[
https://issues.apache.org/jira/browse/GROOVY-12191?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18099448#comment-18099448
]
ASF GitHub Bot commented on GROOVY-12191:
-----------------------------------------
daniellansun commented on code in PR #2736:
URL: https://github.com/apache/groovy/pull/2736#discussion_r3657349536
##########
src/main/java/org/codehaus/groovy/vmplugin/v8/IndyInterface.java:
##########
@@ -472,9 +537,14 @@ private static Object
invokeColdReflective(ColdReflectiveMethodHandleWrapper col
throw ScriptBytecodeAdapter.unwrap(gre);
}
}
- MethodHandle mh = selectMethodHandle(cold.callSite, cold.sender,
cold.methodName, cold.callID,
- cold.safeNavigation, cold.thisCall, cold.spreadCall, 1,
arguments);
- return mh.invokeExact(arguments);
+ // Re-select without the cold tier so an always-invalid SwitchPoint
(or any
+ // permanent cold miss after class-domain failover) cannot recurse
through
+ // tryBuild → invokeColdReflective (GROOVY-12191).
+ MethodHandleWrapper full = fallback(cold.callSite, cold.sender,
cold.methodName, cold.callID,
+ cold.safeNavigation, cold.thisCall, cold.spreadCall, 1,
arguments, false);
+ cold.callSite.put(receiverCacheKey(arguments[0]),
+ full.isCanSetTarget() ? full : NULL_METHOD_HANDLE_WRAPPER);
Review Comment:
Agreed that this needed an explicit comment — the policy is shared with
`fromCacheHandle` / `selectMethodHandle`, not cold-tier-specific invention.
**Why the sentinel is the only safe PIC value when `!canSetTarget`**
1. `canSetTarget == false` means the selection is **not
class-keyed-cacheable** (typical causes: per-instance MetaClass, spread-call →
`selector.cache == false`).
2. Storing that real wrapper under the receiver-class PIC key would pin a
selection that is only valid for one instance or one spread shape, and the next
hit would skip re-selection.
3. `NULL_METHOD_HANDLE_WRAPPER` is the existing PIC sentinel meaning *“do
not cache this receiver shape; re-run fallback on the next hit”* (see
`fromCacheHandle`).
4. The **current** invocation still uses `full.getCachedMethodHandle()` —
the sentinel only shapes the PIC, not the return of this call.
**Action taken**
- Expanded the inline comment at the cold-path PIC write to state the above.
- Added a unit test that asserts the uncacheable → sentinel PIC write policy.
> Scope indy SwitchPoint invalidation
> -----------------------------------
>
> Key: GROOVY-12191
> URL: https://issues.apache.org/jira/browse/GROOVY-12191
> Project: Groovy
> Issue Type: Improvement
> Reporter: Daniel Sun
> Priority: Major
> Fix For: 6.0.0-beta-1
>
>
> h3. Problem
> With invokedynamic enabled (default since Groovy 4), linked MOP call sites
> were guarded by a *single process-wide* {{SwitchPoint}}
> ({{{}IndyInterface.switchPoint{}}}).
> Any MetaClass registry change or category enter/leave invalidated that switch
> point, so *every* linked site fell back and re-linked — including sites whose
> receiver type was unrelated.
> That global invalidation is expensive when MetaClass churn is common (e.g.
> ExpandoMetaClass / mixins on startup or per-request paths, Grails-like
> patterns). Unrelated hot monomorphic sites pay re-link and JIT deopt cost
> they should not.
> h3. Goal
> Keep linked call sites optimized unless the *relevant* MetaClass state for
> that site actually changed.
> h3. Approach
> One SwitchPoint domain {*}per class{*}, stored on {{{}ClassInfo{}}}:
> * MetaClass change for type {{T}} retires {{{}T{}}}'s SwitchPoint *and*
> those of loaded subtypes / implementors (hierarchy fan-out).
> * Unrelated types keep their SwitchPoints; their call sites stay optimized.
> * Category enter/leave (and {{{}VMPlugin.invalidateCallSites(){}}})
> bulk-retire *all* loaded class SwitchPoints so sites re-link under the new
> category state. There is *no* second category SwitchPoint on the hot path.
> * Linked handles always install a *single* class-domain guard via
> {{IndyInvalidation.guardWithMopSwitchPoints(...)}} — same monomorphic guard
> shape as before, without global deopt on unrelated MetaClass churn.
> Final classes short-circuit hierarchy fan-out (no full {{ClassInfo}} scan).
> Non-final types batch retirements with {{{}SwitchPoint.invalidateAll{}}}.
> h3. Invalidation map
> ||Event||What is retired||
> |MetaClass change for type {{T}} (registry /
> {{{}ClassInfo.incVersion{}}})|{{T}} + loaded subtypes / implementors|
> |Category enter/leave, {{invalidateCallSites()}}|All loaded class
> SwitchPoints (bulk)|
> |Unattributed MetaClass registry event|All loaded class SwitchPoints (bulk)|
> |First MetaClass *install* on a class|Version bump only (no linked sites
> yet); replacement / clear retires that class's SwitchPoint|
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
