diff --git a/clang/lib/Driver/Driver.cpp b/clang/lib/Driver/Driver.cpp
--- a/clang/lib/Driver/Driver.cpp
+++ b/clang/lib/Driver/Driver.cpp
@@ -4604,6 +4604,17 @@
   return Args.MakeArgString(Filename.c_str());
 }
 
+static bool HasPreprocessOutput(const Action &JA) {
+  if (isa<PreprocessJobAction>(JA))
+    return true;
+  if (isa<OffloadAction>(JA) && isa<PreprocessJobAction>(JA.getInputs()[0]))
+    return true;
+  if (isa<OffloadBundlingJobAction>(JA) &&
+      HasPreprocessOutput(*(JA.getInputs()[0])))
+    return true;
+  return false;
+}
+
 const char *Driver::GetNamedOutputPath(Compilation &C, const JobAction &JA,
                                        const char *BaseInput,
                                        StringRef BoundArch, bool AtTopLevel,
@@ -4629,8 +4640,9 @@
   }
 
   // Default to writing to stdout?
-  if (AtTopLevel && !CCGenDiagnostics && isa<PreprocessJobAction>(JA))
+  if (AtTopLevel && !CCGenDiagnostics && HasPreprocessOutput(JA)) {
     return "-";
+  }
 
   // Is this the assembly listing for /FA?
   if (JA.getType() == types::TY_PP_Asm &&
diff --git a/clang/test/Driver/hip-output-file-name.hip b/clang/test/Driver/hip-output-file-name.hip
--- a/clang/test/Driver/hip-output-file-name.hip
+++ b/clang/test/Driver/hip-output-file-name.hip
@@ -7,3 +7,45 @@
 // RUN: 2>&1 | FileCheck %s
 
 // CHECK: {{.*}}clang-offload-bundler{{.*}}"-outputs=hip-output-file-name.o"
+
+// Check -E default output is "-" (stdout).
+
+// RUN: %clang -### -E -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=DASH %s
+
+// RUN: %clang -### -E -save-temps -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=DASH %s
+
+// RUN: %clang -### -E --cuda-device-only -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=CLANG-DASH %s
+
+// RUN: %clang -### -E --cuda-host-only -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=CLANG-DASH %s
+
+// DASH: {{.*}}clang-offload-bundler{{.*}}"-outputs=-"
+// CLANG-DASH: {{.*}}clang{{.*}}"-o" "-"
+
+// Check -E with -o.
+
+// RUN: %clang -### -E -o test.cui -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=OUT %s
+
+// RUN: %clang -### -E -o test.cui -save-temps -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=OUT %s
+
+// RUN: %clang -### -E -o test.cui --cuda-device-only -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=CLANG-OUT %s
+
+// RUN: %clang -### -E -o test.cui --cuda-host-only -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s \
+// RUN: 2>&1 | FileCheck -check-prefixes=CLANG-OUT %s
+
+// OUT: {{.*}}clang-offload-bundler{{.*}}"-outputs=test.cui"
+// CLANG-OUT: {{.*}}clang{{.*}}"-o" "test.cui"