perf(runner-shared): buffer the decompressed memtrack event stream

[not-matthias]

Buffer the decompressed stream in MemtrackArtifact::decode_streamed.

rmp_serde reads 1 to 8 bytes per call. Without a buffer, each of those reads goes through zstd's stream API. A BufReader now sits between the zstd decoder and rmp_serde.

On a large memtrack artifact, decoding all events drops from 21-33 s to 11-21 s. Peak memory is unchanged.

Refs COD-3704

[greptile-apps]

Confidence Score: 5/5

[Medium risk] Adds buffering to memory tracking event deserialization.

The PR appears safe to merge.

Summary

The PR adds a buffer between the zstd decoder and the MessagePack deserializer to reduce small reads through the decoder. No actionable issues were identified.

_{Reviews (1) · Last reviewed commit: "perf(runner-shared): buffer the decompre..."}

[codspeed]

Merging this PR will not alter performance

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

✅ 31 untouched benchmarks
⏩ 6 skipped benchmarks¹

---

_{Comparing perf-buffer-memtrack-decode (1cf4bcc) with main (407b28e)}

Footnotes

1. 6 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. ↩