On Mon, 14 Oct 2024 13:08:32 GMT, Aleksey Shipilev <sh...@openjdk.org> wrote:
> This implicitly fixes x86-32 builds that fail at runtime, because late > barrier expansion is not implemented for them. I think about this check as > something that we should have done as part of original JDK-8334060. > > The first version was summarily disabling G1 on platforms not explicitly > listed in arches. Unfortunately, that affects Zero, while the problem we are > working out is G1+C2 support. The second revision of this PR checks for > `compiler2` and x86-32 explicitly, and disables G1 only then. > > Additional testing: > - [x] Linux x86_32 server fastdebug, `make bootcycle-images` now passes > - [x] A matrix of (server, zero) builds on different platforms Even better this way. Should also be reviewed by a build expert. ------------- Marked as reviewed by mdoerr (Reviewer). PR Review: https://git.openjdk.org/jdk/pull/21493#pullrequestreview-2366853428
