8378985: serviceability/sa/TestJhsdbJstackMixedWithXComp.java failed if sender frame is return barrier of Continuation by YaSuenag · Pull Request #30107 · openjdk/jdk

YaSuenag · 2026-03-06T04:25:24Z

We saw the failure in serviceability/sa/TestJhsdbJstackMixedWithXComp.java on Valhalla repo. jhsdb jstack --mixed could not unwind continuation call frames as following:

                   * LingeredAppWithVirtualThread.run() bci:15 line:69 (Interpreted frame)
                   * java.lang.Thread.runWith(java.lang.Object, java.lang.Runnable) bci:5 line:1540 (Compiled frame [deoptimized]; information may be imprecise)
                   * java.lang.VirtualThread.run(java.lang.Runnable) bci:62 line:472 (Compiled frame [deoptimized]; information may be imprecise)
0x00007fdad773cf98 <StubRoutines (continuation stubs)>
0xfefefefefefefefe ????????

I found that frame::sender_for_compiled_frame() in frame_x86.inline.hpp has a special case if sender PC has return barrier entry, but SA does not handle it.

This is not only a problem on Valhalla. Same problem exists on JDK. So I want to fix on JDK.
This PR passed serviceability/sa tests on Linux, and also TestJhsdbJstackMixedWithXComp.java on Valhalla passed 100 times.

This PR is assembled by following commits:

Follows continuation-related code in HotSpot, and use it on AMD64 SA code
- 4af559e
Fix for AArch64
- 2cff0fe
Fix for RISC-V
- fdef038
Fix for PPC64
- 5d9774d

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-8378985: serviceability/sa/TestJhsdbJstackMixedWithXComp.java failed if sender frame is return barrier of Continuation (Bug - P4)

Reviewers

Chris Plummer (@plummercj - Reviewer)
Martin Doerr (@TheRealMDoerr - Reviewer)
Fei Yang (@RealFYang - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/30107/head:pull/30107
$ git checkout pull/30107

Update a local copy of the PR:
$ git checkout pull/30107
$ git pull https://git.openjdk.org/jdk.git pull/30107/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 30107

View PR using the GUI difftool:
$ git pr show -t 30107

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/30107.diff

Using Webrev

Link to Webrev Comment

bridgekeeper · 2026-03-06T04:27:16Z

👋 Welcome back ysuenaga! 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 · 2026-03-06T04:29:01Z

@YaSuenag 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:

8378985: serviceability/sa/TestJhsdbJstackMixedWithXComp.java failed if sender frame is return barrier of Continuation

Reviewed-by: cjplummer, mdoerr, fyang

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 271 new commits pushed to the master branch:

4408e1c: 8380580: Enable TestCombineAddPWithConstantOffsets.java IR tests for RISC-V
5688595: 8380408: GenShen: Fix regression with vmTestbase/gc/memory/Churn/Churn2/TestDescription.java
e8ce930: 8379020: GenShen: Promote all objects when whitebox full GC is requested
... and 268 more: https://git.openjdk.org/jdk/compare/f266079d26ffc1541e5d3146258f7d39188d1a84...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.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

openjdk · 2026-03-06T04:29:57Z

@YaSuenag The following labels will be automatically applied to this pull request:

hotspot-runtime
serviceability

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

mlbridge · 2026-03-06T09:47:57Z

Webrevs

YaSuenag · 2026-03-06T09:48:10Z

Could someone can help for testing on RISC-V and PPC64?

This PR includes changes for them. It affects serviceability/sa tests especially TestJhsdbJstackWithVirtualThread.java. I tested both AMD64 and AArch64 Linux, but I cannot test RISC-V and PPC64 because I do not have them.

plummercj

Overall this looks good and passes my testing. I can only test that it builds on riscv64. I can't run tests on that platform, and can't do builds or testing on s390 or PPC.

plummercj · 2026-03-06T20:30:32Z

  public abstract Address getFP();

+  public void setSP(Address newSP) {
+    throw new UnsupportedPlatformException("Continuation is not yet implemented.");


Although I understand that this method is not overridden on platforms that don't support continuations (s390), the relationship between a method with a generic name like "setSP" and a specific error messages about continuations is not obvious. If setSP() only only useful in the context of continuations, I'd suggest naming it something like setContinuationSP(), or you can leave it as setSP() but get rid of the reference to continuations in the error message.

We can make it to abstract method, so I did it in new commit.

plummercj · 2026-03-07T00:13:12Z

+    public static boolean isSPInContinuation(ContinuationEntry entry, Address sp) {
+        return entry.getEntrySP().greaterThan(sp);
+    }


This could be made private.

It is public member in HotSpot, so I aligned with it. However it does not need to be public in SA.
Should we it to be private?

It's ok to keep public then.

plummercj · 2026-03-07T00:13:36Z

+        return entry.getEntrySP().greaterThan(sp);
+    }
+
+    public static ContinuationEntry getContinuationEntryForSP(JavaThread thread, Address sp) {


This also could be made private.

It is public member in HotSpot, so I aligned with it. However it does not need to be public in SA.
Should we it to be private?

It's ok to keep public then.

plummercj · 2026-03-07T00:17:50Z

+    // HotSpot has following code in frame::sender_raw() at frame_x86.inline.hpp, however
+    // in_cont() should be false.
+    //
+    //   if (map->in_cont()) { // already in an h-stack
+    //     return map->stack_chunk()->sender(*this, map);
+    //   }
+    //
+    // in_cont() returns true if _chunk() is not null.
+    //
+    // frame::next_frame() creates RegisterMap instance with 4 arguments.
+    // It sets RegisterMap::WalkContinuation::skip to final argument (walk_count),
+    // then _chunk would not be initialized following code in c'tor of RegisterMap.
+    //
+    //   if (walk_cont == WalkContinuation::include && thread != nullptr && thread->last_continuation() != nullptr) {
+    //     _chunk = stackChunkHandle(Thread::current()->handle_area()->allocate_null_handle(), true /* dummy */);
+    //   }
+


I don't understand how this comment is relevant to the code below. Why is this here?

SA should follow HotSpot implementation, so that code should be included basically. However it is not needed in this case. Thus I left comment the reason.

I still don't follow. Are you saying that map->in_cont() will always return true so you don't need the walk_cont related code?

I just want to say we don't need to copy return map->stack_chunk()->sender(*this, map) to SA from HotSpot because map->in_cont() should be false in use case of SA.

in_cont() returns true if _chunk is not NULL, but it would not happen in the use case of SA because RegisterMap would not be initialized with WalkContinuation::include.

YaSuenag · 2026-03-07T04:08:57Z

can't do builds or testing on s390 or PPC.

We do not need to consider s390 because SA does not have s390 implementation.

RealFYang · 2026-03-08T01:49:29Z

        return senderForUpcallStub(map, (UpcallStub)cb);
-      } else if (cb.isContinuationStub()) {
-        return senderForContinuationStub(map, cb);
      } else {


Hi,
I tried this on linux-riscv64 platform and I witnessed test failure when running serviceability/sa/TestJhsdbJstackWithVirtualThread.java:

java.lang.RuntimeException: 'must have non-zero frame size' found in stdout at jdk.test.lib.process.OutputAnalyzer.shouldNotContain(OutputAnalyzer.java:299) at TestJhsdbJstackWithVirtualThread.runJstack(TestJhsdbJstackWithVirtualThread.java:63) at TestJhsdbJstackWithVirtualThread.main(TestJhsdbJstackWithVirtualThread.java:74) at java.base/jdk.internal.reflect.DirectMethodHandleAccessor.invoke(DirectMethodHandleAccessor.java:104) at java.base/java.lang.reflect.Method.invoke(Method.java:565) at com.sun.javatest.regtest.agent.MainActionHelper$AgentVMRunnable.run(MainActionHelper.java:335) at java.base/java.lang.Thread.run(Thread.java:1527) JavaTest Message: Test threw exception: java.lang.RuntimeException JavaTest Message: shutting down test

I guess you might want to add following add-on change. I see aarch64 and amd64 has a similar frame size check before invoking senderForCompiledFrame method.

diff --git a/src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/riscv64/RISCV64Frame.java b/src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/riscv64/RISCV64Frame.java index 67b4314a3c7..a35c0735979 100644 --- a/src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/riscv64/RISCV64Frame.java +++ b/src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/riscv64/RISCV64Frame.java @@ -269,7 +269,7 @@ public Frame sender(RegisterMap regMap, CodeBlob cb) { if (cb != null) { if (cb.isUpcallStub()) { return senderForUpcallStub(map, (UpcallStub)cb); - } else { + } else if (cb.getFrameSize() > 0) { return senderForCompiledFrame(map, cb); } }

Thanks a lot! You are right.
I applied that change to both RISC-V and PPC64. Could you check again?

My local hotspot_serviceability and jdk_svc tests on linux-riscv64 are passing using the latest version. Thanks for the update.

…iledFrame()

dholmes-ora · 2026-03-09T06:09:05Z

Just a drive-by comment, but if this is being fixed in mainline then the [lworld] should be removed from the JBS title. Though IIUC no failure has been seen in mainline - is that correct?

YaSuenag · 2026-03-09T07:05:45Z

Fixed title both this PR and JBS.

Though IIUC no failure has been seen in mainline - is that correct?

Yes, however mainline has potential bug, so I want to fix mainline first.
I'm not sure, but this bug may be more likely to occur in Valhalla than in the mainline.

TheRealMDoerr · 2026-03-09T09:35:31Z

Could someone can help for testing on RISC-V and PPC64?

This PR includes changes for them. It affects serviceability/sa tests especially TestJhsdbJstackWithVirtualThread.java. I tested both AMD64 and AArch64 Linux, but I cannot test RISC-V and PPC64 because I do not have them.

Tier1-4 have passed on PPC64. Thanks for implementing it for all SA platforms!

…d64/AMD64Frame.java Co-authored-by: Chris Plummer <chris.plummer@oracle.com>

YaSuenag · 2026-03-23T07:56:25Z

PING: Can I get Reviewers? This PR has passed SA tests on all of supported platforms.

TheRealMDoerr

PPC64 code looks correct to me. Thanks!

TheRealMDoerr · 2026-03-23T10:18:12Z

+      // We assume WalkContinuation is "WalkContinuation::skip".
+      // It is same with c'tor arguments of RegisterMap in frame::next_frame().
+      //
+      // HotSpot code in cpu/riscv/frame_riscv.inline.hpp:


Should better refer to cpu/ppc/frame_ppc.inline.hpp

Thanks! Fixed it in new commit.

TheRealMDoerr · 2026-03-23T10:19:42Z

+
+    @Override
+    public Frame toFrame() {
+        return new AARCH64Frame(getEntrySP(), getEntrySP(), getEntryFP(), getEntryPC());


Do we need the same code replicated for all platforms? (May make senes. Not sure.)

This structure comes from HotSpot.
to_frame() is defined in continuationEntry_<arch>.inline.hpp. So I think it is better to follow it because most of source file in SA follows HotSpot manner.

TheRealMDoerr

PPC64 parts looks good. Only a partial review. Thanks!

plummercj

Looks good.

RealFYang

The RISC-V part of the change looks fine. hotspot_serviceability and jdk_svc still test good on this platform.

YaSuenag · 2026-03-26T02:15:17Z

/integrate

openjdk · 2026-03-26T02:16:04Z

Going to push as commit 69deec2.
Since your change was applied there have been 293 commits pushed to the master branch:

ceb109f: 8380081: jpackage: Incorrect descriptions of win-specific options
a55656d: 8380789: RISC-V: TestOpaqueConstantBoolNodes.java fails when running without RVV
3588fb7: 8380825: AOT tests should run in two step training mode by default
... and 290 more: https://git.openjdk.org/jdk/compare/f266079d26ffc1541e5d3146258f7d39188d1a84...master

Your commit was automatically rebased without conflicts.

openjdk · 2026-03-26T02:16:13Z

@YaSuenag Pushed as commit 69deec2.

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

YaSuenag added 2 commits March 6, 2026 13:13

Implement Continuation.continuationBottomSender

4af559e

Fix for AArch64

2cff0fe

openjdk bot added serviceability serviceability-dev@openjdk.org hotspot-runtime hotspot-runtime-dev@openjdk.org labels Mar 6, 2026

YaSuenag added 3 commits March 6, 2026 15:16

Fix for RISC-V

fdef038

Fix for PPC64

5d9774d

Use create() to instantiate parent instance

50c22d3

YaSuenag marked this pull request as ready for review March 6, 2026 09:41

openjdk bot added the rfr Pull request is ready for review label Mar 6, 2026

plummercj reviewed Mar 7, 2026

View reviewed changes

Make setSP() to abstract method

875a7bd

RealFYang reviewed Mar 8, 2026

View reviewed changes

Frame size of CodeBlob should be checked before calling senderForComp…

4bb73f6

…iledFrame()

YaSuenag changed the title ~~8378985: [lworld] serviceability/sa/TestJhsdbJstackMixedWithXComp.java#xcomp fails with just Test failed!~~ 8378985: serviceability/sa/TestJhsdbJstackMixedWithXComp.java failed if sender frame is return barrier of Continuation Mar 9, 2026

plummercj reviewed Mar 10, 2026

View reviewed changes

Comment thread src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/amd64/AMD64Frame.java Outdated

Comment thread src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/amd64/AMD64Frame.java Outdated

YaSuenag and others added 2 commits March 10, 2026 13:01

Update src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/am…

f1e5e21

…d64/AMD64Frame.java Co-authored-by: Chris Plummer <chris.plummer@oracle.com>

Update src/jdk.hotspot.agent/share/classes/sun/jvm/hotspot/runtime/am…

ffe03c7

…d64/AMD64Frame.java Co-authored-by: Chris Plummer <chris.plummer@oracle.com>

TheRealMDoerr reviewed Mar 23, 2026

View reviewed changes

Fix comments

eedb87d

TheRealMDoerr approved these changes Mar 23, 2026

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Mar 23, 2026

plummercj approved these changes Mar 25, 2026

View reviewed changes

RealFYang approved these changes Mar 25, 2026

View reviewed changes

openjdk bot added the integrated Pull request has been integrated label Mar 26, 2026

openjdk bot closed this Mar 26, 2026

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Mar 26, 2026

YaSuenag deleted the JDK-8378985 branch March 26, 2026 02:16

Conversation

YaSuenag commented Mar 6, 2026 • edited by openjdk bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Progress

Issue

Reviewers

Reviewing

Uh oh!

bridgekeeper bot commented Mar 6, 2026

Uh oh!

openjdk bot commented Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

openjdk bot commented Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mlbridge bot commented Mar 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Webrevs

Uh oh!

YaSuenag commented Mar 6, 2026

Uh oh!

plummercj left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

YaSuenag commented Mar 7, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

RealFYang Mar 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dholmes-ora commented Mar 9, 2026

Uh oh!

YaSuenag commented Mar 9, 2026

Uh oh!

TheRealMDoerr commented Mar 9, 2026

Uh oh!

Uh oh!

Uh oh!

YaSuenag commented Mar 23, 2026

Uh oh!

TheRealMDoerr left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

YaSuenag commented Mar 6, 2026 •

edited by openjdk bot

Loading

openjdk bot commented Mar 6, 2026 •

edited

Loading

openjdk bot commented Mar 6, 2026 •

edited

Loading

mlbridge bot commented Mar 6, 2026 •

edited

Loading

RealFYang Mar 9, 2026 •

edited

Loading