[llc/opt] Add an option to run all passes twice

Summary: Lately, I have submitted a number of patches to fix bugs that
only occurred when using the same pass manager to compile multiple
modules (generally these bugs are failure to reset some persistent
state). Unfortunately I don't think there is currently a way to test
that from the command line. This adds a very simple flag to both llc
and opt, under which the tools will simply re-run their respective
pass pipelines using the same pass manager on (a clone of the same
module). Additionally, we verify that both outputs are bitwise the
same.

Reviewers: yaron.keren

Subscribers: loladiro, yaron.keren, kcc, llvm-commits

Differential Revision: http://reviews.llvm.org/D14965

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@254774 91177308-0d34-0410-b5e6-96231b3b80d8

test/MC/ELF/empty-twice.ll

@@ -0,0 +1,6 @@
+; Check that there is no persistent state in the ELF emitter that crashes us
+; when we try to reuse the pass manager
+; RUN: llc -compile-twice -filetype=obj %s -o -
+
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32-n8:16:32"
+target triple = "i386-pc-linux-gnu"

test/Other/opt-twice.ll

@@ -0,0 +1,14 @@
+; The pass here doesn't matter (we use deadargelim), but test
+; that the -run-twice options exists, generates output, and
+; doesn't crash
+; RUN: opt -run-twice -deadargelim -S < %s | FileCheck %s
+
+; CHECK: define internal void @test
+define internal {} @test() {
+  ret {} undef
+}
+
+define void @caller() {
+  call {} @test()
+  ret void
+}

tools/llc/llc.cpp

@@ -45,6 +45,7 @@
 #include "llvm/Support/ToolOutputFile.h"
 #include "llvm/Target/TargetMachine.h"
 #include "llvm/Target/TargetSubtargetInfo.h"
+#include "llvm/Transforms/Utils/Cloning.h"
 #include <memory>
 using namespace llvm;
 
@@ -96,6 +97,12 @@ static cl::opt<bool> AsmVerbose("asm-verbose",
                                 cl::desc("Add comments to directives."),
                                 cl::init(true));
 
+static cl::opt<bool>
+    CompileTwice("compile-twice", cl::Hidden,
+                 cl::desc("Run everything twice, re-using the same pass "
+                          "manager and verify the the result is the same."),
+                 cl::init(false));
+
 static int compileModule(char **, LLVMContext &);
 
 static std::unique_ptr<tool_output_file>
@@ -325,10 +332,15 @@ static int compileModule(char **argv, LLVMContext &Context) {
 
   {
     raw_pwrite_stream *OS = &Out->os();
-    std::unique_ptr<buffer_ostream> BOS;
-    if (FileType != TargetMachine::CGFT_AssemblyFile &&
-        !Out->os().supportsSeeking()) {
-      BOS = make_unique<buffer_ostream>(*OS);
+
+    // Manually do the buffering rather than using buffer_ostream,
+    // so we can memcmp the contents in CompileTwice mode
+    SmallVector<char, 0> Buffer;
+    std::unique_ptr<raw_svector_ostream> BOS;
+    if ((FileType != TargetMachine::CGFT_AssemblyFile &&
+         !Out->os().supportsSeeking()) ||
+        CompileTwice) {
+      BOS = make_unique<raw_svector_ostream>(Buffer);
       OS = BOS.get();
     }
 
@@ -378,7 +390,39 @@ static int compileModule(char **argv, LLVMContext &Context) {
     // Before executing passes, print the final values of the LLVM options.
     cl::PrintOptionValues();
 
+    // If requested, run the pass manager over the same module again,
+    // to catch any bugs due to persistent state in the passes. Note that
+    // opt has the same functionality, so it may be worth abstracting this out
+    // in the future.
+    SmallVector<char, 0> CompileTwiceBuffer;
+    if (CompileTwice) {
+      std::unique_ptr<Module> M2(llvm::CloneModule(M.get()));
+      PM.run(*M2);
+      CompileTwiceBuffer = Buffer;
+      Buffer.clear();
+    }
+
     PM.run(*M);
+
+    // Compare the two outputs and make sure they're the same
+    if (CompileTwice) {
+      if (Buffer.size() != CompileTwiceBuffer.size() ||
+          (memcmp(Buffer.data(), CompileTwiceBuffer.data(), Buffer.size()) !=
+           0)) {
+        errs()
+            << "Running the pass manager twice changed the output.\n"
+               "Writing the result of the second run to the specified output\n"
+               "To generate the one-run comparison binary, just run without\n"
+               "the compile-twice option\n";
+        Out->os() << Buffer;
+        Out->keep();
+        return 1;
+      }
+    }
+
+    if (BOS) {
+      Out->os() << Buffer;
+    }
   }
 
   // Declare success.

tools/opt/opt.cpp

@@ -28,6 +28,7 @@
 #include "llvm/IR/DebugInfo.h"
 #include "llvm/IR/IRPrintingPasses.h"
 #include "llvm/IR/LLVMContext.h"
+#include "llvm/IR/LegacyPassManager.h"
 #include "llvm/IR/LegacyPassNameParser.h"
 #include "llvm/IR/Module.h"
 #include "llvm/IR/Verifier.h"
@@ -36,7 +37,6 @@
 #include "llvm/LinkAllIR.h"
 #include "llvm/LinkAllPasses.h"
 #include "llvm/MC/SubtargetFeature.h"
-#include "llvm/IR/LegacyPassManager.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/FileSystem.h"
 #include "llvm/Support/Host.h"
@@ -51,6 +51,7 @@
 #include "llvm/Support/ToolOutputFile.h"
 #include "llvm/Target/TargetMachine.h"
 #include "llvm/Transforms/IPO/PassManagerBuilder.h"
+#include "llvm/Transforms/Utils/Cloning.h"
 #include <algorithm>
 #include <memory>
 using namespace llvm;
@@ -190,6 +191,11 @@ static cl::opt<bool> PreserveAssemblyUseListOrder(
     cl::desc("Preserve use-list order when writing LLVM assembly."),
     cl::init(false), cl::Hidden);
 
+static cl::opt<bool>
+    RunTwice("run-twice",
+             cl::desc("Run all passes twice, re-using the same pass manager."),
+             cl::init(false), cl::Hidden);
+
 static inline void addPass(legacy::PassManagerBase &PM, Pass *P) {
   // Add the pass to the pass manager...
   PM.add(P);
@@ -582,14 +588,25 @@ int main(int argc, char **argv) {
   if (!NoVerify && !VerifyEach)
     Passes.add(createVerifierPass());
 
+  // In run twice mode, we want to make sure the output is bit-by-bit
+  // equivalent if we run the pass manager again, so setup two buffers and
+  // a stream to write to them. Note that llc does something similar and it
+  // may be worth to abstract this out in the future.
+  SmallVector<char, 0> Buffer;
+  SmallVector<char, 0> CompileTwiceBuffer;
+  std::unique_ptr<raw_svector_ostream> BOS;
+  raw_ostream *OS = &Out->os();
+  if (RunTwice) {
+    BOS = make_unique<raw_svector_ostream>(Buffer);
+    OS = BOS.get();
+  }
+
   // Write bitcode or assembly to the output as the last step...
   if (!NoOutput && !AnalyzeOnly) {
     if (OutputAssembly)
-      Passes.add(
-          createPrintModulePass(Out->os(), "", PreserveAssemblyUseListOrder));
+      Passes.add(createPrintModulePass(*OS, "", PreserveAssemblyUseListOrder));
     else
-      Passes.add(
-          createBitcodeWriterPass(Out->os(), PreserveBitcodeUseListOrder));
+      Passes.add(createBitcodeWriterPass(*OS, PreserveBitcodeUseListOrder));
   }
 
   // Before executing passes, print the final values of the LLVM options.
@@ -598,6 +615,27 @@ int main(int argc, char **argv) {
   // Now that we have all of the passes ready, run them.
   Passes.run(*M);
 
+  // If requested, run all passes again with the same pass manager to catch
+  // bugs caused by persistent state in the passes
+  if (RunTwice) {
+    CompileTwiceBuffer = Buffer;
+    Buffer.clear();
+    std::unique_ptr<Module> M2(CloneModule(M.get()));
+    Passes.run(*M2);
+    if (Buffer.size() != CompileTwiceBuffer.size() ||
+        (memcmp(Buffer.data(), CompileTwiceBuffer.data(), Buffer.size()) !=
+         0)) {
+      errs() << "Running the pass manager twice changed the output.\n"
+                "Writing the result of the second run to the specified output."
+                "To generate the one-run comparison binary, just run without\n"
+                "the compile-twice option\n";
+      Out->os() << BOS->str();
+      Out->keep();
+      return 1;
+    }
+    Out->os() << BOS->str();
+  }
+
   // Declare success.
   if (!NoOutput || PrintBreakpoints)
     Out->keep();