test(perf): track newly native repository reads in CI benchmarks

Extend the existing `pyperf` journey with submodule inventory, tree/index
revision paths and modes, and raw commit/tree readback. These use fresh public
wrappers on the retained `Repo` and automatically join the existing separate
benchmark CI job, result parity checks and per-measurement reporting.

Read existing object bytes rather than invoking serializers: the fixed fixture
has a signed commit, and benchmarking must not rewrite repository objects.
Add a zero-CLI ceiling for each new measurement while retaining every existing
ceiling. The journey still permits only its single patch-diff fallback launch.

This measurement-only commit reduces backend CLI launches by 0. Its warm
preflight records CLI/Gix counts of 11/0 for submodule inventory, 8/0 for
revision paths and 1/0 for object readback. The complete expanded journey is
43/1. The preceding optimization commits implement these native reductions;
this change keeps them visible and enforceable as coverage grows.

Validation: Gix preflight passed all 14 ceilings and five comparison tests,
then the CLI preflight and five tests passed. All 14 result digests matched.
Repository-wide Ruff lint/format, mypy, basedpyright and `git diff --check`
passed. Timed measurements will be recorded separately after final validation.

Author: codex

doc/gix-backend.md

@@ -186,14 +186,19 @@ environment metadata.
 ### Existing-repository benchmark
 
 [`test/performance/README.md`](../test/performance/README.md) describes the
-`pyperf` harness and fixed GitPython 3.1.45 fixture. It measures eight public-API
+`pyperf` harness and fixed GitPython 3.1.45 fixture. It measures eleven public-API
 operations and their complete journey on one already-open `git.Repo`, plus
 separate direct opening and discovery from `git/objects`. Fresh high-level
 wrappers preserve the cost of actual operations; imports, fixture preparation
 and parity preflight are outside timing. Native repository refresh remains
 inside operation timing, so these measurements include the cost of keeping
 retained state current.
 
+Submodule inventory, revision path/mode lookup and raw commit/tree readback
+extend the already-open journey, each with a zero-CLI ceiling. The readback
+measurement reads existing bytes, including the signed fixture commit, without
+rewriting objects. Historical tables below retain their original workloads.
+
 At `244e418da6cc43de129cbe2908d11be4c5ad457a`, using official GixPython
 0.1.0 and the same existing CPython 3.12.14/macOS arm64 interpreter for both
 installations, with Git 2.54.0 (Apple Git-157), six worker processes with three values each produced the

test/performance/README.md

@@ -1,7 +1,7 @@
 # Backend benchmark
 
 `bench_repository.py` uses `pyperf` to measure a complete read-only journey and
-eight named operations against an existing repository, plus separate direct
+eleven named operations against an existing repository, plus separate direct
 opening and nested-directory discovery measurements. It keeps one `Repo`
 instance open per worker, but creates fresh commit/tree/index wrappers per
 invocation so their cached properties cannot turn the measurement into a no-op.
@@ -72,6 +72,12 @@ at least one parent, a root `README.md` and a `git/objects` directory for the
 nested discovery measurement. The harness never prepares or
 mutates that worktree, and ignores ambient Git environment/config overrides.
 
+The journey also enumerates submodules, resolves tree/index revisions with
+their paths and modes, and reads raw commit/tree bytes through the object
+database. The last operation covers serialization readback without rewriting
+objects; the pinned fixture's signed commit remains intact. These three
+measurements have zero-CLI ceilings and join the existing CI job automatically.
+
 Add a function returning JSON-compatible results to `MEASUREMENTS` in
 `bench_repository.py`. It automatically becomes a separate benchmark and part
 of the journey and parity checks. Include supported and fallback operations;

test/performance/bench_repository.py

@@ -44,6 +44,21 @@ def index(repo):
     return sorted([path, stage, entry.mode, entry.hexsha] for (path, stage), entry in repo.index.entries.items())
 
 
+def submodules(repo):
+    return [[module.name, module.path, module.hexsha, module.url, module.branch_path] for module in repo.submodules]
+
+
+def revision_paths(repo):
+    objects = [repo.rev_parse(revision) for revision in ("HEAD:README.md", "HEAD:git", ":README.md")]
+    return [[obj.hexsha, obj.path, obj.mode] for obj in objects]
+
+
+def object_readback(repo):
+    commit = repo.commit("HEAD")
+    # Read the same bytes used by serialization, without writing objects or stripping signatures.
+    return [sha256(obj.data_stream.read()).hexdigest() for obj in (commit, commit.tree)]
+
+
 def graph(repo):
     tip, parent = repo.commit("HEAD"), repo.commit("HEAD~1")
     return {
@@ -81,6 +96,9 @@ def worktree(repo):
     "history_25": history,
     "browse_tree_and_blob": browse,
     "read_index": index,
+    "submodule_inventory": submodules,
+    "revision_paths": revision_paths,
+    "object_readback": object_readback,
     "revision_graph": graph,
     "diff_and_stats": changes,
     "patch_diff": patch,

test/performance/cli-budget.json

@@ -6,6 +6,9 @@
     "history_25": 0,
     "browse_tree_and_blob": 0,
     "read_index": 0,
+    "submodule_inventory": 0,
+    "revision_paths": 0,
+    "object_readback": 0,
     "revision_graph": 0,
     "diff_and_stats": 0,
     "patch_diff": 1,