Liang Mao

> 我这边程序有个需要交互的控制台，elesticheap的输出经常把有用的信息或者界面冲掉 什么叫把有用的信息冲掉？是指多出来的gc日志信息吗？

> 又是一个很好的例子 >  我了解了，后面我们看看修复一下

Could reviewers please help with the review? The risk is very low as the original wrong encoding is not used in code base. riscv64 build failed because of the risc...

> > Could reviewers please help with the review? The risk is very low as the original wrong encoding is not used in code base. > > There is, therefore,...

> > > > Could reviewers please help with the review? The risk is very low as the original wrong encoding is not used in code base. > > >...

> Inspecting the existing call chains of `set_should_clear_all_soft_refs`, I can see that Parallel full-gc already invokes it via adaptive-size-policy. Yes, set_should_clear_all_soft_refs will be set to true adaptively but never reset...

> > set_should_clear_all_soft_refs will be set to true adaptively but never reset to false. > > I see; does it make sense to reset to false inside `check_gc_overhead_limit` so that...

> `set_should_clear_all_soft_refs` should match in pairs. If it's reset in the destructor, where is corresponding set-to-true? > > (In the Parallel case, the destructor is invoked after `check_gc_overhead_limit` so the...

> In `PSParallelCompact::invoke_no_policy`, the structure is essentially: > > ``` > { > ClearedAllSoftRefs ... > check_gc_overhead_limit; // set-to-true > } > ``` > > The destructor (as in this...

> Personally, I believe ClearedAllSoftRefs will not be necessary in the long run, as it is only used by the full GC of Parallel and G1. > > To address...