[llvm-commits] CVS: llvm/lib/Transforms/Scalar/LCSSA.cpp

Owen Anderson resistor at mac.com
Fri May 26 06:58:51 PDT 2006



Changes in directory llvm/lib/Transforms/Scalar:

LCSSA.cpp added (r1.1)
---
Log message:

Skeletal LCSSA pass.  This is currently non-functional.  Expect functionality
and documentation updates soo.


---
Diffs of the changes:  (+159 -0)

 LCSSA.cpp |  159 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 159 insertions(+)


Index: llvm/lib/Transforms/Scalar/LCSSA.cpp
diff -c /dev/null llvm/lib/Transforms/Scalar/LCSSA.cpp:1.1
*** /dev/null	Fri May 26 08:58:36 2006
--- llvm/lib/Transforms/Scalar/LCSSA.cpp	Fri May 26 08:58:26 2006
***************
*** 0 ****
--- 1,159 ----
+ //===-- LCSSA.cpp - Convert loops into loop-closed SSA form         ------===//
+ //
+ //                     The LLVM Compiler Infrastructure
+ //
+ // This file was developed by Owen Anderson and is distributed under the
+ // University of Illinois Open Source License. See LICENSE.TXT for details.
+ //
+ //===----------------------------------------------------------------------===//
+ //
+ // This pass transforms loops by placing phi nodes at the end of the loops for
+ // all values that are live across the loop boundary.  For example, it turns
+ // the left into the right code:
+ // 
+ // for (...)                for (...)
+ //   if (c)                   if(c)
+ //     X1 = ...                 X1 = ...
+ //   else                     else
+ //     X2 = ...                 X2 = ...
+ //   X3 = phi(X1, X2)         X3 = phi(X1, X2)
+ // ... = X3 + 4              X4 = phi(X3)
+ //                           ... = X4 + 4
+ //
+ // This is still valid LLVM; the extra phi nodes are purely redundant, and will
+ // be trivially eliminated by InstCombine.  The major benefit of this 
+ // transformation is that it makes many other loop optimizations, such as 
+ // LoopUnswitching, simpler.
+ //
+ //===----------------------------------------------------------------------===//
+ 
+ #include "llvm/Pass.h"
+ #include "llvm/Function.h"
+ #include "llvm/Instructions.h"
+ #include "llvm/Analysis/LoopInfo.h"
+ #include "llvm/Support/CFG.h"
+ #include "llvm/Transforms/Scalar.h"
+ #include "llvm/Transforms/Utils/BasicBlockUtils.h"
+ #include <iostream>
+ #include <set>
+ #include <vector>
+ 
+ using namespace llvm;
+ 
+ namespace {
+   class LCSSA : public FunctionPass {
+   public:
+     LoopInfo *LI;  // Loop information
+     
+     // LoopProcessWorklist - List of loops we need to process.
+     std::vector<Loop*> LoopProcessWorklist;
+     
+     virtual bool runOnFunction(Function &F);
+     
+     bool visitLoop(Loop *L, Value *V);
+     
+     /// This transformation requires natural loop information & requires that
+     /// loop preheaders be inserted into the CFG...
+     ///
+     virtual void getAnalysisUsage(AnalysisUsage &AU) const {
+       AU.addRequiredID(LoopSimplifyID);
+       AU.addPreservedID(LoopSimplifyID);
+       AU.addRequired<LoopInfo>();
+       AU.addPreserved<LoopInfo>();
+     }
+   private:
+     void addSubloopsToWorklist(Loop* L);
+     std::set<Value*> loopValuesUsedOutsideLoop(Loop *L);
+   };
+   
+   RegisterOpt<LCSSA> X("lcssa", "Loop-Closed SSA Form Pass");
+ }
+ 
+ FunctionPass *llvm::createLCSSAPass() { return new LCSSA(); }
+ 
+ bool LCSSA::runOnFunction(Function &F) {
+   bool changed = false;
+   LI = &getAnalysis<LoopInfo>();
+     
+   for (LoopInfo::iterator I = LI->begin(), E = LI->end(); I != E; ++I) {
+     addSubloopsToWorklist(*I);
+     LoopProcessWorklist.push_back(*I);
+   }
+       
+   for (std::vector<Loop*>::iterator I = LoopProcessWorklist.begin(),
+        E = LoopProcessWorklist.end(); I != E; ++I) {
+     std::set<Value*> AffectedValues = loopValuesUsedOutsideLoop(*I);
+     if (!AffectedValues.empty()) {
+       for (std::set<Value*>::iterator VI = AffectedValues.begin(),
+            VE = AffectedValues.end(); VI != VE; ++VI)
+         changed |= visitLoop(*I, *VI);
+     }
+   }
+       
+   return changed;
+ }
+ 
+ bool LCSSA::visitLoop(Loop *L, Value* V) {
+   // We will be doing lots of "loop contains block" queries.  Loop::contains is
+   // linear time, use a set to speed this up.
+   std::set<BasicBlock*> LoopBlocks;
+ 
+   for (Loop::block_iterator BB = L->block_begin(), E = L->block_end();
+        BB != E; ++BB)
+     LoopBlocks.insert(*BB);
+   
+   std::vector<BasicBlock*> exitBlocks;
+   L->getExitBlocks(exitBlocks);
+   
+   for (std::vector<BasicBlock*>::iterator BBI = exitBlocks.begin(),
+        BBE = exitBlocks.end(); BBI != BBE; ++BBI) {
+     PHINode *phi = new PHINode(V->getType(), "lcssa");
+     (*BBI)->getInstList().insert((*BBI)->front(), phi);
+     
+     for (pred_iterator PI = pred_begin(*BBI), PE = pred_end(*BBI); PI != PE;
+          ++PI)
+       phi->addIncoming(V, *PI);
+   }
+   
+   for (Value::use_iterator UI = V->use_begin(), UE = V->use_end(); UI != UE;
+        ++UI) {
+     BasicBlock *UserBB = cast<Instruction>(*UI)->getParent();
+     if (!LoopBlocks.count(UserBB))
+       ; // FIXME: This should update the SSA form through the rest of the graph.
+   }
+   
+   return false;
+ }
+ 
+ void LCSSA::addSubloopsToWorklist(Loop* L) {
+   for (Loop::iterator I = L->begin(), E = L->end(); I != E; ++I) {
+     addSubloopsToWorklist(*I);
+     LoopProcessWorklist.push_back(*I);
+   }
+ }
+ 
+ /// loopValuesUsedOutsideLoop - Return true if there are any values defined in
+ /// the loop that are used by instructions outside of it.
+ std::set<Value*> LCSSA::loopValuesUsedOutsideLoop(Loop *L) {
+   std::set<Value*> AffectedValues;
+ 
+   // We will be doing lots of "loop contains block" queries.  Loop::contains is
+   // linear time, use a set to speed this up.
+   std::set<BasicBlock*> LoopBlocks;
+ 
+   for (Loop::block_iterator BB = L->block_begin(), E = L->block_end();
+        BB != E; ++BB)
+     LoopBlocks.insert(*BB);
+   
+   for (Loop::block_iterator BB = L->block_begin(), E = L->block_end();
+        BB != E; ++BB) {
+     for (BasicBlock::iterator I = (*BB)->begin(), E = (*BB)->end(); I != E; ++I)
+       for (Value::use_iterator UI = I->use_begin(), E = I->use_end(); UI != E;
+            ++UI) {
+         BasicBlock *UserBB = cast<Instruction>(*UI)->getParent();
+         if (!LoopBlocks.count(UserBB))
+           AffectedValues.insert(I);
+       }
+   }
+   return AffectedValues;
+ }
\ No newline at end of file






More information about the llvm-commits mailing list