8295303: cleanup debug agent's confusing use of EI_GC_FINISH #10887

plummercj · 2022-10-27T18:06:38Z

This PR renamed EI_GC_FINISH to EI_CLASS_UNLOAD and does some other minor cleanups related to EI_GC_FINISH.

The debug agent deals with 3 types of events: JVMTI (per the spec), JDWP (per the spec), and an event indexing that the debug agent refers to as EI (Event Index) events. We have the following EI_GC_FINISH event. The indexing is into the array of jdwp event handlers which are created when the debugger sends an EventRequest command.

Note there is no EI_CLASS_UNLOAD event that maps to the JDWP CLASS_UNLOAD event, but instead we have:

EI_GC_FINISH = 8,

And these are the mappings between EI_GC_FINISH and JDWP and JVMTI events

  case JVMTI_EVENT_GARBAGE_COLLECTION_FINISH:
      return EI_GC_FINISH;

  case JDWP_EVENT(CLASS_UNLOAD):
      return EI_GC_FINISH;

  index2jvmti[EI_GC_FINISH - EI_min] = JVMTI_EVENT_GARBAGE_COLLECTION_FINISH;

  index2jdwp[EI_GC_FINISH - EI_min] = JDWP_EVENT(CLASS_UNLOAD);

So there is this odd relationship between EI_GC_FINISH and the JDWP CLASS_UNLOAD event. Note that JVMTI does not have a CLASS_UNLOAD (except for unused support in the extension mechanism), and JDWP has no GC_FINISH event.

This relationship between EI_GC_FINISH and the JDWP CLASS_UNLOAD events stems from the fact that at one point JVMTI_EVENT_GARBAGE_COLLECTION_FINISH was used to trigger the synthesizing all of JDWP CLASS_UNLOAD events for classes that unloaded during the last GC. That is no longer the case, and instead each time a JVMTI OBJECT_FREE event is triggered for a Class instance, the JDWP CLASS_UNLOAD is generated.

Since JDWP CLASS_UNLOAD maps to EI_GC_FINISH, we have the following:

node = getHandlerChain(EI_GC_FINISH)->first;

By looking at this line of code, you would never guess that this is how you fetch the event handler chain for JDWP CLASS_UNLOAD EventRequests, but it is.

To clean this up I renamed EI_GC_FINISH to EI_CLASS_UNLOAD. However, that still leaves the EI_GC_FINISH to JVMTI_EVENT_GARBAGE_COLLECTION_FINISH mapping to deal with. It's not needed anymore. When we get a JVMTI_EVENT_GARBAGE_COLLECTION_FINISH, this is all we ever do with it:

static void JNICALL
cbGarbageCollectionFinish(jvmtiEnv *jvmti_env)
{
    LOG_CB(("cbGarbageCollectionFinish"));
    ++garbageCollected;
    LOG_MISC(("END cbGarbageCollectionFinish"));
}

So the event is not passed around at all, and doesn't trigger other events or mapping to a JDWP event. It is never used as an "event index". This means jvmti2EventIndex should assert if it is ever passed in, since following mapping should never be needed:

        case JVMTI_EVENT_GARBAGE_COLLECTION_FINISH:
            return EI_GC_FINISH;

This is accomplished by simply deleting the above code, and failing through to the default error handling code.

Also no longer needed is the following entry:

index2jvmti[EI_GC_FINISH          -EI_min] = JVMTI_EVENT_GARBAGE_COLLECTION_FINISH

For this we just assign the entry to 0, which will result in an error if the entry is ever referenced.

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-8295303: cleanup debug agent's confusing use of EI_GC_FINISH

Reviewers

Alex Menkov (@alexmenkov - Reviewer)
Serguei Spitsyn (@sspitsyn - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk pull/10887/head:pull/10887
$ git checkout pull/10887

Update a local copy of the PR:
$ git checkout pull/10887
$ git pull https://git.openjdk.org/jdk pull/10887/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 10887

View PR using the GUI difftool:
$ git pr show -t 10887

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/10887.diff

…EI_GC_FINISH

bridgekeeper · 2022-10-27T18:07:03Z

👋 Welcome back cjplummer! 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 · 2022-10-27T18:11:12Z

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

serviceability

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 · 2022-11-03T16:26:13Z

Webrevs

alexmenkov · 2022-11-08T01:16:06Z

src/jdk.jdwp.agent/share/native/libjdwp/eventHandler.c

+     * GARBAGE_COLLECTION_FINISH is special since it is not tied to any handlers or an EI,
+     * so it cannot be setup using threadControl_setEventMode(). Use JVMTI API directly.
+     */
+    error = JVMTI_FUNC_PTR(gdata->jvmti,SetEventNotificationMode)


Please add space after the comma:
error = JVMTI_FUNC_PTR(gdata->jvmti, SetEventNotificationMode)

Omitting the space is consistent with other uses of JVMTI_FUNC_PTR.

alexmenkov · 2022-11-08T01:19:06Z

src/jdk.jdwp.agent/share/native/libjdwp/util.c

-    if ( i < EI_min || i > EI_max ) {
-        EXIT_ERROR(AGENT_ERROR_INVALID_INDEX,"bad EventIndex");
+    jdwpEvent event = 0;
+    if (ei >= EI_min || ei >= EI_max) {


Should be "(ei >= EI_min && ei <= EI_max"

alexmenkov · 2022-11-08T01:20:21Z

src/jdk.jdwp.agent/share/native/libjdwp/util.c

-    if ( i < EI_min || i > EI_max ) {
-        EXIT_ERROR(AGENT_ERROR_INVALID_INDEX,"bad EventIndex");
+    jvmtiEvent event = 0;
+    if (ei >= EI_min || ei >= EI_max) {


Should be "(ei >= EI_min && ei <= EI_max"

openjdk · 2022-11-08T02:18:29Z

@plummercj this pull request can not be integrated into master due to one or more merge conflicts. To resolve these merge conflicts and update this pull request you can run the following commands in the local repository for your personal fork:

git checkout 8295303_gc_finish
git fetch https://git.openjdk.org/jdk master
git merge FETCH_HEAD
# resolve conflicts and follow the instructions given by git merge
git commit -m "Merge master"
git push

sspitsyn · 2022-11-08T09:27:33Z

src/jdk.jdwp.agent/share/native/libjdwp/error_messages.c

@@ -232,6 +232,7 @@ eventText(int i)
        CASE_RETURN_TEXT(EI_THREAD_START)
        CASE_RETURN_TEXT(EI_THREAD_END)
        CASE_RETURN_TEXT(EI_CLASS_PREPARE)
+        CASE_RETURN_TEXT(EI_CLASS_UNLOAD)
        CASE_RETURN_TEXT(EI_CLASS_LOAD)


Nit: Would it be a little better to place it after EI_CLASS_UNLOAD?

If you mean put EI_CLASS_UNLOAD after EI_CLASS_LOAD, I tried to keep the current ordering. If you look in util.h (bottom of this page), it used to have:

EI_CLASS_PREPARE = 7, EI_GC_FINISH = 8, EI_CLASS_LOAD = 9,

and now it has:

EI_CLASS_PREPARE = 7, EI_CLASS_UNLOAD = 8, EI_CLASS_LOAD = 9,

There are a couple of other places where this ordering is preserved and I tried to be consistent.

Okay, thanks.
It makes sense then.

sspitsyn

Looks good.
Thanks,
Serguei

openjdk · 2022-11-08T09:30:23Z

@plummercj 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:

8295303: cleanup debug agent's confusing use of EI_GC_FINISH

Reviewed-by: amenkov, sspitsyn

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 4 new commits pushed to the master branch:

0ee25de: 8296504: Memory leak in G1PLABAllocator::PLABData
dd5d4df: 8295658: G1: Refactor G1SegmentedArray to indicate that it is an allocator
cf65605: 8296445: C++ syntax error in jdwpTransport.h
1169dc0: 8296447: RISC-V: Make the operands order of vrsub_vx/vrsub_vi consistent with RVV 1.0 spec

Please see this link for an up-to-date comparison between the source branch of this pull request and the master branch.
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.

plummercj · 2022-11-08T21:35:33Z

Thanks Alex and Serguei!

/integrate

openjdk · 2022-11-08T21:37:33Z

Going to push as commit 74f2b16.
Since your change was applied there have been 4 commits pushed to the master branch:

0ee25de: 8296504: Memory leak in G1PLABAllocator::PLABData
dd5d4df: 8295658: G1: Refactor G1SegmentedArray to indicate that it is an allocator
cf65605: 8296445: C++ syntax error in jdwpTransport.h
1169dc0: 8296447: RISC-V: Make the operands order of vrsub_vx/vrsub_vi consistent with RVV 1.0 spec

Your commit was automatically rebased without conflicts.

openjdk · 2022-11-08T21:37:55Z

@plummercj Pushed as commit 74f2b16.

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

Rename EI_GC_FINISH to EI_CLASS_UNLOAD and other cleanups related to …

aac8bdd

…EI_GC_FINISH

openjdk bot changed the title ~~8295303~~ 8295303: cleanup debug agent's confusing use of EI_GC_FINISH Oct 27, 2022

openjdk bot added the serviceability serviceability-dev@openjdk.org label Oct 27, 2022

fix extra whitespace

ce7a283

openjdk bot added the rfr Pull request is ready for review label Nov 3, 2022

alexmenkov suggested changes Nov 8, 2022

View reviewed changes

Fixed ei range check logic errors.

ec44b9d

openjdk bot added the merge-conflict Pull request has merge conflict with target branch label Nov 8, 2022

merge

ca98ed8

openjdk bot removed the merge-conflict Pull request has merge conflict with target branch label Nov 8, 2022

sspitsyn reviewed Nov 8, 2022

View reviewed changes

sspitsyn approved these changes Nov 8, 2022

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Nov 8, 2022

alexmenkov approved these changes Nov 8, 2022

View reviewed changes

openjdk bot added the integrated Pull request has been integrated label Nov 8, 2022

openjdk bot closed this Nov 8, 2022

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Nov 8, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8295303: cleanup debug agent's confusing use of EI_GC_FINISH #10887

8295303: cleanup debug agent's confusing use of EI_GC_FINISH #10887

plummercj commented Oct 27, 2022 •

edited by openjdk bot

Loading

bridgekeeper bot commented Oct 27, 2022

openjdk bot commented Oct 27, 2022

mlbridge bot commented Nov 3, 2022 •

edited

Loading

alexmenkov Nov 8, 2022

plummercj Nov 8, 2022

alexmenkov Nov 8, 2022

plummercj Nov 8, 2022

alexmenkov Nov 8, 2022

plummercj Nov 8, 2022

openjdk bot commented Nov 8, 2022

sspitsyn Nov 8, 2022

plummercj Nov 8, 2022

sspitsyn Nov 9, 2022

sspitsyn left a comment

openjdk bot commented Nov 8, 2022 •

edited

Loading

plummercj commented Nov 8, 2022

openjdk bot commented Nov 8, 2022

openjdk bot commented Nov 8, 2022

8295303: cleanup debug agent's confusing use of EI_GC_FINISH #10887

8295303: cleanup debug agent's confusing use of EI_GC_FINISH #10887

Conversation

plummercj commented Oct 27, 2022 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Oct 27, 2022

openjdk bot commented Oct 27, 2022

mlbridge bot commented Nov 3, 2022 • edited Loading

Webrevs

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

openjdk bot commented Nov 8, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

sspitsyn left a comment

Choose a reason for hiding this comment

openjdk bot commented Nov 8, 2022 • edited Loading

plummercj commented Nov 8, 2022

openjdk bot commented Nov 8, 2022

openjdk bot commented Nov 8, 2022

plummercj commented Oct 27, 2022 •

edited by openjdk bot

Loading

mlbridge bot commented Nov 3, 2022 •

edited

Loading

openjdk bot commented Nov 8, 2022 •

edited

Loading