Imported Upstream version 5.16.0.100

Former-commit-id: 38faa55fb9669e35e7d8448b15c25dc447f25767

mcs/class/referencesource/mscorlib/system/timespan.cs

@@ -379,6 +379,26 @@ namespace System {
             return t1._ticks >= t2._ticks;
         }
 
+        public static TimeSpan operator / (TimeSpan timeSpan, double divisor)
+        {
+            if (double.IsNaN (divisor)) {
+                throw new ArgumentException (SR.Arg_CannotBeNaN, nameof (divisor));
+            }
+
+            double ticks = Math.Round (timeSpan.Ticks / divisor);
+            if (ticks > long.MaxValue | ticks < long.MinValue || double.IsNaN (ticks)) {
+                throw new OverflowException (SR.Overflow_TimeSpanTooLong);
+            }
+
+            return FromTicks ((long)ticks);
+        }
+
+        // Using floating-point arithmetic directly means that infinities can be returned, which is reasonable
+        // if we consider TimeSpan.FromHours(1) / TimeSpan.Zero asks how many zero-second intervals there are in
+        // an hour for which infinity is the mathematic correct answer. Having TimeSpan.Zero / TimeSpan.Zero return NaN
+        // is perhaps less useful, but no less useful than an exception.
+        public static double operator /(TimeSpan t1, TimeSpan t2) => t1.Ticks / (double)t2.Ticks;
+
 
         //
         // In .NET Framework v1.0 - v3.5 System.TimeSpan did not implement IFormattable