On Thu, 16 Apr 2026 10:03:05 GMT, Roman Kennke <[email protected]> wrote:
>> This change makes UseCompactObjectHeaders default and switches the sense of >> the CDS archive naming. Now classes.jsa is with UseCompressedOops + >> UseCompactObjectHeaders (both default). Then theres: >> classes_nocoops.jsa >> classes_nocoh.jsa >> classes_nocoops_nocoh.jsa >> >> Hopefully someday we won't need the nocoh versions. >> >> Tested with tier1-8. With only one issue also fixed in this change, and no >> other new issues found. >> >> - [x] I confirm that I make this contribution in accordance with the >> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai). > > src/hotspot/share/cds/cdsConfig.cpp line 143: > >> 141: } >> 142: if (!UseCompactObjectHeaders) { >> 143: // Note that generation of xxx_coh.jsa variants require > > Also fix the comment? There is no _coh variant anymore. There is a _nocoh version and I didn't reverse the logic in autoconf/jdk-options.m4. So I fixed the comment to refer to nocoh instead. > src/hotspot/share/runtime/arguments.cpp line 3786: > >> 3784: // If UseCompactObjectHeaders is on the command line, turn on >> UseObjectMonitorTable. >> 3785: if (FLAG_IS_CMDLINE(UseCompactObjectHeaders)) { >> 3786: if (FLAG_IS_CMDLINE(UseObjectMonitorTable)) { > > I think this should not be inside if > (FLAG_IS_CMDLINE(UseCompactObjectHeaders)) { .. }. Even if UCOH is not > specified at all, we want to warn when somebody tries to turn UOMT off. Right? Maybe yes, I was thinking if someone just had -XX:-UseObjectMonitorTable without also specifying -XX:-UseCompactObjectHeaders, they should just get UCOH turned off, but this would only be for out testing, and not a real user, and we can fix our testing. > src/hotspot/share/runtime/arguments.cpp line 3792: > >> 3790: >> 3791: // If UseObjectMonitorTable is on the command line, turn off >> UseCompactObjectHeaders. >> 3792: } else if (FLAG_IS_CMDLINE(UseObjectMonitorTable)) { > > And then this branch would be unnecessary. That does simplify things a lot. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/30322#discussion_r3093033716 PR Review Comment: https://git.openjdk.org/jdk/pull/30322#discussion_r3093053742 PR Review Comment: https://git.openjdk.org/jdk/pull/30322#discussion_r3093090968
