LibJS/Temporal: Perform floating point arithmetic in RoundTime

The valid range for temporal values (`nsMinInstant`/`nsMaxInstant`)
means performing nanosecond-valued integers could lead to an overflow.

NB: Only the `roundingMode: "day"` case was affected, as all others were
already performing the division on floating-point `fractional_second`
values. I'm adding `.0` suffixes everywhere to make this fact clearer.

This adds a few local tests as well, as those are tested with sanitizers
enabled by default, unlike test262.

Userland/Libraries/LibJS/Tests/builtins/Temporal/PlainDateTime/PlainDateTime.prototype.round.js

@@ -88,6 +88,28 @@ describe("correct behavior", () => {
             plainDateTime.round("minute").equals(plainDateTime.round({ smallestUnit: "minute" }))
         ).toBeTrue();
     });
+
+    test("range boundary conditions", () => {
+        // PlainDateTime can represent a point of time ±10**8 days from the epoch.
+        const min = new Temporal.PlainDateTime(-271821, 4, 19, 0, 0, 0, 0, 0, 1);
+        const max = new Temporal.PlainDateTime(275760, 9, 13, 23, 59, 59, 999, 999, 999);
+
+        ["day", "hour", "minute", "second", "millisecond", "microsecond"].forEach(smallestUnit => {
+            expect(() => {
+                min.round({ smallestUnit, roundingMode: "floor" });
+            }).toThrow(RangeError);
+            expect(() => {
+                min.round({ smallestUnit, roundingMode: "ceil" });
+            }).not.toThrow();
+
+            expect(() => {
+                max.round({ smallestUnit, roundingMode: "floor" });
+            }).not.toThrow();
+            expect(() => {
+                max.round({ smallestUnit, roundingMode: "ceil" });
+            }).toThrow(RangeError);
+        });
+    });
 });
 
 describe("errors", () => {