google/fuzzer-test-suite

evaluate libFuzzer's -fork=1 mode

kcc opened this issue · 1 comments

kcc commented

Matt, please try to evaluate -fork=0 (default) vs -fork=1 using the fresh clang

In general fork modes seems to start off much slower than the normal mode, but catches up and wins for benchmarks that crash a lot (e.g., pcre2), while significantly lagging behind for less buggy benchmarks.