On Wed, 13 Apr 2022 12:24:46 GMT, Harold Seigel <hsei...@openjdk.org> wrote:
> Please review this small fix for JDK-8284642. The fix was tested by running > Mach5 tiers 1-2 on Linux, Mac OS, and Windows, and Mach5 tiers 3-5 on Linux > x64. Additionally, the modified test and the test in the bug report were run > locally. > > Thanks, Harold I don't this is the right "fix". There is a bit of a disconnect between the VM option and the system property. Although the default value of the VM flag is zero, that is ignored and the JDK side, not being directed otherwise, determines the max size. If you set the flag explicitly to zero that is passed through to the JDK and the max memory is actually set to zero. I think this is more a documentation issue - MaxDirectMemory has no affect unless explicitly set on the command-line. The default value of the flag is actually irrelevant. ------------- PR: https://git.openjdk.java.net/jdk/pull/8222
