What is the difference between 2 code blocks

package main

import (
   "fmt"
   "github.com/sirupsen/logrus"
   "runtime/debug"
   "time"
)

func main() {

   defer func() {
      if err := recover(); err != nil {
         logrus.Errorln("main", err)
         debug.PrintStack()
      }
   }()
   GoInside(nil)
   time.Sleep(2 * time.Second)
   fmt.Println("After panic")

}

type Test struct {
   X int
}

func GoInside(v *Test) {
   go NullPointer(v.X)
}
func NullPointer(X int) {
   fmt.Println(X)

}

This above code can handle a panic error , but the code below can not handle a panic ==> why 

package main

import (
   "fmt"
   "github.com/sirupsen/logrus"
   "runtime/debug"
   "time"
)

func main() {

   defer func() {
      if err := recover(); err != nil {
         logrus.Errorln("main", err)
         debug.PrintStack()
      }
   }()
   GoInside()
   time.Sleep(2 * time.Second)
   fmt.Println("After panic")

}

type Test struct {
   X int
}

func GoInside() {
   go NullPointer(nil)
}
func NullPointer(v *Test) {
   fmt.Println(v.X)

}

 

Because  in the first code block, we calculated v.X before passing v.X to the go routine => panic happened before start calling to go routine => can handle this panic  

func GoInside(v *Test) {
   go NullPointer(v.X)
}

But int the second code block, we start running go routine first, then inside go routine we calculated v.X => panic happened => main func can not handle this panic