eth/tracers: use non-threaded tracechain (#24283)

This makes non-JS tracers execute all block txs on a single goroutine.
In the previous implementation, we used to prepare every tx pre-state
on one goroutine, and then run the transactions again with tracing enabled.
Native tracers are usually faster, so it is faster overall to use their output as
the pre-state for tracing the next transaction.

Co-authored-by: Sina Mahmoodi <itz.s1na@gmail.com>

eth/tracers/js/goja.go

@@ -45,7 +45,16 @@ func init() {
 	if err != nil {
 		panic(err)
 	}
-	tracers.RegisterLookup(true, newJsTracer)
+	type ctorFn = func(*tracers.Context, json.RawMessage) (tracers.Tracer, error)
+	lookup := func(code string) ctorFn {
+		return func(ctx *tracers.Context, cfg json.RawMessage) (tracers.Tracer, error) {
+			return newJsTracer(code, ctx, cfg)
+		}
+	}
+	for name, code := range assetTracers {
+		tracers.DefaultDirectory.Register(name, lookup(code), true)
+	}
+	tracers.DefaultDirectory.RegisterJSEval(newJsTracer)
 }
 
 // bigIntProgram is compiled once and the exported function mostly invoked to convert
@@ -122,16 +131,14 @@ type jsTracer struct {
 	frameResultValue goja.Value
 }
 
-// newJsTracer instantiates a new JS tracer instance. code is either
-// the name of a built-in JS tracer or a Javascript snippet which
-// evaluates to an expression returning an object with certain methods.
+// newJsTracer instantiates a new JS tracer instance. code is a
+// Javascript snippet which evaluates to an expression returning
+// an object with certain methods:
+//
 // The methods `result` and `fault` are required to be present.
 // The methods `step`, `enter`, and `exit` are optional, but note that
 // `enter` and `exit` always go together.
 func newJsTracer(code string, ctx *tracers.Context, cfg json.RawMessage) (tracers.Tracer, error) {
-	if c, ok := assetTracers[code]; ok {
-		code = c
-	}
 	vm := goja.New()
 	// By default field names are exported to JS as is, i.e. capitalized.
 	vm.SetFieldNameMapper(goja.UncapFieldNameMapper())