RFR: 8300915: G1: incomplete SATB because nmethod entry barriers don't get armed [v2]

[Thomas Schatzl]

On Wed, 25 Jan 2023 17:39:47 GMT, Richard Reingruber <rrich at openjdk.org> wrote:

>> src/hotspot/share/gc/g1/g1ConcurrentMark.cpp line 796:
>> 
>>> 794: 
>>> 795:   CodeCache::on_gc_marking_cycle_start();
>>> 796:   CodeCache::arm_all_nmethods();
>> 
>> Maybe factor this out into a method (`start_codecache_marking_cycle`) and call it above.
>
> Like this right?
> 
> ```c++
> void G1CollectedHeap::start_codecache_marking_cycle() {
>   CodeCache::on_gc_marking_cycle_start();
>   CodeCache::arm_all_nmethods();
> }

Exactly. Also I meant "... to call that method here and above"; i.e. I asked to factor out because this is done twice.

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

PR: https://git.openjdk.org/jdk/pull/12194