perf(new-execution): use page-based approach for merkle tree metering#1770
Conversation
CodSpeed WallTime Performance ReportMerging #1770 will improve performances by 36.24%Comparing
|
| Benchmark | BASE |
HEAD |
Change | |
|---|---|---|---|---|
| ⚡ | benchmark_execute_metered[bubblesort] |
475.4 ms | 417.9 ms | +13.74% |
| ⚡ | benchmark_execute_metered[keccak256] |
594.6 ms | 437.1 ms | +36.04% |
| ⚡ | benchmark_execute_metered[quicksort] |
543.8 ms | 440.1 ms | +23.58% |
| ⚡ | benchmark_execute_metered[sha256] |
597.9 ms | 438.8 ms | +36.24% |
This comment has been minimized.
This comment has been minimized.
CodSpeed Instrumentation Performance ReportMerging #1770 will improve performances by 16.4%Comparing Summary
Benchmarks breakdown
|
4ef0470 to
8d3bfdf
Compare
This comment has been minimized.
This comment has been minimized.
6dae7f2 to
ba1e9a3
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
759ec11 to
65af708
Compare
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
| } | ||
| } | ||
|
|
||
| // TODO(ayush): can segmentation also be triggered by timestamp overflow? should that be tracked? |
There was a problem hiding this comment.
the interaction bound is always stricter than timestamp overflow, so timestamp overflow does not need to be tracked
jonathanpwang
left a comment
There was a problem hiding this comment.
LGTM, although we would still prefer if we could skip the page approach altogether (I'm assuming you tried this one first since it was faster to implement)
Commit: 8488120 |
BitSetof page indicestest_segmentation_retrysince it's not relevant nowTowards INT-4068