RFR: 8333258: C2: high memory usage in PhaseCFG::insert_anti_dependences() [v2]

[Emanuel Peter]

On Tue, 25 Jun 2024 13:51:23 GMT, Emanuel Peter <epeter at openjdk.org> wrote:

>> src/hotspot/share/opto/gcm.cpp line 585:
>> 
>>> 583:     if (worklist_mem.at(j-1) != mem) {
>>> 584:       // We're done with the uses of mem
>>> 585:       return false;
>> 
>> Should this be `continue;` instead to make sure we scan all memory nodes?
>
> Seems my old comments have disappeared, and with them Roland's answers.
> 
> Roland said duplicates are ok. But if you @vnkozlov are asking, then @rwestrel needs to make a comment about why we can early exit here ;)

Ah, it the comment is still here:
https://github.com/openjdk/jdk/pull/19791#issuecomment-2180754800

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

PR Review Comment: https://git.openjdk.org/jdk/pull/19791#discussion_r1652867396