[llvm] Introduce paged vector (PR #66430)

[Jakub Kuderski via llvm-commits]

================
@@ -0,0 +1,133 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors --------*- C++
+//-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines the PagedVector class.
+//
+//===----------------------------------------------------------------------===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include <cassert>
+#include <vector>
+
+namespace llvm {
+// A vector that allocates memory in pages.
+// Order is kept, but memory is allocated only when one element of the page is
+// accessed. This introduces a level of indirection, but it is useful when you
+// have a sparsely initialised vector where the full size is allocated upfront
+// with the default constructor and elements are initialised later, on first
+// access.
+//
+// Notice that this does not have iterators, because if you
+// have iterators it probably means you are going to touch
+// all the memory in any case, so better use a std::vector in
+// the first place.
+template <typename T, int PAGE_SIZE = 1024 / sizeof(T)> class PagedVector {
+  // The actual number of element in the vector which can be accessed.
+  std::size_t Size = 0;
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable std::vector<int> Lookup;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The oder of
+  // the elements however depends on the order of access of the pages.
+  mutable std::vector<T> Data;
+
+public:
+  // Lookup an element at position Index.
+  T &operator[](std::size_t Index) const { return at(Index); }
+
+  // Lookup an element at position i.
+  // If the associated page is not filled, it will be filled with default
+  // constructed elements. If the associated page is filled, return the element.
+  T &at(std::size_t Index) const {
+    assert(Index < Size);
+    assert(Index / PAGE_SIZE < Lookup.size());
+    auto &PageId = Lookup[Index / PAGE_SIZE];
+    // If the range is not filled, fill it
+    if (PageId == -1) {
+      int OldSize = Data.size();
+      PageId = OldSize / PAGE_SIZE;
+      // Allocate the memory
+      Data.resize(OldSize + PAGE_SIZE);
+      // Fill the whole capacity with empty elements
+      for (int I = 0; I < PAGE_SIZE; ++I) {
+        Data[I + OldSize] = T();
+      }
+    }
+    // Calculate the actual position in the Data vector
+    // by taking the start of the page and adding the offset
+    // in the page.
+    std::size_t StoreIndex = Index % PAGE_SIZE + PAGE_SIZE * PageId;
+    // Return the element
+    assert(StoreIndex < Data.size());
+    return Data[StoreIndex];
+  }
+
+  // Return the capacity of the vector. I.e. the maximum size it can be expanded
+  // to with the expand method without allocating more pages.
+  std::size_t capacity() const { return Lookup.size() * PAGE_SIZE; }
+
+  // Return the size of the vector. I.e. the maximum index that can be
+  // accessed, i.e. the maximum value which was used as argument of the
+  // expand method.
+  std::size_t size() const { return Size; }
+
+  // Expands the vector to the given NewSize number of elements.
+  // If the vector was smaller, allocates new pages as needed.
+  // It should be called only with NewSize >= Size.
+  void expand(std::size_t NewSize) {
+    // You cannot shrink the vector, otherwise
+    // one would have to invalidate contents which is expensive and
+    // while giving the false hope that the resize is cheap.
+    if (NewSize <= Size) {
+      return;
+    }
+    // If the capacity is enough, just update the size and continue
+    // with the currently allocated pages.
+    if (NewSize <= capacity()) {
+      Size = NewSize;
+      return;
+    }
+    // The number of pages to allocate. The Remainder is calculated
+    // for the case in which the NewSize is not a multiple of PAGE_SIZE.
+    // In that case we need one more page.
+    auto Pages = NewSize / PAGE_SIZE;
+    auto Remainder = NewSize % PAGE_SIZE;
+    if (Remainder) {
+      Pages += 1;
+    }
+    assert(Pages > Lookup.size());
+    // We use -1 to indicate that a page has not been allocated yet.
+    // This cannot be 0, because 0 is a valid page id.
+    // We use -1 instead of a separate bool to avoid wasting space.
+    Lookup.resize(Pages, -1);
+    Size = NewSize;
+  }
+
+  // Return true if the vector is empty
+  bool empty() const { return Size == 0; }
+
+  /// Clear the vector, i.e. clear the allocated pages, the whole page
+  /// lookup index and reset the size.
+  void clear() {
+    Size = 0;
+    Lookup.clear();
+    Data.clear();
+  }
+
+  /// Return the materialised vector. This is useful if you want to iterate
+  /// in an efficient way over the non default constructed elements.
+  /// It's not called data() because that would be misleading, since only
+  /// elements for pages which have been accessed are actually allocated.
+  std::vector<T> const &materialised() const { return Data; }
----------------
kuhar wrote:

It's up to you to define how to iterate over the collection, I would not expect the default allocator to necessarily visit every (logical) element.

For visiting elements of the underlying storage, we could expose something like `materilized_begin`/`materialized_end`/`materialized` with iterator/range types exposed through typedefs (`using`). This way if someone decided to use the actual type, they can do so without the code being broken when we decide to change the underlying storage type. 

https://github.com/llvm/llvm-project/pull/66430