profiling(ddprof): migrate context bridge to the all-native API (Phase 2)

[rkennke]

What

Migrates the profiler context bridge off the deprecated DirectByteBuffer (DBB) context API onto java-profiler's all-native API (setTraceContext / clearTraceContext / setContextValue / clearContextValue).

This is Phase 2 of the all-native context work (Phase 1 = DataDog/java-profiler#631).

Why

• Fixes the virtual-thread use-after-free: the DBB path cached a DirectByteBuffer over the native OTEP record; on a carrier that a mounted virtual thread later migrated off, the record was freed while the cached buffer kept being written. The all-native API resolves the current carrier's record inside each JNI call — no cached buffer to dangle.
• Per-activation perf: activation collapses from setContext + 2×setContextValue (3 JNI calls) into one setTraceContext (~28% faster on the combined cycle per the design-note benchmark).

Changes

• DatadogProfilingIntegration.activate → one setTraceContext(...) carrying trace/span context + operation & resource attributes; close/clearContext → clearTraceContext().
• DatadogProfiler: setContextValue/clearContextValue/reapplyAppContext/syncNativeAppContext are now all-native. reapplyAppContext uses a per-slot native setContextValue loop — native setContextValue publishes the record (valid=1) even with no active span, so app context stays visible between spans (preserves Restore app context attributes wiped by ddprof setContext on span activation #11646). A native batch reapply is deferred to a measured follow-up (java-profiler PROF-15361).
• AppContextSnapshot simplified to strings-only (the native path resolves each value's encoding via the process-wide cache — no cached id/utf8/snapshotTags DBB read).
• snapshot() uses the new native copyContextTags read (no ThreadContext/DBB, so it observes native writes without resetting the record).
• ContextSetter kept only for offsetOf + size (pure Java); no DBB usage remains.

Behavior notes

• Per-slot reapply is not atomic across app attributes. reapplyAppContext publishes each app-managed attribute with its own native setContextValue (detach→attach), so a profiler sample landing mid-reapply can observe a partial app-attribute set on activation — a window the old batch fast-path (setContextValuesByIdAndBytes, single attach) did not have. This is deliberate for the expected small app-attribute cardinality (0–2); the single-call batch reapply is the measured follow-up (PROF-15361). Trace/span + operation/resource stay atomic (one setTraceContext).
• Defensive zero-span guard. The native setTraceContext rejects spanId==0 (it is the activation path; clearing is clearTraceContext). Span ids are non-zero by construction (IdGenerationStrategy never yields 0, DDSpanId.ZERO means "no span", and DDSpanContext is the only ProfilerContext), so the bridge routes a stray zero span to clearTraceContext() rather than letting the IllegalArgumentException be swallowed over stale context.

Dependency

Requires java-profiler's all-native API, from DataDog/java-profiler#631. Master now pins ddprof = 1.47.0 (#12055), which includes it, so this branch has been rebased and builds against the published release — no local snapshot needed.

Status / verification

• Rebased onto master with the published ddprof:1.47.0 (Bump java-profiler (ddprof) to 1.47.0 #12055); compiles clean (no cannot find symbol on the new calls) and the profiling-ddprof unit test suite is GREEN on Linux post-rebase.
• Previously verified GREEN on Linux against a local 1.47.0-SNAPSHOT (pre-rebase, same code): dd-smoke-tests:profiling-integration-tests (CodeHotspotsTest, JFRBasedProfilingIntegrationTest) — 20 tests, 0 failures (end-to-end agent + profiler + JFR, exercising span/op/resource context). Will re-run against the published release before merge.

🤖 Generated with Claude Code

[dd-octo-sts]

🟢 Java Benchmark SLOs — All performance SLOs passed

Suite	Status
Startup	🟢 pass

SLO thresholds are defined here based on automatically generated metrics. A warning is raised when results are within 5% of the threshold.

PR vs. master results

Scenario	Candidate	master	Δ (95% CI of mean)
startup:insecure-bank:iast:Agent	13.97 s	13.91 s	[-0.1%; +1.0%] (no difference)
startup:insecure-bank:tracing:Agent	12.88 s	12.96 s	[-1.4%; +0.1%] (no difference)
startup:petclinic:appsec:Agent	16.91 s	16.79 s	[-0.2%; +1.6%] (no difference)
startup:petclinic:iast:Agent	16.91 s	16.92 s	[-1.1%; +0.9%] (no difference)
startup:petclinic:profiling:Agent	16.73 s	16.71 s	[-1.4%; +1.6%] (no difference)
startup:petclinic:sca:Agent	16.31 s	16.86 s	[-7.5%; +1.0%] (no difference)
startup:petclinic:tracing:Agent	16.05 s	16.15 s	[-1.5%; +0.2%] (no difference)

Commit: abbe49d3 · CI Pipeline · Benchmarking Platform UI

---

Load and DaCapo benchmarks can be triggered manually in the GitLab pipeline. Results will appear in the Benchmarking Platform UI after completion.

[dd-octo-sts]

Hi! 👋 Thanks for your pull request! 🎉

To help us review it, please make sure to:

• Add at least one type, and one component or instrumentation label to the pull request

If you need help, please check our contributing guidelines.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 07a20b2f98

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[chatgpt-codex-connector]

Keep app snapshots in sync when native writes clear a slot

When an app context attribute already has a snapshot, a later write whose value is rejected by the all-native API (for example UTF-8 >255 bytes, dictionary full, or attrs_data overflow) returns false after clearing the native slot. This branch then leaves appContextValues untouched, so the next reapplyAppContext() on span activation/clear writes the previous value back and profiles regain a stale custom attribute instead of observing that the new value was not applied.

Useful? React with 👍 / 👎.

[chatgpt-codex-connector]

Preserve span-derived attributes over app context

When profiling.context.attributes contains _dd.trace.operation or _dd.trace.resource and the corresponding span/resource context attribute is enabled, that slot is also marked app-owned. The previous activation order set operation/resource after setSpanContext, so the span-derived values won; now setTraceContext writes them first and this reapply immediately overwrites them with the saved app value, causing profiles to be tagged with stale/custom operation/resource names for every span on that thread.

Useful? React with 👍 / 👎.

[datadog-prod-us1-5]

Replacing an existing profiling context value with a >255-byte string now clears the native slot but retains its Java snapshot; the next span activation resurrects the stale prior value. The DBB base path retains the prior value continuously, so samples can lose and later regain incorrect context across this migration.

_{📊 Validated against 10 scenarios · Open Bits AI session}

_{🤖 Datadog Autotest · Commit 07a20b2 · What is Autotest? · @DataDog review to ask questions · Any feedback? Reach out in #autotest}

[datadog-prod-us1-5]

Rejected values resurrect stale app context

Profiler samples can temporarily lose an application context attribute and then be labeled with its stale prior value after the next span activation.

Assertion details

• Input: Set a configured app profiling attribute to a valid value, replace it with a 256-byte UTF-8 string, then activate a span.
• Expected: A rejected replacement should leave the native slot and Java snapshot consistent; the pre-PR DBB path retains the prior value continuously.
• Actual: ddprof 1.47.0 clears the native slot and returns false for the oversized replacement, but this branch keeps the prior Java snapshot. The slot changes from encoding 1 to 0, then setSpanContext reapplies the stale snapshot and changes it back to 1.

Was this helpful? React 👍 or 👎
_{🤖 Datadog Autotest · What is Autotest? · @DataDog review to ask questions · Any feedback? Reach out in #autotest}

Suggested change

	if (profiler.setContextValue(offset, value)) {
	recordAppContextValue(offset, value);
	return true;
	}
	if (profiler.setContextValue(offset, value)) {
	recordAppContextValue(offset, value);
	return true;
	}
	reapplyAppContext();