[llvm] Implement reserveAllocationSpace for SectionMemoryManager (PR #71968)

[Michael Smith via llvm-commits]

================
@@ -18,6 +18,63 @@
 
 namespace llvm {
 
+static uint64_t requiredPageSize(uintptr_t Size, Align Alignment) {
+  static const size_t PageSize = sys::Process::getPageSizeEstimate();
+  // Use the same calculation as allocateSection because we need to be able to
+  // satisfy it.
+  uint64_t RequiredSize = alignTo(Size, Alignment);
+  // Round up to the nearest page size. Blocks must be page-aligned.
+  return PageSize * ((RequiredSize + PageSize - 1) / PageSize);
+}
+
+void SectionMemoryManager::reserveAllocationSpace(
+    uintptr_t CodeSize, Align CodeAlign, uintptr_t RODataSize,
+    Align RODataAlign, uintptr_t RWDataSize, Align RWDataAlign) {
+  if (CodeSize == 0 && RODataSize == 0 && RWDataSize == 0)
+    return;
+
+  // Get space required for each section.
+  uint64_t RequiredCodeSize = requiredPageSize(CodeSize, CodeAlign);
----------------
MikaelSmith wrote:

I have concerns about this and line 46:
```
uintptr_t RequiredSize = Alignment * ((Size + Alignment - 1) / Alignment + 1);
```
That calculation buffers by 1 Alignment, to ensure we always reserve enough space to align on address. It's inefficient, but not terribly.

If this sums up many calls to allocateCodeSection/allocateDataSection, does the caller also sum appropriately to cover those buffers?

https://github.com/llvm/llvm-project/pull/71968