Fast-track @threads when nthreads() == 1

This avoids overhead when threading is disabled, Example benchmark:

```
using BenchmarkTools, Base.Threads

function func(val)
    sum = 0*(1 .^ val)
    for idx in 1:100
        sum += idx.^val
    end
    return sum
end

function func_threaded(val)
    sum = 0*(1 .^ val)
    @threads for idx in 1:100
        sum += idx.^val
    end
    return sum
end

function func_threaded_ref(val)
    sum = Ref(0*(1 .^ val))
    @threads for idx in 1:100
        sum[] += idx.^val
    end
    return sum[]
end

@show @benchmark func(2.0)
@show @benchmark func_threaded(2.0)
@show @benchmark func_threaded_ref(2.0)
```

Before change:
```
```

After change:
```
```

The rest of the overhead is attributed to the `Box` introduced by
#15276

Author: staticfloat

base/threadingconstructs.jl

@@ -25,8 +25,17 @@ function _threadsfor(iter,lbody)
     lidx = iter.args[1]         # index
     range = iter.args[2]
     quote
+        local serialfor_fun
         local threadsfor_fun
+
         let range = $(esc(range))
+        # Fast-track serial execution for case of nthreads == 1
+        function serialfor_fun()
+            for $(esc(lidx)) in range
+                $(esc(lbody))
+            end
+        end
+
         function threadsfor_fun(onethread=false)
             r = range # Load into local variable
             lenr = length(r)
@@ -70,10 +79,14 @@ function _threadsfor(iter,lbody)
             # We are in a nested threaded loop
             Base.invokelatest(threadsfor_fun, true)
         else
-            in_threaded_loop[] = true
-            # the ccall is not expected to throw
-            ccall(:jl_threading_run, Cvoid, (Any,), threadsfor_fun)
-            in_threaded_loop[] = false
+            if nthreads() == 1
+                Base.invokelatest(serialfor_fun)
+            else
+                in_threaded_loop[] = true
+                # the ccall is not expected to throw
+                ccall(:jl_threading_run, Cvoid, (Any,), threadsfor_fun)
+                in_threaded_loop[] = false
+            end
         end
         nothing
     end