RFR: JDK-8304954: SegmentedCodeCache fails when using large pages [v4]

[Thomas Stuefe]

On Tue, 25 Jul 2023 12:29:54 GMT, Damon Fenacci <dfenacci at openjdk.org> wrote:

>> I may be wrong too. Looking closer, I think it aligns up. So, I think you can forget the first of the two points. Sorry for the confusion.
>
> I've changed "reserve" into the more generic "use" in the message.
> To check for the actual large page size I've used the same `page_size` method giving 1 as minimum number of pages (`default_large_page_size` seems to be specific for linux)

Sorry to be a stickler for precise warning messages. Proposal: "Code cache size too small for " PROPER_FMT " pages." - the rest is fine.

If we really wanted, we could also print out the code cache size needed for that page size, but its a bit pointless with 1G pages since that would mean an 8G code cache, which is past max codecache size.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/14903#discussion_r1274705380