db_bench throws an exception when passed an invalid benchmark

[mrambacher]

Expected behavior

I expected an error "no such benchmark"

Actual behavior

Got a stack trace out.

Steps to reproduce the behavior

./db_bench --db=/tmp --use_existing_db=false --benchmarks=fred --compression_type=none

Note that I think this is a regression introduced by a recent change in 6.11 (I do not recall seeing it previously).

[akankshamahajan15]

thanks @mrambacher for raising it. @jay-zhuang Could you take a look? I think it shows data race in Timer called through StatsDumpScheduler.

[akankshamahajan15]

Update: With the latest changes, it shows error in the deletion of unique_ptr Timer.