8366182: Some PKCS11Tests are being skipped when they shouldn't

[mpdonova]

This PR updates PKCS11 tests to better handle NSS version numbers. The previous code treated the version numbers as double values and used comparison operators. The problem is that it incorrectly treats 3.111 as between 3.11 and 3.12. This update parses and compares the major and minor version numbers separately.

---

Progress

• Change must be properly reviewed (1 review required, with at least 1 Reviewer)
• Change must not contain extraneous whitespace
• Commit message must refer to an issue

Issue

• JDK-8366182: Some PKCS11Tests are being skipped when they shouldn't (Bug - P4)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/27095/head:pull/27095
$ git checkout pull/27095

Update a local copy of the PR:
$ git checkout pull/27095
$ git pull https://git.openjdk.org/jdk.git pull/27095/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 27095

View PR using the GUI difftool:
$ git pr show -t 27095

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/27095.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back mdonovan! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

[openjdk]

❗ This change is not yet ready to be integrated.
See the Progress checklist in the description for automated requirements.

[openjdk]

@mpdonova The following label will be automatically applied to this pull request:

• security

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

[mlbridge]

Webrevs

• 01: Full - Incremental (cd250fab)
• 00: Full (c869d478)

[myankelev]

IMO the split between major version and minor is a bit hard to read. Wouldn't it be easier to just get a major.minor version entirely with something like:

String[] splitParts = ver.split("//.");
Double.parseDouble(splitParts.length > 1
        ? splitParts[0] + "." + splitParts[1] 
        : splitParts[0]);

This way it will always take a full double value (1.13 will not be the same as 1.1, as it's now as far as I can see) and would be a bit easier to understand

The checking for only major version could be either doubleVersion<4 && doubleVersion>=3 or even cleaner, using floor function Math.floor(doubleVersion)

And the same for the other files.

What do you think?

[mpdonova]

I may be misunderstanding your comment but it looks like you're using a single, double value to compare versions. The original code did that but it doesn't work in all cases. The problem occurs when checking a range such as

version >= 3.11 && version < 3.12

If the version number is 3.111, then the comparison is true and tests are skipped, even though version 3.111 is "greater" than 3.12. So this code creates two doubles: major and minor to do separate comparisons of the values.

To further complicate things, NSS also has versions of the form x.y.z. The original code combined the 'y' and 'z' components to go from 3.11.9 to 3.119. My change would create major version 3.0 and minor version 11.9.

[myankelev]

Ah, you're right, I didn't think of this case! This approach will be easier then, I agree.

Still, ver.substring(idx+1) may cause errors in cases like this NSS 3.15.3.1 release notes. It will try to convert 15.3.1 to a double. Do you think changing to something like this would be worth it?

String verSubstring = ver.substring(idx+1);
String[] splitParts = verSubstring.split("//.");
Double minor = Double.parseDouble(splitParts.length > 1
                              ? splitParts[0] + "." + splitParts[1] 
                              : splitParts[0]);

[mpdonova]

That's a good point about 4-component version numbers. Frankly, I was hoping that NSS never did that. I refactored the version parsing code to create a Version record with major, minor, and patch versions. This way, individual tests don't have to do the String parsing. If we end up with a test that has to be skipped due to that 4th component, we can update the parsing accordingly. Alternately, I can add it since I'm here now... what is that fourth component called?

[myankelev]

Thank you for the changes!
I'm not sure if it's even needed to be honest, it would be an overkill imo. I can't find the exact naming on the nss website, but I think the common one would be major.minor.patch.hotfix