RFR: 8273143: Transition to _thread_in_vm when handling a polling page exception [v2]

[Patricio Chilano Mateo]

On Tue, 11 Jan 2022 17:49:57 GMT, Daniel D. Daugherty <dcubed at openjdk.org> wrote:

> Thumbs up. It took a bit of digging to determine equivalence of the new code relative to the old code. I do have one query in there about a call to `make_walkable()`.
> 
> It would be good to run at least one Tier7 and one Tier8 on this change. You'll get the longer running stress tests that way.
>
Thanks for the review Dan! I'll go ahead and run it through tier7 and tier8. 

Patricio

> src/hotspot/share/runtime/handshake.cpp line 492:
> 
>> 490:   assert(!_handshakee->is_terminated(), "should not be a terminated thread");
>> 491: 
>> 492:   _handshakee->frame_anchor()->make_walkable(_handshakee);
> 
> The old code in `ThreadInVMForHandshake` protected the `make_walkable()`
> call with `if (thread->has_last_Java_frame()) {`. Why/How did you decide that
> you no longer need that check?

I removed it because has_last_Java_frame() just checks if the _last_Java_sp != NULL, but make_walkable() already checks for that at the beginning (in the platforms where we actually do something).

-------------

PR: https://git.openjdk.java.net/jdk/pull/7009