Eliminate the sizing template parameter N from CoalescingBitVector

[dim]

Summary

Since the parameter is not used anywhere, and the default size of 16
apparently causes PR47359, remove it. This ensures that IntervalMap will
automatically determine the optimal size, using its NodeSizer struct.

[dblaikie]

Fair enough

llvm/include/llvm/ADT/CoalescingBitVector.h

 /// elements into intervals. Good for representing sets which predominantly
 /// contain contiguous ranges. Bad for representing sets with lots of gaps
 /// between elements.
 ///
 /// Compared to SparseBitVector, CoalescingBitVector offers more predictable
 /// performance for non-sequential find() operations.
 ///
 /// \tparam IndexT - The type of the index into the bitvector.
-/// \tparam N - The first N coalesced intervals of set bits are stored in-place.
-template <typename IndexT, unsigned N = 16> class CoalescingBitVector {
+template <typename IndexT> class CoalescingBitVector {
   static_assert(std::is_unsigned<IndexT>::value,
                 "Index must be an unsigned integer.");
 
-  using ThisT = CoalescingBitVector<IndexT, N>;
+  using ThisT = CoalescingBitVector<IndexT>;
 
   /// An interval map for closed integer ranges. The mapped values are unused.
-  using MapT = IntervalMap<IndexT, char, N>;
+  using MapT = IntervalMap<IndexT, char>;
 
   using UnderlyingIterator = typename MapT::const_iterator;
 
   using IntervalT = std::pair<IndexT, IndexT>;
 
 public:
   using Allocator = typename MapT::Allocator;