> The test randomly excludes methods from C1/C2 compilation to stress test the > calling convention. While this has been proven to be extremely helpful for > testing the Virtual Threads implementation, it's also extremely slow - > especially when we are unlucky enough for multiple methods to be excluded > both from C1 and C2 compilation. On slow machines and in combination with > other VM flags, test execution time can easily exceed an hour. > > While we could now try to improve how these methods are selected, I don't > think it's worth it and we risk wasting significant machine hours just > running methods in the interpreter. I therefore disabled this test feature by > default. It can still be re-enabled for debugging / stress testing purposes > on demand. Long-term, we should add this capability to the VM. I filed > [JDK-8381751](https://bugs.openjdk.org/browse/JDK-8381751) for this. > > Thanks, > Tobias
Tobias Hartmann has updated the pull request incrementally with one additional commit since the last revision: Lower iteration count, moved to tier2 ------------- Changes: - all: https://git.openjdk.org/valhalla/pull/2306/files - new: https://git.openjdk.org/valhalla/pull/2306/files/77b38d9d..50d65c8c Webrevs: - full: https://webrevs.openjdk.org/?repo=valhalla&pr=2306&range=01 - incr: https://webrevs.openjdk.org/?repo=valhalla&pr=2306&range=00-01 Stats: 10 lines in 2 files changed: 1 ins; 1 del; 8 mod Patch: https://git.openjdk.org/valhalla/pull/2306.diff Fetch: git fetch https://git.openjdk.org/valhalla.git pull/2306/head:pull/2306 PR: https://git.openjdk.org/valhalla/pull/2306
