Creating an LLVM Project

Brian Gaeke · $1

% cd build
  % ../src/configure --prefix=$CFEINSTALL --disable-nls --disable-shared \
@@ -94,6 +99,24 @@
  % gmake all; gmake install
 
+ % cd build
+ % ../src/configure --prefix=$CFEINSTALL --disable-nls --disable-shared \
+   --enable-languages=c,c++ sparcv9-sun-solaris2.8
+ % gmake all-gcc
+ % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
+ % gmake all; gmake install
+
  % cd build
  % ../src/configure --prefix=$CFEINSTALL --disable-nls --disable-shared \
-   --enable-languages=c,c++ sparcv9-sun-solaris2.8
+   --enable-languages=c,c++ --host=sparcv9-sun-solaris2.8
  % gmake all-gcc
  % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
  % gmake all; gmake install
@@ -203,7 +203,7 @@
   Brian Gaeke
   The LLVM Compiler Infrastructure
   

-  Last modified: $Date: 2003/12/08 19:59:14 $
+  Last modified: $Date: 2003/12/10 15:27:29 $
 
 
 




From criswell at cs.uiuc.edu  Wed Dec 10 09:32:02 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 09:32:02 2003
Subject: [llvm-commits] CVS: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
Message-ID: <200312101531.JAA04677@choi.cs.uiuc.edu>


Changes in directory llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++:

Makefile updated: 1.4 -> 1.5

---
Log message:

Added the -Wno-deprecated flag to avoid spurious error messages.



---
Diffs of the changes:  (+1 -0)

Index: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
diff -u llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.4 llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.5

--- llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.4	Mon Dec  8 17:00:01 2003
+++ llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile	Wed Dec 10 09:31:43 2003
@@ -1,4 +1,5 @@
 LEVEL = ../../../../..
+CXXFLAGS += -Wno-deprecated
 CPPFLAGS += -Wno-deprecated
 LDFLAGS += -lpthread -lstdc++
 LIBS += -lstdc++




From lattner at cs.uiuc.edu  Wed Dec 10 09:35:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 09:35:02 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/AliasAnalysisEvaluator.cpp
Message-ID: <200312101534.JAA05126@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

AliasAnalysisEvaluator.cpp updated: 1.9 -> 1.10

---
Log message:

Finegrainify namespacification
Provide a context module to WriteAsOperand


---
Diffs of the changes:  (+8 -10)

Index: llvm/lib/Analysis/AliasAnalysisEvaluator.cpp
diff -u llvm/lib/Analysis/AliasAnalysisEvaluator.cpp:1.9 llvm/lib/Analysis/AliasAnalysisEvaluator.cpp:1.10
--- llvm/lib/Analysis/AliasAnalysisEvaluator.cpp:1.9	Tue Nov 11 16:41:31 2003
+++ llvm/lib/Analysis/AliasAnalysisEvaluator.cpp	Wed Dec 10 09:33:59 2003
@@ -25,8 +25,7 @@
 #include "llvm/Assembly/Writer.h"
 #include "Support/CommandLine.h"
 #include 
-
-namespace llvm {
+using namespace llvm;
 
 namespace {
   cl::opt PrintNo  ("print-no-aliases", cl::ReallyHidden);
@@ -51,11 +50,12 @@
   X("aa-eval", "Exhaustive Alias Analysis Precision Evaluator");
 }
 
-static inline void PrintResults(const char *Msg, bool P, Value *V1, Value *V2) {
+static inline void PrintResults(const char *Msg, bool P, Value *V1, Value *V2,
+                                Module *M) {
   if (P) {
     std::cerr << "  " << Msg << ":\t";
-    WriteAsOperand(std::cerr, V1) << ", ";
-    WriteAsOperand(std::cerr, V2) << "\n";
+    WriteAsOperand(std::cerr, V1, true, true, M) << ", ";
+    WriteAsOperand(std::cerr, V2, true, true, M) << "\n";
   }
 }
 
@@ -85,13 +85,13 @@
     for (std::set::iterator I2 = Pointers.begin(); I2 != I1; ++I2)
       switch (AA.alias(*I1, 0, *I2, 0)) {
       case AliasAnalysis::NoAlias:
-        PrintResults("No", PrintNo, *I1, *I2);
+        PrintResults("No", PrintNo, *I1, *I2, F.getParent());
         ++No; break;
       case AliasAnalysis::MayAlias:
-        PrintResults("May", PrintMay, *I1, *I2);
+        PrintResults("May", PrintMay, *I1, *I2, F.getParent());
         ++May; break;
       case AliasAnalysis::MustAlias:
-        PrintResults("Must", PrintMust, *I1, *I2);
+        PrintResults("Must", PrintMust, *I1, *I2, F.getParent());
         ++Must; break;
       default:
         std::cerr << "Unknown alias query result!\n";
@@ -116,5 +116,3 @@
             << May*100/Sum << "%/" << Must*100/Sum<<"%\n";
   return false;
 }
-
-} // End llvm namespace




From lattner at cs.uiuc.edu  Wed Dec 10 09:35:05 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 09:35:05 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/AliasAnalysisCounter.cpp
Message-ID: <200312101534.JAA05133@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

AliasAnalysisCounter.cpp updated: 1.6 -> 1.7

---
Log message:

Finegrainify namespacification


---
Diffs of the changes:  (+1 -4)

Index: llvm/lib/Analysis/AliasAnalysisCounter.cpp
diff -u llvm/lib/Analysis/AliasAnalysisCounter.cpp:1.6 llvm/lib/Analysis/AliasAnalysisCounter.cpp:1.7
--- llvm/lib/Analysis/AliasAnalysisCounter.cpp:1.6	Tue Nov 11 16:41:31 2003
+++ llvm/lib/Analysis/AliasAnalysisCounter.cpp	Wed Dec 10 09:34:03 2003
@@ -15,8 +15,7 @@
 #include "llvm/Analysis/AliasAnalysis.h"
 #include "llvm/Pass.h"
 #include 
-
-namespace llvm {
+using namespace llvm;
 
 namespace {
   class AliasAnalysisCounter : public Pass, public AliasAnalysis {
@@ -110,5 +109,3 @@
   X("count-aa", "Count Alias Analysis Query Responses");
   RegisterAnalysisGroup Y;
 }
-
-} // End llvm namespace




From criswell at cs.uiuc.edu  Wed Dec 10 09:44:01 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 09:44:01 2003
Subject: [llvm-commits] CVS: llvm/runtime/GCCLibraries/libpthread/pthread.c
Message-ID: <200312101543.JAA04783@choi.cs.uiuc.edu>


Changes in directory llvm/runtime/GCCLibraries/libpthread:

pthread.c updated: 1.1 -> 1.2

---
Log message:

Removed the pthread_once() function because it doesn't actually do
what it's supposed to do.
Just leave this library as a dummy library for now until we have time to
port a real pthread library (or get glibc working).



---
Diffs of the changes:  (+0 -7)

Index: llvm/runtime/GCCLibraries/libpthread/pthread.c
diff -u llvm/runtime/GCCLibraries/libpthread/pthread.c:1.1 llvm/runtime/GCCLibraries/libpthread/pthread.c:1.2
--- llvm/runtime/GCCLibraries/libpthread/pthread.c:1.1	Wed Dec 10 08:46:36 2003
+++ llvm/runtime/GCCLibraries/libpthread/pthread.c	Wed Dec 10 09:43:07 2003
@@ -1,8 +1 @@
-#include 
-
-int
-pthread_once (pthread_once_t * control, void (*func) (void))
-{
-  return 0;
-}
 




From lattner at cs.uiuc.edu  Wed Dec 10 09:57:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 09:57:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LICM.cpp
Message-ID: <200312101556.JAA05670@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.44 -> 1.45

---
Log message:

Make LICM itself a bit more efficient, and make the generated code more efficient too: don't insert a store in every exit block, because a particular block may be exited to more than once by a loop


---
Diffs of the changes:  (+26 -21)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.44 llvm/lib/Transforms/Scalar/LICM.cpp:1.45
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.44	Wed Dec 10 00:41:05 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Wed Dec 10 09:56:24 2003
@@ -571,9 +571,7 @@
   }
   
   // Scan the basic blocks in the loop, replacing uses of our pointers with
-  // uses of the allocas in question.  If we find a branch that exits the
-  // loop, make sure to put reload code into all of the successors of the
-  // loop.
+  // uses of the allocas in question.
   //
   const std::vector &LoopBBs = CurLoop->getBlocks();
   for (std::vector::const_iterator I = LoopBBs.begin(),
@@ -593,26 +591,33 @@
           S->setOperand(1, I->second);    // Rewrite store instruction...
       }
     }
+  }
+
+  // Now that the body of the loop uses the allocas instead of the original
+  // memory locations, insert code to copy the alloca value back into the
+  // original memory location on all exits from the loop.  Note that we only
+  // want to insert one copy of the code in each exit block, though the loop may
+  // exit to the same block more than once.
+  //
+  std::set ProcessedBlocks;
 
-    // Check to see if any successors of this block are outside of the loop.
-    // If so, we need to copy the value from the alloca back into the memory
-    // location...
-    //
-    for (succ_iterator SI = succ_begin(*I), SE = succ_end(*I); SI != SE; ++SI)
-      if (!CurLoop->contains(*SI)) {
-        // Copy all of the allocas into their memory locations...
-        BasicBlock::iterator BI = (*SI)->begin();
-        while (isa(*BI))
-          ++BI;             // Skip over all of the phi nodes in the block...
-        Instruction *InsertPos = BI;
-        for (unsigned i = 0, e = PromotedValues.size(); i != e; ++i) {
-          // Load from the alloca...
-          LoadInst *LI = new LoadInst(PromotedValues[i].first, "", InsertPos);
-          // Store into the memory we promoted...
-          new StoreInst(LI, PromotedValues[i].second, InsertPos);
-        }
+  const std::vector &ExitBlocks = CurLoop->getExitBlocks();
+  for (unsigned i = 0, e = ExitBlocks.size(); i != e; ++i)
+    if (!ProcessedBlocks.count(ExitBlocks[i])) {
+      ProcessedBlocks.insert(ExitBlocks[i]);
+    
+      // Copy all of the allocas into their memory locations...
+      BasicBlock::iterator BI = ExitBlocks[i]->begin();
+      while (isa(*BI))
+        ++BI;             // Skip over all of the phi nodes in the block...
+      Instruction *InsertPos = BI;
+      for (unsigned i = 0, e = PromotedValues.size(); i != e; ++i) {
+        // Load from the alloca...
+        LoadInst *LI = new LoadInst(PromotedValues[i].first, "", InsertPos);
+        // Store into the memory we promoted...
+        new StoreInst(LI, PromotedValues[i].second, InsertPos);
       }
-  }
+    }
 
   // Now that we have done the deed, use the mem2reg functionality to promote
   // all of the new allocas we just created into real SSA registers...




From lattner at cs.uiuc.edu  Wed Dec 10 10:58:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 10:58:02 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LICM.cpp
Message-ID: <200312101657.KAA06991@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.45 -> 1.46

---
Log message:

Avoid performing two identical lookups when one will suffice


---
Diffs of the changes:  (+1 -3)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.45 llvm/lib/Transforms/Scalar/LICM.cpp:1.46
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.45	Wed Dec 10 09:56:24 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Wed Dec 10 10:57:24 2003
@@ -603,9 +603,7 @@
 
   const std::vector &ExitBlocks = CurLoop->getExitBlocks();
   for (unsigned i = 0, e = ExitBlocks.size(); i != e; ++i)
-    if (!ProcessedBlocks.count(ExitBlocks[i])) {
-      ProcessedBlocks.insert(ExitBlocks[i]);
-    
+    if (ProcessedBlocks.insert(ExitBlocks[i]).second) {
       // Copy all of the allocas into their memory locations...
       BasicBlock::iterator BI = ExitBlocks[i]->begin();
       while (isa(*BI))




From lattner at cs.uiuc.edu  Wed Dec 10 10:59:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 10:59:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LICM.cpp
Message-ID: <200312101658.KAA07325@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.46 -> 1.47

---
Log message:

Simplify code


---
Diffs of the changes:  (+1 -6)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.46 llvm/lib/Transforms/Scalar/LICM.cpp:1.47
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.46	Wed Dec 10 10:57:24 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Wed Dec 10 10:58:24 2003
@@ -437,10 +437,8 @@
       BasicBlock *ExitBlock = ExitBlocks[i];
 
       if (isExitBlockDominatedByBlockInLoop(ExitBlock, InstOrigBB)) {
-        std::set::iterator SI =
-          InsertedBlocks.lower_bound(ExitBlock);
         // If we haven't already processed this exit block, do so now.
-        if (SI == InsertedBlocks.end() || *SI != ExitBlock) {
+        if (InsertedBlocks.insert(ExitBlock).second) {
           // Insert the code after the last PHI node...
           BasicBlock::iterator InsertPt = ExitBlock->begin();
           while (isa(InsertPt)) ++InsertPt;
@@ -461,9 +459,6 @@
           
           // Now that we have inserted the instruction, store it into the alloca
           new StoreInst(New, AI, InsertPt);
-          
-          // Remember we processed this block
-          InsertedBlocks.insert(SI, ExitBlock);
         }
       }
     }




From lattner at cs.uiuc.edu  Wed Dec 10 11:10:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 11:10:01 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll
Message-ID: <200312101709.LAA07432@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/LoopSimplify:

2003-12-10-ExitBlocksProblem.ll added (r1.1)

---
Log message:

New testcase which caused a bunch of tests to break last night with more 
aggressive LICM


---
Diffs of the changes:  (+42 -0)

Index: llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll
diff -c /dev/null llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll:1.1
*** /dev/null	Wed Dec 10 11:09:45 2003
--- llvm/test/Regression/Transforms/LoopSimplify/2003-12-10-ExitBlocksProblem.ll	Wed Dec 10 11:09:35 2003
***************
*** 0 ****
--- 1,42 ----
+ ; LoopSimplify is breaking LICM on this testcase because the exit blocks from
+ ; the loop are reachable from more than just the exit nodes: the exit blocks
+ ; have predecessors from outside of the loop!
+ ;
+ ; This is distilled from a monsterous crafty example.
+ 
+ ; RUN: llvm-as < %s | opt -licm | lli
+ 
+ %G = weak global int 0		;  [#uses=13]
+ 
+ implementation   ; Functions:
+ 
+ int %main() {
+ entry:
+ 	store int 123, int* %G
+ 	br label %loopentry.i
+ 
+ loopentry.i:		; preds = %entry, %endif.1.i
+ 	%tmp.0.i = load int* %G		;  [#uses=1]
+ 	%tmp.1.i = seteq int %tmp.0.i, 123		;  [#uses=1]
+ 	br bool %tmp.1.i, label %Out.i, label %endif.0.i
+ 
+ endif.0.i:		; preds = %loopentry.i
+ 	%tmp.3.i = load int* %G		;  [#uses=1]
+ 	%tmp.4.i = seteq int %tmp.3.i, 126		;  [#uses=1]
+ 	br bool %tmp.4.i, label %ExitBlock.i, label %endif.1.i
+ 
+ endif.1.i:		; preds = %endif.0.i
+ 	%tmp.6.i = load int* %G		;  [#uses=1]
+ 	%inc.i = add int %tmp.6.i, 1		;  [#uses=1]
+ 	store int %inc.i, int* %G
+ 	br label %loopentry.i
+ 
+ Out.i:		; preds = %loopentry.i
+ 	store int 0, int* %G
+ 	br label %ExitBlock.i
+ 
+ ExitBlock.i:		; preds = %endif.0.i, %Out.i
+ 	%tmp.7.i = load int* %G		;  [#uses=1]
+ 	ret int %tmp.7.i
+ }
+ 




From lattner at cs.uiuc.edu  Wed Dec 10 11:24:33 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 11:24:33 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LoopSimplify.cpp
Message-ID: <200312101720.LAA09075@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LoopSimplify.cpp updated: 1.27 -> 1.28

---
Log message:

Finegrainify namespacification
Fix bug: LoopSimplify/2003-12-10-ExitBlocksProblem.ll



---
Diffs of the changes:  (+21 -22)

Index: llvm/lib/Transforms/Scalar/LoopSimplify.cpp
diff -u llvm/lib/Transforms/Scalar/LoopSimplify.cpp:1.27 llvm/lib/Transforms/Scalar/LoopSimplify.cpp:1.28
--- llvm/lib/Transforms/Scalar/LoopSimplify.cpp:1.27	Tue Dec  9 17:12:55 2003
+++ llvm/lib/Transforms/Scalar/LoopSimplify.cpp	Wed Dec 10 11:20:35 2003
@@ -17,8 +17,9 @@
 //
 // Loop exit-block insertion guarantees that all exit blocks from the loop
 // (blocks which are outside of the loop that have predecessors inside of the
-// loop) are dominated by the loop header.  This simplifies transformations such
-// as store-sinking that are built into LICM.
+// loop) only have predecessors from inside of the loop (and are thus dominated
+// by the loop header).  This simplifies transformations such as store-sinking
+// that are built into LICM.
 //
 // This pass also guarantees that loops will have exactly one backedge.
 //
@@ -42,12 +43,11 @@
 #include "Support/SetOperations.h"
 #include "Support/Statistic.h"
 #include "Support/DepthFirstIterator.h"
-
-namespace llvm {
+using namespace llvm;
 
 namespace {
   Statistic<>
-  NumInserted("loopsimplify", "Number of pre-header blocks inserted");
+  NumInserted("loopsimplify", "Number of pre-header or exit blocks inserted");
 
   struct LoopSimplify : public FunctionPass {
     virtual bool runOnFunction(Function &F);
@@ -81,8 +81,8 @@
 }
 
 // Publically exposed interface to pass...
-const PassInfo *LoopSimplifyID = X.getPassInfo();
-Pass *createLoopSimplifyPass() { return new LoopSimplify(); }
+const PassInfo *llvm::LoopSimplifyID = X.getPassInfo();
+Pass *llvm::createLoopSimplifyPass() { return new LoopSimplify(); }
 
 /// runOnFunction - Run down all loops in the CFG (recursively, but we could do
 /// it in any convenient order) inserting preheaders...
@@ -111,18 +111,20 @@
     Changed = true;
   }
 
-  // Regardless of whether or not we added a preheader to the loop we must
-  // guarantee that the preheader dominates all exit nodes.  If there are any
-  // exit nodes not dominated, split them now.
-  DominatorSet &DS = getAnalysis();
-  BasicBlock *Header = L->getHeader();
-  for (unsigned i = 0, e = L->getExitBlocks().size(); i != e; ++i)
-    if (!DS.dominates(Header, L->getExitBlocks()[i])) {
-      RewriteLoopExitBlock(L, L->getExitBlocks()[i]);
-      assert(DS.dominates(Header, L->getExitBlocks()[i]) &&
-             "RewriteLoopExitBlock failed?");
-      NumInserted++;
-      Changed = true;
+  // Next, check to make sure that all exit nodes of the loop only have
+  // predecessors that are inside of the loop.  This check guarantees that the
+  // loop preheader/header will dominate the exit blocks.  If the exit block has
+  // predecessors from outside of the loop, split the edge now.
+  for (unsigned i = 0, e = L->getExitBlocks().size(); i != e; ++i) {
+    BasicBlock *ExitBlock = L->getExitBlocks()[i];
+    for (pred_iterator PI = pred_begin(ExitBlock), PE = pred_end(ExitBlock);
+         PI != PE; ++PI)
+      if (!L->contains(*PI)) {
+        RewriteLoopExitBlock(L, ExitBlock);
+        NumInserted++;
+        Changed = true;
+        break;
+      }
     }
 
   // The preheader may have more than two predecessors at this point (from the
@@ -342,8 +344,6 @@
 
 void LoopSimplify::RewriteLoopExitBlock(Loop *L, BasicBlock *Exit) {
   DominatorSet &DS = getAnalysis();
-  assert(!DS.dominates(L->getHeader(), Exit) &&
-         "Loop already dominates exit block??");
   assert(std::find(L->getExitBlocks().begin(), L->getExitBlocks().end(), Exit)
          != L->getExitBlocks().end() && "Not a current exit block!");
   
@@ -576,4 +576,3 @@
   }
 }
 
-} // End llvm namespace




From lattner at cs.uiuc.edu  Wed Dec 10 11:58:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 11:58:01 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-IndVarDeadCode.ll
Message-ID: <200312101757.LAA13760@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/IndVarsSimplify:

2003-12-10-IndVarDeadCode.ll added (r1.1)

---
Log message:

New testcase


---
Diffs of the changes:  (+23 -0)

Index: llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-IndVarDeadCode.ll
diff -c /dev/null llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-IndVarDeadCode.ll:1.1
*** /dev/null	Wed Dec 10 11:57:48 2003
--- llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-IndVarDeadCode.ll	Wed Dec 10 11:57:38 2003
***************
*** 0 ****
--- 1,23 ----
+ ; The induction variable canonicalization pass shouldn't leave dead
+ ; instructions laying around!
+ ;
+ ; RUN: llvm-as < %s | opt -indvars | llvm-dis | not grep '#uses=0'
+ 
+ int %mul(int %x, int %y) {
+ entry:
+ 	br label %tailrecurse
+ 
+ tailrecurse:		; preds = %entry, %endif
+ 	%accumulator.tr = phi int [ %x, %entry ], [ %tmp.9, %endif ]		;  [#uses=2]
+ 	%y.tr = phi int [ %y, %entry ], [ %tmp.8, %endif ]		;  [#uses=2]
+ 	%tmp.1 = seteq int %y.tr, 0		;  [#uses=1]
+ 	br bool %tmp.1, label %return, label %endif
+ 
+ endif:		; preds = %tailrecurse
+ 	%tmp.8 = add int %y.tr, -1		;  [#uses=1]
+ 	%tmp.9 = add int %accumulator.tr, %x		;  [#uses=1]
+ 	br label %tailrecurse
+ 
+ return:		; preds = %tailrecurse
+ 	ret int %accumulator.tr
+ }




From criswell at cs.uiuc.edu  Wed Dec 10 12:07:01 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 12:07:01 2003
Subject: [llvm-commits] CVS: llvm/lib/CWriter/Writer.cpp
Message-ID: <200312101806.MAA06186@choi.cs.uiuc.edu>


Changes in directory llvm/lib/CWriter:

Writer.cpp updated: 1.148 -> 1.149

---
Log message:

Fixed the CBE on Solaris/Sparc.  We need to define the return value of
the write() system call because it returns 64 bits on Solaris 64 bit,
and an implicit return value of int says it returns 32 bits.
Admittedly, this is a bit of a hack.




---
Diffs of the changes:  (+8 -1)

Index: llvm/lib/CWriter/Writer.cpp
diff -u llvm/lib/CWriter/Writer.cpp:1.148 llvm/lib/CWriter/Writer.cpp:1.149
--- llvm/lib/CWriter/Writer.cpp:1.148	Tue Nov 25 18:09:17 2003
+++ llvm/lib/CWriter/Writer.cpp	Wed Dec 10 12:06:34 2003
@@ -1087,13 +1087,20 @@
 
 
 void CWriter::visitUnwindInst(UnwindInst &I) {
+  // Determine the return size of write() based on the data model.
+#ifdef _LP64
+      const char * writedecl = "    extern signed long long write();\n";
+#else
+      const char * writedecl = "    extern write();\n";
+#endif
+
   // The unwind instructions causes a control flow transfer out of the current
   // function, unwinding the stack until a caller who used the invoke
   // instruction is found.  In this context, we code generated the invoke
   // instruction to add an entry to the top of the jmpbuf_list.  Thus, here we
   // just have to longjmp to the specified handler.
   Out << "  if (__llvm_jmpbuf_list == 0) {  /* unwind */\n"
-      << "    extern write();\n"
+      << writedecl
       << "    ((void (*)(int, void*, unsigned))write)(2,\n"
       << "           \"throw found with no handler!\\n\", 31); abort();\n"
       << "  }\n"




From lattner at cs.uiuc.edu  Wed Dec 10 12:07:04 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 12:07:04 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/IndVarSimplify.cpp
Message-ID: <200312101806.MAA14848@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

IndVarSimplify.cpp updated: 1.42 -> 1.43

---
Log message:

Finegrainify namespacification
Reorder #includes
Implement: IndVarsSimplify/2003-12-10-IndVarDeadCode.ll


---
Diffs of the changes:  (+27 -8)

Index: llvm/lib/Transforms/Scalar/IndVarSimplify.cpp
diff -u llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.42 llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.43
--- llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.42	Tue Nov 11 16:41:33 2003
+++ llvm/lib/Transforms/Scalar/IndVarSimplify.cpp	Wed Dec 10 12:06:47 2003
@@ -15,18 +15,18 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/Transforms/Scalar.h"
-#include "llvm/Analysis/InductionVariable.h"
-#include "llvm/Analysis/LoopInfo.h"
+#include "llvm/Constants.h"
+#include "llvm/Type.h"
 #include "llvm/iPHINode.h"
 #include "llvm/iOther.h"
-#include "llvm/Type.h"
-#include "llvm/Constants.h"
+#include "llvm/Analysis/InductionVariable.h"
+#include "llvm/Analysis/LoopInfo.h"
 #include "llvm/Support/CFG.h"
+#include "llvm/Transforms/Utils/Local.h"
 #include "Support/Debug.h"
 #include "Support/Statistic.h"
 #include "Support/STLExtras.h"
-
-namespace llvm {
+using namespace llvm;
 
 namespace {
   Statistic<> NumRemoved ("indvars", "Number of aux indvars removed");
@@ -185,8 +185,28 @@
       IV->Phi->setName("");
       Val->setName(OldName);
 
+      // Get the incoming values used by the PHI node
+      std::vector PHIOps;
+      PHIOps.reserve(IV->Phi->getNumIncomingValues());
+      for (unsigned i = 0, e = IV->Phi->getNumIncomingValues(); i != e; ++i)
+        PHIOps.push_back(IV->Phi->getIncomingValue(i));
+
       // Delete the old, now unused, phi node...
       Header->getInstList().erase(IV->Phi);
+
+      // If the PHI is the last user of any instructions for computing PHI nodes
+      // that are irrelevant now, delete those instructions.
+      while (!PHIOps.empty()) {
+        Instruction *MaybeDead = dyn_cast(PHIOps.back());
+        PHIOps.pop_back();
+
+        if (MaybeDead && isInstructionTriviallyDead(MaybeDead)) {
+          PHIOps.insert(PHIOps.end(), MaybeDead->op_begin(),
+                        MaybeDead->op_end());
+          MaybeDead->getParent()->getInstList().erase(MaybeDead);
+        }
+      }
+
       Changed = true;
       ++NumRemoved;
     }
@@ -216,8 +236,7 @@
                                            "Canonicalize Induction Variables");
 }
 
-Pass *createIndVarSimplifyPass() {
+Pass *llvm::createIndVarSimplifyPass() {
   return new InductionVariableSimplify();
 }
 
-} // End llvm namespace




From criswell at cs.uiuc.edu  Wed Dec 10 12:19:00 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 12:19:00 2003
Subject: [llvm-commits] CVS: llvm/lib/CWriter/Writer.cpp
Message-ID: <200312101818.MAA06492@choi.cs.uiuc.edu>


Changes in directory llvm/lib/CWriter:

Writer.cpp updated: 1.149 -> 1.150

---
Log message:

Upon Chris's suggestion, moved the #ifdef's to the generated C code.
This makes LLVM and the generated C code more portable.



---
Diffs of the changes:  (+5 -8)

Index: llvm/lib/CWriter/Writer.cpp
diff -u llvm/lib/CWriter/Writer.cpp:1.149 llvm/lib/CWriter/Writer.cpp:1.150
--- llvm/lib/CWriter/Writer.cpp:1.149	Wed Dec 10 12:06:34 2003
+++ llvm/lib/CWriter/Writer.cpp	Wed Dec 10 12:17:57 2003
@@ -1087,20 +1087,17 @@
 
 
 void CWriter::visitUnwindInst(UnwindInst &I) {
-  // Determine the return size of write() based on the data model.
-#ifdef _LP64
-      const char * writedecl = "    extern signed long long write();\n";
-#else
-      const char * writedecl = "    extern write();\n";
-#endif
-
   // The unwind instructions causes a control flow transfer out of the current
   // function, unwinding the stack until a caller who used the invoke
   // instruction is found.  In this context, we code generated the invoke
   // instruction to add an entry to the top of the jmpbuf_list.  Thus, here we
   // just have to longjmp to the specified handler.
   Out << "  if (__llvm_jmpbuf_list == 0) {  /* unwind */\n"
-      << writedecl
+      << "#ifdef _LP64\n"
+      << "    extern signed long long write();\n"
+      << "#else\n"
+      << "    extern write();\n"
+      << "#endif\n"
       << "    ((void (*)(int, void*, unsigned))write)(2,\n"
       << "           \"throw found with no handler!\\n\", 31); abort();\n"
       << "  }\n"




From gaeke at cs.uiuc.edu  Wed Dec 10 12:42:04 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 12:42:04 2003
Subject: [llvm-commits] CVS: llvm/Makefile
Message-ID: <200312101841.MAA06003@zion.cs.uiuc.edu>


Changes in directory llvm:

Makefile updated: 1.13 -> 1.14

---
Log message:

Support 'make install' for LLVM include files.


---
Diffs of the changes:  (+10 -0)

Index: llvm/Makefile
diff -u llvm/Makefile:1.13 llvm/Makefile:1.14
--- llvm/Makefile:1.13	Tue Oct 21 13:08:10 2003
+++ llvm/Makefile	Wed Dec 10 12:41:20 2003
@@ -34,3 +34,13 @@
 include/Config/config.h.in: autoconf/configure.ac autoconf/aclocal.m4
 	$(AUTOHEADER) -I autoconf autoconf/configure.ac
 
+# Install support for llvm include files.
+
+.PHONY: install-includes
+
+install-includes:
+	$(MKDIR) $(includedir)/llvm
+	cd include && find * '!' '(' -name '*~' -o -name .cvsignore ')' -print | grep -v CVS | pax -rwdvpe $(includedir)/llvm
+
+install:: install-includes
+




From lattner at cs.uiuc.edu  Wed Dec 10 13:23:06 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 13:23:06 2003
Subject: [llvm-commits] CVS: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp
Message-ID: <200312101922.NAA19196@zion.cs.uiuc.edu>


Changes in directory llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++:

lists1.cpp updated: 1.2 -> 1.3

---
Log message:

Fix a bug in the benchmark.  Apparently these things haven't been tested with the default input sizes


---
Diffs of the changes:  (+2 -2)

Index: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp
diff -u llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp:1.2 llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp:1.3
--- llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp:1.2	Mon Oct  6 20:03:17 2003
+++ llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/lists1.cpp	Wed Dec 10 13:22:22 2003
@@ -1,5 +1,5 @@
 // -*- mode: c++ -*-
-// $Id: lists1.cpp,v 1.2 2003/10/07 01:03:17 lattner Exp $
+// $Id: lists1.cpp,v 1.3 2003/12/10 19:22:22 lattner Exp $
 // http://www.bagley.org/~doug/shootout/
 
 #include 
@@ -19,7 +19,7 @@
 void list_print_n (list L, int n) {
     int c, lastc = n - 1;
     list::iterator i;
-    for (c = 0, i = L.begin(); i != L.end(), c < n; ++i, ++c) {
+    for (c = 0, i = L.begin(); i != L.end() && c < n; ++i, ++c) {
 	cout << (*i);
 	if (c < lastc) cout << " ";
     }




From criswell at cs.uiuc.edu  Wed Dec 10 14:05:02 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 14:05:02 2003
Subject: [llvm-commits] CVS: llvm/docs/CFEBuildInstrs.html
Message-ID: <200312102004.OAA08770@choi.cs.uiuc.edu>


Changes in directory llvm/docs:

CFEBuildInstrs.html updated: 1.9 -> 1.10

---
Log message:

Fixed a comma error.
Added the --disable-threads option to GCC configure (to avoid pthread errors
when building code with LLVM).



---
Diffs of the changes:  (+4 -4)

Index: llvm/docs/CFEBuildInstrs.html
diff -u llvm/docs/CFEBuildInstrs.html:1.9 llvm/docs/CFEBuildInstrs.html:1.10
--- llvm/docs/CFEBuildInstrs.html:1.9	Wed Dec 10 09:27:29 2003
+++ llvm/docs/CFEBuildInstrs.html	Wed Dec 10 14:04:13 2003
@@ -84,7 +84,7 @@
  % set CFEINSTALL = `pwd`/install
 
  % cd build
- % ../src/configure --prefix=$CFEINSTALL --disable-nls --disable-shared \
+ % ../src/configure --prefix=$CFEINSTALL --disable-threads --disable-nls --disable-shared \
    --enable-languages=c,c++
  % gmake all-gcc
  % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
@@ -110,7 +110,7 @@
 
   % cd build
- % ../src/configure --prefix=$CFEINSTALL --disable-nls --disable-shared \
+ % ../src/configure --prefix=$CFEINSTALL --disable-threads --disable-nls --disable-shared \
    --enable-languages=c,c++ --host=sparcv9-sun-solaris2.8
  % gmake all-gcc
  % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
@@ -203,7 +203,7 @@
   Brian Gaeke
   The LLVM Compiler Infrastructure
   

-  Last modified: $Date: 2003/12/10 15:27:29 $
+  Last modified: $Date: 2003/12/10 20:04:13 $
 
 
 




From lattner at cs.uiuc.edu  Wed Dec 10 14:15:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:15:02 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/LowerInvoke/
Message-ID: <200312102014.OAA28996@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/LowerInvoke:


---
Log message:

Directory /home/vadve/shared/PublicCVS/llvm/test/Regression/Transforms/LowerInvoke added to the repository


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



From lattner at cs.uiuc.edu  Wed Dec 10 14:19:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:19:01 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/LowerInvoke/2003-12-10-Crash.llx
Message-ID: <200312102018.OAA30062@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/LowerInvoke:

2003-12-10-Crash.llx added (r1.1)

---
Log message:

new testcase


---
Diffs of the changes:  (+24 -0)

Index: llvm/test/Regression/Transforms/LowerInvoke/2003-12-10-Crash.llx
diff -c /dev/null llvm/test/Regression/Transforms/LowerInvoke/2003-12-10-Crash.llx:1.1
*** /dev/null	Wed Dec 10 14:18:19 2003
--- llvm/test/Regression/Transforms/LowerInvoke/2003-12-10-Crash.llx	Wed Dec 10 14:18:09 2003
***************
*** 0 ****
--- 1,24 ----
+ ; This testcase was reduced from Shootout-C++/reversefile.cpp by bugpoint
+ 
+ ; RUN: llvm-as < %s | opt -lowerinvoke -disable-output
+ 
+ declare void %baz()
+ 
+ declare void %bar()
+ 
+ void %foo() {
+ then:		; preds = %entry
+ 	invoke void %baz( )
+ 			to label %invoke_cont.0 except label %try_catch
+ 
+ invoke_cont.0:		; preds = %then
+ 	invoke void %bar( )
+ 			to label %try_exit except label %try_catch
+ 
+ try_catch:		; preds = %then, %invoke_cont.0
+ 	%__tmp.0 = phi int* [ null, %invoke_cont.0 ], [ null, %then ]
+ 	ret void
+ 
+ try_exit:		; preds = %invoke_cont.0
+ 	ret void
+ }




From lattner at cs.uiuc.edu  Wed Dec 10 14:23:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:23:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LowerInvoke.cpp
Message-ID: <200312102022.OAA30871@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LowerInvoke.cpp updated: 1.3 -> 1.4

---
Log message:

Finegrainify namespacification
Fix bug: LowerInvoke/2003-12-10-Crash.llx


---
Diffs of the changes:  (+11 -11)

Index: llvm/lib/Transforms/Scalar/LowerInvoke.cpp
diff -u llvm/lib/Transforms/Scalar/LowerInvoke.cpp:1.3 llvm/lib/Transforms/Scalar/LowerInvoke.cpp:1.4
--- llvm/lib/Transforms/Scalar/LowerInvoke.cpp:1.3	Tue Nov 11 16:41:33 2003
+++ llvm/lib/Transforms/Scalar/LowerInvoke.cpp	Wed Dec 10 14:22:42 2003
@@ -22,8 +22,7 @@
 #include "llvm/Type.h"
 #include "llvm/Constant.h"
 #include "Support/Statistic.h"
-
-namespace llvm {
+using namespace llvm;
 
 namespace {
   Statistic<> NumLowered("lowerinvoke", "Number of invoke & unwinds replaced");
@@ -40,7 +39,7 @@
 }
 
 // Public Interface To the LowerInvoke pass.
-FunctionPass *createLowerInvokePass() { return new LowerInvoke(); }
+FunctionPass *llvm::createLowerInvokePass() { return new LowerInvoke(); }
 
 // doInitialization - Make sure that there is a prototype for abort in the
 // current module.
@@ -51,8 +50,8 @@
 
 bool LowerInvoke::runOnFunction(Function &F) {
   bool Changed = false;
-  for (Function::iterator I = F.begin(), E = F.end(); I != E; ++I)
-    if (InvokeInst *II = dyn_cast(I->getTerminator())) {
+  for (Function::iterator BB = F.begin(), E = F.end(); BB != E; ++BB)
+    if (InvokeInst *II = dyn_cast(BB->getTerminator())) {
       // Insert a normal call instruction...
       std::string Name = II->getName(); II->setName("");
       Value *NewCall = new CallInst(II->getCalledValue(),
@@ -60,14 +59,17 @@
                                                         II->op_end()), Name,II);
       II->replaceAllUsesWith(NewCall);
       
-      // Insert an unconditional branch to the normal destination
+      // Insert an unconditional branch to the normal destination.
       new BranchInst(II->getNormalDest(), II);
 
+      // Remove any PHI node entries from the exception destination.
+      II->getExceptionalDest()->removePredecessor(BB);
+
       // Remove the invoke instruction now.
-      I->getInstList().erase(II);
+      BB->getInstList().erase(II);
 
       ++NumLowered; Changed = true;
-    } else if (UnwindInst *UI = dyn_cast(I->getTerminator())) {
+    } else if (UnwindInst *UI = dyn_cast(BB->getTerminator())) {
       // Insert a call to abort()
       new CallInst(AbortFn, std::vector(), "", UI);
 
@@ -76,11 +78,9 @@
                             Constant::getNullValue(F.getReturnType()), UI);
 
       // Remove the unwind instruction now.
-      I->getInstList().erase(UI);
+      BB->getInstList().erase(UI);
 
       ++NumLowered; Changed = true;
     }
   return Changed;
 }
-
-} // End llvm namespace




From lattner at cs.uiuc.edu  Wed Dec 10 14:30:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:30:02 2003
Subject: [llvm-commits] CVS: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
Message-ID: <200312102029.OAA32270@zion.cs.uiuc.edu>


Changes in directory llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++:

Makefile updated: 1.5 -> 1.6

---
Log message:

Eliminate the one test that uses threads, we aren't ready for it yet.
Also, don't link in the native libstdc++, as we aren't abi compatible with it,
and don't need it.  We now pass all but one of the tests.


---
Diffs of the changes:  (+1 -1)

Index: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
diff -u llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.5 llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.6
--- llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.5	Wed Dec 10 09:31:43 2003
+++ llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile	Wed Dec 10 14:29:07 2003
@@ -1,7 +1,7 @@
 LEVEL = ../../../../..
 CXXFLAGS += -Wno-deprecated
 CPPFLAGS += -Wno-deprecated
-LDFLAGS += -lpthread -lstdc++
+LDFLAGS += -lm
 LIBS += -lstdc++
 
 include $(LEVEL)/test/Programs/SingleSource/Makefile.singlesrc




From lattner at cs.uiuc.edu  Wed Dec 10 14:41:03 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:41:03 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-RemoveInstrCrash.llx
Message-ID: <200312102040.OAA00873@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/IndVarsSimplify:

2003-12-10-RemoveInstrCrash.llx added (r1.1)

---
Log message:

A new testcase, reduced by bugpoint


---
Diffs of the changes:  (+17 -0)

Index: llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-RemoveInstrCrash.llx
diff -c /dev/null llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-RemoveInstrCrash.llx:1.1
*** /dev/null	Wed Dec 10 14:40:04 2003
--- llvm/test/Regression/Transforms/IndVarsSimplify/2003-12-10-RemoveInstrCrash.llx	Wed Dec 10 14:39:53 2003
***************
*** 0 ****
--- 1,17 ----
+ ; RUN: llvm-as < %s | opt -indvars -disable-output
+ 
+ void %test() {
+ entry:
+ 	%inc.2 = add int 1, 1		;  [#uses=1]
+ 	br bool false, label %no_exit, label %loopexit
+ 
+ no_exit:		; preds = %entry, %no_exit
+ 	%j.0.pn = phi int [ %inc.3, %no_exit ], [ %inc.2, %entry ]		;  [#uses=1]
+ 	%k.0.pn = phi int [ %inc.4, %no_exit ], [ 1, %entry ]		;  [#uses=1]
+ 	%inc.3 = add int %j.0.pn, 1		;  [#uses=1]
+ 	%inc.4 = add int %k.0.pn, 1		;  [#uses=1]
+ 	br bool false, label %no_exit, label %loopexit
+ 
+ loopexit:		; preds = %entry, %no_exit
+ 	ret void
+ }




From lattner at cs.uiuc.edu  Wed Dec 10 14:44:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:44:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/IndVarSimplify.cpp
Message-ID: <200312102043.OAA01435@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

IndVarSimplify.cpp updated: 1.43 -> 1.44

---
Log message:

Fix bug: IndVarsSimplify/2003-12-10-RemoveInstrCrash.llx



---
Diffs of the changes:  (+5 -0)

Index: llvm/lib/Transforms/Scalar/IndVarSimplify.cpp
diff -u llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.43 llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.44
--- llvm/lib/Transforms/Scalar/IndVarSimplify.cpp:1.43	Wed Dec 10 12:06:47 2003
+++ llvm/lib/Transforms/Scalar/IndVarSimplify.cpp	Wed Dec 10 14:43:04 2003
@@ -140,6 +140,8 @@
 
     DEBUG(IV->print(std::cerr));
 
+    while (isa(AfterPHIIt)) ++AfterPHIIt;
+
     // Don't do math with pointers...
     const Type *IVTy = IV->Phi->getType();
     if (isa(IVTy)) IVTy = Type::ULongTy;
@@ -204,6 +206,9 @@
           PHIOps.insert(PHIOps.end(), MaybeDead->op_begin(),
                         MaybeDead->op_end());
           MaybeDead->getParent()->getInstList().erase(MaybeDead);
+
+          // Erasing the instruction could invalidate the AfterPHI iterator!
+          AfterPHIIt = Header->begin();
         }
       }
 




From lattner at cs.uiuc.edu  Wed Dec 10 14:44:05 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 14:44:05 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LICM.cpp
Message-ID: <200312102043.OAA01679@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.47 -> 1.48

---
Log message:

Don't allow dead instructions to stop sinking early.


---
Diffs of the changes:  (+9 -10)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.47 llvm/lib/Transforms/Scalar/LICM.cpp:1.48
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.47	Wed Dec 10 10:58:24 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Wed Dec 10 14:43:29 2003
@@ -366,7 +366,12 @@
   DEBUG(std::cerr << "LICM sinking instruction: " << I);
 
   const std::vector &ExitBlocks = CurLoop->getExitBlocks();
-  
+  std::vector Operands(I.op_begin(), I.op_end());
+
+  if (isa(I)) ++NumMovedLoads;
+  ++NumSunk;
+  Changed = true;
+
   // The case where there is only a single exit node of this loop is common
   // enough that we handle it as a special (more efficient) case.  It is more
   // efficient to handle because there are no PHI nodes that need to be placed.
@@ -386,7 +391,6 @@
   } else if (ExitBlocks.size() == 0) {
     // The instruction is actually dead if there ARE NO exit blocks.
     I.getParent()->getInstList().erase(&I);
-    return;   // Don't count this as a sunk instruction, don't check operands.
   } else {
     // Otherwise, if we have multiple exits, use the PromoteMem2Reg function to
     // do all of the hard work of inserting PHI nodes as necessary.  We convert
@@ -469,17 +473,12 @@
     PromoteMemToReg(Allocas, *DT, *DF, AA->getTargetData());
   }
   
-  if (isa(I)) ++NumMovedLoads;
-  ++NumSunk;
-  Changed = true;
-
   // Since we just sunk an instruction, check to see if any other instructions
   // used by this instruction are now sinkable.  If so, sink them too.
-  for (unsigned i = 0, e = I.getNumOperands(); i != e; ++i)
-    if (Instruction *OpI = dyn_cast(I.getOperand(i)))
+  for (unsigned i = 0, e = Operands.size(); i != e; ++i)
+    if (Instruction *OpI = dyn_cast(Operands[i]))
       if (CurLoop->contains(OpI->getParent()) && canSinkOrHoistInst(*OpI) &&
-          isNotUsedInLoop(*OpI) &&
-          isSafeToExecuteUnconditionally(*OpI))
+          isNotUsedInLoop(*OpI) && isSafeToExecuteUnconditionally(*OpI))
         sink(*OpI);
 }
 




From criswell at cs.uiuc.edu  Wed Dec 10 14:55:01 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 14:55:01 2003
Subject: [llvm-commits] CVS: llvm/test/QMTest/expectations.sunos.qmr
Message-ID: <200312102054.OAA29808@choi.cs.uiuc.edu>


Changes in directory llvm/test/QMTest:

expectations.sunos.qmr updated: 1.3 -> 1.4

---
Log message:

The following tests are now expected to pass:
Regression.CFrontend.2003-08-17-DeadCodeShortCircuit
Regression.CFrontend.2003-08-30-LargeIntegerBitfieldMember
Regression.Transforms.ConstProp.logicaltest
Regression.Transforms.FunctionResolve.2003-04-18-ForwardDeclGlobal
Regression.Transforms.IndVarsSimplify.2003-09-23-NotAtTop
Regression.Transforms.Inline.alloca_test
Regression.Transforms.InstCombine.cast-set
Regression.Transforms.LICM.2003-08-04-TrappingInstHoist
Regression.Transforms.LICM.2003-08-04-TrappingInstOkHoist

The following tests are now expected to fail:
Regression.C++Frontend.2003-12-08-ArrayOfPtrToMemberFunc
Regression.CFrontend.2003-08-30-AggregateInitializer
Regression.Transforms.BasicAA.2003-11-04-SimpleCases
Regression.Transforms.InstCombine.bit-tracking



---
Diffs of the changes:  (+1 -1)

Index: llvm/test/QMTest/expectations.sunos.qmr
diff -u llvm/test/QMTest/expectations.sunos.qmr:1.3 llvm/test/QMTest/expectations.sunos.qmr:1.4
--- llvm/test/QMTest/expectations.sunos.qmr:1.3	Fri Dec  5 17:51:44 2003
+++ llvm/test/QMTest/expectations.sunos.qmr	Wed Dec 10 14:54:09 2003
@@ -3,13 +3,13 @@
 qoq}q(U
_Result__kindqUtestqU_Result__outcomeqUPASSqU_Result__annotationsq}U_Result__idq	U0Regression.Assembler.2002-08-16-ConstExprInlinedq
 U_Result__contextq(cqm.test.context
 Context
-qo}q
(U_Context__propertiesq}U_Context__temporariesq}ubub.(hoq}q(hhhhh}h	U(Regression.Transforms.PruneEH.simpletestqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	U/Regression.CBackend.2002-08-19-HardConstantExprqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	URegression.Jello.test-shiftqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	U9Regression.Transforms.Mem2Reg.2003-10-05-DeadPHIInsertionqh(ho}q(h}h}ubub.(hoq }q!(hhhUFAILq"h}h	U.Regression.Transforms.CorrelatedExprs.looptestq#h(ho}q$(h}h}ubub.(hoq%}q&(hhhhh}h	U-Regression.Linker.2003-08-23-GlobalVarLinkingq'h(ho}q((h}h}ubub.(hoq)}q*(hhhhh}h	U?Regression.Transforms.Inline.2003-09-22-PHINodesInExceptionDestq+h(ho}q,(h}h}ubub.(hoq-}q.(hhhh"h}h	U,Regression.CFrontend.2003-02-12-NonlocalGotoq/h(ho}q0(h}h}ubub.(hoq1}q2(hhhhh}h	U2Regression.Transforms.FunctionResolve.retmismatch1q3h(ho}q4(h}h}ubub.(hoq5}q6(hhhhh}h	U8Regression.Transforms.ADCE.2003-01-22-Prede!
 cessorProblemq7h(ho}q8(h}h}ubub.(hoq9}q:(hhhhh}h	U=Regression.Transforms.LevelRaise.2002-10-08-VarArgCallInfLoopq;h(ho}q<(h}h}ubub.(hoq=}q>(hhhhh}h	U Regression.Reoptimizer.ticm.ticmq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-07-DominatorProblemq_!
 h(ho}q`(h}h}ubub.(hoqa}qb(hhhhh}h	U3Regression.Transforms.Pi
NodeInserter.substitutetestqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U-Regression.CFrontend.2002-02-18-64bitConstantqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U(Regression.Reoptimizer.BinInterface.testqsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U-Regression.C++Frontend.2003-08-28-SaveExprBugqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U"Regression.Jello.2003-06-05-PHIBugqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq_h(ho}q`(h}h}ubub.(!
 hoqa}qb(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-
07-DominatorProblemqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U3Regression.Transforms.PiNodeInserter.substitutetestqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U$Regression.Jello.2003-01-04-LoopTestqoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U)Regression.CFrontend.2002-07-14-MiscTestsqwh(ho}qx(h}h}ubub.(hoqy}qz(hhhhh}h	U


Changes in directory llvm/test/QMTest:

expectations.linux.qmr updated: 1.3 -> 1.4

---
Log message:

The following tests are now expected to pass:
Regression.CFrontend.2003-01-30-UnionInit

The following tests are now expected to fail:
Regression.C++Frontend.2003-12-08-ArrayOfPtrToMemberFunc
Regression.Transforms.LICM.sink_multiple_exits
Regression.Transforms.TailCallElim.tail_call_with_branch



---
Diffs of the changes:  (+1 -1)

Index: llvm/test/QMTest/expectations.linux.qmr
diff -u llvm/test/QMTest/expectations.linux.qmr:1.3 llvm/test/QMTest/expectations.linux.qmr:1.4
--- llvm/test/QMTest/expectations.linux.qmr:1.3	Tue Dec  2 16:21:59 2003
+++ llvm/test/QMTest/expectations.linux.qmr	Wed Dec 10 15:27:38 2003
@@ -3,13 +3,13 @@
 qoq}q(U
_Result__kindqUtestqU_Result__outcomeqUPASSqU_Result__annotationsq}U_Result__idq	U0Regression.Assembler.2002-08-16-ConstExprInlinedq
 U_Result__contextq(cqm.test.context
 Context
-qo}q
(U_Context__propertiesq}U_Context__temporariesq}ubub.(hoq}q(hhhhh}h	U(Regression.Transforms.PruneEH.simpletestqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	U/Regression.CBackend.2002-08-19-HardConstantExprqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	URegression.Jello.test-shiftqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	U9Regression.Transforms.Mem2Reg.2003-10-05-DeadPHIInsertionqh(ho}q(h}h}ubub.(hoq }q!(hhhUFAILq"h}h	U.Regression.Transforms.CorrelatedExprs.looptestq#h(ho}q$(h}h}ubub.(hoq%}q&(hhhhh}h	U-Regression.Linker.2003-08-23-GlobalVarLinkingq'h(ho}q((h}h}ubub.(hoq)}q*(hhhhh}h	U?Regression.Transforms.Inline.2003-09-22-PHINodesInExceptionDestq+h(ho}q,(h}h}ubub.(hoq-}q.(hhhh"h}h	U,Regression.CFrontend.2003-02-12-NonlocalGotoq/h(ho}q0(h}h}ubub.(hoq1}q2(hhhhh}h	U2Regression.Transforms.FunctionResolve.retmismatch1q3h(ho}q4(h}h}ubub.(hoq5}q6(hhhhh}h	U8Regression.Transforms.ADCE.2003-01-22-Prede!
 cessorProblemq7h(ho}q8(h}h}ubub.(hoq9}q:(hhhhh}h	U=Regression.Transforms.LevelRaise.2002-10-08-VarArgCallInfLoopq;h(ho}q<(h}h}ubub.(hoq=}q>(hhhhh}h	U Regression.Reoptimizer.ticm.ticmq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-07-DominatorProblemq_!
 h(ho}q`(h}h}ubub.(hoqa}qb(hhhhh}h	U3Regression.Transforms.Pi
NodeInserter.substitutetestqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U-Regression.CFrontend.2002-02-18-64bitConstantqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U(Regression.Reoptimizer.BinInterface.testqsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U-Regression.C++Frontend.2003-08-28-SaveExprBugqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U"Regression.Jello.2003-06-05-PHIBugqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq_h(ho}q`(h}h}ubub.(!
 hoqa}qb(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-
07-DominatorProblemqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U3Regression.Transforms.PiNodeInserter.substitutetestqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U$Regression.Jello.2003-01-04-LoopTestqoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U)Regression.CFrontend.2002-07-14-MiscTestsqwh(ho}qx(h}h}ubub.(hoqy}qz(hhhhh}h	U


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.48 -> 1.49

---
Log message:

Fix bug: LICM/sink_multiple_exits.ll

Thanks for pointing this out John  :)



---
Diffs of the changes:  (+1 -1)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.48 llvm/lib/Transforms/Scalar/LICM.cpp:1.49
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.48	Wed Dec 10 14:43:29 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Wed Dec 10 16:35:56 2003
@@ -451,7 +451,7 @@
           // instruction, otherwise clone the original instruction and insert
           // the copy.
           Instruction *New;
-          if (InsertedBlocks.empty()) {
+          if (InsertedBlocks.size() == 1) {
             I.getParent()->getInstList().remove(&I);
             ExitBlock->getInstList().insert(InsertPt, &I);
             New = &I;




From criswell at cs.uiuc.edu  Wed Dec 10 16:52:01 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Wed Dec 10 16:52:01 2003
Subject: [llvm-commits] CVS: llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp
Message-ID: <200312102251.QAA10996@choi.cs.uiuc.edu>


Changes in directory llvm/lib/CodeGen/InstrSelection:

InstrSelectionSupport.cpp updated: 1.58 -> 1.59

---
Log message:

This appears to fix Bug 172 and does not break any other feature tests or
regression tests.



---
Diffs of the changes:  (+6 -0)

Index: llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp
diff -u llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp:1.58 llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp:1.59
--- llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp:1.58	Tue Nov 11 16:41:33 2003
+++ llvm/lib/CodeGen/InstrSelection/InstrSelectionSupport.cpp	Wed Dec 10 16:51:41 2003
@@ -167,6 +167,12 @@
             constantThatMustBeLoaded = true;
         }
       } else {
+        //
+        // If the operand is from the constant pool, don't try to change it.
+        //
+        if (mop.getType() == MachineOperand::MO_ConstantPoolIndex) {
+          continue;
+        }
         assert(mop.isImmediate());
         bool isSigned = mop.getType() == MachineOperand::MO_SignExtendedImmed;
 




From gaeke at cs.uiuc.edu  Wed Dec 10 18:24:03 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 18:24:03 2003
Subject: [llvm-commits] CVS: llvm/lib/ExecutionEngine/Interpreter/Execution.cpp
Message-ID: <200312110023.SAA08022@zion.cs.uiuc.edu>


Changes in directory llvm/lib/ExecutionEngine/Interpreter:

Execution.cpp updated: 1.116 -> 1.117

---
Log message:

Add support for --debug-only=interpreter, to print out instrs before
interpreting them.

Move support for getting the value of a ConstantExpr into
getConstantExprValue(), and add support for the rest of the different
kinds of ConstantExprs.  (I don't think I like ConstantExprs!)
This requires separate procedures executeShlInst() and executeShrInst().

Reduce the number of references to TheEE.

Get rid of an old comment mentioning annotations.

Fix exitCalled(), which was crashing the Interpreter. This was a
leftover from the return-value code refactoring.


---
Diffs of the changes:  (+144 -40)

Index: llvm/lib/ExecutionEngine/Interpreter/Execution.cpp
diff -u llvm/lib/ExecutionEngine/Interpreter/Execution.cpp:1.116 llvm/lib/ExecutionEngine/Interpreter/Execution.cpp:1.117
--- llvm/lib/ExecutionEngine/Interpreter/Execution.cpp:1.116	Tue Nov 25 14:44:56 2003
+++ llvm/lib/ExecutionEngine/Interpreter/Execution.cpp	Wed Dec 10 18:22:59 2003
@@ -11,12 +11,15 @@
 //
 //===----------------------------------------------------------------------===//
 
+#define DEBUG_TYPE "interpreter"
+
 #include "Interpreter.h"
 #include "llvm/Instructions.h"
 #include "llvm/DerivedTypes.h"
 #include "llvm/Constants.h"
 #include "llvm/Support/GetElementPtrTypeIterator.h"
 #include "Support/Statistic.h"
+#include "Support/Debug.h"
 #include   // For fmod
 using namespace llvm;
 
@@ -31,34 +34,126 @@
 //===----------------------------------------------------------------------===//
 //                     Value Manipulation code
 //===----------------------------------------------------------------------===//
-
-// Operations used by constant expr implementations...
-static GenericValue executeCastOperation(Value *Src, const Type *DestTy,
-                                         ExecutionContext &SF);
 static GenericValue executeAddInst(GenericValue Src1, GenericValue Src2, 
 				   const Type *Ty);
+static GenericValue executeSubInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeMulInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeRemInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeDivInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeAndInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeOrInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeXorInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetEQInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetNEInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetLTInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetGTInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetLEInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeSetGEInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeShlInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+static GenericValue executeShrInst(GenericValue Src1, GenericValue Src2, 
+				   const Type *Ty);
+                                   
+GenericValue Interpreter::getConstantExprValue (ConstantExpr *CE,
+                                                ExecutionContext &SF) {
+  switch (CE->getOpcode()) {
+  case Instruction::Cast:
+    return executeCastOperation(CE->getOperand(0), CE->getType(), SF);
+  case Instruction::GetElementPtr:
+    return executeGEPOperation(CE->getOperand(0), gep_type_begin(CE),
+                               gep_type_end(CE), SF);
+  case Instruction::Add:
+    return executeAddInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Sub:
+    return executeSubInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Mul:
+    return executeMulInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Div:
+    return executeDivInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Rem:
+    return executeRemInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::And:
+    return executeAndInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Or:
+    return executeOrInst(getOperandValue(CE->getOperand(0), SF),
+                         getOperandValue(CE->getOperand(1), SF),
+                         CE->getOperand(0)->getType());
+  case Instruction::Xor:
+    return executeXorInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::SetEQ:
+    return executeSetEQInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::SetNE:
+    return executeSetNEInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::SetLE:
+    return executeSetLEInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::SetGE:
+    return executeSetGEInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::SetLT:
+    return executeSetLTInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::SetGT:
+    return executeSetGTInst(getOperandValue(CE->getOperand(0), SF),
+                            getOperandValue(CE->getOperand(1), SF),
+                            CE->getOperand(0)->getType());
+  case Instruction::Shl:
+    return executeShlInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  case Instruction::Shr:
+    return executeShrInst(getOperandValue(CE->getOperand(0), SF),
+                          getOperandValue(CE->getOperand(1), SF),
+                          CE->getOperand(0)->getType());
+  
+  default:
+    std::cerr << "Unhandled ConstantExpr: " << CE << "\n";
+    abort();
+    return GenericValue();
+  }
+}
 
 GenericValue Interpreter::getOperandValue(Value *V, ExecutionContext &SF) {
   if (ConstantExpr *CE = dyn_cast(V)) {
-    switch (CE->getOpcode()) {
-    case Instruction::Cast:
-      return executeCastOperation(CE->getOperand(0), CE->getType(), SF);
-    case Instruction::GetElementPtr:
-      return TheEE->executeGEPOperation(CE->getOperand(0), gep_type_begin(CE),
-					gep_type_end(CE), SF);
-    case Instruction::Add:
-      return executeAddInst(getOperandValue(CE->getOperand(0), SF),
-                            getOperandValue(CE->getOperand(1), SF),
-                            CE->getType());
-    default:
-      std::cerr << "Unhandled ConstantExpr: " << CE << "\n";
-      abort();
-      return GenericValue();
-    }
+    return getConstantExprValue(CE, SF);
   } else if (Constant *CPV = dyn_cast(V)) {
-    return TheEE->getConstantValue(CPV);
+    return getConstantValue(CPV);
   } else if (GlobalValue *GV = dyn_cast(V)) {
-    return PTOGV(TheEE->getPointerToGlobal(GV));
+    return PTOGV(getPointerToGlobal(GV));
   } else {
     return SF.Values[V];
   }
@@ -68,10 +163,6 @@
   SF.Values[V] = Val;
 }
 
-//===----------------------------------------------------------------------===//
-//                    Annotation Wrangling code
-//===----------------------------------------------------------------------===//
-
 void Interpreter::initializeExecutionEngine() {
   TheEE = this;
 }
@@ -432,8 +523,8 @@
 //===----------------------------------------------------------------------===//
 
 void Interpreter::exitCalled(GenericValue GV) {
-  ExitCode = GV.SByteVal;
-  ECStack.clear();
+  runAtExitHandlers ();
+  exit (GV.IntVal);
 }
 
 /// Pop the last stack frame off of ECStack and then copy the result
@@ -711,13 +802,9 @@
 #define IMPLEMENT_SHIFT(OP, TY) \
    case Type::TY##TyID: Dest.TY##Val = Src1.TY##Val OP Src2.UByteVal; break
 
-void Interpreter::visitShl(ShiftInst &I) {
-  ExecutionContext &SF = ECStack.back();
-  const Type *Ty    = I.getOperand(0)->getType();
-  GenericValue Src1 = getOperandValue(I.getOperand(0), SF);
-  GenericValue Src2 = getOperandValue(I.getOperand(1), SF);
+static GenericValue executeShlInst(GenericValue Src1, GenericValue Src2,
+                                   const Type *Ty) {
   GenericValue Dest;
-
   switch (Ty->getPrimitiveID()) {
     IMPLEMENT_SHIFT(<<, UByte);
     IMPLEMENT_SHIFT(<<, SByte);
@@ -730,16 +817,12 @@
   default:
     std::cout << "Unhandled type for Shl instruction: " << *Ty << "\n";
   }
-  SetValue(&I, Dest, SF);
+  return Dest;
 }
 
-void Interpreter::visitShr(ShiftInst &I) {
-  ExecutionContext &SF = ECStack.back();
-  const Type *Ty    = I.getOperand(0)->getType();
-  GenericValue Src1 = getOperandValue(I.getOperand(0), SF);
-  GenericValue Src2 = getOperandValue(I.getOperand(1), SF);
+static GenericValue executeShrInst(GenericValue Src1, GenericValue Src2,
+                                   const Type *Ty) {
   GenericValue Dest;
-
   switch (Ty->getPrimitiveID()) {
     IMPLEMENT_SHIFT(>>, UByte);
     IMPLEMENT_SHIFT(>>, SByte);
@@ -753,6 +836,26 @@
     std::cout << "Unhandled type for Shr instruction: " << *Ty << "\n";
     abort();
   }
+  return Dest;
+}
+
+void Interpreter::visitShl(ShiftInst &I) {
+  ExecutionContext &SF = ECStack.back();
+  const Type *Ty    = I.getOperand(0)->getType();
+  GenericValue Src1 = getOperandValue(I.getOperand(0), SF);
+  GenericValue Src2 = getOperandValue(I.getOperand(1), SF);
+  GenericValue Dest;
+  Dest = executeShlInst (Src1, Src2, Ty);
+  SetValue(&I, Dest, SF);
+}
+
+void Interpreter::visitShr(ShiftInst &I) {
+  ExecutionContext &SF = ECStack.back();
+  const Type *Ty    = I.getOperand(0)->getType();
+  GenericValue Src1 = getOperandValue(I.getOperand(0), SF);
+  GenericValue Src2 = getOperandValue(I.getOperand(1), SF);
+  GenericValue Dest;
+  Dest = executeShrInst (Src1, Src2, Ty);
   SetValue(&I, Dest, SF);
 }
 
@@ -919,6 +1022,7 @@
     // Track the number of dynamic instructions executed.
     ++NumDynamicInsts;
 
+    DEBUG(std::cerr << "About to interpret: " << I);
     visit(I);   // Dispatch to one of the visit* methods...
   }
 }




From gaeke at cs.uiuc.edu  Wed Dec 10 18:24:08 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 18:24:08 2003
Subject: [llvm-commits] CVS: llvm/lib/ExecutionEngine/Interpreter/Interpreter.h
Message-ID: <200312110023.SAA08059@zion.cs.uiuc.edu>


Changes in directory llvm/lib/ExecutionEngine/Interpreter:

Interpreter.h updated: 1.56 -> 1.57

---
Log message:

Fix typo in comment.  Add prototype for getConstantExprValue().


---
Diffs of the changes:  (+3 -1)

Index: llvm/lib/ExecutionEngine/Interpreter/Interpreter.h
diff -u llvm/lib/ExecutionEngine/Interpreter/Interpreter.h:1.56 llvm/lib/ExecutionEngine/Interpreter/Interpreter.h:1.57
--- llvm/lib/ExecutionEngine/Interpreter/Interpreter.h:1.56	Mon Dec  8 02:23:04 2003
+++ llvm/lib/ExecutionEngine/Interpreter/Interpreter.h	Wed Dec 10 18:23:28 2003
@@ -26,9 +26,10 @@
 
 struct FunctionInfo;        // Defined in ExecutionAnnotations.h
 class gep_type_iterator;
+class ConstantExpr;
 
 // AllocaHolder - Object to track all of the blocks of memory allocated by
-// alloca.  When the function returns, this object is poped off the execution
+// alloca.  When the function returns, this object is popped off the execution
 // stack, which causes the dtor to be run, which frees all the alloca'd memory.
 //
 class AllocaHolder {
@@ -174,6 +175,7 @@
 
   void initializeExecutionEngine();
   void initializeExternalFunctions();
+  GenericValue getConstantExprValue(ConstantExpr *CE, ExecutionContext &SF);
   GenericValue getOperandValue(Value *V, ExecutionContext &SF);
   GenericValue executeCastOperation(Value *SrcVal, const Type *Ty,
 				    ExecutionContext &SF);




From gaeke at cs.uiuc.edu  Wed Dec 10 18:25:01 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 18:25:01 2003
Subject: [llvm-commits] CVS: llvm/lib/CWriter/Writer.cpp
Message-ID: <200312110024.SAA08251@zion.cs.uiuc.edu>


Changes in directory llvm/lib/CWriter:

Writer.cpp updated: 1.150 -> 1.151

---
Log message:

Turn off "attribute weak" to pacify Mac OS X's system compiler, which prints a
warning whenever it sees it.


---
Diffs of the changes:  (+25 -4)

Index: llvm/lib/CWriter/Writer.cpp
diff -u llvm/lib/CWriter/Writer.cpp:1.150 llvm/lib/CWriter/Writer.cpp:1.151
--- llvm/lib/CWriter/Writer.cpp:1.150	Wed Dec 10 12:17:57 2003
+++ llvm/lib/CWriter/Writer.cpp	Wed Dec 10 18:24:36 2003
@@ -585,7 +585,28 @@
   // If we aren't being compiled with GCC, just drop these attributes.
   Out << "#ifndef __GNUC__  /* Can only support \"linkonce\" vars with GCC */\n"
       << "#define __attribute__(X)\n"
-      << "#endif\n";
+      << "#endif\n\n";
+
+#if 0
+  // At some point, we should support "external weak" vs. "weak" linkages.
+  // On Mac OS X, "external weak" is spelled "__attribute__((weak_import))".
+  Out << "#if defined(__GNUC__) && defined(__APPLE_CC__)\n"
+      << "#define __EXTERNAL_WEAK__ __attribute__((weak_import))\n"
+      << "#elif defined(__GNUC__)\n"
+      << "#define __EXTERNAL_WEAK__ __attribute__((weak))\n"
+      << "#else\n"
+      << "#define __EXTERNAL_WEAK__\n"
+      << "#endif\n\n";
+#endif
+
+  // For now, turn off the weak linkage attribute on Mac OS X. (See above.)
+  Out << "#if defined(__GNUC__) && defined(__APPLE_CC__)\n"
+      << "#define __ATTRIBUTE_WEAK__\n"
+      << "#elif defined(__GNUC__)\n"
+      << "#define __ATTRIBUTE_WEAK__ __attribute__((weak))\n"
+      << "#else\n"
+      << "#define __ATTRIBUTE_WEAK__\n"
+      << "#endif\n\n";
 }
 
 // generateProcessorSpecificCode - This is where we add conditional compilation
@@ -683,7 +704,7 @@
       if ((I->hasInternalLinkage() || !MangledGlobals.count(I)) &&
           !I->getIntrinsicID()) {
         printFunctionSignature(I, true);
-        if (I->hasWeakLinkage()) Out << " __attribute__((weak))";
+        if (I->hasWeakLinkage()) Out << " __ATTRIBUTE_WEAK__";
         Out << ";\n";
       }
     }
@@ -706,7 +727,7 @@
         if (I->hasLinkOnceLinkage())
           Out << " __attribute__((common))";
         else if (I->hasWeakLinkage())
-          Out << " __attribute__((weak))";
+          Out << " __ATTRIBUTE_WEAK__";
         Out << ";\n";
       }
   }
@@ -722,7 +743,7 @@
         if (I->hasLinkOnceLinkage())
           Out << " __attribute__((common))";
         else if (I->hasWeakLinkage())
-          Out << " __attribute__((weak))";
+          Out << " __ATTRIBUTE_WEAK__";
 
         // If the initializer is not null, emit the initializer.  If it is null,
         // we try to avoid emitting large amounts of zeros.  The problem with




From gaeke at cs.uiuc.edu  Wed Dec 10 18:39:03 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 18:39:03 2003
Subject: [llvm-commits] CVS: llvm/lib/Bytecode/Reader/ArchiveReader.cpp
Message-ID: <200312110038.SAA11592@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Bytecode/Reader:

ArchiveReader.cpp updated: 1.9 -> 1.10

---
Log message:

In getObjectType(), doxygenify comment, and take a pointer to the
beginning of the archive member data as an argument.

Get rid of ParseLongFilenameSection(), which is dead.

In ReadArchiveBuffer(), implement support for 4.4BSD/MacOSX long filenames.
This is kind of invasive, because they prepend the long filename to the archive 
member data, and then lie about the size. So we have to keep track of the real
size.


---
Diffs of the changes:  (+43 -50)

Index: llvm/lib/Bytecode/Reader/ArchiveReader.cpp
diff -u llvm/lib/Bytecode/Reader/ArchiveReader.cpp:1.9 llvm/lib/Bytecode/Reader/ArchiveReader.cpp:1.10
--- llvm/lib/Bytecode/Reader/ArchiveReader.cpp:1.9	Sun Nov 16 17:08:48 2003
+++ llvm/lib/Bytecode/Reader/ArchiveReader.cpp	Wed Dec 10 18:38:04 2003
@@ -44,10 +44,12 @@
   };
 }
 
-// getObjectType - Determine the type of object that this header represents.
-// This is capable of parsing the variety of special sections used for various
-// purposes.
-static enum ObjectType getObjectType(ar_hdr *H, unsigned Size) {
+/// getObjectType - Determine the type of object that this header represents.
+/// This is capable of parsing the variety of special sections used for various
+/// purposes.
+///
+static enum ObjectType getObjectType(ar_hdr *H, unsigned char *MemberData,
+                                     unsigned Size) {
   // Check for sections with special names...
   if (!memcmp(H->name, "__.SYMDEF       ", 16))
     return ArchiveSymbolTable;
@@ -55,44 +57,17 @@
     return SVR4LongFilename;
 
   // Check to see if it looks like an llvm object file...
-  if (Size >= 4 && !memcmp(H+1, "llvm", 4))
+  if (Size >= 4 && !memcmp(MemberData, "llvm", 4))
     return UserObject;
 
   return Unknown;
 }
 
-
 static inline bool Error(std::string *ErrorStr, const char *Message) {
   if (ErrorStr) *ErrorStr = Message;
   return true;
 }
 
-static bool ParseLongFilenameSection(unsigned char *Buffer, unsigned Size,
-                                     std::vector &LongFilenames,
-                                     std::string *S) {
-  if (!LongFilenames.empty())
-    return Error(S, "archive file contains multiple long filename entries");
-                 
-  while (Size) {
-    // Long filename entries are newline delimited to keep the archive readable.
-    unsigned char *Ptr = (unsigned char*)memchr(Buffer, '\n', Size);
-    if (Ptr == 0)
-      return Error(S, "archive long filename entry doesn't end with newline!");
-    assert(*Ptr == '\n');
-
-    if (Ptr == Buffer) break;  // Last entry contains just a newline.
-
-    unsigned char *End = Ptr;
-    if (End[-1] == '/') --End; // Remove trailing / from name
-    
-    LongFilenames.push_back(std::string(Buffer, End));
-    Size -= Ptr-Buffer+1;
-    Buffer = Ptr+1;
-  }
-  
-  return false;
-}
-
 static bool ParseSymbolTableSection(unsigned char *Buffer, unsigned Size,
                                     std::string *S) {
   // Currently not supported (succeeds without doing anything)
@@ -111,40 +86,58 @@
 
   while (Length >= sizeof(ar_hdr)) {
     ar_hdr *Hdr = (ar_hdr*)Buffer;
-    unsigned Size = atoi(Hdr->size);
-    if (Size+sizeof(ar_hdr) > Length)
+    unsigned SizeFromHeader = atoi(Hdr->size);
+    if (SizeFromHeader + sizeof(ar_hdr) > Length)
       return Error(ErrorStr, "invalid record length in archive file!");
 
+    unsigned char *MemberData = Buffer + sizeof(ar_hdr);
+    unsigned MemberSize = SizeFromHeader;
     // Get name of archive member.
     char *startp = Hdr->name;
-    char *endp = strchr (startp, '/');
-    if (startp == endp && isdigit (Hdr->name[1])) {
-      // Long filenames are abbreviated as "/I", where I is a decimal
-      // index into the LongFilenames vector.
-      unsigned Index = atoi (&Hdr->name[1]);
-      assert (LongFilenames.size () > Index
-              && "Long filename for archive member not found");
-      startp = &LongFilenames[Index];
+    char *endp = (char *) memchr (startp, '/', sizeof(ar_hdr));
+    if (memcmp (Hdr->name, "#1/", 3) == 0) {
+      // 4.4BSD/MacOSX long filenames are abbreviated as "#1/L", where L is an
+      // ASCII-coded decimal number representing the length of the name buffer,
+      // which is prepended to the archive member's contents.
+      unsigned NameLength = atoi (&Hdr->name[3]);
+      startp = (char *) MemberData;
+      endp = startp + NameLength;
+      MemberData += NameLength;
+      MemberSize -= NameLength;
+    } else if (startp == endp && isdigit (Hdr->name[1])) {
+      // SVR4 long filenames are abbreviated as "/I", where I is
+      // an ASCII-coded decimal index into the LongFilenames vector.
+      unsigned NameIndex = atoi (&Hdr->name[1]);
+      assert (LongFilenames.size () > NameIndex
+              && "SVR4-style long filename for archive member not found");
+      startp = &LongFilenames[NameIndex];
       endp = strchr (startp, '/');
     }
+    if (!endp) {
+      // 4.4BSD/MacOSX *short* filenames are not guaranteed to have a
+      // terminator. Start at the end of the field and backtrack over spaces.
+      endp = startp + sizeof(Hdr->name);
+      while (endp[-1] == ' ')
+        --endp;
+    }
     std::string MemberName (startp, endp);
     std::string FullMemberName = ArchiveName + "(" + MemberName + ")";
 
-    switch (getObjectType(Hdr, Size)) {
+    switch (getObjectType(Hdr, MemberData, MemberSize)) {
     case SVR4LongFilename:
       // If this is a long filename section, read all of the file names into the
       // LongFilenames vector.
-      LongFilenames.assign (Buffer+sizeof(ar_hdr), Buffer+sizeof(ar_hdr)+Size);
+      LongFilenames.assign (MemberData, MemberData + MemberSize);
       break;
     case UserObject: {
-      Module *M = ParseBytecodeBuffer(Buffer+sizeof(ar_hdr), Size,
+      Module *M = ParseBytecodeBuffer(MemberData, MemberSize,
                                       FullMemberName, ErrorStr);
       if (!M) return true;
       Objects.push_back(M);
       break;
     }
     case ArchiveSymbolTable:
-      if (ParseSymbolTableSection(Buffer+sizeof(ar_hdr), Size, ErrorStr))
+      if (ParseSymbolTableSection(MemberData, MemberSize, ErrorStr))
         return true;
       break;
     default:
@@ -153,10 +146,10 @@
       break;   // Just ignore unknown files.
     }
 
-    // Round Size up to an even number...
-    Size = (Size+1)/2*2;
-    Buffer += sizeof(ar_hdr)+Size;   // Move to the next entry
-    Length -= sizeof(ar_hdr)+Size;
+    // Round SizeFromHeader up to an even number...
+    SizeFromHeader = (SizeFromHeader+1)/2*2;
+    Buffer += sizeof(ar_hdr)+SizeFromHeader;   // Move to the next entry
+    Length -= sizeof(ar_hdr)+SizeFromHeader;
   }
 
   return Length != 0;




From lattner at cs.uiuc.edu  Wed Dec 10 18:59:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 18:59:01 2003
Subject: [llvm-commits] CVS: llvm/utils/TableGen/InstrSelectorEmitter.cpp
Message-ID: <200312110058.SAA17616@zion.cs.uiuc.edu>


Changes in directory llvm/utils/TableGen:

InstrSelectorEmitter.cpp updated: 1.36 -> 1.37

---
Log message:

Stop laughing Misha


---
Diffs of the changes:  (+1 -1)

Index: llvm/utils/TableGen/InstrSelectorEmitter.cpp
diff -u llvm/utils/TableGen/InstrSelectorEmitter.cpp:1.36 llvm/utils/TableGen/InstrSelectorEmitter.cpp:1.37
--- llvm/utils/TableGen/InstrSelectorEmitter.cpp:1.36	Tue Nov 11 16:41:34 2003
+++ llvm/utils/TableGen/InstrSelectorEmitter.cpp	Wed Dec 10 18:58:34 2003
@@ -58,7 +58,7 @@
     return true;
   }
 
-  throw "Type inferfence contradiction found for pattern " + RecName;
+  throw "Type inference contradiction found for pattern " + RecName;
 }
 
 /// InstantiateNonterminals - If this pattern refers to any nonterminals which




From gaeke at cs.uiuc.edu  Wed Dec 10 22:42:01 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 22:42:01 2003
Subject: [llvm-commits] CVS: llvm/docs/ReleaseNotes.html
Message-ID: <200312110441.WAA20024@zion.cs.uiuc.edu>


Changes in directory llvm/docs:

ReleaseNotes.html updated: 1.72 -> 1.73

---
Log message:

Add more fixed bugs. Keep llvm core section sorted by pr#. (Is that ok?)


---
Diffs of the changes:  (+6 -2)

Index: llvm/docs/ReleaseNotes.html
diff -u llvm/docs/ReleaseNotes.html:1.72 llvm/docs/ReleaseNotes.html:1.73
--- llvm/docs/ReleaseNotes.html:1.72	Mon Dec  8 00:33:19 2003
+++ llvm/docs/ReleaseNotes.html	Wed Dec 10 22:40:58 2003
@@ -148,6 +148,9 @@
 llvm-nm cannot read archive
 files
 
+Interpreter does not handle
+setne constant expression
+
 
 
 
@@ -206,6 +209,7 @@
 [bcreader] Cannot read shift constant expressions from bytecode file
 [lowersetjmp] Lowersetjmp pass breaks dominance properties!
 SymbolTable::getUniqueName is very inefficient
+bugpoint must not pass -R<directory> to Mach-O linker
 [buildscripts] Building into objdir with .o in it fails
 [setjmp/longjmp] Linking C programs which use setjmp/longjmp sometimes fail with references to the C++ runtime library!
 AsmParser Misses Symbol Redefinition Error
@@ -214,7 +218,7 @@
 [adce] ADCE considers blocks without postdominators to be unreachable
 [X86] div and rem constant exprs invalidate iterators!
 [vmcore] Symbol table doesn't rename colliding variables during type resolution
-bugpoint must not pass -R<directory> to Mach-O linker
+Archive reader does not understand 4.4BSD/Mac OS X long filenames
 gccld produces a runner script that includes command-line options to load the necessary shared objects
 
 
@@ -648,7 +652,7 @@
   src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" />
 
   The LLVM Compiler Infrastructure

-  Last modified: $Date: 2003/12/08 06:33:19 $
+  Last modified: $Date: 2003/12/11 04:40:58 $
 
 
 




From gaeke at cs.uiuc.edu  Wed Dec 10 22:49:02 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 22:49:02 2003
Subject: [llvm-commits] CVS: llvm/test/Makefile
Message-ID: <200312110448.WAA20275@zion.cs.uiuc.edu>


Changes in directory llvm/test:

Makefile updated: 1.51 -> 1.52

---
Log message:

Separate out target for cleaning up qmtest garbage.


---
Diffs of the changes:  (+3 -1)

Index: llvm/test/Makefile
diff -u llvm/test/Makefile:1.51 llvm/test/Makefile:1.52
--- llvm/test/Makefile:1.51	Thu Nov 20 19:12:47 2003
+++ llvm/test/Makefile	Wed Dec 10 22:48:38 2003
@@ -96,7 +96,9 @@
 	$(QMTEST) gui --no-browser --daemon
 
 # Also get rid of qmtest garbage when we 'make clean' in this directory.
-clean::
+clean:: qmtest-clean
+
+qmtest-clean:
 	$(RM) -rf $(LLVM_OBJ_ROOT)/test/tmp
 	$(RM) -f $(LLVM_SRC_ROOT)/test/QMTest/*.pyo \
 		$(LLVM_OBJ_ROOT)/test/QMTest/*.pyo




From gaeke at cs.uiuc.edu  Wed Dec 10 23:06:01 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 23:06:01 2003
Subject: [llvm-commits] CVS: llvm/include/llvm/Analysis/AliasAnalysis.h
Message-ID: <200312110505.XAA20514@zion.cs.uiuc.edu>


Changes in directory llvm/include/llvm/Analysis:

AliasAnalysis.h updated: 1.7 -> 1.8

---
Log message:

IncludeFile hack to pull in BasicAliasAnalysis whenever AliasAnalysis.h
is included.


---
Diffs of the changes:  (+9 -0)

Index: llvm/include/llvm/Analysis/AliasAnalysis.h
diff -u llvm/include/llvm/Analysis/AliasAnalysis.h:1.7 llvm/include/llvm/Analysis/AliasAnalysis.h:1.8
--- llvm/include/llvm/Analysis/AliasAnalysis.h:1.7	Tue Nov 11 16:41:30 2003
+++ llvm/include/llvm/Analysis/AliasAnalysis.h	Wed Dec 10 23:05:02 2003
@@ -31,6 +31,7 @@
 #define LLVM_ANALYSIS_ALIAS_ANALYSIS_H
 
 #include "llvm/Support/CallSite.h"
+#include "llvm/Pass.h"
 
 namespace llvm {
 
@@ -158,6 +159,14 @@
   bool canInstructionRangeModify(const Instruction &I1, const Instruction &I2,
                                  const Value *Ptr, unsigned Size);
 };
+
+// Because of the way .a files work, we must force the BasicAA implementation to
+// be pulled in if the AliasAnalysis header is included.  Otherwise we run
+// the risk of AliasAnalysis being used, but the default implementation not
+// being linked into the tool that uses it.
+//
+extern void BasicAAStub();
+static IncludeFile HDR_INCLUDE_BASICAA_CPP((void*)&BasicAAStub);
 
 } // End llvm namespace
 




From gaeke at cs.uiuc.edu  Wed Dec 10 23:07:00 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 23:07:00 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/ValueNumbering.cpp
Message-ID: <200312110506.XAA20561@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

ValueNumbering.cpp updated: 1.8 -> 1.9

---
Log message:

IncludeFile hack to pull in BasicValueNumbering whenever ValueNumbering.h
is included.


---
Diffs of the changes:  (+2 -0)

Index: llvm/lib/Analysis/ValueNumbering.cpp
diff -u llvm/lib/Analysis/ValueNumbering.cpp:1.8 llvm/lib/Analysis/ValueNumbering.cpp:1.9
--- llvm/lib/Analysis/ValueNumbering.cpp:1.8	Tue Nov 25 14:11:00 2003
+++ llvm/lib/Analysis/ValueNumbering.cpp	Wed Dec 10 23:05:56 2003
@@ -191,4 +191,6 @@
       }
 }
 
+void BasicValueNumberingStub() { }
+
 } // End llvm namespace




From gaeke at cs.uiuc.edu  Wed Dec 10 23:07:04 2003
From: gaeke at cs.uiuc.edu (Brian Gaeke)
Date: Wed Dec 10 23:07:04 2003
Subject: [llvm-commits] CVS: llvm/include/llvm/Analysis/ValueNumbering.h
Message-ID: <200312110506.XAA20558@zion.cs.uiuc.edu>


Changes in directory llvm/include/llvm/Analysis:

ValueNumbering.h updated: 1.3 -> 1.4

---
Log message:

IncludeFile hack to pull in BasicValueNumbering whenever ValueNumbering.h
is included.


---
Diffs of the changes:  (+5 -0)

Index: llvm/include/llvm/Analysis/ValueNumbering.h
diff -u llvm/include/llvm/Analysis/ValueNumbering.h:1.3 llvm/include/llvm/Analysis/ValueNumbering.h:1.4
--- llvm/include/llvm/Analysis/ValueNumbering.h:1.3	Tue Nov 11 16:41:31 2003
+++ llvm/include/llvm/Analysis/ValueNumbering.h	Wed Dec 10 23:05:56 2003
@@ -21,6 +21,7 @@
 #define LLVM_ANALYSIS_VALUE_NUMBERING_H
 
 #include 
+#include "llvm/Pass.h"
 
 namespace llvm {
 
@@ -36,6 +37,10 @@
 
   virtual ~ValueNumbering();    // We want to be subclassed
 };
+
+extern void BasicValueNumberingStub();
+static IncludeFile
+HDR_INCLUDE_VALUENUMBERING_CPP((void*)&BasicValueNumberingStub);
 
 } // End llvm namespace
 




From lattner at cs.uiuc.edu  Wed Dec 10 23:46:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Wed Dec 10 23:46:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/BasicAliasAnalysis.cpp
Message-ID: <200312110545.XAA22609@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

BasicAliasAnalysis.cpp updated: 1.19 -> 1.20

---
Log message:

With Brian's change to AA.h we can now clean out this uglyness


---
Diffs of the changes:  (+1 -1)

Index: llvm/lib/Analysis/BasicAliasAnalysis.cpp
diff -u llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.19 llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.20
--- llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.19	Tue Nov 25 14:11:47 2003
+++ llvm/lib/Analysis/BasicAliasAnalysis.cpp	Wed Dec 10 23:44:59 2003
@@ -25,7 +25,7 @@
 using namespace llvm;
 
 // Make sure that anything that uses AliasAnalysis pulls in this file...
-namespace llvm { void BasicAAStub() {} }
+void llvm::BasicAAStub() {}
 
 namespace {
   struct BasicAliasAnalysis : public ImmutablePass, public AliasAnalysis {




From lattner at cs.uiuc.edu  Thu Dec 11 00:03:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 00:03:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/BasicAliasAnalysis.cpp
Message-ID: <200312110602.AAA23155@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

BasicAliasAnalysis.cpp updated: 1.20 -> 1.21

---
Log message:

Realize the gep P,  must aliases P.
This is a partial fix for PR 86


---
Diffs of the changes:  (+11 -1)

Index: llvm/lib/Analysis/BasicAliasAnalysis.cpp
diff -u llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.20 llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.21
--- llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.20	Wed Dec 10 23:44:59 2003
+++ llvm/lib/Analysis/BasicAliasAnalysis.cpp	Thu Dec 11 00:02:00 2003
@@ -175,12 +175,22 @@
         // If there is at least one non-zero constant index, we know they cannot
         // alias.
         bool ConstantFound = false;
+        bool AllZerosFound = true;
         for (unsigned i = 1, e = GEP->getNumOperands(); i != e; ++i)
-          if (const Constant *C = dyn_cast(GEP->getOperand(i)))
+          if (const Constant *C = dyn_cast(GEP->getOperand(i))) {
             if (!C->isNullValue()) {
               ConstantFound = true;
               break;
+            }
+          } else {
+            AllZerosFound = false;
           }
+
+        // If we have getelementptr , 0, 0, 0, 0, ... and V2 must aliases
+        // the ptr, the end result is a must alias also.
+        if (AllZerosFound)
+          return MustAlias;
+
         if (ConstantFound) {
           if (V2Size <= 1 && V1Size <= 1)  // Just pointer check?
             return NoAlias;




From lattner at cs.uiuc.edu  Thu Dec 11 00:07:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 00:07:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/BasicAliasAnalysis.cpp
Message-ID: <200312110606.AAA23568@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

BasicAliasAnalysis.cpp updated: 1.21 -> 1.22

---
Log message:

Fix a glaring bug


---
Diffs of the changes:  (+1 -0)

Index: llvm/lib/Analysis/BasicAliasAnalysis.cpp
diff -u llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.21 llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.22
--- llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.21	Thu Dec 11 00:02:00 2003
+++ llvm/lib/Analysis/BasicAliasAnalysis.cpp	Thu Dec 11 00:06:28 2003
@@ -180,6 +180,7 @@
           if (const Constant *C = dyn_cast(GEP->getOperand(i))) {
             if (!C->isNullValue()) {
               ConstantFound = true;
+              AllZerosFound = false;
               break;
             }
           } else {




From lattner at cs.uiuc.edu  Thu Dec 11 11:23:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 11:23:01 2003
Subject: [llvm-commits] CVS: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp
Message-ID: <200312111722.LAA11799@zion.cs.uiuc.edu>


Changes in directory llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++:

ary2.cpp updated: 1.2 -> 1.3

---
Log message:

Hrm, reading off the end of an array is _bad_.  PLEASE don't do that!  :)


---
Diffs of the changes:  (+2 -2)

Index: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp
diff -u llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp:1.2 llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp:1.3
--- llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp:1.2	Mon Oct  6 19:53:07 2003
+++ llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/ary2.cpp	Thu Dec 11 11:22:38 2003
@@ -1,5 +1,5 @@
 // -*- mode: c++ -*-
-// $Id: ary2.cpp,v 1.2 2003/10/07 00:53:07 lattner Exp $
+// $Id: ary2.cpp,v 1.3 2003/12/11 17:22:38 lattner Exp $
 // http://www.bagley.org/~doug/shootout/
 
 #include 
@@ -7,7 +7,7 @@
 
 int
 main(int argc, char *argv[]) {
-    int i, n = ((argc == 2) ? atoi(argv[1]) : 1);
+    int i, n = 10*((argc == 2) ? atoi(argv[1]) : 1);
     typedef std::vector ARY;
     ARY x(n);
     ARY y(n);




From lattner at cs.uiuc.edu  Thu Dec 11 11:51:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 11:51:01 2003
Subject: [llvm-commits] CVS: llvm/tools/gccas/gccas.cpp
Message-ID: <200312111750.LAA17183@zion.cs.uiuc.edu>


Changes in directory llvm/tools/gccas:

gccas.cpp updated: 1.88 -> 1.89

---
Log message:

Run tailcall elimination in a more logical place


---
Diffs of the changes:  (+1 -1)

Index: llvm/tools/gccas/gccas.cpp
diff -u llvm/tools/gccas/gccas.cpp:1.88 llvm/tools/gccas/gccas.cpp:1.89
--- llvm/tools/gccas/gccas.cpp:1.88	Sat Dec  6 19:44:18 2003
+++ llvm/tools/gccas/gccas.cpp	Thu Dec 11 11:50:32 2003
@@ -74,11 +74,11 @@
   addPass(PM, createTailDuplicationPass());      // Simplify cfg by copying code
   addPass(PM, createCFGSimplificationPass());    // Merge & remove BBs
   addPass(PM, createScalarReplAggregatesPass()); // Break up aggregate allocas
-  addPass(PM, createTailCallEliminationPass());  // Eliminate tail calls
   addPass(PM, createInstructionCombiningPass()); // Combine silly seq's
 
   addPass(PM, createReassociatePass());          // Reassociate expressions
   addPass(PM, createInstructionCombiningPass()); // Combine silly seq's
+  addPass(PM, createTailCallEliminationPass());  // Eliminate tail calls
   addPass(PM, createCFGSimplificationPass());    // Merge & remove BBs
   addPass(PM, createLICMPass());                 // Hoist loop invariants
   addPass(PM, createLoadValueNumberingPass());   // GVN for load instructions




From lattner at cs.uiuc.edu  Thu Dec 11 13:17:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 13:17:02 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll
Message-ID: <200312111916.NAA12055@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/BasicAA:

2003-12-11-ConstExprGEP.ll added (r1.1)

---
Log message:

New testcase for PR86


---
Diffs of the changes:  (+17 -0)

Index: llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll
diff -c /dev/null llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll:1.1
*** /dev/null	Thu Dec 11 13:16:41 2003
--- llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll	Thu Dec 11 13:16:30 2003
***************
*** 0 ****
--- 1,17 ----
+ ; This testcase consists of alias relations which should be completely
+ ; resolvable by basicaa, but require analysis of getelementptr constant exprs.
+ 
+ ; RUN: llvm-as < %s | opt -aa-eval -print-may-aliases 2>&1 -disable-output | not grep May:
+ 
+ %T = type { uint, [10 x ubyte] }
+ 
+ %G = external global %T
+ 
+ void %test() {
+   %D = getelementptr %T* %G, long 0, ubyte 0
+   %E = getelementptr %T* %G, long 0, ubyte 1, long 5
+   %F = getelementptr uint* getelementptr (%T* %G, long 0, ubyte 0), long 0
+   %G = getelementptr [10 x ubyte]* getelementptr (%T* %G, long 0, ubyte 1), long 0, long 5
+ 
+   ret void
+ }




From lattner at cs.uiuc.edu  Thu Dec 11 15:07:03 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 15:07:03 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll
Message-ID: <200312112106.PAA13145@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/LICM:

2003-08-04-TrappingInstOkHoist.ll updated: 1.3 -> 1.4

---
Log message:

Fix test that was failing because we were optimizing the program better than expected :)


---
Diffs of the changes:  (+2 -2)

Index: llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll
diff -u llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll:1.3 llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll:1.4
--- llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll:1.3	Tue Dec  9 10:49:12 2003
+++ llvm/test/Regression/Transforms/LICM/2003-08-04-TrappingInstOkHoist.ll	Thu Dec 11 15:06:43 2003
@@ -4,14 +4,14 @@
 ; RUN: llvm-as < %s | opt -licm | llvm-dis | grep -C2 "test" | grep div
 
 %X = global int 0
-declare void %foo()
+declare void %foo(int)
 
 int %test(bool %c) {
 	%A = load int *%X
 	br label %Loop
 Loop:
-	call void %foo()
 	%B = div int 4, %A  ;; Should have hoisted this div!
+	call void %foo(int %B)
         br bool %c, label %Loop, label %Out
 
 Out:




From criswell at cs.uiuc.edu  Thu Dec 11 15:22:04 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Thu Dec 11 15:22:04 2003
Subject: [llvm-commits] CVS: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
Message-ID: <200312112121.PAA26909@choi.cs.uiuc.edu>


Changes in directory llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++:

Makefile updated: 1.6 -> 1.7

---
Log message:

Added libsocket because the except program seems to want it on Sparc.
I'll have to dig into that sometime.



---
Diffs of the changes:  (+1 -1)

Index: llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile
diff -u llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.6 llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.7
--- llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile:1.6	Wed Dec 10 14:29:07 2003
+++ llvm/test/Programs/SingleSource/Benchmarks/Shootout-C++/Makefile	Thu Dec 11 15:21:08 2003
@@ -1,7 +1,7 @@
 LEVEL = ../../../../..
 CXXFLAGS += -Wno-deprecated
 CPPFLAGS += -Wno-deprecated
-LDFLAGS += -lm
+LDFLAGS += -lm -lsocket
 LIBS += -lstdc++
 
 include $(LEVEL)/test/Programs/SingleSource/Makefile.singlesrc




From lattner at cs.uiuc.edu  Thu Dec 11 15:48:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 15:48:02 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/LevelRaise.cpp
Message-ID: <200312112147.PAA15781@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms:

LevelRaise.cpp updated: 1.88 -> 1.89

---
Log message:

verifyFunction depends on dominator info, which levelraise does not declare
that it needs.  This is pretty scary code!  This fixes

Regression.Transforms.LevelRaise.2002-07-16-SourceAndDestCrash
Regression.Transforms.LevelRaise.2002-07-31-AssertionFailure



---
Diffs of the changes:  (+0 -5)

Index: llvm/lib/Transforms/LevelRaise.cpp
diff -u llvm/lib/Transforms/LevelRaise.cpp:1.88 llvm/lib/Transforms/LevelRaise.cpp:1.89
--- llvm/lib/Transforms/LevelRaise.cpp:1.88	Tue Nov 11 16:41:33 2003
+++ llvm/lib/Transforms/LevelRaise.cpp	Thu Dec 11 15:47:37 2003
@@ -21,7 +21,6 @@
 #include "llvm/Pass.h"
 #include "llvm/ConstantHandling.h"
 #include "llvm/Analysis/Expressions.h"
-#include "llvm/Analysis/Verifier.h"
 #include "llvm/Transforms/Utils/BasicBlockUtils.h"
 #include "Support/CommandLine.h"
 #include "Support/Debug.h"
@@ -295,8 +294,6 @@
                           << BB->getParent());
         }
 
-        DEBUG(assert(verifyFunction(*BB->getParent()) == false &&
-                     "Function broken!"));
         BI = BB->begin();  // Rescan basic block.  BI might be invalidated.
         ++NumExprTreesConv;
         return true;
@@ -320,8 +317,6 @@
         PRINT_PEEPHOLE1("CAST-DEST-EXPR-CONV:out", Src);
         DEBUG(std::cerr << "DONE CONVERTING EXPR TYPE: \n\n" << BB->getParent());
 
-        DEBUG(assert(verifyFunction(*BB->getParent()) == false &&
-                     "Function broken!"));
         BI = BB->begin();  // Rescan basic block.  BI might be invalidated.
         ++NumExprTreesConv;
         return true;




From lattner at cs.uiuc.edu  Thu Dec 11 15:49:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 15:49:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/CFGPrinter.cpp
Message-ID: <200312112148.PAA15802@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

CFGPrinter.cpp updated: 1.4 -> 1.5

---
Log message:

Finegrainify namespacification
Add new -print-cfg-only pass


---
Diffs of the changes:  (+23 -5)

Index: llvm/lib/Analysis/CFGPrinter.cpp
diff -u llvm/lib/Analysis/CFGPrinter.cpp:1.4 llvm/lib/Analysis/CFGPrinter.cpp:1.5
--- llvm/lib/Analysis/CFGPrinter.cpp:1.4	Tue Nov 11 16:41:31 2003
+++ llvm/lib/Analysis/CFGPrinter.cpp	Thu Dec 11 15:48:18 2003
@@ -25,8 +25,7 @@
 #include "llvm/Support/CFG.h"
 #include 
 #include 
-
-namespace llvm {
+using namespace llvm;
 
 /// CFGOnly flag - This is used to control whether or not the CFG graph printer
 /// prints out the contents of basic blocks or not.  This is acceptable because
@@ -34,6 +33,7 @@
 ///
 static bool CFGOnly = false;
 
+namespace llvm {
 template<>
 struct DOTGraphTraits : public DefaultDOTGraphTraits {
   static std::string getGraphName(const Function *F) {
@@ -87,6 +87,7 @@
     return "";
   }
 };
+}
 
 namespace {
   struct CFGPrinter : public FunctionPass {
@@ -112,7 +113,26 @@
 
   RegisterAnalysis P1("print-cfg",
                                   "Print CFG of function to 'dot' file");
-};
+
+  struct CFGOnlyPrinter : public CFGPrinter {
+    virtual bool runOnFunction(Function &F) {
+      bool OldCFGOnly = CFGOnly;
+      CFGOnly = true;
+      CFGPrinter::runOnFunction(F);
+      CFGOnly = OldCFGOnly;
+      return false;
+    }
+    void print(std::ostream &OS) const {}
+    
+    virtual void getAnalysisUsage(AnalysisUsage &AU) const {
+      AU.setPreservesAll();
+    }
+  };
+
+  RegisterAnalysis
+  P2("print-cfg-only",
+     "Print CFG of function to 'dot' file (with no function bodies)");
+}
 
 /// viewCFG - This function is meant for use from the debugger.  You can just
 /// say 'call F->viewCFG()' and a ghostview window should pop up from the
@@ -153,5 +173,3 @@
   viewCFG();
   CFGOnly = false;
 }
-
-} // End llvm namespace




From lattner at cs.uiuc.edu  Thu Dec 11 16:16:03 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 16:16:03 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/LICM/2003-12-11-SinkingToPHI.ll
Message-ID: <200312112215.QAA16757@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/LICM:

2003-12-11-SinkingToPHI.ll added (r1.1)

---
Log message:

New testcase that LICM is breaking crafty on


---
Diffs of the changes:  (+20 -0)

Index: llvm/test/Regression/Transforms/LICM/2003-12-11-SinkingToPHI.ll
diff -c /dev/null llvm/test/Regression/Transforms/LICM/2003-12-11-SinkingToPHI.ll:1.1
*** /dev/null	Thu Dec 11 16:15:21 2003
--- llvm/test/Regression/Transforms/LICM/2003-12-11-SinkingToPHI.ll	Thu Dec 11 16:15:11 2003
***************
*** 0 ****
--- 1,20 ----
+ ; RUN: llvm-as < %s | opt -licm | lli
+ 
+ implementation   ; Functions:
+ 
+ int %main() {
+ entry:
+ 	br label %Loop
+ 
+ Loop:
+ 	br bool true, label %LoopCont, label %Out
+ LoopCont:
+ 	%X = add int 1, 0
+ 	br bool true, label %Out, label %Loop
+ 
+ Out:
+ 	%V = phi int [ 2, %Loop], [ %X, %LoopCont]
+ 	%V2 = sub int %V, 1
+ 	ret int %V2
+ }
+ 




From lattner at cs.uiuc.edu  Thu Dec 11 16:24:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 16:24:01 2003
Subject: [llvm-commits] CVS: llvm/lib/Transforms/Scalar/LICM.cpp
Message-ID: <200312112223.QAA17603@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Transforms/Scalar:

LICM.cpp updated: 1.49 -> 1.50

---
Log message:

Fix LICM/2003-12-11-SinkingToPHI.ll, and quite possibly all of the other known problems in the universe.


---
Diffs of the changes:  (+11 -2)

Index: llvm/lib/Transforms/Scalar/LICM.cpp
diff -u llvm/lib/Transforms/Scalar/LICM.cpp:1.49 llvm/lib/Transforms/Scalar/LICM.cpp:1.50
--- llvm/lib/Transforms/Scalar/LICM.cpp:1.49	Wed Dec 10 16:35:56 2003
+++ llvm/lib/Transforms/Scalar/LICM.cpp	Thu Dec 11 16:23:32 2003
@@ -337,9 +337,18 @@
 /// exit blocks of the loop.
 ///
 bool LICM::isNotUsedInLoop(Instruction &I) {
-  for (Value::use_iterator UI = I.use_begin(), E = I.use_end(); UI != E; ++UI)
-    if (CurLoop->contains(cast(*UI)->getParent()))
+  for (Value::use_iterator UI = I.use_begin(), E = I.use_end(); UI != E; ++UI) {
+    Instruction *User = cast(*UI);
+    if (PHINode *PN = dyn_cast(User)) {
+      // PHI node uses occur in predecessor blocks!
+      for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i)
+        if (PN->getIncomingValue(i) == &I)
+          if (CurLoop->contains(PN->getIncomingBlock(i)))
+            return false;
+    } else if (CurLoop->contains(User->getParent())) {
       return false;
+    }
+  }
   return true;
 }
 




From lattner at cs.uiuc.edu  Thu Dec 11 16:45:02 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 16:45:02 2003
Subject: [llvm-commits] CVS: llvm/lib/Analysis/BasicAliasAnalysis.cpp
Message-ID: <200312112244.QAA19505@zion.cs.uiuc.edu>


Changes in directory llvm/lib/Analysis:

BasicAliasAnalysis.cpp updated: 1.22 -> 1.23

---
Log message:

Fix PR86.  This makes basicaa _SIGNIFICANLY_ more aggressive with getelementptr's


---
Diffs of the changes:  (+251 -121)

Index: llvm/lib/Analysis/BasicAliasAnalysis.cpp
diff -u llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.22 llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.23
--- llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.22	Thu Dec 11 00:06:28 2003
+++ llvm/lib/Analysis/BasicAliasAnalysis.cpp	Thu Dec 11 16:44:13 2003
@@ -41,11 +41,14 @@
     AliasResult alias(const Value *V1, unsigned V1Size,
                       const Value *V2, unsigned V2Size);
   private:
-    // CheckGEPInstructions - Check two GEP instructions of compatible types and
-    // equal number of arguments.  This checks to see if the index expressions
+    // CheckGEPInstructions - Check two GEP instructions with known
+    // must-aliasing base pointers.  This checks to see if the index expressions
     // preclude the pointers from aliasing...
-    AliasResult CheckGEPInstructions(GetElementPtrInst *GEP1, unsigned G1Size,
-                                     GetElementPtrInst *GEP2, unsigned G2Size);
+    AliasResult
+    CheckGEPInstructions(const Type* BasePtr1Ty, std::vector &GEP1Ops,
+                         unsigned G1Size,
+                         const Type *BasePtr2Ty, std::vector &GEP2Ops,
+                         unsigned G2Size);
   };
  
   // Register this pass...
@@ -89,6 +92,13 @@
   return 0;
 }
 
+static const User *isGEP(const Value *V) {
+  if (isa(V) ||
+      (isa(V) &&
+       cast(V)->getOpcode() == Instruction::GetElementPtr))
+    return cast(V);
+  return 0;
+}
 
 // alias - Provide a bunch of ad-hoc rules to disambiguate in common cases, such
 // as array references.  Note that this function is heavily tail recursive.
@@ -97,6 +107,14 @@
 AliasAnalysis::AliasResult
 BasicAliasAnalysis::alias(const Value *V1, unsigned V1Size,
                           const Value *V2, unsigned V2Size) {
+  // Strip off any constant expression casts if they exist
+  if (const ConstantExpr *CE = dyn_cast(V1))
+    if (CE->getOpcode() == Instruction::Cast)
+      V1 = CE->getOperand(0);
+  if (const ConstantExpr *CE = dyn_cast(V2))
+    if (CE->getOpcode() == Instruction::Cast)
+      V2 = CE->getOperand(0);
+
   // Strip off constant pointer refs if they exist
   if (const ConstantPointerRef *CPR = dyn_cast(V1))
     V1 = CPR->getValue();
@@ -145,19 +163,67 @@
     return NoAlias;                    // Unique values don't alias null
   }
 
-  // If we have two gep instructions with identical indices, return an alias
-  // result equal to the alias result of the original pointer...
+  // If we have two gep instructions with must-alias'ing base pointers, figure
+  // out if the indexes to the GEP tell us anything about the derived pointer.
+  // Note that we also handle chains of getelementptr instructions as well as
+  // constant expression getelementptrs here.
   //
-  if (const GetElementPtrInst *GEP1 = dyn_cast(V1))
-    if (const GetElementPtrInst *GEP2 = dyn_cast(V2))
-      if (GEP1->getNumOperands() == GEP2->getNumOperands() &&
-          GEP1->getOperand(0)->getType() == GEP2->getOperand(0)->getType()) {
-        AliasResult GAlias =
-          CheckGEPInstructions((GetElementPtrInst*)GEP1, V1Size,
-                               (GetElementPtrInst*)GEP2, V2Size);
-        if (GAlias != MayAlias)
-          return GAlias;
+  if (isGEP(V1) && isGEP(V2)) {
+    // Drill down into the first non-gep value, to test for must-aliasing of
+    // the base pointers.
+    const Value *BasePtr1 = V1, *BasePtr2 = V2;
+    do {
+      BasePtr1 = cast(BasePtr1)->getOperand(0);
+    } while (isGEP(BasePtr1) &&
+             cast(BasePtr1)->getOperand(1) == 
+       Constant::getNullValue(cast(BasePtr1)->getOperand(1)->getType()));
+    do {
+      BasePtr2 = cast(BasePtr2)->getOperand(0);
+    } while (isGEP(BasePtr2) &&
+             cast(BasePtr2)->getOperand(1) == 
+       Constant::getNullValue(cast(BasePtr2)->getOperand(1)->getType()));
+
+    // Do the base pointers alias?
+    AliasResult BaseAlias = alias(BasePtr1, V1Size, BasePtr2, V2Size);
+    if (BaseAlias == NoAlias) return NoAlias;
+    if (BaseAlias == MustAlias) {
+      // If the base pointers alias each other exactly, check to see if we can
+      // figure out anything about the resultant pointers, to try to prove
+      // non-aliasing.
+
+      // Collect all of the chained GEP operands together into one simple place
+      std::vector GEP1Ops(cast(V1)->op_begin()+1,
+                                  cast(V1)->op_end());
+      std::vector GEP2Ops(cast(V2)->op_begin()+1,
+                                  cast(V2)->op_end());
+
+      // Accumulate all of the chained indexes into the operand arrays
+      BasePtr1 = cast(V1)->getOperand(0);
+      BasePtr2 = cast(V2)->getOperand(0);
+      while (const User *G = isGEP(BasePtr1)) {
+        if (!isa(GEP1Ops[0]) ||
+            !cast(GEP1Ops[0])->isNullValue())
+          break;  // Don't handle folding arbitrary pointer offsets yet...
+        GEP1Ops.erase(GEP1Ops.begin());
+        GEP1Ops.insert(GEP1Ops.begin(), G->op_begin()+1, G->op_end());
+        BasePtr1 = G->getOperand(0);
       }
+      while (const User *G = isGEP(BasePtr2)) {
+        if (!isa(GEP2Ops[0]) ||
+            !cast(GEP2Ops[0])->isNullValue())
+          break;  // Don't handle folding arbitrary pointer offsets yet...
+        GEP2Ops.erase(GEP2Ops.begin());
+        GEP2Ops.insert(GEP2Ops.begin(), G->op_begin()+1, G->op_end());
+        BasePtr2 = G->getOperand(0);
+      }
+      
+      AliasResult GAlias =
+        CheckGEPInstructions(BasePtr1->getType(), GEP1Ops, V1Size,
+                             BasePtr2->getType(), GEP2Ops, V2Size);
+      if (GAlias != MayAlias)
+        return GAlias;
+    }
+  }
 
   // Check to see if these two pointers are related by a getelementptr
   // instruction.  If one pointer is a GEP with a non-zero index of the other
@@ -219,45 +285,60 @@
   return MayAlias;
 }
 
-static Value *CheckArrayIndicesForOverflow(const Type *PtrTy,
-                                           const std::vector &Indices,
-                                           const ConstantInt *Idx) {
-  if (const ConstantSInt *IdxS = dyn_cast(Idx)) {
-    if (IdxS->getValue() < 0)   // Underflow on the array subscript?
-      return Constant::getNullValue(Type::LongTy);
-    else {                       // Check for overflow
-      const ArrayType *ATy =
-        cast(GetElementPtrInst::getIndexedType(PtrTy, Indices,true));
-      if (IdxS->getValue() >= (int64_t)ATy->getNumElements())
-        return ConstantSInt::get(Type::LongTy, ATy->getNumElements()-1);
+/// CheckGEPInstructions - Check two GEP instructions with known must-aliasing
+/// base pointers.  This checks to see if the index expressions preclude the
+/// pointers from aliasing...
+AliasAnalysis::AliasResult BasicAliasAnalysis::
+CheckGEPInstructions(const Type* BasePtr1Ty, std::vector &GEP1Ops,
+                     unsigned G1S,
+                     const Type *BasePtr2Ty, std::vector &GEP2Ops,
+                     unsigned G2S) {
+  // We currently can't handle the case when the base pointers have different
+  // primitive types.  Since this is uncommon anyway, we are happy being
+  // extremely conservative.
+  if (BasePtr1Ty != BasePtr2Ty)
+    return MayAlias;
+
+  const Type *GEPPointerTy = BasePtr1Ty;
+
+  // Find the (possibly empty) initial sequence of equal values... which are not
+  // necessarily constants.
+  unsigned NumGEP1Operands = GEP1Ops.size(), NumGEP2Operands = GEP2Ops.size();
+  unsigned MinOperands = std::min(NumGEP1Operands, NumGEP2Operands);
+  unsigned MaxOperands = std::max(NumGEP1Operands, NumGEP2Operands);
+  unsigned UnequalOper = 0;
+  while (UnequalOper != MinOperands &&
+         GEP1Ops[UnequalOper] == GEP2Ops[UnequalOper]) {
+    // Advance through the type as we go...
+    ++UnequalOper;
+    if (const CompositeType *CT = dyn_cast(BasePtr1Ty))
+      BasePtr1Ty = CT->getTypeAtIndex(GEP1Ops[UnequalOper-1]);
+    else {
+      // If all operands equal each other, then the derived pointers must
+      // alias each other...
+      BasePtr1Ty = 0;
+      assert(UnequalOper == NumGEP1Operands && UnequalOper == NumGEP2Operands &&
+             "Ran out of type nesting, but not out of operands?");
+      return MustAlias;
     }
   }
-  return (Value*)Idx;  // Everything is acceptable.
-}
 
-// CheckGEPInstructions - Check two GEP instructions of compatible types and
-// equal number of arguments.  This checks to see if the index expressions
-// preclude the pointers from aliasing...
-//
-AliasAnalysis::AliasResult
-BasicAliasAnalysis::CheckGEPInstructions(GetElementPtrInst *GEP1, unsigned G1S, 
-                                         GetElementPtrInst *GEP2, unsigned G2S){
-  // Do the base pointers alias?
-  AliasResult BaseAlias = alias(GEP1->getOperand(0), G1S,
-                                GEP2->getOperand(0), G2S);
-  if (BaseAlias != MustAlias)   // No or May alias: We cannot add anything...
-    return BaseAlias;
-  
-  // Find the (possibly empty) initial sequence of equal values...
-  unsigned NumGEPOperands = GEP1->getNumOperands();
-  unsigned UnequalOper = 1;
-  while (UnequalOper != NumGEPOperands &&
-         GEP1->getOperand(UnequalOper) == GEP2->getOperand(UnequalOper))
-    ++UnequalOper;
+  // If we have seen all constant operands, and run out of indexes on one of the
+  // getelementptrs, check to see if the tail of the leftover one is all zeros.
+  // If so, return mustalias.
+  if (UnequalOper == MinOperands && MinOperands != MaxOperands) {
+    if (GEP1Ops.size() < GEP2Ops.size()) std::swap(GEP1Ops, GEP2Ops);
     
-  // If all operands equal each other, then the derived pointers must
-  // alias each other...
-  if (UnequalOper == NumGEPOperands) return MustAlias;
+    bool AllAreZeros = true;
+    for (unsigned i = UnequalOper; i != MaxOperands; ++i)
+      if (!isa(GEP1Ops[i]) ||
+          !cast(GEP1Ops[i])->isNullValue()) {
+        AllAreZeros = false;
+        break;
+      }
+    if (AllAreZeros) return MustAlias;
+  }
+
     
   // So now we know that the indexes derived from the base pointers,
   // which are known to alias, are different.  We can still determine a
@@ -271,101 +352,150 @@
   // Scan for the first operand that is constant and unequal in the
   // two getelemenptrs...
   unsigned FirstConstantOper = UnequalOper;
-  for (; FirstConstantOper != NumGEPOperands; ++FirstConstantOper) {
-    const Value *G1Oper = GEP1->getOperand(FirstConstantOper);
-    const Value *G2Oper = GEP2->getOperand(FirstConstantOper);
+  for (; FirstConstantOper != MinOperands; ++FirstConstantOper) {
+    const Value *G1Oper = GEP1Ops[FirstConstantOper];
+    const Value *G2Oper = GEP2Ops[FirstConstantOper];
+    
     if (G1Oper != G2Oper &&   // Found non-equal constant indexes...
         isa(G1Oper) && isa(G2Oper)) {
-      // Make sure they are comparable...  and make sure the GEP with
-      // the smaller leading constant is GEP1.
-      ConstantBool *Compare =
-        *cast(GEP1->getOperand(FirstConstantOper)) >
-        *cast(GEP2->getOperand(FirstConstantOper));
+      // Make sure they are comparable (ie, not constant expressions)...  and
+      // make sure the GEP with the smaller leading constant is GEP1.
+      ConstantBool *Compare = *cast(G1Oper) > *cast(G2Oper);
       if (Compare) {  // If they are comparable...
         if (Compare->getValue())
-          std::swap(GEP1, GEP2);  // Make GEP1 < GEP2
+          std::swap(GEP1Ops, GEP2Ops);  // Make GEP1 < GEP2
         break;
       }
     }
+    BasePtr1Ty = cast(BasePtr1Ty)->getTypeAtIndex(G1Oper);
   }
   
-  // No constant operands, we cannot tell anything...
-  if (FirstConstantOper == NumGEPOperands) return MayAlias;
+  // No shared constant operands, and we ran out of common operands.  At this
+  // point, the GEP instructions have run through all of their operands, and we
+  // haven't found evidence that there are any deltas between the GEP's.
+  // However, one GEP may have more operands than the other.  If this is the
+  // case, there may still be hope.  This this now.
+  if (FirstConstantOper == MinOperands) {
+    // Make GEP1Ops be the longer one if there is a longer one.
+    if (GEP1Ops.size() < GEP2Ops.size())
+      std::swap(GEP1Ops, GEP2Ops);
+
+    // Is there anything to check?
+    if (GEP1Ops.size() > MinOperands) {
+      for (unsigned i = FirstConstantOper; i != MaxOperands; ++i)
+        if (isa(GEP1Ops[i]) && !isa(GEP1Ops[i]) &&
+            !cast(GEP1Ops[i])->isNullValue()) {
+          // Yup, there's a constant in the tail.  Set all variables to
+          // constants in the GEP instruction to make it suiteable for
+          // TargetData::getIndexedOffset.
+          for (i = 0; i != MaxOperands; ++i)
+            if (!isa(GEP1Ops[i]) || isa(GEP1Ops[i]))
+              GEP1Ops[i] = Constant::getNullValue(GEP1Ops[i]->getType());
+          // Okay, now get the offset.  This is the relative offset for the full
+          // instruction.
+          const TargetData &TD = getTargetData();
+          int64_t Offset1 = TD.getIndexedOffset(GEPPointerTy, GEP1Ops);
+
+          // Now crop off any constants from the end...
+          GEP1Ops.resize(MinOperands);
+          int64_t Offset2 = TD.getIndexedOffset(GEPPointerTy, GEP1Ops);
+        
+          // If the tail provided a bit enough offset, return noalias!
+          if ((uint64_t)(Offset2-Offset1) >= SizeMax)
+            return NoAlias;
+        }
+    }
+    
+    // Couldn't find anything useful.
+    return MayAlias;
+  }
 
   // If there are non-equal constants arguments, then we can figure
   // out a minimum known delta between the two index expressions... at
   // this point we know that the first constant index of GEP1 is less
   // than the first constant index of GEP2.
-  //
-  std::vector Indices1;
-  Indices1.reserve(NumGEPOperands-1);
-  
-  for (gep_type_iterator I = gep_type_begin(GEP1);
-       I.getOperandNum() != FirstConstantOper; ++I)
-    if (isa(*I))
-      Indices1.push_back(I.getOperand());
-    else
-      Indices1.push_back(Constant::getNullValue(Type::LongTy));
-
-  std::vector Indices2;
-  Indices2.reserve(NumGEPOperands-1);
-  Indices2 = Indices1;           // Copy the zeros prefix...
-  
-  // Add the two known constant operands...
-  Indices1.push_back((Value*)GEP1->getOperand(FirstConstantOper));
-  Indices2.push_back((Value*)GEP2->getOperand(FirstConstantOper));
+
+  // Advance BasePtr[12]Ty over this first differing constant operand.
+  BasePtr2Ty = cast(BasePtr1Ty)->getTypeAtIndex(GEP2Ops[FirstConstantOper]);
+  BasePtr1Ty = cast(BasePtr1Ty)->getTypeAtIndex(GEP1Ops[FirstConstantOper]);
   
-  const Type *GEPPointerTy = GEP1->getOperand(0)->getType();
+  // We are going to be using TargetData::getIndexedOffset to determine the
+  // offset that each of the GEP's is reaching.  To do this, we have to convert
+  // all variable references to constant references.  To do this, we convert the
+  // initial equal sequence of variables into constant zeros to start with.
+  for (unsigned i = 0; i != FirstConstantOper; ++i) {
+    if (!isa(GEP1Ops[i]) || isa(GEP1Ops[i]) ||
+        !isa(GEP2Ops[i]) || isa(GEP2Ops[i])) {
+      GEP1Ops[i] = Constant::getNullValue(GEP1Ops[i]->getType());
+      GEP2Ops[i] = Constant::getNullValue(GEP2Ops[i]->getType());
+    }
+  }
+
+  // We know that GEP1Ops[FirstConstantOper] & GEP2Ops[FirstConstantOper] are ok
+
   
   // Loop over the rest of the operands...
-  for (unsigned i = FirstConstantOper+1; i != NumGEPOperands; ++i) {
-    const Value *Op1 = GEP1->getOperand(i);
-    const Value *Op2 = GEP2->getOperand(i);
-    if (Op1 == Op2) {   // If they are equal, use a zero index...
-      if (!isa(Op1)) {
-        Indices1.push_back(Constant::getNullValue(Op1->getType()));
-        Indices2.push_back(Indices1.back());
-      } else {
-        Indices1.push_back((Value*)Op1);
-        Indices2.push_back((Value*)Op2);
-      }
+  for (unsigned i = FirstConstantOper+1; i != MaxOperands; ++i) {
+    const Value *Op1 = i < GEP1Ops.size() ? GEP1Ops[i] : 0;
+    const Value *Op2 = i < GEP2Ops.size() ? GEP2Ops[i] : 0;
+    // If they are equal, use a zero index...
+    if (Op1 == Op2 && BasePtr1Ty == BasePtr2Ty) {
+      if (!isa(Op1) || isa(Op1))
+        GEP1Ops[i] = GEP2Ops[i] = Constant::getNullValue(Op1->getType());
+      // Otherwise, just keep the constants we have.
     } else {
-      if (const ConstantInt *Op1C = dyn_cast(Op1)) {
-        // If this is an array index, make sure the array element is in range...
-        if (i != 1)   // The pointer index can be "out of range"
-          Op1 = CheckArrayIndicesForOverflow(GEPPointerTy, Indices1, Op1C);
-
-        Indices1.push_back((Value*)Op1);
-      } else {
-        // GEP1 is known to produce a value less than GEP2.  To be
-        // conservatively correct, we must assume the largest possible constant
-        // is used in this position.  This cannot be the initial index to the
-        // GEP instructions (because we know we have at least one element before
-        // this one with the different constant arguments), so we know that the
-        // current index must be into either a struct or array.  Because we know
-        // it's not constant, this cannot be a structure index.  Because of
-        // this, we can calculate the maximum value possible.
-        //
-        const ArrayType *ElTy =
-          cast(GEP1->getIndexedType(GEPPointerTy, Indices1, true));
-        Indices1.push_back(ConstantSInt::get(Type::LongTy,
-                                             ElTy->getNumElements()-1));
+      if (Op1) {
+        if (const ConstantInt *Op1C = dyn_cast(Op1)) {
+          // If this is an array index, make sure the array element is in range.
+          if (const ArrayType *AT = dyn_cast(BasePtr1Ty))
+            if (Op1C->getRawValue() >= AT->getNumElements())
+              return MayAlias;  // Be conservative with out-of-range accesses
+          
+        } else {
+          // GEP1 is known to produce a value less than GEP2.  To be
+          // conservatively correct, we must assume the largest possible
+          // constant is used in this position.  This cannot be the initial
+          // index to the GEP instructions (because we know we have at least one
+          // element before this one with the different constant arguments), so
+          // we know that the current index must be into either a struct or
+          // array.  Because we know it's not constant, this cannot be a
+          // structure index.  Because of this, we can calculate the maximum
+          // value possible.
+          //
+          if (const ArrayType *AT = dyn_cast(BasePtr1Ty))
+            GEP1Ops[i] = ConstantSInt::get(Type::LongTy,AT->getNumElements()-1);
+        }
       }
       
-      if (const ConstantInt *Op1C = dyn_cast(Op2)) {
-        // If this is an array index, make sure the array element is in range...
-        if (i != 1)   // The pointer index can be "out of range"
-          Op1 = CheckArrayIndicesForOverflow(GEPPointerTy, Indices2, Op1C);
-
-        Indices2.push_back((Value*)Op2);
+      if (Op2) {
+        if (const ConstantInt *Op2C = dyn_cast(Op2)) {
+          // If this is an array index, make sure the array element is in range.
+          if (const ArrayType *AT = dyn_cast(BasePtr1Ty))
+            if (Op2C->getRawValue() >= AT->getNumElements())
+              return MayAlias;  // Be conservative with out-of-range accesses
+        } else {  // Conservatively assume the minimum value for this index
+          GEP2Ops[i] = Constant::getNullValue(Op2->getType());
+        }
       }
-      else // Conservatively assume the minimum value for this index
-        Indices2.push_back(Constant::getNullValue(Op2->getType()));
+    }
+
+    if (BasePtr1Ty && Op1) {
+      if (const CompositeType *CT = dyn_cast(BasePtr1Ty))
+        BasePtr1Ty = CT->getTypeAtIndex(GEP1Ops[i]);
+      else
+        BasePtr1Ty = 0;
+    }
+
+    if (BasePtr2Ty && Op2) {
+      if (const CompositeType *CT = dyn_cast(BasePtr2Ty))
+        BasePtr2Ty = CT->getTypeAtIndex(GEP2Ops[i]);
+      else
+        BasePtr2Ty = 0;
     }
   }
   
-  int64_t Offset1 = getTargetData().getIndexedOffset(GEPPointerTy, Indices1);
-  int64_t Offset2 = getTargetData().getIndexedOffset(GEPPointerTy, Indices2);
+  int64_t Offset1 = getTargetData().getIndexedOffset(GEPPointerTy, GEP1Ops);
+  int64_t Offset2 = getTargetData().getIndexedOffset(GEPPointerTy, GEP2Ops);
   assert(Offset1 < Offset2 &&"There is at least one different constant here!");
 
   if ((uint64_t)(Offset2-Offset1) >= SizeMax) {




From lattner at cs.uiuc.edu  Thu Dec 11 16:50:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 16:50:01 2003
Subject: [llvm-commits] CVS: llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll
Message-ID: <200312112249.QAA19608@zion.cs.uiuc.edu>


Changes in directory llvm/test/Regression/Transforms/BasicAA:

2003-12-11-ConstExprGEP.ll updated: 1.1 -> 1.2

---
Log message:

Rename variable to avoid confusing it with the global G


---
Diffs of the changes:  (+1 -1)

Index: llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll
diff -u llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll:1.1 llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll:1.2
--- llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll:1.1	Thu Dec 11 13:16:30 2003
+++ llvm/test/Regression/Transforms/BasicAA/2003-12-11-ConstExprGEP.ll	Thu Dec 11 16:49:32 2003
@@ -11,7 +11,7 @@
   %D = getelementptr %T* %G, long 0, ubyte 0
   %E = getelementptr %T* %G, long 0, ubyte 1, long 5
   %F = getelementptr uint* getelementptr (%T* %G, long 0, ubyte 0), long 0
-  %G = getelementptr [10 x ubyte]* getelementptr (%T* %G, long 0, ubyte 1), long 0, long 5
+  %X = getelementptr [10 x ubyte]* getelementptr (%T* %G, long 0, ubyte 1), long 0, long 5
 
   ret void
 }




From criswell at cs.uiuc.edu  Thu Dec 11 17:25:04 2003
From: criswell at cs.uiuc.edu (John Criswell)
Date: Thu Dec 11 17:25:04 2003
Subject: [llvm-commits] CVS: llvm/test/QMTest/expectations.linux.qmr expectations.sunos.qmr
Message-ID: <200312112318.RAA02335@choi.cs.uiuc.edu>


Changes in directory llvm/test/QMTest:

expectations.linux.qmr updated: 1.4 -> 1.5
expectations.sunos.qmr updated: 1.4 -> 1.5

---
Log message:

Updated the expectations to match the current results of the LLVM tree.



---
Diffs of the changes:  (+2 -2)

Index: llvm/test/QMTest/expectations.linux.qmr
diff -u llvm/test/QMTest/expectations.linux.qmr:1.4 llvm/test/QMTest/expectations.linux.qmr:1.5
--- llvm/test/QMTest/expectations.linux.qmr:1.4	Wed Dec 10 15:27:38 2003
+++ llvm/test/QMTest/expectations.linux.qmr	Thu Dec 11 17:18:33 2003
@@ -3,13 +3,13 @@
 qoq}q(U
_Result__kindqUtestqU_Result__outcomeqUPASSqU_Result__annotationsq}U_Result__idq	U0Regression.Assembler.2002-08-16-ConstExprInlinedq
 U_Result__contextq(cqm.test.context
 Context
-qo}q
(U_Context__propertiesq}U_Context__temporariesq}ubub.(hoq}q(hhhhh}h	U(Regression.Transforms.PruneEH.simpletestqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	U/Regression.CBackend.2002-08-19-HardConstantExprqh(ho}q(h}h}ubub.(hoq}q(hhhhh}h	URegression.Jello.test-shiftqh(ho}q(h}h}ubub.(hoq}q(hhhUFAILqh}h	U.Regression.Transforms.CorrelatedExprs.looptestqh(ho}q (h}h}ubub.(hoq!}q"(hhhhh}h	U-Regression.Linker.2003-08-23-GlobalVarLinkingq#h(ho}q$(h}h}ubub.(hoq%}q&(hhhhh}h	U?Regression.Transforms.Inline.2003-09-22-PHINodesInExceptionDestq'h(ho}q((h}h}ubub.(hoq)}q*(hhhhh}h	U,Regression.CFrontend.2003-02-12-NonlocalGotoq+h(ho}q,(h}h}ubub.(hoq-}q.(hhhhh}h	U2Regression.Transforms.FunctionResolve.retmismatch1q/h(ho}q0(h}h}ubub.(hoq1}q2(hhhhh}h	U8Regression.Transforms.ADCE.2003-01-22-PredecessorProblemq3h(ho}q4(h}h}ubub.(hoq5}q6(hhhhh}h	U=Regression.Transforms.LevelRaise.2002-10-08-!
 VarArgCallInfLoopq7h(ho}q8(h}h}ubub.(hoq9}q:(hhhhh}h	U5Regression.CFrontend.2003-07-22-ArrayAccessTypeSafetyq;h(ho}q<(h}h}ubub.(hoq=}q>(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U-Regression.C++Frontend.2003-08-28-SaveExprBugqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U"Regression.Jello.2003-06-05-PHIBugqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq_h(ho}q`(h}h}ubub.(!
 hoqa}qb(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-
07-DominatorProblemqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U3Regression.Transforms.PiNodeInserter.substitutetestqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U$Regression.Jello.2003-01-04-LoopTestqoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U)Regression.CFrontend.2002-07-14-MiscTestsqwh(ho}qx(h}h}ubub.(hoqy}qz(hhhhh}h	U(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U-Regression.C++Frontend.2003-08-28-SaveExprBugqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U"Regression.Jello.2003-06-05-PHIBugqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq_h(ho}q`(h}h}ubub.(!
 hoqa}qb(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-
07-DominatorProblemqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U>Regression.Transforms.InstCombine.2003-06-22-ConstantExprCrashqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U3Regression.Transforms.PiNodeInserter.substitutetestqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U$Regression.Jello.2003-01-04-LoopTestqsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qwh(ho}qx(h}h}ubub.(hoqy}qz(hhhhh}h	U)Regression.CFrontend.2002-07-14-MiscTestsq{h(ho}q|(h}h}ubub.(hoq}}q~(hhhhh}h	U(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U-Regression.C++Frontend.2003-08-28-SaveExprBugqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U"Regression.Jello.2003-06-05-PHIBugqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq_h(ho}q`(h}h}ubub.(!
 hoqa}qb(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-
07-DominatorProblemqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U3Regression.Transforms.PiNodeInserter.substitutetestqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U$Regression.Jello.2003-01-04-LoopTestqoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U)Regression.CFrontend.2002-07-14-MiscTestsqwh(ho}qx(h}h}ubub.(hoqy}qz(hhhhh}h	U(hhhhh}h	U Regression.Reoptimizer.ticm.ticmq?h(ho}q@(h}h}ubub.(hoqA}qB(hhhhh}h	U8Regression.C++Frontend.2003-09-29-ArgumentNumberMismatchqCh(ho}qD(h}h}ubub.(hoqE}qF(hhhhh}h	U+Regression.CFrontend.2002-09-19-StarInLabelqGh(ho}qH(h}h}ubub.(hoqI}qJ(hhhhh}h	U.Regression.CFrontend.2003-08-23-LocalUnionTestqKh(ho}qL(h}h}ubub.(hoqM}qN(hhhhh}h	U/Regression.Linker.2003-04-26-NullPtrLinkProblemqOh(ho}qP(h}h}ubub.(hoqQ}qR(hhhhh}h	U)Regression.Transforms.Reassociate.subtestqSh(ho}qT(h}h}ubub.(hoqU}qV(hhhhh}h	U)Regression.Linker.2002-08-20-ConstantExprqWh(ho}qX(h}h}ubub.(hoqY}qZ(hhhhh}h	U=Regression.Transforms.Reassociate.2002-05-15-AgressiveSubMoveq[h(ho}q\(h}h}ubub.(hoq]}q^(hhhhh}h	UARegression.Transforms.CorrelatedExprs.2002-10-07-DominatorProblemq_!
 h(ho}q`(h}h}ubub.(hoqa}qb(hhhhh}h	U3Regression.Transforms.Pi
NodeInserter.substitutetestqch(ho}qd(h}h}ubub.(hoqe}qf(hhhhh}h	U4Regression.Transforms.SCCP.2003-08-26-InvokeHandlingqgh(ho}qh(h}h}ubub.(hoqi}qj(hhhhh}h	U-Regression.CFrontend.2002-02-18-64bitConstantqkh(ho}ql(h}h}ubub.(hoqm}qn(hhhhh}h	U1Regression.Assembler.2002-04-04-PureVirtMethCall2qoh(ho}qp(h}h}ubub.(hoqq}qr(hhhhh}h	U(Regression.Reoptimizer.BinInterface.testqsh(ho}qt(h}h}ubub.(hoqu}qv(hhhhh}h	U


Changes in directory llvm/lib/Analysis:

BasicAliasAnalysis.cpp updated: 1.23 -> 1.24

---
Log message:

Factor out some duplicated code, implement the rest of the cases in
BasicAA/2003-12-11-ConstExprGEP.ll


---
Diffs of the changes:  (+40 -43)

Index: llvm/lib/Analysis/BasicAliasAnalysis.cpp
diff -u llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.23 llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.24
--- llvm/lib/Analysis/BasicAliasAnalysis.cpp:1.23	Thu Dec 11 16:44:13 2003
+++ llvm/lib/Analysis/BasicAliasAnalysis.cpp	Thu Dec 11 17:20:16 2003
@@ -100,6 +100,26 @@
   return 0;
 }
 
+static const Value *GetGEPOperands(const Value *V, std::vector &GEPOps){
+  assert(GEPOps.empty() && "Expect empty list to populate!");
+  GEPOps.insert(GEPOps.end(), cast(V)->op_begin()+1,
+                cast(V)->op_end());
+
+  // Accumulate all of the chained indexes into the operand array
+  V = cast(V)->getOperand(0);
+
+  while (const User *G = isGEP(V)) {
+    if (!isa(GEPOps[0]) ||
+        !cast(GEPOps[0])->isNullValue())
+      break;  // Don't handle folding arbitrary pointer offsets yet...
+    GEPOps.erase(GEPOps.begin());   // Drop the zero index
+    GEPOps.insert(GEPOps.begin(), G->op_begin()+1, G->op_end());
+    V = G->getOperand(0);
+  }
+  return V;
+}
+
+
 // alias - Provide a bunch of ad-hoc rules to disambiguate in common cases, such
 // as array references.  Note that this function is heavily tail recursive.
 // Hopefully we have a smart C++ compiler.  :)
@@ -192,31 +212,10 @@
       // non-aliasing.
 
       // Collect all of the chained GEP operands together into one simple place
-      std::vector GEP1Ops(cast(V1)->op_begin()+1,
-                                  cast(V1)->op_end());
-      std::vector GEP2Ops(cast(V2)->op_begin()+1,
-                                  cast(V2)->op_end());
-
-      // Accumulate all of the chained indexes into the operand arrays
-      BasePtr1 = cast(V1)->getOperand(0);
-      BasePtr2 = cast(V2)->getOperand(0);
-      while (const User *G = isGEP(BasePtr1)) {
-        if (!isa(GEP1Ops[0]) ||
-            !cast(GEP1Ops[0])->isNullValue())
-          break;  // Don't handle folding arbitrary pointer offsets yet...
-        GEP1Ops.erase(GEP1Ops.begin());
-        GEP1Ops.insert(GEP1Ops.begin(), G->op_begin()+1, G->op_end());
-        BasePtr1 = G->getOperand(0);
-      }
-      while (const User *G = isGEP(BasePtr2)) {
-        if (!isa(GEP2Ops[0]) ||
-            !cast(GEP2Ops[0])->isNullValue())
-          break;  // Don't handle folding arbitrary pointer offsets yet...
-        GEP2Ops.erase(GEP2Ops.begin());
-        GEP2Ops.insert(GEP2Ops.begin(), G->op_begin()+1, G->op_end());
-        BasePtr2 = G->getOperand(0);
-      }
-      
+      std::vector GEP1Ops, GEP2Ops;
+      BasePtr1 = GetGEPOperands(V1, GEP1Ops);
+      BasePtr2 = GetGEPOperands(V2, GEP2Ops);
+
       AliasResult GAlias =
         CheckGEPInstructions(BasePtr1->getType(), GEP1Ops, V1Size,
                              BasePtr2->getType(), GEP2Ops, V2Size);
@@ -229,21 +228,24 @@
   // instruction.  If one pointer is a GEP with a non-zero index of the other
   // pointer, we know they cannot alias.
   //
-  if (isa(V2)) {
+  if (isGEP(V2)) {
     std::swap(V1, V2);
     std::swap(V1Size, V2Size);
   }
 
   if (V1Size != ~0U && V2Size != ~0U)
-    if (const GetElementPtrInst *GEP = dyn_cast(V1)) {
-      AliasResult R = alias(GEP->getOperand(0), V1Size, V2, V2Size);
+    if (const User *GEP = isGEP(V1)) {
+      std::vector GEPOperands;
+      const Value *BasePtr = GetGEPOperands(V1, GEPOperands);
+
+      AliasResult R = alias(BasePtr, V1Size, V2, V2Size);
       if (R == MustAlias) {
         // If there is at least one non-zero constant index, we know they cannot
         // alias.
         bool ConstantFound = false;
         bool AllZerosFound = true;
-        for (unsigned i = 1, e = GEP->getNumOperands(); i != e; ++i)
-          if (const Constant *C = dyn_cast(GEP->getOperand(i))) {
+        for (unsigned i = 0, e = GEPOperands.size(); i != e; ++i)
+          if (const Constant *C = dyn_cast(GEPOperands[i])) {
             if (!C->isNullValue()) {
               ConstantFound = true;
               AllZerosFound = false;
@@ -266,17 +268,13 @@
           // the size of the argument... build an index vector that is equal to
           // the arguments provided, except substitute 0's for any variable
           // indexes we find...
-          
-          std::vector Indices;
-          Indices.reserve(GEP->getNumOperands()-1);
-          for (unsigned i = 1; i != GEP->getNumOperands(); ++i)
-            if (const Constant *C = dyn_cast(GEP->getOperand(i)))
-              Indices.push_back((Value*)C);
-            else
-              Indices.push_back(Constant::getNullValue(Type::LongTy));
-          const Type *Ty = GEP->getOperand(0)->getType();
-          int Offset = getTargetData().getIndexedOffset(Ty, Indices);
-          if (Offset >= (int)V2Size || Offset <= -(int)V1Size)
+          for (unsigned i = 0; i != GEPOperands.size(); ++i)
+            if (!isa(GEPOperands[i]) ||
+                isa(GEPOperands[i]))
+              GEPOperands[i] =Constant::getNullValue(GEPOperands[i]->getType());
+          int64_t Offset = getTargetData().getIndexedOffset(BasePtr->getType(),
+                                                            GEPOperands);
+          if (Offset >= (int64_t)V2Size || Offset <= -(int64_t)V1Size)
             return NoAlias;
         }
       }
@@ -326,7 +324,7 @@
   // If we have seen all constant operands, and run out of indexes on one of the
   // getelementptrs, check to see if the tail of the leftover one is all zeros.
   // If so, return mustalias.
-  if (UnequalOper == MinOperands && MinOperands != MaxOperands) {
+  if (UnequalOper == MinOperands) {
     if (GEP1Ops.size() < GEP2Ops.size()) std::swap(GEP1Ops, GEP2Ops);
     
     bool AllAreZeros = true;
@@ -432,7 +430,6 @@
   }
 
   // We know that GEP1Ops[FirstConstantOper] & GEP2Ops[FirstConstantOper] are ok
-
   
   // Loop over the rest of the operands...
   for (unsigned i = FirstConstantOper+1; i != MaxOperands; ++i) {




From lattner at cs.uiuc.edu  Thu Dec 11 17:32:01 2003
From: lattner at cs.uiuc.edu (Chris Lattner)
Date: Thu Dec 11 17:32:01 2003
Subject: [llvm-commits] CVS: llvm/docs/ReleaseNotes.html
Message-ID: <200312112331.RAA23182@zion.cs.uiuc.edu>


Changes in directory llvm/docs:

ReleaseNotes.html updated: 1.73 -> 1.74

---
Log message:

Update release notes for PR86 and C++ support


---
Diffs of the changes:  (+9 -8)

Index: llvm/docs/ReleaseNotes.html
diff -u llvm/docs/ReleaseNotes.html:1.73 llvm/docs/ReleaseNotes.html:1.74
--- llvm/docs/ReleaseNotes.html:1.73	Wed Dec 10 22:40:58 2003
+++ llvm/docs/ReleaseNotes.html	Thu Dec 11 17:31:21 2003
@@ -76,9 +76,9 @@
 includes a few new features, such as a simple profiler, support for Mac OS/X,
 and better interoperability with external source bases.
 
-At this time, LLVM is known to correctly compile the SPEC CPU2000 benchmarks
-(X86 only), the Olden benchmarks, and the Ptrdist benchmarks along with
-many other programs.  LLVM now also works with a fairly broad variety of
+
At this time, LLVM is known to correctly compile the C & C++ SPEC CPU2000
+benchmarks (X86 only), the Olden benchmarks, and the Ptrdist benchmarks along
+with many other programs.  LLVM now also works with a broad variety of
 C++ programs, though it has still received much less testing than the C
 front-end.
 
@@ -126,10 +126,11 @@
   
 
configure
llvm/projects
configure
llvm/projects
-basicaa
configure
llvm/projects
-basicaa
-licm
-basicaa
longjmp
longjmp
vector_size
dllimport
dllexport
transparent_union
unused
deprecated
may_alias
deprecated
may_alias
aligned
packed
configure
llvm/projects
configure
llvm/projects
-tailcallelim
-licm
-licm
-basicaa
llvm/doc/
longjmp
253.perlbmk
longjmp
253.perlbmk
llvm/doc/
vfwscanf
wcstof
sparcv9-sun-solaris2.8/libstdc++-v3/config.h
_GLIBCXX_USE_WCHAR_T
+ % gmake all
+ % gmake install
 
--with-llvmgccdir=directory
--enable-spec2000=directory
+Copyright (c) 1994
+Hewlett-Packard Company
+
+Permission to use, copy, modify, distribute and sell this software
+and its documentation for any purpose is hereby granted without fee,
+provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear
+in supporting documentation.  Hewlett-Packard Company makes no
+representations about the suitability of this software for any
+purpose.  It is provided "as is" without express or implied warranty.
+
+Copyright (c) 1996, 1997, 1998, 1999
+Silicon Graphics Computer Systems, Inc.
+
+Permission to use, copy, modify, distribute and sell this software
+and its documentation for any purpose is hereby granted without fee,
+provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear
+in supporting documentation.  Silicon Graphics makes no
+representations about the suitability of this software for any
+purpose.  It is provided "as is" without express or implied warranty.
+
getModRefInfo
-load-vn
AliasSetTracker
-no-aa
-print-alias-sets
-count-aa
-aa-eval
AliasResult
ModRefResult
call
invoke
+   int i;
+   char C[2];
+   char A[10]; 
+   /* ... */
+   for (i = 0; i != 10; ++i) {
+     C[0] = A[i];          /* One byte store */
+     C[1] = A[9-i];        /* One byte store */
+   }
+ 
basicaa
C[0]
C[1]
+   int i;
+   char C[2];
+   char A[10]; 
+   /* ... */
+   for (i = 0; i != 10; ++i) {
+     ((short*)C)[0] = A[i];  /* Two byte store! */
+     C[1] = A[9-i];          /* One byte store */
+   }
+ 
&C[0]
getModRefInfo
getModRefInfo
no-aa
Pass
FunctionPass
BasicBlockPass
ImmutablePass
AliasAnalysis
RegisterAnalysisGroup
AliasAnalysis
getAnalysisUsage
InitializeAliasAnalysis
getAnalysisUsage
AliasAnalysis::getAnalysisUsage
+     void getAnalysisUsage(AnalysisUsage &AU) const {
+       AliasAnalysis::getAnalysisUsage(AU);
+       // declare your dependencies here.
+     }
+ 
InitializeAliasAnalysis
run
Pass
runOnFunction
FunctionPass
runOnBasicBlock
BasicBlockPass
InitializeAliasAnalysis
ImmutablePass
Pass
+     bool run(Module &M) {
+       InitializeAliasAnalysis(this);
+       // Perform analysis here...
+       return false;
+     }
+ 
basicaa
no-aa
-basic-aa -ds-aa -andersens-aa -licm
getAnalysisUsage
-load-vn
load-vn
load
load-vn
AliasSetTracker
AliasSetTracker
add
begin()
end()
AliasSet
AliasSetTracker
-no-aa
-no-aa
basicaa
-print-alias-sets
-print-alias-sets
analyze
AliasSetTracker
AliasSetTracker
-count-aa
-count-aa
+   $ opt -basicaa -count-aa -ds-aa -count-aa -licm
+ 
-licm
-ds-aa
-basicaa
-ds-aa
-aa-eval
-aa-eval
+  % cd llvm
+  % ./configure [options...]
+  % gmake tools-only
+ 
+  % set path = ( `cd llvm/tools/Debug && pwd` $path )
+ 
+  % pwd
+  /usr/local/example/cfrontend/src
+  % cd ..
+  % mkdir build install
+  % set CFEINSTALL = `pwd`/install
+ 
+  % cd build
+  % ../src/configure --prefix=$CFEINSTALL --disable-threads --disable-nls --disable-shared \
+    --enable-languages=c,c++
+  % gmake all-gcc
+  % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
+  % gmake all; gmake install
+ 
+  % cd build
+  % ../src/configure --prefix=$CFEINSTALL --disable-threads --disable-nls --disable-shared \
+    --enable-languages=c,c++ --host=sparcv9-sun-solaris2.8
+  % gmake all-gcc
+  % setenv LLVM_LIB_SEARCH_PATH `pwd`/gcc 
+  % gmake all
+ 
vfwscanf
wcstof
sparcv9-sun-solaris2.8/libstdc++-v3/config.h
_GLIBCXX_USE_WCHAR_T
+  % gmake all
+  % gmake install
+ 
gccas
gccld
.init
.fini
-L
gccld
MULTILIB_OPTIONS
+  % gmake -C runtime
+  % mkdir $CFEINSTALL/bytecode-libs
+  % gmake -C runtime install
+  % setenv LLVM_LIB_SEARCH_PATH $CFEINSTALL/bytecode-libs
+ 
test/Programs
+ Copyright (c) 1994
+ Hewlett-Packard Company
+ 
+ Permission to use, copy, modify, distribute and sell this software
+ and its documentation for any purpose is hereby granted without fee,
+ provided that the above copyright notice appear in all copies and
+ that both that copyright notice and this permission notice appear
+ in supporting documentation.  Hewlett-Packard Company makes no
+ representations about the suitability of this software for any
+ purpose.  It is provided "as is" without express or implied warranty.
+ 
+ Copyright (c) 1996, 1997, 1998, 1999
+ Silicon Graphics Computer Systems, Inc.
+ 
+ Permission to use, copy, modify, distribute and sell this software
+ and its documentation for any purpose is hereby granted without fee,
+ provided that the above copyright notice appear in all copies and
+ that both that copyright notice and this permission notice appear
+ in supporting documentation.  Silicon Graphics makes no
+ representations about the suitability of this software for any
+ purpose.  It is provided "as is" without express or implied warranty.
+ 
+ //===-- llvm/Instruction.h - Instruction class definition -------*- C++ -*-===//
+ //
+ // This file contains the declaration of the Instruction class, which is the
+ // base class for all of the VM instructions.
+ //
+ //===----------------------------------------------------------------------===//
+ 
-*- C++
+ -*-
//
/* */
#if 0
#endif
gcc
gcc
+   if (V = getValue()) {
+     ..
+   }
+ 
gcc
==
+   if ((V = getValue())) {
+     ..
+   }
+ 
gcc
gcc
gcc
-Wall
+ -Winline -W -Wwrite-strings -Wno-unused
.cpp
.cpp
#include
#include
.cpp
assert
<cassert>
+   inline Value *getOperand(unsigned i) { 
+     assert(i < Operands.size() && "getOperand() out of range!");
+     return Operands[i]; 
+   }
+ 
+   assert(Ty->isPointerType() && "Can't allocate a non pointer type!");
+ 
+   assert((Opcode == Shl || Opcode == Shr) && "ShiftInst Opcode invalid!");
+ 
+   assert(idx < getNumSuccessors() && "Successor # out of range!");
+ 
+   assert(V1.getType() == V2.getType() && "Constant types must be identical!");
+ 
+   assert(isa<PHINode>(Succ->front()) && "Only works on PHId BBs!");
+ 
endl
+   cout << endl;
+   cout << "\n" << flush;
+ 
"\n"
+ From: Ross Smith <ross.s at ihug.co.nz>
+ Newsgroups: comp.lang.c++.moderated
+ Subject: Writing iterators (was: Re: Non-template functions that take iterators)
+ Date: 28 Jun 2001 12:07:10 -0400
+ 
+ Andre Majorel wrote:
+ > Any pointers handy on "writing STL-compatible iterators for
+ > dummies ?"
+ 
+ I'll give it a try...
+ 
+ The usual situation requiring user-defined iterators is that you have
+ a type that bears some resemblance to an STL container, and you want
+ to provide iterators so it can be used with STL algorithms. You need
+ to ask three questions:
+ 
+ First, is this simply a wrapper for an underlying collection of
+ objects that's held somewhere as a real STL container, or is it a
+ "virtual container" for which iteration is (under the hood) more
+ complicated than simply incrementing some underlying iterator (or
+ pointer or index or whatever)? In the former case you can frequently
+ get away with making your container's iterators simply typedefs for
+ those of the underlying container; your begin() function would call
+ member_container.begin(), and so on.
+ 
+ Second, do you only need read-only iterators, or do you need separate
+ read-only (const) and read-write (non-const) iterators?
+ 
+ Third, which kind of iterator (input, output, forward, bidirectional,
+ or random access) is appropriate? If you're familiar with the
+ properties of the iterator types (if not, visit
+ http://www.sgi.com/tech/stl/), the appropriate choice should be
+ obvious from the semantics of the container.
+ 
+ I'll start with forward iterators, as the simplest case that's likely
+ to come up in normal code. Input and output iterators have some odd
+ properties and rarely need to be implemented in user code; I'll leave
+ them out of discussion. Bidirectional and random access iterators are
+ covered below.
+ 
+ The exact behaviour of a forward iterator is spelled out in the
+ Standard in terms of a set of expressions with specified behaviour,
+ rather than a set of member functions, which leaves some leeway in how
+ you actually implement it. Typically it looks something like this
+ (I'll start with the const-iterator-only situation):
+ 
+   #include <iterator>
+ 
+   class container {
+     public:
+       typedef something_or_other value_type;
+       class const_iterator:
+         public std::iterator<std::forward_iterator_tag, value_type> {
+           friend class container;
+           public:
+             const value_type& operator*() const;
+             const value_type* operator->() const;
+             const_iterator& operator++();
+             const_iterator operator++(int);
+             friend bool operator==(const_iterator lhs,
+                                    const_iterator rhs);
+             friend bool operator!=(const_iterator lhs,
+                                    const_iterator rhs);
+           private:
+             //...
+         };
+       //...
+   };
+ 
+ An iterator should always be derived from an instantiation of the
+ std::iterator template. The iterator's life cycle functions
+ (constructors, destructor, and assignment operator) aren't declared
+ here; in most cases the compiler-generated ones are sufficient. The
+ container needs to be a friend of the iterator so that the container's
+ begin() and end() functions can fill in the iterator's private members
+ with the appropriate values.
+ 
+ [Chris's Note: I prefer to not make my iterators friends.  Instead, two
+ ctor's are provided for the iterator class: one to start at the end of the
+ container, and one at the beginning.  Typically this is done by providing
+ two constructors with different signatures.]
+ 
+ There are normally only three member functions that need nontrivial
+ implementations; the rest are just boilerplate.
+ 
+   const container::value_type&
+     container::const_iterator::operator*() const {
+       // find the element and return a reference to it
+     }
+ 
+   const container::value_type*
+     container::const_iterator::operator->() const {
+       return &**this;
+     }
+ 
+ If there's an underlying real container, operator*() can just return a
+ reference to the appropriate element. If there's no actual container
+ and the elements need to be generated on the fly -- what I think of as
+ a "virtual container" -- things get a bit more complicated; you'll
+ probably need to give the iterator a value_type member object, and
+ fill it in when you need to. This might be done as part of the
+ increment operator (below), or if the operation is nontrivial, you
+ might choose the "lazy" approach and only generate the actual value
+ when one of the dereferencing operators is called.
+ 
+ The operator->() function is just boilerplate around a call to
+ operator*().
+ 
+   container::const_iterator&
+     container::const_iterator::operator++() {
+       // the incrementing logic goes here
+       return *this;
+     }
+ 
+   container::const_iterator
+     container::const_iterator::operator++(int) {
+       const_iterator old(*this);
+       ++*this;
+       return old;
+     }
+ 
+ Again, the incrementing logic will usually be trivial if there's a
+ real container involved, more complicated if you're working with a
+ virtual container. In particular, watch out for what happens when you
+ increment past the last valid item -- this needs to produce an
+ iterator that will compare equal to container.end(), and making this
+ work is often nontrivial for virtual containers.
+ 
+ The post-increment function is just boilerplate again (and
+ incidentally makes it obvious why all the experts recommend using
+ pre-increment wherever possible).
+ 
+   bool operator==(container::const_iterator lhs,
+                   container::const_iterator rhs) {
+     // equality comparison goes here
+   }
+ 
+   bool operator!=(container::const_iterator lhs,
+                   container::const_iterator rhs) {
+     return !(lhs == rhs);
+   }
+ 
+ For a real container, the equality comparison will usually just
+ compare the underlying iterators (or pointers or indices or whatever).
+ The semantics of comparisons for virtual container iterators are often
+ tricky. Remember that iterator comparison only needs to be defined for
+ iterators into the same container, so you can often simplify things by
+ taking for granted that lhs and rhs both point into the same container
+ object. Again, the second function is just boilerplate.
+ 
+ It's a matter of taste whether iterator arguments are passed by value
+ or reference; I've shown tham passed by value to reduce clutter, but
+ if the iterator contains several data members, passing by reference
+ may be better.
+ 
+ That convers the const-iterator-only situation. When we need separate
+ const and mutable iterators, one small complication is added beyond
+ the simple addition of a second class.
+ 
+   class container {
+     public:
+       typedef something_or_other value_type;
+       class const_iterator;
+       class iterator:
+         public std::iterator<std::forward_iterator_tag, value_type> {
+           friend class container;
+           friend class container::const_iterator;
+           public:
+             value_type& operator*() const;
+             value_type* operator->() const;
+             iterator& operator++();
+             iterator operator++(int);
+             friend bool operator==(iterator lhs, iterator rhs);
+             friend bool operator!=(iterator lhs, iterator rhs);
+           private:
+             //...
+         };
+       class const_iterator:
+         public std::iterator<std::forward_iterator_tag, value_type> {
+           friend class container;
+           public:
+             const_iterator();
+             const_iterator(const iterator& i);
+             const value_type& operator*() const;
+             const value_type* operator->() const;
+             const_iterator& operator++();
+             const_iterator operator++(int);
+             friend bool operator==(const_iterator lhs,
+                                    const_iterator rhs);
+             friend bool operator!=(const_iterator lhs,
+                                    const_iterator rhs);
+           private:
+             //...
+         };
+       //...
+   };
+ 
+ There needs to be a conversion from iterator to const_iterator (so
+ that mixed-type operations, such as comparison between an iterator and
+ a const_iterator, will work). This is done here by giving
+ const_iterator a conversion constructor from iterator (equivalently,
+ we could have given iterator an operator const_iterator()), which
+ requires const_iterator to be a friend of iterator, so it can copy its
+ data members. (It also requires the addition of an explicit default
+ constructor to const_iterator, since the existence of another
+ user-defined constructor inhibits the compiler-defined one.)
+ 
+ Bidirectional iterators add just two member functions to forward
+ iterators:
+ 
+   class iterator:
+     public std::iterator<std::bidirectional_iterator_tag, value_type> {
+       public:
+         //...
+         iterator& operator--();
+         iterator operator--(int);
+         //...
+     };
+ 
+ I won't detail the implementations, they're obvious variations on
+ operator++().
+ 
+ Random access iterators add several more member and friend functions:
+ 
+   class iterator:
+     public std::iterator<std::random_access_iterator_tag, value_type> {
+       public:
+         //...
+         iterator& operator+=(difference_type rhs);
+         iterator& operator-=(difference_type rhs);
+         friend iterator operator+(iterator lhs, difference_type rhs);
+         friend iterator operator+(difference_type lhs, iterator rhs);
+         friend iterator operator-(iterator lhs, difference_type rhs);
+         friend difference_type operator-(iterator lhs, iterator rhs);
+         friend bool operator<(iterator lhs, iterator rhs);
+         friend bool operator>(iterator lhs, iterator rhs);
+         friend bool operator<=(iterator lhs, iterator rhs);
+         friend bool operator>=(iterator lhs, iterator rhs);
+         //...
+     };
+ 
+   container::iterator&
+     container::iterator::operator+=(container::difference_type rhs) {
+       // add rhs to iterator position
+       return *this;
+     }
+ 
+   container::iterator&
+     container::iterator::operator-=(container::difference_type rhs) {
+       // subtract rhs from iterator position
+       return *this;
+     }
+ 
+   container::iterator operator+(container::iterator lhs,
+                                 container::difference_type rhs) {
+     return iterator(lhs) += rhs;
+   }
+ 
+   container::iterator operator+(container::difference_type lhs,
+                                 container::iterator rhs) {
+     return iterator(rhs) += lhs;
+   }
+ 
+   container::iterator operator-(container::iterator lhs,
+                                 container::difference_type rhs) {
+     return iterator(lhs) -= rhs;
+   }
+ 
+   container::difference_type operator-(container::iterator lhs,
+                                        container::iterator rhs) {
+     // calculate distance between iterators
+   }
+ 
+   bool operator<(container::iterator lhs, container::iterator rhs) {
+     // perform less-than comparison
+   }
+ 
+   bool operator>(container::iterator lhs, container::iterator rhs) {
+     return rhs < lhs;
+   }
+ 
+   bool operator<=(container::iterator lhs, container::iterator rhs) {
+     return !(rhs < lhs);
+   }
+ 
+   bool operator>=(container::iterator lhs, container::iterator rhs) {
+     return !(lhs < rhs);
+   }
+ 
+ Four of the functions (operator+=(), operator-=(), the second
+ operator-(), and operator<()) are nontrivial; the rest are
+ boilerplate.
+ 
+ One feature of the above code that some experts may disapprove of is
+ the declaration of all the free functions as friends, when in fact
+ only a few of them need direct access to the iterator's private data.
+ I originally got into the habit of doing this simply to keep the
+ declarations together; declaring some functions inside the class and
+ some outside seemed awkward. Since then, though, I've been told that
+ there's a subtle difference in the way name lookup works for functions
+ declared inside a class (as friends) and outside, so keeping them
+ together in the class is probably a good idea for practical as well as
+ aesthetic reasons.
+ 
+ I hope all this is some help to anyone who needs to write their own
+ STL-like containers and iterators.
+ 
+ -- 
+ Ross Smith <ross.s at ihug.co.nz> The Internet Group, Auckland, New Zealand
+ 
cl::ConsumeAfter
--help
cl::ParseCommandLineOptions
cl::ParseEnvironmentOptions
cl::opt
cl::list
cl::alias
parser<t>
parser<bool>
parser<string>
parser<int>
parser<double>
parser<float>
--help
ls
ls
+ -lad
ld
-lmalloc -L/usr/lib
+   #include "Support/CommandLine.h"
+ 
+ int main(int argc, char **argv) {
+   cl::ParseCommandLineOptions(argc, argv);
+   ...
+ }
+ 
-o <filename>

+ 
+ cl::opt<string> OutputFilename("o", cl::desc("Specify output filename"), cl::value_desc("filename"));
+ 
OutputFilename
o
cl::opt
cl::list
--help
+ USAGE: compiler [options]
+ 
+ OPTIONS:
+   -help             - display available options (--help-hidden for more)
+   -o <filename>     - Specify output filename
+ 
string
+   ...
+   ofstream Output(OutputFilename.c_str());
+   if (Out.good()) ...
+   ...
+ 
cl::desc(...)
-filename.c
+ cl::opt<string> InputFilename(cl::Positional, cl::desc("<input file>"), cl::init("-"));
+ 
cl::init
cl::init
cl::Required
cl::init
+ cl::opt<string> InputFilename(cl::Positional, cl::desc("<input file>"), cl::Required);
+ 
+ cl::opt<string> InputFilename(cl::Positional, cl::Required, cl::desc("<input file>"));
+ 
cl::Required
--help
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   -help             - display available options (--help-hidden for more)
+   -o <filename>     - Specify output filename
+ 
-f
--quiet
-q
+ cl::opt<bool> Force ("f", cl::desc("Overwrite output files"));
+ cl::opt<bool> Quiet ("quiet", cl::desc("Don't print informational messages"));
+ cl::opt<bool> Quiet2("q", cl::desc("Don't print informational messages"), cl::Hidden);
+ 
Force
Quiet
Quiet2
-q
cl::Hidden
--help
--help-hidden
true
false
+  compiler -f          # No value, 'Force' == true
+  compiler -f=true     # Value specified, 'Force' == true
+  compiler -f=TRUE     # Value specified, 'Force' == true
+  compiler -f=FALSE    # Value specified, 'Force' == false
+ 
compiler
+ -f=foo
strtol
strtod
compiler --help
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   -f     - Overwrite output files
+   -o     - Override output filename
+   -quiet - Don't print informational messages
+   -help  - display available options (--help-hidden for more)
+ 
opt --help-hidden
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   -f     - Overwrite output files
+   -o     - Override output filename
+   -q     - Don't print informational messages
+   -quiet - Don't print informational messages
+   -help  - display available options (--help-hidden for more)
+ 
cl::opt
+ ...
+   if (!Quiet && !Quiet2) printInformationalMessage(...);
+ ...
+ 
cl::alias
-q
-quiet
+ cl::opt<bool> Force ("f", cl::desc("Overwrite output files"));
+ cl::opt<bool> Quiet ("quiet", cl::desc("Don't print informational messages"));
+ cl::alias     QuietA("q", cl::desc("Alias for -quiet"), cl::aliasopt(Quiet));
+ 
-q
Quiet
cl::aliasopt
Quiet
-help
--help-hidden
+ output
+ ...
+   if (!Quiet) printInformationalMessage(...);
+ ...
+ 
cl::alias
std::string
bool
int
int*
-g
-O0
-O1
-O2
opt -O3 -O2
-O1
+ enum OptLevel {
+   g, O1, O2, O3
+ };
+ 
+ cl::opt<OptLevel> OptimizationLevel(cl::desc("Choose optimization level:"),
+   cl::values(
+     clEnumVal(g , "No optimizations, enable debugging"),
+     clEnumVal(O1, "Enable trivial optimizations"),
+     clEnumVal(O2, "Enable default optimizations"),
+     clEnumVal(O3, "Enable expensive optimizations"),
+    0));
+ 
+ ...
+   if (OptimizationLevel >= O2) doPartialRedundancyElimination(...);
+ ...
+ 
OptimizationLevel
OptLevel
0
clEnumVal
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   Choose optimization level:
+     -g          - No optimizations, enable debugging
+     -O1         - Enable trivial optimizations
+     -O2         - Enable default optimizations
+     -O3         - Enable expensive optimizations
+   -f            - Overwrite output files
+   -help         - display available options (--help-hidden for more)
+   -o <filename> - Specify output filename
+   -quiet        - Don't print informational messages
+ 
g
+ enum OptLevel {
+   Debug, O1, O2, O3
+ };
+ 
+ cl::opt<OptLevel> OptimizationLevel(cl::desc("Choose optimization level:"),
+   cl::values(
+    clEnumValN(Debug, "g", "No optimizations, enable debugging"),
+     clEnumVal(O1        , "Enable trivial optimizations"),
+     clEnumVal(O2        , "Enable default optimizations"),
+     clEnumVal(O3        , "Enable expensive optimizations"),
+    0));
+ 
+ ...
+   if (OptimizationLevel == Debug) outputDebugInfo(...);
+ ...
+ 
clEnumValN
clEnumVal
--debug-level=none
--debug-level=quick
--debug-level=detailed
+ enum DebugLev {
+   nodebuginfo, quick, detailed
+ };
+ 
+ // Enable Debug Options to be specified on the command line
+ cl::opt<DebugLev> DebugLevel("debug_level", cl::desc("Set the debugging level:"),
+   cl::values(
+     clEnumValN(nodebuginfo, "none", "disable debug information"),
+      clEnumVal(quick,               "enable quick debug information"),
+      clEnumVal(detailed,            "enable detailed debug information"),
+     0));
+ 
enum
+ DebugLev
--help
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   Choose optimization level:
+     -g          - No optimizations, enable debugging
+     -O1         - Enable trivial optimizations
+     -O2         - Enable default optimizations
+     -O3         - Enable expensive optimizations
+   -debug_level  - Set the debugging level:
+     =none       - disable debug information
+     =quick      - enable quick debug information
+     =detailed   - enable detailed debug information
+   -f            - Overwrite output files
+   -help         - display available options (--help-hidden for more)
+   -o <filename> - Specify output filename
+   -quiet        - Don't print informational messages
+ 
"debug_level"
compiler -dce -constprop -inline -dce -strip
cl::list
+ enum Opts {
+   // 'inline' is a C++ keyword, so name it 'inlining'
+   dce, constprop, inlining, strip
+ };
+ 
cl::list
+ cl::list<Opts> OptimizationList(cl::desc("Available Optimizations:"),
+   cl::values(
+     clEnumVal(dce               , "Dead Code Elimination"),
+     clEnumVal(constprop         , "Constant Propagation"),
+    clEnumValN(inlining, "inline", "Procedure Integration"),
+     clEnumVal(strip             , "Strip Symbols"),
+   0));
+ 
std::vector<enum Opts>
+   for (unsigned i = 0; i != OptimizationList.size(); ++i)
+     switch (OptimizationList[i])
+        ...
+ 
cl::list
cl::opt
.o
+ ...
+ cl::list<std::string> InputFilenames(cl::Positional, cl::desc("<Input files>"), cl::OneOrMore);
+ ...
+ 
vector<string>
cl::OneOrMore
.o
man
man
cl::ParseCommandLineOptions
+ int main(int argc, char **argv) {
+   cl::ParseCommandLineOptions(argc, argv, " CommandLine compiler example\n\n"
+                               "  This program blah blah blah...\n");
+   ...
+ }
+ 
+ OVERVIEW: CommandLine compiler example
+ 
+   This program blah blah blah...
+ 
+ USAGE: compiler [options] <input file>
+ 
+ OPTIONS:
+   ...
+   -help             - display available options (--help-hidden for more)
+   -o <filename>     - Specify output filename
+ 
grep
+ cl::opt<string> Regex   (cl::Positional, cl::desc("<regular expression>"), cl::Required);
+ cl::opt<string> Filename(cl::Positional, cl::desc("<input file>"), cl::init("-"));
+ 
--help
+ USAGE: spiffygrep [options] <regular expression> <input file>
+ 
+ OPTIONS:
+   -help - display available options (--help-hidden for more)
+ 
grep
-foo
-foo
grep
+   $ spiffygrep '-foo' test.txt
+   Unknown command line argument '-foo'.  Try: spiffygrep --help'
+ 
+   $ grep '-foo' test.txt
+   grep: illegal option -- f
+   grep: illegal option -- o
+   grep: illegal option -- o
+   Usage: grep -hblcnsviw pattern file . . .
+ 
--
--
--
+   $ spiffygrep -- -foo test.txt
+     ...output...
+ 
cl::ConsumeAfter
cl::ConsumeAfter
/bin/sh
/bin/sh
-x
+ cl::opt<string> Script(cl::Positional, cl::desc("<input script>"), cl::init("-"));
+ cl::list<string>  Argv(cl::ConsumeAfter, cl::desc("<program arguments>..."));
+ cl::opt<bool>    Trace("x", cl::desc("Enable trace output"));
+ 
+ USAGE: spiffysh [options] <input script> <program arguments>...
+ 
+ OPTIONS:
+   -help - display available options (--help-hidden for more)
+   -x    - Enable trace output
+ 
spiffysh -x test.sh
+ -a -x -y bar
Trace
Script
test.sh
Argv
["-a", "-x", "-y", "bar"]
cl::ConsumeAfter
cl::ConsumeAfter
cl::ConsumeAfter
-debug
CommandLine.h
+ // DebugFlag.h - Get access to the '-debug' command line option
+ //
+ 
+ // DebugFlag - This boolean is set to true if the '-debug' command line option
+ // is specified.  This should probably not be referenced directly, instead, use
+ // the DEBUG macro below.
+ //
+ extern bool DebugFlag;
+ 
+ // DEBUG macro - This macro should be used by code to emit debug information.
+ // In the '-debug' option is specified on the command line, and if this is a
+ // debug build, then the code specified as the option to the macro will be
+ // executed.  Otherwise it will not be.  Example:
+ //
+ // DEBUG(cerr << "Bitset contains: " << Bitset << "\n");
+ //
+ #ifdef NDEBUG
+ #define DEBUG(X)
+ #else
+ #define DEBUG(X) \
+   do { if (DebugFlag) { X; } } while (0)
+ #endif
+ 
DEBUG()
DebugFlag
DebugFlag
+ bool DebugFlag;      // the actual value
+ static cl::opt<bool, true>       // The parser
+ Debug("debug", cl::desc("Enable debug output"), cl::Hidden,
+       cl::location(DebugFlag));
+ 
true
DebugFlag
+ cl::opt<bool> Quiet("quiet");
+ 
cl::desc
--help
cl::value_desc
--help
cl::init
cl::init
cl::location
cl::location
cl::init
cl::location
cl::aliasopt
cl::values
--help
clEnumVal
clEnumValN
cl::opt
cl::list
--help
--help
--help
cl::NotHidden
cl::Hidden
cl::ReallyHidden
--help
--help-hidden
cl::NotHidden
cl::opt
cl::list
cl::Hidden
cl::alias
--help
--help-hidden
cl::ReallyHidden
cl::Optional
cl::opt
cl::alias
cl::ZeroOrMore
cl::list
cl::Required
cl::OneOrMore
cl::ConsumeAfter
cl::init
cl::init
cl::opt
-index-depth=17
-o a.out
cl::ValueOptional
bool
-foo=true
-foo true
cl::ValueRequired
-o
+ a.out
cl::ValueDisallowed
-foo=true
cl::NormalFormatting
cl::Positional
cl::ConsumeAfter
cl::Prefix
-lmalloc -L/usr/lib'
l
L
cl::Grouping
ls
ls -labF
parse(string OrigInput) {
string input = OrigInput;
if (isOption(input)) return getOption(input).parse();
while (!isOption(input) && !input.empty()) input.pop_back();
if (input.empty()) return error();
if (getOption(input).isPrefix())

+   return getOption(input).parse(input);
while (!input.empty()) {    // Must be grouping options

+   getOption(input).parse();

+   OrigInput.erase(OrigInput.begin(), OrigInput.begin()+input.length());

+   input = OrigInput;

+   while (!isOption(input) && !input.empty()) input.pop_back();

+ }
if (!OrigInput.empty()) error();
}
cl::CommaSeparated
cl::CommaSeparated
-foo=a -foo=b -foo=c
-foo=a,b,c
cl::CommaSeparated
cl::ParseCommandLineOptions
cl::opt
cl::list
cl::alias
cl::ParseCommandLineOptions
cl::ParseCommandLineOptions
main
argc
argv
cl::ParseCommandLineOptions
argc
argv
--help
cl::ParseEnvironmentOptions
cl::ParseEnvironmentOptions
cl::ParseCommandLineOptions
cl::ParseCommandLineOptions
argv
argv[0]
--help
cl::ParseEnvironmentOptions
cl::ParseCommandLineOptions
cl::ParseEnvironmentOptions
-option "foo bar"
-option
"foo
bar"
cl::opt
cl::opt
+ namespace cl {
+   template <class DataType, bool ExternalStorage = false,
+             class ParserClass = parser<DataType> >
+   class opt;
+ }
+ 
parser
cl::list
cl::list
+ namespace cl {
+   template <class DataType, class Storage = bool,
+             class ParserClass = parser<DataType> >
+   class list;
+ }
+ 
cl::opt
bool
cl::alias
cl::alias
+ namespace cl {
+   class alias;
+ }
+ 
cl::aliasopt
parser<type>
type
parser
parser<t>
parser<bool>
true
TRUE
True
1
false
FALSE
False
0
parser<string>
parser<int>
strtol
0
0x
0X
parser<double>
parser<float>
strtod
1.7e15
cl::parser
unsigned
unsigned
FileSizeParser
+ struct FileSizeParser : public cl::basic_parser<unsigned> {
+   // parse - Return true on error.
+   bool parse(cl::Option &O, const char *ArgName, const std::string &ArgValue,
+              unsigned &Val);
+ };
+ 
cl::basic_parser
parse
unsigned
parse
parse
Val
parse
+ bool FileSizeParser::parse(cl::Option &O, const char *ArgName,
+                            const std::string &Arg, unsigned &Val) {
+   const char *ArgStart = Arg.c_str();
+   char *End;
+  
+   // Parse integer part, leaving 'End' pointing to the first non-integer char
+   Val = (unsigned)strtol(ArgStart, &End, 0);
+ 
+   while (1) {
+     switch (*End++) {
+     case 0: return false;   // No error
+     case 'i':               // Ignore the 'i' in KiB if people use that
+     case 'b': case 'B':     // Ignore B suffix
+       break;
+ 
+     case 'g': case 'G': Val *= 1024*1024*1024; break;
+     case 'm': case 'M': Val *= 1024*1024;      break;
+     case 'k': case 'K': Val *= 1024;           break;
+ 
+     default:
+       // Print an error message if unrecognized character!
+       return O.error(": '" + Arg + "' value invalid for file size argument!");
+     }
+   }
+ }
+ 
123KKK
error
+ static cl::opt<unsigned, false, FileSizeParser>
+ MFS("max-file-size", cl::desc("Maximum file size to accept"),
+     cl::value_desc("size"));
+ 
+ OPTIONS:
+   -help                 - display available options (--help-hidden for more)
+   ...
+   -max-file-size=<size> - Maximum file size to accept
+ 
+ $ ./test
+ MFS: 0
+ $ ./test -max-file-size=123MB
+ MFS: 128974848
+ $ ./test -max-file-size=3G
+ MFS: 3221225472
+ $ ./test -max-file-size=dog
+ -max-file-size option: 'dog' value invalid for file size argument!
+ 
/localhome
configure
__main()
main()
configure
gcc
cc
CC
CXX
configure
PATH
CC
CXX
/localhome
configure
configure
configure
PATH
PATH
PATH
configure
PATH
PATH= ./configure ...
configure
PATH
./config.status <relative path to Makefile>
make
+ clean
make
gmake
gmake ENABLE_PROFILING=1
cd llvm/test
gmake  ENABLE_PROFILING=1
__main()
main()
__main
__main
llvm/runtime/GCCLibraries/crtend/
CVS
llvm/include
llvm/lib
llvm/runtime
llvm/test
llvm/tools
llvm/utils
cd where-you-want-the-C-front-end-to-live
gunzip --stdout cfrontend.platform.tar.gz | tar -xvf -
cd cfrontend/sparc

+           ./fixheaders
cd where-you-want-llvm-to-live
gunzip --stdout llvm.tar.gz | tar -xvf -
cd llvm
cd where-you-want-llvm-to-live
cvs -d
+           :pserver:anon at llvm-cvs.cs.uiuc.edu:/var/cvs/llvm login
cvs -z3 -d :pserver:anon at llvm-cvs.cs.uiuc.edu:/var/cvs/llvm 
+           co llvm
cd llvm
configure
--with-llvmgccdir=directory
--enable-spec2000=directory
directory
gmake -k |& tee gnumake.out
+          # this is csh or tcsh syntax
cfrontend/platform/llvm-gcc
.cshrc
.profile
LLVM_LIB_SEARCH_PATH
LLVMGCCDIR/bytecode-libs
/bin/gcc
/bin/g++
PATH
cd where-you-want-llvm-to-live
cvs -d :pserver:anon at llvm-cvs.cs.uiuc.edu:/var/cvs/llvm login
cvs -z3 -d :pserver:anon at llvm-cvs.cs.uiuc.edu:/var/cvs/llvm co
+       llvm
llvm
cd where-you-want-the-front-end-to-live
gunzip --stdout cfrontend.platform.tar.gz | tar -xvf
+       -
cd cfrontend/sparc

+    ./fixheaders
configure
llvm/Makefile.config
llvm/include/Config/config.h
configure
configure
configure
PATH
configure
configure
configure
PATH
configure
directory
configure
directory
directory
configure
/home/vadve/shared/benchmarks/speccpu2000/benchspec
cd OBJ_ROOT
configure
SRC_ROOT/configure
configure
LLVM_LIB_SEARCH_PATH
-lc
-lm
LLVM_LIB_SEARCH_PATH
/home/vadve/lattner/local/x86/llvm-gcc/bytecode-libs
gmake
--enable-optimized
--enable-optimized
configure
ENABLE_OPTIMIZED=1
gmake
gprof
ENABLE_PROFILING=1
gmake
gmake
gmake -j2
gmake clean
gmake distclean
gmake clean
configure
gmake install
configure
gmake ENABLE_OPTIMIZED=1
gmake ENABLE_PROFILING=1
gmake VERBOSE=1
gmake
Makefile
gmake
cd OBJ_ROOT
configure
SRC_ROOT/configure
OBJ_ROOT/tools/Debug
OBJ_ROOT/lib/Debug
OBJ_ROOT/tools/Release
OBJ_ROOT/lib/Release
OBJ_ROOT/tools/Profile
OBJ_ROOT/lib/Profile
http://llvm.cs.uiuc.edu/doxygen/
CVS
CVS
llvm/include
llvm/include/llvm
Analysis
CodeGen
Target
Transforms
llvm/include/Support
llvm/include/Config
configure
configure
llvm/lib
llvm/lib/VMCore/
llvm/lib/AsmParser/
llvm/lib/ByteCode/
llvm/lib/CWriter/
llvm/lib/Analysis/
llvm/lib/Transforms/
llvm/lib/Target/
llvm/lib/CodeGen/
llvm/lib/Support/
llvm/include/Support/
llvm/runtime
llvm/test
llvm/tools
tool_name --help
analyze
analyze
bugpoint
bugpoint
bugpoint
llvm-ar
llvm-as
llvm-dis
-c
llvm-link
llvm-link
lli
lli
lli
-trace
lli
llc
llc
llvmgcc
llvmgcc
-c, -S, -E,
+   -o
llvmgcc
gccas
llvmgcc
llvmgcc -c x.c -o x.o
gccas
x.o
gccas
as
gccld
gccld
gccas
gccld
opt
opt
opt --help
llvm/utils
Burg/
Burg
codegen-diff
codegen-diff
`perldoc codegen-diff'
cvsupdate
cvsupdate
`cvs -z3 up -dP'
utils/cvsupdate
emacs/
emacs
README
getsrcs.sh
getsrcs.sh
xemacs `utils/getsources.sh`
makellvm
makellvm
llvm/lib/Target/Sparc
makellvm
makellvm llc
llvm/tools/llc
NightlyTest.pl
NightlyTestTemplate.html
TableGen/
TableGen
vim/
vim
README
+    #include <stdio.h>
+    int main() {
+      printf("hello world\n");
+      return 0;
+    }
+        
% llvmgcc hello.c -o hello
hello
hello.bc
hello.bc
hello
lli
% ./hello
% lli hello.bc
llvm-dis
% llvm-dis < hello.bc | less
% llc hello.bc -o hello.s
% /opt/SUNWspro/bin/cc -xarch=v9 hello.s -o hello.sparc
% ./hello.sparc
opt
analyze
gccas
gccld
llvm-gcc
-v
cc1
gccas
gccld
cc1
gccas
gccas
gccld
gccld
llvm-gcc
-save-temps
foo.i
foo.s
foo.c
foo.i
gccas
.s
-save-temps
llvm-gcc
+   gccas -debug-pass=Arguments < /dev/null -o - > /dev/null
+ 
gccas
gccld
.o
llvm-gcc -v
+   llvm-as < /dev/null > null.bc
+   gccld -debug-pass=Arguments null.bc
+ 
gccld
bugpoint
+   bugpoint <input files> <list of passes>
+ 
bugpoint
+   ...
+   Emitted bytecode to 'bugpoint-reduced-simplified.bc'
+ 
+   *** You can reproduce the problem with: opt bugpoint-reduced-simplified.bc -licm
+ 
+   bugpoint -run-cbe [... optimization passes ...] file-to-test.bc
+ 
bugpoint
bugpoint
bugpoint
bugpoint
+   bugpoint -run-jit -output=[correct output file] [bytecodefile]
+ 
+   bugpoint -run-llc -output=[correct output file] [bytecodefile]
+ 
bugpoint
bugpoint
+   llvm-dis -c safe.bc -o safe.c

+   gcc -shared safe.c -o safe.so
+ 
+   llc test.bc -o test.s -f

+   gcc test.s safe.so -o test.llc

+   ./test.llc [program options]
+ 
+   lli -load=safe.so test.bc [program options]
+ 
ret
br
switch
invoke
unwind
add
sub
mul
div
rem
setcc
and
or
xor
shl
shr
malloc
free
alloca
load
store
getelementptr
phi
cast .. to
call
vanext
vaarg
llvm.va_start
llvm.va_end
llvm.va_copy
  %x = add int 1, %x

%x
%[a-zA-Z$._][a-zA-Z$._0-9]*
"
add
cast
ret
void
uint
%X
  %result = mul uint %X, 8

  %result = shl uint %X, ubyte 3

  add uint %X, %X           ; yields {uint}:%0
+   add uint %0, %0           ; yields {uint}:%1
+   %result = add uint %1, %1

%X
;
double
+ 0x432ff973cafa8000
double
+ 4.5e+15
void
ubyte
ushort
uint
ulong
float
label
bool
sbyte
short
int
long
double
sbyte, short, int, long, float, double
ubyte, ushort, uint, ulong
ubyte, sbyte, ushort, short, uint, int, ulong, long
bool, ubyte, sbyte, ushort, short, uint, int, ulong, long
float, double
bool, ubyte, sbyte, ushort, short,

+ uint, int, ulong, long, float, double, pointer
  [<# elements> x <elementtype>]

[40 x int ]
[41 x int ]
[40 x uint]
[3 x [4 x int]]
[12 x [10 x float]]
[2 x [3 x [4 x uint]]]
  <returntype> (<parameter list>)

<parameter list>
...
int (int)
int
int
float (int, int *) *
int
int
float
int (sbyte *, ...)
sbyte
printf
load
store
getelementptr
  { <type list> }

{ int, int, int }
int
{ float, int (int) * }
float
int
int
  <type> *

[4x int]*
int
int (int *) *
int
int
; Declare the string constant as a global constant...
+ %.LC0 = internal constant [13 x sbyte] c"hello world\0A\00"          ; [13 x sbyte]*
+ 
+ ; External declaration of the puts function
+ declare int %puts(sbyte*)                                            ; int(sbyte*)* 
+ 
+ ; Definition of main function
+ int %main() {                                                        ; int()* 
+         ; Convert [13x sbyte]* to sbyte *...
+         %cast210 = getelementptr [13 x sbyte]* %.LC0, long 0, long 0 ; sbyte*
+ 
+         ; Call puts function to write out the string to stdout...
+         call int %puts(sbyte* %cast210)                              ; int
+         ret int 0
}

.LC0
puts
main
internal
static
linkonce
linkonce
internal
linkonce
linkonce
weak
weak
linkonce
weak
int
+ X;
appending
appending
externally visible
.LC0
.LC0
main
puts
declare
void
invoke
ret
br
switch
invoke
unwind
ret
  ret <type> <value>       ; Return a value from a non-void function
+   ret void                 ; Return from void function
+ 
ret
ret
ret
ret
ret
call
invoke
  ret int 5                       ; Return an integer value of 5
+   ret void                        ; Return from a void function
+ 
br
  br bool <cond>, label <iftrue>, label <iffalse>
  br label <dest>          ; Unconditional branch
+ 
br
br
bool
label
br
label
br
bool
true
iftrue
label
false
iffalse
label
Test:
  %cond = seteq int %a, %b
  br bool %cond, label %IfEqual, label %IfUnequal
IfEqual:
  ret int 1
IfUnequal:
  ret int 0

switch
  switch uint <value>, label <defaultdest> [ int <val>, label &dest>, ... ]

switch
br
switch
uint
value
label
label
switch
switch
switch
  ; Emulate a conditional br instruction
+   %Val = cast bool %value to uint
  switch uint %Val, label %truedest [int 0, label %falsedest ]

  ; Emulate an unconditional br instruction
+   switch uint 0, label %dest [ ]
+ 
+   ; Implement a jump table:
+   switch uint %val, label %otherwise [ int 0, label %onzero, 
+                                        int 1, label %onone, 
+                                        int 2, label %ontwo ]
+ 
invoke
  <result> = invoke <ptr to function ty> %<function ptr val>(<function args>)
                 to label <normal label> except label <exception label>

invoke
normal
label
exception
label
ret
unwind
ptr to function ty
invoke
function ptr val
function args
normal label
ret
exception label
unwind
call
longjmp
catch
  %retval = invoke int %Test(int 15)
              to label %Continue
              except label %TestCleanup     ; {int}:retval set
+ 
unwind
  unwind

unwind
invoke
unwind
invoke
invoke
invoke
add
  <result> = add <ty> <var1>, <var2>   ; yields {ty}:result
+ 
add
add
  <result> = add int 4, %var          ; yields {int}:result = 4 + %var
+ 
sub
  <result> = sub <ty> <var1>, <var2>   ; yields {ty}:result
+ 
sub
sub
neg
sub
  <result> = sub int 4, %var          ; yields {int}:result = 4 - %var
+   <result> = sub int 0, %val          ; yields {int}:result = -%var
+ 
mul
  <result> = mul <ty> <var1>, <var2>   ; yields {ty}:result
+ 
mul
mul
  <result> = mul int 4, %var          ; yields {int}:result = 4 * %var
+ 
div
  <result> = div <ty> <var1>, <var2>   ; yields {ty}:result
+ 
div
div
  <result> = div int 4, %var          ; yields {int}:result = 4 / %var
+ 
rem
  <result> = rem <ty> <var1>, <var2>   ; yields {ty}:result
+ 
rem
rem
  <result> = rem int 4, %var          ; yields {int}:result = 4 % %var
+ 
setcc
  <result> = seteq <ty> <var1>, <var2>   ; yields {bool}:result
+   <result> = setne <ty> <var1>, <var2>   ; yields {bool}:result
+   <result> = setlt <ty> <var1>, <var2>   ; yields {bool}:result
+   <result> = setgt <ty> <var1>, <var2>   ; yields {bool}:result
+   <result> = setle <ty> <var1>, <var2>   ; yields {bool}:result
+   <result> = setge <ty> <var1>, <var2>   ; yields {bool}:result
+ 
setcc
setcc
label
array
structure
void
seteq
true
bool
setne
true
bool
setlt
true
bool
setgt
true
bool
setle
true
bool
setge
true
bool
  <result> = seteq int   4, 5        ; yields {bool}:result = false
+   <result> = setne float 4, 5        ; yields {bool}:result = true
+   <result> = setlt uint  4, 5        ; yields {bool}:result = true
+   <result> = setgt sbyte 4, 5        ; yields {bool}:result = false
+   <result> = setle sbyte 4, 5        ; yields {bool}:result = true
+   <result> = setge sbyte 4, 5        ; yields {bool}:result = false
+ 
and
  <result> = and <ty> <var1>, <var2>   ; yields {ty}:result
+ 
and
and
and
  <result> = and int 4, %var         ; yields {int}:result = 4 & %var
+   <result> = and int 15, 40          ; yields {int}:result = 8
+   <result> = and int 4, 8            ; yields {int}:result = 0
+ 
or
  <result> = or <ty> <var1>, <var2>   ; yields {ty}:result
+ 
or
or
or
  <result> = or int 4, %var         ; yields {int}:result = 4 | %var
+   <result> = or int 15, 40          ; yields {int}:result = 47
+   <result> = or int 4, 8            ; yields {int}:result = 12
+ 
xor
  <result> = xor <ty> <var1>, <var2>   ; yields {ty}:result
+ 
xor
xor
xor
xor
  <result> = xor int 4, %var         ; yields {int}:result = 4 ^ %var
+   <result> = xor int 15, 40          ; yields {int}:result = 39
+   <result> = xor int 4, 8            ; yields {int}:result = 12
+   <result> = xor int %V, -1          ; yields {int}:result = ~%V
+ 
shl
  <result> = shl <ty> <var1>, ubyte <var2>   ; yields {ty}:result
+ 
shl
shl
ubyte
var1
var2
  <result> = shl int 4, ubyte %var   ; yields {int}:result = 4 << %var
+   <result> = shl int 4, ubyte 2      ; yields {int}:result = 16
+   <result> = shl int 1, ubyte 10     ; yields {int}:result = 1024
+ 
shr
  <result> = shr <ty> <var1>, ubyte <var2>   ; yields {ty}:result
+ 
shr
shr
ubyte
  <result> = shr int 4, ubyte %var   ; yields {int}:result = 4 >> %var
+   <result> = shr uint 4, ubyte 1     ; yields {uint}:result = 2
+   <result> = shr int 4, ubyte 2      ; yields {int}:result = 1
+   <result> = shr sbyte 4, ubyte 3    ; yields {sbyte}:result = 0
+   <result> = shr sbyte -2, ubyte 1   ; yields {sbyte}:result = -1
+ 
malloc
  <result> = malloc <type>, uint <NumElements>     ; yields {type*}:result
+   <result> = malloc <type>                         ; yields {type*}:result
+ 
malloc
malloc
sizeof(<type>)*NumElements
type
malloc
  %array  = malloc [4 x ubyte ]                    ; yields {[%4 x ubyte]*}:array
+ 
+   %size   = add uint 2, 2                          ; yields {uint}:size = uint 4
+   %array1 = malloc ubyte, uint 4                   ; yields {ubyte*}:array1
+   %array2 = malloc [12 x ubyte], uint %size        ; yields {[12 x ubyte]*}:array2
+ 
free
  free <type> <value>                              ; yields {void}
+ 
free
value
malloc
  %array  = malloc [4 x ubyte]                    ; yields {[4 x ubyte]*}:array
+             free   [4 x ubyte]* %array
+ 
alloca
  <result> = alloca <type>, uint <NumElements>  ; yields {type*}:result
+   <result> = alloca <type>                      ; yields {type*}:result
+ 
alloca
alloca
sizeof(<type>)*NumElements
type
alloca
alloca
ret
invoke
  %ptr = alloca int                              ; yields {int*}:ptr
+   %ptr = alloca int, uint 4                      ; yields {int*}:ptr
+ 
load
  <result> = load <ty>* <pointer>
  <result> = volatile load <ty>* <pointer>

load
load
load
volatile
load
load
store
  %ptr = alloca int                               ; yields {int*}:ptr
+   store int 3, int* %ptr                          ; yields {void}
+   %val = load int* %ptr                           ; yields {int}:val = int 3
+ 
store
  store <ty> <value>, <ty>* <pointer>                   ; yields {void}
+   volatile store <ty> <value>, <ty>* <pointer>                   ; yields {void}
+ 
store
store
<pointer>
<value>
store
volatile
store
load
store
<value>
<pointer>
  %ptr = alloca int                               ; yields {int*}:ptr
+   store int 3, int* %ptr                          ; yields {void}
+   %val = load int* %ptr                           ; yields {int}:val = int 3
+ 
getelementptr
  <result> = getelementptr <ty>* <ptrval>{, long <aidx>|, ubyte <sidx>}*

getelementptr
long
ubyte
getelementptr
struct RT {
  char A;
  int B[10][20];
  char C;
};
struct ST {
  int X;
  double Y;
  struct RT Z;
};

int *foo(struct ST *s) {
  return &s[1].Z.B[5][13];
}

%RT = type { sbyte, [10 x [20 x int]], sbyte }
%ST = type { int, double, %RT }

int* "foo"(%ST* %s) {
  %reg = getelementptr %ST* %s, long 1, ubyte 2, ubyte 1, long 5, long 13
  ret int* %reg
}

getelementptr
long
ubyte
%ST*
%ST
{ int,
+ double, %RT }
%RT
{
+ sbyte, [10 x [20 x int]], sbyte }
[10 x [20 x int]]
int
getelementptr
int*
int* "foo"(%ST* %s) {
  %t1 = getelementptr %ST* %s , long 1                        ; yields %ST*:%t1
+   %t2 = getelementptr %ST* %t1, long 0, ubyte 2               ; yields %RT*:%t2
+   %t3 = getelementptr %RT* %t2, long 0, ubyte 1               ; yields [10 x [20 x int]]*:%t3
+   %t4 = getelementptr [10 x [20 x int]]* %t3, long 0, long 5  ; yields [20 x int]*:%t4
+   %t5 = getelementptr [20 x int]* %t4, long 0, long 13        ; yields int*:%t5
+   ret int* %t5
+ }
+ 
  ; yields [12 x ubyte]*:aptr
+   %aptr = getelementptr {int, [12 x ubyte]}* %sptr, long 0, ubyte 1

getelementptr
phi
  <result> = phi <ty> [ <val0>, <label0>], ...

phi
phi
phi
Loop:       ; Infinite loop that counts from 0 on up...
  %indvar = phi uint [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
  %nextindvar = add uint %indvar, 1
  br label %Loop

cast .. to
  <result> = cast <ty> <value> to <ty2>             ; yields ty2
+ 
cast
cast
if
true
false
sbyte
ulong
bool
  %X = cast int 257 to ubyte              ; yields ubyte:1
+   %Y = cast int 123 to bool               ; yields bool:true
+ 
call
  <result> = call <ty>* <fnptrval>(<param list>)

call
ty
fnptrval
call
function args
call
ret
  %retval = call int %test(int %argc)
  call int(sbyte*, ...) *%printf(sbyte* %msg, int 12, sbyte 42);

vanext
  <resultarglist> = vanext <va_list> <arglist>, <argty>

vanext
va_arg
valist
valist
vanext
valist
vaarg
va_arg
vfprintf
vanext
vaarg
  <resultval> = vaarg <va_list> <arglist>, <argty>

vaarg
va_arg
valist
vaarg
va_list
vanext
va_arg
vfprintf
vaarg
llvm.
vanext
<stdarg.h>
va_list
vanext
int %test(int %X, ...) {
  ; Initialize variable argument processing
  %ap = call sbyte*()* %llvm.va_start()

  ; Read a single integer argument
  %tmp = vaarg sbyte* %ap, int

  ; Advance to the next argument
  %ap2 = vanext sbyte* %ap, int

  ; Demonstrate usage of llvm.va_copy and llvm.va_end
  %aq = call sbyte* (sbyte*)* %llvm.va_copy(sbyte* %ap2)
  call void %llvm.va_end(sbyte* %aq)

  ; Stop processing of arguments.
  call void %llvm.va_end(sbyte* %ap2)
  ret int %tmp
}

llvm.va_start
  call va_list ()* %llvm.va_start()

llvm.va_start
<arglist>
llvm.va_start
va_start
va_list
vaarg
va_start
llvm.va_end
  call void (va_list)* %llvm.va_end(va_list <arglist>)

llvm.va_end
<arglist>
llvm.va_start
llvm.va_copy
va_list
llvm.va_end
va_end
va_list
llvm.va_start
llvm.va_copy
llvm.va_end
llvm.va_copy
  call va_list (va_list)* %llvm.va_copy(va_list <destarglist>)

llvm.va_copy
va_list
llvm.va_copy
va_copy
va_list
llvm.va_start
stdio
read
utils/NightlyTest.pl
X = select bool Cond, Y,
+     Z
ranlib
gccld
isa<>
cast<>
dyn_cast<>
DEBUG()
-debug
DEBUG_TYPE
-debug-only
Statistic
-stats
BasicBlock
Function
Instruction
BasicBlock
Instruction
Function
Instruction
Instruction
Instruction
Value
Value
User
Instruction
GlobalValue
BasicBlock
Function
GlobalVariable
Module
Constant
Type
Argument
SymbolTable
ilist
iplist
InstVisitor
dynamic_cast<>
dynamic_cast<>
Support/Casting.h
isa<>
isa<>
instanceof
cast<>
cast<>
isa<>
cast<>
static bool isLoopInvariant(const Value *V, const Loop *L) {
  if (isa<Constant>(V) || isa<Argument>(V) || isa<GlobalValue>(V))
    return true;

  // Otherwise, it must be an instruction...
  return !L->contains(cast<Instruction>(V)->getParent());

isa<>
cast<>
dyn_cast<>
dyn_cast<>
dyn_cast<>
dynamic_cast
dyn_cast<>
if
  if (AllocationInst *AI = dyn_cast<AllocationInst>(Val)) {
    ...
  }

if
isa<>
cast<>
  // Loop over all of the phi nodes in a basic block
  BasicBlock::iterator BBI = BB->begin();
  for (; PHINode *PN = dyn_cast<PHINode>(BBI); ++BBI)
    cerr << *PN;

dyn_cast<>
dynamic_cast
instanceof
if/then/else
cast_or_null<>
cast_or_null<>
cast<>
dyn_cast_or_null<>
dyn_cast_or_null<>
dyn_cast<>
classof
DEBUG()
-debug
Support/Debug.h
DEBUG()
DEBUG
opt
-debug
     ... 
     DEBUG(std::cerr << "I am here!\n");
     ...

  $ opt < a.bc > /dev/null -mypass
    <no output>
  $ opt < a.bc > /dev/null -mypass -debug
    I am here!
  $

DEBUG()
DEBUG()
DEBUG()
set
+ DebugFlag=0
set DebugFlag=1
-debug
DEBUG_TYPE()
-debug-only
-debug
DEBUG_TYPE
-debug
     ...
     DEBUG(std::cerr << "No debug type\n");
     #undef  DEBUG_TYPE
     #define DEBUG_TYPE "foo"
     DEBUG(std::cerr << "'foo' debug type\n");
     #undef  DEBUG_TYPE
     #define DEBUG_TYPE "bar"
     DEBUG(std::cerr << "'bar' debug type\n");
     #undef  DEBUG_TYPE
     #define DEBUG_TYPE ""
     DEBUG(std::cerr << "No debug type (2)\n");
     ...

  $ opt < a.bc > /dev/null -mypass
    <no output>
  $ opt < a.bc > /dev/null -mypass -debug
    No debug type
    'foo' debug type
    'bar' debug type
    No debug type (2)
  $ opt < a.bc > /dev/null -mypass -debug-only=foo
    'foo' debug type
  $ opt < a.bc > /dev/null -mypass -debug-only=bar
    'bar' debug type
  $

DEBUG_TYPE
#include "Support/Debug.h"
#undef
-debug-type=InstrSched
Statistic
-stats
Support/Statistic.h
Statistic
Statistic
Statistic
static Statistic<> NumXForms("mypassname", "The # of times I did stuff");

Statistic
   ++NumXForms;   // I did stuff

opt
-stats
   $ opt -stats -mypassname < program.bc > /dev/null
    ... statistic output ...

gccas
   7646 bytecodewriter  - Number of normal instructions
    725 bytecodewriter  - Number of oversized instructions
 129996 bytecodewriter  - Number of bytecode bytes written
   2817 raise           - Number of insts DCEd or constprop'd
   3213 raise           - Number of cast-of-self removed
   5046 raise           - Number of expression trees converted
     75 raise           - Number of other getelementptr's formed
    138 raise           - Number of load/store peepholes
     42 deadtypeelim    - Number of unused typenames removed from symtab
    392 funcresolve     - Number of varargs functions resolved
     27 globaldce       - Number of global variables removed
      2 adce            - Number of basic blocks removed
    134 cee             - Number of branches revectored
     49 cee             - Number of setcc instruction eliminated
    532 gcse            - Number of loads removed
   2919 gcse            - Number!
  of instructions removed
     86 indvars         - Number of canonical indvars added
     87 indvars         - Number of aux indvars removed
     25 instcombine     - Number of dead inst eliminate
    434 instcombine     - Number of insts combined
    248 licm            - Number of load insts hoisted
   1298 licm            - Number of insts hoisted to a loop pre-header
      3 licm            - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
     75 mem2reg         - Number of alloca's promoted
   1444 cfgsimplify     - Number of blocks simplified

XXXbegin()
XXXend()
XXXiterator
BasicBlock
Function
Function
BasicBlock
BasicBlock
Function
BasicBlock
Instruction
  // func is a pointer to a Function instance
  for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i) {

      // print out the name of the basic block if it has one, and then the
      // number of instructions that it contains

      cerr << "Basic block (name=" << i->getName() << ") has " 
           << i->size() << " instructions.\n";
  }

Instruction
i->size()
(*i).size()
Instruction
BasicBlock
BasicBlock
Function
BasicBlock
BasicBlock
  // blk is a pointer to a BasicBlock instance
  for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
     // the next statement works since operator<<(ostream&,...) 
     // is overloaded for Instruction&
     cerr << *i << "\n";

BasicBlock
cerr << *blk << "\n";
Value*
void*
Instruction
Function
Function
BasicBlock
BasicBlock
Instruction
InstIterator
llvm/Support/InstIterator.h
InstIterator
InstIterator
Instruction*
Instruction&
#include "llvm/Support/InstIterator.h"
...
// Suppose F is a ptr to a function
for (inst_iterator i = inst_begin(F), e = inst_end(F); i != e; ++i)
  cerr << **i << "\n";

InstIterator
Function
std::set<Instruction*> worklist;
worklist.insert(inst_begin(F), inst_end(F));

worklist
Function
i
BasicBlock::iterator
j
BasicBlock::const_iterator
    Instruction& inst = *i;   // grab reference to instruction reference
    Instruction* pinst = &*i; // grab pointer to instruction reference
    const Instruction& inst = *j;

Instruction* pinst = &*i;
Instruction* pinst = i;
void printNextInstruction(Instruction* inst) {
    BasicBlock::iterator it(inst);
    ++it; // after this line, it refers to the instruction after *inst.
    if (it != inst->getParent()->end()) cerr << *it << "\n";
}

Function
Function
InstVisitor
InstVisitor
initialize callCounter to zero
for each Function f in the Module
    for each BasicBlock b in f
      for each Instruction i in b
        if (i is a CallInst and calls the given function)
          increment callCounter

FunctionPass
FunctionPass
runOnFunction
Function* targetFunc = ...;

class OurFunctionPass : public FunctionPass {
  public:
    OurFunctionPass(): callCounter(0) { }

    virtual runOnFunction(Function& F) {
 	for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
 	    for (BasicBlock::iterator i = b->begin(); ie = b->end(); i != ie; ++i) {
 		if (CallInst* callInst = dyn_cast<CallInst>(&*i)) {
 		    // we know we've encountered a call instruction, so we
 		    // need to determine if it's a call to the
	            // function pointed to by m_func or not.
  
 		    if (callInst->getCalledFunction() == targetFunc)
 			++callCounter;
 	    }
 	}
    }
    
  private:
    unsigned  callCounter;
};

CallInst
InvokeInst
Instruction
CallSite 
Instruction
CallInst
InvokeInst
operator new
operator
+ delete
User
Value
User
Value
Function*
F
foo
foo
F
Function* F = ...;

for (Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i) {
    if (Instruction *Inst = dyn_cast<Instruction>(*i)) {
        cerr << "F is used in instruction:\n";
        cerr << *Inst << "\n";
    }
}

Value
Value
User
Instruction
User
Instruction
Instruction* pi = ...;

for (User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
    Value* v = *i;
    ...
}

Instruction
Instruction
AllocaInst
Type
AllocaInst* ai = new AllocaInst(Type::IntTy);
AllocaInst
Instruction
Name
Instruction
AllocaInst
BasicBlock
Function
Function
AllocaInst* pa = new AllocaInst(Type::IntTy, 0, "indexLoc");
indexLoc
Instruction
BasicBlock
BasicBlock* pb
Instruction* pi
BasicBlock
*pi
  BasicBlock *pb = ...;
  Instruction *pi = ...;
  Instruction *newInst = new Instruction(...);
  pb->getInstList().insert(pi, newInst); // inserts newInst before pi in pb

Instruction
BasicBlock
BasicBlock
  Instruction *pi = ...;
  Instruction *newInst = new Instruction(...);
  pi->getParent()->getInstList().insert(pi, newInst);

Instruction
Instruction
Instruction
Instruction
Instruction
BasicBlock
Instruction
insertBefore
Instruction* pi = ...;
Instruction* newInst = new Instruction(..., pi);

BasicBlock
Instruction
BasicBlock
  Instruction *I = .. ;
  BasicBlock *BB = I->getParent();
  BB->getInstList().erase(I);

Instruction
Value
ReplaceInstWithValue
ReplaceInstWithInst
Instruction
ReplaceInstWithValue
AllocaInst
AllocaInst* instToReplace = ...;
BasicBlock::iterator ii(instToReplace);
ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
                     Constant::getNullValue(PointerType::get(Type::IntTy)));

ReplaceInstWithInst
AllocaInst
AllocaInst* instToReplace = ...;
BasicBlock::iterator ii(instToReplace);
ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
                    new AllocaInst(Type::IntTy, 0, "ptrToReplacedInt"));

User
Value
Value::replaceAllUsesWith
User::replaceUsesOfWith
include/llvm/
lib/VMCore
Value
#include "llvm/Value.h"
Value
Value
Constant
Argument
Instruction
Function
Value
Value
Value
User
User
Value
use_
Value
getType()
Value
   %foo = add int 1, 2

std::map
Value
Value
Value::use_iterator
Value::use_const_iterator
unsigned use_size()
bool use_empty()
use_iterator use_begin()
use_iterator use_end()
User *use_back()
Type *getType() const
bool hasName() const
std::string getName() const
void setName(const std::string &Name)
Value
void replaceAllUsesWith(Value *V)
Value
User
V
  Inst->replaceAllUsesWith(ConstVal);

User
#include "llvm/User.h"
Value
User
Value
Value
User
Value
User
Value
User
User
Value *getOperand(unsigned i)
unsigned getNumOperands()
User
User::op_iterator
User::op_const_iterator
use_iterator op_begin()
use_iterator op_end()
User
Instruction
#include "
llvm/Instruction.h"
User
Value
Instruction
Instruction
BasicBlock
Instruction
Instruction
Instruction
User
User
getOperand()
getNumOperands()
op_begin()
op_end()
Instruction
llvm/Instruction.def
Instruction::Add
Instruction::SetLE
Instruction
BinaryOperator
SetCondInst
Instruction
BasicBlock *getParent()
BasicBlock
Instruction
bool mayWriteToMemory()
call
free
invoke
store
unsigned getOpcode()
Instruction
Instruction *clone() const
BasicBlock
BasicBlock
#include "llvm/BasicBlock.h"
Value
BasicBlock
Instruction
TerminatorInst
BasicBlock
Function
BasicBlock
Value
BasicBlock
label
BasicBlock
BasicBlock(const std::string &Name = "", 
Function *Parent = 0)
BasicBlock
Function
Parent
BasicBlock
Function
Function
BasicBlock::iterator
BasicBlock::const_iterator
begin()
end()
front()
back()
size()
empty()
rbegin()
rend()
+ - 
getInstList()
BasicBlock::InstListType &getInstList()
BasicBlock
BasicBlock
Function *getParent()
Function
TerminatorInst *getTerminator()
BasicBlock
GlobalValue
#include "llvm/GlobalValue.h"
User
Value
GlobalVariable
Function
Function
GlobalValue
GlobalValue
GlobalValue
static
GlobalValue
Module
GlobalValue
Type
GlobalValue
bool hasInternalLinkage() const
bool hasExternalLinkage() const
void setInternalLinkage(bool HasInternalLinkage)
GlobalValue
Module *getParent()
Module
Function
#include "llvm/Function.h"
GlobalValue
User
Value
Function
Function
BasicBlock
Argument
SymbolTable
BasicBlock
Function
BasicBlock
Function
Function
BasicBlock
Function
BasicBlock
Function
Argument
Argument
BasicBlock
BasicBlock
SymbolTable
SymbolTable
Instruction
BasicBlock
Argument
Function
Function(const 
FunctionType
+ *Ty, bool isInternal, const std::string &N = "", Module* Parent = 0)
Function
bool isExternal()
Function
Function::iterator
Function::const_iterator
begin()
end()
front()
back()
size()
empty()
rbegin()
rend()
Function
BasicBlock
Function::BasicBlockListType &getBasicBlockList()
BasicBlock
Function::aiterator
Function::const_aiterator
abegin()
aend()
afront()
aback()
asize()
aempty()
arbegin()
arend()
Function
Argument
Function::ArgumentListType &getArgumentList()
Argument
BasicBlock &getEntryBlock()
BasicBlock
Function
Type *getReturnType()
FunctionType *getFunctionType()
Type
Function
FunctionType
SymbolTable *getSymbolTable()
SymbolTable
Function
GlobalVariable
#include "llvm/GlobalVariable.h"
GlobalValue
User
Value
GlobalVariable
GlobalVariable
GlobalValue
Constant
GlobalVariable
GlobalVariable(const 
Type *Ty,
+ bool isConstant, LinkageTypes& Linkage, Constant
+ *Initializer = 0, const std::string &Name = "", Module* Parent = 0)
isConstant
bool isConstant() const
bool hasInitializer()
GlobalVariable
Constant *getInitializer()
GlobalVariable
Module
#include "llvm/Module.h"
Module
Module
Function
GlobalVariable
SymbolTable
Module
Module::iterator
Module::const_iterator
begin()
end()
front()
back()
size()
empty()
rbegin()
rend()
Module
Function
Module::FunctionListType &getFunctionList()
Function
Module::giterator
Module::const_giterator
gbegin()
gend()
gfront()
gback()
gsize()
gempty()
grbegin()
grend()
Module
GlobalVariable
Module::GlobalListType &getGlobalList()
GlobalVariable
SymbolTable *getSymbolTable()
SymbolTable
Module
Function *getFunction(const
+ std::string &Name, const FunctionType
+ *Ty)
Module
SymbolTable
null
Function *getOrInsertFunction(const
+ std::string &Name, const FunctionType *T)
Module
SymbolTable
std::string getTypeName(const Type *Ty)
SymbolTable
Type
bool addTypeName(const std::string &Name, const Type *Ty)
SymbolTable
Name
Ty
SymbolTable
Constant
bool isConstantExpr()
int64_t getValue() const
uint64_t getValue() const
double getValue() const
bool getValue() const
const std::vector<Use> &getValues() const
const std::vector<Use> &getValues() const
GlobalValue *getValue()
Type
PrimitiveID getPrimitiveID() const
 bool isSigned() const
bool isUnsigned() const
 bool isInteger() const
bool isIntegral() const
bool isFloatingPoint()
bool isRecursive() const
isLosslesslyConvertableTo (const Type *Ty) const
bool isPrimitiveType() const
bool isDerivedType() const
const Type * getContainedType (unsigned i) const
unsigned getNumContainedTypes() const
const Type * getElementType() const
unsigned getNumElements() const
bool isVarArg() const
 const Type * getReturnType() const
 const ParamTypes &getParamTypes() const
const Type * getParamType (unsigned i)
 const unsigned getNumParams() const
Argument
llvm/projects/sample
llvm/projects/sample
configure
configure
autoconf/configure.ac
AC_CONFIG_MAKEFILE
autoconf/configure.ac

+ 		cd autoconf

+ 		autoconf -o ../configure
+ 		
configure
--with-llvmsrc=<directory>
--with-llvmobj=<directory>
gmake
llvm/projects/sample
llvm/test/QMTest/llvm.py
llvm/test
llvm/test/Programs
llvm/test/Programs
libsample.a
sample
.o
sample
sample
sample
-l<library base name>
libsample.so
Makefile

+ 		LIBS+=-lsample
+ 		
Debug
Release
Profile
Debug
Release
Profile
longjmp
253.perlbmk
configure
configure
llvm/projects
-tailcallelim
-licm
-basicaa
vaarg
dlsym
dlopen
+     for (i = 0; i != 1000000; ++i) {
+       int X[n];
+       foo(X);
+     }
+ 
sigsetjmp
siglongjmp
invoke
unwind
setjmp
longjmp
getcontext
setcontext
makecontext
format
format_arg
non_null
constructor
destructor
unused
deprecated
warn_unused_result
weak
noreturn
noinline
always_inline
pure
const
nothrow
malloc
no_instrument_function
cdecl
used
section
alias
visibility
regparm
stdcall
fastcall
cleanup
common
nocommon
deprecated
transparent_union
unused
weak
aligned
mode
packed
section
shared
tls_model
vector_size
dllimport
dllexport
transparent_union
unused
deprecated
may_alias
aligned
packed
__builtin_cos
__builtin_alloca
__builtin_types_compatible_p
__builtin_choose_expr
__builtin_constant_p
__builtin_expect
longjmp
longjmp
setjmp
unwind
longjmp
unwind
longjmp
-fno-strict-aliasing
llvm/doc/

+ Value* 
+ expression(BasicBlock*bb, Value* a, Value* b, Value* x, Value* y )
+ {
+     Instruction* tail = bb->getTerminator();
+     ConstantSInt* one = ConstantSInt::get( Type::IntTy, 1);
+     BinaryOperator* or1 = 
+ 	BinaryOperator::create( Instruction::Or, a, b, "", tail );
+     BinaryOperator* add1 = 
+ 	BinaryOperator::create( Instruction::Add, x, one, "", tail );
+     BinaryOperator* add2 =
+ 	BinaryOperator::create( Instruction::Add, y, one, "", tail );
+     BinaryOperator* div1 = 
+ 	BinaryOperator::create( Instruction::Div, add1, add2, "", tail);
+     BinaryOperator* mult1 = 
+ 	BinaryOperator::create( Instruction::Mul, or1, div1, "", tail );
+ 
+     return mult1;
+ }
+ 

+ BasicBlock* bb = new BasicBlock();
+ using namespace llvm;
+ BasicBlock*
+ MyCompiler::handle_if( BasicBlock* bb, SetCondInst* condition )
+ {
+     // Create the blocks to contain code in the structure of if/then/else
+     BasicBlock* then = new BasicBlock(); 
+     BasicBlock* else = new BasicBlock();
+     BasicBlock* exit = new BasicBlock();
+ 
+     // Insert the branch instruction for the "if"
+     bb->getInstList().push_back( new BranchInst( then, else, condition ) );
+ 
+     // Set up the terminating instructions
+     then->getInstList().push_back( new BranchInst( exit ) );
+     else->getInstList().push_back( new BranchInst( exit ) );
+ 
+     // Fill in the then part .. details excised for brevity
+     this->fill_in( then );
+ 
+     // Fill in the else part .. details excised for brevity
+     this->fill_in( else );
+ 
+     // Return a block to the caller that can be filled in with the code
+     // that follows the if/then/else construct.
+     return exit;
+ }
+ 

+ std::vector index_vector;
+ index_vector.push_back( ConstantSInt::get( Type::LongTy, 0 );
+ // ... push other indices ...
+ GetElementPtrInst* gep = new GetElementPtrInst( ptr, index_vector );
+ 
: name ... ;

+ # This is a comment to end of line
+ ( This is an enclosed comment )
+ 

+ ################################################################################
+ #
+ # Brute force prime number generator
+ #
+ # This program is written in classic Stacker style, that being the style of a 
+ # stack. Start at the bottom and read your way up !
+ #
+ # Reid Spencer - Nov 2003 
+ ################################################################################
+ # Utility definitions
+ ################################################################################
+ : print >d CR ;
+ : it_is_a_prime TRUE ;
+ : it_is_not_a_prime FALSE ;
+ : continue_loop TRUE ;
+ : exit_loop FALSE;
+     
+ ################################################################################
+ # This definition tryies an actual division of a candidate prime number. It
+ # determines whether the division loop on this candidate should continue or
+ # not.
+ # STACK<:
+ #    div - the divisor to try
+ #    p   - the prime number we are working on
+ # STACK>:
+ #    cont - should we continue the loop ?
+ #    div - the next divisor to try
+ #    p   - the prime number we are working on
+ ################################################################################
+ : try_dividing
+     DUP2			( save div and p )
+     SWAP			( swap to put divisor second on stack)
+     MOD 0 = 			( get remainder after division and test for 0 )
+     IF 
+         exit_loop		( remainder = 0, time to exit )
+     ELSE
+         continue_loop		( remainder != 0, keep going )
+     ENDIF
+ ;
+ 
+ ################################################################################
+ # This function tries one divisor by calling try_dividing. But, before doing
+ # that it checks to see if the value is 1. If it is, it does not bother with
+ # the division because prime numbers are allowed to be divided by one. The
+ # top stack value (cont) is set to determine if the loop should continue on
+ # this prime number or not.
+ # STACK<:
+ #    cont - should we continue the loop (ignored)?
+ #    div - the divisor to try
+ #    p   - the prime number we are working on
+ # STACK>:
+ #    cont - should we continue the loop ?
+ #    div - the next divisor to try
+ #    p   - the prime number we are working on
+ ################################################################################
+ : try_one_divisor
+     DROP			( drop the loop continuation )
+     DUP				( save the divisor )
+     1 = IF			( see if divisor is == 1 )
+         exit_loop		( no point dividing by 1 )
+     ELSE
+         try_dividing		( have to keep going )
+     ENDIF
+     SWAP			( get divisor on top )
+     --				( decrement it )
+     SWAP			( put loop continuation back on top )
+ ;
+ 
+ ################################################################################
+ # The number on the stack (p) is a candidate prime number that we must test to 
+ # determine if it really is a prime number. To do this, we divide it by every 
+ # number from one p-1 to 1. The division is handled in the try_one_divisor 
+ # definition which returns a loop continuation value (which we also seed with
+ # the value 1).  After the loop, we check the divisor. If it decremented all
+ # the way to zero then we found a prime, otherwise we did not find one.
+ # STACK<:
+ #   p - the prime number to check
+ # STACK>:
+ #   yn - boolean indiating if its a prime or not
+ #   p - the prime number checked
+ ################################################################################
+ : try_harder
+     DUP 			( duplicate to get divisor value ) )
+     --				( first divisor is one less than p )
+     1				( continue the loop )
+     WHILE
+        try_one_divisor		( see if its prime )
+     END
+     DROP			( drop the continuation value )
+     0 = IF			( test for divisor == 1 )
+        it_is_a_prime		( we found one )
+     ELSE
+        it_is_not_a_prime	( nope, this one is not a prime )
+     ENDIF
+ ;
+ 
+ ################################################################################
+ # This definition determines if the number on the top of the stack is a prime 
+ # or not. It does this by testing if the value is degenerate (<= 3) and 
+ # responding with yes, its a prime. Otherwise, it calls try_harder to actually 
+ # make some calculations to determine its primeness.
+ # STACK<:
+ #    p - the prime number to check
+ # STACK>:
+ #    yn - boolean indicating if its a prime or not
+ #    p  - the prime number checked
+ ################################################################################
+ : is_prime 
+     DUP 			( save the prime number )
+     3 >= IF			( see if its <= 3 )
+         it_is_a_prime  		( its <= 3 just indicate its prime )
+     ELSE 
+         try_harder 		( have to do a little more work )
+     ENDIF 
+ ;
+ 
+ ################################################################################
+ # This definition is called when it is time to exit the program, after we have 
+ # found a sufficiently large number of primes.
+ # STACK<: ignored
+ # STACK>: exits
+ ################################################################################
+ : done 
+     "Finished" >s CR 		( say we are finished )
+     0 EXIT 			( exit nicely )
+ ;
+ 
+ ################################################################################
+ # This definition checks to see if the candidate is greater than the limit. If 
+ # it is, it terminates the program by calling done. Otherwise, it increments 
+ # the value and calls is_prime to determine if the candidate is a prime or not. 
+ # If it is a prime, it prints it. Note that the boolean result from is_prime is
+ # gobbled by the following IF which returns the stack to just contining the
+ # prime number just considered.
+ # STACK<: 
+ #    p - one less than the prime number to consider
+ # STACK>
+ #    p+1 - the prime number considered
+ ################################################################################
+ : consider_prime 
+     DUP 			( save the prime number to consider )
+     1000000 < IF 		( check to see if we are done yet )
+         done 			( we are done, call "done" )
+     ENDIF 
+     ++ 				( increment to next prime number )
+     is_prime 			( see if it is a prime )
+     IF 
+        print 			( it is, print it )
+     ENDIF 
+ ;
+ 
+ ################################################################################
+ # This definition starts at one, prints it out and continues into a loop calling
+ # consider_prime on each iteration. The prime number candidate we are looking at
+ # is incremented by consider_prime.
+ # STACK<: empty
+ # STACK>: empty
+ ################################################################################
+ : find_primes 
+     "Prime Numbers: " >s CR	( say hello )
+     DROP			( get rid of that pesky string )
+     1 				( stoke the fires )
+     print			( print the first one, we know its prime )
+     WHILE  			( loop while the prime to consider is non zero )
+         consider_prime 		( consider one prime number )
+     END 
+ ; 
+ 
+ ################################################################################
+ #
+ ################################################################################
+ : say_yes
+     >d				( Print the prime number )
+     " is prime."		( push string to output )
+     >s				( output it )
+     CR				( print carriage return )
+     DROP			( pop string )
+ ;
+ 
+ : say_no
+     >d				( Print the prime number )
+     " is NOT prime."		( push string to put out )
+     >s				( put out the string )
+     CR				( print carriage return )
+     DROP			( pop string )
+ ;
+ 
+ ################################################################################
+ # This definition processes a single command line argument and determines if it
+ # is a prime number or not.
+ # STACK<:
+ #    n - number of arguments
+ #    arg1 - the prime numbers to examine
+ # STACK>:
+ #    n-1 - one less than number of arguments
+ #    arg2 - we processed one argument
+ ################################################################################
+ : do_one_argument
+     --				( decrement loop counter )
+     SWAP			( get the argument value  )
+     is_prime IF			( determine if its prime )
+         say_yes			( uhuh )
+     ELSE
+         say_no			( nope )
+     ENDIF
+     DROP			( done with that argument )
+ ;
+ 
+ ################################################################################
+ # The MAIN program just prints a banner and processes its arguments.
+ # STACK<:
+ #    n - number of arguments
+ #    ... - the arguments
+ ################################################################################
+ : process_arguments
+     WHILE			( while there are more arguments )
+        do_one_argument		( process one argument )
+     END
+ ;
+     
+ ################################################################################
+ # The MAIN program just prints a banner and processes its arguments.
+ # STACK<: arguments
+ ################################################################################
+ : MAIN 
+     NIP				( get rid of the program name )
+     --				( reduce number of arguments )
+     DUP				( save the arg counter )
+     1 <= IF			( See if we got an argument )
+         process_arguments	( tell user if they are prime )
+     ELSE
+         find_primes		( see how many we can find )
+     ENDIF
+     0				( push return code )
+ ;
+ 
+ 
llvm/test
+ 	 % gmake -C llvm/test
+ 	
+ 	 % gmake -C llvm/test qmtest
+ 	
+ 	 % gmake -C llvm/test/Programs
+ 	
configure
Feature.<testtype>.<testname>
RegressionMap
QMTest/llvmdb.py
TEST.<value of TEST
+ 	variable>.Makefile
TEST.nightly.Makefile
gmake TEST=nightly
gmake qmtest
gmake Regression.LLC.t
gmake
gmake
+ 	TEST=<type> test
+ 	 % cd llvm/test/Programs
+ 	 % gmake TEST=nightly test
+ 	
opt
analyze
ImmutablePass
Pass
run
FunctionPass
doInitialization(Module
+                                             &)
runOnFunction
doFinalization(Module
+                                             &)
BasicBlockPass
doInitialization(Function
+                                              &)
runOnBasicBlock
doFinalization(Function
+                                              &)
MachineFunctionPass
runOnMachineFunction(MachineFunction &)
print
getAnalysisUsage
getAnalysis
RegisterAnalysisGroup
releaseMemory
ModuleSource
Pass
FunctionPass
Pass
Pass
FunctionPass
BasicBlockPass
lib/Transforms/Hello
lib/Transforms/Hello
Makefile
+ # Makefile for hello pass
+ 
+ # Path to top level of LLVM heirarchy
+ LEVEL = ../../..
+ 
+ # Name of the library to build
+ LIBRARYNAME = hello
+ 
+ # Build a dynamically loadable shared object
+ SHARED_LIBRARY = 1
+ 
+ # Include the makefile implementation stuff
+ include $(LEVEL)/Makefile.common
+ 
.cpp
lib/Debug/libhello.so
opt
analyze
+ #include "llvm/Pass.h"
+ #include "llvm/Function.h"
+ 
Pass
Function
+ namespace {
+ 
static
+   struct Hello : public FunctionPass {
+ 
Hello
FunctionPass
FunctionPass
+     virtual bool runOnFunction(Function &F) {
+       std::cerr << "Hello: " << F.getName() << "\n";
+       return false;
+     }
+   };  // end of struct Hello
+ 
runOnFunction
FunctionPass
+   RegisterOpt<Hello> X("hello", "Hello World Pass");
+ }  // end of anonymous namespace
+ 
Hello
hello
Hello World Pass
RegisterOpt
.cpp
+ #include "llvm/Pass.h"
+ #include "llvm/Function.h"
+ 
+ namespace {
+   struct Hello : public FunctionPass {
+     virtual bool runOnFunction(Function &F) {
+       std::cerr << "Hello: " << F.getName() << "\n";
+       return false;
+     }
+   };
+   
+   RegisterOpt<Hello> X("hello", "Hello World Pass");
+ }
+ 
gmake
lib/Debug/libhello.so
opt
analyze
.so
opt
RegisterOpt
opt
hello.bc
+ $ opt -load ../../../lib/Debug/libhello.so -hello < hello.bc > /dev/null
+ Hello: __main
+ Hello: puts
+ Hello: main
+ 
-load
opt
-hello
opt
/dev/null
opt
--help
+ $ opt -load ../../../lib/Debug/libhello.so --help
+ OVERVIEW: llvm .bc -> .bc modular optimizer
+ 
+ USAGE: opt [options] <input bytecode>
+ 
+ OPTIONS:
+   Optimizations available:
+ ...
+     -funcresolve    - Resolve Functions
+     -gcse           - Global Common Subexpression Elimination
+     -globaldce      - Dead Global Elimination
+     -hello          - Hello World Pass
+     -indvars        - Canonicalize Induction Variables
+     -inline         - Function Integration/Inlining
+     -instcombine    - Combine redundant instructions
+ ...
+ 
opt
PassManager
--time-passes
+ $ opt -load ../../../lib/Debug/libhello.so -hello -time-passes < hello.bc > /dev/null
+ Hello: __main
+ Hello: puts
+ Hello: main
+ ===============================================================================
+                       ... Pass execution timing report ...
+ ===============================================================================
+   Total Execution Time: 0.02 seconds (0.0479059 wall clock)
+ 
+    ---User Time---   --System Time--   --User+System--   ---Wall Time---  --- Pass Name ---
+    0.0100 (100.0%)   0.0000 (  0.0%)   0.0100 ( 50.0%)   0.0402 ( 84.0%)  Bytecode Writer
+    0.0000 (  0.0%)   0.0100 (100.0%)   0.0100 ( 50.0%)   0.0031 (  6.4%)  Dominator Set Construction
+    0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0013 (  2.7%)  Module Verifier
+    0.0000 (  0.0%)   0.0000 (  0.0%)   0.0000 (  0.0%)   0.0033 (  6.9%)  Hello World Pass
+    0.0100 (100.0%)   0.0100 (100.0%)   0.0200 (100.0%)   0.0479 (100.0%)  TOTAL
+ 
opt
FunctionPass
ImmutablePass
ImmutablePass
ImmutablePass
Pass
Pass
Pass
Pass
Pass
Pass
run
run
+   virtual bool run(Module &M) = 0;
+ 
run
FunctionPass
Pass

FunctionPass
FunctionPass
FunctionPass
FunctionPass
FunctionPass
runOnFunction
FunctionPass
FunctionPass
doInitialization(Module &)
+   virtual bool doInitialization(Module &M);
+ 
doIninitialize
FunctionPass
doInitialization
doInitialization
malloc
free
malloc()
free()
doInitialization
runOnFunction
+   virtual bool runOnFunction(Function &F) = 0;
+ 
runOnFunction
doFinalization(Module &)
+   virtual bool doFinalization(Module &M);
+ 
doFinalization
runOnFunction
BasicBlockPass
BasicBlockPass
FunctionPass
runOnBasicBlock
FunctionPass
BasicBlockPass
doInitialization(Module &)
doFinalization(Module &)
FunctionPass
doInitialization(Function &)
+   virtual bool doInitialization(Function &F);
+ 
doIninitialize
BasicBlockPass
FunctionPass
doInitialization
doInitialization
runOnBasicBlock
+   virtual bool runOnBasicBlock(BasicBlock &BB) = 0;
+ 
BasicBlockPass
doFinalization(Function
+ &)
+   virtual bool doFinalization(Function &F);
+ 
doFinalization
runOnBasicBlock
MachineFunctionPass
MachineFunctionPass
MachineFunctionPass
FunctionPass
FunctionPass
MachineFunctionPass
MachineFunctionPass
runOnMachineFunction
runOnMachineFunction(MachineFunction &MF)
+   virtual bool runOnMachineFunction(MachineFunction &MF) = 0;
+ 
runOnMachineFunction
MachineFunctionPass
MachineFunctionPass
runOnMachineFunction
MachineFunction
Module
MachineFunctionPass
Function
MachineFunction
MachineFunction
getFunction()
Function
MachineFunctionPass
RegisterOpt
opt
RegisterAnalysis
analysis
RegisterLLC
llc
RegisterPass
Register*
opt
analyze
--help
--debug-pass
analyze
print
print
+   virtual void print(std::ostream &O, const Module *M) const;
+ 
print
analyze
ostream
Module
Pass::dump()
PassManager
PassManager
getAnalysisUsage
getAnalysisUsage
+   virtual void getAnalysisUsage(AnalysisUsage &Info) const;
+ 
getAnalysisUsage

AnalysisUsage
AnalysisUsage
+   // addRequires - Add the specified pass to the required set for your pass.
+   template<class PassClass>
+   AnalysisUsage &AnalysisUsage::addRequired();
+ 
+   // addPreserved - Add the specified pass to the set of analyses preserved by
+   // this pass
+   template<class PassClass>
+   AnalysisUsage &AnalysisUsage::addPreserved();
+ 
+   // setPreservesAll - Call this if the pass does not modify its input at all
+   void AnalysisUsage::setPreservesAll();
+ 
+   // setPreservesCFG - This function should be called by the pass, iff they do not:
+   //
+   //  1. Add or remove basic blocks from the function
+   //  2. Modify terminator instructions in any way.
+   //
+   //  This is automatically implied for BasicBlockPass's
+   //
+   void AnalysisUsage::setPreservesCFG();
+ 
+   // This is an example implementation from an analysis, which does not modify
+   // the program at all, yet has a prerequisite.
+   void PostDominanceFrontier::getAnalysisUsage(AnalysisUsage &AU) const {
+     AU.setPreservesAll();
+     AU.addRequired<PostDominatorTree>();
+   }
+ 
+   // This example modifies the program, but does not modify the CFG
+   void LICM::getAnalysisUsage(AnalysisUsage &AU) const {
+     AU.setPreservesCFG();
+     AU.addRequired<LoopInfo>();
+   }
+ 
getAnalysis<>
Pass::getAnalysis<>
getAnalysisUsage
+   template<typename PassClass>
+   AnalysisType &getAnalysis();
+ 
getAnalysisUsage
run*
run*
Pass
AnalysisUsage::addRequired()
Pass::getAnalysis()
RegisterAnalysisGroup
RegisterAnalysisGroup
basicaa
AliasAnalysis
gcse
opt -gcse ...
basicaa
opt
+ -somefancyaa -gcse ...
gcse
somefancyaa
RegisterAnalysisGroup
RegisterAnalysisGroup
+   static RegisterAnalysisGroup<AliasAnalysis> A("Alias Analysis");
+ 
+ namespace {
+   // Analysis Group implementations must be registered normally...
+   RegisterOpt<FancyAA>
+   B("somefancyaa", "A more complex alias analysis implementation");
+ 
+   // Declare that we implement the AliasAnalysis interface
+   RegisterAnalysisGroup<AliasAnalysis, FancyAA> C;
+ }
+ 
FancyAA
RegisterAnalysisGroup
AliasAnalysis
+ namespace {
+   // Analysis Group implementations must be registered normally...
+   RegisterOpt<BasicAliasAnalysis>
+   D("basicaa", "Basic Alias Analysis (default AA impl)");
+ 
+   // Declare that we implement the AliasAnalysis interface
+   RegisterAnalysisGroup<AliasAnalysis, BasicAliasAnalysis, true> E;
+ }
+ 
RegisterAnalysisGroup
BasicAliasAnalysis
PassManager
PassManager
PassManager
PassManager
PassManager
FunctionPass
FunctionPass
FunctionPass
DominatorSet
PassManager
getAnalysisUsage
PassManager
--debug-pass
--debug-pass
opt --help-hidden
--debug-pass=Structure
gcse
licm
+ $ opt -load ../../../lib/Debug/libhello.so -gcse -licm --debug-pass=Structure < hello.bc > /dev/null
+ Module Pass Manager
+   Function Pass Manager
+     Dominator Set Construction
+     Immediate Dominators Construction
+     Global Common Subexpression Elimination
+ --  Immediate Dominators Construction
+ --  Global Common Subexpression Elimination
+     Natural Loop Construction
+     Loop Invariant Code Motion
+ --  Natural Loop Construction
+ --  Loop Invariant Code Motion
+     Module Verifier
+ --  Dominator Set Construction
+ --  Module Verifier
+   Bytecode Writer
+ --Bytecode Writer
+ 
--
opt
+ $ opt -load ../../../lib/Debug/libhello.so -gcse -hello -licm --debug-pass=Structure < hello.bc > /dev/null
+ Module Pass Manager
+   Function Pass Manager
+     Dominator Set Construction
+     Immediate Dominators Construction
+     Global Common Subexpression Elimination
+ --  Dominator Set Construction
+ --  Immediate Dominators Construction
+ --  Global Common Subexpression Elimination
+     Hello World Pass
+ --  Hello World Pass
+     Dominator Set Construction
+     Natural Loop Construction
+     Loop Invariant Code Motion
+ --  Natural Loop Construction
+ --  Loop Invariant Code Motion
+     Module Verifier
+ --  Dominator Set Construction
+ --  Module Verifier
+   Bytecode Writer
+ --Bytecode Writer
+ Hello: __main
+ Hello: puts
+ Hello: main
+ 
getAnalysisUsage
+     // We don't modify the program, so we preserve all analyses
+     virtual void getAnalysisUsage(AnalysisUsage &AU) const {
+       AU.setPreservesAll();
+     }
+ 
+ $ opt -load ../../../lib/Debug/libhello.so -gcse -hello -licm --debug-pass=Structure < hello.bc > /dev/null
+ Pass Arguments:  -gcse -hello -licm
+ Module Pass Manager
+   Function Pass Manager
+     Dominator Set Construction
+     Immediate Dominators Construction
+     Global Common Subexpression Elimination
+ --  Immediate Dominators Construction
+ --  Global Common Subexpression Elimination
+     Hello World Pass
+ --  Hello World Pass
+     Natural Loop Construction
+     Loop Invariant Code Motion
+ --  Loop Invariant Code Motion
+ --  Natural Loop Construction
+     Module Verifier
+ --  Dominator Set Construction
+ --  Module Verifier
+   Bytecode Writer
+ --Bytecode Writer
+ Hello: __main
+ Hello: puts
+ Hello: main
+ 
releaseMemory
+   virtual void releaseMemory();
+ 
PassManager
releaseMemory
releaseMEmory
run*
run*
opt
gdb
opt
+ $ gdb opt
+ GNU gdb 5.0
+ Copyright 2000 Free Software Foundation, Inc.
+ GDB is free software, covered by the GNU General Public License, and you are
+ welcome to change it and/or distribute copies of it under certain conditions.
+ Type "show copying" to see the conditions.
+ There is absolutely no warranty for GDB.  Type "show warranty" for details.
+ This GDB was configured as "sparc-sun-solaris2.6"...
+ (gdb)
+ 
opt
PassManager::run
+ (gdb) break PassManager::run
+ Breakpoint 1 at 0x2413bc: file Pass.cpp, line 70.
+ (gdb) run test.bc -load $(LLVMTOP)/llvm/lib/Debug/[libname].so -[passoption]
+ Starting program: opt test.bc -load $(LLVMTOP)/llvm/lib/Debug/[libname].so -[passoption]
+ Breakpoint 1, PassManager::run (this=0xffbef174, M=@0x70b298) at Pass.cpp:70
+ 70      bool PassManager::run(Module &M) { return PM->run(M); }
+ (gdb)
+ 
opt
PassManager::run
run
delete
PassManager::run
run*
PassManager
ModuleSource
PassManager
run

Module
PassManager
PassManger
ModuleSource
Pass
FunctionPass
Pass
FunctionPass
FunctionPass
FunctionPass
Pass
analyze
analyze
analyze [options] [filename]
analyze
analyze
analyze
analyze -load <plugin.so> -help
analyze
opt
bugpoint
bugpoint
bugpoint [options] [input LLVM ll/bc files] [LLVM passes] --args <program arguments>...
bugpoint
gccas
bugpoint
bugpoint
bugpoint
.bc
.ll
bugpoint
-output
bugpoint
bugpoint
bugpoint
bugpoint
bugpoint
bugpoint
gccas
bugpoint
bugpoint
bugpoint

opt
analyze
-run-llc
-run-jit
bugpoint
bugpoint
bugpoint
bugpoint
bugpoint
bugpoint
bugpoint
-load
bugpoint
bugpoint  ..... |& tee bugpoint.log
bugpoint
bugpoint.log
bugpoint
bugpoint
llvm-link -v
-additional-so <library.so>
<library.so>
-args <program args>
-args
<program args>
bugpoint <bugpoint args> -args -- <program args>
--
-args
bugpoint
-
-args
bugpoint
-disable-{adce,dce,simplifycfg}
bugpoint
bugpoint
-help
-input <filename>
<filename>
-load <plugin.so>
<plugin.so>
bugpoint
bugpoint -load <plugin.so> -help
-output <filename>
<filename>
bugpoint
-run-{int,jit,llc,cbe}
bugpoint
bugpoint
opt
analyze
extract
extract
extract [options] [filename]
extract
extract
extract
extract
bugpoint
gccas
gccas
gccas [options] < filename>
gccas
gccas
gccas
llvm-as
gccld
gccld
gccld
gccld [options] < filename> [ filename ...]
gccld
gccld
gccld
gccld
gccld
LLVM_LIB_SEARCH_PATH
gccld
gccld
LLVM_LIB_SEARCH_PATH
gccld
gccld
gccld
gccld
gccld
gccld
gccld
llvm-link
gccas
llvmprof.out
llc
llc
llc [options] [filename]
llc
llc
llc
llc
llc
llc
llc
lli
lli
lli
lli [options] [filename] [args ...]
lli
lli
llc
lli
-help
-stats
-time-passes
-march=<arch>
lli
-force-interpreter={false,true}
-f=<name>
<name>
int
+     <name> (int, char **, char **)
lli
llc
llvm-as
llvm-as
llvm-as [options] [filename]
llvm-as
llvm-as
-o
llvm-as
llvm-as
llvm-as
llvm-as
llvm-as
llvm-dis
gccas
llvm-dis
llvm-dis
llvm-dis [options] [filename]
llvm-dis
llvm-dis
llvm-dis
llvm-dis
llvm-dis
llvm-dis
llvm-dis
llvm-dis
llvm-as
llvm-link
llvm-link
llvm-link [options] <filename> [filename ...]
llvm-link
llvm-link
llvm-link
llvm-link
llvm-link
llvm-link
llvm-link
llvm-link
llvm-link
gccld
llvm-nm
llvm-nm
llvm-nm [options] [filenames...]
llvm-nm
ar(1)
llvm-nm
llvm-nm
nm(1)
llvm-nm
llvm-nm
nm(1)
llvm-nm
llvm-dis
ar(1)
nm(1)
llvm-prof
llvm-prof
llvm-prof [options] [bytecode file] [LLVM passes]
llvm-prof
llvmprof.out
utils/profile.sh
llvm-prof
utils/profile.sh
--help
--annotated-llvm
-A
--print-all-code
--annotated-llvm
llvm-prof
llvmgcc
llvmgcc
llvmgcc [options] filename
llvmgcc
-S
llvmgcc
gccas
-c
llvmgcc
gccld
gccld
llvmgcc
llvmg++
gccas
gccld
gcc
llvmg++
llvmg++
llvmg++ [options] filename
llvmg++
-S
llvmg++
gccas
-c
llvmg++
gccld
llvmg++
llvmg++
llvmg++
gccas
gccld
opt
opt
opt [options] [filename]
opt
opt
-load
-help
opt
-o
opt
opt
opt
DEBUG
opt -load  <plugin.so> -help
opt
analyze
configure
configure
llvm/projects
-tailcallelim
-licm
-basicaa
vaarg

First Name (required):
Last Name (required):
Email Address (required):
Organizational Affiliation:
Position/Title:

signed	`sbyte, short, int, long, float, double`
unsigned	`ubyte, ushort, uint, ulong`
integer	`ubyte, sbyte, ushort, short, uint, int, ulong, long`
integral	`bool, ubyte, sbyte, ushort, short, uint, int, ulong, long`
floating point	`float, double`
first class	`bool, ubyte, sbyte, ushort, short, + uint, int, ulong, long, float, double, pointer`

`[3 x [4 x int]]`	: 3x4 array integer values.
`[12 x [10 x float]]`	: 12x10 array of single precision floating point values.
`[2 x [3 x [4 x uint]]]`	: 2x3x4 array of unsigned integer values.

`int (int)`	: function taking an `int`, returning an `int`
`float (int, int ) `	: Pointer to a function that takes + an `int` and a pointer to `int`, + returning `float`.
`int (sbyte *, ...)`	: A vararg function that takes at least one pointer to `sbyte` (signed char in C), + which returns an integer. This is the signature for `printf` + in LLVM.

`{ int, int, int }`	: a triple of three `int` values
`{ float, int (int) * }`	: A pair, where the first element is a `float` and the + second element is a pointer to a function that takes an `int`, returning + an `int`.

Variable	Purpose
CC	Tells `configure` which C compiler to use. By default, + `configure` will look for the first GCC C compiler in + `PATH`. Use this variable to override + `configure`'s default behavior.
CXX	Tells `configure` which C++ compiler to use. By default, + `configure` will look for the first GCC C++ compiler in + `PATH`. Use this variable to override + `configure`'s default behavior.

`[4x int]*`	: pointer to array + of four `int` values
`int (int ) `	: A pointer to a function that takes an `int`, returning + an `int`.

Definition Of Operation Of Built In Words
LOGICAL OPERATIONS
Word	Name	Operation	Description
<	LT	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is less than w2, TRUE is pushed back on + the stack, otherwise FALSE is pushed back on the stack.
>	GT	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is greater than w2, TRUE is pushed back on + the stack, otherwise FALSE is pushed back on the stack.
>=	GE	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is greater than or equal to w2, TRUE is + pushed back on the stack, otherwise FALSE is pushed back + on the stack.
<=	LE	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is less than or equal to w2, TRUE is + pushed back on the stack, otherwise FALSE is pushed back + on the stack.
=	EQ	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is equal to w2, TRUE is + pushed back on the stack, otherwise FALSE is pushed back +
<>	NE	w1 w2 -- b	Two values (w1 and w2) are popped off the stack and + compared. If w1 is equal to w2, TRUE is + pushed back on the stack, otherwise FALSE is pushed back +
FALSE	FALSE	-- b	The boolean value FALSE (0) is pushed onto the stack.
TRUE	TRUE	-- b	The boolean value TRUE (-1) is pushed onto the stack.
BITWISE OPERATIONS
Word	Name	Operation	Description
<<	SHL	w1 w2 -- w1<<w2	Two values (w1 and w2) are popped off the stack. The w2 + operand is shifted left by the number of bits given by the + w1 operand. The result is pushed back to the stack.
>>	SHR	w1 w2 -- w1>>w2	Two values (w1 and w2) are popped off the stack. The w2 + operand is shifted right by the number of bits given by the + w1 operand. The result is pushed back to the stack.
OR	OR	w1 w2 -- w2\|w1	Two values (w1 and w2) are popped off the stack. The values + are bitwise OR'd together and pushed back on the stack. This is + not a logical OR. The sequence 1 2 OR yields 3 not 1.
AND	AND	w1 w2 -- w2&w1	Two values (w1 and w2) are popped off the stack. The values + are bitwise AND'd together and pushed back on the stack. This is + not a logical AND. The sequence 1 2 AND yields 0 not 1.
XOR	XOR	w1 w2 -- w2^w1	Two values (w1 and w2) are popped off the stack. The values + are bitwise exclusive OR'd together and pushed back on the stack. + For example, The sequence 1 3 XOR yields 2.
ARITHMETIC OPERATIONS
Word	Name	Operation	Description
ABS	ABS	w -- \|w\|	One value s popped off the stack; its absolute value is computed + and then pushed onto the stack. If w1 is -1 then w2 is 1. If w1 is + 1 then w2 is also 1.
NEG	NEG	w -- -w	One value is popped off the stack which is negated and then + pushed back onto the stack. If w1 is -1 then w2 is 1. If w1 is + 1 then w2 is -1.
+	ADD	w1 w2 -- w2+w1	Two values are popped off the stack. Their sum is pushed back + onto the stack
-	SUB	w1 w2 -- w2-w1	Two values are popped off the stack. Their difference is pushed back + onto the stack
*	MUL	w1 w2 -- w2*w1	Two values are popped off the stack. Their product is pushed back + onto the stack
/	DIV	w1 w2 -- w2/w1	Two values are popped off the stack. Their quotient is pushed back + onto the stack
MOD	MOD	w1 w2 -- w2%w1	Two values are popped off the stack. Their remainder after division + of w1 by w2 is pushed back onto the stack
*/	STAR_SLAH	w1 w2 w3 -- (w3*w2)/w1	Three values are popped off the stack. The product of w1 and w2 is + divided by w3. The result is pushed back onto the stack.
++	INCR	w -- w+1	One value is popped off the stack. It is incremented by one and then + pushed back onto the stack.
--	DECR	w -- w-1	One value is popped off the stack. It is decremented by one and then + pushed back onto the stack.
MIN	MIN	w1 w2 -- (w2<w1?w2:w1)	Two values are popped off the stack. The larger one is pushed back + onto the stack.
MAX	MAX	w1 w2 -- (w2>w1?w2:w1)	Two values are popped off the stack. The larger value is pushed back + onto the stack.
STACK MANIPULATION OPERATIONS
Word	Name	Operation	Description
DROP	DROP	w --	One value is popped off the stack.
DROP2	DROP2	w1 w2 --	Two values are popped off the stack.
NIP	NIP	w1 w2 -- w2	The second value on the stack is removed from the stack. That is, + a value is popped off the stack and retained. Then a second value is + popped and the retained value is pushed.
NIP2	NIP2	w1 w2 w3 w4 -- w3 w4	The third and fourth values on the stack are removed from it. That is, + two values are popped and retained. Then two more values are popped and + the two retained values are pushed back on.
DUP	DUP	w1 -- w1 w1	One value is popped off the stack. That value is then pushed onto + the stack twice to duplicate the top stack vaue.
DUP2	DUP2	w1 w2 -- w1 w2 w1 w2	The top two values on the stack are duplicated. That is, two vaues + are popped off the stack. They are alternately pushed back on the + stack twice each.
SWAP	SWAP	w1 w2 -- w2 w1	The top two stack items are reversed in their order. That is, two + values are popped off the stack and pushed back onto the stack in + the opposite order they were popped.
SWAP2	SWAP2	w1 w2 w3 w4 -- w3 w4 w2 w1	The top four stack items are swapped in pairs. That is, two values + are popped and retained. Then, two more values are popped and retained. + The values are pushed back onto the stack in the reverse order but + in pairs. +
OVER	OVER	w1 w2-- w1 w2 w1	Two values are popped from the stack. They are pushed back + onto the stack in the order w1 w2 w1. This seems to cause the + top stack element to be duplicated "over" the next value.
OVER2	OVER2	w1 w2 w3 w4 -- w1 w2 w3 w4 w1 w2	The third and fourth values on the stack are replicated onto the + top of the stack
ROT	ROT	w1 w2 w3 -- w2 w3 w1	The top three values are rotated. That is, three value are popped + off the stack. They are pushed back onto the stack in the order + w1 w3 w2.
ROT2	ROT2	w1 w2 w3 w4 w5 w6 -- w3 w4 w5 w6 w1 w2	Like ROT but the rotation is done using three pairs instead of + three singles.
RROT	RROT	w1 w2 w3 -- w2 w3 w1	Reverse rotation. Like ROT, but it rotates the other way around. + Essentially, the third element on the stack is moved to the top + of the stack.
RROT2	RROT2	w1 w2 w3 w4 w5 w6 -- w3 w4 w5 w6 w1 w2	Double reverse rotation. Like RROT but the rotation is done using + three pairs instead of three singles. The fifth and sixth stack + elements are moved to the first and second positions
TUCK	TUCK	w1 w2 -- w2 w1 w2	Similar to OVER except that the second operand is being + replicated. Essentially, the first operand is being "tucked" + in between two instances of the second operand. Logically, two + values are popped off the stack. They are placed back on the + stack in the order w2 w1 w2.
TUCK2	TUCK2	w1 w2 w3 w4 -- w3 w4 w1 w2 w3 w4	Like TUCK but a pair of elements is tucked over two pairs. + That is, the top two elements of the stack are duplicated and + inserted into the stack at the fifth and positions.
PICK	PICK	x0 ... Xn n -- x0 ... Xn x0	The top of the stack is used as an index into the remainder of + the stack. The element at the nth position replaces the index + (top of stack). This is useful for cycling through a set of + values. Note that indexing is zero based. So, if n=0 then you + get the second item on the stack. If n=1 you get the third, etc. + Note also that the index is replaced by the n'th value.
SELECT	SELECT	m n X0..Xm Xm+1 .. Xn -- Xm	This is like PICK but the list is removed and you need to specify + both the index and the size of the list. Careful with this one, + the wrong value for n can blow away a huge amount of the stack.
ROLL	ROLL	x0 x1 .. xn n -- x1 .. xn x0	Not Implemented. This one has been left as an exercise to + the student. See Exercise. ROLL requires + a value, "n", to be on the top of the stack. This value specifies how + far into the stack to "roll". The n'th value is moved (not + copied) from its location and replaces the "n" value on the top of the + stack. In this way, all the values between "n" and x0 roll up the stack. + The operation of ROLL is a generalized ROT. The "n" value specifies + how much to rotate. That is, ROLL with n=1 is the same as ROT and + ROLL with n=2 is the same as ROT2.
MEMORY OPERATIONS
Word	Name	Operation	Description
MALLOC	MALLOC	w1 -- p	One value is popped off the stack. The value is used as the size + of a memory block to allocate. The size is in bytes, not words. + The memory allocation is completed and the address of the memory + block is pushed onto the stack.
FREE	FREE	p --	One pointer value is popped off the stack. The value should be + the address of a memory block created by the MALLOC operation. The + associated memory block is freed. Nothing is pushed back on the + stack. Many bugs can be created by attempting to FREE something + that isn't a pointer to a MALLOC allocated memory block. Make + sure you know what's on the stack. One way to do this is with + the following idiom: + `64 MALLOC DUP DUP (use ptr) DUP (use ptr) ... FREE` + This ensures that an extra copy of the pointer is placed on + the stack (for the FREE at the end) and that every use of the + pointer is preceded by a DUP to retain the copy for FREE.
GET	GET	w1 p -- w2 p	An integer index and a pointer to a memory block are popped of + the block. The index is used to index one byte from the memory + block. That byte value is retained, the pointer is pushed again + and the retained value is pushed. Note that the pointer value + s essentially retained in its position so this doesn't count + as a "use ptr" in the FREE idiom.
PUT	PUT	w1 w2 p -- p	An integer value is popped of the stack. This is the value to + be put into a memory block. Another integer value is popped of + the stack. This is the indexed byte in the memory block. A + pointer to the memory block is popped off the stack. The + first value (w1) is then converted to a byte and written + to the element of the memory block(p) at the index given + by the second value (w2). The pointer to the memory block is + pushed back on the stack so this doesn't count as a "use ptr" + in the FREE idiom.
CONTROL FLOW OPERATIONS
Word	Name	Operation	Description
RETURN	RETURN	--	The currently executing definition returns immediately to its caller. + Note that there is an implicit `RETURN` at the end of each + definition, logically located at the semi-colon. The sequence + `RETURN ;` is valid but redundant.
EXIT	EXIT	w1 --	A return value for the program is popped off the stack. The program is + then immediately terminated. This is normally an abnormal exit from the + program. For a normal exit (when `MAIN` finishes), the exit + code will always be zero in accordance with UNIX conventions.
RECURSE	RECURSE	--	The currently executed definition is called again. This operation is + needed since the definition of a word doesn't exist until the semi colon + is reacher. Attempting something like: + `: recurser recurser ;` will yield and error saying that + "recurser" is not defined yet. To accomplish the same thing, change this + to: + `: recurser RECURSE ;`
IF (words...) ENDIF	IF (words...) ENDIF	b --	A boolean value is popped of the stack. If it is non-zero then the "words..." + are executed. Otherwise, execution continues immediately following the ENDIF.
IF (words...) ELSE (words...) ENDIF	IF (words...) ELSE (words...) ENDIF	b --	A boolean value is popped of the stack. If it is non-zero then the "words..." + between IF and ELSE are executed. Otherwise the words between ELSE and ENDIF are + executed. In either case, after the (words....) have executed, execution continues + immediately following the ENDIF.
WHILE (words...) END	WHILE (words...) END	b -- b	The boolean value on the top of the stack is examined. If it is non-zero then the + "words..." between WHILE and END are executed. Execution then begins again at the WHILE where another + boolean is popped off the stack. To prevent this operation from eating up the entire + stack, you should push onto the stack (just before the END) a boolean value that indicates + whether to terminate. Note that since booleans and integers can be coerced you can + use the following "for loop" idiom: + `(push count) WHILE (words...) -- END` + For example: + `10 WHILE DUP >d -- END` + This will print the numbers from 10 down to 1. 10 is pushed on the stack. Since that is + non-zero, the while loop is entered. The top of the stack (10) is duplicated and then + printed out with >d. The top of the stack is decremented, yielding 9 and control is + transfered back to the WHILE keyword. The process starts all over again and repeats until + the top of stack is decremented to 0 at which the WHILE test fails and control is + transfered to the word after the END.
INPUT & OUTPUT OPERATIONS
Word	Name	Operation	Description
SPACE	SPACE	--	A space character is put out. There is no stack effect.
TAB	TAB	--	A tab character is put out. There is no stack effect.
CR	CR	--	A carriage return character is put out. There is no stack effect.
>s	OUT_STR	--	A string pointer is popped from the stack. It is put out.
>d	OUT_STR	--	A value is popped from the stack. It is put out as a decimal integer.
>c	OUT_CHR	--	A value is popped from the stack. It is put out as an ASCII character.
<s	IN_STR	-- s	A string is read from the input via the scanf(3) format string " %as". The + resulting string is pushed onto the stack.
<d	IN_STR	-- w	An integer is read from the input via the scanf(3) format string " %d". The + resulting value is pushed onto the stack
<c	IN_CHR	-- w	A single character is read from the input via the scanf(3) format string + " %c". The value is converted to an integer and pushed onto the stack.
DUMP	DUMP	--	The stack contents are dumped to standard output. This is useful for + debugging your definitions. Put DUMP at the beginning and end of a definition + to see instantly the net effect of the definition.

U	Named object is referenced but undefined in this + bytecode file
C	Common (multiple defs link together into one + def)
W	Weak reference (multiple defs link together into zero or + one defs)
t	Local function (text) object
T	Global function (text) object
d	Local data object
D	Global data object
?	Something unrecognizable

In0	In1	Out
0	0	0
0	1	0
1	0	0
1	1	1

In0	In1	Out
0	0	0
0	1	1
1	0	1
1	1	1

In0	In1	Out
0	0	0
0	1	1
1	0	1
1	1	0

') + print ('Error in form:' + error) + print ('

File Headers

Class overviews

Method information

Overview:

Syntax:

Examples:

Overview:

Syntax:

Examples:

Overview:

Syntax:

Examples:

Overview:

Syntax:

Examples:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Implementation:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Semantics:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example:

Syntax:

Overview:

Arguments:

Semantics:

Example: