[PassManager] First Pass implementation at -O1 pass pipeline

[echristo]

Summary

As a follow-up to my initial mail to llvm-dev here's a first pass at the O1 described there.

Some rough internal testing using a bootstrap and test of clang has shown a combined build and test time for clang with nearly equivalent performance to O3 and quite a speedup over O0 - it's currently a little slower than the existing O1, likely due to the clang+llvm testsuite use of the same binaries many times rather than a few for individual tests. Build time is a bit better. For a larger build and smaller test time (think a couple of unittests), this is a bit better than either O3, O0, or O1. Overall binary size drops significantly compared to O0.

This change doesn't include any change to move from selection dag to fast isel and that will come with other numbers that should help inform that decision. I also haven't done any real debuggability studies with this pipeline yet, I wanted to get the initial start done so that people could see it and we could start tweaking after.

Test updates: Outside of the newpm tests most of the updates are coming from either optimization passes not run anymore (and without a compelling argument at the moment) that were largely used for canonicalization in clang.

Original post:

http://lists.llvm.org/pipermail/llvm-dev/2019-April/131494.html

[hfinkel]

Thanks for starting on this. Can you go ahead and replace the sroa calls with mem2reg calls for O1 and then see what that does to the performance? That strikes me as a major change, but certainly one that potentially makes sense, so I'd rather we go ahead and test it now before we make decisions about other adjustments.

FWIW, I thought that we might run InstCombine less often (or maybe replace it with InstSimplify, in some places). Did you try that?

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

356

By definition, this loses information from the call stack, no?

427

Yes, I'd fall back to using regular DCE.

[chandlerc]

One high level point that is at least worth clarifying, and maybe others will want to suggest a different approach:

The overall approach here is to have as small of a difference between the O1 and O2 pipelines as possible.

An alternative approach that we could take would be to design a focused O1 pipeline without regard to how much it diverges from the O2 pipeline.

Which approach is used somewhat depends on the goals. I feel like the goal here is to get as close to the level of optimization at O2 as possible without losing compile time or coherent backtraces for test / assertion failures. For that goal, the approach taken makes sense. But it seems important to clarify that goal as otherwise I think we'd want to go in very different directions.

In D65410#1613555, @hfinkel wrote:

Thanks for starting on this. Can you go ahead and replace the sroa calls with mem2reg calls for O1 and then see what that does to the performance? That strikes me as a major change, but certainly one that potentially makes sense, so I'd rather we go ahead and test it now before we make decisions about other adjustments.

I really think we need mem2reg at least at -O1... In fact, I really think we need SROA at O1. If it is actually a compile time problem, I'd like to fix that in SROA. I don't really expect it to be though.

FWIW, I thought that we might run InstCombine less often (or maybe replace it with InstSimplify, in some places). Did you try that?

I think the biggest thing to do would be to avoid repeated runs of instcombine over the same code. I suspect we want at least one run after inliner and inside the CGSCC walk for canonicalization. But it'd be great to limit it to exactly one or maaaaybe one before and one after the loop pipeline.

llvm/lib/Passes/PassBuilder.cpp

412–421

I think you can merge all of these?

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

356

Yeah, I'd have really expected this to be skipped.

427

+1

[probinson]

The goal from the original email was:

The design goal is to rewrite the O1
optimization and code generation pipeline to include the set of
optimizations that minimizes build and test time while retaining our
ability to debug.

"Retaining our ability to debug" is a more constraining goal than Chandler's "coherent backtraces for test / assertion failures." A good debugging experience comes from good variable-location information, and minimal reordering of instructions. This is different from "don't mess with my stack frames" which lets you do pretty much anything that doesn't involve a call.

Without presuming to speak for Eric, I don't think there was an explicit goal to make O1 look like a stripped-down O2? But certain pass orderings make sense, and so where the same passes occur, the same ordering would be pretty likely. So, building an O1 pipeline by erasing some stuff from the O2 sequence makes some sense, because it's the practical thing and not because it's a goal to do it that way.

As for specific pass choices: Based on the data from Greg Bedwell's lightning talk last year (https://llvm.org/devmtg/2018-04/talks.html#Lightning_11) the three worst offenders for debuggability were LICM, InstCombine, and SROA. It looks like this patch does exclude SROA but not the others? Or at least not all cases of LICM.

I'd suspect that SROA's main problem is that we don't do a good job of tracking broken-up fragments (the original SROA rewrite didn't even try, IIRC), and if the pass has a good compile-time versus run-time benefit, it's worth investing in making that work better.
LICM just inherently interferes with smooth debugging; it's hard to do anything useful with loops that remains nicely debuggable.
I don't know what InstCombine's problem is.

[echristo]

In D65410#1617366, @probinson wrote:

The goal from the original email was:

The design goal is to rewrite the O1
optimization and code generation pipeline to include the set of
optimizations that minimizes build and test time while retaining our
ability to debug.

"Retaining our ability to debug" is a more constraining goal than Chandler's "coherent backtraces for test / assertion failures." A good debugging experience comes from good variable-location information, and minimal reordering of instructions. This is different from "don't mess with my stack frames" which lets you do pretty much anything that doesn't involve a call.

Without presuming to speak for Eric, I don't think there was an explicit goal to make O1 look like a stripped-down O2? But certain pass orderings make sense, and so where the same passes occur, the same ordering would be pretty likely. So, building an O1 pipeline by erasing some stuff from the O2 sequence makes some sense, because it's the practical thing and not because it's a goal to do it that way.

Not an explicit goal of a stripped down O2, but honestly probably not too far off.

(further below)...

As for specific pass choices: Based on the data from Greg Bedwell's lightning talk last year (https://llvm.org/devmtg/2018-04/talks.html#Lightning_11) the three worst offenders for debuggability were LICM, InstCombine, and SROA. It looks like this patch does exclude SROA but not the others? Or at least not all cases of LICM.

I'd suspect that SROA's main problem is that we don't do a good job of tracking broken-up fragments (the original SROA rewrite didn't even try, IIRC), and if the pass has a good compile-time versus run-time benefit, it's worth investing in making that work better.
LICM just inherently interferes with smooth debugging; it's hard to do anything useful with loops that remains nicely debuggable.
I don't know what InstCombine's problem is.

Smooth is going to be somewhat subjective and there's some work in instcombine (rnk took a stab at it and is currently reviewing a pretty good patch I think in this area). SROA just needs some tracking work if there are still problems.

Now LICM - this is one of those I've gone back and forth on and I'd like to see what debugging and performance look like with and without. I like the general perspective of "remove the abstraction penalties" as an O1/Og sort of thing. That said, ultimately something that people are able to use to debug their code is the ultimate goal - it won't be 100%, but as good as we can make it in the general case.

At any rate, this is just a first pass through both the individual passes are going to need work for debugging as is the particular pipeline.

-eric

[echristo]

In D65410#1613555, @hfinkel wrote:

Thanks for starting on this. Can you go ahead and replace the sroa calls with mem2reg calls for O1 and then see what that does to the performance? That strikes me as a major change, but certainly one that potentially makes sense, so I'd rather we go ahead and test it now before we make decisions about other adjustments.

I'll give it a shot. I think we'll want it in the long run, but happy to run it through the performance blender just to get an idea of what we're getting for our complexity.

FWIW, I thought that we might run InstCombine less often (or maybe replace it with InstSimplify, in some places). Did you try that?

I haven't. It's one part of pass ordering that we'll want to look at, but I figured some optimizing here could happen later. Happy to try a few things though.

[echristo]

I've gone ahead and enabled SROA here. In the testing I've done so far it's helped execute time quite a bit and compile time/object size as well. It'll be really good for use with the trivial auto var initialization option also. SROA is a bit of a worry right now for debugging, but it's an area that's been improved upon significantly and I'm not worried about it getting a lot better.

Future plans here are likely going to be moving it to a more separate pass pipeline so I can pull out things like superfluous instcombines etc while continuing to do incremental performance improvements. In addition, I'm going to do experiments with enabling fast-isel at O1 so we can look at the compile time and performance impact of individual changes there - and hopefully some debugging analysis in the near term.

[echristo]

Update to remove comments around SROA addition.

[echristo]

Ping :)

[mehdi_amini]

LGTM to land this and iterate, but you likely want someone else to confirm :)

llvm/test/Other/new-pm-defaults.ll

230

Just a drive-by idea: you could have simplified the changes here with a new prefix: "CHECK-O23sz" that you could add to the non-O1 invocations.

[chandlerc]

Minor nits around redundant predicates for SROA. With thouse fixed, LGTM.

I'd really love to find a way to make TCO debuggable so that we don't lose that. I'm particularly worried about code that relies on it to not run out of stack. Not sure what the best thing to do here is though. Anyways, not relevant for this iteration. I mostly feel bad for a potential future re-churn of all the tests. ;]

llvm/lib/Passes/PassBuilder.cpp

397

We know O0 isn't used here, so this should be a no-op.

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

263

We early exit at O0 above, so this is a no-op.

293

We only reach here if OptLevel > 0 so this should be redundant?

324

This doesn't have the assert, but I believe this is only used above O0 as well. Maybe just add the assert?

llvm/test/Other/new-pm-defaults.ll

230

Good idea!

[omjavaid]

Re-opening this because I have reverted the commit due to failures seen on LLDB AArch64 buildbot with this commit.

clang/test/CodeGen/2008-07-30-implicit-initialization.c

-// RUN: %clang_cc1 -triple i386-unknown-unknown -O1 -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -triple i386-unknown-unknown -O2 -emit-llvm -o - %s | FileCheck %s
 // CHECK-LABEL: define i32 @f0()
 // CHECK:   ret i32 0
 // CHECK-LABEL: define i32 @f1()
 // CHECK:   ret i32 0
 // CHECK-LABEL: define i32 @f2()
 // CHECK:   ret i32 0
 // <rdar://problem/6113085>
 

clang/test/CodeGen/arm-fp16-arguments.c

-// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi soft -fallow-half-arguments-and-returns -emit-llvm -o - -O1 %s | FileCheck %s --check-prefix=CHECK --check-prefix=SOFT
-// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi hard -fallow-half-arguments-and-returns -emit-llvm -o - -O1 %s | FileCheck %s --check-prefix=CHECK --check-prefix=HARD
-// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi soft -fnative-half-arguments-and-returns -emit-llvm -o - -O1 %s | FileCheck %s --check-prefix=NATIVE
+// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi soft -fallow-half-arguments-and-returns -emit-llvm -o - -O2 %s | FileCheck %s --check-prefix=CHECK --check-prefix=SOFT
+// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi hard -fallow-half-arguments-and-returns -emit-llvm -o - -O2 %s | FileCheck %s --check-prefix=CHECK --check-prefix=HARD
+// RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs -mfloat-abi soft -fnative-half-arguments-and-returns -emit-llvm -o - -O2 %s | FileCheck %s --check-prefix=NATIVE
 
 __fp16 g;
 
 void t1(__fp16 a) { g = a; }
 // SOFT: define void @t1(i32 [[PARAM:%.*]])
 // SOFT: [[TRUNC:%.*]] = trunc i32 [[PARAM]] to i16
 // HARD: define arm_aapcs_vfpcc void @t1(float [[PARAM:%.*]])
 // HARD: [[BITCAST:%.*]] = bitcast float [[PARAM]] to i32

clang/test/CodeGen/arm-vfp16-arguments2.cpp

 // RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs \
-// RUN:   -mfloat-abi soft -target-feature +neon -emit-llvm -o - -O1 %s \
+// RUN:   -mfloat-abi soft -target-feature +neon -emit-llvm -o - -O2 %s \
 // RUN:   | FileCheck %s --check-prefix=CHECK-SOFT
 // RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs \
-// RUN:   -mfloat-abi hard -target-feature +neon -emit-llvm -o - -O1 %s \
+// RUN:   -mfloat-abi hard -target-feature +neon -emit-llvm -o - -O2 %s \
 // RUN:   | FileCheck %s --check-prefix=CHECK-HARD
 // RUN: %clang_cc1 -triple armv7a--none-eabi -target-abi aapcs \
 // RUN:   -mfloat-abi hard -target-feature +neon -target-feature +fullfp16 \
-// RUN:   -emit-llvm -o - -O1 %s \
+// RUN:   -emit-llvm -o - -O2 %s \
 // RUN:   | FileCheck %s --check-prefix=CHECK-FULL
 
 typedef float float32_t;
 typedef __fp16 float16_t;
 typedef __attribute__((neon_vector_type(2))) float32_t float32x2_t;
 typedef __attribute__((neon_vector_type(4))) float16_t float16x4_t;
 
 struct S1 {

clang/test/CodeGen/atomic-ops-libcall.c

 // RUN: %clang_cc1 < %s -triple armv5e-none-linux-gnueabi -emit-llvm -O1 | FileCheck %s
 
 // FIXME: This file should not be checking -O1 output.
 // Ie, it is testing many IR optimizer passes as part of front-end verification.
 
 enum memory_order {
   memory_order_relaxed, memory_order_consume, memory_order_acquire,
   memory_order_release, memory_order_acq_rel, memory_order_seq_cst
 };
 
 int *test_c11_atomic_fetch_add_int_ptr(_Atomic(int *) *p) {
   // CHECK: test_c11_atomic_fetch_add_int_ptr
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 12, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 12, i32 5)
   return __c11_atomic_fetch_add(p, 3, memory_order_seq_cst);
 }
 
 int *test_c11_atomic_fetch_sub_int_ptr(_Atomic(int *) *p) {
   // CHECK: test_c11_atomic_fetch_sub_int_ptr
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 20, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 20, i32 5)
   return __c11_atomic_fetch_sub(p, 5, memory_order_seq_cst);
 }
 
 int test_c11_atomic_fetch_add_int(_Atomic(int) *p) {
   // CHECK: test_c11_atomic_fetch_add_int
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 3, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 3, i32 5)
   return __c11_atomic_fetch_add(p, 3, memory_order_seq_cst);
 }
 
 int test_c11_atomic_fetch_sub_int(_Atomic(int) *p) {
   // CHECK: test_c11_atomic_fetch_sub_int
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 5, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 5, i32 5)
   return __c11_atomic_fetch_sub(p, 5, memory_order_seq_cst);
 }
 
 int *fp2a(int **p) {
   // CHECK: @fp2a
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 4, i32 0)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 4, i32 0)
   // Note, the GNU builtins do not multiply by sizeof(T)!
   return __atomic_fetch_sub(p, 4, memory_order_relaxed);
 }
 
 int test_atomic_fetch_add(int *p) {
   // CHECK: test_atomic_fetch_add
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_add(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_fetch_sub(int *p) {
   // CHECK: test_atomic_fetch_sub
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_sub(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_fetch_and(int *p) {
   // CHECK: test_atomic_fetch_and
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_and_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_and_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_and(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_fetch_or(int *p) {
   // CHECK: test_atomic_fetch_or
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_or_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_or_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_or(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_fetch_xor(int *p) {
   // CHECK: test_atomic_fetch_xor
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_xor_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_xor_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_xor(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_fetch_nand(int *p) {
   // CHECK: test_atomic_fetch_nand
-  // CHECK: {{%[^ ]*}} = tail call i32 @__atomic_fetch_nand_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: {{%[^ ]*}} = call i32 @__atomic_fetch_nand_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   return __atomic_fetch_nand(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_add_fetch(int *p) {
   // CHECK: test_atomic_add_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_add_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // CHECK: {{%[^ ]*}} = add i32 [[CALL]], 55
   return __atomic_add_fetch(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_sub_fetch(int *p) {
   // CHECK: test_atomic_sub_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_sub_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // CHECK: {{%[^ ]*}} = add i32 [[CALL]], -55
   return __atomic_sub_fetch(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_and_fetch(int *p) {
   // CHECK: test_atomic_and_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_and_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_and_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // CHECK: {{%[^ ]*}} = and i32 [[CALL]], 55
   return __atomic_and_fetch(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_or_fetch(int *p) {
   // CHECK: test_atomic_or_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_or_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_or_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // CHECK: {{%[^ ]*}} = or i32 [[CALL]], 55
   return __atomic_or_fetch(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_xor_fetch(int *p) {
   // CHECK: test_atomic_xor_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_xor_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_xor_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // CHECK: {{%[^ ]*}} = xor i32 [[CALL]], 55
   return __atomic_xor_fetch(p, 55, memory_order_seq_cst);
 }
 
 int test_atomic_nand_fetch(int *p) {
   // CHECK: test_atomic_nand_fetch
-  // CHECK: [[CALL:%[^ ]*]] = tail call i32 @__atomic_fetch_nand_4(i8* {{%[0-9]+}}, i32 55, i32 5)
+  // CHECK: [[CALL:%[^ ]*]] = call i32 @__atomic_fetch_nand_4(i8* {{%[0-9]+}}, i32 55, i32 5)
   // FIXME: We should not be checking optimized IR. It changes independently of clang.
   // FIXME-CHECK: [[AND:%[^ ]*]] = and i32 [[CALL]], 55
   // FIXME-CHECK: {{%[^ ]*}} = xor i32 [[AND]], -1
   return __atomic_nand_fetch(p, 55, memory_order_seq_cst);
 }

clang/test/CodeGenCXX/atomicinit.cpp

struct B {
   B(int x) : i(x) {}
 };
 
 _Atomic(B) b;
 
 // CHECK-LABEL: define void @_Z11atomic_initR1Ai
 void atomic_init(A& a, int i) {
   // CHECK-NOT: atomic
-  // CHECK: tail call void @_ZN1BC1Ei
+  // CHECK: call void @_ZN1BC1Ei
   __c11_atomic_init(&b, B(i));
   // CHECK-NEXT: ret void
 }
 
 // CHECK-LABEL: define void @_Z16atomic_init_boolPU7_Atomicbb
 void atomic_init_bool(_Atomic(bool) *ab, bool b) {
   // CHECK-NOT: atomic
   // CHECK: {{zext i1.*to i8}}

clang/test/CodeGenCXX/auto-var-init.cpp

 // CHECK-NEXT:  call void @{{.*}}used{{.*}}%uninit)
 // PATTERN-LABEL: @test_smallpartinit_uninit()
 // PATTERN-O0: call void @llvm.memcpy{{.*}} @__const.test_smallpartinit_uninit.uninit
 // PATTERN-O1: store i8 [[I8]], {{.*}} align 1
 // PATTERN-O1: store i8 42, {{.*}} align 1
 // ZERO-LABEL: @test_smallpartinit_uninit()
 // ZERO-O0: call void @llvm.memset{{.*}}, i8 0,
 // ZERO-O1-LEGACY: store i16 0, i16* %uninit, align 2
-// ZERO-O1-NEWPM: store i16 42, i16* %uninit, align 2
+// ZERO-O1-NEWPM: store i16 0, i16* %uninit, align 2
 
 TEST_BRACES(smallpartinit, smallpartinit);
 // CHECK-LABEL: @test_smallpartinit_braces()
 // CHECK:       %braces = alloca %struct.smallpartinit, align [[ALIGN:[0-9]*]]
 // CHECK-NEXT:  %[[C:[^ ]*]] = getelementptr inbounds %struct.smallpartinit, %struct.smallpartinit* %braces, i32 0, i32 0
 // CHECK-NEXT:  store i8 42, i8* %[[C]], align [[ALIGN]]
 // CHECK-NEXT:  %[[D:[^ ]*]] = getelementptr inbounds %struct.smallpartinit, %struct.smallpartinit* %braces, i32 0, i32 1
 // CHECK-NEXT:  store i8 0, i8* %[[D]], align [[ALIGN]]

 TEST_UNINIT(paddednullinit, paddednullinit);
 // CHECK-LABEL: @test_paddednullinit_uninit()
 // CHECK:       %uninit = alloca %struct.paddednullinit, align
 // CHECK-NEXT:  call void @{{.*}}paddednullinit{{.*}}%uninit)
 // CHECK-NEXT:  call void @{{.*}}used{{.*}}%uninit)
 // PATTERN-LABEL: @test_paddednullinit_uninit()
 // PATTERN-O0: call void @llvm.memcpy{{.*}} @__const.test_paddednullinit_uninit.uninit
 // PATTERN-O1-LEGACY: store i64 [[I64]], i64* %uninit, align 8
-// PATTERN-O1-NEWPM: store i64 2863311360, i64* %uninit, align 8
+// PATTERN-O1-NEWPM: store i64 [[I64]], i64* %uninit, align 8
 // ZERO-LABEL: @test_paddednullinit_uninit()
 // ZERO-O0: call void @llvm.memset{{.*}}, i8 0,
 // ZERO-O1: store i64 0, i64* %uninit, align 8
 
 TEST_BRACES(paddednullinit, paddednullinit);
 // CHECK-LABEL: @test_paddednullinit_braces()
 // CHECK:       %braces = alloca %struct.paddednullinit, align [[ALIGN:[0-9]*]]
 // CHECK-NEXT:  %[[C:[^ ]*]] = getelementptr inbounds %struct.paddednullinit, %struct.paddednullinit* %braces, i32 0, i32 0

 // CHECK:       %uninit = alloca %struct.virtualderived, align
 // CHECK-NEXT:  call void @{{.*}}virtualderived{{.*}}%uninit)
 // CHECK-NEXT:  call void @{{.*}}used{{.*}}%uninit)
 // PATTERN-LABEL: @test_virtualderived_uninit()
 // PATTERN-O0: call void @llvm.memcpy{{.*}} @__const.test_virtualderived_uninit.uninit
 // ZERO-LABEL: @test_virtualderived_uninit()
 // ZERO-O0: call void @llvm.memset{{.*}}, i8 0,
 // ZERO-O1-LEGACY: call void @llvm.memset{{.*}}, i8 0,
-// ZERO-O1-NEWPM: [[FIELD1:%.*]] = getelementptr inbounds %struct.virtualderived, %struct.virtualderived* %uninit, i64 0, i32 1, i32 0, i32 0
-// ZERO-O1-NEWPM: [[FIELD0:%.*]] = getelementptr inbounds %struct.virtualderived, %struct.virtualderived* %uninit, i64 0, i32 0, i32 0
-// ZERO-O1-NEWPM: store i32 (...)** bitcast (i8** getelementptr inbounds ({ [7 x i8*], [5 x i8*] }, { [7 x i8*], [5 x i8*] }* @_ZTV14virtualderived, i64 0, inrange i32 0, i64 5) to i32 (...)**), i32 (...)*** [[FIELD0]], align 8
-// ZERO-O1-NEWPM: store i32 (...)** bitcast (i8** getelementptr inbounds ({ [7 x i8*], [5 x i8*] }, { [7 x i8*], [5 x i8*] }* @_ZTV14virtualderived, i64 0, inrange i32 1, i64 3) to i32 (...)**), i32 (...)*** [[FIELD1]], align 8
+// ZERO-O1-NEWPM: call void @llvm.memset{{.*}}, i8 0,
 
 TEST_BRACES(virtualderived, virtualderived);
 // CHECK-LABEL: @test_virtualderived_braces()
 // CHECK:       %braces = alloca %struct.virtualderived, align [[ALIGN:[0-9]*]]
 // CHECK-NEXT:  bitcast
 // CHECK-NEXT:  call void @llvm.memset{{.*}}(i8* align [[ALIGN]] %{{.*}}, i8 0, i64 16, i1 false)
 // CHECK-NEXT:  call void @{{.*}}virtualderived{{.*}}%braces)
 // CHECK-NEXT:  call void @{{.*}}used{{.*}}%braces)

clang/test/CodeGenCXX/discard-name-values.cpp

 // RUN: %clang_cc1 -emit-llvm  -triple=armv7-apple-darwin -std=c++11 %s -o - -O1 \
 // RUN:    | FileCheck %s
 // RUN: %clang_cc1 -emit-llvm -triple=armv7-apple-darwin -std=c++11 %s -o - -O1 \
 // RUN:    -discard-value-names | FileCheck %s --check-prefix=DISCARDVALUE
 
 extern "C" void branch();
 
 bool test(bool pred) {
   // DISCARDVALUE: br i1 %0, label %2, label %3
   // CHECK: br i1 %pred, label %if.then, label %if.end
 
   if (pred) {
     // DISCARDVALUE: 2:
-    // DISCARDVALUE-NEXT: tail call void @branch()
+    // DISCARDVALUE-NEXT: call void @branch()
     // DISCARDVALUE-NEXT: br label %3
 
     // CHECK: if.then:
-    // CHECK-NEXT: tail call void @branch()
+    // CHECK-NEXT: call void @branch()
     // CHECK-NEXT: br label %if.end
     branch();
   }
 
   // DISCARDVALUE: 3:
   // DISCARDVALUE-NEXT: ret i1 %0
 
   // CHECK: if.end:
   // CHECK-NEXT: ret i1 %pred
   return pred;
 }

clang/test/CodeGenCXX/microsoft-abi-dynamic-cast.cpp

 // RUN: %clang_cc1 -emit-llvm -O1 -o - -fexceptions -triple=i386-pc-win32 %s | FileCheck %s
 
 struct S { char a; };
 struct V { virtual void f(); };
 struct A : virtual V {};
 struct B : S, virtual V {};
 struct T {};
 
 T* test0() { return dynamic_cast<T*>((B*)0); }
 // CHECK-LABEL: define dso_local noalias %struct.T* @"?test0@@YAPAUT@@XZ"()
 // CHECK:   ret %struct.T* null
 
 T* test1(V* x) { return &dynamic_cast<T&>(*x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test1@@YAPAUT@@PAUV@@@Z"(%struct.V* %x)
 // CHECK:        [[CAST:%.*]] = bitcast %struct.V* %x to i8*
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* [[CAST]], i32 0, i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUV@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* [[CAST]], i32 0, i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUV@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 T* test2(A* x) { return &dynamic_cast<T&>(*x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test2@@YAPAUT@@PAUA@@@Z"(%struct.A* %x)
 // CHECK:        [[CAST:%.*]] = bitcast %struct.A* %x to i8*
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = getelementptr %struct.A, %struct.A* %x, i32 0, i32 0
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[CAST]], i32 [[VBOFFS]]
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[VBOFFS]], i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUA@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[VBOFFS]], i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUA@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 T* test3(B* x) { return &dynamic_cast<T&>(*x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test3@@YAPAUT@@PAUB@@@Z"(%struct.B* %x)
 // CHECK:        [[VOIDP:%.*]] = getelementptr %struct.B, %struct.B* %x, i32 0, i32 0, i32 0
 // CHECK-NEXT:   [[VBPTR:%.*]] = getelementptr inbounds i8, i8* [[VOIDP]], i32 4
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = bitcast i8* [[VBPTR:%.*]] to i32**
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[DELTA:%.*]] = add nsw i32 [[VBOFFS]], 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[VOIDP]], i32 [[DELTA]]
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[DELTA]], i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUB@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[DELTA]], i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUB@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 1)
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 T* test4(V* x) { return dynamic_cast<T*>(x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test4@@YAPAUT@@PAUV@@@Z"(%struct.V* %x)
 // CHECK:        [[CAST:%.*]] = bitcast %struct.V* %x to i8*
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* [[CAST]], i32 0, i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUV@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* [[CAST]], i32 0, i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUV@@@8" to i8*), i8* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 T* test5(A* x) { return dynamic_cast<T*>(x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test5@@YAPAUT@@PAUA@@@Z"(%struct.A* %x)
 // CHECK:        [[CHECK:%.*]] = icmp eq %struct.A* %x, null
 // CHECK-NEXT:   br i1 [[CHECK]]
 // CHECK:        [[VOIDP:%.*]] = bitcast %struct.A* %x to i8*
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = getelementptr %struct.A, %struct.A* %x, i32 0, i32 0
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[VOIDP]], i32 [[VBOFFS]]
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* nonnull [[ADJ]], i32 [[VBOFFS]], i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUA@@@8" to i8*), i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* nonnull [[ADJ]], i32 [[VBOFFS]], i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUA@@@8" to i8*), i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
 // CHECK-NEXT:   [[RES:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   br label
 // CHECK:        [[RET:%.*]] = phi %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 T* test6(B* x) { return dynamic_cast<T*>(x); }
 // CHECK-LABEL: define dso_local %struct.T* @"?test6@@YAPAUT@@PAUB@@@Z"(%struct.B* %x)
 // CHECK:        [[CHECK:%.*]] = icmp eq %struct.B* %x, null
 // CHECK-NEXT:   br i1 [[CHECK]]
 // CHECK:        [[CAST:%.*]] = getelementptr %struct.B, %struct.B* %x, i32 0, i32 0, i32 0
 // CHECK-NEXT:   [[VBPTR:%.*]] = getelementptr inbounds i8, i8* [[CAST]], i32 4
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = bitcast i8* [[VBPTR]] to i32**
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[DELTA:%.*]] = add nsw i32 [[VBOFFS]], 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[CAST]], i32 [[DELTA]]
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[DELTA]], i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUB@@@8" to i8*), i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTDynamicCast(i8* [[ADJ]], i32 [[DELTA]], i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUB@@@8" to i8*), i8* {{.*}}bitcast (%rtti.TypeDescriptor7* @"??_R0?AUT@@@8" to i8*), i32 0)
 // CHECK-NEXT:   [[RES:%.*]] = bitcast i8* [[CALL]] to %struct.T*
 // CHECK-NEXT:   br label
 // CHECK:        [[RET:%.*]] = phi %struct.T*
 // CHECK-NEXT:   ret %struct.T* [[RET]]
 
 void* test7(V* x) { return dynamic_cast<void*>(x); }
 // CHECK-LABEL: define dso_local i8* @"?test7@@YAPAXPAUV@@@Z"(%struct.V* %x)
 // CHECK:        [[CAST:%.*]] = bitcast %struct.V* %x to i8*
-// CHECK-NEXT:   [[RET:%.*]] = tail call i8* @__RTCastToVoid(i8* [[CAST]])
+// CHECK-NEXT:   [[RET:%.*]] = call i8* @__RTCastToVoid(i8* [[CAST]])
 // CHECK-NEXT:   ret i8* [[RET]]
 
 void* test8(A* x) { return dynamic_cast<void*>(x); }
 // CHECK-LABEL: define dso_local i8* @"?test8@@YAPAXPAUA@@@Z"(%struct.A* %x)
 // CHECK:        [[CHECK:%.*]] = icmp eq %struct.A* %x, null
 // CHECK-NEXT:   br i1 [[CHECK]]
 // CHECK:        [[VOIDP:%.*]] = bitcast %struct.A* %x to i8*
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = getelementptr %struct.A, %struct.A* %x, i32 0, i32 0
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[VOIDP]], i32 [[VBOFFS]]
-// CHECK-NEXT:   [[RES:%.*]] = tail call i8* @__RTCastToVoid(i8* nonnull [[ADJ]])
+// CHECK-NEXT:   [[RES:%.*]] = call i8* @__RTCastToVoid(i8* nonnull [[ADJ]])
 // CHECK-NEXT:   br label
 // CHECK:        [[RET:%.*]] = phi i8*
 // CHECK-NEXT:   ret i8* [[RET]]
 
 void* test9(B* x) { return dynamic_cast<void*>(x); }
 // CHECK-LABEL: define dso_local i8* @"?test9@@YAPAXPAUB@@@Z"(%struct.B* %x)
 // CHECK:        [[CHECK:%.*]] = icmp eq %struct.B* %x, null
 // CHECK-NEXT:   br i1 [[CHECK]]
 // CHECK:        [[CAST:%.*]] = getelementptr %struct.B, %struct.B* %x, i32 0, i32 0, i32 0
 // CHECK-NEXT:   [[VBPTR:%.*]] = getelementptr inbounds i8, i8* [[CAST]], i32 4
 // CHECK-NEXT:   [[VBPTRPTR:%.*]] = bitcast i8* [[VBPTR]] to i32**
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBPTRPTR]], align 4
 // CHECK-NEXT:   [[VBOFFP:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBOFFS:%.*]] = load i32, i32* [[VBOFFP]], align 4
 // CHECK-NEXT:   [[DELTA:%.*]] = add nsw i32 [[VBOFFS]], 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[CAST]], i32 [[DELTA]]
-// CHECK-NEXT:   [[CALL:%.*]] = tail call i8* @__RTCastToVoid(i8* [[ADJ]])
+// CHECK-NEXT:   [[CALL:%.*]] = call i8* @__RTCastToVoid(i8* [[ADJ]])
 // CHECK-NEXT:   br label
 // CHECK:        [[RET:%.*]] = phi i8*
 // CHECK-NEXT:   ret i8* [[RET]]
 
 namespace PR25606 {
 struct Cleanup {
   ~Cleanup();
 };

clang/test/CodeGenCXX/microsoft-abi-typeid.cpp

 // CHECK:   ret %struct.type_info* bitcast (%rtti.TypeDescriptor7* @"??_R0?AUA@@@8" to %struct.type_info*)
 
 const std::type_info* test2_typeid() { return &typeid(&a); }
 // CHECK-LABEL: define dso_local %struct.type_info* @"?test2_typeid@@YAPBUtype_info@@XZ"()
 // CHECK:   ret %struct.type_info* bitcast (%rtti.TypeDescriptor7* @"??_R0PAUA@@@8" to %struct.type_info*)
 
 const std::type_info* test3_typeid() { return &typeid(*fn()); }
 // CHECK-LABEL: define dso_local %struct.type_info* @"?test3_typeid@@YAPBUtype_info@@XZ"()
-// CHECK:        [[CALL:%.*]] = tail call %struct.A* @"?fn@@YAPAUA@@XZ"()
+// CHECK:        [[CALL:%.*]] = call %struct.A* @"?fn@@YAPAUA@@XZ"()
 // CHECK-NEXT:   [[CMP:%.*]] = icmp eq %struct.A* [[CALL]], null
 // CHECK-NEXT:   br i1 [[CMP]]
-// CHECK:        tail call i8* @__RTtypeid(i8* null)
+// CHECK:        call i8* @__RTtypeid(i8* null)
 // CHECK-NEXT:   unreachable
 // CHECK:        [[THIS:%.*]] = bitcast %struct.A* [[CALL]] to i8*
 // CHECK-NEXT:   [[VBTBLP:%.*]] = getelementptr %struct.A, %struct.A* [[CALL]], i32 0, i32 0
 // CHECK-NEXT:   [[VBTBL:%.*]] = load i32*, i32** [[VBTBLP]], align 4
 // CHECK-NEXT:   [[VBSLOT:%.*]] = getelementptr inbounds i32, i32* [[VBTBL]], i32 1
 // CHECK-NEXT:   [[VBASE_OFFS:%.*]] = load i32, i32* [[VBSLOT]], align 4
 // CHECK-NEXT:   [[ADJ:%.*]] = getelementptr inbounds i8, i8* [[THIS]], i32 [[VBASE_OFFS]]
-// CHECK-NEXT:   [[RT:%.*]] = tail call i8* @__RTtypeid(i8* nonnull [[ADJ]])
+// CHECK-NEXT:   [[RT:%.*]] = call i8* @__RTtypeid(i8* nonnull [[ADJ]])
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[RT]] to %struct.type_info*
 // CHECK-NEXT:   ret %struct.type_info* [[RET]]
 
 const std::type_info* test4_typeid() { return &typeid(b); }
 // CHECK: define dso_local %struct.type_info* @"?test4_typeid@@YAPBUtype_info@@XZ"()
 // CHECK:   ret %struct.type_info* bitcast (%rtti.TypeDescriptor2* @"??_R0H@8" to %struct.type_info*)
 
 const std::type_info* test5_typeid() { return &typeid(v); }
 // CHECK: define dso_local %struct.type_info* @"?test5_typeid@@YAPBUtype_info@@XZ"()
-// CHECK:        [[RT:%.*]] = tail call i8* @__RTtypeid(i8* bitcast (%struct.V* @"?v@@3UV@@A" to i8*))
+// CHECK:        [[RT:%.*]] = call i8* @__RTtypeid(i8* bitcast (%struct.V* @"?v@@3UV@@A" to i8*))
 // CHECK-NEXT:   [[RET:%.*]] = bitcast i8* [[RT]] to %struct.type_info*
 // CHECK-NEXT:   ret %struct.type_info* [[RET]]
 
 namespace PR26329 {
 struct Polymorphic {
   virtual ~Polymorphic();
 };
 

clang/test/CodeGenCXX/nrvo.cpp

X test0() {
   // CHECK-EH:       call {{.*}} @_ZN1XC1Ev
   // CHECK-EH-NEXT:  ret void
   return x;
 }
 
 // CHECK-LABEL: define void @_Z5test1b(
 // CHECK-EH-LABEL: define void @_Z5test1b(
 X test1(bool B) {
-  // CHECK:      tail call {{.*}} @_ZN1XC1Ev
+  // CHECK:      call {{.*}} @_ZN1XC1Ev
   // CHECK-NEXT: ret void
   X x;
   if (B)
     return (x);
   return x;
-  // CHECK-EH:      tail call {{.*}} @_ZN1XC1Ev
+  // CHECK-EH:      call {{.*}} @_ZN1XC1Ev
   // CHECK-EH-NEXT: ret void
 }
 
 // CHECK-LABEL: define void @_Z5test2b
 // CHECK-EH-LABEL: define void @_Z5test2b
 // CHECK-EH-SAME:  personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*)
 X test2(bool B) {
   // No NRVO.
X test2(bool B) {
   // CHECK-EH-03-NEXT: [[T1:%.*]] = extractvalue { i8*, i32 } [[T0]], 0
   // CHECK-EH-03-NEXT: call void @__clang_call_terminate(i8* [[T1]]) [[NR_NUW:#[0-9]+]]
   // CHECK-EH-03-NEXT: unreachable
 
 }
 
 // CHECK-LABEL: define void @_Z5test3b
 X test3(bool B) {
-  // CHECK: tail call {{.*}} @_ZN1XC1Ev
+  // CHECK: call {{.*}} @_ZN1XC1Ev
   // CHECK-NOT: call {{.*}} @_ZN1XC1ERKS_
   // CHECK: call {{.*}} @_ZN1XC1Ev
   // CHECK: call {{.*}} @_ZN1XC1ERKS_
   if (B) {
     X y;
     return y;
   }
   // FIXME: we should NRVO this variable too.
   X x;
   return x;
 }
 
 extern "C" void exit(int) throw();
 
 // CHECK-LABEL: define void @_Z5test4b
 X test4(bool B) {
   {
-    // CHECK: tail call {{.*}} @_ZN1XC1Ev
+    // CHECK: call {{.*}} @_ZN1XC1Ev
     X x;
     // CHECK: br i1
     if (B)
       return x;
   }
-  // CHECK: tail call {{.*}} @_ZN1XD1Ev
-  // CHECK: tail call void @exit(i32 1)
+  // CHECK: call {{.*}} @_ZN1XD1Ev
+  // CHECK: call void @exit(i32 1)
   exit(1);
 }
 
 #ifdef __EXCEPTIONS
 // CHECK-EH-LABEL: define void @_Z5test5
 void may_throw();
 X test5() {
   try {
X test6() {
   // CHECK-NEXT: call {{.*}} @_ZN1XC1ERKS_([[X]]* {{%.*}}, [[X]]* nonnull dereferenceable({{[0-9]+}}) [[A]])
   // CHECK-NEXT: call {{.*}} @_ZN1XD1Ev([[X]]* nonnull [[A]])
   // CHECK-NEXT: call void @llvm.lifetime.end.p0i8(i64 1, i8* nonnull [[PTR]])
   // CHECK-NEXT: ret void
 }
 
 // CHECK-LABEL: define void @_Z5test7b
 X test7(bool b) {
-  // CHECK: tail call {{.*}} @_ZN1XC1Ev
+  // CHECK: call {{.*}} @_ZN1XC1Ev
   // CHECK-NEXT: ret
   if (b) {
     X x;
     return x;
   }
   return X();
 }
 
 // CHECK-LABEL: define void @_Z5test8b
 X test8(bool b) {
-  // CHECK: tail call {{.*}} @_ZN1XC1Ev
+  // CHECK: call {{.*}} @_ZN1XC1Ev
   // CHECK-NEXT: ret
   if (b) {
     X x;
     return x;
   } else {
     X y;
     return y;
   }
 }
 
 Y<int> test9() {
   Y<int>::f();
 }
 
 // CHECK-LABEL: define linkonce_odr void @_ZN1YIiE1fEv
-// CHECK: tail call {{.*}} @_ZN1YIiEC1Ev
+// CHECK: call {{.*}} @_ZN1YIiEC1Ev
 
 // CHECK-EH-03: attributes [[NR_NUW]] = { noreturn nounwind }

clang/test/CodeGenCXX/stack-reuse.cpp

-// RUN: %clang_cc1 -triple armv7-unknown-linux-gnueabihf %s -o - -emit-llvm -O1 | FileCheck %s
+// RUN: %clang_cc1 -triple armv7-unknown-linux-gnueabihf %s -o - -emit-llvm -O2 | FileCheck %s
 
 // Stack should be reused when possible, no need to allocate two separate slots
 // if they have disjoint lifetime.
 
 // Sizes of objects are related to previously existed threshold of 32.  In case
 // of S_large stack size is rounded to 40 bytes.
 
 // 32B

clang/test/CodeGenCXX/wasm-args-returns.cpp

 
 struct one_field {
   double d;
 };
 test(one_field);
 // CHECK: define double @_Z7forward9one_field(double returned %{{.*}})
 //
 // CHECK: define void @_Z14test_one_fieldv()
-// CHECK: %[[call:.*]] = tail call double @_Z13def_one_fieldv()
-// CHECK: tail call void @_Z3use9one_field(double %[[call]])
+// CHECK: %[[call:.*]] = call double @_Z13def_one_fieldv()
+// CHECK: call void @_Z3use9one_field(double %[[call]])
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use9one_field(double)
 // CHECK: declare double @_Z13def_one_fieldv()
 
 struct two_fields {
   double d, e;
 };

 // CHECK: declare void @_Z3use17aligned_copy_ctor(%struct.aligned_copy_ctor*)
 // CHECK: declare void @_Z21def_aligned_copy_ctorv(%struct.aligned_copy_ctor* sret)
 
 struct empty {};
 test(empty);
 // CHECK: define void @_Z7forward5empty()
 //
 // CHECK: define void @_Z10test_emptyv()
-// CHECK: tail call void @_Z9def_emptyv()
-// CHECK: tail call void @_Z3use5empty()
+// CHECK: call void @_Z9def_emptyv()
+// CHECK: call void @_Z3use5empty()
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use5empty()
 // CHECK: declare void @_Z9def_emptyv()
 
 struct one_bitfield {
   int d : 3;
 };
 test(one_bitfield);
 // CHECK: define i32 @_Z7forward12one_bitfield(i32 returned %{{.*}})
 //
 // CHECK: define void @_Z17test_one_bitfieldv()
-// CHECK: %[[call:.*]] = tail call i32 @_Z16def_one_bitfieldv()
-// CHECK: tail call void @_Z3use12one_bitfield(i32 %[[call]])
+// CHECK: %[[call:.*]] = call i32 @_Z16def_one_bitfieldv()
+// CHECK: call void @_Z3use12one_bitfield(i32 %[[call]])
 // CHECK: ret void
 //
 // CHECK: declare void @_Z3use12one_bitfield(i32)
 // CHECK: declare i32 @_Z16def_one_bitfieldv()

clang/test/CodeGenObjCXX/arc-blocks.mm

 
 // CHECK: [[EH_RESUME]]:
 // CHECK: resume { i8*, i32 }
 
 // CHECK: [[TERMINATE_LPAD]]:
 // CHECK: call void @__clang_call_terminate(
 
 // CHECK-O1-LABEL: define linkonce_odr hidden void @__copy_helper_block_ea8_32s40r48w56c15_ZTSN5test12S0E60c15_ZTSN5test12S0E(
-// CHECK-O1: tail call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
+// CHECK-O1: call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
 // CHECK-NOEXCP: define linkonce_odr hidden void @__copy_helper_block_8_32s40r48w56c15_ZTSN5test12S0E60c15_ZTSN5test12S0E(
 
 // CHECK: define linkonce_odr hidden void @__destroy_helper_block_ea8_32s40r48w56c15_ZTSN5test12S0E60c15_ZTSN5test12S0E(
 // CHECK: %[[BLOCK:.*]] = bitcast i8* %{{.*}} to <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>*
 // CHECK: %[[V4:.*]] = getelementptr inbounds <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>, <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>* %[[BLOCK]], i32 0, i32 5
 // CHECK: %[[V2:.*]] = getelementptr inbounds <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>, <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>* %[[BLOCK]], i32 0, i32 6
 // CHECK: %[[V3:.*]] = getelementptr inbounds <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>, <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>* %[[BLOCK]], i32 0, i32 7
 // CHECK: %[[V5:.*]] = getelementptr inbounds <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>, <{ i8*, i32, i32, i8*, %[[STRUCT_BLOCK_DESCRIPTOR]]*, i8*, i8*, i8*, %[[STRUCT_TEST1_S0]], %[[STRUCT_TEST1_S0]], %[[STRUCT_TRIVIAL_INTERNAL]] }>* %[[BLOCK]], i32 0, i32 8

 
 // CHECK: [[EH_RESUME]]:
 // CHECK: resume { i8*, i32 }
 
 // CHECK: [[TERMINATE_LPAD]]:
 // CHECK: call void @__clang_call_terminate(
 
 // CHECK-O1-LABEL: define linkonce_odr hidden void @__destroy_helper_block_ea8_32s40r48w56c15_ZTSN5test12S0E60c15_ZTSN5test12S0E(
-// CHECK-O1: tail call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
-// CHECK-O1: tail call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
+// CHECK-O1: call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
+// CHECK-O1: call void @llvm.objc.release({{.*}}) {{.*}} !clang.imprecise_release
 // CHECK-NOEXCP: define linkonce_odr hidden void @__destroy_helper_block_8_32s40r48w56c15_ZTSN5test12S0E60c15_ZTSN5test12S0E(
 
 namespace {
 struct TrivialInternal {
   int a;
 };
 }
 

clang/test/CodeGenObjCXX/nrvo.mm

 // RUN: %clang_cc1 -emit-llvm -o - -fblocks %s -O1 -fno-experimental-new-pass-manager -triple x86_64-apple-darwin10.0.0 -fobjc-runtime=macosx-fragile-10.5 | FileCheck %s
 
 // PR10835 / <rdar://problem/10050178>
 struct X {
   X();
   X(const X&);
   ~X();
 };
 
 @interface NRVO
 @end
 
 @implementation NRVO
 // CHECK: define internal void @"\01-[NRVO getNRVO]"
 - (X)getNRVO { 
   X x;
-  // CHECK: tail call void @_ZN1XC1Ev
+  // CHECK: call void @_ZN1XC1Ev
   // CHECK-NEXT: ret void
   return x;
 }
 @end
 
 X blocksNRVO() {
   return ^{
     // CHECK-LABEL: define internal void @___Z10blocksNRVOv_block_invoke
     X x;
-    // CHECK: tail call void @_ZN1XC1Ev
+    // CHECK: call void @_ZN1XC1Ev
     // CHECK-NEXT: ret void
     return x;
   }() ;
 }
 

clang/test/Lexer/minimize_source_to_dependency_directives_invalid_error.c

-// Test CF+LF are properly handled along with quoted, multi-line #error
-// RUN: %clang_cc1 -DOTHER -print-dependency-directives-minimized-source %s 2>&1 | FileCheck %s
-
-#ifndef TEST
-#error "message \
-   more message \
-   even more"
-#endif
-
-#ifdef OTHER
-#include <string>
-#endif
-
-// CHECK:      #ifdef OTHER
-// CHECK-NEXT: #include <string>
-// CHECK-NEXT: #endif
+// Test CF+LF are properly handled along with quoted, multi-line #error
+// RUN: %clang_cc1 -DOTHER -print-dependency-directives-minimized-source %s 2>&1 | FileCheck %s
+
+#ifndef TEST
+#error "message \
+   more message \
+   even more"
+#endif
+
+#ifdef OTHER
+#include <string>
+#endif
+
+// CHECK:      #ifdef OTHER
+// CHECK-NEXT: #include <string>
+// CHECK-NEXT: #endif

clang/test/PCH/no-escaping-block-tail-calls.cpp

-// RUN: %clang_cc1 -x c++-header -triple x86_64-apple-darwin11 -emit-pch -O1 -fblocks -fno-escaping-block-tail-calls -o %t %S/no-escaping-block-tail-calls.h
-// RUN: %clang_cc1 -triple x86_64-apple-darwin11 -include-pch %t -emit-llvm -O1 -fblocks -fno-escaping-block-tail-calls -o - %s | FileCheck %s
+// RUN: %clang_cc1 -x c++-header -triple x86_64-apple-darwin11 -emit-pch -O2 -fblocks -fno-escaping-block-tail-calls -o %t %S/no-escaping-block-tail-calls.h
+// RUN: %clang_cc1 -triple x86_64-apple-darwin11 -include-pch %t -emit-llvm -O2 -fblocks -fno-escaping-block-tail-calls -o - %s | FileCheck %s
 
 // Check that -fno-escaping-block-tail-calls doesn't disable tail-call
 // optimization if the block is non-escaping.
 
 // CHECK-LABEL: define internal i32 @___ZN1S1mEv_block_invoke(
 // CHECK: %[[CALL:.*]] = tail call i32 @_ZN1S3fooER2S0(
 // CHECK-NEXT: ret i32 %[[CALL]]
 
 void test() {
   S s;
   s.m();
 }

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/ambiguous_tail_call_seq1/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/ambiguous_tail_call_seq2/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/disambiguate_call_site/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/disambiguate_paths_to_common_sink/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/disambiguate_tail_call_seq/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/inlining_and_tail_calls/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/sbapi_support/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/thread_step_out_message/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/thread_step_out_or_return/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

lldb/packages/Python/lldbsuite/test/functionalities/tail_call_frames/unambiguous_sequence/Makefile

 CXX_SOURCES := main.cpp
 
-CXXFLAGS_EXTRAS := -g -O1 -glldb
+CXXFLAGS_EXTRAS := -g -O2 -glldb
 include Makefile.rules

llvm/include/llvm/Passes/PassBuilder.h

public:
   enum OptimizationLevel {
     /// Disable as many optimizations as possible. This doesn't completely
     /// disable the optimizer in all cases, for example always_inline functions
     /// can be required to be inlined for correctness.
     O0,
 
     /// Optimize quickly without destroying debuggability.
     ///
-    /// FIXME: The current and historical behavior of this level does *not*
-    /// agree with this goal, but we would like to move toward this goal in the
-    /// future.
-    ///
     /// This level is tuned to produce a result from the optimizer as quickly
     /// as possible and to avoid destroying debuggability. This tends to result
     /// in a very good development mode where the compiled code will be
     /// immediately executed as part of testing. As a consequence, where
     /// possible, we would like to produce efficient-to-execute code, but not
     /// if it significantly slows down compilation or would prevent even basic
     /// debugging of the resulting binary.
     ///
     /// As an example, complex loop transformations such as versioning,
-    /// vectorization, or fusion might not make sense here due to the degree to
-    /// which the executed code would differ from the source code, and the
-    /// potential compile time cost.
+    /// vectorization, or fusion don't make sense here due to the degree to
+    /// which the executed code differs from the source code, and the compile time
+    /// cost.
     O1,
 
     /// Optimize for fast execution as much as possible without triggering
     /// significant incremental compile time or code size growth.
     ///
     /// The key idea is that optimizations at this level should "pay for
     /// themselves". So if an optimization increases compile time by 5% or
     /// increases code size by 5% for a particular benchmark, that benchmark

llvm/lib/Passes/PassBuilder.cpp

 PassBuilder::buildFunctionSimplificationPipeline(OptimizationLevel Level,
                                                  ThinLTOPhase Phase,
                                                  bool DebugLogging) {
   assert(Level != O0 && "Must request optimizations!");
   FunctionPassManager FPM(DebugLogging);
 
   // Form SSA out of local memory accesses after breaking apart aggregates into
   // scalars.
   FPM.addPass(SROA());
 
   // Catch trivial redundancies
   FPM.addPass(EarlyCSEPass(true /* Enable mem-ssa. */));
 
   // Hoisting of scalars and load expressions.
+  if (Level > O1) {
     if (EnableGVNHoist)
       FPM.addPass(GVNHoistPass());
 
     // Global value numbering based sinking.
     if (EnableGVNSink) {
       FPM.addPass(GVNSinkPass());
       FPM.addPass(SimplifyCFGPass());
     }
+  }
 
   // Speculative execution if the target has divergent branches; otherwise nop.
+  if (Level > O1) {
     FPM.addPass(SpeculativeExecutionPass());
 
     // Optimize based on known information about branches, and cleanup afterward.
     FPM.addPass(JumpThreadingPass());
     FPM.addPass(CorrelatedValuePropagationPass());
+  }
   FPM.addPass(SimplifyCFGPass());
   if (Level == O3)
     FPM.addPass(AggressiveInstCombinePass());
   FPM.addPass(InstCombinePass());
 
   if (!isOptimizingForSize(Level))
     FPM.addPass(LibCallsShrinkWrapPass());
 
   invokePeepholeEPCallbacks(FPM, Level);
 
   // For PGO use pipeline, try to optimize memory intrinsics such as memcpy
   // using the size value profile. Don't perform this when optimizing for size.
   if (PGOOpt && PGOOpt->Action == PGOOptions::IRUse &&
-      !isOptimizingForSize(Level))
+      !isOptimizingForSize(Level) && Level > O1)
     FPM.addPass(PGOMemOPSizeOpt());
 
+  // TODO: Investigate the cost/benefit of tail call elimination on debugging.
+  if (Level > O1)
     FPM.addPass(TailCallElimPass());
   FPM.addPass(SimplifyCFGPass());
 
   // Form canonically associated expression trees, and simplify the trees using
   // basic mathematical properties. For example, this will form (nearly)
   // minimal multiplication trees.
   FPM.addPass(ReassociatePass());
 
   // Add the primary loop simplification pipeline.
PassBuilder::buildFunctionSimplificationPipeline(OptimizationLevel Level,
   // Simplify the loop body. We do this initially to clean up after other loop
   // passes run, either when iterating on a loop or on inner loops with
   // implications on the outer loop.
   LPM1.addPass(LoopInstSimplifyPass());
   LPM1.addPass(LoopSimplifyCFGPass());
 
   // Rotate Loop - disable header duplication at -Oz
   LPM1.addPass(LoopRotatePass(Level != Oz));
+  // TODO: Investigate promotion cap for O1.
   LPM1.addPass(LICMPass(PTO.LicmMssaOptCap, PTO.LicmMssaNoAccForPromotionCap));
   LPM1.addPass(SimpleLoopUnswitchPass());
   LPM2.addPass(IndVarSimplifyPass());
   LPM2.addPass(LoopIdiomRecognizePass());
 
   for (auto &C : LateLoopOptimizationsEPCallbacks)
     C(LPM2, Level);
 
PassBuilder::buildFunctionSimplificationPipeline(OptimizationLevel Level,
 
   // Run instcombine after redundancy and dead bit elimination to exploit
   // opportunities opened up by them.
   FPM.addPass(InstCombinePass());
   invokePeepholeEPCallbacks(FPM, Level);
 
   // Re-consider control flow based optimizations after redundancy elimination,
   // redo DCE, etc.
+  if (Level > O1) {
     FPM.addPass(JumpThreadingPass());
     FPM.addPass(CorrelatedValuePropagationPass());
     FPM.addPass(DSEPass());
     FPM.addPass(createFunctionToLoopPassAdaptor(
         LICMPass(PTO.LicmMssaOptCap, PTO.LicmMssaNoAccForPromotionCap),
         EnableMSSALoopDependency, DebugLogging));
+  }
 
   for (auto &C : ScalarOptimizerLateEPCallbacks)
     C(FPM, Level);
 
   // Finally, do an expensive DCE pass to catch all the dead code exposed by
   // the simplifications and basic cleanup after all the simplifications.
+  // TODO: Investigate if this is too expensive.
   FPM.addPass(ADCEPass());
   FPM.addPass(SimplifyCFGPass());
   FPM.addPass(InstCombinePass());
   invokePeepholeEPCallbacks(FPM, Level);
 
   if (EnableCHR && Level == O3 && PGOOpt &&
       (PGOOpt->Action == PGOOptions::IRUse ||
        PGOOpt->Action == PGOOptions::SampleUse))

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

void PassManagerBuilder::populateFunctionPassManager(
   if (LibraryInfo)
     FPM.add(new TargetLibraryInfoWrapperPass(*LibraryInfo));
 
   if (OptLevel == 0) return;
 
   addInitialAliasAnalysisPasses(FPM);
 
   FPM.add(createCFGSimplificationPass());
   FPM.add(createSROAPass());
   FPM.add(createEarlyCSEPass());
   FPM.add(createLowerExpectIntrinsicPass());
 }
 
 // Do PGO instrumentation generation or use pass as the option specified.
 void PassManagerBuilder::addPGOInstrPasses(legacy::PassManagerBase &MPM,
                                            bool IsCS = false) {
   if (IsCS) {
if (OptLevel > 0 && SizeLevel == 0 && !DisablePreInliner &&
     // care about are DefaultThreshold and HintThreshold.
     InlineParams IP;
     IP.DefaultThreshold = PreInlineThreshold;
     // FIXME: The hint threshold has the same value used by the regular inliner.
     // This should probably be lowered after performance testing.
     IP.HintThreshold = 325;
 
     MPM.add(createFunctionInliningPass(IP));
     MPM.add(createSROAPass());
     MPM.add(createEarlyCSEPass());             // Catch trivial redundancies
     MPM.add(createCFGSimplificationPass());    // Merge & remove BBs
     MPM.add(createInstructionCombiningPass()); // Combine silly seq's
     addExtensionsToPM(EP_Peephole, MPM);
   }
   if ((EnablePGOInstrGen && !IsCS) || (EnablePGOCSInstrGen && IsCS)) {
     MPM.add(createPGOInstrumentationGenLegacyPass(IsCS));
     // Add the profile lowering pass.
void PassManagerBuilder::addPGOInstrPasses(legacy::PassManagerBase &MPM,
   if (OptLevel > 0 && !IsCS)
     MPM.add(
         createPGOIndirectCallPromotionLegacyPass(false, !PGOSampleUse.empty()));
 }
 void PassManagerBuilder::addFunctionSimplificationPasses(
     legacy::PassManagerBase &MPM) {
   // Start of function pass.
   // Break up aggregate allocas, using SSAUpdater.
+  assert(OptLevel >= 1 && "Calling function optimizer with no optimization level!");
   MPM.add(createSROAPass());
   MPM.add(createEarlyCSEPass(true /* Enable mem-ssa. */)); // Catch trivial redundancies
+
+  if (OptLevel > 1) {
     if (EnableGVNHoist)
       MPM.add(createGVNHoistPass());
     if (EnableGVNSink) {
       MPM.add(createGVNSinkPass());
       MPM.add(createCFGSimplificationPass());
     }
+  }
 
+  if (OptLevel > 1) {
     // Speculative execution if the target has divergent branches; otherwise nop.
     MPM.add(createSpeculativeExecutionIfHasBranchDivergencePass());
+
     MPM.add(createJumpThreadingPass());         // Thread jumps.
     MPM.add(createCorrelatedValuePropagationPass()); // Propagate conditionals
+  }
   MPM.add(createCFGSimplificationPass());     // Merge & remove BBs
   // Combine silly seq's
   if (OptLevel > 2)
     MPM.add(createAggressiveInstCombinerPass());
   addInstructionCombiningPass(MPM);
   if (SizeLevel == 0 && !DisableLibCallsShrinkWrap)
     MPM.add(createLibCallsShrinkWrapPass());
   addExtensionsToPM(EP_Peephole, MPM);
 
   // Optimize memory intrinsic calls based on the profiled size information.
   if (SizeLevel == 0)
     MPM.add(createPGOMemOPSizeOptLegacyPass());
 
+  // TODO: Investigate the cost/benefit of tail call elimination on debugging.
+  if (OptLevel > 1)
     MPM.add(createTailCallEliminationPass()); // Eliminate tail calls
-  MPM.add(createCFGSimplificationPass());     // Merge & remove BBs
+  MPM.add(createCFGSimplificationPass());      // Merge & remove BBs
   MPM.add(createReassociatePass());           // Reassociate expressions
 
   // Begin the loop pass pipeline.
   if (EnableSimpleLoopUnswitch) {
     // The simple loop unswitch pass relies on separate cleanup passes. Schedule
     // them first so when we re-process a loop they run before other loop
     // passes.
     MPM.add(createLoopInstSimplifyPass());
     MPM.add(createLoopSimplifyCFGPass());
   }
   // Rotate Loop - disable header duplication at -Oz
   MPM.add(createLoopRotatePass(SizeLevel == 2 ? 0 : -1));
+  // TODO: Investigate promotion cap for O1.
   MPM.add(createLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap));
   if (EnableSimpleLoopUnswitch)
     MPM.add(createSimpleLoopUnswitchLegacyPass());
   else
     MPM.add(createLoopUnswitchPass(SizeLevel || OptLevel < 3, DivergentTarget));
   // FIXME: We break the loop pass pipeline here in order to do full
   // simplify-cfg. Eventually loop-simplifycfg should be enhanced to replace the
   // need for this.
void PassManagerBuilder::addFunctionSimplificationPasses(
   // computations, and then ADCE will run later to exploit any new DCE
   // opportunities that creates).
   MPM.add(createBitTrackingDCEPass());        // Delete dead bit computations
 
   // Run instcombine after redundancy elimination to exploit opportunities
   // opened up by them.
   addInstructionCombiningPass(MPM);
   addExtensionsToPM(EP_Peephole, MPM);
+  if (OptLevel > 1) {
     MPM.add(createJumpThreadingPass());         // Thread jumps
     MPM.add(createCorrelatedValuePropagationPass());
     MPM.add(createDeadStoreEliminationPass());  // Delete dead stores
     MPM.add(createLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap));
+  }
 
   addExtensionsToPM(EP_ScalarOptimizerLate, MPM);
 
   if (RerollLoops)
     MPM.add(createLoopRerollPass());
 
+  // TODO: Investigate if this is too expensive at O1.
   MPM.add(createAggressiveDCEPass());         // Delete dead instructions
   MPM.add(createCFGSimplificationPass()); // Merge & remove BBs
   // Clean up after everything.
   addInstructionCombiningPass(MPM);
   addExtensionsToPM(EP_Peephole, MPM);
 
   if (EnableCHR && OptLevel >= 3 &&
       (!PGOInstrUse.empty() || !PGOSampleUse.empty() || EnablePGOCSInstrGen))
void PassManagerBuilder::addLTOOptimizationPasses(legacy::PassManagerBase &PM) {
   addExtensionsToPM(EP_Peephole, PM);
   PM.add(createJumpThreadingPass());
 
   // Break up allocas
   PM.add(createSROAPass());
 
   // LTO provides additional opportunities for tailcall elimination due to
   // link-time inlining, and visibility of nocapture attribute.
+  if (OptLevel > 1)
     PM.add(createTailCallEliminationPass());
 
   // Infer attributes on declarations, call sites, arguments, etc.
   PM.add(createPostOrderFunctionAttrsLegacyPass()); // Add nocapture.
   // Run a few AA driven optimizations here and now, to cleanup the code.
   PM.add(createGlobalsAAWrapperPass()); // IP alias analysis.
 
   PM.add(createLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap));
   PM.add(createMergedLoadStoreMotionPass()); // Merge ld/st in diamonds.

llvm/test/CodeGen/AMDGPU/simplify-libcalls.ll

 ; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-simplify-libcall < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-POSTLINK %s
 ; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-simplify-libcall -amdgpu-prelink  <%s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-PRELINK %s
 ; RUN: opt -S -O1 -mtriple=amdgcn-- -amdgpu-use-native -amdgpu-prelink < %s | FileCheck -enable-var-scope -check-prefix=GCN -check-prefix=GCN-NATIVE %s
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos
-; GCN-POSTLINK: tail call fast float @_Z3sinf(
-; GCN-POSTLINK: tail call fast float @_Z3cosf(
+; GCN-POSTLINK: call fast float @_Z3sinf(
+; GCN-POSTLINK: call fast float @_Z3cosf(
 ; GCN-PRELINK: call fast float @_Z6sincosfPf(
-; GCN-NATIVE: tail call fast float @_Z10native_sinf(
-; GCN-NATIVE: tail call fast float @_Z10native_cosf(
+; GCN-NATIVE: call fast float @_Z10native_sinf(
+; GCN-NATIVE: call fast float @_Z10native_cosf(
 define amdgpu_kernel void @test_sincos(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3sinf(float %tmp)
+  %call = call fast float @_Z3sinf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
-  %call2 = tail call fast float @_Z3cosf(float %tmp)
+  %call2 = call fast float @_Z3cosf(float %tmp)
   %arrayidx3 = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   store float %call2, float addrspace(1)* %arrayidx3, align 4
   ret void
 }
 
 declare float @_Z3sinf(float)
 
 declare float @_Z3cosf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos_v2
-; GCN-POSTLINK: tail call fast <2 x float> @_Z3sinDv2_f(
-; GCN-POSTLINK: tail call fast <2 x float> @_Z3cosDv2_f(
+; GCN-POSTLINK: call fast <2 x float> @_Z3sinDv2_f(
+; GCN-POSTLINK: call fast <2 x float> @_Z3cosDv2_f(
 ; GCN-PRELINK: call fast <2 x float> @_Z6sincosDv2_fPS_(
-; GCN-NATIVE: tail call fast <2 x float> @_Z10native_sinDv2_f(
-; GCN-NATIVE: tail call fast <2 x float> @_Z10native_cosDv2_f(
+; GCN-NATIVE: call fast <2 x float> @_Z10native_sinDv2_f(
+; GCN-NATIVE: call fast <2 x float> @_Z10native_cosDv2_f(
 define amdgpu_kernel void @test_sincos_v2(<2 x float> addrspace(1)* nocapture %a) {
 entry:
   %tmp = load <2 x float>, <2 x float> addrspace(1)* %a, align 8
-  %call = tail call fast <2 x float> @_Z3sinDv2_f(<2 x float> %tmp)
+  %call = call fast <2 x float> @_Z3sinDv2_f(<2 x float> %tmp)
   store <2 x float> %call, <2 x float> addrspace(1)* %a, align 8
-  %call2 = tail call fast <2 x float> @_Z3cosDv2_f(<2 x float> %tmp)
+  %call2 = call fast <2 x float> @_Z3cosDv2_f(<2 x float> %tmp)
   %arrayidx3 = getelementptr inbounds <2 x float>, <2 x float> addrspace(1)* %a, i64 1
   store <2 x float> %call2, <2 x float> addrspace(1)* %arrayidx3, align 8
   ret void
 }
 
 declare <2 x float> @_Z3sinDv2_f(<2 x float>)
 
 declare <2 x float> @_Z3cosDv2_f(<2 x float>)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos_v3
-; GCN-POSTLINK: tail call fast <3 x float> @_Z3sinDv3_f(
-; GCN-POSTLINK: tail call fast <3 x float> @_Z3cosDv3_f(
+; GCN-POSTLINK: call fast <3 x float> @_Z3sinDv3_f(
+; GCN-POSTLINK: call fast <3 x float> @_Z3cosDv3_f(
 ; GCN-PRELINK: call fast <3 x float> @_Z6sincosDv3_fPS_(
-; GCN-NATIVE: tail call fast <3 x float> @_Z10native_sinDv3_f(
-; GCN-NATIVE: tail call fast <3 x float> @_Z10native_cosDv3_f(
+; GCN-NATIVE: call fast <3 x float> @_Z10native_sinDv3_f(
+; GCN-NATIVE: call fast <3 x float> @_Z10native_cosDv3_f(
 define amdgpu_kernel void @test_sincos_v3(<3 x float> addrspace(1)* nocapture %a) {
 entry:
   %castToVec4 = bitcast <3 x float> addrspace(1)* %a to <4 x float> addrspace(1)*
   %loadVec4 = load <4 x float>, <4 x float> addrspace(1)* %castToVec4, align 16
   %extractVec4 = shufflevector <4 x float> %loadVec4, <4 x float> undef, <3 x i32> <i32 0, i32 1, i32 2>
-  %call = tail call fast <3 x float> @_Z3sinDv3_f(<3 x float> %extractVec4)
+  %call = call fast <3 x float> @_Z3sinDv3_f(<3 x float> %extractVec4)
   %extractVec6 = shufflevector <3 x float> %call, <3 x float> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 undef>
   store <4 x float> %extractVec6, <4 x float> addrspace(1)* %castToVec4, align 16
-  %call11 = tail call fast <3 x float> @_Z3cosDv3_f(<3 x float> %extractVec4)
+  %call11 = call fast <3 x float> @_Z3cosDv3_f(<3 x float> %extractVec4)
   %arrayidx12 = getelementptr inbounds <3 x float>, <3 x float> addrspace(1)* %a, i64 1
   %extractVec13 = shufflevector <3 x float> %call11, <3 x float> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 undef>
   %storetmp14 = bitcast <3 x float> addrspace(1)* %arrayidx12 to <4 x float> addrspace(1)*
   store <4 x float> %extractVec13, <4 x float> addrspace(1)* %storetmp14, align 16
   ret void
 }
 
 declare <3 x float> @_Z3sinDv3_f(<3 x float>)
 
 declare <3 x float> @_Z3cosDv3_f(<3 x float>)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos_v4
-; GCN-POSTLINK: tail call fast <4 x float> @_Z3sinDv4_f(
-; GCN-POSTLINK: tail call fast <4 x float> @_Z3cosDv4_f(
+; GCN-POSTLINK: call fast <4 x float> @_Z3sinDv4_f(
+; GCN-POSTLINK: call fast <4 x float> @_Z3cosDv4_f(
 ; GCN-PRELINK: call fast <4 x float> @_Z6sincosDv4_fPS_(
-; GCN-NATIVE: tail call fast <4 x float> @_Z10native_sinDv4_f(
-; GCN-NATIVE: tail call fast <4 x float> @_Z10native_cosDv4_f(
+; GCN-NATIVE: call fast <4 x float> @_Z10native_sinDv4_f(
+; GCN-NATIVE: call fast <4 x float> @_Z10native_cosDv4_f(
 define amdgpu_kernel void @test_sincos_v4(<4 x float> addrspace(1)* nocapture %a) {
 entry:
   %tmp = load <4 x float>, <4 x float> addrspace(1)* %a, align 16
-  %call = tail call fast <4 x float> @_Z3sinDv4_f(<4 x float> %tmp)
+  %call = call fast <4 x float> @_Z3sinDv4_f(<4 x float> %tmp)
   store <4 x float> %call, <4 x float> addrspace(1)* %a, align 16
-  %call2 = tail call fast <4 x float> @_Z3cosDv4_f(<4 x float> %tmp)
+  %call2 = call fast <4 x float> @_Z3cosDv4_f(<4 x float> %tmp)
   %arrayidx3 = getelementptr inbounds <4 x float>, <4 x float> addrspace(1)* %a, i64 1
   store <4 x float> %call2, <4 x float> addrspace(1)* %arrayidx3, align 16
   ret void
 }
 
 declare <4 x float> @_Z3sinDv4_f(<4 x float>)
 
 declare <4 x float> @_Z3cosDv4_f(<4 x float>)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos_v8
-; GCN-POSTLINK: tail call fast <8 x float> @_Z3sinDv8_f(
-; GCN-POSTLINK: tail call fast <8 x float> @_Z3cosDv8_f(
+; GCN-POSTLINK: call fast <8 x float> @_Z3sinDv8_f(
+; GCN-POSTLINK: call fast <8 x float> @_Z3cosDv8_f(
 ; GCN-PRELINK: call fast <8 x float> @_Z6sincosDv8_fPS_(
-; GCN-NATIVE: tail call fast <8 x float> @_Z10native_sinDv8_f(
-; GCN-NATIVE: tail call fast <8 x float> @_Z10native_cosDv8_f(
+; GCN-NATIVE: call fast <8 x float> @_Z10native_sinDv8_f(
+; GCN-NATIVE: call fast <8 x float> @_Z10native_cosDv8_f(
 define amdgpu_kernel void @test_sincos_v8(<8 x float> addrspace(1)* nocapture %a) {
 entry:
   %tmp = load <8 x float>, <8 x float> addrspace(1)* %a, align 32
-  %call = tail call fast <8 x float> @_Z3sinDv8_f(<8 x float> %tmp)
+  %call = call fast <8 x float> @_Z3sinDv8_f(<8 x float> %tmp)
   store <8 x float> %call, <8 x float> addrspace(1)* %a, align 32
-  %call2 = tail call fast <8 x float> @_Z3cosDv8_f(<8 x float> %tmp)
+  %call2 = call fast <8 x float> @_Z3cosDv8_f(<8 x float> %tmp)
   %arrayidx3 = getelementptr inbounds <8 x float>, <8 x float> addrspace(1)* %a, i64 1
   store <8 x float> %call2, <8 x float> addrspace(1)* %arrayidx3, align 32
   ret void
 }
 
 declare <8 x float> @_Z3sinDv8_f(<8 x float>)
 
 declare <8 x float> @_Z3cosDv8_f(<8 x float>)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_sincos_v16
-; GCN-POSTLINK: tail call fast <16 x float> @_Z3sinDv16_f(
-; GCN-POSTLINK: tail call fast <16 x float> @_Z3cosDv16_f(
+; GCN-POSTLINK: call fast <16 x float> @_Z3sinDv16_f(
+; GCN-POSTLINK: call fast <16 x float> @_Z3cosDv16_f(
 ; GCN-PRELINK: call fast <16 x float> @_Z6sincosDv16_fPS_(
-; GCN-NATIVE: tail call fast <16 x float> @_Z10native_sinDv16_f(
-; GCN-NATIVE: tail call fast <16 x float> @_Z10native_cosDv16_f(
+; GCN-NATIVE: call fast <16 x float> @_Z10native_sinDv16_f(
+; GCN-NATIVE: call fast <16 x float> @_Z10native_cosDv16_f(
 define amdgpu_kernel void @test_sincos_v16(<16 x float> addrspace(1)* nocapture %a) {
 entry:
   %tmp = load <16 x float>, <16 x float> addrspace(1)* %a, align 64
-  %call = tail call fast <16 x float> @_Z3sinDv16_f(<16 x float> %tmp)
+  %call = call fast <16 x float> @_Z3sinDv16_f(<16 x float> %tmp)
   store <16 x float> %call, <16 x float> addrspace(1)* %a, align 64
-  %call2 = tail call fast <16 x float> @_Z3cosDv16_f(<16 x float> %tmp)
+  %call2 = call fast <16 x float> @_Z3cosDv16_f(<16 x float> %tmp)
   %arrayidx3 = getelementptr inbounds <16 x float>, <16 x float> addrspace(1)* %a, i64 1
   store <16 x float> %call2, <16 x float> addrspace(1)* %arrayidx3, align 64
   ret void
 }
 
 declare <16 x float> @_Z3sinDv16_f(<16 x float>)
 
 declare <16 x float> @_Z3cosDv16_f(<16 x float>)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_native_recip
 ; GCN: store float 0x3FD5555560000000, float addrspace(1)* %a
 define amdgpu_kernel void @test_native_recip(float addrspace(1)* nocapture %a) {
 entry:
-  %call = tail call fast float @_Z12native_recipf(float 3.000000e+00)
+  %call = call fast float @_Z12native_recipf(float 3.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z12native_recipf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_half_recip
 ; GCN: store float 0x3FD5555560000000, float addrspace(1)* %a
 define amdgpu_kernel void @test_half_recip(float addrspace(1)* nocapture %a) {
 entry:
-  %call = tail call fast float @_Z10half_recipf(float 3.000000e+00)
+  %call = call fast float @_Z10half_recipf(float 3.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z10half_recipf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_native_divide
 ; GCN: fmul fast float %tmp, 0x3FD5555560000000
 define amdgpu_kernel void @test_native_divide(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z13native_divideff(float %tmp, float 3.000000e+00)
+  %call = call fast float @_Z13native_divideff(float %tmp, float 3.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z13native_divideff(float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_half_divide
 ; GCN: fmul fast float %tmp, 0x3FD5555560000000
 define amdgpu_kernel void @test_half_divide(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z11half_divideff(float %tmp, float 3.000000e+00)
+  %call = call fast float @_Z11half_divideff(float %tmp, float 3.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z11half_divideff(float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_0f
 ; GCN: store float 1.000000e+00, float addrspace(1)* %a
 define amdgpu_kernel void @test_pow_0f(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 0.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 0.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3powff(float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_0i
 ; GCN: store float 1.000000e+00, float addrspace(1)* %a
 define amdgpu_kernel void @test_pow_0i(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 0.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 0.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_1f
 ; GCN: %tmp = load float, float addrspace(1)* %arrayidx, align 4
 ; GCN: store float %tmp, float addrspace(1)* %a, align 4
 define amdgpu_kernel void @test_pow_1f(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 1.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 1.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_1i
 ; GCN: %tmp = load float, float addrspace(1)* %arrayidx, align 4
 ; GCN: store float %tmp, float addrspace(1)* %a, align 4
 define amdgpu_kernel void @test_pow_1i(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 1.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 1.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_2f
 ; GCN: %tmp = load float, float addrspace(1)* %a, align 4
 ; GCN: %__pow2 = fmul fast float %tmp, %tmp
 define amdgpu_kernel void @test_pow_2f(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 2.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 2.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_2i
 ; GCN: %tmp = load float, float addrspace(1)* %a, align 4
 ; GCN: %__pow2 = fmul fast float %tmp, %tmp
 define amdgpu_kernel void @test_pow_2i(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 2.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float 2.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_m1f
 ; GCN: %tmp = load float, float addrspace(1)* %arrayidx, align 4
 ; GCN: %__powrecip = fdiv fast float 1.000000e+00, %tmp
 define amdgpu_kernel void @test_pow_m1f(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float -1.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float -1.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_m1i
 ; GCN: %tmp = load float, float addrspace(1)* %arrayidx, align 4
 ; GCN: %__powrecip = fdiv fast float 1.000000e+00, %tmp
 define amdgpu_kernel void @test_pow_m1i(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float -1.000000e+00)
+  %call = call fast float @_Z3powff(float %tmp, float -1.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_half
-; GCN-POSTLINK: tail call fast float @_Z3powff(float %tmp, float 5.000000e-01)
-; GCN-PRELINK: %__pow2sqrt = tail call fast float @_Z4sqrtf(float %tmp)
+; GCN-POSTLINK: call fast float @_Z3powff(float %tmp, float 5.000000e-01)
+; GCN-PRELINK: %__pow2sqrt = call fast float @_Z4sqrtf(float %tmp)
 define amdgpu_kernel void @test_pow_half(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 5.000000e-01)
+  %call = call fast float @_Z3powff(float %tmp, float 5.000000e-01)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_mhalf
-; GCN-POSTLINK: tail call fast float @_Z3powff(float %tmp, float -5.000000e-01)
-; GCN-PRELINK: %__pow2rsqrt = tail call fast float @_Z5rsqrtf(float %tmp)
+; GCN-POSTLINK: call fast float @_Z3powff(float %tmp, float -5.000000e-01)
+; GCN-PRELINK: %__pow2rsqrt = call fast float @_Z5rsqrtf(float %tmp)
 define amdgpu_kernel void @test_pow_mhalf(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float -5.000000e-01)
+  %call = call fast float @_Z3powff(float %tmp, float -5.000000e-01)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow_c
 ; GCN: %__powx2 = fmul fast float %tmp, %tmp
 ; GCN: %__powx21 = fmul fast float %__powx2, %__powx2
 ; GCN: %__powx22 = fmul fast float %__powx2, %tmp
 ; GCN: %[[r0:.*]] = fmul fast float %__powx21, %__powx21
 ; GCN: %__powprod3 = fmul fast float %[[r0]], %__powx22
 define amdgpu_kernel void @test_pow_c(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 1.100000e+01)
+  %call = call fast float @_Z3powff(float %tmp, float 1.100000e+01)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_powr_c
 ; GCN: %__powx2 = fmul fast float %tmp, %tmp
 ; GCN: %__powx21 = fmul fast float %__powx2, %__powx2
 ; GCN: %__powx22 = fmul fast float %__powx2, %tmp
 ; GCN: %[[r0:.*]] = fmul fast float %__powx21, %__powx21
 ; GCN: %__powprod3 = fmul fast float %[[r0]], %__powx22
 define amdgpu_kernel void @test_powr_c(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z4powrff(float %tmp, float 1.100000e+01)
+  %call = call fast float @_Z4powrff(float %tmp, float 1.100000e+01)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z4powrff(float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pown_c
 ; GCN: %__powx2 = fmul fast float %tmp, %tmp
 ; GCN: %__powx21 = fmul fast float %__powx2, %__powx2
 ; GCN: %__powx22 = fmul fast float %__powx2, %tmp
 ; GCN: %[[r0:.*]] = fmul fast float %__powx21, %__powx21
 ; GCN: %__powprod3 = fmul fast float %[[r0]], %__powx22
 define amdgpu_kernel void @test_pown_c(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z4pownfi(float %tmp, i32 11)
+  %call = call fast float @_Z4pownfi(float %tmp, i32 11)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z4pownfi(float, i32)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pow
-; GCN-POSTLINK: tail call fast float @_Z3powff(float %tmp, float 1.013000e+03)
-; GCN-PRELINK: %__fabs = tail call fast float @_Z4fabsf(float %tmp)
-; GCN-PRELINK: %__log2 = tail call fast float @_Z4log2f(float %__fabs)
+; GCN-POSTLINK: call fast float @_Z3powff(float %tmp, float 1.013000e+03)
+; GCN-PRELINK: %__fabs = call fast float @_Z4fabsf(float %tmp)
+; GCN-PRELINK: %__log2 = call fast float @_Z4log2f(float %__fabs)
 ; GCN-PRELINK: %__ylogx = fmul fast float %__log2, 1.013000e+03
-; GCN-PRELINK: %__exp2 = tail call fast float @_Z4exp2f(float %__ylogx)
+; GCN-PRELINK: %__exp2 = call fast float @_Z4exp2f(float %__ylogx)
 ; GCN-PRELINK: %[[r0:.*]] = bitcast float %tmp to i32
 ; GCN-PRELINK: %__pow_sign = and i32 %[[r0]], -2147483648
 ; GCN-PRELINK: %[[r1:.*]] = bitcast float %__exp2 to i32
 ; GCN-PRELINK: %[[r2:.*]] = or i32 %__pow_sign, %[[r1]]
 ; GCN-PRELINK: %[[r3:.*]] = bitcast float addrspace(1)* %a to i32 addrspace(1)*
 ; GCN-PRELINK: store i32 %[[r2]], i32 addrspace(1)* %[[r3]], align 4
 define amdgpu_kernel void @test_pow(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3powff(float %tmp, float 1.013000e+03)
+  %call = call fast float @_Z3powff(float %tmp, float 1.013000e+03)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_powr
-; GCN-POSTLINK: tail call fast float @_Z4powrff(float %tmp, float %tmp1)
-; GCN-PRELINK: %__log2 = tail call fast float @_Z4log2f(float %tmp)
+; GCN-POSTLINK: call fast float @_Z4powrff(float %tmp, float %tmp1)
+; GCN-PRELINK: %__log2 = call fast float @_Z4log2f(float %tmp)
 ; GCN-PRELINK: %__ylogx = fmul fast float %__log2, %tmp1
-; GCN-PRELINK: %__exp2 = tail call fast float @_Z4exp2f(float %__ylogx)
+; GCN-PRELINK: %__exp2 = call fast float @_Z4exp2f(float %__ylogx)
 ; GCN-PRELINK: store float %__exp2, float addrspace(1)* %a, align 4
-; GCN-NATIVE:  %__log2 = tail call fast float @_Z11native_log2f(float %tmp)
+; GCN-NATIVE:  %__log2 = call fast float @_Z11native_log2f(float %tmp)
 ; GCN-NATIVE:  %__ylogx = fmul fast float %__log2, %tmp1
-; GCN-NATIVE:  %__exp2 = tail call fast float @_Z11native_exp2f(float %__ylogx)
+; GCN-NATIVE:  %__exp2 = call fast float @_Z11native_exp2f(float %__ylogx)
 ; GCN-NATIVE:  store float %__exp2, float addrspace(1)* %a, align 4
 define amdgpu_kernel void @test_powr(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
   %arrayidx1 = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp1 = load float, float addrspace(1)* %arrayidx1, align 4
-  %call = tail call fast float @_Z4powrff(float %tmp, float %tmp1)
+  %call = call fast float @_Z4powrff(float %tmp, float %tmp1)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_pown
-; GCN-POSTLINK: tail call fast float @_Z4pownfi(float %tmp, i32 %conv)
+; GCN-POSTLINK: call fast float @_Z4pownfi(float %tmp, i32 %conv)
 ; GCN-PRELINK: %conv = fptosi float %tmp1 to i32
-; GCN-PRELINK: %__fabs = tail call fast float @_Z4fabsf(float %tmp)
-; GCN-PRELINK: %__log2 = tail call fast float @_Z4log2f(float %__fabs)
+; GCN-PRELINK: %__fabs = call fast float @_Z4fabsf(float %tmp)
+; GCN-PRELINK: %__log2 = call fast float @_Z4log2f(float %__fabs)
 ; GCN-PRELINK: %pownI2F = sitofp i32 %conv to float
 ; GCN-PRELINK: %__ylogx = fmul fast float %__log2, %pownI2F
-; GCN-PRELINK: %__exp2 = tail call fast float @_Z4exp2f(float %__ylogx)
+; GCN-PRELINK: %__exp2 = call fast float @_Z4exp2f(float %__ylogx)
 ; GCN-PRELINK: %__yeven = shl i32 %conv, 31
 ; GCN-PRELINK: %[[r0:.*]] = bitcast float %tmp to i32
 ; GCN-PRELINK: %__pow_sign = and i32 %__yeven, %[[r0]]
 ; GCN-PRELINK: %[[r1:.*]] = bitcast float %__exp2 to i32
 ; GCN-PRELINK: %[[r2:.*]] = or i32 %__pow_sign, %[[r1]]
 ; GCN-PRELINK: %[[r3:.*]] = bitcast float addrspace(1)* %a to i32 addrspace(1)*
 ; GCN-PRELINK: store i32 %[[r2]], i32 addrspace(1)* %[[r3]], align 4
 define amdgpu_kernel void @test_pown(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
   %arrayidx1 = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp1 = load float, float addrspace(1)* %arrayidx1, align 4
   %conv = fptosi float %tmp1 to i32
-  %call = tail call fast float @_Z4pownfi(float %tmp, i32 %conv)
+  %call = call fast float @_Z4pownfi(float %tmp, i32 %conv)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_rootn_1
 ; GCN: %tmp = load float, float addrspace(1)* %arrayidx, align 4
 ; GCN: store float %tmp, float addrspace(1)* %a, align 4
 define amdgpu_kernel void @test_rootn_1(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
-  %call = tail call fast float @_Z5rootnfi(float %tmp, i32 1)
+  %call = call fast float @_Z5rootnfi(float %tmp, i32 1)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z5rootnfi(float, i32)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_rootn_2
-; GCN-POSTLINK: tail call fast float @_Z5rootnfi(float %tmp, i32 2)
-; GCN-PRELINK: %__rootn2sqrt = tail call fast float @_Z4sqrtf(float %tmp)
+; GCN-POSTLINK: call fast float @_Z5rootnfi(float %tmp, i32 2)
+; GCN-PRELINK: %__rootn2sqrt = call fast float @_Z4sqrtf(float %tmp)
 define amdgpu_kernel void @test_rootn_2(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5rootnfi(float %tmp, i32 2)
+  %call = call fast float @_Z5rootnfi(float %tmp, i32 2)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_rootn_3
-; GCN-POSTLINK: tail call fast float @_Z5rootnfi(float %tmp, i32 3)
-; GCN-PRELINK: %__rootn2cbrt = tail call fast float @_Z4cbrtf(float %tmp)
+; GCN-POSTLINK: call fast float @_Z5rootnfi(float %tmp, i32 3)
+; GCN-PRELINK: %__rootn2cbrt = call fast float @_Z4cbrtf(float %tmp)
 define amdgpu_kernel void @test_rootn_3(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5rootnfi(float %tmp, i32 3)
+  %call = call fast float @_Z5rootnfi(float %tmp, i32 3)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_rootn_m1
 ; GCN: fdiv fast float 1.000000e+00, %tmp
 define amdgpu_kernel void @test_rootn_m1(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5rootnfi(float %tmp, i32 -1)
+  %call = call fast float @_Z5rootnfi(float %tmp, i32 -1)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_rootn_m2
-; GCN-POSTLINK: tail call fast float @_Z5rootnfi(float %tmp, i32 -2)
-; GCN-PRELINK: %__rootn2rsqrt = tail call fast float @_Z5rsqrtf(float %tmp)
+; GCN-POSTLINK: call fast float @_Z5rootnfi(float %tmp, i32 -2)
+; GCN-PRELINK: %__rootn2rsqrt = call fast float @_Z5rsqrtf(float %tmp)
 define amdgpu_kernel void @test_rootn_m2(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5rootnfi(float %tmp, i32 -2)
+  %call = call fast float @_Z5rootnfi(float %tmp, i32 -2)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_fma_0x
 ; GCN: store float %y, float addrspace(1)* %a
 define amdgpu_kernel void @test_fma_0x(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3fmafff(float 0.000000e+00, float %tmp, float %y)
+  %call = call fast float @_Z3fmafff(float 0.000000e+00, float %tmp, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3fmafff(float, float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_fma_x0
 ; GCN: store float %y, float addrspace(1)* %a
 define amdgpu_kernel void @test_fma_x0(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3fmafff(float %tmp, float 0.000000e+00, float %y)
+  %call = call fast float @_Z3fmafff(float %tmp, float 0.000000e+00, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_mad_0x
 ; GCN: store float %y, float addrspace(1)* %a
 define amdgpu_kernel void @test_mad_0x(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3madfff(float 0.000000e+00, float %tmp, float %y)
+  %call = call fast float @_Z3madfff(float 0.000000e+00, float %tmp, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3madfff(float, float, float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_mad_x0
 ; GCN: store float %y, float addrspace(1)* %a
 define amdgpu_kernel void @test_mad_x0(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3madfff(float %tmp, float 0.000000e+00, float %y)
+  %call = call fast float @_Z3madfff(float %tmp, float 0.000000e+00, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_fma_x1y
 ; GCN: %fmaadd = fadd fast float %tmp, %y
 define amdgpu_kernel void @test_fma_x1y(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3fmafff(float %tmp, float 1.000000e+00, float %y)
+  %call = call fast float @_Z3fmafff(float %tmp, float 1.000000e+00, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_fma_1xy
 ; GCN: %fmaadd = fadd fast float %tmp, %y
 define amdgpu_kernel void @test_fma_1xy(float addrspace(1)* nocapture %a, float %y) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3fmafff(float 1.000000e+00, float %tmp, float %y)
+  %call = call fast float @_Z3fmafff(float 1.000000e+00, float %tmp, float %y)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_fma_xy0
 ; GCN: %fmamul = fmul fast float %tmp1, %tmp
 define amdgpu_kernel void @test_fma_xy0(float addrspace(1)* nocapture %a) {
 entry:
   %arrayidx = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp = load float, float addrspace(1)* %arrayidx, align 4
   %tmp1 = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3fmafff(float %tmp, float %tmp1, float 0.000000e+00)
+  %call = call fast float @_Z3fmafff(float %tmp, float %tmp1, float 0.000000e+00)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_exp
-; GCN-NATIVE: tail call fast float @_Z10native_expf(float %tmp)
+; GCN-NATIVE: call fast float @_Z10native_expf(float %tmp)
 define amdgpu_kernel void @test_use_native_exp(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3expf(float %tmp)
+  %call = call fast float @_Z3expf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3expf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_exp2
-; GCN-NATIVE: tail call fast float @_Z11native_exp2f(float %tmp)
+; GCN-NATIVE: call fast float @_Z11native_exp2f(float %tmp)
 define amdgpu_kernel void @test_use_native_exp2(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z4exp2f(float %tmp)
+  %call = call fast float @_Z4exp2f(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z4exp2f(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_exp10
-; GCN-NATIVE: tail call fast float @_Z12native_exp10f(float %tmp)
+; GCN-NATIVE: call fast float @_Z12native_exp10f(float %tmp)
 define amdgpu_kernel void @test_use_native_exp10(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5exp10f(float %tmp)
+  %call = call fast float @_Z5exp10f(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z5exp10f(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_log
-; GCN-NATIVE: tail call fast float @_Z10native_logf(float %tmp)
+; GCN-NATIVE: call fast float @_Z10native_logf(float %tmp)
 define amdgpu_kernel void @test_use_native_log(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3logf(float %tmp)
+  %call = call fast float @_Z3logf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3logf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_log2
-; GCN-NATIVE: tail call fast float @_Z11native_log2f(float %tmp)
+; GCN-NATIVE: call fast float @_Z11native_log2f(float %tmp)
 define amdgpu_kernel void @test_use_native_log2(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z4log2f(float %tmp)
+  %call = call fast float @_Z4log2f(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z4log2f(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_log10
-; GCN-NATIVE: tail call fast float @_Z12native_log10f(float %tmp)
+; GCN-NATIVE: call fast float @_Z12native_log10f(float %tmp)
 define amdgpu_kernel void @test_use_native_log10(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5log10f(float %tmp)
+  %call = call fast float @_Z5log10f(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z5log10f(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_powr
 ; GCN-NATIVE: %tmp1 = load float, float addrspace(1)* %arrayidx1, align 4
-; GCN-NATIVE: %__log2 = tail call fast float @_Z11native_log2f(float %tmp)
+; GCN-NATIVE: %__log2 = call fast float @_Z11native_log2f(float %tmp)
 ; GCN-NATIVE: %__ylogx = fmul fast float %__log2, %tmp1
-; GCN-NATIVE: %__exp2 = tail call fast float @_Z11native_exp2f(float %__ylogx)
+; GCN-NATIVE: %__exp2 = call fast float @_Z11native_exp2f(float %__ylogx)
 ; GCN-NATIVE: store float %__exp2, float addrspace(1)* %a, align 4
 define amdgpu_kernel void @test_use_native_powr(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
   %arrayidx1 = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp1 = load float, float addrspace(1)* %arrayidx1, align 4
-  %call = tail call fast float @_Z4powrff(float %tmp, float %tmp1)
+  %call = call fast float @_Z4powrff(float %tmp, float %tmp1)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_sqrt
-; GCN-NATIVE: tail call fast float @_Z11native_sqrtf(float %tmp)
+; GCN-NATIVE: call fast float @_Z11native_sqrtf(float %tmp)
 define amdgpu_kernel void @test_use_native_sqrt(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z4sqrtf(float %tmp)
+  %call = call fast float @_Z4sqrtf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_dont_use_native_sqrt_fast_f64
-; GCN: tail call fast double @_Z4sqrtd(double %tmp)
+; GCN: call fast double @_Z4sqrtd(double %tmp)
 define amdgpu_kernel void @test_dont_use_native_sqrt_fast_f64(double addrspace(1)* nocapture %a) {
 entry:
   %tmp = load double, double addrspace(1)* %a, align 8
-  %call = tail call fast double @_Z4sqrtd(double %tmp)
+  %call = call fast double @_Z4sqrtd(double %tmp)
   store double %call, double addrspace(1)* %a, align 8
   ret void
 }
 
 declare float @_Z4sqrtf(float)
 declare double @_Z4sqrtd(double)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_rsqrt
-; GCN-NATIVE: tail call fast float @_Z12native_rsqrtf(float %tmp)
+; GCN-NATIVE: call fast float @_Z12native_rsqrtf(float %tmp)
 define amdgpu_kernel void @test_use_native_rsqrt(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z5rsqrtf(float %tmp)
+  %call = call fast float @_Z5rsqrtf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z5rsqrtf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_tan
-; GCN-NATIVE: tail call fast float @_Z10native_tanf(float %tmp)
+; GCN-NATIVE: call fast float @_Z10native_tanf(float %tmp)
 define amdgpu_kernel void @test_use_native_tan(float addrspace(1)* nocapture %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
-  %call = tail call fast float @_Z3tanf(float %tmp)
+  %call = call fast float @_Z3tanf(float %tmp)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z3tanf(float)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_use_native_sincos
-; GCN-NATIVE: tail call float @_Z10native_sinf(float %tmp)
-; GCN-NATIVE: tail call float @_Z10native_cosf(float %tmp)
+; GCN-NATIVE: call float @_Z10native_sinf(float %tmp)
+; GCN-NATIVE: call float @_Z10native_cosf(float %tmp)
 define amdgpu_kernel void @test_use_native_sincos(float addrspace(1)* %a) {
 entry:
   %tmp = load float, float addrspace(1)* %a, align 4
   %arrayidx1 = getelementptr inbounds float, float addrspace(1)* %a, i64 1
   %tmp1 = addrspacecast float addrspace(1)* %arrayidx1 to float*
-  %call = tail call fast float @_Z6sincosfPf(float %tmp, float* %tmp1)
+  %call = call fast float @_Z6sincosfPf(float %tmp, float* %tmp1)
   store float %call, float addrspace(1)* %a, align 4
   ret void
 }
 
 declare float @_Z6sincosfPf(float, float*)
 
 %opencl.pipe_t = type opaque
 %opencl.reserve_id_t = type opaque
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_read_pipe(%opencl.pipe_t addrspace(1)* %p, i32 addrspace(1)* %ptr)
 ; GCN-PRELINK: call i32 @__read_pipe_2_4(%opencl.pipe_t addrspace(1)* %{{.*}}, i32* %{{.*}}) #[[$NOUNWIND:[0-9]+]]
 ; GCN-PRELINK: call i32 @__read_pipe_4_4(%opencl.pipe_t addrspace(1)* %{{.*}}, %opencl.reserve_id_t addrspace(5)* %{{.*}}, i32 2, i32* %{{.*}}) #[[$NOUNWIND]]
 define amdgpu_kernel void @test_read_pipe(%opencl.pipe_t addrspace(1)* %p, i32 addrspace(1)* %ptr) local_unnamed_addr {
 entry:
   %tmp = bitcast i32 addrspace(1)* %ptr to i8 addrspace(1)*
   %tmp1 = addrspacecast i8 addrspace(1)* %tmp to i8*
-  %tmp2 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p, i8* %tmp1, i32 4, i32 4) #0
-  %tmp3 = tail call %opencl.reserve_id_t addrspace(5)* @__reserve_read_pipe(%opencl.pipe_t addrspace(1)* %p, i32 2, i32 4, i32 4)
-  %tmp4 = tail call i32 @__read_pipe_4(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 2, i8* %tmp1, i32 4, i32 4) #0
-  tail call void @__commit_read_pipe(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 4, i32 4)
+  %tmp2 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p, i8* %tmp1, i32 4, i32 4) #0
+  %tmp3 = call %opencl.reserve_id_t addrspace(5)* @__reserve_read_pipe(%opencl.pipe_t addrspace(1)* %p, i32 2, i32 4, i32 4)
+  %tmp4 = call i32 @__read_pipe_4(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 2, i8* %tmp1, i32 4, i32 4) #0
+  call void @__commit_read_pipe(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 4, i32 4)
   ret void
 }
 
 declare i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)*, i8*, i32, i32)
 
 declare %opencl.reserve_id_t addrspace(5)* @__reserve_read_pipe(%opencl.pipe_t addrspace(1)*, i32, i32, i32)
 
 declare i32 @__read_pipe_4(%opencl.pipe_t addrspace(1)*, %opencl.reserve_id_t addrspace(5)*, i32, i8*, i32, i32)
 
 declare void @__commit_read_pipe(%opencl.pipe_t addrspace(1)*, %opencl.reserve_id_t addrspace(5)*, i32, i32)
 
 ; GCN-LABEL: {{^}}define amdgpu_kernel void @test_write_pipe(%opencl.pipe_t addrspace(1)* %p, i32 addrspace(1)* %ptr)
 ; GCN-PRELINK: call i32 @__write_pipe_2_4(%opencl.pipe_t addrspace(1)* %{{.*}}, i32* %{{.*}}) #[[$NOUNWIND]]
 ; GCN-PRELINK: call i32 @__write_pipe_4_4(%opencl.pipe_t addrspace(1)* %{{.*}}, %opencl.reserve_id_t addrspace(5)* %{{.*}}, i32 2, i32* %{{.*}}) #[[$NOUNWIND]]
 define amdgpu_kernel void @test_write_pipe(%opencl.pipe_t addrspace(1)* %p, i32 addrspace(1)* %ptr) local_unnamed_addr {
 entry:
   %tmp = bitcast i32 addrspace(1)* %ptr to i8 addrspace(1)*
   %tmp1 = addrspacecast i8 addrspace(1)* %tmp to i8*
-  %tmp2 = tail call i32 @__write_pipe_2(%opencl.pipe_t addrspace(1)* %p, i8* %tmp1, i32 4, i32 4) #0
-  %tmp3 = tail call %opencl.reserve_id_t addrspace(5)* @__reserve_write_pipe(%opencl.pipe_t addrspace(1)* %p, i32 2, i32 4, i32 4) #0
-  %tmp4 = tail call i32 @__write_pipe_4(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 2, i8* %tmp1, i32 4, i32 4) #0
-  tail call void @__commit_write_pipe(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 4, i32 4) #0
+  %tmp2 = call i32 @__write_pipe_2(%opencl.pipe_t addrspace(1)* %p, i8* %tmp1, i32 4, i32 4) #0
+  %tmp3 = call %opencl.reserve_id_t addrspace(5)* @__reserve_write_pipe(%opencl.pipe_t addrspace(1)* %p, i32 2, i32 4, i32 4) #0
+  %tmp4 = call i32 @__write_pipe_4(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 2, i8* %tmp1, i32 4, i32 4) #0
+  call void @__commit_write_pipe(%opencl.pipe_t addrspace(1)* %p, %opencl.reserve_id_t addrspace(5)* %tmp3, i32 4, i32 4) #0
   ret void
 }
 
 declare i32 @__write_pipe_2(%opencl.pipe_t addrspace(1)*, i8*, i32, i32) local_unnamed_addr
 
 declare %opencl.reserve_id_t addrspace(5)* @__reserve_write_pipe(%opencl.pipe_t addrspace(1)*, i32, i32, i32) local_unnamed_addr
 
 declare i32 @__write_pipe_4(%opencl.pipe_t addrspace(1)*, %opencl.reserve_id_t addrspace(5)*, i32, i8*, i32, i32) local_unnamed_addr

 ; GCN-PRELINK: call i32 @__read_pipe_2_16(%opencl.pipe_t addrspace(1)* %{{.*}}, <2 x i64>* %{{.*}}) #[[$NOUNWIND]]
 ; GCN-PRELINK: call i32 @__read_pipe_2_32(%opencl.pipe_t addrspace(1)* %{{.*}}, <4 x i64>* %{{.*}} #[[$NOUNWIND]]
 ; GCN-PRELINK: call i32 @__read_pipe_2_64(%opencl.pipe_t addrspace(1)* %{{.*}}, <8 x i64>* %{{.*}} #[[$NOUNWIND]]
 ; GCN-PRELINK: call i32 @__read_pipe_2_128(%opencl.pipe_t addrspace(1)* %{{.*}}, <16 x i64>* %{{.*}} #[[$NOUNWIND]]
 ; GCN-PRELINK: call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %{{.*}}, i8* %{{.*}} i32 400, i32 4) #[[$NOUNWIND]]
 define amdgpu_kernel void @test_pipe_size(%opencl.pipe_t addrspace(1)* %p1, i8 addrspace(1)* %ptr1, %opencl.pipe_t addrspace(1)* %p2, i16 addrspace(1)* %ptr2, %opencl.pipe_t addrspace(1)* %p4, i32 addrspace(1)* %ptr4, %opencl.pipe_t addrspace(1)* %p8, i64 addrspace(1)* %ptr8, %opencl.pipe_t addrspace(1)* %p16, <2 x i64> addrspace(1)* %ptr16, %opencl.pipe_t addrspace(1)* %p32, <4 x i64> addrspace(1)* %ptr32, %opencl.pipe_t addrspace(1)* %p64, <8 x i64> addrspace(1)* %ptr64, %opencl.pipe_t addrspace(1)* %p128, <16 x i64> addrspace(1)* %ptr128, %opencl.pipe_t addrspace(1)* %pu, %struct.S addrspace(1)* %ptru) local_unnamed_addr #0 {
 entry:
   %tmp = addrspacecast i8 addrspace(1)* %ptr1 to i8*
-  %tmp1 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p1, i8* %tmp, i32 1, i32 1) #0
+  %tmp1 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p1, i8* %tmp, i32 1, i32 1) #0
   %tmp2 = bitcast i16 addrspace(1)* %ptr2 to i8 addrspace(1)*
   %tmp3 = addrspacecast i8 addrspace(1)* %tmp2 to i8*
-  %tmp4 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p2, i8* %tmp3, i32 2, i32 2) #0
+  %tmp4 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p2, i8* %tmp3, i32 2, i32 2) #0
   %tmp5 = bitcast i32 addrspace(1)* %ptr4 to i8 addrspace(1)*
   %tmp6 = addrspacecast i8 addrspace(1)* %tmp5 to i8*
-  %tmp7 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p4, i8* %tmp6, i32 4, i32 4) #0
+  %tmp7 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p4, i8* %tmp6, i32 4, i32 4) #0
   %tmp8 = bitcast i64 addrspace(1)* %ptr8 to i8 addrspace(1)*
   %tmp9 = addrspacecast i8 addrspace(1)* %tmp8 to i8*
-  %tmp10 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p8, i8* %tmp9, i32 8, i32 8) #0
+  %tmp10 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p8, i8* %tmp9, i32 8, i32 8) #0
   %tmp11 = bitcast <2 x i64> addrspace(1)* %ptr16 to i8 addrspace(1)*
   %tmp12 = addrspacecast i8 addrspace(1)* %tmp11 to i8*
-  %tmp13 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p16, i8* %tmp12, i32 16, i32 16) #0
+  %tmp13 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p16, i8* %tmp12, i32 16, i32 16) #0
   %tmp14 = bitcast <4 x i64> addrspace(1)* %ptr32 to i8 addrspace(1)*
   %tmp15 = addrspacecast i8 addrspace(1)* %tmp14 to i8*
-  %tmp16 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p32, i8* %tmp15, i32 32, i32 32) #0
+  %tmp16 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p32, i8* %tmp15, i32 32, i32 32) #0
   %tmp17 = bitcast <8 x i64> addrspace(1)* %ptr64 to i8 addrspace(1)*
   %tmp18 = addrspacecast i8 addrspace(1)* %tmp17 to i8*
-  %tmp19 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p64, i8* %tmp18, i32 64, i32 64) #0
+  %tmp19 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p64, i8* %tmp18, i32 64, i32 64) #0
   %tmp20 = bitcast <16 x i64> addrspace(1)* %ptr128 to i8 addrspace(1)*
   %tmp21 = addrspacecast i8 addrspace(1)* %tmp20 to i8*
-  %tmp22 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p128, i8* %tmp21, i32 128, i32 128) #0
+  %tmp22 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %p128, i8* %tmp21, i32 128, i32 128) #0
   %tmp23 = bitcast %struct.S addrspace(1)* %ptru to i8 addrspace(1)*
   %tmp24 = addrspacecast i8 addrspace(1)* %tmp23 to i8*
-  %tmp25 = tail call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %pu, i8* %tmp24, i32 400, i32 4) #0
+  %tmp25 = call i32 @__read_pipe_2(%opencl.pipe_t addrspace(1)* %pu, i8* %tmp24, i32 400, i32 4) #0
   ret void
 }
 
 ; GCN-PRELINK: declare float @_Z4fabsf(float) local_unnamed_addr #[[$NOUNWIND_READONLY:[0-9]+]]
 ; GCN-PRELINK: declare float @_Z4cbrtf(float) local_unnamed_addr #[[$NOUNWIND_READONLY]]
 ; GCN-PRELINK: declare float @_Z11native_sqrtf(float) local_unnamed_addr #[[$NOUNWIND_READONLY]]
 
 ; CGN-PRELINK: attributes #[[$NOUNWIND]] = { nounwind }
 ; GCN-PRELINK: attributes #[[$NOUNWIND_READONLY]] = { nounwind readonly }
 attributes #0 = { nounwind }

llvm/test/Feature/optnone-opt.ll

 attributes #0 = { optnone noinline }
 
 ; Nothing that runs at -O0 gets skipped.
 ; OPT-O0-NOT: Skipping pass
 
 ; IR passes run at -O1 and higher.
 ; OPT-O1-DAG: Skipping pass 'Aggressive Dead Code Elimination'
 ; OPT-O1-DAG: Skipping pass 'Combine redundant instructions'
-; OPT-O1-DAG: Skipping pass 'Dead Store Elimination'
 ; OPT-O1-DAG: Skipping pass 'Early CSE'
-; OPT-O1-DAG: Skipping pass 'Jump Threading'
-; OPT-O1-DAG: Skipping pass 'MemCpy Optimization'
 ; OPT-O1-DAG: Skipping pass 'Reassociate expressions'
 ; OPT-O1-DAG: Skipping pass 'Simplify the CFG'
 ; OPT-O1-DAG: Skipping pass 'Sparse Conditional Constant Propagation'
-; OPT-O1-DAG: Skipping pass 'SROA'
-; OPT-O1-DAG: Skipping pass 'Tail Call Elimination'
-; OPT-O1-DAG: Skipping pass 'Value Propagation'
 
 ; Additional IR passes run at -O2 and higher.
 ; OPT-O2O3-DAG: Skipping pass 'Global Value Numbering'
 ; OPT-O2O3-DAG: Skipping pass 'SLP Vectorizer'
 
 ; Additional IR passes that opt doesn't turn on by default.
 ; OPT-MORE-DAG: Skipping pass 'Dead Code Elimination'
 ; OPT-MORE-DAG: Skipping pass 'Dead Instruction Elimination'

llvm/test/Other/new-pm-defaults.ll

 ; The IR below was crafted so as:
 ; 1) To have a loop, so we create a loop pass manager
 ; 2) To be "immutable" in the sense that no pass in the standard
 ;    pipeline will modify it.
 ; Since no transformations take place, we don't expect any analyses
 ; to be invalidated.
 ; Any invalidation that shows up here is a bug, unless we started modifying
 ; the IR, in which case we need to make it immutable harder.
 
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='default<O1>' -S %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O1
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='default<O2>' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O2
+; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O2 \
+; RUN:     --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3
+; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
+; RUN:     --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='default<Os>' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-Os
+; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-Os \
+; RUN:     --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='default<Oz>' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-Oz
+; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-Oz \
+; RUN:     --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='lto-pre-link<O2>' -S %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O2 \
-; RUN:     --check-prefix=CHECK-O2-LTO
+; RUN:     --check-prefix=CHECK-O2-LTO --check-prefix=CHECK-O23SZ
 
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-peephole='no-op-function' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-PEEPHOLE
+; RUN:     --check-prefix=CHECK-EP-PEEPHOLE --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-late-loop-optimizations='no-op-loop' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-LOOP-LATE
+; RUN:     --check-prefix=CHECK-EP-LOOP-LATE --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-loop-optimizer-end='no-op-loop' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-LOOP-END
+; RUN:     --check-prefix=CHECK-EP-LOOP-END --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-scalar-optimizer-late='no-op-function' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-SCALAR-LATE
+; RUN:     --check-prefix=CHECK-EP-SCALAR-LATE --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-cgscc-optimizer-late='no-op-cgscc' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-CGSCC-LATE
+; RUN:     --check-prefix=CHECK-EP-CGSCC-LATE --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-vectorizer-start='no-op-function' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-VECTORIZER-START
+; RUN:     --check-prefix=CHECK-EP-VECTORIZER-START --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-pipeline-start='no-op-module' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-PIPELINE-START
+; RUN:     --check-prefix=CHECK-EP-PIPELINE-START --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-pipeline-start='no-op-module' \
 ; RUN:     -passes='lto-pre-link<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-PIPELINE-START
+; RUN:     --check-prefix=CHECK-EP-PIPELINE-START --check-prefix=CHECK-O23SZ
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes-ep-optimizer-last='no-op-function' \
 ; RUN:     -passes='default<O3>' -S  %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefix=CHECK-O --check-prefix=CHECK-O3 \
-; RUN:     --check-prefix=CHECK-EP-OPTIMIZER-LAST
+; RUN:     --check-prefix=CHECK-EP-OPTIMIZER-LAST --check-prefix=CHECK-O23SZ
 
 ; CHECK-O: Running analysis: PassInstrumentationAnalysis
 ; CHECK-O-NEXT: Starting llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: PassManager<{{.*}}Module{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: ForceFunctionAttrsPass
 ; CHECK-EP-PIPELINE-START-NEXT: Running pass: NoOpModulePass
 ; CHECK-O-NEXT: Running pass: PassManager<{{.*}}Module{{.*}}>

 ; CHECK-O-NEXT: Running pass: InlinerPass
 ; CHECK-O-NEXT: Running pass: PostOrderFunctionAttrsPass
 ; CHECK-O3-NEXT: Running pass: ArgumentPromotionPass
 ; CHECK-O-NEXT: Running pass: CGSCCToFunctionPassAdaptor<{{.*}}PassManager{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: SROA
 ; CHECK-O-NEXT: Running pass: EarlyCSEPass
 ; CHECK-O-NEXT: Running analysis: MemorySSAAnalysis
-; CHECK-O-NEXT: Running pass: SpeculativeExecutionPass
-; CHECK-O-NEXT: Running pass: JumpThreadingPass
-; CHECK-O-NEXT: Running analysis: LazyValueAnalysis
-; CHECK-O-NEXT: Running pass: CorrelatedValuePropagationPass
+; CHECK-O23SZ-NEXT: Running pass: SpeculativeExecutionPass
+; CHECK-O23SZ-NEXT: Running pass: JumpThreadingPass
+; CHECK-O23SZ-NEXT: Running analysis: LazyValueAnalysis
+; CHECK-O23SZ-NEXT: Running pass: CorrelatedValuePropagationPass
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O3-NEXT: AggressiveInstCombinePass
 ; CHECK-O-NEXT: Running pass: InstCombinePass
 ; CHECK-O1-NEXT: Running pass: LibCallsShrinkWrapPass
 ; CHECK-O2-NEXT: Running pass: LibCallsShrinkWrapPass
 ; CHECK-O3-NEXT: Running pass: LibCallsShrinkWrapPass
 ; CHECK-EP-PEEPHOLE-NEXT: Running pass: NoOpFunctionPass
-; CHECK-O-NEXT: Running pass: TailCallElimPass
+; CHECK-O23SZ-NEXT: Running pass: TailCallElimPass
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O-NEXT: Running pass: ReassociatePass
 ; CHECK-O-NEXT: Running pass: RequireAnalysisPass<{{.*}}OptimizationRemarkEmitterAnalysis
 ; CHECK-O-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LoopStandardAnalysisResults{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: LoopSimplifyPass
 ; CHECK-O-NEXT: Running analysis: LoopAnalysis
 ; CHECK-O-NEXT: Running pass: LCSSAPass

 ; CHECK-O-NEXT: Running pass: IndVarSimplifyPass
 ; CHECK-O-NEXT: Running pass: LoopIdiomRecognizePass
 ; CHECK-EP-LOOP-LATE-NEXT: Running pass: NoOpLoopPass
 ; CHECK-O-NEXT: Running pass: LoopDeletionPass
 ; CHECK-O-NEXT: Running pass: LoopFullUnrollPass
 ; CHECK-EP-LOOP-END-NEXT: Running pass: NoOpLoopPass
 ; CHECK-O-NEXT: Finished Loop pass manager run.
 ; CHECK-O-NEXT: Running pass: SROA on foo
-; CHECK-Os-NEXT: Running pass: MergedLoadStoreMotionPass
-; CHECK-Os-NEXT: Running pass: GVN
-; CHECK-Os-NEXT: Running analysis: MemoryDependenceAnalysis
-; CHECK-Os-NEXT: Running analysis: PhiValuesAnalysis
-; CHECK-Oz-NEXT: Running pass: MergedLoadStoreMotionPass
-; CHECK-Oz-NEXT: Running pass: GVN
-; CHECK-Oz-NEXT: Running analysis: MemoryDependenceAnalysis
-; CHECK-Oz-NEXT: Running analysis: PhiValuesAnalysis
-; CHECK-O2-NEXT: Running pass: MergedLoadStoreMotionPass
-; CHECK-O2-NEXT: Running pass: GVN
-; CHECK-O2-NEXT: Running analysis: MemoryDependenceAnalysis
-; CHECK-O2-NEXT: Running analysis: PhiValuesAnalysis
-; CHECK-O3-NEXT: Running pass: MergedLoadStoreMotionPass
-; CHECK-O3-NEXT: Running pass: GVN
-; CHECK-O3-NEXT: Running analysis: MemoryDependenceAnalysis
-; CHECK-O3-NEXT: Running analysis: PhiValuesAnalysis
+; CHECK-O23SZ-NEXT: Running pass: MergedLoadStoreMotionPass
+; CHECK-O23SZ-NEXT: Running pass: GVN
+; CHECK-O23SZ-NEXT: Running analysis: MemoryDependenceAnalysis
+; CHECK-O23SZ-NEXT: Running analysis: PhiValuesAnalysis
 ; CHECK-O-NEXT: Running pass: MemCpyOptPass
 ; CHECK-O1-NEXT: Running analysis: MemoryDependenceAnalysis
 ; CHECK-O1-NEXT: Running analysis: PhiValuesAnalysis
 ; CHECK-O-NEXT: Running pass: SCCPPass
 ; CHECK-O-NEXT: Running pass: BDCEPass
 ; CHECK-O-NEXT: Running analysis: DemandedBitsAnalysis
 ; CHECK-O-NEXT: Running pass: InstCombinePass
 ; CHECK-EP-PEEPHOLE-NEXT: Running pass: NoOpFunctionPass
-; CHECK-O-NEXT: Running pass: JumpThreadingPass
-; CHECK-O-NEXT: Running pass: CorrelatedValuePropagationPass
-; CHECK-O-NEXT: Running pass: DSEPass
-; CHECK-O-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LICMPass{{.*}}>
-; CHECK-O-NEXT: Starting llvm::Function pass manager run.
-; CHECK-O-NEXT: Running pass: LoopSimplifyPass
-; CHECK-O-NEXT: Running pass: LCSSAPass
-; CHECK-O-NEXT: Finished llvm::Function pass manager run.
+; CHECK-O23SZ-NEXT: Running pass: JumpThreadingPass
+; CHECK-O23SZ-NEXT: Running pass: CorrelatedValuePropagationPass
+; CHECK-O23SZ-NEXT: Running pass: DSEPass
+; CHECK-O23SZ-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LICMPass{{.*}}>
+; CHECK-O23SZ-NEXT: Starting llvm::Function pass manager run.
+; CHECK-O23SZ-NEXT: Running pass: LoopSimplifyPass
+; CHECK-O23SZ-NEXT: Running pass: LCSSAPass
+; CHECK-O23SZ-NEXT: Finished llvm::Function pass manager run.
 ; CHECK-EP-SCALAR-LATE-NEXT: Running pass: NoOpFunctionPass
 ; CHECK-O-NEXT: Running pass: ADCEPass
 ; CHECK-O-NEXT: Running analysis: PostDominatorTreeAnalysis
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O-NEXT: Running pass: InstCombinePass
 ; CHECK-EP-PEEPHOLE-NEXT: Running pass: NoOpFunctionPass
 ; CHECK-O-NEXT: Finished llvm::Function pass manager run.
 ; CHECK-EP-CGSCC-LATE-NEXT: Running pass: NoOpCGSCCPass
 ; CHECK-O-NEXT: Finished CGSCC pass manager run.
 ; CHECK-O-NEXT: Finished llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: PassManager<{{.*}}Module{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: GlobalOptPass
 ; CHECK-O-NEXT: Running pass: GlobalDCEPass
 ; CHECK-O2-LTO-NOT: Running pass: EliminateAvailableExternallyPass
 ; CHECK-O: Running pass: ReversePostOrderFunctionAttrsPass
 ; CHECK-O-NEXT: Running pass: RequireAnalysisPass<{{.*}}GlobalsAA
 ; CHECK-O-NEXT: Running pass: ModuleToFunctionPassAdaptor<{{.*}}PassManager{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: Float2IntPass
 ; CHECK-O-NEXT: Running pass: LowerConstantIntrinsicsPass on foo
 ; CHECK-EP-VECTORIZER-START-NEXT: Running pass: NoOpFunctionPass
 ; CHECK-O-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LoopRotatePass
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: LoopSimplifyPass
 ; CHECK-O-NEXT: Running pass: LCSSAPass
 ; CHECK-O-NEXT: Finished llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: LoopDistributePass
 ; CHECK-O-NEXT: Running pass: LoopVectorizePass
 ; CHECK-O-NEXT: Running analysis: BlockFrequencyAnalysis
 ; CHECK-O-NEXT: Running analysis: BranchProbabilityAnalysis
 ; CHECK-O-NEXT: Running pass: LoopLoadEliminationPass

llvm/test/Other/new-pm-thinlto-defaults.ll

 ; The IR below was crafted so as:
 ; 1) To have a loop, so we create a loop pass manager
 ; 2) To be "immutable" in the sense that no pass in the standard
 ;    pipeline will modify it.
 ; Since no transformations take place, we don't expect any analyses
 ; to be invalidated.
 ; Any invalidation that shows up here is a bug, unless we started modifying
 ; the IR, in which case we need to make it immutable harder.
 ;
 ; Prelink pipelines:
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto-pre-link<O1>,name-anon-globals' -S %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O1,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-O1
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto-pre-link<O2>,name-anon-globals' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-O2
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-O23SZ,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-O2
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto-pre-link<O3>,name-anon-globals' -S -passes-ep-pipeline-start='no-op-module' %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O3,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-O3,CHECK-EP-PIPELINE-START
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O3,CHECK-O23SZ,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-O3,CHECK-EP-PIPELINE-START
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto-pre-link<Os>,name-anon-globals' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Os,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-Os
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Os,CHECK-O23SZ,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-Os
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto-pre-link<Oz>,name-anon-globals' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Oz,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-Oz
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Oz,CHECK-O23SZ,CHECK-PRELINK-O,CHECK-PRELINK-O-NODIS,CHECK-PRELINK-Oz
 ; RUN: opt -disable-verify -debug-pass-manager -new-pm-debug-info-for-profiling \
 ; RUN:     -passes='thinlto-pre-link<O2>,name-anon-globals' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-DIS,CHECK-O,CHECK-O2,CHECK-PRELINK-O,CHECK-PRELINK-O2
+; RUN:     | FileCheck %s --check-prefixes=CHECK-DIS,CHECK-O,CHECK-O2,CHECK-O23SZ,CHECK-PRELINK-O,CHECK-PRELINK-O2
 ;
 ; Postlink pipelines:
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto<O1>' -S %s 2>&1 \
 ; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O1,CHECK-POSTLINK-O,CHECK-POSTLINK-O1
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto<O2>' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-POSTLINK-O,CHECK-POSTLINK-O2
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-O23SZ,CHECK-POSTLINK-O,CHECK-POSTLINK-O2
 ; RUN: opt -disable-verify -debug-pass-manager -passes-ep-pipeline-start='no-op-module' \
 ; RUN:     -passes='thinlto<O3>' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O3,CHECK-POSTLINK-O,CHECK-POSTLINK-O3
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O3,CHECK-O23SZ,CHECK-POSTLINK-O,CHECK-POSTLINK-O3
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto<Os>' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Os,CHECK-POSTLINK-O,CHECK-POSTLINK-Os
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Os,CHECK-O23SZ,CHECK-POSTLINK-O,CHECK-POSTLINK-Os
 ; RUN: opt -disable-verify -debug-pass-manager \
 ; RUN:     -passes='thinlto<Oz>' -S %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Oz,CHECK-POSTLINK-O,CHECK-POSTLINK-Oz
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-Oz,CHECK-O23SZ,CHECK-POSTLINK-O,CHECK-POSTLINK-Oz
 ; RUN: opt -disable-verify -debug-pass-manager -new-pm-debug-info-for-profiling \
 ; RUN:     -passes='thinlto<O2>' -S  %s 2>&1 \
-; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-POSTLINK-O,CHECK-POSTLINK-O2
+; RUN:     | FileCheck %s --check-prefixes=CHECK-O,CHECK-O2,CHECK-O23SZ,CHECK-POSTLINK-O,CHECK-POSTLINK-O2
 ;
 ; CHECK-O: Running analysis: PassInstrumentationAnalysis
 ; CHECK-O-NEXT: Starting llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: PassManager<{{.*}}Module{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Module pass manager run.
 ; CHECK-O-NEXT: Running pass: ForceFunctionAttrsPass
 ; CHECK-EP-PIPELINE-START-NEXT: Running pass: NoOpModulePass
 ; CHECK-DIS-NEXT: Running pass: ModuleToFunctionPassAdaptor<{{.*}}AddDiscriminatorsPass{{.*}}>

 ; CHECK-O-NEXT: Running pass: InlinerPass
 ; CHECK-O-NEXT: Running pass: PostOrderFunctionAttrsPass
 ; CHECK-O3-NEXT: Running pass: ArgumentPromotionPass
 ; CHECK-O-NEXT: Running pass: CGSCCToFunctionPassAdaptor<{{.*}}PassManager{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run.
 ; CHECK-O-NEXT: Running pass: SROA
 ; CHECK-O-NEXT: Running pass: EarlyCSEPass
 ; CHECK-O-NEXT: Running analysis: MemorySSAAnalysis
-; CHECK-O-NEXT: Running pass: SpeculativeExecutionPass
-; CHECK-O-NEXT: Running pass: JumpThreadingPass
-; CHECK-O-NEXT: Running analysis: LazyValueAnalysis
-; CHECK-O-NEXT: Running pass: CorrelatedValuePropagationPass
+; CHECK-O23SZ-NEXT: Running pass: SpeculativeExecutionPass
+; CHECK-O23SZ-NEXT: Running pass: JumpThreadingPass
+; CHECK-O23SZ-NEXT: Running analysis: LazyValueAnalysis
+; CHECK-O23SZ-NEXT: Running pass: CorrelatedValuePropagationPass
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O3-NEXT: Running pass: AggressiveInstCombinePass
 ; CHECK-O-NEXT: Running pass: InstCombinePass
 ; CHECK-O1-NEXT: Running pass: LibCallsShrinkWrapPass
 ; CHECK-O2-NEXT: Running pass: LibCallsShrinkWrapPass
 ; CHECK-O3-NEXT: Running pass: LibCallsShrinkWrapPass
-; CHECK-O-NEXT: Running pass: TailCallElimPass
+; CHECK-O23SZ-NEXT: Running pass: TailCallElimPass
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O-NEXT: Running pass: ReassociatePass
 ; CHECK-O-NEXT: Running pass: RequireAnalysisPass<{{.*}}OptimizationRemarkEmitterAnalysis
 ; CHECK-O-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LoopStandardAnalysisResults{{.*}}>
 ; CHECK-O-NEXT: Starting llvm::Function pass manager run
 ; CHECK-O-NEXT: Running pass: LoopSimplifyPass
 ; CHECK-O-NEXT: Running analysis: LoopAnalysis
 ; CHECK-O-NEXT: Running pass: LCSSAPass

 ; CHECK-O3-NEXT: Running analysis: PhiValuesAnalysis
 ; CHECK-O-NEXT: Running pass: MemCpyOptPass
 ; CHECK-O1-NEXT: Running analysis: MemoryDependenceAnalysis
 ; CHECK-O1-NEXT: Running analysis: PhiValuesAnalysis
 ; CHECK-O-NEXT: Running pass: SCCPPass
 ; CHECK-O-NEXT: Running pass: BDCEPass
 ; CHECK-O-NEXT: Running analysis: DemandedBitsAnalysis
 ; CHECK-O-NEXT: Running pass: InstCombinePass
-; CHECK-O-NEXT: Running pass: JumpThreadingPass
-; CHECK-O-NEXT: Running pass: CorrelatedValuePropagationPass
-; CHECK-O-NEXT: Running pass: DSEPass
-; CHECK-O-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LICMPass{{.*}}>
-; CHECK-O-NEXT: Starting llvm::Function pass manager run
-; CHECK-O-NEXT: Running pass: LoopSimplifyPass
-; CHECK-O-NEXT: Running pass: LCSSAPass
-; CHECK-O-NEXT: Finished llvm::Function pass manager run
+; CHECK-O23SZ-NEXT: Running pass: JumpThreadingPass
+; CHECK-O23SZ-NEXT: Running pass: CorrelatedValuePropagationPass
+; CHECK-O23SZ-NEXT: Running pass: DSEPass
+; CHECK-O23SZ-NEXT: Running pass: FunctionToLoopPassAdaptor<{{.*}}LICMPass{{.*}}>
+; CHECK-O23SZ-NEXT: Starting llvm::Function pass manager run
+; CHECK-O23SZ-NEXT: Running pass: LoopSimplifyPass
+; CHECK-O23SZ-NEXT: Running pass: LCSSAPass
+; CHECK-O23SZ-NEXT: Finished llvm::Function pass manager run
 ; CHECK-O-NEXT: Running pass: ADCEPass
 ; CHECK-O-NEXT: Running analysis: PostDominatorTreeAnalysis
 ; CHECK-O-NEXT: Running pass: SimplifyCFGPass
 ; CHECK-O-NEXT: Running pass: InstCombinePass
 ; CHECK-O-NEXT: Finished llvm::Function pass manager run.
 ; CHECK-O-NEXT: Finished CGSCC pass manager run.
 ; CHECK-O-NEXT: Finished llvm::Module pass manager run.
 ; CHECK-PRELINK-O-NEXT: Running pass: GlobalOptPass

llvm/test/Transforms/MemCpyOpt/lifetime.ll

-; RUN: opt < %s -O1 -S | FileCheck %s
+; RUN: opt < %s -O2 -S | FileCheck %s
 
 ; performCallSlotOptzn in MemCpy should not exchange the calls to
 ; @llvm.lifetime.start and @llvm.memcpy.
 
 declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
 declare void @llvm.lifetime.start.p0i8(i64, i8* nocapture) #1
 declare void @llvm.lifetime.end.p0i8(i64, i8* nocapture) #1
 

llvm/test/Transforms/PhaseOrdering/simplifycfg-options.ll

 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -O1 -S < %s                    | FileCheck %s --check-prefix=ALL --check-prefix=OLDPM
 ; RUN: opt -passes='default<O1>' -S < %s  | FileCheck %s --check-prefix=ALL --check-prefix=NEWPM
 
 ; Don't simplify unconditional branches from empty blocks in simplifyCFG
 ; until late in the pipeline because it can destroy canonical loop structure.
 
 define i1 @PR33605(i32 %a, i32 %b, i32* %c) {
 ; ALL-LABEL: @PR33605(
-; ALL-NEXT:  for.body:
+; ALL-NEXT:  entry:
 ; ALL-NEXT:    [[OR:%.*]] = or i32 [[B:%.*]], [[A:%.*]]
 ; ALL-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[C:%.*]], i64 1
 ; ALL-NEXT:    [[TMP0:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
 ; ALL-NEXT:    [[CMP:%.*]] = icmp eq i32 [[OR]], [[TMP0]]
 ; ALL-NEXT:    br i1 [[CMP]], label [[IF_END:%.*]], label [[IF_THEN:%.*]]
 ; ALL:       if.then:
 ; ALL-NEXT:    store i32 [[OR]], i32* [[ARRAYIDX]], align 4
-; ALL-NEXT:    tail call void @foo()
+; ALL-NEXT:    call void @foo()
 ; ALL-NEXT:    br label [[IF_END]]
 ; ALL:       if.end:
-; ALL-NEXT:    [[CHANGED_1_OFF0:%.*]] = phi i1 [ true, [[IF_THEN]] ], [ false, [[FOR_BODY:%.*]] ]
+; ALL-NEXT:    [[CHANGED_1_OFF0:%.*]] = phi i1 [ true, [[IF_THEN]] ], [ false, [[ENTRY:%.*]] ]
 ; ALL-NEXT:    [[TMP1:%.*]] = load i32, i32* [[C]], align 4
 ; ALL-NEXT:    [[CMP_1:%.*]] = icmp eq i32 [[OR]], [[TMP1]]
 ; ALL-NEXT:    br i1 [[CMP_1]], label [[IF_END_1:%.*]], label [[IF_THEN_1:%.*]]
 ; ALL:       if.then.1:
 ; ALL-NEXT:    store i32 [[OR]], i32* [[C]], align 4
-; ALL-NEXT:    tail call void @foo()
+; ALL-NEXT:    call void @foo()
 ; ALL-NEXT:    br label [[IF_END_1]]
 ; ALL:       if.end.1:
 ; ALL-NEXT:    [[CHANGED_1_OFF0_1:%.*]] = phi i1 [ true, [[IF_THEN_1]] ], [ [[CHANGED_1_OFF0]], [[IF_END]] ]
 ; ALL-NEXT:    ret i1 [[CHANGED_1_OFF0_1]]
 ;
 entry:
   br label %for.cond
 

llvm/test/Transforms/PhaseOrdering/two-shifts-by-sext.ll

;
   %len.reloaded.2 = load i32, i32* %wide_len, align 4
   %shr = ashr i32 %shl, %len.reloaded.2
   ret i32 %shr
 }
 
 define i32 @two_shifts_by_sext_with_extra_use(i32 %val, i8 signext %len) {
 ; CHECK-LABEL: @two_shifts_by_sext_with_extra_use(
 ; CHECK-NEXT:    [[CONV:%.*]] = sext i8 [[LEN:%.*]] to i32
-; CHECK-NEXT:    tail call void @use_int32(i32 [[CONV]])
+; CHECK-NEXT:    call void @use_int32(i32 [[CONV]])
 ; CHECK-NEXT:    [[SHL:%.*]] = shl i32 [[VAL:%.*]], [[CONV]]
 ; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[SHL]], [[CONV]]
 ; CHECK-NEXT:    ret i32 [[SHR]]
 ;
   %val.addr = alloca i32, align 4
   %len.addr = alloca i8, align 1
   store i32 %val, i32* %val.addr, align 4
   store i8 %len, i8* %len.addr, align 1
;
   ret i32 %shr
 }
 
 declare void @use_int32(i32)
 
 define i32 @two_shifts_by_same_sext_with_extra_use(i32 %val, i8 signext %len) {
 ; CHECK-LABEL: @two_shifts_by_same_sext_with_extra_use(
 ; CHECK-NEXT:    [[CONV:%.*]] = sext i8 [[LEN:%.*]] to i32
-; CHECK-NEXT:    tail call void @use_int32(i32 [[CONV]])
+; CHECK-NEXT:    call void @use_int32(i32 [[CONV]])
 ; CHECK-NEXT:    [[SHL:%.*]] = shl i32 [[VAL:%.*]], [[CONV]]
 ; CHECK-NEXT:    [[SHR:%.*]] = ashr i32 [[SHL]], [[CONV]]
 ; CHECK-NEXT:    ret i32 [[SHR]]
 ;
   %val.addr = alloca i32, align 4
   %len.addr = alloca i8, align 1
   %wide_len = alloca i32, align 4
   store i32 %val, i32* %val.addr, align 4