
||
|
||
The contributor already corrected both earlier review findings: summaries retain
|
||
an issue count rather than the issue array, and `generic_method_issues` owns
|
||
selected/paged issue details; method generic containers are read on demand from
|
||
the shared record-layout owner. No per-method generic-container vector is built
|
||
when generic instances are disabled. Tests inspect the actual link-fact columns,
|
||
retained views, historical omissions, malformed references, budgets and stable
|
||
input identity. These are verification of the contributor's changes, not a new
|
||
implementation of those fixes.
|
||
|
||
A further producing-boundary defect was reproduced: unreadable method-spec bytes
|
||
became the sentinel `-1`, conflating unavailable data with a recorded negative
|
||
index. The corrected producer preserves `method: null` when unreadable and keeps
|
||
an actual recorded `-1`. The named-instance contract requires a known method;
|
||
unknown unnamed rows retain null and cannot match a method-index selector.
|
||
The producer regression fails before correction (112 pass, one fail).
|
||
An intermediate new view expectation omitted an existing parsed offset; that test
|
||
expectation was corrected without changing the product contract.
|
||
|
||
All **285 selected checks pass** after the correction, covering native registration,
|
||
metadata containers, generic resolution, domain views, public CLI and actual
|
||
production-stdio MCP with advertised schemas. The complete task takes 164.798 s,
|
||
149.772 CPU s and 2,441.5 MiB sampled peak process-family RSS. Type/lint,
|
||
full formatting and generated-document checks also pass at that source pin.
|
||
An earlier integrated selection passed all 280 existing checks before adding the
|
||
five new unreadable/index-contract cases. No simulated provider establishes Unity
|
||
runtime dispatch, and no real Ghidra import of generated seeds was executed.
|
||
|
||
## Completed MCP query comparison
|
||
|
||
Analyst question: obtain instance/issue counts and states, the first two instance
|
||
identities and RVAs, and the first duplicate-spec issue, retaining both input
|
||
artifact identities. Two authored version-31 inputs each contain three instances
|
||
and two issues: one code row and two unknown duplicate-spec rows. The long case
|
||
uses a 4,096-character type name; these are authored format fixtures, not binaries
|
||
produced by Unity.
|
||
|
||
| Artifact | Bytes | SHA-256 |
|
||
| ---------------------- | ----: | ----------------------------------------------------------------------- |
|
||
| Small metadata | 1,176 | `743ef4841be52734a349d9094fff971a8161ba35e9644d393c5803f84f71dc7b6cd7b` |
|
||
| Long-name metadata | 5,273 | `031d4e1f08219d7da4ff3fc459f8791a7e8c041234a156358361a3d3a7b6cd27` |
|
||
| Both native ELF inputs | 5,072 | `a6bf6e29397a389ae50f251b4167983b62cf1d9812c9108f6a17965eaa821259` |
|
||
|
||
`complete` makes one inspection call with complete detail. `paged` makes a summary
|
||
inspection followed by two retained views: two selected instance rows/fields and
|
||
one issue. Three alternating pairs per input give **12 completed tasks**. Every
|
||
answer agrees on the requested facts and identities; file digests are unchanged.
|
||
Complete and paged Evidence objects are different workflows, so full Evidence/ID
|
||
equality is not claimed. The actual SDK-converted output schemas validate with
|
||
Ajv 2020. Two earlier probes used a different input-schema profile and are excluded
|
||
from the paired results.
|
||
|
||
All runs use pinned Node 24.22.1, npm 11.16.0, the same locked dependencies, two
|
||
allowed CPUs, a 2 GiB Node heap and production stdio with the default schema
|
||
profiles. Every task starts a fresh SDK client and server. OS/file caches were
|
||
primed by probes; persistent compilation-cache state was not isolated. Client and
|
||
server startup, lazy schema validation, RPC trace writing and final file hashing
|
||
are included in elapsed/CPU/memory costs. The first small pair was substantially
|
||
slower; no cause is assigned and those completed runs remain in the ranges.
|
||
|
||
Token estimates use tokenx 1.3.0 on every reserialized JSON-RPC request,
|
||
notification, catalog and reply, counting both text and structured carriers.
|
||
They are estimates of transmitted payloads, **not billed tokens or measured model
|
||
context**. Transport framing, stderr, fixture generation, token-estimation CPU,
|
||
prompts, agent reasoning and model-specific conversion are outside this measure.
|
||
The full catalog is 222 tools and estimates 1,003,790 tokens, including advertised
|
||
output schemas. Its name/description/input-schema subset alone estimates 72,079
|
||
tokens (259,949 UTF-8 bytes); neither establishes what a particular agent sends to
|
||
its model. The catalog is unchanged across these query workflows.
|
||
|
||
| Input / workflow | Calls | Entire RPC token estimate | Tool-result estimate | Elapsed s, median (range) | CPU s, median (range) | Peak family RSS MiB, median (range) |
|
||
| ---------------- | ----: | ------------------------: | -------------------: | ------------------------- | ---------------------- | ----------------------------------- |
|
||
| Small / complete | 1 | 1,012,270 | 7,582 | 20.216 (19.957–39.633) | 19.992 (19.186–36.537) | 946.1 (927.5–947.8) |
|
||
| Small / paged | 3 | 1,018,217 | 13,317 | 25.325 (24.214–37.530) | 24.853 (24.471–36.119) | 1,072.7 (1,066.3–1,222.9) |
|
||
| Long / complete | 1 | 1,025,889 | 21,201 | 19.211 (17.895–20.017) | 19.762 (19.351–20.314) | 935.7 (934.2–941.8) |
|
||
| Long / paged | 3 | 1,018,217 | 13,317 | 23.724 (23.712–26.661) | 23.907 (23.097–24.567) | 1,108.8 (1,093.8–1,123.2) |
|
||
|
||
Paging increases the small-task estimate by 5,947 (0.59%). For the long case it
|
||
saves 7,672 (0.75% of the whole RPC estimate), while using more calls, CPU and
|
||
memory and taking longer. The 37.2% reduction in long-case tool-result estimates
|
||
is not a whole-task/context reduction. There is no general performance win here;
|
||
choose complete output for small questions spanning both collections and selected
|
||
pages when irrelevant retained text would otherwise dominate. The measured data
|
||
does not qualify large real tables or the opt-out allocation savings numerically.
|
||
|
||
CPU is waited-child accounting. Family RSS sums recursively observed processes
|
||
at 50 ms intervals; it may double-count shared pages and miss short-lived children.
|
||
It is not a cgroup memory high-water mark. Host load and primed file caches limit
|
||
comparability; no statistical speedup is asserted.
|
||
|
||
## Real-artifact limitations
|
||
|
||
The contributor reports two private, self-built Unity 2022.3.62f3 Android arm64
|
||
version-31 builds (ordinary and identical-code folding), compared in both
|
||
directions with Il2CppDumper: 5,329 instance rows, 5,285 code, 44 no-body and no
|
||
issues, using 16-byte table rows. Native hashes are
|
||
`580ba23b41b9470bff8ec1a06edf06294221d9b363afba3b7337632a5cef2415` and
|
||
`9a0708bdbbc579f183856d3b633a16429558cd58db9863b05acbc6dcc8ae51ce`;
|
||
metadata is `2eb6338d81a68dffd34234d26ec1bf82c3e00c435e4d14f349ba03acc8e48b70`.
|
||
The oracle's Il2CppDumper version was not recorded. Those samples are not
|
||
redistributed and this maintenance run **did not reproduce that comparison**.
|
||
Other historical contributor platform checks were not rerun. Metadata 27/29 and
|
||
the new unreadable/negative-index cases have authored boundary coverage only.
|
||
No authentic Unity artifact, game execution, real target dispatch or Ghidra import
|
||
is qualified by the 12 query tasks above.
|
||
|
||
## Final integration repeat
|
||
|
||
Final executed source pin `7fdb5a29885ce05ec9181a609a37bf008b94e5df`
|
||
incorporates main `98ec00f4068bab894639d0d1d83ac0bfc428bcbf` and the author's
|
||
later cleanup `3dd803b81ccfb6b75afaa8c616693e8c4e7183d1`, preserving both authors'
|
||
changes. All **285** selected checks pass again: 183.057 s
|
||
elapsed, 187.472 CPU s and
|
||
2,583.3 MiB sampled peak process-family
|
||
RSS. Type/lint, full formatting and generated-document checks pass. This is a
|
||
completed validation cost with primed file caches, not a speed comparison.
|
||
The 12 paired query measurements remain pinned to the earlier source revision
|
||
and are not re-labelled as measurements of this integration repeat. The final
|
||
report update changes documentation only.
|