[clang][FatLTO] Avoid UnifiedLTO until it can support WPD/CFI (#79061)

Currently, the UnifiedLTO pipeline seems to have trouble with several
LTO features, like SplitLTO units, which means we cannot use important
optimizations like Whole Program Devirtualization or security hardening
instrumentation like CFI.

This patch reverts FatLTO to using distinct pipelines for Full LTO and
ThinLTO. It still avoids module cloning, since that was error prone.

14 files changed, 92 insertions, 72 deletions

diff --git a/clang/lib/CodeGen/BackendUtil.cpp b/clang/lib/CodeGen/BackendUtil.cpp
index a6142d99f3b6..ec203f6f28bc 100644
--- a/clang/lib/CodeGen/BackendUtil.cpp
+++ b/clang/lib/CodeGen/BackendUtil.cpp
@@ -1001,8 +1001,9 @@ void EmitAssemblyHelper::RunOptimizationPipeline(
     }
 
     if (CodeGenOpts.FatLTO) {
-      assert(CodeGenOpts.UnifiedLTO && "FatLTO requires UnifiedLTO");
-      MPM.addPass(PB.buildFatLTODefaultPipeline(Level));
+      MPM.addPass(PB.buildFatLTODefaultPipeline(
+          Level, PrepareForThinLTO,
+          PrepareForThinLTO || shouldEmitRegularLTOSummary()));
     } else if (PrepareForThinLTO) {
       MPM.addPass(PB.buildThinLTOPreLinkDefaultPipeline(Level));
     } else if (PrepareForLTO) {
@@ -1073,8 +1074,7 @@ void EmitAssemblyHelper::RunOptimizationPipeline(
     if (!TheModule->getModuleFlag("EnableSplitLTOUnit"))
       TheModule->addModuleFlag(llvm::Module::Error, "EnableSplitLTOUnit",
                                uint32_t(CodeGenOpts.EnableSplitLTOUnit));
-    // FatLTO always means UnifiedLTO
-    if (!TheModule->getModuleFlag("UnifiedLTO"))
+    if (CodeGenOpts.UnifiedLTO && !TheModule->getModuleFlag("UnifiedLTO"))
       TheModule->addModuleFlag(llvm::Module::Error, "UnifiedLTO", uint32_t(1));
   }
 
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp
index bcba7cbbdb58..5dc614e11aab 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -4854,9 +4854,7 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
   bool UnifiedLTO = false;
   if (IsUsingLTO) {
     UnifiedLTO = Args.hasFlag(options::OPT_funified_lto,
-                              options::OPT_fno_unified_lto, Triple.isPS()) ||
-                 Args.hasFlag(options::OPT_ffat_lto_objects,
-                              options::OPT_fno_fat_lto_objects, false);
+                              options::OPT_fno_unified_lto, Triple.isPS());
     if (UnifiedLTO)
       CmdArgs.push_back("-funified-lto");
   }
diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp
index 7edea7798af1..feb4de2084b8 100644
--- a/clang/lib/Frontend/CompilerInvocation.cpp
+++ b/clang/lib/Frontend/CompilerInvocation.cpp
@@ -1861,20 +1861,6 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions &Opts, ArgList &Args,
     if (Args.hasArg(OPT_funified_lto))
       Opts.PrepareForThinLTO = true;
   }
-  if (Arg *A = Args.getLastArg(options::OPT_ffat_lto_objects,
-                               options::OPT_fno_fat_lto_objects)) {
-    if (A->getOption().matches(options::OPT_ffat_lto_objects)) {
-      if (Arg *Uni = Args.getLastArg(options::OPT_funified_lto,
-                                     options::OPT_fno_unified_lto)) {
-        if (Uni->getOption().matches(options::OPT_fno_unified_lto))
-          Diags.Report(diag::err_drv_incompatible_options)
-              << A->getAsString(Args) << "-fno-unified-lto";
-      } else
-        Diags.Report(diag::err_drv_argument_only_allowed_with)
-            << A->getAsString(Args) << "-funified-lto";
-    }
-  }
-
   if (Arg *A = Args.getLastArg(OPT_fthinlto_index_EQ)) {
     if (IK.getLanguage() != Language::LLVM_IR)
       Diags.Report(diag::err_drv_argument_only_allowed_with)
diff --git a/clang/test/CodeGen/fat-lto-objects.c b/clang/test/CodeGen/fat-lto-objects.c
index 5c8ad1fd93c4..afce798c5c81 100644
--- a/clang/test/CodeGen/fat-lto-objects.c
+++ b/clang/test/CodeGen/fat-lto-objects.c
@@ -1,34 +1,37 @@
 // REQUIRES: x86-registered-target
 
-// RUN: not %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -fsplit-lto-unit -emit-llvm < %s 2>&1 | FileCheck %s --check-prefixes=NO-UNIFIED
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -fsplit-lto-unit -emit-llvm < %s  | FileCheck %s --check-prefixes=FULL,SPLIT
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=FULL,SPLIT
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -funified-lto -ffat-lto-objects -fsplit-lto-unit -emit-llvm < %s  | FileCheck %s --check-prefixes=FULL,SPLIT,UNIFIED
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -funified-lto -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=FULL,NOSPLIT,UNIFIED
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -fsplit-lto-unit -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=THIN,SPLIT
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=THIN,NOSPLIT
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -funified-lto -fsplit-lto-unit -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=THIN,SPLIT,UNIFIED
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -funified-lto -ffat-lto-objects -emit-llvm < %s  | FileCheck %s --check-prefixes=THIN,NOSPLIT,UNIFIED
-
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -funified-lto -ffat-lto-objects -fsplit-lto-unit -emit-obj < %s -o %t.full.split.o
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -fsplit-lto-unit -emit-obj < %s -o %t.full.split.o
 // RUN: llvm-readelf -S %t.full.split.o | FileCheck %s --check-prefixes=ELF
 // RUN: llvm-objcopy --dump-section=.llvm.lto=%t.full.split.bc %t.full.split.o
-// RUN: llvm-dis %t.full.split.bc -o - | FileCheck %s --check-prefixes=THIN,SPLIT,UNIFIED
+// RUN: llvm-dis %t.full.split.bc -o - | FileCheck %s --check-prefixes=FULL,SPLIT,NOUNIFIED
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -funified-lto -ffat-lto-objects -emit-obj < %s -o %t.full.nosplit.o
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -emit-obj < %s -o %t.full.nosplit.o
 // RUN: llvm-readelf -S %t.full.nosplit.o | FileCheck %s --check-prefixes=ELF
 // RUN: llvm-objcopy --dump-section=.llvm.lto=%t.full.nosplit.bc %t.full.nosplit.o
-// RUN: llvm-dis %t.full.nosplit.bc -o - | FileCheck %s --check-prefixes=THIN,NOSPLIT,UNIFIED
+// RUN: llvm-dis %t.full.nosplit.bc -o - | FileCheck %s --check-prefixes=FULL,NOSPLIT,NOUNIFIED
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -funified-lto -fsplit-lto-unit -ffat-lto-objects -emit-obj < %s -o %t.thin.split.o
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -fsplit-lto-unit -ffat-lto-objects -emit-obj < %s -o %t.thin.split.o
 // RUN: llvm-readelf -S %t.thin.split.o | FileCheck %s --check-prefixes=ELF
 // RUN: llvm-objcopy --dump-section=.llvm.lto=%t.thin.split.bc %t.thin.split.o
-// RUN: llvm-dis %t.thin.split.bc -o - | FileCheck %s --check-prefixes=THIN,SPLIT,UNIFIED
+// RUN: llvm-dis %t.thin.split.bc -o - | FileCheck %s --check-prefixes=THIN,SPLIT,NOUNIFIED
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -funified-lto -ffat-lto-objects -emit-obj < %s -o %t.thin.nosplit.o
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -ffat-lto-objects -emit-obj < %s -o %t.thin.nosplit.o
 // RUN: llvm-readelf -S %t.thin.nosplit.o | FileCheck %s --check-prefixes=ELF
 // RUN: llvm-objcopy --dump-section=.llvm.lto=%t.thin.nosplit.bc %t.thin.nosplit.o
-// RUN: llvm-dis %t.thin.nosplit.bc -o -  | FileCheck %s --check-prefixes=THIN,NOSPLIT,UNIFIED
+// RUN: llvm-dis %t.thin.nosplit.bc -o -  | FileCheck %s --check-prefixes=THIN,NOSPLIT,NOUNIFIED
+
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=thin -funified-lto -ffat-lto-objects -emit-obj < %s -o %t.unified.o
+// RUN: llvm-readelf -S %t.unified.o | FileCheck %s --check-prefixes=ELF
+// RUN: llvm-objcopy --dump-section=.llvm.lto=%t.unified.bc %t.unified.o
+// RUN: llvm-dis %t.unified.bc -o - | FileCheck %s --check-prefixes=THIN,NOSPLIT,UNIFIED
 
-// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -funified-lto -ffat-lto-objects -fsplit-lto-unit -S < %s -o - \
+// RUN: %clang -cc1 -triple x86_64-unknown-linux-gnu -flto=full -ffat-lto-objects -fsplit-lto-unit -S < %s -o - \
 // RUN: | FileCheck %s --check-prefixes=ASM
 
 /// Be sure we enable split LTO units correctly under -ffat-lto-objects.
@@ -38,9 +41,8 @@
 // FULL-NOT: ![[#]] = !{i32 1, !"ThinLTO", i32 0}
 // THIN-NOT: ![[#]] = !{i32 1, !"ThinLTO", i32 0}
 
-/// FatLTO always uses UnifiedLTO. It's an error if they aren't set together
-//    UNIFIED: ![[#]] = !{i32 1, !"UnifiedLTO", i32 1}
-// NO-UNIFIED: error: invalid argument '-ffat-lto-objects' only allowed with '-funified-lto'
+// UNIFIED: ![[#]] = !{i32 1, !"UnifiedLTO", i32 1}
+// NOUNIFIED-NOT: ![[#]] = !{i32 1, !"UnifiedLTO", i32 1}
 
 // ELF: .llvm.lto
 
diff --git a/clang/test/Driver/fat-lto-objects.c b/clang/test/Driver/fat-lto-objects.c
index 203175d61b73..97002db6edc5 100644
--- a/clang/test/Driver/fat-lto-objects.c
+++ b/clang/test/Driver/fat-lto-objects.c
@@ -1,6 +1,5 @@
 // RUN: %clang --target=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -### %s -c 2>&1 | FileCheck %s -check-prefix=CHECK-CC
 // CHECK-CC: -cc1
-// CHECK-CC-SAME: -funified-lto
 // CHECK-CC-SAME: -emit-obj
 // CHECK-CC-SAME: -ffat-lto-objects
 
@@ -15,21 +14,18 @@
 /// When fat LTO is enabled with -S, we expect asm output and -ffat-lto-objects to be passed to cc1.
 // RUN: %clang --target=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -### %s -S 2>&1 | FileCheck %s -check-prefix=CHECK-CC-S-LTO
 // CHECK-CC-S-LTO: -cc1
-// CHECK-CC-S-LTO-SAME: -funified-lto
 // CHECK-CC-S-NOT: -emit-llvm
 // CHECK-CC-S-LTO-SAME: -ffat-lto-objects
 
 /// When fat LTO is enabled with -S and -emit-llvm, we expect IR output and -ffat-lto-objects to be passed to cc1.
 // RUN: %clang --target=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -### %s -S -emit-llvm 2>&1 | FileCheck %s -check-prefix=CHECK-CC-S-EL-LTO
 // CHECK-CC-S-EL-LTO: -cc1
-// CHECK-CC-S-EL-LTO-SAME: -funified-lto
 // CHECK-CC-S-EL-LTO-SAME: -emit-llvm
 // CHECK-CC-S-EL-LTO-SAME: -ffat-lto-objects
 
 /// When fat LTO is enabled wihtout -S we expect native object output and -ffat-lto-object to be passed to cc1.
 // RUN: %clang --target=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -### %s -c 2>&1 | FileCheck %s -check-prefix=CHECK-CC-C-LTO
 // CHECK-CC-C-LTO: -cc1
-// CHECK-CC-C-LTO: -funified-lto
 // CHECK-CC-C-LTO: -emit-obj
 // CHECK-CC-C-LTO: -ffat-lto-objects
 
@@ -47,13 +43,3 @@
 // RUN:   -fuse-ld=lld -fno-lto -ffat-lto-objects -### 2>&1 | FileCheck --check-prefix=NOLTO %s
 // LTO: "--fat-lto-objects"
 // NOLTO-NOT: "--fat-lto-objects"
-
-/// Make sure that incompatible options emit the correct diagnostics, since -ffat-lto-objects requires -funified-lto
-// RUN: %clang -cc1 -triple=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -funified-lto -emit-llvm-only %s  2>&1 | FileCheck %s -check-prefix=UNIFIED --allow-empty
-// UNIFIED-NOT: error:
-
-// RUN: not %clang -cc1 -triple=x86_64-unknown-linux-gnu -flto -ffat-lto-objects -emit-llvm-only %s  2>&1 | FileCheck %s -check-prefix=MISSING_UNIFIED
-// MISSING_UNIFIED: error: invalid argument '-ffat-lto-objects' only allowed with '-funified-lto'
-
-// RUN: not %clang -cc1 -triple=x86_64-unknown-linux-gnu -flto -fno-unified-lto -ffat-lto-objects -emit-llvm-only %s  2>&1 | FileCheck %s -check-prefix=NO-UNIFIED
-// NO-UNIFIED: error: the combination of '-ffat-lto-objects' and '-fno-unified-lto' is incompatible
diff --git a/llvm/docs/FatLTO.rst b/llvm/docs/FatLTO.rst
index 76b849b16fc9..5864944332fc 100644
--- a/llvm/docs/FatLTO.rst
+++ b/llvm/docs/FatLTO.rst
@@ -29,9 +29,9 @@ Overview
 Within LLVM, FatLTO is supported by choosing the ``FatLTODefaultPipeline``.
 This pipeline will:
 
-#) Run the pre-link UnifiedLTO pipeline on the current module.
+#) Run the pre-link (Thin)LTO pipeline on the current module.
 #) Embed the pre-link bitcode in a special ``.llvm.lto`` section.
-#) Finish optimizing the module using the post-link ThinLTO pipeline.
+#) Finish optimizing the module using the ModuleOptimization pipeline.
 #) Emit the object file, including the new ``.llvm.lto`` section.
 
 .. NOTE
diff --git a/llvm/include/llvm/Passes/PassBuilder.h b/llvm/include/llvm/Passes/PassBuilder.h
index 33cf8af87381..7339b8a98823 100644
--- a/llvm/include/llvm/Passes/PassBuilder.h
+++ b/llvm/include/llvm/Passes/PassBuilder.h
@@ -250,12 +250,9 @@ public:
   ///
   /// This builds a pipeline that runs the LTO/ThinLTO  pre-link pipeline, and
   /// emits a section containing the pre-link bitcode along side the object code
-  /// generated by running the PerModuleDefaultPipeline, used when compiling
-  /// without LTO. It clones the module and runs the LTO/non-LTO pipelines
-  /// separately to avoid any inconsistencies with an ad-hoc pipeline that tries
-  /// to approximate the PerModuleDefaultPipeline from the pre-link LTO
-  /// pipelines.
-  ModulePassManager buildFatLTODefaultPipeline(OptimizationLevel Level);
+  /// generated in non-LTO compilation.
+  ModulePassManager buildFatLTODefaultPipeline(OptimizationLevel Level,
+                                               bool ThinLTO, bool EmitSummary);
 
   /// Build a pre-link, ThinLTO-targeting default optimization pipeline to
   /// a pass manager.
diff --git a/llvm/include/llvm/Transforms/IPO/EmbedBitcodePass.h b/llvm/include/llvm/Transforms/IPO/EmbedBitcodePass.h
index 2bb7d5f1fcf1..12bf0dd6581c 100644
--- a/llvm/include/llvm/Transforms/IPO/EmbedBitcodePass.h
+++ b/llvm/include/llvm/Transforms/IPO/EmbedBitcodePass.h
@@ -24,13 +24,25 @@ namespace llvm {
 class Module;
 class Pass;
 
+struct EmbedBitcodeOptions {
+  EmbedBitcodeOptions() : EmbedBitcodeOptions(false, false) {}
+  EmbedBitcodeOptions(bool IsThinLTO, bool EmitLTOSummary)
+      : IsThinLTO(IsThinLTO), EmitLTOSummary(EmitLTOSummary) {}
+  bool IsThinLTO;
+  bool EmitLTOSummary;
+};
+
 /// Pass embeds a copy of the module optimized with the provided pass pipeline
 /// into a global variable.
 class EmbedBitcodePass : public PassInfoMixin<EmbedBitcodePass> {
-  ModulePassManager MPM;
+  bool IsThinLTO;
+  bool EmitLTOSummary;
 
 public:
-  EmbedBitcodePass() {}
+  EmbedBitcodePass(EmbedBitcodeOptions Opts)
+      : EmbedBitcodePass(Opts.IsThinLTO, Opts.EmitLTOSummary) {}
+  EmbedBitcodePass(bool IsThinLTO, bool EmitLTOSummary)
+      : IsThinLTO(IsThinLTO), EmitLTOSummary(EmitLTOSummary) {}
 
   PreservedAnalyses run(Module &M, ModuleAnalysisManager &);
 
diff --git a/llvm/lib/Passes/PassBuilder.cpp b/llvm/lib/Passes/PassBuilder.cpp
index 19fb136f3756..000594f0e7f4 100644
--- a/llvm/lib/Passes/PassBuilder.cpp
+++ b/llvm/lib/Passes/PassBuilder.cpp
@@ -752,6 +752,26 @@ Expected<HWAddressSanitizerOptions> parseHWASanPassOptions(StringRef Params) {
   return Result;
 }
 
+Expected<EmbedBitcodeOptions> parseEmbedBitcodePassOptions(StringRef Params) {
+  EmbedBitcodeOptions Result;
+  while (!Params.empty()) {
+    StringRef ParamName;
+    std::tie(ParamName, Params) = Params.split(';');
+
+    if (ParamName == "thinlto") {
+      Result.IsThinLTO = true;
+    } else if (ParamName == "emit-summary") {
+      Result.EmitLTOSummary = true;
+    } else {
+      return make_error<StringError>(
+          formatv("invalid EmbedBitcode pass parameter '{0}' ", ParamName)
+              .str(),
+          inconvertibleErrorCode());
+    }
+  }
+  return Result;
+}
+
 Expected<MemorySanitizerOptions> parseMSanPassOptions(StringRef Params) {
   MemorySanitizerOptions Result;
   while (!Params.empty()) {
diff --git a/llvm/lib/Passes/PassBuilderPipelines.cpp b/llvm/lib/Passes/PassBuilderPipelines.cpp
index 525b83dee79b..6ede86382912 100644
--- a/llvm/lib/Passes/PassBuilderPipelines.cpp
+++ b/llvm/lib/Passes/PassBuilderPipelines.cpp
@@ -1531,14 +1531,17 @@ PassBuilder::buildPerModuleDefaultPipeline(OptimizationLevel Level,
 }
 
 ModulePassManager
-PassBuilder::buildFatLTODefaultPipeline(OptimizationLevel Level) {
+PassBuilder::buildFatLTODefaultPipeline(OptimizationLevel Level, bool ThinLTO,
+                                        bool EmitSummary) {
   ModulePassManager MPM;
-  // FatLTO always uses UnifiedLTO, so use the ThinLTOPreLink pipeline
-  MPM.addPass(buildThinLTOPreLinkDefaultPipeline(Level));
-  MPM.addPass(EmbedBitcodePass());
+  if (ThinLTO)
+    MPM.addPass(buildThinLTOPreLinkDefaultPipeline(Level));
+  else
+    MPM.addPass(buildLTOPreLinkDefaultPipeline(Level));
+  MPM.addPass(EmbedBitcodePass(ThinLTO, EmitSummary));
 
-  // Use the ThinLTO post-link pipeline with sample profiling, other
-  if (PGOOpt && PGOOpt->Action == PGOOptions::SampleUse)
+  // Use the ThinLTO post-link pipeline with sample profiling
+  if (ThinLTO && PGOOpt && PGOOpt->Action == PGOOptions::SampleUse)
     MPM.addPass(buildThinLTODefaultPipeline(Level, /*ImportSummary=*/nullptr));
   else {
     // otherwise, just use module optimization
diff --git a/llvm/lib/Passes/PassRegistry.def b/llvm/lib/Passes/PassRegistry.def
index 5f2530209112..e59795c7b084 100644
--- a/llvm/lib/Passes/PassRegistry.def
+++ b/llvm/lib/Passes/PassRegistry.def
@@ -58,7 +58,6 @@ MODULE_PASS("dfsan", DataFlowSanitizerPass())
 MODULE_PASS("dot-callgraph", CallGraphDOTPrinterPass())
 MODULE_PASS("dxil-upgrade", DXILUpgradePass())
 MODULE_PASS("elim-avail-extern", EliminateAvailableExternallyPass())
-MODULE_PASS("embed-bitcode", EmbedBitcodePass())
 MODULE_PASS("extract-blocks", BlockExtractorPass({}, false))
 MODULE_PASS("forceattrs", ForceFunctionAttrsPass())
 MODULE_PASS("function-import", FunctionImportPass())
@@ -162,6 +161,10 @@ MODULE_PASS_WITH_PARAMS(
     "no-group-by-use;no-ignore-single-use;no-merge-const;no-merge-external;"
     "size-only")
 MODULE_PASS_WITH_PARAMS(
+    "embed-bitcode", "EmbedBitcodePass",
+    [](EmbedBitcodeOptions Opts) { return EmbedBitcodePass(Opts); },
+    parseEmbedBitcodePassOptions, "thinlto;emit-summary")
+MODULE_PASS_WITH_PARAMS(
     "globaldce", "GlobalDCEPass",
     [](bool InLTOPostLink) { return GlobalDCEPass(InLTOPostLink); },
     parseGlobalDCEPassOptions, "in-lto-post-link")
diff --git a/llvm/lib/Transforms/IPO/EmbedBitcodePass.cpp b/llvm/lib/Transforms/IPO/EmbedBitcodePass.cpp
index 48ef0772e800..6af3a45701bc 100644
--- a/llvm/lib/Transforms/IPO/EmbedBitcodePass.cpp
+++ b/llvm/lib/Transforms/IPO/EmbedBitcodePass.cpp
@@ -7,6 +7,8 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/Transforms/IPO/EmbedBitcodePass.h"
+#include "llvm/Bitcode/BitcodeWriter.h"
+#include "llvm/Bitcode/BitcodeWriterPass.h"
 #include "llvm/IR/PassManager.h"
 #include "llvm/Pass.h"
 #include "llvm/Support/ErrorHandling.h"
@@ -16,6 +18,7 @@
 #include "llvm/Transforms/IPO/ThinLTOBitcodeWriter.h"
 #include "llvm/Transforms/Utils/ModuleUtils.h"
 
+#include <memory>
 #include <string>
 
 using namespace llvm;
@@ -30,9 +33,16 @@ PreservedAnalyses EmbedBitcodePass::run(Module &M, ModuleAnalysisManager &AM) {
     report_fatal_error(
         "EmbedBitcode pass currently only supports ELF object format",
         /*gen_crash_diag=*/false);
+
   std::string Data;
   raw_string_ostream OS(Data);
-  ThinLTOBitcodeWriterPass(OS, /*ThinLinkOS=*/nullptr).run(M, AM);
+  if (IsThinLTO)
+    ThinLTOBitcodeWriterPass(OS, /*ThinLinkOS=*/nullptr).run(M, AM);
+  else
+    BitcodeWriterPass(OS, /*ShouldPreserveUseListOrder=*/false, EmitLTOSummary)
+        .run(M, AM);
+
   embedBufferInModule(M, MemoryBufferRef(Data, "ModuleData"), ".llvm.lto");
+
   return PreservedAnalyses::all();
 }
diff --git a/llvm/test/CodeGen/X86/fat-lto-section.ll b/llvm/test/CodeGen/X86/fat-lto-section.ll
index 9a4359bab6b5..30c56229a0e2 100644
--- a/llvm/test/CodeGen/X86/fat-lto-section.ll
+++ b/llvm/test/CodeGen/X86/fat-lto-section.ll
@@ -1,5 +1,5 @@
 ;; Ensure that the .llvm.lto section has SHT_EXCLUDE set.
-; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode" -S \
+; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode<thinlto;emit-summary>" -S \
 ; RUN:   | llc --mtriple x86_64-unknown-linux-gnu -filetype=obj \
 ; RUN:   | llvm-readelf - --sections \
 ; RUN:   | FileCheck %s --check-prefix=EXCLUDE
diff --git a/llvm/test/Transforms/EmbedBitcode/embed.ll b/llvm/test/Transforms/EmbedBitcode/embed.ll
index 734bf5274a5f..dffb5cf75547 100644
--- a/llvm/test/Transforms/EmbedBitcode/embed.ll
+++ b/llvm/test/Transforms/EmbedBitcode/embed.ll
@@ -1,4 +1,7 @@
 ; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode" -S | FileCheck %s
+; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode<thinlto>" -S | FileCheck %s
+; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode<emit-summary>" -S | FileCheck %s
+; RUN: opt --mtriple x86_64-unknown-linux-gnu < %s -passes="embed-bitcode<thinlto;emit-summary>" -S | FileCheck %s
 
 @a = global i32 1