Fix DDTraceId/DD64bTraceId class-initialization deadlock

[bm1549]

What Does This Do

Fixes a class-initialization deadlock between DDTraceId and DD64bTraceId that can hang trace creation at startup. DDTraceId.ZERO/ONE are now backed by a private sibling type instead of DD64bTraceId, so DDTraceId.<clinit> no longer initializes its own subclass. The public DDTraceId.ZERO/ONE fields are unchanged (no binary-incompatible change). A new value-based DDTraceId.isZero() replaces the == DDTraceId.ZERO sentinel checks.

Motivation

DD64bTraceId extends DDTraceId, so the JVM initializes DDTraceId first. But DDTraceId.<clinit> built its ZERO/ONE constants via DD64bTraceId.from(...), which initializes the subclass while the DDTraceId init lock is held. When the two classes are first touched concurrently from opposite ends, each thread ends up holding one class-init lock and waiting for the other:

• dd-task-scheduler: the service-discovery task added in Add support for service discovery using JNA #9705 runs muteTracing() -> blackholeSpan() -> DDTraceId.ZERO
• main: the application's first span runs IdGenerationStrategy.generateTraceId() -> DD64bTraceId.from()

Trace creation then hangs. This surfaced as recurring ~30s LogInjectionSmokeTest timeouts on master (traceCount=0, process.alive=true, RC polls received: ~135). The forked-process thread dumps added in #11400 confirmed the cycle, and it reproduces deterministically.

Additional Notes

• Approach: break the cycle at its source. ZERO/ONE stay public static final DDTraceId fields (the surface deliberately restored in [6to7] Restore public DDTraceId class API #5021), but are now instances of a private DDTraceId subtype, ConstantId, that is a sibling of DD64bTraceId. Because DDTraceId.<clinit> no longer references the subclass, the deadlock cannot happen regardless of timing.
• Zero checks now use a value-based DDTraceId.isZero() instead of == DDTraceId.ZERO. The identity checks assumed every zero id was the single ZERO instance; isZero() recognizes a zero id of any concrete type, so the factories no longer special-case 0 and a zero parsed via the direct 64-bit factories (DD64bTraceId.fromHex in the XRay/Haystack codecs) is handled correctly. It also recognizes an all-zero 128-bit id, which == ZERO silently missed.
• DDTraceIdClinitDeadlockForkedTest runs in a fresh JVM and initializes the two classes concurrently from opposite ends; it deadlocks without the fix and passes with it. TraceIdIsZeroTest and DDTraceIdConstantsTest cover isZero() and the constants across the DDTraceId subtypes.
• The deadlock has been latent since Add support for service discovery using JNA #9705 (Oct 2025) added the scheduled muteTracing() task; it began manifesting recently as startup timing shifted.

Contributor Checklist

• Title formatted per the contribution guidelines
• type: and comp: labels assigned
• No issue-linking keywords used
• CODEOWNERS update not required (no file addition/migration/deletion)
• Public documentation update not required (no new configuration or behavior)

Jira ticket: N/A

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 56ea720eb8

ℹ️ About Codex in GitHub

Your team has set up Codex to 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 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[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	14.08 s	14.01 s	[-0.7%; +1.7%] (no difference)
startup:insecure-bank:tracing:Agent	12.90 s	12.93 s	[-1.1%; +0.6%] (no difference)
startup:petclinic:appsec:Agent	16.57 s	15.53 s	[-2.4%; +15.7%] (unstable)
startup:petclinic:iast:Agent	16.41 s	16.46 s	[-1.5%; +0.9%] (no difference)
startup:petclinic:profiling:Agent	16.40 s	15.62 s	[-3.8%; +13.7%] (unstable)
startup:petclinic:tracing:Agent	15.65 s	14.87 s	[-3.3%; +13.7%] (unstable)

Commit: ded2c7c7 · 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.

[dougqh]

Overall, it looks good to me.
But before merging, double check that the Codex comment isn't a problem