On Fri, 18 Sep 2020 17:11:04 GMT, Phil Race <p...@openjdk.org> wrote:
> This test is being marked intermittent, although at the same time I am trying > to make it less likely to fail. > However since we have a known issue around NIO mmap'd files not being > directly unmappable, the > deletes the font system make may be stymied on Windows. So marking it > intermittent is probably for the best > One other thing is that I changed it so that the tmp files created are now of > different sizes so we can now > tell which createFont() call resulted in the font that can't be deleted. If > it is always the Type 1 fonts then > that will be good evidence it is mmap that is the problem. > We likely need to stop using mmap for this reason. Marked as reviewed by serb (Reviewer). ------------- PR: https://git.openjdk.java.net/jdk/pull/256
