Remove forgotten dead code from #13 and #14

[franferrax]

Remove forgotten dead code from #13 and #14

#13 isn't a perfect revert of 0af22dc limited to SSLContextImpl.java and SunJSSE.java, since it doesn't remove the SharedSecrets import. This was already in this way in rh2020290-support_tls_1_3_in_fips.v1.patch, I'm now realizing this when doing the OpenJDK 11 backport (rh-openjdk/jdk11u#7) and retrying the same approach in OpenJDK 17:

# Revert #13
git show 0bd5ca9ccc5890b009b927b541424d0aacbb0463 | git apply -R
# Redo #13 by reverting 0af22dc in SSLContextImpl.java and SunJSSE.java
git show 0af22dcc391ff38f65ff7d7cfcc2f933e29e9126 |
  git apply -R --include=src/java.base/share/classes/sun/security/ssl/*

---

In #14, I forgot to delete the DHKF and DHKFLock static variables from FIPSKeyImporter, which are no longer used, see this #14 comment.

[gnu-andrew]

Looks fine to me.

[gnu-andrew]

Any idea why we still have a test failure on Linux? Seems unrelated to this change.

[franferrax]

Any idea why we still have a test failure on Linux? Seems unrelated to this change.

Yes, the Linux failure is the currently known one (-Djava.security.debug=properties makes SimpleFormatterFormat fail, see the 1ˢᵗ comment of #11).

Windows MineField.sh and wcMineField.sh failures were already present in bb46af0, but I haven't analyzed them:

• jdk/test/langtools/tools/javac/Paths/MineField.sh

  Line 219 in c1a8749

  Success env CLASSPATH="GooZip/Lib.zip${PS}BadJar/Lib.jar" "$javac" ${TESTTOOLVMOPTS} Main.java

  • Fails in Windows x64 (langtools/tier1) > Run tests > 16646 (line 17714 of the raw logs)

• jdk/test/langtools/tools/javac/Paths/MineField.sh

  Line 221 in c1a8749

  Success env CLASSPATH="${PS}GooZip/Lib.zip${PS}BadJar/Lib.jar" "$java" ${TESTVMOPTS} Main

  • Fails in Windows x64 (langtools/tier1) > Run tests > 16652 (line 17720 of the raw logs)

• jdk/test/langtools/tools/javac/Paths/wcMineField.sh

  Line 196 in c1a8749

  Success env CLASSPATH="GooJar/SubDir/*" "$javac" ${TESTTOOLVMOPTS} Main1.java

  • Fails in Windows x64 (langtools/tier1) > Run tests > 17008 (line 18076 of the raw logs)

• jdk/test/langtools/tools/javac/Paths/wcMineField.sh

  Line 197 in c1a8749

  Success env CLASSPATH="GooJar/SubDir/*${PS}." "$java" ${TESTVMOPTS} Main1

  • Fails in Windows x64 (langtools/tier1) > Run tests > 17011 (line 18079 of the raw logs)

[gnu-andrew]

Any idea why we still have a test failure on Linux? Seems unrelated to this change.

Yes, the Linux failure is the currently known one (-Djava.security.debug=properties makes SimpleFormatterFormat fail, see the 1ˢᵗ comment of #11).

Ah, for some reason, I thought we'd fixed that. I'll open a PR that removes it. We can always add the line temporarily if we need the debugging info for a specific case.

Windows MineField.sh and wcMineField.sh failures were already present in bb46af0, but I haven't analyzed them:

As both were present before, I don't see them as a blocker for this. I don't see this failure in upstream 17u so I'll sync with the latest release and see if that fixes it.

We may as well get this change in along with the NSS DB support.

[gnu-andrew]

I've removed the -Djava.security.debug=properties as part of the 17.0.5 merge. That refactored all the GitHub files so it didn't seem worth trying to get the argument working again in a new place, just to remove it again.

[gnu-andrew]

From the 17.0.5 run, all the tests seem to be passing: https://github.com/gnu-andrew/jdk/actions/runs/3528808451