41f19708c8
The benchmark creates a set of DSOs that mimic the work involved in loading the current version of libandroid_servers.so. The synthetic benchmark has roughly the same number of libraries with roughly the same relocations. Currently, on a local aosp_walleye build that includes recent performance improvements (including the Neon-based CL I3983bca1dddc9241bb70290ad3651d895f046660), using the "performance" governor, the benchmark reports these scores: $ adb shell taskset 10 \ /data/benchmarktest64/linker-reloc-bench/linker-reloc-bench \ --benchmark_repetitions=20 --benchmark_display_aggregates_only=true ... -------------------------------------------------------------------------------- Benchmark Time CPU Iterations -------------------------------------------------------------------------------- BM_linker_relocation/real_time_mean 70048 us 465 us 20 BM_linker_relocation/real_time_median 70091 us 466 us 20 BM_linker_relocation/real_time_stddev 329 us 8.29 us 20 $ adb shell taskset 10 \ /data/benchmarktest/linker-reloc-bench/linker-reloc-bench \ --benchmark_repetitions=20 --benchmark_display_aggregates_only=true ... -------------------------------------------------------------------------------- Benchmark Time CPU Iterations -------------------------------------------------------------------------------- BM_linker_relocation/real_time_mean 83051 us 462 us 20 BM_linker_relocation/real_time_median 83069 us 464 us 20 BM_linker_relocation/real_time_stddev 184 us 8.91 us 20 Test: manual Bug: none Change-Id: I6dac66978f8666f95c76387093bda6be0151bfce
1 KiB
Dynamic Linker Relocation Benchmark
This benchmark measures the time spent in the dynamic linker to load a large set
of shared objects with many symbols and relocations. It mimics the work involved
in loading a fixed version of libandroid_servers.so.
Running the benchmark
To run the benchmark, build the linker-reloc-bench target, sync data, and
run the benchmark from /data/benchmarktest[64]/linker-reloc-bench.
There is also a run_bench_with_ninja.sh script that uses the
gen_bench.py --ninja mode to generate a benchmark. It's useful for
experimentation. The --cc and --linker flags allow swapping out different
static and dynamic linkers.
Regenerating the synthetic benchmark
regen/dump_relocs.py scans an ELF file and its dependencies, outputting a JSON
dump, then regen/gen_bench.py processes the JSON file into benchmark code.
gen_bench.py has two modes:
- (default) generate Android.bp and source files
- with
--ninja: generate a build.ninja instead, and build a set of ELF file outputs