mirror of
https://github.com/rui314/mold.git
synced 2026-11-07 09:16:04 -10:00
Rui Ueyama
f41bfcd5c7
C++ reference: mold fbaaa75f9f
Benchmark protocol: 32 threads; one C++ warmup followed by five
interleaved measured runs per linker; median reported. Every linker
invocation was preceded by sync and followed by sync; sleep 1.
Workload C++ (s) Rust (s) Rust vs C++
blender-debug 0.821 0.770 -6.21%
blender-release 0.187 0.165 -11.76%
chrome-debug-arm64 1.714 1.815 +5.89%
chrome-debug-x86_64 1.602 1.643 +2.56%
chrome-release-arm64 0.713 0.751 +5.33%
chrome-release-x86_64 0.608 0.636 +4.61%
clang-debug 1.303 1.139 -12.59%
clang-release 0.097 0.097 +0.00%
clickhouse-debug 0.917 0.906 -1.20%
clickhouse-release 0.379 0.386 +1.85%
firefox-debug-arm64 0.918 0.935 +1.85%
firefox-debug 0.787 0.802 +1.91%
firefox-release 0.200 0.206 +3.00%
godot-debug 0.445 0.418 -6.07%
godot-release 0.066 0.075 +13.64%
libmergedlo.so-debug 0.424 0.439 +3.54%
libmergedlo.so-release 0.184 0.171 -7.07%
pytorch-debug 0.794 0.736 -7.30%
pytorch-release 0.142 0.136 -4.23%
tensorflow-debug 3.043 2.808 -7.72%
tensorflow-release 0.657 0.610 -7.15%
Validation: cargo build --release; ./run-tests.sh --all (pass 5360,
skip 636, fail 0); cargo clippy --release --all-targets.
3 lines
31 B
TOML
3 lines
31 B
TOML
[toolchain]
|
|
channel = "stable"
|