RFR: 8315916: assert(C->live_nodes() <= C->max_node_limit()) failed: Live Node limit exceeded [v7]

Emanuel Peter epeter at openjdk.org
Tue Apr 22 15:22:57 UTC 2025


On Tue, 22 Apr 2025 14:49:39 GMT, Dhamoder Nalla <dhanalla at openjdk.org> wrote:

>>> @dhanalla Are you still making changes or is this ready to review? (if not ready just make it a draft ;) )
>> 
>> @eme64, This is ready for review.
>
>> @dhanalla Do you want us to continue reviewing? It is usually good to ping people again after making changes. Otherwise, we don't know if you are still working on it and we should wait.
> @eme64 yes, please review.

@dhanalla 
You now first perform the transformation, and then fail because there are too many nodes.
At that point, the compilation is basically in a bad state and cannot be finished.

Is there no alternative where we could check first if we would exceed the node limit, and then just avoid scalarizing such very large arrays, but continue with the compilation? It seems to be a shame to give up on escape analysis entirely, if it maybe was just a single array. But maybe @vnkozlov should say more about this, I only have a very rudimentary understanding of escape analysis.

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

PR Comment: https://git.openjdk.org/jdk/pull/20504#issuecomment-2821679100


More information about the hotspot-compiler-dev mailing list