RFR: 8351833: Unexpected increase in live nodes when splitting Phis through MergeMems in PhiNode::Ideal

Mon Apr 7 12:05:48 UTC 2025

On Mon, 31 Mar 2025 12:25:48 GMT, Daniel Lundén <dlunden at openjdk.org> wrote:

> This motivates that the dramatic live node increase seen in this issue is an edge case and is not a problem in general.

Thanks for the elaborate analysis, Daniel! I agree with your conclusion from the results you show, but if possible it would be good, for completeness, to study how many (if any) of those bailouts in Renaissance, SPECjvm, and SPECjbb are due to excessive IGVN node counts, at least for the "baseline" and "target" configurations.

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

PR Comment: https://git.openjdk.org/jdk/pull/24325#issuecomment-2783101024