Index: llvm/docs/LoopTerminology.rst
===================================================================
--- llvm/docs/LoopTerminology.rst
+++ llvm/docs/LoopTerminology.rst
@@ -169,11 +169,13 @@
 A program is in Loop Closed SSA Form if it is in SSA form
 and all values that are defined in a loop are used only inside
 this loop.
+
 Programs written in LLVM IR are always in SSA form but not necessarily
 in LCSSA. To achieve the latter, single entry PHI nodes are inserted
-at the end of the loops for all values that are live
-across the loop boundary [#lcssa-construction]_.
-In particular, consider the following loop:
+at the end of the loop (specifically at its exit blocks) for all values
+that are live across the loop boundary [#lcssa-construction]_
+in order to "close" these values inside it.  In particular, consider the
+following loop:
 
 .. code-block:: C
 
@@ -214,6 +216,19 @@
 After the loop optimizations are done, these extra phi nodes
 will be deleted by :ref:`-instcombine <passes-instcombine>`.
 
+Note that an exit block is outside of a loop, so how can such a phi "close"
+the value inside the loop since it uses it outside of it ? For phi nodes, as
+`mentioned in the LangRef <https://llvm.org/docs/LangRef.html#id311>`_:
+"the use of each incoming value is deemed to occur on the edge from the
+corresponding predecessor block to the current block". To that end,
+formally the use truly happens at the boundary of the loop and for
+the purposes of LCSSA, we consider it happens inside it.
+
+However, an edge doesn't actually contain any IR, so in source code,
+we have to choose a convention of whether the use happens in
+the current block or in the respective predecessor.
+The latter is chosen (across all LLVM source code) for practical purposes.
+
 The major benefit of this transformation is that it makes many other
 loop optimizations simpler.
 
Index: llvm/lib/Analysis/LoopInfo.cpp
===================================================================
--- llvm/lib/Analysis/LoopInfo.cpp
+++ llvm/lib/Analysis/LoopInfo.cpp
@@ -431,6 +431,10 @@
     for (const Use &U : I.uses()) {
       const Instruction *UI = cast<Instruction>(U.getUser());
       const BasicBlock *UserBB = UI->getParent();
+
+      // For practical purposes, we consider that the use in a PHI
+      // occurs in the respective predecessor block. For more info,
+      // see the `phi` doc in LangRef and the LCSSA doc.
       if (const PHINode *P = dyn_cast<PHINode>(UI))
         UserBB = P->getIncomingBlock(U);
 
Index: llvm/lib/Transforms/Utils/LCSSA.cpp
===================================================================
--- llvm/lib/Transforms/Utils/LCSSA.cpp
+++ llvm/lib/Transforms/Utils/LCSSA.cpp
@@ -111,6 +111,10 @@
     for (Use &U : I->uses()) {
       Instruction *User = cast<Instruction>(U.getUser());
       BasicBlock *UserBB = User->getParent();
+
+      // For practical purposes, we consider that the use in a PHI
+      // occurs in the respective predecessor block. For more info,
+      // see the `phi` doc in LangRef and the LCSSA doc.
       if (auto *PN = dyn_cast<PHINode>(User))
         UserBB = PN->getIncomingBlock(U);
 
@@ -160,7 +164,12 @@
                                       I->getName() + ".lcssa");
       // Get the debug location from the original instruction.
       PN->setDebugLoc(I->getDebugLoc());
-      // Add inputs from inside the loop for this PHI.
+
+      // Add inputs from inside the loop for this PHI. This is valid
+      // because `I` dominates `ExitBB` (checked above).  This implies
+      // that every incoming block/edge is dominated by `I` as well,
+      // i.e. we can add uses of `I` to those incoming edges/append to the the incoming
+      // blocks without violating the SSA dominance property.
       for (BasicBlock *Pred : PredCache.get(ExitBB)) {
         PN->addIncoming(I, Pred);
 
@@ -194,15 +203,19 @@
     // Rewrite all uses outside the loop in terms of the new PHIs we just
     // inserted.
     for (Use *UseToRewrite : UsesToRewrite) {
-      // If this use is in an exit block, rewrite to use the newly inserted PHI.
-      // This is required for correctness because SSAUpdate doesn't handle uses
-      // in the same block.  It assumes the PHI we inserted is at the end of the
-      // block.
       Instruction *User = cast<Instruction>(UseToRewrite->getUser());
       BasicBlock *UserBB = User->getParent();
+
+      // For practical purposes, we consider that the use in a PHI
+      // occurs in the respective predecessor block. For more info,
+      // see the `phi` doc in LangRef and the LCSSA doc.
       if (auto *PN = dyn_cast<PHINode>(User))
         UserBB = PN->getIncomingBlock(*UseToRewrite);
 
+      // If this use is in an exit block, rewrite to use the newly inserted PHI.
+      // This is required for correctness because SSAUpdate doesn't handle uses
+      // in the same block.  It assumes the PHI we inserted is at the end of the
+      // block.
       if (isa<PHINode>(UserBB->begin()) && isExitBlock(UserBB, ExitBlocks)) {
         UseToRewrite->set(&UserBB->front());
         continue;