Added a bump allocator to test it's perf again pooling,
turned out for unpooled request, the time is decreased by 30%,
and as expected, for pooled request, time consume goes up by 6%,
which indicate bump allocator is better suit for case where pooling request is hard, allocation is small and phased
Added a bump allocator to test it's perf again pooling,
turned out for unpooled request, the time is decreased by 30%,
and as expected, for pooled request, time consume goes up by 6%,
which indicate bump allocator is better suit for case where pooling request is hard, allocation is small and phased