RFR: 8337505: Footprint and startup regressions up to 20% in GUI apps

[Sergey Bylokhov]

On Mon, 28 Oct 2024 21:26:38 GMT, Phil Race <prr at openjdk.org> wrote:

> https://bugs.openjdk.org/browse/JDK-8338677 already improved things for this so that's good.
> 
> This fix adds to it lazy initialisation of VarHandles in StrikeCache at the cost of some extra code.
> Since these VarHandles get used more or less immediately on Linux this new fix won't further improve matters there
> But should help on Mac where they aren't usually needed at startup
> And Windows is somewhere in between.

Do we have a chance to reproduce the 20% numbers, and how it is comparable to the initial unsafe implementation? As a mentioned in the initial PR the new ffm implementation may add up to 100ms on startup(I tested it in the cmm). So taking that into account and 20% on some perftests(which one?) I assume migration back to unsafe could be an option as well.

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

PR Comment: https://git.openjdk.org/jdk/pull/21748#issuecomment-2588024176