perf(concat): hoist per-token chunk state out of the per-iteration closure#331
Merged
Conversation
Merging this PR will improve performance by 1.97%
|
| Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|
| ⚡ | add_sourcemap_loop |
343.3 µs | 336.8 µs | +1.92% |
| ⚡ | from_sourcemaps |
325.4 µs | 319 µs | +2.02% |
Tip
Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.
Comparing perf/concat-hoist-state (3959542) with main (db883f9)
Footnotes
-
5 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
Boshen
force-pushed
the
perf/concat-hoist-state
branch
2 times, most recently
from
15efed1 to
a6cbefa
Compare
…osure `add_sourcemap`'s per-token loop captured `&mut self` inside the `get_source_id().map(...)` / `get_name_id().map(...)` closures so it could update `token_chunk_prev_source_id` / `token_chunk_prev_name_id` on every iteration. That forced the compiler to round-trip the prev_* fields through memory (load, mutate, store via the `self` pointer) for every token rather than keeping them in registers. Track the running prev values in plain locals, then write them back to `self` once at the end. Also handle the first token separately so the `i == 0` dedup check no longer runs every iteration. Behavior, including the duplicate-token de-duplication at the concatenation seam, is unchanged.
Boshen
force-pushed
the
perf/concat-hoist-state
branch
from
a6cbefa to
3959542
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
`add_sourcemap`'s per-token loop captured `&mut self` inside the `get_source_id().map(...)` / `get_name_id().map(...)` closures so it could update `token_chunk_prev_source_id` / `token_chunk_prev_name_id` on every iteration. That forced the compiler to round-trip the `prev_*` fields through memory (load, mutate, store via the `self` pointer) for every token rather than keeping them in registers.
Track the running prev values in plain locals, then write them back to `self` once at the end. Also handle the first token separately so the `i == 0` dedup check no longer runs every iteration.
Behavior, including the duplicate-token de-duplication at the concatenation seam, is unchanged.