Virtual threads support for jvmtiGetFrameCount #15753

thallium · 2022-08-19T14:44:23Z

Adds walkContinuationStackFrames to internal functions so jvmti code can use it.
Adds a check for whether it's a virtual thread and not yielded or terminated.
If it is, use walkContinuationStackFrames to walk the continuation object.
Issue: #15183 #15759

babsingh · 2022-08-19T14:48:48Z

The current problem is that this doesn't compile

Try to fetch the new changes from #15690.

If it still fails to compile, can you post the compilation error here?

runtime/jvmti/jvmtiStackFrame.c

thallium · 2022-08-19T16:57:24Z

The current problem is that this doesn't compile

Try to fetch the new changes from #15690.

If it still fails to compile, can you post the compilation error here?

Actually the error was undefined reference to walkContinuationStackFrames, ended up adding it to internalVMFunctions.

runtime/jvmti/jvmtiStackFrame.c

babsingh · 2022-08-20T00:18:14Z

If needed, request @EricYangIBM's help for running the below tests:

getstacktrace tests with virtual threads: https://github.com/ibmruntimes/openj9-openjdk-jdk19/tree/openj9/test/hotspot/jtreg/serviceability/jvmti/thread/GetStackTrace
getframecount test with virtual threads: https://github.com/ibmruntimes/openj9-openjdk-jdk19/tree/openj9/test/hotspot/jtreg/serviceability/jvmti/thread/GetFrameCount/framecnt01

Instructions which were used for https://github.com/ibmruntimes/openj9-openjdk-jdk19/tree/openj9/test/hotspot/jtreg/serviceability/jvmti/thread/GetThreadState

Get the testimage from the latest AdoptOpenJDK build.
Example: https://ci.adoptopenjdk.net/job/build-scripts/job/jobs/job/jdk19/job/jdk19-linux-x64-temurin/12/artifact/workspace/target/OpenJDK19U-testimage_x64_linux_hotspot_2022-06-23-03-30.tar.gz

# Unzip the testimage
tar -xvzf OpenJDK19U-testimage_x64_linux_hotspot_2022-06-23-03-30.tar.gz

# Copy all shared libraries into the build directory
cp /root/openj9-openjdk-jdk19/jdk-19+27-test-image/hotspot/jtreg/native/* /root/openj9-openjdk-jdk19/build/linux-x86_64-server-release/images/test/hotspot/jtreg/native/.

# Run test
make test TEST="jtreg:test/hotspot/jtreg/serviceability/jvmti/thread/GetThreadState" JTREG="JAVA_OPTIONS=--enable-preview"

fengxue-IS · 2022-08-25T17:33:14Z

@thallium Can you rebase this change on the new changes from #15766, and test again

thallium · 2022-08-25T20:08:46Z

back to seg fault now:

Testing:
JVMTI Stack Trace for thread VirtualThread-Live: frame count: 8
 0: framecnt01: checkFrames0(Ljava/lang/Thread;ZI)Z
 1: framecnt01: checkFrames(Ljava/lang/Thread;ZI)V
 2: framecnt01: lambda$main$0()V
 3: framecnt01$$Lambda$1/0x00000000ec556238: run()V
 4: java/lang/VirtualThread: run(Ljava/lang/Runnable;)V
 5: java/lang/VirtualThread$VThreadContinuation: lambda$new$0(Ljava/lang/VirtualThread;Ljava/lang/Runnable;)V
 6: java/lang/VirtualThread$VThreadContinuation$$Lambda$7/0x000000009406a100: run()V
 7: jdk/internal/vm/Continuation: execute(Ljdk/internal/vm/Continuation;)V

Testing:
STDERR:
Unhandled exception
Type=Segmentation error vmState=0x00040000
J9Generic_Signal_Number=00000018 Signal_Number=0000000b Error_Value=00000000 Signal_Code=00000001
Handler1=00007FA6F10D9BA0 Handler2=00007FA6F068DE30 InaccessibleAddress=0000000000000090
RDI=00000000001A8900 RSI=0000000000000000 RAX=0000000000000020 RBX=00000000001A8900
RCX=0000000000001020 RDX=0000000000000020 R8=0000000000000000 R9=0000000000000000
R10=00007FA690004618 R11=0000000000000000 R12=0000000000000000 R13=00007FA6F149D3E0
R14=00007FA6CC593010 R15=0000000000000000
RIP=00007FA6F11240E3 GS=0000 FS=0000 RSP=00007FA6CC592EE0
EFlags=0000000000010246 CS=0033 RBP=00000000001A8998 ERR=0000000000000004
TRAPNO=000000000000000E OLDMASK=0000000000000000 CR2=0000000000000090
xmm0 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm1 50554f52475f4441 (f: 1197425792.000000, d: 9.870048e+78)
xmm2 ffffffffffffffff (f: 4294967296.000000, d: -nan)
xmm3 ff00000000000000 (f: 0.000000, d: -5.486124e+303)
xmm4 61657268545f676e (f: 1415538560.000000, d: 1.507621e+161)
xmm5 6176614a00494a5f (f: 4803167.000000, d: 3.146433e+161)
xmm6 6c735f6461657268 (f: 1634038400.000000, d: 2.608708e+214)
xmm7 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm8 00000000000ce1b0 (f: 844208.000000, d: 4.170942e-318)
xmm9 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm10 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm11 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm12 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm13 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm14 0000000000000000 (f: 0.000000, d: 0.000000e+00)
xmm15 0000000000000000 (f: 0.000000, d: 0.000000e+00)
Module=/root/openj9-openjdk-jdk19/build/linux-x86_64-server-release/images/jdk/lib/default/libj9vm29.so
Module_base_address=00007FA6F109B000
Target=2_90_20220825_000000 (Linux 5.15.0-46-generic)
CPU=amd64 (8 logical CPUs) (0x3e8091000 RAM)
----------- Stack Backtrace -----------
haltThreadForInspection+0xa3 (0x00007FA6F11240E3 [libj9vm29.so+0x890e3])
jvmtiGetStackTrace+0x159 (0x00007FA6E9ADDE29 [libj9jvmti29.so+0x28e29])
_ZL17print_stack_traceP9_jvmtiEnvP7JNIEnv_P8_jobject+0x61 (0x00007FA6F2782D41 [libframecnt01.so+0x1d41])
Java_framecnt01_checkFrames0+0x51 (0x00007FA6F2783A31 [libframecnt01.so+0x2a31])
ffi_call_unix64+0x52 (0x00007FA6F124BFAA [libj9vm29.so+0x1b0faa])
ffi_call_int+0x1fc (0x00007FA6F124A28C [libj9vm29.so+0x1af28c])
_ZN32VM_BytecodeInterpreterCompressed3runEP10J9VMThread+0x11ec9 (0x00007FA6F1140469 [libj9vm29.so+0xa5469])
bytecodeLoopCompressed+0xad (0x00007FA6F112E58D [libj9vm29.so+0x9358d])
 (0x00007FA6F11D88A2 [libj9vm29.so+0x13d8a2])

fengxue-IS · 2022-08-25T22:17:37Z

the crash is expected as jvmtiGetStackTrace have not been updated to accept a VirtualThread

runtime/vm/intfunc.c

runtime/jvmti/jvmtiStackFrame.c

thallium · 2022-08-29T18:39:12Z

with #15788, the GetFrameCount test passed.

fengxue-IS

LGTM

fengxue-IS · 2022-08-29T19:32:09Z

@thallium can you remove the WIP in heading and mark this as ready for review

fengxue-IS · 2022-08-29T19:32:32Z

@hangshao0 can you take a look

hangshao0 · 2022-08-30T15:35:18Z

Can you update the commit message and the description here including more detailed explanation about the solution ?

thallium · 2022-08-30T15:45:08Z

Done

hangshao0 · 2022-08-30T18:04:56Z

Done

Did you push the change ? I don't see any change.

thallium · 2022-08-30T18:08:13Z

Done

Did you push the change ? I don't see any change.

Sorry I thought you only want to update the PR message, but I don't see any need to update the commit message? You mean adding the message related to walkContinuationStackFrames?

hangshao0 · 2022-08-30T18:17:00Z

You need to update both the commit message and the description of this PR. You can look at other PRs as an example. Also I suggest to mention the issue number in both places. If there is no issue for jvmtiGetFrameCount, you can use #15183.

thallium · 2022-08-30T18:29:24Z

Is it good now?

tajila

one minor comment

runtime/jvmti/jvmtiStackFrame.c

tajila · 2022-08-31T22:04:26Z

runtime/jvmti/jvmtiStackFrame.c

+					j9object_t contObject = (j9object_t)J9VMJAVALANGVIRTUALTHREAD_CONT(currentThread, threadObject);
+					internalVMFunctions->walkContinuationStackFrames(currentThread, contObject, &walkState);
+				} else {
+					Assert_JVMTI_true(0);


you can add a new Assert_JVMTI_unreachable to j9jvmti.tdf

Adds `Assert_JVMTI_unreachable` to mark places as unreachable Adds `walkContinuationStackFrames` to internal functions so jvmti code can use it. Adds a check for whether it's a virtual thread and not yielded or terminated. If it is, use `walkContinuationStackFrames` to walk the continuation object. Issue: eclipse-openj9#15183 eclipse-openj9#15759 Signed-off-by: Gengchen Tuo <gengchen.tuo@ibm.com>

tajila · 2022-08-31T23:52:16Z

jenkins compile win jdk8

tajila · 2022-08-31T23:52:35Z

jenkins test sanity amac jdk19

tajila · 2022-08-31T23:52:54Z

jenkins test sanity plinux jdk19

runtime/jvmti/j9jvmti.tdf

runtime/jvmti/jvmtiStackFrame.c

* move new tracepoint to the end * fix illegal C code (move declaration) Signed-off-by: Keith W. Campbell <keithc@ca.ibm.com>

keithc-ca · 2022-09-01T13:52:49Z

My concerns above are addressed by #15813.

babsingh · 2022-09-01T14:08:31Z

re #15753 (comment):

@thallium Did the GetFrameCount/framecnt01 test in the above comment pass with these changes?

thallium · 2022-09-01T14:13:18Z

re #15753 (comment):

@thallium Did the GetFrameCount/framecnt01 test in the above comment pass with these changes?

It’s mostly due to getStackTrace hasn’t been updated, it passed with my PR on updating getStackTrace

Fix issues in #15753

fengxue-IS suggested changes Aug 19, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

fengxue-IS suggested changes Aug 19, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

thallium force-pushed the jvmtiGetFrameCount branch from 9f741ca to 5751c74 Compare August 19, 2022 22:08

babsingh reviewed Aug 20, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

babsingh mentioned this pull request Aug 20, 2022

Loom: Enable JVMTI serviceability tests from the extensions repo #15748

Closed

thallium force-pushed the jvmtiGetFrameCount branch from 5751c74 to 1afcf92 Compare August 23, 2022 15:00

thallium force-pushed the jvmtiGetFrameCount branch 2 times, most recently from 4986887 to a086b7a Compare August 26, 2022 21:05

fengxue-IS reviewed Aug 26, 2022

View reviewed changes

runtime/vm/intfunc.c Outdated Show resolved Hide resolved

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

thallium force-pushed the jvmtiGetFrameCount branch from a086b7a to f951801 Compare August 26, 2022 21:30

fengxue-IS approved these changes Aug 29, 2022

View reviewed changes

fengxue-IS requested a review from hangshao0 August 29, 2022 19:32

thallium changed the title ~~[WIP]Virtual threads support for jvmtiGetFrameCount~~ Virtual threads support for jvmtiGetFrameCount Aug 29, 2022

thallium marked this pull request as ready for review August 29, 2022 19:34

thallium force-pushed the jvmtiGetFrameCount branch from f951801 to 8fea479 Compare August 30, 2022 18:27

hangshao0 added comp:vm project:loom Used to track Project Loom related work labels Aug 30, 2022

hangshao0 added this to the Java 19 milestone Aug 30, 2022

hangshao0 added the jdk19 label Aug 30, 2022

thallium force-pushed the jvmtiGetFrameCount branch from c91dc35 to e282302 Compare August 31, 2022 14:41

hangshao0 requested a review from tajila August 31, 2022 15:25

tajila reviewed Aug 31, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Show resolved Hide resolved

thallium force-pushed the jvmtiGetFrameCount branch from e282302 to 7955359 Compare August 31, 2022 21:47

thallium requested a review from tajila August 31, 2022 21:56

tajila reviewed Aug 31, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Outdated Show resolved Hide resolved

tajila reviewed Aug 31, 2022

View reviewed changes

thallium force-pushed the jvmtiGetFrameCount branch from 7955359 to 2aa6979 Compare August 31, 2022 22:19

tajila approved these changes Aug 31, 2022

View reviewed changes

tajila merged commit cfbfdeb into eclipse-openj9:master Sep 1, 2022

thallium deleted the jvmtiGetFrameCount branch September 1, 2022 13:02

keithc-ca reviewed Sep 1, 2022

View reviewed changes

runtime/jvmti/j9jvmti.tdf Show resolved Hide resolved

keithc-ca reviewed Sep 1, 2022

View reviewed changes

runtime/jvmti/jvmtiStackFrame.c Show resolved Hide resolved

keithc-ca added a commit to keithc-ca/openj9 that referenced this pull request Sep 1, 2022

Fix issues in eclipse-openj9#15753

2a0fb45

* move new tracepoint to the end * fix illegal C code (move declaration) Signed-off-by: Keith W. Campbell <keithc@ca.ibm.com>

thallium mentioned this pull request Sep 2, 2022

Update jvmtiGetFrameLocation to support virtual threads #15820

Merged

tajila added a commit that referenced this pull request Sep 2, 2022

Merge pull request #15813 from keithc-ca/fix15753

87c808f

Fix issues in #15753

fengxue-IS mentioned this pull request Oct 25, 2022

[JDK19] exclude JTReg JVMTI serviceability tests due to implementation differences #16185

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Virtual threads support for jvmtiGetFrameCount #15753

Virtual threads support for jvmtiGetFrameCount #15753

thallium commented Aug 19, 2022 •

edited

Loading

babsingh commented Aug 19, 2022 •

edited

Loading

thallium commented Aug 19, 2022

babsingh commented Aug 20, 2022 •

edited

Loading

fengxue-IS commented Aug 25, 2022 •

edited

Loading

thallium commented Aug 25, 2022

fengxue-IS commented Aug 25, 2022

thallium commented Aug 29, 2022

fengxue-IS left a comment

fengxue-IS commented Aug 29, 2022

fengxue-IS commented Aug 29, 2022

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022 •

edited

Loading

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022

tajila left a comment

tajila Aug 31, 2022

tajila commented Aug 31, 2022

tajila commented Aug 31, 2022

tajila commented Aug 31, 2022

keithc-ca commented Sep 1, 2022

babsingh commented Sep 1, 2022

thallium commented Sep 1, 2022

Virtual threads support for jvmtiGetFrameCount #15753

Virtual threads support for jvmtiGetFrameCount #15753

Conversation

thallium commented Aug 19, 2022 • edited Loading

babsingh commented Aug 19, 2022 • edited Loading

thallium commented Aug 19, 2022

babsingh commented Aug 20, 2022 • edited Loading

fengxue-IS commented Aug 25, 2022 • edited Loading

thallium commented Aug 25, 2022

fengxue-IS commented Aug 25, 2022

thallium commented Aug 29, 2022

fengxue-IS left a comment

Choose a reason for hiding this comment

fengxue-IS commented Aug 29, 2022

fengxue-IS commented Aug 29, 2022

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022 • edited Loading

hangshao0 commented Aug 30, 2022

thallium commented Aug 30, 2022

tajila left a comment

Choose a reason for hiding this comment

tajila Aug 31, 2022

Choose a reason for hiding this comment

tajila commented Aug 31, 2022

tajila commented Aug 31, 2022

tajila commented Aug 31, 2022

keithc-ca commented Sep 1, 2022

babsingh commented Sep 1, 2022

thallium commented Sep 1, 2022

thallium commented Aug 19, 2022 •

edited

Loading

babsingh commented Aug 19, 2022 •

edited

Loading

babsingh commented Aug 20, 2022 •

edited

Loading

fengxue-IS commented Aug 25, 2022 •

edited

Loading

thallium commented Aug 30, 2022 •

edited

Loading