Move thread-ownership optimization into OptimizedTagMap

Per review feedback, move the lock-skipping optimization from
DDSpanContext into OptimizedTagMap itself. This keeps the optimization
invisible to callers — DDSpanContext no longer needs synchronized blocks
around tag operations, and developers adding new tag operations don't
need to think about locking.

OptimizedTagMap now has a volatile Thread ownerThread field. Core methods
(getAndSet, getAndRemove, getEntry, putAll, forEach, copy, etc.) check
ownership: owner thread skips the lock, non-owner threads synchronize
and permanently transition to shared mode.

DDSpanContext changes: removed all 27 synchronized(unsafeTags) blocks,
added setOwnerThread(current) in constructor, transitionToShared()
delegates to TagMap.

Also adds @threads(8) JMH benchmark variants and 5 new concurrency
tests (mixed read/write, fuzz, value consistency, finish race,
concurrent metrics).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: bm1549

dd-trace-core/src/jmh/java/datadog/trace/core/SpanTagBenchmark.java

@@ -86,6 +86,43 @@ public AgentSpan setStringTag_crossThread(SharedSpan state) {
     return state.span;
   }
 
+  // Multi-threaded owner-thread benchmarks: 8 threads each working on their own span.
+  // Measures the macro impact of the optimization when many threads tag spans concurrently,
+  // the realistic web-server scenario.
+
+  @Benchmark
+  @Threads(8)
+  @OutputTimeUnit(NANOSECONDS)
+  public AgentSpan setStringTag_ownerThread_8T(SpanPerThread state) {
+    state.span.setTag("key", "value");
+    return state.span;
+  }
+
+  @Benchmark
+  @Threads(8)
+  @OutputTimeUnit(NANOSECONDS)
+  public AgentSpan setIntTag_ownerThread_8T(SpanPerThread state) {
+    state.span.setTag("key", 42);
+    return state.span;
+  }
+
+  @Benchmark
+  @Threads(8)
+  @OutputTimeUnit(NANOSECONDS)
+  public AgentSpan setTenTags_ownerThread_8T(SpanPerThread state) {
+    state.span.setTag("k0", "v0");
+    state.span.setTag("k1", "v1");
+    state.span.setTag("k2", "v2");
+    state.span.setTag("k3", "v3");
+    state.span.setTag("k4", "v4");
+    state.span.setTag("k5", 5);
+    state.span.setTag("k6", 6L);
+    state.span.setTag("k7", 7.0);
+    state.span.setTag("k8", true);
+    state.span.setTag("k9", "v9");
+    return state.span;
+  }
+
   @Benchmark
   @Threads(1)
   @OutputTimeUnit(MICROSECONDS)
@@ -103,4 +140,22 @@ public AgentSpan fullLifecycle_tenTags(SpanPerThread state) {
     state.span.finish();
     return state.span;
   }
+
+  @Benchmark
+  @Threads(8)
+  @OutputTimeUnit(MICROSECONDS)
+  public AgentSpan fullLifecycle_tenTags_8T(SpanPerThread state) {
+    state.span.setTag("k0", "v0");
+    state.span.setTag("k1", "v1");
+    state.span.setTag("k2", "v2");
+    state.span.setTag("k3", "v3");
+    state.span.setTag("k4", "v4");
+    state.span.setTag("k5", 5);
+    state.span.setTag("k6", 6L);
+    state.span.setTag("k7", 7.0);
+    state.span.setTag("k8", true);
+    state.span.setTag("k9", "v9");
+    state.span.finish();
+    return state.span;
+  }
 }