[clang][NewPM] Fixing remaining -O0 tests that are broken under new PM

[leonardchan]

Summary

This is a patch that should go on top of D58375 which addresses/fixes tests that fail under new PM with -O0.

Current reasons for fixes (from other child patches attached to this one):

• Pass false to the new PM AlwaysInliner to prevent emission of lifetime start/end intrinsics.
• Add -fno-experimental-new-pass-manager to tests that use -O (but did not specify a number)
• Add an AssumptionCache callback to the InlineFuntionInfo used for the AlwaysInlinerPass to match codegen of the AlwaysInlinerLegacyPass to generate llvm.assume for -O0 tests.
• Some tests explicitly test for cases between both PMs but not all of the legacy PM ones specify -fno-experimental-new-pass-manager.
• Some New PM debug pass manager output is different from that of the legacy PM. Added a separate run for tests which check that certain passes are run for PGO.
• Add a check that the dwarf fission mode is split when generating a split dwarf file.
• Some of the avx-builtin tests emit extra bitcasts which seem to add unnecessary bitcasts. Logically, the behavior is still the same with these bitcasts, but we just check for them if running with the new PM. I couldn’t come up with a better way to ignore/remove them.
• Some of the stores and allocas in CodeGen/avx512-reduceMinMaxIntrin.c were out of order under new PM, but the behavior of the function is still the same. For this, we just ignore pointers that are meant to be loaded/stored in specific lines. I'm not sure if this is the best solution for this tho.

Fixes in this patch:

• CodeGen/flatten.c will fail under new PM becausec the new PM AlwaysInliner seems to intentionally inline functions but not call sites marked with alwaysinline (D23299)
• Tests that check remarks happen to check them for the inliner which is not turned on at O0. These tests just check that remarks work, but we can make separate tests for the new PM with -O1 so we can turn on the inliner and check the remarks with minimal changes.

[echristo]

clang/lib/CodeGen/BackendUtil.cpp

1104–1107

Can you elaborate more here? We do turn on the always inliner at O0 which makes this comment a bit confusing.

clang/test/CodeGen/aarch64-neon-fma.c

235

Do we really need to check for it or does the autogeneration of testcases do some of this?

[serge-sans-paille]

clang/lib/CodeGen/BackendUtil.cpp

1104–1107

I guess he means

We always pass false here since according to the legacy PM logic for enabling lifetime intrinsics, they are not required with O0

[leonardchan]

clang/lib/CodeGen/BackendUtil.cpp

1104–1107

Yup, my bad. This is what I meant with this comment. Always inlining is used. It's the lifetime intrinsics that aren't always used.

clang/test/CodeGen/aarch64-neon-fma.c

235

I do not know if this is actually an autogenerated test, but if so then probably not. Is there a way to check this?

[leonardchan]

OK, this should be all the tests and now requesting reviews. Most of the tests just require checking different output that is slightly different between pass managers whereas others were functional changes that involved adding some passes to the new PM pipeline or changing the AlwaysInliner somehow.

[efriedma]

Are the behavior differences between the newpm alwaysinliner and the oldpm alwaysinliner intentional? Specifically, the differences in pass remarks, and the differences in the treatment of the alwaysinline attribute seem suspect. (I'm not that interested in the different bitcasts and different order of inline allocas, although that might be nice to fix if it's a small change.)

[leonardchan]

In D62225#1519394, @efriedma wrote:

Are the behavior differences between the newpm alwaysinliner and the oldpm alwaysinliner intentional? Specifically, the differences in pass remarks, and the differences in the treatment of the alwaysinline attribute seem suspect. (I'm not that interested in the different bitcasts and different order of inline allocas, although that might be nice to fix if it's a small change.)

This is something for @chandlerc to answer, but I think based off D23299 the different treatment of alwaysinline was intentional and the other aspects of it that this patch addresses like the AddAlignmentAssumptions are probably things that were overlooked at the time.

[leonardchan]

For now, I made the avx512-reduceMinMaxIntrin.c test only run under the legacy PM since there were already too many changes in this patch. The changes were lit expecting allocas for temporary variables to be in specific orders which weren't respected in the new PM for some reason, but still followed the same logic. I can update this in a later patch separated from this one to make reviewing easier.

[chandlerc]

I think this ultimately needs to be split up into smaller patches. A bunch of these things can be landed independently. Here is my first cut at things to split out, each one into its own patch.

1. the LLVM change to the always inliner
2. the Clang change to how we build the always inliner
3. the PGO pipeline changes (which I have to admit I still don't fully understand)
4. The additions of -fno-experimental-new-pass-manager for test cases that are explicitly testing legacy PM behavior
5. switching tests to be resilient to changes in attribute group numbering (and adding a RUN line w/ the new PM to ensure we don't regress)

for #5 (or others) where *some* testing needs to be working before they can land, just sequence them after whatever they depend on

Other things I think can also be split out, but I suspect into *different* changes from what you have here:

6. Instead of passing -fno-experimental-new-pass-manager for tests that use -O but don't specify a number, Clang should pick a consistent value for the level I think

I'd be interested to then see what is left here.

clang/lib/CodeGen/BackendUtil.cpp

1104–1107

I don't think explaining it in terms of one pass manager or another is the righ thing to do.

Instead, I'd say what the desired result is:

Build a minimal pipeline based on the semantics required by Clang,
which is just that always inlining occurs. Further, disable generating
lifetime intrinsics to avoid enabling further optimizations during
code generation.

[leonardchan]

In D62225#1537248, @chandlerc wrote:

I think this ultimately needs to be split up into smaller patches. A bunch of these things can be landed independently. Here is my first cut at things to split out, each one into its own patch.

Done. I split them into 6 other patches plus this one.

1. the LLVM change to the always inliner

Addressed in D63170.

2. the Clang change to how we build the always inliner

Addressed in D63153.

3. the PGO pipeline changes (which I have to admit I still don't fully understand)

Addressed in D63155. There aren't any logical changes to PGO in this. This just adds PGOInstrumentationGenCreateVar to the pipeline such that ./default.profraw is generated and allows Profile/gcc-flag-compatibility.c to pass.

4. The additions of -fno-experimental-new-pass-manager for test cases that are explicitly testing legacy PM behavior

Addressed in D63156. I also combined #6 in this one since those -O tests are pretty much -O2 tests.

5. switching tests to be resilient to changes in attribute group numbering (and adding a RUN line w/ the new PM to ensure we don't regress)

for #5 (or others) where *some* testing needs to be working before they can land, just sequence them after whatever they depend on

Addressed in D63174. The tests that were failing due to attribute numbering don't seem to fail anymore, so someone must've addressed that in a previous patch. This one now just contains the tests that produces slightly different IR.

Other things I think can also be split out, but I suspect into *different* changes from what you have here:

6. Instead of passing -fno-experimental-new-pass-manager for tests that use -O but don't specify a number, Clang should pick a consistent value for the level I think

D63168 seems separate from the rest and fixes CodeGen/split-debug-single-file.c.

I'd be interested to then see what is left here.

The remaining ones here are the flatten tests and optimization remark tests. The flatten tests fail since the new PM AlwaysInliner seems to inline functions only and not calls (based off the description in D23299), so for now they are marked as UNSUPPORTED.

The optimization remark tests fail since it seems that the new PM inliner is not enabled at -O0 and emit slightly different remarks. This patch forces those breaking runs to run with legacy PM only and adds separate new PM tests with -O1 to enable the inliner.

[chandlerc]

FWIW, this LGTM. We may want to tweak the behavior around flattening, but that can happen as a follow-up, and this seems to get us to a better state.

clang/lib/CodeGen/BackendUtil.cpp

if (CodeGenOpts.OptimizationLevel == 0) {
       if (Optional<GCOVOptions> Options = getGCOVOptions(CodeGenOpts))
         MPM.addPass(GCOVProfilerPass(*Options));
       if (Optional<InstrProfOptions> Options =
               getInstrProfOptions(CodeGenOpts, LangOpts))
         MPM.addPass(InstrProfiling(*Options, false));
 
       // Build a minimal pipeline based on the semantics required by Clang,
       // which is just that always inlining occurs.
-      MPM.addPass(AlwaysInlinerPass());
+      // We always pass false here since according to the legacy PM logic for
+      // enabling lifetime intrinsics, they are not required with O0.
+      MPM.addPass(AlwaysInlinerPass(/*InsertLifetimeIntrinsics=*/false));
 
       // At -O0 we directly run necessary sanitizer passes.
       if (LangOpts.Sanitize.has(SanitizerKind::LocalBounds))
         MPM.addPass(createModuleToFunctionPassAdaptor(BoundsCheckingPass()));
 
       // Lastly, add semantically necessary passes for LTO.
       if (IsLTO || IsThinLTO) {
         MPM.addPass(CanonicalizeAliasesPass());
         MPM.addPass(NameAnonGlobalPass());

clang/test/CodeGen/aarch64-neon-across.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
 // RUN:  -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define i16 @test_vaddlv_s8(<8 x i8> %a) #0 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR:#[0-9]+]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDLV_I]] to i16
 // CHECK:   ret i16 [[TMP0]]
 int16_t test_vaddlv_s8(int8x8_t a) {
   return vaddlv_s8(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddlv_s16(<4 x i16> %a) #0 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDLV_I]]
 int32_t test_vaddlv_s16(int16x4_t a) {
   return vaddlv_s16(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddlv_u8(<8 x i8> %a) #0 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDLV_I]] to i16
 // CHECK:   ret i16 [[TMP0]]
 uint16_t test_vaddlv_u8(uint8x8_t a) {
   return vaddlv_u8(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddlv_u16(<4 x i16> %a) #0 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDLV_I]]
 uint32_t test_vaddlv_u16(uint16x4_t a) {
   return vaddlv_u16(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddlvq_s8(<16 x i8> %a) #1 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDLV_I]] to i16
 // CHECK:   ret i16 [[TMP0]]
 int16_t test_vaddlvq_s8(int8x16_t a) {
   return vaddlvq_s8(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddlvq_s16(<8 x i16> %a) #1 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.saddlv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDLV_I]]
 int32_t test_vaddlvq_s16(int16x8_t a) {
   return vaddlvq_s16(a);
 }
 
 // CHECK-LABEL: define i64 @test_vaddlvq_s32(<4 x i32> %a) #1 {
-// CHECK:   [[VADDLVQ_S32_I:%.*]] = call i64 @llvm.aarch64.neon.saddlv.i64.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VADDLVQ_S32_I:%.*]] = call i64 @llvm.aarch64.neon.saddlv.i64.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VADDLVQ_S32_I]]
 int64_t test_vaddlvq_s32(int32x4_t a) {
   return vaddlvq_s32(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddlvq_u8(<16 x i8> %a) #1 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDLV_I]] to i16
 // CHECK:   ret i16 [[TMP0]]
 uint16_t test_vaddlvq_u8(uint8x16_t a) {
   return vaddlvq_u8(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddlvq_u16(<8 x i16> %a) #1 {
-// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VADDLV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddlv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDLV_I]]
 uint32_t test_vaddlvq_u16(uint16x8_t a) {
   return vaddlvq_u16(a);
 }
 
 // CHECK-LABEL: define i64 @test_vaddlvq_u32(<4 x i32> %a) #1 {
-// CHECK:   [[VADDLVQ_U32_I:%.*]] = call i64 @llvm.aarch64.neon.uaddlv.i64.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VADDLVQ_U32_I:%.*]] = call i64 @llvm.aarch64.neon.uaddlv.i64.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VADDLVQ_U32_I]]
 uint64_t test_vaddlvq_u32(uint32x4_t a) {
   return vaddlvq_u32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vmaxv_s8(<8 x i8> %a) #0 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMAXV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vmaxv_s8(int8x8_t a) {
   return vmaxv_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vmaxv_s16(<4 x i16> %a) #0 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMAXV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vmaxv_s16(int16x4_t a) {
   return vmaxv_s16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vmaxv_u8(<8 x i8> %a) #0 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMAXV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vmaxv_u8(uint8x8_t a) {
   return vmaxv_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vmaxv_u16(<4 x i16> %a) #0 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMAXV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vmaxv_u16(uint16x4_t a) {
   return vmaxv_u16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vmaxvq_s8(<16 x i8> %a) #1 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMAXV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vmaxvq_s8(int8x16_t a) {
   return vmaxvq_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vmaxvq_s16(<8 x i16> %a) #1 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMAXV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vmaxvq_s16(int16x8_t a) {
   return vmaxvq_s16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vmaxvq_s32(<4 x i32> %a) #1 {
-// CHECK:   [[VMAXVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VMAXVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.smaxv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VMAXVQ_S32_I]]
 int32_t test_vmaxvq_s32(int32x4_t a) {
   return vmaxvq_s32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vmaxvq_u8(<16 x i8> %a) #1 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMAXV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vmaxvq_u8(uint8x16_t a) {
   return vmaxvq_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vmaxvq_u16(<8 x i16> %a) #1 {
-// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VMAXV_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMAXV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vmaxvq_u16(uint16x8_t a) {
   return vmaxvq_u16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vmaxvq_u32(<4 x i32> %a) #1 {
-// CHECK:   [[VMAXVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VMAXVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.umaxv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VMAXVQ_U32_I]]
 uint32_t test_vmaxvq_u32(uint32x4_t a) {
   return vmaxvq_u32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vminv_s8(<8 x i8> %a) #0 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMINV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vminv_s8(int8x8_t a) {
   return vminv_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vminv_s16(<4 x i16> %a) #0 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMINV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vminv_s16(int16x4_t a) {
   return vminv_s16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vminv_u8(<8 x i8> %a) #0 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMINV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vminv_u8(uint8x8_t a) {
   return vminv_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vminv_u16(<4 x i16> %a) #0 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMINV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vminv_u16(uint16x4_t a) {
   return vminv_u16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vminvq_s8(<16 x i8> %a) #1 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMINV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vminvq_s8(int8x16_t a) {
   return vminvq_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vminvq_s16(<8 x i16> %a) #1 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMINV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vminvq_s16(int16x8_t a) {
   return vminvq_s16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vminvq_s32(<4 x i32> %a) #1 {
-// CHECK:   [[VMINVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VMINVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.sminv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VMINVQ_S32_I]]
 int32_t test_vminvq_s32(int32x4_t a) {
   return vminvq_s32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vminvq_u8(<16 x i8> %a) #1 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VMINV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vminvq_u8(uint8x16_t a) {
   return vminvq_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vminvq_u16(<8 x i16> %a) #1 {
-// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VMINV_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VMINV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vminvq_u16(uint16x8_t a) {
   return vminvq_u16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vminvq_u32(<4 x i32> %a) #1 {
-// CHECK:   [[VMINVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VMINVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.uminv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VMINVQ_U32_I]]
 uint32_t test_vminvq_u32(uint32x4_t a) {
   return vminvq_u32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vaddv_s8(<8 x i8> %a) #0 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vaddv_s8(int8x8_t a) {
   return vaddv_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddv_s16(<4 x i16> %a) #0 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VADDV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vaddv_s16(int16x4_t a) {
   return vaddv_s16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vaddv_u8(<8 x i8> %a) #0 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v8i8(<8 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vaddv_u8(uint8x8_t a) {
   return vaddv_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddv_u16(<4 x i16> %a) #0 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v4i16(<4 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VADDV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vaddv_u16(uint16x4_t a) {
   return vaddv_u16(a);
 }
 
 // CHECK-LABEL: define i8 @test_vaddvq_s8(<16 x i8> %a) #1 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 int8_t test_vaddvq_s8(int8x16_t a) {
   return vaddvq_s8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddvq_s16(<8 x i16> %a) #1 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VADDV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 int16_t test_vaddvq_s16(int16x8_t a) {
   return vaddvq_s16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddvq_s32(<4 x i32> %a) #1 {
-// CHECK:   [[VADDVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VADDVQ_S32_I:%.*]] = call i32 @llvm.aarch64.neon.saddv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDVQ_S32_I]]
 int32_t test_vaddvq_s32(int32x4_t a) {
   return vaddvq_s32(a);
 }
 
 // CHECK-LABEL: define i8 @test_vaddvq_u8(<16 x i8> %a) #1 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v16i8(<16 x i8> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP0:%.*]] = trunc i32 [[VADDV_I]] to i8
 // CHECK:   ret i8 [[TMP0]]
 uint8_t test_vaddvq_u8(uint8x16_t a) {
   return vaddvq_u8(a);
 }
 
 // CHECK-LABEL: define i16 @test_vaddvq_u16(<8 x i16> %a) #1 {
-// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> %a) #3
+// CHECK:   [[VADDV_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v8i16(<8 x i16> %a) [[NOUNWIND_ATTR]]
 // CHECK:   [[TMP2:%.*]] = trunc i32 [[VADDV_I]] to i16
 // CHECK:   ret i16 [[TMP2]]
 uint16_t test_vaddvq_u16(uint16x8_t a) {
   return vaddvq_u16(a);
 }
 
 // CHECK-LABEL: define i32 @test_vaddvq_u32(<4 x i32> %a) #1 {
-// CHECK:   [[VADDVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> %a) #3
+// CHECK:   [[VADDVQ_U32_I:%.*]] = call i32 @llvm.aarch64.neon.uaddv.i32.v4i32(<4 x i32> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VADDVQ_U32_I]]
 uint32_t test_vaddvq_u32(uint32x4_t a) {
   return vaddvq_u32(a);
 }
 
 // CHECK-LABEL: define float @test_vmaxvq_f32(<4 x float> %a) #1 {
-// CHECK:   [[VMAXVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fmaxv.f32.v4f32(<4 x float> %a) #3
+// CHECK:   [[VMAXVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fmaxv.f32.v4f32(<4 x float> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret float [[VMAXVQ_F32_I]]
 float32_t test_vmaxvq_f32(float32x4_t a) {
   return vmaxvq_f32(a);
 }
 
 // CHECK-LABEL: define float @test_vminvq_f32(<4 x float> %a) #1 {
-// CHECK:   [[VMINVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fminv.f32.v4f32(<4 x float> %a) #3
+// CHECK:   [[VMINVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fminv.f32.v4f32(<4 x float> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret float [[VMINVQ_F32_I]]
 float32_t test_vminvq_f32(float32x4_t a) {
   return vminvq_f32(a);
 }
 
 // CHECK-LABEL: define float @test_vmaxnmvq_f32(<4 x float> %a) #1 {
-// CHECK:   [[VMAXNMVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fmaxnmv.f32.v4f32(<4 x float> %a) #3
+// CHECK:   [[VMAXNMVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fmaxnmv.f32.v4f32(<4 x float> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret float [[VMAXNMVQ_F32_I]]
 float32_t test_vmaxnmvq_f32(float32x4_t a) {
   return vmaxnmvq_f32(a);
 }
 
 // CHECK-LABEL: define float @test_vminnmvq_f32(<4 x float> %a) #1 {
-// CHECK:   [[VMINNMVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fminnmv.f32.v4f32(<4 x float> %a) #3
+// CHECK:   [[VMINNMVQ_F32_I:%.*]] = call float @llvm.aarch64.neon.fminnmv.f32.v4f32(<4 x float> %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret float [[VMINNMVQ_F32_I]]
 float32_t test_vminnmvq_f32(float32x4_t a) {
   return vminnmvq_f32(a);
 }
 
 // CHECK: attributes #0 ={{.*}}"min-legal-vector-width"="64"
 // CHECK: attributes #1 ={{.*}}"min-legal-vector-width"="128"
+// CHECK: attributes [[NOUNWIND_ATTR]] = { nounwind }

clang/test/CodeGen/aarch64-neon-fcvt-intrinsics.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
 // RUN:  -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define float @test_vcvtxd_f32_f64(double %a) #0 {
-// CHECK:   [[VCVTXD_F32_F64_I:%.*]] = call float @llvm.aarch64.sisd.fcvtxn(double %a) #2
+// CHECK:   [[VCVTXD_F32_F64_I:%.*]] = call float @llvm.aarch64.sisd.fcvtxn(double %a) [[NOUNWIND_ATTR:#[0-9]+]]
 // CHECK:   ret float [[VCVTXD_F32_F64_I]]
 float32_t test_vcvtxd_f32_f64(float64_t a) {
   return (float32_t)vcvtxd_f32_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtas_s32_f32(float %a) #0 {
-// CHECK:   [[VCVTAS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtas.i32.f32(float %a) #2
+// CHECK:   [[VCVTAS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtas.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTAS_S32_F32_I]]
 int32_t test_vcvtas_s32_f32(float32_t a) {
   return (int32_t)vcvtas_s32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_test_vcvtad_s64_f64(double %a) #0 {
-// CHECK:   [[VCVTAD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtas.i64.f64(double %a) #2
+// CHECK:   [[VCVTAD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtas.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTAD_S64_F64_I]]
 int64_t test_test_vcvtad_s64_f64(float64_t a) {
   return (int64_t)vcvtad_s64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtas_u32_f32(float %a) #0 {
-// CHECK:   [[VCVTAS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtau.i32.f32(float %a) #2
+// CHECK:   [[VCVTAS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtau.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTAS_U32_F32_I]]
 uint32_t test_vcvtas_u32_f32(float32_t a) {
   return (uint32_t)vcvtas_u32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtad_u64_f64(double %a) #0 {
-// CHECK:   [[VCVTAD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtau.i64.f64(double %a) #2
+// CHECK:   [[VCVTAD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtau.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTAD_U64_F64_I]]
 uint64_t test_vcvtad_u64_f64(float64_t a) {
   return (uint64_t)vcvtad_u64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtms_s32_f32(float %a) #0 {
-// CHECK:   [[VCVTMS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtms.i32.f32(float %a) #2
+// CHECK:   [[VCVTMS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtms.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTMS_S32_F32_I]]
 int32_t test_vcvtms_s32_f32(float32_t a) {
   return (int32_t)vcvtms_s32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtmd_s64_f64(double %a) #0 {
-// CHECK:   [[VCVTMD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtms.i64.f64(double %a) #2
+// CHECK:   [[VCVTMD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtms.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTMD_S64_F64_I]]
 int64_t test_vcvtmd_s64_f64(float64_t a) {
   return (int64_t)vcvtmd_s64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtms_u32_f32(float %a) #0 {
-// CHECK:   [[VCVTMS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtmu.i32.f32(float %a) #2
+// CHECK:   [[VCVTMS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtmu.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTMS_U32_F32_I]]
 uint32_t test_vcvtms_u32_f32(float32_t a) {
   return (uint32_t)vcvtms_u32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtmd_u64_f64(double %a) #0 {
-// CHECK:   [[VCVTMD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtmu.i64.f64(double %a) #2
+// CHECK:   [[VCVTMD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtmu.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTMD_U64_F64_I]]
 uint64_t test_vcvtmd_u64_f64(float64_t a) {
   return (uint64_t)vcvtmd_u64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtns_s32_f32(float %a) #0 {
-// CHECK:   [[VCVTNS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtns.i32.f32(float %a) #2
+// CHECK:   [[VCVTNS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtns.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTNS_S32_F32_I]]
 int32_t test_vcvtns_s32_f32(float32_t a) {
   return (int32_t)vcvtns_s32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtnd_s64_f64(double %a) #0 {
-// CHECK:   [[VCVTND_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtns.i64.f64(double %a) #2
+// CHECK:   [[VCVTND_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtns.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTND_S64_F64_I]]
 int64_t test_vcvtnd_s64_f64(float64_t a) {
   return (int64_t)vcvtnd_s64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtns_u32_f32(float %a) #0 {
-// CHECK:   [[VCVTNS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtnu.i32.f32(float %a) #2
+// CHECK:   [[VCVTNS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtnu.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTNS_U32_F32_I]]
 uint32_t test_vcvtns_u32_f32(float32_t a) {
   return (uint32_t)vcvtns_u32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtnd_u64_f64(double %a) #0 {
-// CHECK:   [[VCVTND_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtnu.i64.f64(double %a) #2
+// CHECK:   [[VCVTND_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtnu.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTND_U64_F64_I]]
 uint64_t test_vcvtnd_u64_f64(float64_t a) {
   return (uint64_t)vcvtnd_u64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtps_s32_f32(float %a) #0 {
-// CHECK:   [[VCVTPS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtps.i32.f32(float %a) #2
+// CHECK:   [[VCVTPS_S32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtps.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTPS_S32_F32_I]]
 int32_t test_vcvtps_s32_f32(float32_t a) {
   return (int32_t)vcvtps_s32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtpd_s64_f64(double %a) #0 {
-// CHECK:   [[VCVTPD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtps.i64.f64(double %a) #2
+// CHECK:   [[VCVTPD_S64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtps.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTPD_S64_F64_I]]
 int64_t test_vcvtpd_s64_f64(float64_t a) {
   return (int64_t)vcvtpd_s64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvtps_u32_f32(float %a) #0 {
-// CHECK:   [[VCVTPS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtpu.i32.f32(float %a) #2
+// CHECK:   [[VCVTPS_U32_F32_I:%.*]] = call i32 @llvm.aarch64.neon.fcvtpu.i32.f32(float %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i32 [[VCVTPS_U32_F32_I]]
 uint32_t test_vcvtps_u32_f32(float32_t a) {
   return (uint32_t)vcvtps_u32_f32(a);
 }
 
 // CHECK-LABEL: define i64 @test_vcvtpd_u64_f64(double %a) #0 {
-// CHECK:   [[VCVTPD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtpu.i64.f64(double %a) #2
+// CHECK:   [[VCVTPD_U64_F64_I:%.*]] = call i64 @llvm.aarch64.neon.fcvtpu.i64.f64(double %a) [[NOUNWIND_ATTR]]
 // CHECK:   ret i64 [[VCVTPD_U64_F64_I]]
 uint64_t test_vcvtpd_u64_f64(float64_t a) {
   return (uint64_t)vcvtpd_u64_f64(a);
 }
 
 // CHECK-LABEL: define i32 @test_vcvts_s32_f32(float %a) #0 {
 // CHECK:   [[TMP0:%.*]] = fptosi float %a to i32
 // CHECK:   ret i32 [[TMP0]]

 }
 
 // CHECK-LABEL: define i64 @test_vcvtd_u64_f64(double %a) #0 {
 // CHECK:   [[TMP0:%.*]] = fptoui double %a to i64
 // CHECK:   ret i64 [[TMP0]]
 uint64_t test_vcvtd_u64_f64(float64_t a) {
   return (uint64_t)vcvtd_u64_f64(a);
 }
+
+// CHECK: attributes [[NOUNWIND_ATTR]] = { nounwind }

clang/test/CodeGen/aarch64-neon-fma.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon -S -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define <2 x float> @test_vmla_n_f32(<2 x float> %a, <2 x float> %b, float %c) #0 {
 // CHECK:   [[VECINIT_I:%.*]] = insertelement <2 x float> undef, float %c, i32 0

 float64x2_t test_vfmaq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
   return vfmaq_n_f64(a, b, c);
 }
 
 // CHECK-LABEL: define <2 x double> @test_vfmsq_n_f64(<2 x double> %a, <2 x double> %b, double %c) #1 {
 // CHECK:   [[SUB_I:%.*]] = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %b
 // CHECK:   [[VECINIT_I:%.*]] = insertelement <2 x double> undef, double %c, i32 0
 // CHECK:   [[VECINIT1_I:%.*]] = insertelement <2 x double> [[VECINIT_I]], double %c, i32 1
-// CHECK:   [[TMP6:%.*]] = call <2 x double> @llvm.fma.v2f64(<2 x double> [[SUB_I]], <2 x double> [[VECINIT1_I]], <2 x double> %a) #3
+// CHECK:   [[TMP6:%.*]] = call <2 x double> @llvm.fma.v2f64(<2 x double> [[SUB_I]], <2 x double> [[VECINIT1_I]], <2 x double> %a) [[NOUNWIND_ATTR:#[0-9]+]]
 // CHECK:   ret <2 x double> [[TMP6]]
 float64x2_t test_vfmsq_n_f64(float64x2_t a, float64x2_t b, float64_t c) {
   return vfmsq_n_f64(a, b, c);
 }
 
 // CHECK: attributes #0 ={{.*}}"min-legal-vector-width"="64"
 // CHECK: attributes #1 ={{.*}}"min-legal-vector-width"="128"
+// CHECK: attributes [[NOUNWIND_ATTR]] = { nounwind }

clang/test/CodeGen/aarch64-neon-perm.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
 // RUN: -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 #include <arm_neon.h>
 
 // CHECK-LABEL: @test_vuzp1_s8(
 // CHECK:   [[SHUFFLE_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>

clang/test/CodeGen/aarch64-neon-tbl.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
 // RUN: -disable-O0-optnone  -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 // Test new aarch64 intrinsics and types
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define <8 x i8> @test_vtbl1_s8(<8 x i8> %a, <8 x i8> %b) #0 {

clang/test/CodeGen/aarch64-poly128.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // REQUIRES: aarch64-registered-target
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
 // RUN: -disable-O0-optnone -ffp-contract=fast -emit-llvm -o - %s | opt -S -mem2reg \
 // RUN:  | FileCheck %s
 
 // Test new aarch64 intrinsics with poly128
 // FIXME: Currently, poly128_t equals to uint128, which will be spilt into
 // two 64-bit GPR(eg X0, X1). Now moving data from X0, X1 to FPR128 will

clang/test/CodeGen/aarch64-v8.2a-neon-intrinsics.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon -target-feature +fullfp16 -target-feature +v8.2a\
 // RUN: -fallow-half-arguments-and-returns -S -disable-O0-optnone -emit-llvm -o - %s \
 // RUN: | opt -S -mem2reg \
 // RUN: | FileCheck %s
 
 // REQUIRES: aarch64-registered-target
 
 #include <arm_neon.h>

clang/test/CodeGen/aggregate-assign-call.c

-// UNSUPPORTED: experimental-new-pass-manager
-
-// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -O1 -S -emit-llvm -o - %s | FileCheck %s --check-prefix=O1
+// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -O1 -fno-experimental-new-pass-manager -S -emit-llvm -o - %s | FileCheck %s --check-prefix=O1
 // RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -O0 -S -emit-llvm -o - %s | FileCheck %s --check-prefix=O0
 //
 // Ensure that we place appropriate lifetime markers around indirectly returned
 // temporaries, and that the lifetime.ends appear in a timely manner.
 //
 // -O1 is used so lifetime markers actually get emitted.
 
 struct S {

clang/test/CodeGen/arm-neon-fma.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple thumbv7-none-linux-gnueabihf \
 // RUN:   -target-abi aapcs \
 // RUN:   -target-cpu cortex-a7 \
 // RUN:   -mfloat-abi hard \
 // RUN:   -ffreestanding \
 // RUN:   -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
 
 #include <arm_neon.h>

clang/test/CodeGen/arm-neon-numeric-maxmin.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck %s
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define <2 x float> @test_vmaxnm_f32(<2 x float> %a, <2 x float> %b) #0 {
 // CHECK:   [[VMAXNM_V2_I:%.*]] = call <2 x float> @llvm.arm.neon.vmaxnm.v2f32(<2 x float> %a, <2 x float> %b) #3
 // CHECK:   ret <2 x float> [[VMAXNM_V2_I]]
 float32x2_t test_vmaxnm_f32(float32x2_t a, float32x2_t b) {

clang/test/CodeGen/arm-neon-vcvtX.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -triple thumbv8-linux-gnueabihf -target-cpu cortex-a57 -ffreestanding -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck %s
 
 #include <arm_neon.h>
 
 // CHECK-LABEL: define <2 x i32> @test_vcvta_s32_f32(<2 x float> %a) #0 {
 // CHECK:   [[VCVTA_S32_V1_I:%.*]] = call <2 x i32> @llvm.arm.neon.vcvtas.v2i32.v2f32(<2 x float> %a) #3
 // CHECK:   ret <2 x i32> [[VCVTA_S32_V1_I]]
 int32x2_t test_vcvta_s32_f32(float32x2_t a) {

clang/test/CodeGen/arm_acle.c

-// UNSUPPORTED: experimental-new-pass-manager
-
-// RUN: %clang_cc1 -ffreestanding -triple armv8-eabi -target-cpu cortex-a57 -O -S -emit-llvm -o - %s | FileCheck %s -check-prefix=ARM -check-prefix=AArch32
-// RUN: %clang_cc1 -ffreestanding -triple aarch64-eabi -target-cpu cortex-a57 -target-feature +neon -target-feature +crc -target-feature +crypto -O -S -emit-llvm -o - %s | FileCheck %s -check-prefix=ARM -check-prefix=AArch64
+// RUN: %clang_cc1 -ffreestanding -triple armv8-eabi -target-cpu cortex-a57 -O -fno-experimental-new-pass-manager -S -emit-llvm -o - %s | FileCheck %s -check-prefix=ARM -check-prefix=AArch32
+// RUN: %clang_cc1 -ffreestanding -triple aarch64-eabi -target-cpu cortex-a57 -target-feature +neon -target-feature +crc -target-feature +crypto -O -fno-experimental-new-pass-manager -S -emit-llvm -o - %s | FileCheck %s -check-prefix=ARM -check-prefix=AArch64
 
 #include <arm_acle.h>
 
 /* 8 SYNCHRONIZATION, BARRIER AND HINT INTRINSICS */
 /* 8.3 Memory Barriers */
 // ARM-LABEL: test_dmb
 // AArch32: call void @llvm.arm.dmb(i32 1)
 // AArch64: call void @llvm.aarch64.dmb(i32 1)

clang/test/CodeGen/available-externally-suppress.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -emit-llvm -o - -triple x86_64-apple-darwin10 %s | FileCheck %s
-// RUN: %clang_cc1 -O2 -fno-inline -emit-llvm -o - -triple x86_64-apple-darwin10 %s | FileCheck %s
-// RUN: %clang_cc1 -flto -O2 -fno-inline -emit-llvm -o - -triple x86_64-apple-darwin10 %s | FileCheck %s -check-prefix=LTO
+// RUN: %clang_cc1 -O2 -fno-experimental-new-pass-manager -fno-inline -emit-llvm -o - -triple x86_64-apple-darwin10 %s | FileCheck %s
+// RUN: %clang_cc1 -flto -O2 -fno-experimental-new-pass-manager -fno-inline -emit-llvm -o - -triple x86_64-apple-darwin10 %s | FileCheck %s -check-prefix=LTO
 
 // Ensure that we don't emit available_externally functions at -O0.
 // Also should not emit them at -O2, unless -flto is present in which case
 // we should preserve them for link-time inlining decisions.
 int x;
 
 inline void f0(int y) { x = y; }
 

clang/test/CodeGen/avx-builtins.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx -emit-llvm -o - -Wall -Werror | FileCheck %s
 // RUN: %clang_cc1 -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +avx -fno-signed-char -emit-llvm -o - -Wall -Werror | FileCheck %s
 // RUN: %clang_cc1 -fms-extensions -fms-compatibility -ffreestanding %s -triple=x86_64-windows-msvc -target-feature +avx -emit-llvm -o - -Wall -Werror | FileCheck %s
 
 
 #include <immintrin.h>
 
 // NOTE: This should match the tests in llvm/test/CodeGen/X86/sse-intrinsics-fast-isel.ll

clang/test/CodeGen/builtin-movdir.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // RUN: %clang_cc1 -ffreestanding -Wall -pedantic -triple x86_64-unknown-unknown -target-feature +movdiri -target-feature +movdir64b %s -emit-llvm -o - | FileCheck %s --check-prefix=X86_64 --check-prefix=CHECK
 // RUN: %clang_cc1 -ffreestanding -Wall -pedantic -triple i386-unknown-unknown -target-feature +movdiri -target-feature +movdir64b %s -emit-llvm -o - | FileCheck %s --check-prefix=X86 --check-prefix=CHECK
 
 #include <immintrin.h>
 #include <stdint.h>
 
 void test_directstore32(void *dst, uint32_t value) {
   // CHECK-LABEL: test_directstore32

clang/test/CodeGen/builtins-ppc-p9vector.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -target-feature +altivec -target-feature +power9-vector \
 // RUN:   -triple powerpc64-unknown-unknown -emit-llvm %s \
 // RUN:   -o - | FileCheck %s -check-prefix=CHECK-BE
 
 // RUN: %clang_cc1 -target-feature +altivec -target-feature +power9-vector \
 // RUN:   -triple powerpc64le-unknown-unknown -emit-llvm %s \
 // RUN:   -o - | FileCheck %s

clang/test/CodeGen/builtins-ppc-vsx.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // REQUIRES: powerpc-registered-target
 // RUN: %clang_cc1 -target-feature +altivec -target-feature +vsx -triple powerpc64-unknown-unknown -emit-llvm %s -o - | FileCheck %s
 // RUN: %clang_cc1 -target-feature +altivec -target-feature +vsx -triple powerpc64le-unknown-unknown -emit-llvm %s -o - | FileCheck %s -check-prefix=CHECK-LE
 #include <altivec.h>
 
 vector bool char vbc = { 0, 1, 0, 1, 0, 1, 0, 1,
                          0, 1, 0, 1, 0, 1, 0, 1 };
 vector signed char vsc = { -8,  9, -10, 11, -12, 13, -14, 15,

clang/test/CodeGen/flatten.c

 // UNSUPPORTED: experimental-new-pass-manager
+// Currently, different code seems to be intentionally generated under the new
+// PM since we alwaysinline functions and not callsites under new PM.
+// Under new PM, f() will not be inlined from g() since f is not marked as
+// alwaysinline.
 
 // RUN: %clang_cc1 -triple=x86_64-linux-gnu %s -emit-llvm -o - | FileCheck %s
 
 void f(void) {}
 
 __attribute__((noinline)) void ni(void) {}
 
 __attribute__((flatten))

clang/test/CodeGen/lto-newpm-pipeline.c

 // RUN:   -check-prefix=CHECK-THIN-OPTIMIZED
 // RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm-bc -o /dev/null -fexperimental-new-pass-manager -fdebug-pass-manager -flto=full -Oz %s 2>&1 | FileCheck %s \
 // RUN:   -check-prefix=CHECK-FULL-OPTIMIZED
 // RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm-bc -o /dev/null -fexperimental-new-pass-manager -fdebug-pass-manager -flto=thin -Oz %s 2>&1 | FileCheck %s \
 // RUN:   -check-prefix=CHECK-THIN-OPTIMIZED
 
 // CHECK-FULL-O0: Starting llvm::Module pass manager run.
 // CHECK-FULL-O0: Running pass: AlwaysInlinerPass
+// CHECK-FULL-O0-NEXT: Running analysis: InnerAnalysisManagerProxy
 // CHECK-FULL-O0-NEXT: Running pass: CanonicalizeAliasesPass
 // CHECK-FULL-O0-NEXT: Running pass: NameAnonGlobalPass
 // CHECK-FULL-O0-NEXT: Running pass: BitcodeWriterPass
 // CHECK-FULL-O0: Finished llvm::Module pass manager run.
 
 // CHECK-THIN-O0: Starting llvm::Module pass manager run.
 // CHECK-THIN-O0: Running pass: AlwaysInlinerPass
+// CHECK-THIN-O0-NEXT: Running analysis: InnerAnalysisManagerProxy
 // CHECK-THIN-O0-NEXT: Running pass: CanonicalizeAliasesPass
 // CHECK-THIN-O0-NEXT: Running pass: NameAnonGlobalPass
 // CHECK-THIN-O0-NEXT: Running pass: ThinLTOBitcodeWriterPass
 // CHECK-THIN-O0: Finished llvm::Module pass manager run.
 
 // TODO: The LTO pre-link pipeline currently invokes
 //       buildPerModuleDefaultPipeline(), which contains LoopVectorizePass.
 //       This may change as the pipeline gets implemented.

clang/test/CodeGen/pgo-instrumentation.c

-// UNSUPPORTED: experimental-new-pass-manager
-
 // Test if PGO instrumentation and use pass are invoked.
 //
 // Ensure Pass PGOInstrumentationGenPass is invoked.
-// RUN: %clang_cc1 -O2 -fprofile-instrument=llvm %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN --check-prefix=CHECK-INSTRPROF
-// RUN: %clang_cc1 -O2 -fprofile-instrument=llvm %s  -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN-NEWPM --check-prefix=CHECK-INSTRPROF-NEWPM
+// RUN: %clang_cc1 -O2 -fprofile-instrument=llvm %s -fno-experimental-new-pass-manager -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN --check-prefix=CHECK-INSTRPROF
+// RUN: %clang_cc1 -O2 -fprofile-instrument=llvm %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN-NEWPM --check-prefix=CHECK-INSTRPROF-NEWPM
 // CHECK-PGOGENPASS-INVOKED-INSTR-GEN: PGOInstrumentationGenPass
 // CHECK-INSTRPROF: Frontend instrumentation-based coverage lowering
 // CHECK-PGOGENPASS-INVOKED-INSTR-GEN-NEWPM: Running pass: PGOInstrumentationGen on
 // CHECK-INSTRPROF-NEWPM: Running pass: InstrProfiling on
 //
 // Ensure Pass PGOInstrumentationGenPass is not invoked.
 // RUN: %clang_cc1 -O2 -fprofile-instrument=clang %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN-CLANG
 // RUN: %clang_cc1 -O2 -fprofile-instrument=clang %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOGENPASS-INVOKED-INSTR-GEN-CLANG-NEWPM
 // CHECK-PGOGENPASS-INVOKED-INSTR-GEN-CLANG-NOT: PGOInstrumentationGenPass
 // CHECK-PGOGENPASS-INVOKED-INSTR-GEN-CLANG-NEWPM-NOT: Running pass: PGOInstrumentationGen on
 
-// RUN: %clang_cc1 -O2 -fprofile-instrument=clang %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF
+// RUN: %clang_cc1 -O2 -fprofile-instrument=clang %s -fno-experimental-new-pass-manager -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF
 // RUN: %clang_cc1 -O2 -fprofile-instrument=clang %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF-NEWPM
-// RUN: %clang_cc1 -O0 -fprofile-instrument=clang %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF
+// RUN: %clang_cc1 -O0 -fprofile-instrument=clang %s -fno-experimental-new-pass-manager -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF
 // RUN: %clang_cc1 -O0 -fprofile-instrument=clang %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s --check-prefix=CHECK-CLANG-INSTRPROF-NEWPM
 // CHECK-CLANG-INSTRPROF: Frontend instrumentation-based coverage lowering
 // CHECK-CLANG-INSTRPROF-NEWPM: Running pass: InstrProfiling on
 
 // Ensure Pass PGOInstrumentationUsePass is invoked.
 // RUN: llvm-profdata merge -o %t.profdata %S/Inputs/pgotestir.profraw
-// RUN: %clang_cc1 -O2 -fprofile-instrument-use-path=%t.profdata %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOUSEPASS-INVOKED-INSTR-USE
+// RUN: %clang_cc1 -O2 -fprofile-instrument-use-path=%t.profdata %s -fno-experimental-new-pass-manager -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOUSEPASS-INVOKED-INSTR-USE
 // RUN: %clang_cc1 -O2 -fprofile-instrument-use-path=%t.profdata %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOUSEPASS-INVOKED-INSTR-USE-NEWPM
 // CHECK-PGOUSEPASS-INVOKED-INSTR-USE: PGOInstrumentationUsePass
 // CHECK-PGOUSEPASS-INVOKED-INSTR-USE-NEWPM: Running pass: PGOInstrumentationUse on
 //
 // Ensure Pass PGOInstrumentationUsePass is not invoked.
 // RUN: llvm-profdata merge -o %t.profdata %S/Inputs/pgotestclang.profraw
 // RUN: %clang_cc1 -O2 -fprofile-instrument-use-path=%t.profdata %s -mllvm -debug-pass=Structure -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOUSEPASS-INVOKED-USE-CLANG
 // RUN: %clang_cc1 -O2 -fprofile-instrument-use-path=%t.profdata %s -fexperimental-new-pass-manager -fdebug-pass-manager -emit-llvm -o - 2>&1 | FileCheck %s -check-prefix=CHECK-PGOUSEPASS-INVOKED-USE-CLANG-NEWPM
 // CHECK-PGOUSEPASS-INVOKED-USE-CLANG-NOT: PGOInstrumentationUsePass
 // CHECK-PGOUSEPASS-INVOKED-USE-CLANG-NEWPM-NOT: Running pass: PGOInstrumentationUse on

llvm/lib/Transforms/IPO/AlwaysInliner.cpp

 #include "llvm/Transforms/IPO/Inliner.h"
 #include "llvm/Transforms/Utils/Cloning.h"
 #include "llvm/Transforms/Utils/ModuleUtils.h"
 
 using namespace llvm;
 
 #define DEBUG_TYPE "inline"
 
-PreservedAnalyses AlwaysInlinerPass::run(Module &M, ModuleAnalysisManager &) {
-  InlineFunctionInfo IFI;
+PreservedAnalyses AlwaysInlinerPass::run(Module &M, ModuleAnalysisManager &AM) {
+  // Attach an inline assumption cache such that AddAlignmentAssumptions() is
+  // called to match the codegen for the legacy PM.
+  FunctionAnalysisManager &FAM =
+      AM.getResult<FunctionAnalysisManagerModuleProxy>(M).getManager();
+  std::function<AssumptionCache &(Function &)> GetAssumptionCache =
+      [&](Function &F) -> AssumptionCache & {
+    return FAM.getResult<AssumptionAnalysis>(F);
+  };
+  InlineFunctionInfo IFI(/*cg=*/nullptr, &GetAssumptionCache);
+
   SmallSetVector<CallSite, 16> Calls;
   bool Changed = false;
   SmallVector<Function *, 16> InlinedFunctions;
   for (Function &F : M)
     if (!F.isDeclaration() && F.hasFnAttribute(Attribute::AlwaysInline) &&
         isInlineViable(F)) {
       Calls.clear();