Un-implement dividingFullWidth for 64bit types on 32bit platforms

The implementation uses DoubleWidth<T> that needs to be removed due to
library binary size issues. Even extracting the single DoubleWidth
method that's used in it would likely result in having to copy a
large portion of the rest of DoubleWidth implementation along with it,
this defeating the purpose.

This change restores the behavior of dividingFullWidth as it was shipped
with Swift 4.1, therefore it's not a breaking change.

Author: Max Moiseev

stdlib/public/core/Integers.swift.gyb

@@ -3552,23 +3552,11 @@ ${assignmentOperatorComment(x.operator, True)}
   ) -> (quotient: ${Self}, remainder: ${Self}) {
     // FIXME(integers): tests
 %   # 128-bit types are not provided by the 32-bit LLVM
-%   #if word_bits == 32 and bits == 64:
+%   if word_bits == 32 and bits == 64:
 %   # FIXME(integers): uncomment the above after using the right conditional
 %   # compilation block to exclude 64-bit Windows, which does not support
 %   # 128-bit operations
-%   if bits == 64:
-%     HalfWidth = 'Int32' if signed else 'UInt32'
-    let lhsHigh =
-      unsafeBitCast(dividend.high, to: DoubleWidth<${HalfWidth}>.self)
-    let lhsLow = unsafeBitCast(dividend.low, to: DoubleWidth<UInt32>.self)
-    let rhs_ = unsafeBitCast(self, to: DoubleWidth<${HalfWidth}>.self)
-
-    let (quotient_, remainder_) = rhs_.dividingFullWidth((lhsHigh, lhsLow))
-
-    let quotient = unsafeBitCast(quotient_, to: ${Self}.self)
-    let remainder = unsafeBitCast(remainder_, to: ${Self}.self)
-
-    return (quotient: quotient, remainder: remainder)
+    fatalError("Operation is not supported")
 %   else:
     // FIXME(integers): handle division by zero and overflows
     _precondition(self != 0, "Division by zero")