8352595: Regression of JDK-8314999 in IR matching #24163

marc-chevalier · 2025-03-21T15:41:23Z

A lot of tests for the IR framework used ALLOC and friends as a check that would run on the Opto assembly by default, but can also run earlier, but that's no longer the case.

There were two kinds of tests to fix: the ones rather about ALLOC, where the used or expected compile phases have to change, and the tests where ALLOC were just a check that would run on opto assembly. For this, I tried to keep the spirit of the test using other regexes made for this stage.

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-8352595: Regression of JDK-8314999 in IR matching (Bug - P3)

Reviewers

Christian Hagedorn (@chhagedorn - Reviewer)
Tobias Hartmann (@TobiHartmann - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/24163/head:pull/24163
$ git checkout pull/24163

Update a local copy of the PR:
$ git checkout pull/24163
$ git pull https://git.openjdk.org/jdk.git pull/24163/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 24163

View PR using the GUI difftool:
$ git pr show -t 24163

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/24163.diff

Using Webrev

Link to Webrev Comment

bridgekeeper · 2025-03-21T15:42:28Z

👋 Welcome back marc-chevalier! 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 · 2025-03-21T15:42:42Z

@marc-chevalier 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 details.

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

8352595: Regression of JDK-8314999 in IR matching

Reviewed-by: chagedorn, thartmann

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 19 new commits pushed to the master branch:

48fac66: 8347406: [REDO] C1/C2 don't handle allocation failure properly during initialization (RuntimeStub::new_runtime_stub fatal crash)
99c8a6e: 8350463: AArch64: Add vector rearrange support for small lane count vectors
b2da0d3: 8352289: [macos] Review skipped tests in tools/jpackage/macosx/SigningPackage*
ba658a7: 8349522: AArch64: Add backend implementation for new unsigned and saturating vector operations
5625b43: 8350429: runtime/NMT/CheckForProperDetailStackTrace.java should only run for debug JVM
2c60fc5: 8352176: Automate setting up environment for mac signing tests
6e6a39d: 8347321: [ubsan] CGGlyphImages.m:553:30: runtime error: nan is outside the range of representable values of type 'unsigned long'
b84b292: 8352615: [Test] RISC-V: TestVectorizationMultiInvar.java fails on riscv64 without rvv support
a54445f: 8350609: Cleanup unknown unwind opcode (0xB) for windows
c87e1be: 8349582: APX NDD code generation for OpenJDK
... and 9 more: https://git.openjdk.org/jdk/compare/e23e0f85ef0f959a68adda0cff9e721ba2173ffc...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 (@chhagedorn, @TobiHartmann) 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).

openjdk · 2025-03-21T15:43:22Z

@marc-chevalier To determine the appropriate audience for reviewing this pull request, one or more labels corresponding to different subsystems will normally be applied automatically. However, no automatic labelling rule matches the changes in this pull request. In order to have an "RFR" email sent to the correct mailing list, you will need to add one or more applicable labels manually using the /label pull request command.

Applicable Labels

build
client
compiler
core-libs
graal
hotspot
hotspot-compiler
hotspot-gc
hotspot-jfr
hotspot-runtime
i18n
ide-support
javadoc
jdk
jmx
kulla
net
nio
security
serviceability
shenandoah

chhagedorn

Thanks for addressing this. I have some comments.

chhagedorn · 2025-03-24T07:48:59Z

test/hotspot/jtreg/testlibrary_tests/ir_framework/tests/TestPhaseIRMatching.java

    @ExpectedFailure(ruleId = 4, phase = CompilePhase.PRINT_IDEAL, failOn = 1, counts = {1, 4})

    @IR(counts = {IRNode.ALLOC, "2", IRNode.ALLOC_OF, "Object", "1"})
-    @ExpectedFailure(ruleId = 5, phase = CompilePhase.PRINT_OPTO_ASSEMBLY, counts = 1)
+    @ExpectedFailure(ruleId = 5, phase = CompilePhase.BEFORE_MACRO_EXPANSION, counts = 1)
    public void defaultOnBoth() {


I think for this test you should add a matching on some PrintOptoAssembly as well since it tries to verify ideal and opto matching.

chhagedorn · 2025-03-24T08:01:26Z

test/hotspot/jtreg/testlibrary_tests/ir_framework/tests/flag/TestCompilePhaseCollector.java

+        @IR(counts = {IRNode.CBZ_HI, "> 1"})
        public void test4() {}

-
        @Test
-        @IR(failOn = IRNode.FIELD_ACCESS)
-        @IR(counts = {IRNode.CHECKCAST_ARRAY, "2"})
+        @IR(failOn = {IRNode.CBNZW_HI})
+        @IR(counts = {IRNode.CBZ_LS, "2", IRNode.CBZW_LS, "> 1"})
        public void test5() {}

        @Test
-        @IR(failOn = {IRNode.CHECKCAST_ARRAYCOPY, IRNode.CHECKCAST_ARRAY_OF, "Foo"})
-        @IR(counts = {IRNode.ALLOC, "2", IRNode.ALLOC_ARRAY_OF, "Foo", "> 1"})
+        @IR(failOn = {IRNode.CBNZW_HI})
+        @IR(counts = {IRNode.CBZW_LS, "> 1"})


I've just noticed that these AArch64 specific IRNode entries do not have a check that they are only supported on AArch64. In the past, I've added some platform checks for IRNode entries only being supported on certain platforms:

jdk/test/hotspot/jtreg/compiler/lib/ir_framework/IRNode.java

Lines 2905 to 2925 in e23e0f8

public static void checkIRNodeSupported(String node) throws CheckedTestFrameworkException {

switch (node) {

case INTRINSIC_OR_TYPE_CHECKED_INLINING_TRAP -> {

if (!WhiteBox.getWhiteBox().isJVMCISupportedByGC()) {

throw new CheckedTestFrameworkException("INTRINSIC_OR_TYPE_CHECKED_INLINING_TRAP is unsupported " +

"in builds without JVMCI.");

}

}

case CHECKCAST_ARRAYCOPY -> {

if (Platform.isS390x()) {

throw new CheckedTestFrameworkException("CHECKCAST_ARRAYCOPY is unsupported on s390.");

}

}

case IS_FINITE_D, IS_FINITE_F -> {

if (!Platform.isRISCV64()) {

throw new CheckedTestFrameworkException("IS_FINITE_* is only supported on riscv64.");

}

}

// default: do nothing -> IR node is supported and can be used by the user.

}

}

We should probably add a similar check for these CB* entries. But that's something that can be done separately.

For this test here, we only collect compile phases and it does not matter on what platforms these IRNode entries are eventually used. So, no action to be done in your changes here.

chhagedorn · 2025-03-24T08:34:58Z

test/hotspot/jtreg/testlibrary_tests/ir_framework/tests/flag/TestCompilePhaseCollector.java

@@ -113,12 +113,12 @@ public void testMixedPhases() {
        assertContainsOnly(methodToCompilePhases, testClass, "mix7", PHASEIDEALLOOP1, PHASEIDEALLOOP2, FINAL_CODE,
                           OPTIMIZE_FINISHED, PRINT_IDEAL);
        assertContainsOnly(methodToCompilePhases, testClass, "mix8", PHASEIDEALLOOP1, PHASEIDEALLOOP2, FINAL_CODE,
-                           OPTIMIZE_FINISHED, PRINT_IDEAL, PRINT_OPTO_ASSEMBLY);
+                           OPTIMIZE_FINISHED, PRINT_IDEAL);


Note that the tests on this file only collect compile phases and actually do not perform any IR matching. So, the IR rules do not need work. This means that for all tests where we use the default compile phase of ALLOC (which is PRINT_OPTO_ASSEMBLY), we can replace ALLOC with just something else that defaults on PRINT_OPTO_ASSEMBLY.

This just means that we should not replace PRINT_OPTO_ASSEMBLY with BEFORE_MACRO_EXPANSION here but instead use a different IRNode in the IR rules at these tests that default on PRINT_OPTO_ASSEMBLY. I've made some comments further down where this applies. The other changes in the file look fine.

test/hotspot/jtreg/testlibrary_tests/ir_framework/tests/flag/TestCompilePhaseCollector.java

marc-chevalier · 2025-03-24T09:38:50Z

/label add hotspot-compiler

openjdk · 2025-03-24T09:39:36Z

@marc-chevalier
The hotspot-compiler label was successfully added.

mlbridge · 2025-03-24T09:42:27Z

Webrevs

marc-chevalier · 2025-03-24T14:05:03Z

Yes, I indeed misunderstood which tests were specifically for ALLOC, and which were about OptoAssembly, and ALLOC is just a mean. The new diff makes more sense, indeed.

chhagedorn

Thanks for the update, looks good!

TobiHartmann

Looks good to me too. Ship it!

marc-chevalier · 2025-03-25T07:26:27Z

/integrate

Let's indeed!
Thanks!

openjdk · 2025-03-25T07:26:51Z

@marc-chevalier
Your change (at version c7c0cfa) is now ready to be sponsored by a Committer.

chhagedorn · 2025-03-25T07:27:23Z

/sponsor

openjdk · 2025-03-25T07:28:09Z

Going to push as commit c94bc74.
Since your change was applied there have been 19 commits pushed to the master branch:

48fac66: 8347406: [REDO] C1/C2 don't handle allocation failure properly during initialization (RuntimeStub::new_runtime_stub fatal crash)
99c8a6e: 8350463: AArch64: Add vector rearrange support for small lane count vectors
b2da0d3: 8352289: [macos] Review skipped tests in tools/jpackage/macosx/SigningPackage*
ba658a7: 8349522: AArch64: Add backend implementation for new unsigned and saturating vector operations
5625b43: 8350429: runtime/NMT/CheckForProperDetailStackTrace.java should only run for debug JVM
2c60fc5: 8352176: Automate setting up environment for mac signing tests
6e6a39d: 8347321: [ubsan] CGGlyphImages.m:553:30: runtime error: nan is outside the range of representable values of type 'unsigned long'
b84b292: 8352615: [Test] RISC-V: TestVectorizationMultiInvar.java fails on riscv64 without rvv support
a54445f: 8350609: Cleanup unknown unwind opcode (0xB) for windows
c87e1be: 8349582: APX NDD code generation for OpenJDK
... and 9 more: https://git.openjdk.org/jdk/compare/e23e0f85ef0f959a68adda0cff9e721ba2173ffc...master

Your commit was automatically rebased without conflicts.

openjdk · 2025-03-25T07:28:16Z

@chhagedorn @marc-chevalier Pushed as commit c94bc74.

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

Fix TestBadFormat

4c7ed17

marc-chevalier added 2 commits March 21, 2025 19:41

Fix TestPhaseIRMatching.java

ee3cc6f

Fix TestCompilePhaseCollector.java

17e11d4

marc-chevalier force-pushed the fix/fix-IRframework-test branch from 98f574b to 17e11d4 Compare March 21, 2025 18:43

marc-chevalier marked this pull request as ready for review March 24, 2025 07:21

openjdk bot added the rfr Pull request is ready for review label Mar 24, 2025

chhagedorn suggested changes Mar 24, 2025

View reviewed changes

openjdk bot added the hotspot-compiler [email protected] label Mar 24, 2025

marc-chevalier added 2 commits March 24, 2025 14:57

wip

98febc9

wip

b125df8

Merge branch 'master' into fix/fix-IRframework-test

c7c0cfa

chhagedorn approved these changes Mar 25, 2025

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Mar 25, 2025

TobiHartmann approved these changes Mar 25, 2025

View reviewed changes

openjdk bot added the sponsor Pull request is ready to be sponsored label Mar 25, 2025

openjdk bot added the integrated Pull request has been integrated label Mar 25, 2025

openjdk bot closed this Mar 25, 2025

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Mar 25, 2025

openjdk bot removed the sponsor Pull request is ready to be sponsored label Mar 25, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8352595: Regression of JDK-8314999 in IR matching #24163

8352595: Regression of JDK-8314999 in IR matching #24163

marc-chevalier commented Mar 21, 2025 •

edited by openjdk bot

Loading

bridgekeeper bot commented Mar 21, 2025

openjdk bot commented Mar 21, 2025 •

edited

Loading

openjdk bot commented Mar 21, 2025

chhagedorn left a comment

chhagedorn Mar 24, 2025 •

edited

Loading

chhagedorn Mar 24, 2025 •

edited

Loading

chhagedorn Mar 24, 2025

marc-chevalier commented Mar 24, 2025

openjdk bot commented Mar 24, 2025

mlbridge bot commented Mar 24, 2025 •

edited

Loading

marc-chevalier commented Mar 24, 2025

chhagedorn left a comment

TobiHartmann left a comment

marc-chevalier commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

chhagedorn commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

	public static void checkIRNodeSupported(String node) throws CheckedTestFrameworkException {
	switch (node) {
	case INTRINSIC_OR_TYPE_CHECKED_INLINING_TRAP -> {
	if (!WhiteBox.getWhiteBox().isJVMCISupportedByGC()) {
	throw new CheckedTestFrameworkException("INTRINSIC_OR_TYPE_CHECKED_INLINING_TRAP is unsupported " +
	"in builds without JVMCI.");
	}
	}
	case CHECKCAST_ARRAYCOPY -> {
	if (Platform.isS390x()) {
	throw new CheckedTestFrameworkException("CHECKCAST_ARRAYCOPY is unsupported on s390.");
	}
	}
	case IS_FINITE_D, IS_FINITE_F -> {
	if (!Platform.isRISCV64()) {
	throw new CheckedTestFrameworkException("IS_FINITE_* is only supported on riscv64.");
	}
	}
	// default: do nothing -> IR node is supported and can be used by the user.
	}
	}

8352595: Regression of JDK-8314999 in IR matching #24163

8352595: Regression of JDK-8314999 in IR matching #24163

Conversation

marc-chevalier commented Mar 21, 2025 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Mar 21, 2025

openjdk bot commented Mar 21, 2025 • edited Loading

openjdk bot commented Mar 21, 2025

chhagedorn left a comment

Choose a reason for hiding this comment

chhagedorn Mar 24, 2025 • edited Loading

Choose a reason for hiding this comment

chhagedorn Mar 24, 2025 • edited Loading

Choose a reason for hiding this comment

chhagedorn Mar 24, 2025

Choose a reason for hiding this comment

marc-chevalier commented Mar 24, 2025

openjdk bot commented Mar 24, 2025

mlbridge bot commented Mar 24, 2025 • edited Loading

Webrevs

marc-chevalier commented Mar 24, 2025

chhagedorn left a comment

Choose a reason for hiding this comment

TobiHartmann left a comment

Choose a reason for hiding this comment

marc-chevalier commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

chhagedorn commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

openjdk bot commented Mar 25, 2025

marc-chevalier commented Mar 21, 2025 •

edited by openjdk bot

Loading

openjdk bot commented Mar 21, 2025 •

edited

Loading

chhagedorn Mar 24, 2025 •

edited

Loading

chhagedorn Mar 24, 2025 •

edited

Loading

mlbridge bot commented Mar 24, 2025 •

edited

Loading