8252583: Clean up unused phi-to-copy degradation mechanism

[robcasloz]

Remove unused notion of "PhiNode-to-copy degradation", where PhiNodes can be degraded to copies by setting their RegionNode to NULL. Remove corresponding PhiNode::is_copy() test, which always returned NULL (false). Assert that PhiNodes have an associated RegionNode in PhiNode::Ideal().

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• Change must be properly reviewed

Issue

• JDK-8252583: Clean up unused phi-to-copy degradation mechanism

Reviewers

• Tobias Hartmann (@TobiHartmann - Reviewer) ⚠️ Review applies to 91f30ba
• Vladimir Kozlov (@vnkozlov - Reviewer) ⚠️ Review applies to 91f30ba

Download

$ git fetch https://git.openjdk.java.net/jdk pull/275/head:pull/275
$ git checkout pull/275

[bridgekeeper]

👋 Welcome back robcasloz! 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.

[robcasloz]

/summary
Convert PhiNode::is_copy() into an actual, debug-only predicate. Replace calls
to is_copy() from non-debug code with explicit assertions. Remove dead loop in
debug-only MergeMemStream::match_memory().

[openjdk]

@robcasloz Setting summary to:

Convert PhiNode::is_copy() into an actual, debug-only predicate. Replace calls
to is_copy() from non-debug code with explicit assertions. Remove dead loop in
debug-only MergeMemStream::match_memory().

[openjdk]

@robcasloz The following label will be automatically applied to this pull request: hotspot-compiler.

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 (add|remove) "label" command.

[mlbridge]

Webrevs

• 04: Full - Incremental (b7656e1)
• 03: Full - Incremental (91f30ba)
• 02: Full - Incremental (844a31a)
• 01: Full - Incremental (5396298)
• 00: Full (6175d16)

[robcasloz]

Thanks Nils! I just updated the pull request.

[vnkozlov]

Sorry, I am not sure this is better than original code where assert is in one place - in PhiNode::is_copy().
The method is in .hpp file and is inlined - NULL check will be eliminated. It will only be executed in slowdebug build.
May be we should "bite the bullet" and remove this method at all - we don't hit the assert in years.
We can replace is_copy() check with assert in PhiNode::Ideal() - that should be enough to guarantee that we don't create Phi with NULL control edge.

[navyxliu]

Sorry, I am not sure this is better than original code where assert is in one place - in PhiNode::is_copy().
The method is in .hpp file and is inlined - NULL check will be eliminated. It will only be executed in slowdebug build.
May be we should "bite the bullet" and remove this method at all - we don't hit the assert in years.
We can replace is_copy() check with assert in PhiNode::Ideal() - that should be enough to guarantee that we don't create Phi with NULL control edge.

Another thing is that a node has a member called uint Node::is_Copy() const. PhiNode/Region nodes both have a member callled "is_copy()". Is it intentional? IMHO, it's not good in style.

[robcasloz]

Thanks Tobias and Xin for the reviews! I updated the PR for the record, even if we might end up moving the assertions somewhere else as per Vladimir's suggestion.

[robcasloz]

/summary
Remove unused notion of 'PhiNode-to-copy degradation', where PhiNodes can be
degraded to copies by setting their RegionNode to NULL. Remove corresponding
PhiNode::is_copy() test, which always returned NULL (false). Assert that
PhiNodes have an associated RegionNode in PhiNode::Ideal().

[openjdk]

@robcasloz Updating existing summary to:

Remove unused notion of 'PhiNode-to-copy degradation', where PhiNodes can be
degraded to copies by setting their RegionNode to NULL. Remove corresponding
PhiNode::is_copy() test, which always returned NULL (false). Assert that
PhiNodes have an associated RegionNode in PhiNode::Ideal().

[robcasloz]

Tested successfully on hs-tier1, hs-tier2, and hs-tier3.

[vnkozlov]

Good.

[openjdk]

@robcasloz This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for more details.

After integration, the commit message for the final commit will be:

8252583: Clean up unused phi-to-copy degradation mechanism

Remove unused notion of 'PhiNode-to-copy degradation', where PhiNodes can be
degraded to copies by setting their RegionNode to NULL. Remove corresponding
PhiNode::is_copy() test, which always returned NULL (false). Assert that
PhiNodes have an associated RegionNode in PhiNode::Ideal().

Reviewed-by: thartmann, kvn

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 39 new commits pushed to the master branch:

• 4440bda: 8219586: CodeHeap State Analytics processes dead nmethods
• 154b8cf: 8253014: OopStorage bulk release logs entries multiple times
• ba174af: 8253048: AArch64: When CallLeaf, no need to preserve callee-saved registers in caller
• 3320fc0: 8253372: [TESTBUG] update tests which require jvmti - hotspot
• f765a7f: 8252712: move doclint to jdk.javadoc module
• c21690b: 8253464: ARM32 Zero: atomic_copy64 is incorrect, breaking volatile stores
• 0bc01da: 8250635: MethodArityHistogram should use Compile_lock in favour of fancy checks
• 812b39f: 8252739: Deflater.setDictionary(byte[], int off, int len) ignores the starting offset for the dictionary
• 5f1d612: 8253397: Ensure LogTag types are sorted
• b8ea80a: 8253457: Remove unimplemented register stack functions
• ... and 29 more: https://git.openjdk.java.net/jdk/compare/2e30ff61b0394793655d6c6a593a5de8eb9c07b9...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

As you do not have Committer status in this project an existing Committer must agree to sponsor your change. Possible candidates are the reviewers of this PR (@neliasso, @TobiHartmann, @vnkozlov) but any other Committer may sponsor as well.

➡️ To flag this PR as ready for integration with the above commit message, type /integrate in a new comment. (Afterwards, your sponsor types /sponsor in a new comment to perform the integration).

[TobiHartmann]

Looks good to me.

[robcasloz]

Looks good to me.

Thanks Tobias!

[robcasloz]

/integrate

[openjdk]

@robcasloz
Your change (at version b7656e1) is now ready to be sponsored by a Committer.

[TobiHartmann]

/sponsor

[openjdk]

@TobiHartmann @robcasloz Since your change was applied there have been 39 commits pushed to the master branch:

• 4440bda: 8219586: CodeHeap State Analytics processes dead nmethods
• 154b8cf: 8253014: OopStorage bulk release logs entries multiple times
• ba174af: 8253048: AArch64: When CallLeaf, no need to preserve callee-saved registers in caller
• 3320fc0: 8253372: [TESTBUG] update tests which require jvmti - hotspot
• f765a7f: 8252712: move doclint to jdk.javadoc module
• c21690b: 8253464: ARM32 Zero: atomic_copy64 is incorrect, breaking volatile stores
• 0bc01da: 8250635: MethodArityHistogram should use Compile_lock in favour of fancy checks
• 812b39f: 8252739: Deflater.setDictionary(byte[], int off, int len) ignores the starting offset for the dictionary
• 5f1d612: 8253397: Ensure LogTag types are sorted
• b8ea80a: 8253457: Remove unimplemented register stack functions
• ... and 29 more: https://git.openjdk.java.net/jdk/compare/2e30ff61b0394793655d6c6a593a5de8eb9c07b9...master

Your commit was automatically rebased without conflicts.

Pushed as commit f3ea0d3.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.