perf(pm): update resolver cache ownership docs#3048
Conversation
elrrrrrrr
added
A-Pkg Manager
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates documentation comments in manifest.rs and registry.rs to reflect architectural changes regarding the BFS resolver's de-duplication of in-flight jobs and the intended usage of the registry URL. Feedback was provided to clarify that de-duplication prevents redundant parsing rather than bounding the cost of a single call, and to replace the informal term 'regular trait methods' with specific high-level resolution methods in the trait documentation.
| /// The BFS resolver de-duplicates in-flight extract jobs per key, so the | ||
| /// per-call full-tree parse cost is bounded. |
There was a problem hiding this comment.
The phrasing "per-call full-tree parse cost is bounded" is somewhat ambiguous. Since the method performs a full JSON parse of the raw bytes on every invocation, the cost of a single call is always proportional to the manifest size. The de-duplication mentioned actually limits the number of times this cost is incurred for concurrent requests. Consider rephrasing to clarify that the de-duplication prevents redundant parsing of the same version.
/// The BFS resolver de-duplicates in-flight extract jobs per key, ensuring
/// this expensive full-tree parse is not performed redundantly.
| /// | ||
| /// Implementations without a concrete URL can keep the default. | ||
| /// Implementations that cannot expose a concrete URL can keep the default; | ||
| /// callers should fall back to the regular trait methods when this is empty. |
There was a problem hiding this comment.
The term "regular trait methods" is a bit informal. Explicitly mentioning that callers should use high-level resolution or fetching methods (like resolve_package) would improve clarity for implementers and consumers of the trait, especially when they need to obtain absolute URLs for raw fetches.
/// callers should use high-level resolution methods (e.g., resolve_package)
/// to obtain absolute URLs when this is empty.
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-remove-memory-cache
branch
from
f029690 to
0225a9c
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-doc-cleanup
branch
from
953e26b to
9b50ad8
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-doc-cleanup
branch
from
9b50ad8 to
9a0819b
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-remove-memory-cache
branch
from
0225a9c to
83396aa
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-doc-cleanup
branch
from
9a0819b to
bb23373
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-remove-memory-cache
branch
from
83396aa to
8962d83
Compare
elrrrrrrr
force-pushed
the
perf/pm-split-resolver-doc-cleanup
branch
from
bb23373 to
711835d
Compare
📊 pm-bench-phases ·
|
| PM | wall | ±σ | user | sys | RSS | pgMinor |
|---|---|---|---|---|---|---|
| bun | 9.17s | 0.26s | 10.38s | 10.38s | 773M | 345.8K |
| utoo-next | 8.80s | 1.13s | 10.67s | 12.47s | 992M | 119.0K |
| utoo-npm | 8.34s | 0.07s | 10.77s | 12.53s | 947M | 128.2K |
| utoo | 8.14s | 0.44s | 11.18s | 12.39s | 978M | 137.9K |
| PM | vCtx | iCtx | netRX | netTX | cache | node_mod | lock |
|---|---|---|---|---|---|---|---|
| bun | 14.1K | 17.7K | 1.19G | 6M | 1.88G | 1.76G | 1M |
| utoo-next | 116.7K | 84.3K | 1.16G | 5M | 1.71G | 1.70G | 2M |
| utoo-npm | 117.3K | 84.9K | 1.16G | 5M | 1.71G | 1.70G | 2M |
| utoo | 104.7K | 60.7K | 1.16G | 6M | 1.71G | 1.70G | 2M |
p1_resolve
| PM | wall | ±σ | user | sys | RSS | pgMinor |
|---|---|---|---|---|---|---|
| bun | 2.03s | 0.13s | 4.02s | 1.06s | 518M | 172.1K |
| utoo-next | 2.86s | 0.10s | 5.40s | 1.70s | 608M | 82.3K |
| utoo-npm | 3.06s | 0.08s | 5.61s | 2.08s | 610M | 89.3K |
| utoo | 2.36s | 0.06s | 5.97s | 1.66s | 653M | 120.9K |
| PM | vCtx | iCtx | netRX | netTX | cache | node_mod | lock |
|---|---|---|---|---|---|---|---|
| bun | 8.2K | 4.6K | 202M | 3M | 107M | - | 1M |
| utoo-next | 46.3K | 67.5K | 200M | 2M | 7M | 3M | 2M |
| utoo-npm | 71.2K | 88.3K | 200M | 2M | 7M | 3M | 2M |
| utoo | 14.8K | 18.2K | 202M | 3M | 7M | 3M | 2M |
p3_cold_install
| PM | wall | ±σ | user | sys | RSS | pgMinor |
|---|---|---|---|---|---|---|
| bun | 6.89s | 0.49s | 6.32s | 10.11s | 627M | 207.4K |
| utoo-next | 5.81s | 0.11s | 5.00s | 11.05s | 484M | 61.0K |
| utoo-npm | 7.02s | 2.13s | 5.21s | 11.50s | 493M | 60.2K |
| utoo | 6.94s | 2.17s | 5.00s | 11.01s | 508M | 67.2K |
| PM | vCtx | iCtx | netRX | netTX | cache | node_mod | lock |
|---|---|---|---|---|---|---|---|
| bun | 4.8K | 6.6K | 1019M | 3M | 1.76G | 1.76G | 1M |
| utoo-next | 95.8K | 50.0K | 989M | 3M | 1.70G | 1.70G | 2M |
| utoo-npm | 108.7K | 50.7K | 989M | 3M | 1.70G | 1.70G | 2M |
| utoo | 102.1K | 51.4K | 989M | 3M | 1.70G | 1.70G | 2M |
p4_warm_link
| PM | wall | ±σ | user | sys | RSS | pgMinor |
|---|---|---|---|---|---|---|
| bun | 3.35s | 0.06s | 0.22s | 2.35s | 134M | 33.0K |
| utoo-next | 2.37s | 0.21s | 0.54s | 3.93s | 78M | 18.5K |
| utoo-npm | 2.30s | 0.05s | 0.53s | 3.86s | 78M | 18.2K |
| utoo | 2.32s | 0.06s | 0.54s | 3.95s | 79M | 18.7K |
| PM | vCtx | iCtx | netRX | netTX | cache | node_mod | lock |
|---|---|---|---|---|---|---|---|
| bun | 301 | 26 | 5M | 15K | 1.91G | 1.75G | 1M |
| utoo-next | 42.1K | 19.3K | 3K | 29K | 1.70G | 1.70G | 2M |
| utoo-npm | 42.0K | 19.8K | 1K | 6K | 1.70G | 1.70G | 2M |
| utoo | 43.5K | 20.6K | 1K | 5K | 1.71G | 1.70G | 2M |
npmmirror.com: no output captured.
|
Closing this temporary fine-grained draft split. We are replacing it with a smaller 9-PR review stack based on the source PRs. |
1 participant
Summary
RegistryClient::registry_urlfallback behavior for raw-fetch callersValidation
cargo fmtcargo check -p utoo-ruboristcargo clippy --all-targets -- -D warnings --no-depsSplit Plan
Part of the resolver stack split from source PR #3028. This is the final documentation cleanup after cache ownership moved out of
UnifiedRegistry.