This is an archive of the discontinued LLVM Phabricator instance.

Differential D74183

[IRGen] Add an alignment attribute to underaligned sret parameters
ClosedPublic

Authored by erik.pilkington on Feb 6 2020, 4:52 PM.

Details

Reviewers
rjmccall
scanon
aartbik
Commits
rGfe7314d3bb1b: [CodeGen] Add an alignment attribute to all sret parameters
rGde98cf92e301: [CodeGen] Add an alignment attribute to all sret parameters
Summary

Without this, LLVM will assume the pointer has the type's alignment, which is undefined if the value is actually underaligned.

rdar://58316406

Diff Detail

Event Timeline

erik.pilkington created this revision.Feb 6 2020, 4:52 PM
Herald added subscribers: ributzka, dexonsmith, jkorous. · View Herald TranscriptFeb 6 2020, 4:52 PM
rjmccall added inline comments.Feb 6 2020, 6:44 PM
clang/lib/CodeGen/CGCall.cpp
2077

Why only when under-aligned? Just to avoid churning tests? I think we should apply this unconditionally.

scanon added inline comments.Feb 7 2020, 8:18 AM
clang/lib/CodeGen/CGCall.cpp
2077

On mainstream architectures today, there's rarely a benefit to knowing about higher alignment (e.g. MOVUPS is just as fast as MOVAPS if the address is actually aligned), so we won't see significant perf wins from preserving over-alignment in most cases, but it also doesn't cost us anything AFAICT and could deliver wins in some specific cases (e.g. AVX on SNB and IVB, where I think we split underaligned 256b stores into two 128b chunks).

So, yeah, I think we ought to simply unconditionally add the alignment to the sret.

dexonsmith added inline comments.Feb 7 2020, 10:30 AM
clang/lib/CodeGen/CGCall.cpp
2077

@rjmccall, are you seeing a reason to add the attribute when the implicit one is correct (neither over-aligned nor under-aligned)? If so, it seems to me like the added noise would make the IR harder to read.

rjmccall added inline comments.Feb 7 2020, 10:50 AM
clang/lib/CodeGen/CGCall.cpp
2077

Well, first, I think we're going to end up needing an alignment there in all cases eventually because of opaque pointer types. But I also think it's just cleaner and more testable to add the attribute in all cases instead of leaving it off when the IR type happens to have the right alignment, which can be very sensitive to the target.

rjmccall added inline comments.Feb 7 2020, 10:57 AM
clang/lib/CodeGen/CGCall.cpp
2077

In general, I think frontends should *never* be leaving it up to LLVM to infer alignment based on IR types, and this is part-and-parcel with that.

dexonsmith added inline comments.Feb 7 2020, 10:58 AM
clang/lib/CodeGen/CGCall.cpp
2077

I think we're going to end up needing an alignment there in all cases eventually because of opaque pointer types.

That's a great point. In that case I don't have a strong opinion.

erik.pilkington updated this revision to Diff 251806.Mar 20 2020, 5:15 PM

Add alignment attribute to all sret parameters.

Herald added a reviewer: aartbik. · View Herald TranscriptMar 20 2020, 5:15 PM
Herald added subscribers: kerbowa, luismarques, apazos and 29 others. · View Herald Transcript
rjmccall accepted this revision.Mar 23 2020, 11:05 AM

I'm just going to assume the test changes look good; thank you for taking the time to do this.

This revision is now accepted and ready to land.Mar 23 2020, 11:05 AM
Herald added a subscriber: • wuzish. · View Herald TranscriptMar 23 2020, 11:05 AM
efriedma added a subscriber: efriedma.Mar 23 2020, 11:13 AM

I think we should remove the LangRef rule that says "sret" pointers have to be dereferenceable/naturally aligned, and let the frontend add explicit aligned/dereferenceable markings where appropriate. (At that point, sret has no target-independent meaning; it's just to manipulate the target ABI.) It would make the IR easier to understand, and resolves the interaction with opaque pointers.

That isn't to say we shouldn't make this change in clang; clang should do this, but at that point it would just be a performance enhancement, not required for correctness.

rjmccall added a comment.Mar 23 2020, 12:52 PM
In D74183#1937426, @efriedma wrote:

I think we should remove the LangRef rule that says "sret" pointers have to be dereferenceable/naturally aligned, and let the frontend add explicit aligned/dereferenceable markings where appropriate. (At that point, sret has no target-independent meaning; it's just to manipulate the target ABI.) It would make the IR easier to understand, and resolves the interaction with opaque pointers.

That might be reasonable, yeah. And yeah, making the dereferenceable assumption explicit is going to be necessary with opaque pointer types eventually.

Closed by commit rGde98cf92e301: [CodeGen] Add an alignment attribute to all sret parameters (authored by erik.pilkington). · Explain WhyMar 24 2020, 12:55 PM
This revision was automatically updated to reflect the committed changes.
Herald added a project: Restricted Project. · View Herald TranscriptMar 24 2020, 12:55 PM
Herald added a subscriber: jrtc27. · View Herald Transcript
nikic added a subscriber: nikic.Mar 24 2020, 1:55 PM

For the record, this change caused an 1.5% compile-time regression on tramp3d-v4 (http://llvm-compile-time-tracker.com/compare.php?from=43a6d285bfead762ac472a6e62beedc9f88bce89&to=de98cf92e301ab559a7417f1eca5cfa53624c9e1&stat=instructions). As there was also a 0.9% increase in code size, I assume that adding the alignment ends up having a non-trivial impact on optimization behavior for this benchmark and we end up generating more IR. So most likely this is fine and there's nothing to be done about it.

efriedma added a comment.Mar 25 2020, 10:49 AM

If it's just tramp3d-v4, I'm not that concerned... but that's a weird result. On x86 in particular, alignment markings have almost no effect on optimization, generally.

nikic added a comment.Mar 25 2020, 2:38 PM
In D74183#1941741, @efriedma wrote:

If it's just tramp3d-v4, I'm not that concerned... but that's a weird result. On x86 in particular, alignment markings have almost no effect on optimization, generally.

I've just looked at the IR diff for tramp3d-v4 and it turns out that the root cause is an old friend of mine: The insertion of alignment assumptions during inlining (https://github.com/llvm/llvm-project/blob/b58902bc72c2b479b5ed27ec0d3422ba9782edbb/llvm/lib/Transforms/Utils/InlineFunction.cpp#L1139-L1173). That is, the IR now contains many instances of this sequence:

%ptrint = ptrtoint %class.GuardLayers* %guards_m to i64
%maskedptr = and i64 %ptrint, 3
%maskcond = icmp eq i64 %maskedptr, 0
tail call void @llvm.assume(i1 %maskcond)

to preserve the alignment information. From a cursory look I cannot tell whether these additional assumes also regress optimization (due to multi-use), but given the size increase on the final binary it seems pretty likely that this is the case.

This preservation of alignment during inlining is the reason why we used to not emit alignment information for pointer arguments in Rust for a long time: It caused serious regressions in optimization and increased compile-time. Nowadays we do emit alignment information, but set -preserve-alignment-assumptions-during-inlining=false to prevent this inlining behavior.

I think for the purposes of this revision, this means that we should probably either a) default preserve-alignment-assumptions-during-inlining to false (I would prefer this) or b) not emit the alignment unless it is smaller than the ABI alignment (I guess this was what this patch originally did?)

rjmccall added a comment.Mar 25 2020, 3:06 PM

Yeah, if emitting alignment assumptions in inlining is causing regressions when frontends provide better information, those assumptions need to be reverted until they can be fixed.

efriedma added subscribers: hfinkel, jdoerfert.Mar 25 2020, 3:07 PM

That makes sense.

I would slightly lean towards not generating the assumptions, given the current state of assumptions and the likely benefit in this context.

nikic mentioned this in D76886: [InlineFunction] Disable emission of alignment assumptions by default.Mar 26 2020, 1:31 PM
hgreving added a subscriber: hgreving.EditedApr 1 2020, 4:28 PM
In D74183#1942262, @nikic wrote:
In D74183#1941741, @efriedma wrote:

If it's just tramp3d-v4, I'm not that concerned... but that's a weird result. On x86 in particular, alignment markings have almost no effect on optimization, generally.

I've just looked at the IR diff for tramp3d-v4 and it turns out that the root cause is an old friend of mine: The insertion of alignment assumptions during inlining (https://github.com/llvm/llvm-project/blob/b58902bc72c2b479b5ed27ec0d3422ba9782edbb/llvm/lib/Transforms/Utils/InlineFunction.cpp#L1139-L1173). That is, the IR now contains many instances of this sequence:

%ptrint = ptrtoint %class.GuardLayers* %guards_m to i64
%maskedptr = and i64 %ptrint, 3
%maskcond = icmp eq i64 %maskedptr, 0
tail call void @llvm.assume(i1 %maskcond)

to preserve the alignment information. From a cursory look I cannot tell whether these additional assumes also regress optimization (due to multi-use), but given the size increase on the final binary it seems pretty likely that this is the case.

This preservation of alignment during inlining is the reason why we used to not emit alignment information for pointer arguments in Rust for a long time: It caused serious regressions in optimization and increased compile-time. Nowadays we do emit alignment information, but set -preserve-alignment-assumptions-during-inlining=false to prevent this inlining behavior.

I think for the purposes of this revision, this means that we should probably either a) default preserve-alignment-assumptions-during-inlining to false (I would prefer this) or b) not emit the alignment unless it is smaller than the ABI alignment (I guess this was what this patch originally did?)

We are having a problem with this very issue on a target not supporting a stack, with sroa bailing due to above, in our case causing a crash. Our only workaround for this is currently preserve-alignment-assumptions-during-inlining to false. We were actually wondering if this is causing performance issues on targets that do support a stack.

nikic mentioned this in rGb74c6d2c9d8e: [InlineFunction] Disable emission of alignment assumptions by default.Apr 30 2020, 2:33 PM

Revision Contents

PathSize
clang/
lib/
CodeGen/
1 line
test/
CodeGen/
4 lines
4 lines
8 lines
10 lines
arc/
8 lines
2 lines
8 lines
144 lines
18 lines
6 lines
2 lines
10 lines
4 lines
34 lines
2 lines
108 lines
2 lines
4 lines
4 lines
4 lines
2 lines
4 lines
8 lines
2 lines
6 lines
4 lines
12 lines
2 lines
2 lines
46 lines
4 lines
12 lines
4 lines
2 lines
18 lines
2 lines
2 lines
6 lines
8 lines
4 lines
6 lines
2 lines
2 lines
6 lines
2 lines
2 lines
4 lines
124 lines
90 lines
4 lines
2 lines
4 lines
12 lines
4 lines
2 lines
18 lines
2 lines
2 lines
2 lines
6 lines
CodeGenCXX/
2 lines
20 lines
4 lines
8 lines
4 lines
4 lines
6 lines
14 lines
4 lines
4 lines
4 lines
8 lines
4 lines
64 lines
2 lines
8 lines
2 lines
2 lines
12 lines
4 lines
4 lines
6 lines
8 lines
2 lines
18 lines
8 lines
4 lines
CodeGenCoroutines/
10 lines
10 lines
CodeGenObjC/
4 lines
2 lines
4 lines
6 lines
8 lines
2 lines
CodeGenObjCXX/
2 lines
CodeGenOpenCL/
6 lines
6 lines
CodeGenOpenCLCXX/
4 lines
Modules/
2 lines

Diff 251806

clang/lib/CodeGen/CGCall.cpp

Show First 20 Lines • Show All 489 Lines • ▼ Show 20 Lines
bool hasUsedSRet = false; bool hasUsedSRet = false;
SmallVector<llvm::AttributeSet, 4> ArgAttrs(IRFunctionArgs.totalIRArgs()); SmallVector<llvm::AttributeSet, 4> ArgAttrs(IRFunctionArgs.totalIRArgs());
// Attach attributes to sret. // Attach attributes to sret.
if (IRFunctionArgs.hasSRetArg()) { if (IRFunctionArgs.hasSRetArg()) {
llvm::AttrBuilder SRETAttrs; llvm::AttrBuilder SRETAttrs;
SRETAttrs.addAttribute(llvm::Attribute::StructRet); SRETAttrs.addAttribute(llvm::Attribute::StructRet);
hasUsedSRet = true; hasUsedSRet = true;
rjmccallUnsubmitted
Not Done
ReplyInline Actions

Why only when under-aligned? Just to avoid churning tests? I think we should apply this unconditionally.

rjmccall: Why only when under-aligned? Just to avoid churning tests? I think we should apply this…
scanonUnsubmitted
Not Done
ReplyInline Actions

On mainstream architectures today, there's rarely a benefit to knowing about higher alignment (e.g. MOVUPS is just as fast as MOVAPS if the address is actually aligned), so we won't see significant perf wins from preserving over-alignment in most cases, but it also doesn't cost us anything AFAICT and could deliver wins in some specific cases (e.g. AVX on SNB and IVB, where I think we split underaligned 256b stores into two 128b chunks).

So, yeah, I think we ought to simply unconditionally add the alignment to the sret.

scanon: On mainstream architectures today, there's rarely a benefit to knowing about higher alignment…
dexonsmithUnsubmitted
Not Done
ReplyInline Actions

@rjmccall, are you seeing a reason to add the attribute when the implicit one is correct (neither over-aligned nor under-aligned)? If so, it seems to me like the added noise would make the IR harder to read.

dexonsmith: @rjmccall, are you seeing a reason to add the attribute when the implicit one is correct…
rjmccallUnsubmitted
Not Done
ReplyInline Actions

Well, first, I think we're going to end up needing an alignment there in all cases eventually because of opaque pointer types. But I also think it's just cleaner and more testable to add the attribute in all cases instead of leaving it off when the IR type happens to have the right alignment, which can be very sensitive to the target.

rjmccall: Well, first, I think we're going to end up needing an alignment there in all cases eventually…
rjmccallUnsubmitted
Not Done
ReplyInline Actions

In general, I think frontends should *never* be leaving it up to LLVM to infer alignment based on IR types, and this is part-and-parcel with that.

rjmccall: In general, I think frontends should *never* be leaving it up to LLVM to infer alignment based…
dexonsmithUnsubmitted
Not Done
ReplyInline Actions

I think we're going to end up needing an alignment there in all cases eventually because of opaque pointer types.

That's a great point. In that case I don't have a strong opinion.

dexonsmith: > I think we're going to end up needing an alignment there in all cases eventually because of…
if (RetAI.getInReg()) if (RetAI.getInReg())
SRETAttrs.addAttribute(llvm::Attribute::InReg); SRETAttrs.addAttribute(llvm::Attribute::InReg);
SRETAttrs.addAlignmentAttr(RetAI.getIndirectAlign().getQuantity());
ArgAttrs[IRFunctionArgs.getSRetArgNo()] = ArgAttrs[IRFunctionArgs.getSRetArgNo()] =
llvm::AttributeSet::get(getLLVMContext(), SRETAttrs); llvm::AttributeSet::get(getLLVMContext(), SRETAttrs);
} }
// Attach attributes to inalloca argument. // Attach attributes to inalloca argument.
if (IRFunctionArgs.hasInallocaArg()) { if (IRFunctionArgs.hasInallocaArg()) {
llvm::AttrBuilder Attrs; llvm::AttrBuilder Attrs;
Attrs.addAttribute(llvm::Attribute::InAlloca); Attrs.addAttribute(llvm::Attribute::InAlloca);
▲ Show 20 Lines • Show All 492 Lines • Show Last 20 Lines

clang/test/CodeGen/2006-05-19-SingleEltReturn.c

Show All 18 Lines
struct Y bar() { struct Y bar() {
struct Y a; struct Y a;
a.x.D = 0; a.x.D = 0;
return a; return a;
} }
// X86_32: define void @foo(%struct.Y* %P) // X86_32: define void @foo(%struct.Y* %P)
// X86_32: call void @bar(%struct.Y* sret %{{[^),]*}}) // X86_32: call void @bar(%struct.Y* sret align 4 %{{[^),]*}})
// X86_32: define void @bar(%struct.Y* noalias sret %{{[^,)]*}}) // X86_32: define void @bar(%struct.Y* noalias sret align 4 %{{[^,)]*}})
// X86_32: ret void // X86_32: ret void

clang/test/CodeGen/aarch64-varargs.c

Show First 20 Lines • Show All 492 Lines • ▼ Show 20 Lines
// CHECK: [[ADDR:%[a-z._0-9]+]] = phi %struct.underaligned_int128_struct* [ [[FROMREG_ADDR]], %[[VAARG_IN_REG]] ], [ [[FROMSTACK_ADDR]], %[[VAARG_ON_STACK]] ] // CHECK: [[ADDR:%[a-z._0-9]+]] = phi %struct.underaligned_int128_struct* [ [[FROMREG_ADDR]], %[[VAARG_IN_REG]] ], [ [[FROMSTACK_ADDR]], %[[VAARG_ON_STACK]] ]
} }
// Overaligning to 32 bytes causes it to be passed indirectly via a pointer // Overaligning to 32 bytes causes it to be passed indirectly via a pointer
typedef struct __attribute__((aligned(32))) { typedef struct __attribute__((aligned(32))) {
__int128 val; __int128 val;
} overaligned_int128_struct; } overaligned_int128_struct;
overaligned_int128_struct overaligned_int128_struct_test() { overaligned_int128_struct overaligned_int128_struct_test() {
// CHECK-LABEL: define void @overaligned_int128_struct_test(%struct.overaligned_int128_struct* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_int128_struct_test(%struct.overaligned_int128_struct* noalias sret align 32 %agg.result)
return va_arg(the_list, overaligned_int128_struct); return va_arg(the_list, overaligned_int128_struct);
// CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3) // CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3)
// CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0 // CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0
// CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK:[a-z_.0-9]+]], label %[[VAARG_MAYBE_REG:[a-z_.0-9]+]] // CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK:[a-z_.0-9]+]], label %[[VAARG_MAYBE_REG:[a-z_.0-9]+]]
// CHECK: [[VAARG_MAYBE_REG]] // CHECK: [[VAARG_MAYBE_REG]]
// CHECK: [[NEW_REG_OFFS:%[a-z_0-9]+]] = add i32 [[GR_OFFS]], 8 // CHECK: [[NEW_REG_OFFS:%[a-z_0-9]+]] = add i32 [[GR_OFFS]], 8
// CHECK: store i32 [[NEW_REG_OFFS]], i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3) // CHECK: store i32 [[NEW_REG_OFFS]], i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3)
▲ Show 20 Lines • Show All 197 Lines • ▼ Show 20 Lines
// CHECK: [[ADDR:%[a-z._0-9]+]] = phi %struct.underaligned_int128_struct_member* [ [[FROMREG_ADDR]], %[[VAARG_IN_REG]] ], [ [[FROMSTACK_ADDR]], %[[VAARG_ON_STACK]] ] // CHECK: [[ADDR:%[a-z._0-9]+]] = phi %struct.underaligned_int128_struct_member* [ [[FROMREG_ADDR]], %[[VAARG_IN_REG]] ], [ [[FROMSTACK_ADDR]], %[[VAARG_ON_STACK]] ]
} }
// Overaligning to 32 bytes causes it to be passed indirectly via a pointer // Overaligning to 32 bytes causes it to be passed indirectly via a pointer
typedef struct { typedef struct {
__int128 val __attribute__((aligned(32))); __int128 val __attribute__((aligned(32)));
} overaligned_int128_struct_member; } overaligned_int128_struct_member;
overaligned_int128_struct_member overaligned_int128_struct_member_test() { overaligned_int128_struct_member overaligned_int128_struct_member_test() {
// CHECK-LABEL: define void @overaligned_int128_struct_member_test(%struct.overaligned_int128_struct_member* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_int128_struct_member_test(%struct.overaligned_int128_struct_member* noalias sret align 32 %agg.result)
return va_arg(the_list, overaligned_int128_struct_member); return va_arg(the_list, overaligned_int128_struct_member);
// CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3) // CHECK: [[GR_OFFS:%[a-z_0-9]+]] = load i32, i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3)
// CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0 // CHECK: [[EARLY_ONSTACK:%[a-z_0-9]+]] = icmp sge i32 [[GR_OFFS]], 0
// CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK:[a-z_.0-9]+]], label %[[VAARG_MAYBE_REG:[a-z_.0-9]+]] // CHECK: br i1 [[EARLY_ONSTACK]], label %[[VAARG_ON_STACK:[a-z_.0-9]+]], label %[[VAARG_MAYBE_REG:[a-z_.0-9]+]]
// CHECK: [[VAARG_MAYBE_REG]] // CHECK: [[VAARG_MAYBE_REG]]
// CHECK: [[NEW_REG_OFFS:%[a-z_0-9]+]] = add i32 [[GR_OFFS]], 8 // CHECK: [[NEW_REG_OFFS:%[a-z_0-9]+]] = add i32 [[GR_OFFS]], 8
// CHECK: store i32 [[NEW_REG_OFFS]], i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3) // CHECK: store i32 [[NEW_REG_OFFS]], i32* getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 3)
Show All 31 Lines

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

Show First 20 Lines • Show All 56 Lines • ▼ Show 20 Lines do {
// O1-LEGACY: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* %[[P]]) // O1-LEGACY: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* %[[P]])
// O1-NEWPM: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* nonnull %[[P]]) // O1-NEWPM: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* nonnull %[[P]])
// //
// O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP1_ALLOCA]] to i8* // O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP1_ALLOCA]] to i8*
// O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]]) // O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]])
// O1-NEWPM: %[[TMP3:.*]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8* // O1-NEWPM: %[[TMP3:.*]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8*
// O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[P]]) // O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[P]])
// //
// O1-LEGACY: call void @foo_int(%struct.S* sret %[[TMP1_ALLOCA]], // O1-LEGACY: call void @foo_int(%struct.S* sret align 4 %[[TMP1_ALLOCA]],
// O1-NEWPM: call void @foo_int(%struct.S* nonnull sret %[[TMP1_ALLOCA]], // O1-NEWPM: call void @foo_int(%struct.S* nonnull sret align 4 %[[TMP1_ALLOCA]],
// O1: call void @llvm.memcpy // O1: call void @llvm.memcpy
// O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP1_ALLOCA]] to i8* // O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP1_ALLOCA]] to i8*
// O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]]) // O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]])
// O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[P]]) // O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[P]])
// O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8* // O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8*
// O1-LEGACY: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* %[[P]]) // O1-LEGACY: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* %[[P]])
// O1-NEWPM: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* nonnull %[[TMP3]]) // O1-NEWPM: call void @llvm.lifetime.start.p0i8({{[^,]*}}, i8* nonnull %[[TMP3]])
// O1-LEGACY: call void @foo_int(%struct.S* sret %[[TMP2_ALLOCA]], // O1-LEGACY: call void @foo_int(%struct.S* sret align 4 %[[TMP2_ALLOCA]],
// O1-NEWPM: call void @foo_int(%struct.S* nonnull sret %[[TMP2_ALLOCA]], // O1-NEWPM: call void @foo_int(%struct.S* nonnull sret align 4 %[[TMP2_ALLOCA]],
// O1: call void @llvm.memcpy // O1: call void @llvm.memcpy
// O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8* // O1-LEGACY: %[[P:[^ ]+]] = bitcast %struct.S* %[[TMP2_ALLOCA]] to i8*
// O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]]) // O1-LEGACY: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* %[[P]])
// O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[TMP3]]) // O1-NEWPM: call void @llvm.lifetime.end.p0i8({{[^,]*}}, i8* nonnull %[[TMP3]])
r = foo_int(({ r = foo_int(({
if (*j) if (*j)
break; break;
i++; i++;
})); }));
r = foo_int(i++); r = foo_int(i++);
} while (1); } while (1);
return r; return r;
} }

clang/test/CodeGen/aligned-sret.c

  • This file was added.
// RUN: %clang_cc1 -triple x86_64-apple-macos %s -S -emit-llvm -o- | FileCheck %s
typedef __attribute__((__ext_vector_type__(4),__aligned__(16))) double simd_double4;
typedef struct { simd_double4 columns[4]; } simd_double4x4;
typedef simd_double4x4 matrix_double4x4;
// CHECK: define void @ident(%struct.simd_double4x4* noalias sret align 16 %agg.result
matrix_double4x4 ident(matrix_double4x4 x) {
return x;
}

clang/test/CodeGen/arc/arguments.c

Show All 16 Linestypedef struct {
char aa; char bb; char cc; char dd; char aa; char bb; char cc; char dd;
} cs1; } cs1;
// CHECK: define void @cf1(i32 inreg %i.coerce) // CHECK: define void @cf1(i32 inreg %i.coerce)
void cf1(cs1 i) {} void cf1(cs1 i) {}
typedef struct { typedef struct {
int cc; int cc;
} s2; } s2;
// CHECK: define void @f2(%struct.s2* noalias sret %agg.result) // CHECK: define void @f2(%struct.s2* noalias sret align 4 %agg.result)
s2 f2() { s2 f2() {
s2 foo; s2 foo;
return foo; return foo;
} }
typedef struct { typedef struct {
int cc; int cc;
int dd; int dd;
} s3; } s3;
// CHECK: define void @f3(%struct.s3* noalias sret %agg.result) // CHECK: define void @f3(%struct.s3* noalias sret align 4 %agg.result)
s3 f3() { s3 f3() {
s3 foo; s3 foo;
return foo; return foo;
} }
// CHECK: define void @f4(i64 inreg %i) // CHECK: define void @f4(i64 inreg %i)
void f4(long long i) {} void f4(long long i) {}
▲ Show 20 Lines • Show All 79 Lines • ▼ Show 20 Lines
// CHECK: define void @st2(i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %a.coerce2, i32 inreg %a.coerce3, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3) // CHECK: define void @st2(i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %a.coerce2, i32 inreg %a.coerce3, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3)
// Use 8-byte struct 3 times, gets 8 registers, 1 byval struct argument. // Use 8-byte struct 3 times, gets 8 registers, 1 byval struct argument.
void st3(s16 a, s16 b, s16 c) {} void st3(s16 a, s16 b, s16 c) {}
// CHECK: define void @st3(i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %a.coerce2, i32 inreg %a.coerce3, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce) // CHECK: define void @st3(i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %a.coerce2, i32 inreg %a.coerce3, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
// 1 sret + 1 i32 + 2*(i32 coerce) + 4*(i32 coerce) + 1 byval // 1 sret + 1 i32 + 2*(i32 coerce) + 4*(i32 coerce) + 1 byval
s16 st4(int x, s8 a, s16 b, s16 c) { return b; } s16 st4(int x, s8 a, s16 b, s16 c) { return b; }
// CHECK: define void @st4(%struct.s16* noalias sret %agg.result, i32 inreg %x, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce) // CHECK: define void @st4(%struct.s16* noalias sret align 4 %agg.result, i32 inreg %x, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)
// 1 sret + 2*(i32 coerce) + 4*(i32 coerce) + 4*(i32 coerce) // 1 sret + 2*(i32 coerce) + 4*(i32 coerce) + 4*(i32 coerce)
s16 st5(s8 a, s16 b, s16 c) { return b; } s16 st5(s8 a, s16 b, s16 c) { return b; }
// CHECK: define void @st5(%struct.s16* noalias sret %agg.result, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce) // CHECK: define void @st5(%struct.s16* noalias sret align 4 %agg.result, i32 inreg %a.coerce0, i32 inreg %a.coerce1, i32 inreg %b.coerce0, i32 inreg %b.coerce1, i32 inreg %b.coerce2, i32 inreg %b.coerce3, { i32, i32, i32, i32 } %c.coerce)

clang/test/CodeGen/arm-aapcs-vfp.c

Show First 20 Lines • Show All 119 Lines • ▼ Show 20 Lines
typedef struct { long long x; int y; } struct_long_long_int; typedef struct { long long x; int y; } struct_long_long_int;
// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_1(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, i64 %k, i32 %l) // CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_1(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, i64 %k, i32 %l)
void test_vfp_stack_gpr_split_1(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, long long k, int l) {} void test_vfp_stack_gpr_split_1(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, long long k, int l) {}
// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_2(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, [2 x i64] %k.coerce) // CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_2(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, [2 x i64] %k.coerce)
void test_vfp_stack_gpr_split_2(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, struct_long_long_int k) {} void test_vfp_stack_gpr_split_2(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, struct_long_long_int k) {}
// CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_3(%struct.struct_long_long_int* noalias sret %agg.result, double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, [2 x i64] %k.coerce) // CHECK: define arm_aapcs_vfpcc void @test_vfp_stack_gpr_split_3(%struct.struct_long_long_int* noalias sret align 8 %agg.result, double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, [2 x i64] %k.coerce)
struct_long_long_int test_vfp_stack_gpr_split_3(double a, double b, double c, double d, double e, double f, double g, double h, double i, struct_long_long_int k) {} struct_long_long_int test_vfp_stack_gpr_split_3(double a, double b, double c, double d, double e, double f, double g, double h, double i, struct_long_long_int k) {}
typedef struct { int a; int b:4; int c; } struct_int_bitfield_int; typedef struct { int a; int b:4; int c; } struct_int_bitfield_int;
// CHECK: define arm_aapcs_vfpcc void @test_test_vfp_stack_gpr_split_bitfield(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, i32 %k, [3 x i32] %l.coerce) // CHECK: define arm_aapcs_vfpcc void @test_test_vfp_stack_gpr_split_bitfield(double %a, double %b, double %c, double %d, double %e, double %f, double %g, double %h, double %i, i32 %j, i32 %k, [3 x i32] %l.coerce)
void test_test_vfp_stack_gpr_split_bitfield(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, int k, struct_int_bitfield_int l) {} void test_test_vfp_stack_gpr_split_bitfield(double a, double b, double c, double d, double e, double f, double g, double h, double i, int j, int k, struct_int_bitfield_int l) {}
// Note: this struct requires internal padding // Note: this struct requires internal padding
typedef struct { int x; long long y; } struct_int_long_long; typedef struct { int x; long long y; } struct_int_long_long;
Show All 13 Lines

clang/test/CodeGen/arm-homogenous.c

Show All 21 Lines
void test_union_with_first_floats(void) { void test_union_with_first_floats(void) {
takes_union_with_first_floats(g_u_f); takes_union_with_first_floats(g_u_f);
} }
// CHECK: declare arm_aapcs_vfpcc void @takes_union_with_first_floats([4 x i32]) // CHECK: declare arm_aapcs_vfpcc void @takes_union_with_first_floats([4 x i32])
void test_return_union_with_first_floats(void) { void test_return_union_with_first_floats(void) {
g_u_f = returns_union_with_first_floats(); g_u_f = returns_union_with_first_floats();
} }
// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_first_floats(%union.union_with_first_floats* sret) // CHECK: declare arm_aapcs_vfpcc void @returns_union_with_first_floats(%union.union_with_first_floats* sret align 4)
/* This is not a homogenous aggregate - fundamental types are different */ /* This is not a homogenous aggregate - fundamental types are different */
typedef union { typedef union {
uint32_t i[4]; uint32_t i[4];
float f[4]; float f[4];
} union_with_non_first_floats; } union_with_non_first_floats;
union_with_non_first_floats g_u_nf_f; union_with_non_first_floats g_u_nf_f;
extern void takes_union_with_non_first_floats(union_with_non_first_floats a); extern void takes_union_with_non_first_floats(union_with_non_first_floats a);
extern union_with_non_first_floats returns_union_with_non_first_floats(void); extern union_with_non_first_floats returns_union_with_non_first_floats(void);
void test_union_with_non_first_floats(void) { void test_union_with_non_first_floats(void) {
takes_union_with_non_first_floats(g_u_nf_f); takes_union_with_non_first_floats(g_u_nf_f);
} }
// CHECK: declare arm_aapcs_vfpcc void @takes_union_with_non_first_floats([4 x i32]) // CHECK: declare arm_aapcs_vfpcc void @takes_union_with_non_first_floats([4 x i32])
void test_return_union_with_non_first_floats(void) { void test_return_union_with_non_first_floats(void) {
g_u_nf_f = returns_union_with_non_first_floats(); g_u_nf_f = returns_union_with_non_first_floats();
} }
// CHECK: declare arm_aapcs_vfpcc void @returns_union_with_non_first_floats(%union.union_with_non_first_floats* sret) // CHECK: declare arm_aapcs_vfpcc void @returns_union_with_non_first_floats(%union.union_with_non_first_floats* sret align 4)
/* This is not a homogenous aggregate - fundamental types are different */ /* This is not a homogenous aggregate - fundamental types are different */
typedef struct { typedef struct {
float a; float a;
union_with_first_floats b; union_with_first_floats b;
} struct_with_union_with_first_floats; } struct_with_union_with_first_floats;
struct_with_union_with_first_floats g_s_f; struct_with_union_with_first_floats g_s_f;
extern void takes_struct_with_union_with_first_floats(struct_with_union_with_first_floats a); extern void takes_struct_with_union_with_first_floats(struct_with_union_with_first_floats a);
extern struct_with_union_with_first_floats returns_struct_with_union_with_first_floats(void); extern struct_with_union_with_first_floats returns_struct_with_union_with_first_floats(void);
void test_struct_with_union_with_first_floats(void) { void test_struct_with_union_with_first_floats(void) {
takes_struct_with_union_with_first_floats(g_s_f); takes_struct_with_union_with_first_floats(g_s_f);
} }
// CHECK: declare arm_aapcs_vfpcc void @takes_struct_with_union_with_first_floats([5 x i32]) // CHECK: declare arm_aapcs_vfpcc void @takes_struct_with_union_with_first_floats([5 x i32])
void test_return_struct_with_union_with_first_floats(void) { void test_return_struct_with_union_with_first_floats(void) {
g_s_f = returns_struct_with_union_with_first_floats(); g_s_f = returns_struct_with_union_with_first_floats();
} }
// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_first_floats(%struct.struct_with_union_with_first_floats* sret) // CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_first_floats(%struct.struct_with_union_with_first_floats* sret align 4)
/* This is not a homogenous aggregate - fundamental types are different */ /* This is not a homogenous aggregate - fundamental types are different */
typedef struct { typedef struct {
float a; float a;
union_with_non_first_floats b; union_with_non_first_floats b;
} struct_with_union_with_non_first_floats; } struct_with_union_with_non_first_floats;
struct_with_union_with_non_first_floats g_s_nf_f; struct_with_union_with_non_first_floats g_s_nf_f;
extern void takes_struct_with_union_with_non_first_floats(struct_with_union_with_non_first_floats a); extern void takes_struct_with_union_with_non_first_floats(struct_with_union_with_non_first_floats a);
extern struct_with_union_with_non_first_floats returns_struct_with_union_with_non_first_floats(void); extern struct_with_union_with_non_first_floats returns_struct_with_union_with_non_first_floats(void);
void test_struct_with_union_with_non_first_floats(void) { void test_struct_with_union_with_non_first_floats(void) {
takes_struct_with_union_with_non_first_floats(g_s_nf_f); takes_struct_with_union_with_non_first_floats(g_s_nf_f);
} }
// CHECK: declare arm_aapcs_vfpcc void @takes_struct_with_union_with_non_first_floats([5 x i32]) // CHECK: declare arm_aapcs_vfpcc void @takes_struct_with_union_with_non_first_floats([5 x i32])
void test_return_struct_with_union_with_non_first_floats(void) { void test_return_struct_with_union_with_non_first_floats(void) {
g_s_nf_f = returns_struct_with_union_with_non_first_floats(); g_s_nf_f = returns_struct_with_union_with_non_first_floats();
} }
// CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_non_first_floats(%struct.struct_with_union_with_non_first_floats* sret) // CHECK: declare arm_aapcs_vfpcc void @returns_struct_with_union_with_non_first_floats(%struct.struct_with_union_with_non_first_floats* sret align 4)
/* Plain array is not a homogenous aggregate */ /* Plain array is not a homogenous aggregate */
extern void takes_array_of_floats(float a[4]); extern void takes_array_of_floats(float a[4]);
void test_array_of_floats(void) { void test_array_of_floats(void) {
float a[4] = {1.0, 2.0, 3.0, 4.0}; float a[4] = {1.0, 2.0, 3.0, 4.0};
takes_array_of_floats(a); takes_array_of_floats(a);
} }
// CHECK: declare arm_aapcs_vfpcc void @takes_array_of_floats(float*) // CHECK: declare arm_aapcs_vfpcc void @takes_array_of_floats(float*)
▲ Show 20 Lines • Show All 132 Lines • Show Last 20 Lines

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

// RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \ // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
// RUN: -S -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | \ // RUN: -S -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | \
// RUN: FileCheck -check-prefixes=CHECK,CHECK-A64 %s // RUN: FileCheck -check-prefixes=CHECK,CHECK-A64 %s
// RUN: %clang_cc1 -triple armv8-none-linux-gnueabi -target-feature +neon \ // RUN: %clang_cc1 -triple armv8-none-linux-gnueabi -target-feature +neon \
// RUN: -target-feature +fp16 -S -disable-O0-optnone -emit-llvm -o - %s | \ // RUN: -target-feature +fp16 -S -disable-O0-optnone -emit-llvm -o - %s | \
// RUN: opt -S -mem2reg | FileCheck -check-prefixes=CHECK,CHECK-A32 %s // RUN: opt -S -mem2reg | FileCheck -check-prefixes=CHECK,CHECK-A32 %s
#include <arm_neon.h> #include <arm_neon.h>
// CHECK-LABEL: @test_vld1_f16_x2( // CHECK-LABEL: @test_vld1_f16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x2_t, align 8
// CHECK-A32: %struct.float16x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x4x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF:(half|i16)]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF:(half|i16)]]*
// CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v4f16.p0f16|arm.neon.vld1x2.v4i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v4f16.p0f16|arm.neon.vld1x2.v4i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]> }*
// CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]] // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x2_t, %struct.float16x4x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x2_t, %struct.float16x4x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float16x4x2_t [[TMP6]] // CHECK-A64: ret %struct.float16x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x4x2_t test_vld1_f16_x2(float16_t const *a) { float16x4x2_t test_vld1_f16_x2(float16_t const *a) {
return vld1_f16_x2(a); return vld1_f16_x2(a);
} }
// CHECK-LABEL: @test_vld1_f16_x3( // CHECK-LABEL: @test_vld1_f16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x3_t, align 8
// CHECK-A32: %struct.float16x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x4x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]*
// CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v4f16.p0f16|arm.neon.vld1x3.v4i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v4f16.p0f16|arm.neon.vld1x3.v4i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }*
// CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]] // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x3_t, %struct.float16x4x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x3_t, %struct.float16x4x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float16x4x3_t [[TMP6]] // CHECK-A64: ret %struct.float16x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x4x3_t test_vld1_f16_x3(float16_t const *a) { float16x4x3_t test_vld1_f16_x3(float16_t const *a) {
return vld1_f16_x3(a); return vld1_f16_x3(a);
} }
// CHECK-LABEL: @test_vld1_f16_x4( // CHECK-LABEL: @test_vld1_f16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x4x4_t, align 8
// CHECK-A32: %struct.float16x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x4x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float16x4x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]*
// CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v4f16.p0f16|arm.neon.vld1x4.v4i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v4f16.p0f16|arm.neon.vld1x4.v4i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }*
// CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]] // CHECK: store { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> } [[VLD1XN]], { <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]>, <4 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x4_t, %struct.float16x4x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x4x4_t, %struct.float16x4x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float16x4x4_t [[TMP6]] // CHECK-A64: ret %struct.float16x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x4x4_t test_vld1_f16_x4(float16_t const *a) { float16x4x4_t test_vld1_f16_x4(float16_t const *a) {
return vld1_f16_x4(a); return vld1_f16_x4(a);
} }
// CHECK-LABEL: @test_vld1_f32_x2( // CHECK-LABEL: @test_vld1_f32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8
// CHECK-A32: %struct.float32x2x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x2x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x2x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float> }*
// CHECK: store { <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float> }* [[TMP3]] // CHECK: store { <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float32x2x2_t [[TMP6]] // CHECK-A64: ret %struct.float32x2x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float32x2x2_t test_vld1_f32_x2(float32_t const *a) { float32x2x2_t test_vld1_f32_x2(float32_t const *a) {
return vld1_f32_x2(a); return vld1_f32_x2(a);
} }
// CHECK-LABEL: @test_vld1_f32_x3( // CHECK-LABEL: @test_vld1_f32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x3_t, align 8
// CHECK-A32: %struct.float32x2x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x2x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x2x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float>, <2 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float>, <2 x float> }*
// CHECK: store { <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float>, <2 x float> }* [[TMP3]] // CHECK: store { <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float>, <2 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x3_t, %struct.float32x2x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x3_t, %struct.float32x2x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float32x2x3_t [[TMP6]] // CHECK-A64: ret %struct.float32x2x3_t [[TMP6]]
float32x2x3_t test_vld1_f32_x3(float32_t const *a) { float32x2x3_t test_vld1_f32_x3(float32_t const *a) {
return vld1_f32_x3(a); return vld1_f32_x3(a);
} }
// CHECK-LABEL: @test_vld1_f32_x4( // CHECK-LABEL: @test_vld1_f32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x2x4_t, align 8
// CHECK-A32: %struct.float32x2x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x2x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x2x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.float32x2x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x float>, <2 x float>, <2 x float>, <2 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float>, <2 x float>, <2 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x float>, <2 x float>, <2 x float>, <2 x float> }*
// CHECK: store { <2 x float>, <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float>, <2 x float>, <2 x float> }* [[TMP3]] // CHECK: store { <2 x float>, <2 x float>, <2 x float>, <2 x float> } [[VLD1XN]], { <2 x float>, <2 x float>, <2 x float>, <2 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x2x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x4_t, %struct.float32x2x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x2x4_t, %struct.float32x2x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.float32x2x4_t [[TMP6]] // CHECK-A64: ret %struct.float32x2x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float32x2x4_t test_vld1_f32_x4(float32_t const *a) { float32x2x4_t test_vld1_f32_x4(float32_t const *a) {
return vld1_f32_x4(a); return vld1_f32_x4(a);
} }
// CHECK-LABEL: @test_vld1_p16_x2( // CHECK-LABEL: @test_vld1_p16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8
// CHECK-A32: %struct.poly16x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly16x4x2_t [[TMP6]] // CHECK-A64: ret %struct.poly16x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x4x2_t test_vld1_p16_x2(poly16_t const *a) { poly16x4x2_t test_vld1_p16_x2(poly16_t const *a) {
return vld1_p16_x2(a); return vld1_p16_x2(a);
} }
// CHECK-LABEL: @test_vld1_p16_x3( // CHECK-LABEL: @test_vld1_p16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x3_t, align 8
// CHECK-A32: %struct.poly16x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x3_t, %struct.poly16x4x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x3_t, %struct.poly16x4x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly16x4x3_t [[TMP6]] // CHECK-A64: ret %struct.poly16x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x4x3_t test_vld1_p16_x3(poly16_t const *a) { poly16x4x3_t test_vld1_p16_x3(poly16_t const *a) {
return vld1_p16_x3(a); return vld1_p16_x3(a);
} }
// CHECK-LABEL: @test_vld1_p16_x4( // CHECK-LABEL: @test_vld1_p16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x4x4_t, align 8
// CHECK-A32: %struct.poly16x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly16x4x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x4_t, %struct.poly16x4x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x4x4_t, %struct.poly16x4x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly16x4x4_t [[TMP6]] // CHECK-A64: ret %struct.poly16x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x4x4_t test_vld1_p16_x4(poly16_t const *a) { poly16x4x4_t test_vld1_p16_x4(poly16_t const *a) {
return vld1_p16_x4(a); return vld1_p16_x4(a);
} }
// CHECK-LABEL: @test_vld1_p8_x2( // CHECK-LABEL: @test_vld1_p8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8
// CHECK-A32: %struct.poly8x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly8x8x2_t [[TMP4]] // CHECK-A64: ret %struct.poly8x8x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x8x2_t test_vld1_p8_x2(poly8_t const *a) { poly8x8x2_t test_vld1_p8_x2(poly8_t const *a) {
return vld1_p8_x2(a); return vld1_p8_x2(a);
} }
// CHECK-LABEL: @test_vld1_p8_x3( // CHECK-LABEL: @test_vld1_p8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x3_t, align 8
// CHECK-A32: %struct.poly8x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x3_t, %struct.poly8x8x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x3_t, %struct.poly8x8x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly8x8x3_t [[TMP4]] // CHECK-A64: ret %struct.poly8x8x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x8x3_t test_vld1_p8_x3(poly8_t const *a) { poly8x8x3_t test_vld1_p8_x3(poly8_t const *a) {
return vld1_p8_x3(a); return vld1_p8_x3(a);
} }
// CHECK-LABEL: @test_vld1_p8_x4( // CHECK-LABEL: @test_vld1_p8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x8x4_t, align 8
// CHECK-A32: %struct.poly8x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.poly8x8x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x4_t, %struct.poly8x8x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x8x4_t, %struct.poly8x8x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.poly8x8x4_t [[TMP4]] // CHECK-A64: ret %struct.poly8x8x4_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x8x4_t test_vld1_p8_x4(poly8_t const *a) { poly8x8x4_t test_vld1_p8_x4(poly8_t const *a) {
return vld1_p8_x4(a); return vld1_p8_x4(a);
} }
// CHECK-LABEL: @test_vld1_s16_x2( // CHECK-LABEL: @test_vld1_s16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8
// CHECK-A32: %struct.int16x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x4x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int16x4x2_t [[TMP6]] // CHECK-A64: ret %struct.int16x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x4x2_t test_vld1_s16_x2(int16_t const *a) { int16x4x2_t test_vld1_s16_x2(int16_t const *a) {
return vld1_s16_x2(a); return vld1_s16_x2(a);
} }
// CHECK-LABEL: @test_vld1_s16_x3( // CHECK-LABEL: @test_vld1_s16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x3_t, align 8
// CHECK-A32: %struct.int16x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x4x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x3_t, %struct.int16x4x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x3_t, %struct.int16x4x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int16x4x3_t [[TMP6]] // CHECK-A64: ret %struct.int16x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x4x3_t test_vld1_s16_x3(int16_t const *a) { int16x4x3_t test_vld1_s16_x3(int16_t const *a) {
return vld1_s16_x3(a); return vld1_s16_x3(a);
} }
// CHECK-LABEL: @test_vld1_s16_x4( // CHECK-LABEL: @test_vld1_s16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x4x4_t, align 8
// CHECK-A32: %struct.int16x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x4x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int16x4x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x4_t, %struct.int16x4x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x4x4_t, %struct.int16x4x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int16x4x4_t [[TMP6]] // CHECK-A64: ret %struct.int16x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x4x4_t test_vld1_s16_x4(int16_t const *a) { int16x4x4_t test_vld1_s16_x4(int16_t const *a) {
return vld1_s16_x4(a); return vld1_s16_x4(a);
} }
// CHECK-LABEL: @test_vld1_s32_x2( // CHECK-LABEL: @test_vld1_s32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8
// CHECK-A32: %struct.int32x2x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x2x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x2x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int32x2x2_t [[TMP6]] // CHECK-A64: ret %struct.int32x2x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x2x2_t test_vld1_s32_x2(int32_t const *a) { int32x2x2_t test_vld1_s32_x2(int32_t const *a) {
return vld1_s32_x2(a); return vld1_s32_x2(a);
} }
// CHECK-LABEL: @test_vld1_s32_x3( // CHECK-LABEL: @test_vld1_s32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x3_t, align 8
// CHECK-A32: %struct.int32x2x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x2x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x2x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x3_t, %struct.int32x2x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x3_t, %struct.int32x2x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int32x2x3_t [[TMP6]] // CHECK-A64: ret %struct.int32x2x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x2x3_t test_vld1_s32_x3(int32_t const *a) { int32x2x3_t test_vld1_s32_x3(int32_t const *a) {
return vld1_s32_x3(a); return vld1_s32_x3(a);
} }
// CHECK-LABEL: @test_vld1_s32_x4( // CHECK-LABEL: @test_vld1_s32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x2x4_t, align 8
// CHECK-A32: %struct.int32x2x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x2x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x2x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int32x2x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x2x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x4_t, %struct.int32x2x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x2x4_t, %struct.int32x2x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int32x2x4_t [[TMP6]] // CHECK-A64: ret %struct.int32x2x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x2x4_t test_vld1_s32_x4(int32_t const *a) { int32x2x4_t test_vld1_s32_x4(int32_t const *a) {
return vld1_s32_x4(a); return vld1_s32_x4(a);
} }
// CHECK-LABEL: @test_vld1_s64_x2( // CHECK-LABEL: @test_vld1_s64_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x2_t, align 8
// CHECK-A32: %struct.int64x1x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x1x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x1x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x2_t, %struct.int64x1x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x2_t, %struct.int64x1x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int64x1x2_t [[TMP6]] // CHECK-A64: ret %struct.int64x1x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x1x2_t test_vld1_s64_x2(int64_t const *a) { int64x1x2_t test_vld1_s64_x2(int64_t const *a) {
return vld1_s64_x2(a); return vld1_s64_x2(a);
} }
// CHECK-LABEL: @test_vld1_s64_x3( // CHECK-LABEL: @test_vld1_s64_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x3_t, align 8
// CHECK-A32: %struct.int64x1x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x1x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x1x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x3_t, %struct.int64x1x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x3_t, %struct.int64x1x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int64x1x3_t [[TMP6]] // CHECK-A64: ret %struct.int64x1x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x1x3_t test_vld1_s64_x3(int64_t const *a) { int64x1x3_t test_vld1_s64_x3(int64_t const *a) {
return vld1_s64_x3(a); return vld1_s64_x3(a);
} }
// CHECK-LABEL: @test_vld1_s64_x4( // CHECK-LABEL: @test_vld1_s64_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x1x4_t, align 8
// CHECK-A32: %struct.int64x1x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x1x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x1x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int64x1x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x1x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x1x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x1x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x4_t, %struct.int64x1x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x1x4_t, %struct.int64x1x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int64x1x4_t [[TMP6]] // CHECK-A64: ret %struct.int64x1x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x1x4_t test_vld1_s64_x4(int64_t const *a) { int64x1x4_t test_vld1_s64_x4(int64_t const *a) {
return vld1_s64_x4(a); return vld1_s64_x4(a);
} }
// CHECK-LABEL: @test_vld1_s8_x2( // CHECK-LABEL: @test_vld1_s8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8
// CHECK-A32: %struct.int8x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x8x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int8x8x2_t [[TMP4]] // CHECK-A64: ret %struct.int8x8x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x8x2_t test_vld1_s8_x2(int8_t const *a) { int8x8x2_t test_vld1_s8_x2(int8_t const *a) {
return vld1_s8_x2(a); return vld1_s8_x2(a);
} }
// CHECK-LABEL: @test_vld1_s8_x3( // CHECK-LABEL: @test_vld1_s8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x3_t, align 8
// CHECK-A32: %struct.int8x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x8x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x3_t, %struct.int8x8x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x3_t, %struct.int8x8x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int8x8x3_t [[TMP4]] // CHECK-A64: ret %struct.int8x8x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x8x3_t test_vld1_s8_x3(int8_t const *a) { int8x8x3_t test_vld1_s8_x3(int8_t const *a) {
return vld1_s8_x3(a); return vld1_s8_x3(a);
} }
// CHECK-LABEL: @test_vld1_s8_x4( // CHECK-LABEL: @test_vld1_s8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x8x4_t, align 8
// CHECK-A32: %struct.int8x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x8x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.int8x8x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x4_t, %struct.int8x8x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x8x4_t, %struct.int8x8x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.int8x8x4_t [[TMP4]] // CHECK-A64: ret %struct.int8x8x4_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x8x4_t test_vld1_s8_x4(int8_t const *a) { int8x8x4_t test_vld1_s8_x4(int8_t const *a) {
return vld1_s8_x4(a); return vld1_s8_x4(a);
} }
// CHECK-LABEL: @test_vld1_u16_x2( // CHECK-LABEL: @test_vld1_u16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8
// CHECK-A32: %struct.uint16x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint16x4x2_t [[TMP6]] // CHECK-A64: ret %struct.uint16x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x4x2_t test_vld1_u16_x2(uint16_t const *a) { uint16x4x2_t test_vld1_u16_x2(uint16_t const *a) {
return vld1_u16_x2(a); return vld1_u16_x2(a);
} }
// CHECK-LABEL: @test_vld1_u16_x3( // CHECK-LABEL: @test_vld1_u16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x3_t, align 8
// CHECK-A32: %struct.uint16x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x3_t, %struct.uint16x4x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x3_t, %struct.uint16x4x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint16x4x3_t [[TMP6]] // CHECK-A64: ret %struct.uint16x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x4x3_t test_vld1_u16_x3(uint16_t const *a) { uint16x4x3_t test_vld1_u16_x3(uint16_t const *a) {
return vld1_u16_x3(a); return vld1_u16_x3(a);
} }
// CHECK-LABEL: @test_vld1_u16_x4( // CHECK-LABEL: @test_vld1_u16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x4x4_t, align 8
// CHECK-A32: %struct.uint16x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint16x4x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }*
// CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]] // CHECK: store { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> } [[VLD1XN]], { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x4_t, %struct.uint16x4x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x4x4_t, %struct.uint16x4x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint16x4x4_t [[TMP6]] // CHECK-A64: ret %struct.uint16x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x4x4_t test_vld1_u16_x4(uint16_t const *a) { uint16x4x4_t test_vld1_u16_x4(uint16_t const *a) {
return vld1_u16_x4(a); return vld1_u16_x4(a);
} }
// CHECK-LABEL: @test_vld1_u32_x2( // CHECK-LABEL: @test_vld1_u32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8
// CHECK-A32: %struct.uint32x2x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x2x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint32x2x2_t [[TMP6]] // CHECK-A64: ret %struct.uint32x2x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x2x2_t test_vld1_u32_x2(uint32_t const *a) { uint32x2x2_t test_vld1_u32_x2(uint32_t const *a) {
return vld1_u32_x2(a); return vld1_u32_x2(a);
} }
// CHECK-LABEL: @test_vld1_u32_x3( // CHECK-LABEL: @test_vld1_u32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x3_t, align 8
// CHECK-A32: %struct.uint32x2x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x2x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x3_t, %struct.uint32x2x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x3_t, %struct.uint32x2x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint32x2x3_t [[TMP6]] // CHECK-A64: ret %struct.uint32x2x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x2x3_t test_vld1_u32_x3(uint32_t const *a) { uint32x2x3_t test_vld1_u32_x3(uint32_t const *a) {
return vld1_u32_x3(a); return vld1_u32_x3(a);
} }
// CHECK-LABEL: @test_vld1_u32_x4( // CHECK-LABEL: @test_vld1_u32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x2x4_t, align 8
// CHECK-A32: %struct.uint32x2x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x2x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint32x2x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }*
// CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]] // CHECK: store { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> } [[VLD1XN]], { <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x2x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x4_t, %struct.uint32x2x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x2x4_t, %struct.uint32x2x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint32x2x4_t [[TMP6]] // CHECK-A64: ret %struct.uint32x2x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x2x4_t test_vld1_u32_x4(uint32_t const *a) { uint32x2x4_t test_vld1_u32_x4(uint32_t const *a) {
return vld1_u32_x4(a); return vld1_u32_x4(a);
} }
// CHECK-LABEL: @test_vld1_u64_x2( // CHECK-LABEL: @test_vld1_u64_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x2_t, align 8
// CHECK-A32: %struct.uint64x1x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x1x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x2_t, %struct.uint64x1x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x2_t, %struct.uint64x1x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint64x1x2_t [[TMP6]] // CHECK-A64: ret %struct.uint64x1x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x1x2_t test_vld1_u64_x2(uint64_t const *a) { uint64x1x2_t test_vld1_u64_x2(uint64_t const *a) {
return vld1_u64_x2(a); return vld1_u64_x2(a);
} }
// CHECK-LABEL: @test_vld1_u64_x3( // CHECK-LABEL: @test_vld1_u64_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x3_t, align 8
// CHECK-A32: %struct.uint64x1x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x1x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x3_t, %struct.uint64x1x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x3_t, %struct.uint64x1x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint64x1x3_t [[TMP6]] // CHECK-A64: ret %struct.uint64x1x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x1x3_t test_vld1_u64_x3(uint64_t const *a) { uint64x1x3_t test_vld1_u64_x3(uint64_t const *a) {
return vld1_u64_x3(a); return vld1_u64_x3(a);
} }
// CHECK-LABEL: @test_vld1_u64_x4( // CHECK-LABEL: @test_vld1_u64_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x1x4_t, align 8
// CHECK-A32: %struct.uint64x1x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x1x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint64x1x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x1x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v1i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }*
// CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD1XN]], { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x1x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x1x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x4_t, %struct.uint64x1x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x1x4_t, %struct.uint64x1x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint64x1x4_t [[TMP6]] // CHECK-A64: ret %struct.uint64x1x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x1x4_t test_vld1_u64_x4(uint64_t const *a) { uint64x1x4_t test_vld1_u64_x4(uint64_t const *a) {
return vld1_u64_x4(a); return vld1_u64_x4(a);
} }
// CHECK-LABEL: @test_vld1_u8_x2( // CHECK-LABEL: @test_vld1_u8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8
// CHECK-A32: %struct.uint8x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x2_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x2_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 16, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint8x8x2_t [[TMP4]] // CHECK-A64: ret %struct.uint8x8x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint8x8x2_t test_vld1_u8_x2(uint8_t const *a) { uint8x8x2_t test_vld1_u8_x2(uint8_t const *a) {
return vld1_u8_x2(a); return vld1_u8_x2(a);
} }
// CHECK-LABEL: @test_vld1_u8_x3( // CHECK-LABEL: @test_vld1_u8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x3_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x3_t, align 8
// CHECK-A32: %struct.uint8x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x3_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x3_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 24, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x3_t, %struct.uint8x8x3_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x3_t, %struct.uint8x8x3_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint8x8x3_t [[TMP4]] // CHECK-A64: ret %struct.uint8x8x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint8x8x3_t test_vld1_u8_x3(uint8_t const *a) { uint8x8x3_t test_vld1_u8_x3(uint8_t const *a) {
return vld1_u8_x3(a); return vld1_u8_x3(a);
} }
// CHECK-LABEL: @test_vld1_u8_x4( // CHECK-LABEL: @test_vld1_u8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x4_t, align 8 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x8x4_t, align 8
// CHECK-A32: %struct.uint8x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x4_t, align 8 // CHECK: [[__RET:%.*]] = alloca %struct.uint8x8x4_t, align 8
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }*
// CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]] // CHECK: store { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> } [[VLD1XN]], { <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align 8 [[TMP2]], i8* align 8 [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x4_t, %struct.uint8x8x4_t* [[RETVAL]], align 8 // CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x8x4_t, %struct.uint8x8x4_t* [[RETVAL]], align 8
// CHECK-A64: ret %struct.uint8x8x4_t [[TMP4]] // CHECK-A64: ret %struct.uint8x8x4_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint8x8x4_t test_vld1_u8_x4(uint8_t const *a) { uint8x8x4_t test_vld1_u8_x4(uint8_t const *a) {
return vld1_u8_x4(a); return vld1_u8_x4(a);
} }
// CHECK-LABEL: @test_vld1q_f16_x2( // CHECK-LABEL: @test_vld1q_f16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x2_t, align 16
// CHECK-A32: %struct.float16x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x8x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]*
// CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v8f16.p0f16|arm.neon.vld1x2.v8i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x2.v8f16.p0f16|arm.neon.vld1x2.v8i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]> }*
// CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]] // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x2_t, %struct.float16x8x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x2_t, %struct.float16x8x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float16x8x2_t [[TMP6]] // CHECK-A64: ret %struct.float16x8x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x8x2_t test_vld1q_f16_x2(float16_t const *a) { float16x8x2_t test_vld1q_f16_x2(float16_t const *a) {
return vld1q_f16_x2(a); return vld1q_f16_x2(a);
} }
// CHECK-LABEL: @test_vld1q_f16_x3( // CHECK-LABEL: @test_vld1q_f16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x3_t, align 16
// CHECK-A32: %struct.float16x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x8x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]*
// CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v8f16.p0f16|arm.neon.vld1x3.v8i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x3.v8f16.p0f16|arm.neon.vld1x3.v8i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }*
// CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]] // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x3_t, %struct.float16x8x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x3_t, %struct.float16x8x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float16x8x3_t [[TMP6]] // CHECK-A64: ret %struct.float16x8x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x8x3_t test_vld1q_f16_x3(float16_t const *a) { float16x8x3_t test_vld1q_f16_x3(float16_t const *a) {
return vld1q_f16_x3(a); return vld1q_f16_x3(a);
} }
// CHECK-LABEL: @test_vld1q_f16_x4( // CHECK-LABEL: @test_vld1q_f16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float16x8x4_t, align 16
// CHECK-A32: %struct.float16x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float16x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float16x8x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float16x8x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float16x8x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast half* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast half* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to [[HALF]]*
// CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v8f16.p0f16|arm.neon.vld1x4.v8i16.p0i16}}([[HALF]]* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } @llvm.{{aarch64.neon.ld1x4.v8f16.p0f16|arm.neon.vld1x4.v8i16.p0i16}}([[HALF]]* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }*
// CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]] // CHECK: store { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> } [[VLD1XN]], { <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]>, <8 x [[HALF]]> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float16x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float16x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x4_t, %struct.float16x8x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float16x8x4_t, %struct.float16x8x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float16x8x4_t [[TMP6]] // CHECK-A64: ret %struct.float16x8x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float16x8x4_t test_vld1q_f16_x4(float16_t const *a) { float16x8x4_t test_vld1q_f16_x4(float16_t const *a) {
return vld1q_f16_x4(a); return vld1q_f16_x4(a);
} }
// CHECK-LABEL: @test_vld1q_f32_x2( // CHECK-LABEL: @test_vld1q_f32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16
// CHECK-A32: %struct.float32x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x4x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float> }*
// CHECK: store { <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float> }* [[TMP3]] // CHECK: store { <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float32x4x2_t [[TMP6]] // CHECK-A64: ret %struct.float32x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float32x4x2_t test_vld1q_f32_x2(float32_t const *a) { float32x4x2_t test_vld1q_f32_x2(float32_t const *a) {
return vld1q_f32_x2(a); return vld1q_f32_x2(a);
} }
// CHECK-LABEL: @test_vld1q_f32_x3( // CHECK-LABEL: @test_vld1q_f32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x3_t, align 16
// CHECK-A32: %struct.float32x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x4x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float>, <4 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float>, <4 x float> }*
// CHECK: store { <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float>, <4 x float> }* [[TMP3]] // CHECK: store { <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float>, <4 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x3_t, %struct.float32x4x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x3_t, %struct.float32x4x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float32x4x3_t [[TMP6]] // CHECK-A64: ret %struct.float32x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float32x4x3_t test_vld1q_f32_x3(float32_t const *a) { float32x4x3_t test_vld1q_f32_x3(float32_t const *a) {
return vld1q_f32_x3(a); return vld1q_f32_x3(a);
} }
// CHECK-LABEL: @test_vld1q_f32_x4( // CHECK-LABEL: @test_vld1q_f32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.float32x4x4_t, align 16
// CHECK-A32: %struct.float32x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.float32x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.float32x4x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.float32x4x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast float* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast float* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to float*
// CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4f32.p0f32(float* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x float>, <4 x float>, <4 x float>, <4 x float> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4f32.p0f32(float* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float>, <4 x float>, <4 x float> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x float>, <4 x float>, <4 x float>, <4 x float> }*
// CHECK: store { <4 x float>, <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float>, <4 x float>, <4 x float> }* [[TMP3]] // CHECK: store { <4 x float>, <4 x float>, <4 x float>, <4 x float> } [[VLD1XN]], { <4 x float>, <4 x float>, <4 x float>, <4 x float> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.float32x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x4_t, %struct.float32x4x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.float32x4x4_t, %struct.float32x4x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.float32x4x4_t [[TMP6]] // CHECK-A64: ret %struct.float32x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
float32x4x4_t test_vld1q_f32_x4(float32_t const *a) { float32x4x4_t test_vld1q_f32_x4(float32_t const *a) {
return vld1q_f32_x4(a); return vld1q_f32_x4(a);
} }
// CHECK-LABEL: @test_vld1q_p16_x2( // CHECK-LABEL: @test_vld1q_p16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16
// CHECK-A32: %struct.poly16x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly16x8x2_t [[TMP6]] // CHECK-A64: ret %struct.poly16x8x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x8x2_t test_vld1q_p16_x2(poly16_t const *a) { poly16x8x2_t test_vld1q_p16_x2(poly16_t const *a) {
return vld1q_p16_x2(a); return vld1q_p16_x2(a);
} }
// CHECK-LABEL: @test_vld1q_p16_x3( // CHECK-LABEL: @test_vld1q_p16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x3_t, align 16
// CHECK-A32: %struct.poly16x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x3_t, %struct.poly16x8x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x3_t, %struct.poly16x8x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly16x8x3_t [[TMP6]] // CHECK-A64: ret %struct.poly16x8x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x8x3_t test_vld1q_p16_x3(poly16_t const *a) { poly16x8x3_t test_vld1q_p16_x3(poly16_t const *a) {
return vld1q_p16_x3(a); return vld1q_p16_x3(a);
} }
// CHECK-LABEL: @test_vld1q_p16_x4( // CHECK-LABEL: @test_vld1q_p16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly16x8x4_t, align 16
// CHECK-A32: %struct.poly16x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly16x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly16x8x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.poly16x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x4_t, %struct.poly16x8x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.poly16x8x4_t, %struct.poly16x8x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly16x8x4_t [[TMP6]] // CHECK-A64: ret %struct.poly16x8x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly16x8x4_t test_vld1q_p16_x4(poly16_t const *a) { poly16x8x4_t test_vld1q_p16_x4(poly16_t const *a) {
return vld1q_p16_x4(a); return vld1q_p16_x4(a);
} }
// CHECK-LABEL: @test_vld1q_p8_x2( // CHECK-LABEL: @test_vld1q_p8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16
// CHECK-A32: %struct.poly8x16x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x16x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly8x16x2_t [[TMP4]] // CHECK-A64: ret %struct.poly8x16x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x16x2_t test_vld1q_p8_x2(poly8_t const *a) { poly8x16x2_t test_vld1q_p8_x2(poly8_t const *a) {
return vld1q_p8_x2(a); return vld1q_p8_x2(a);
} }
// CHECK-LABEL: @test_vld1q_p8_x3( // CHECK-LABEL: @test_vld1q_p8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x3_t, align 16
// CHECK-A32: %struct.poly8x16x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x16x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x3_t, %struct.poly8x16x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x3_t, %struct.poly8x16x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly8x16x3_t [[TMP4]] // CHECK-A64: ret %struct.poly8x16x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x16x3_t test_vld1q_p8_x3(poly8_t const *a) { poly8x16x3_t test_vld1q_p8_x3(poly8_t const *a) {
return vld1q_p8_x3(a); return vld1q_p8_x3(a);
} }
// CHECK-LABEL: @test_vld1q_p8_x4( // CHECK-LABEL: @test_vld1q_p8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.poly8x16x4_t, align 16
// CHECK-A32: %struct.poly8x16x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.poly8x16x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.poly8x16x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.poly8x16x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x4_t, %struct.poly8x16x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.poly8x16x4_t, %struct.poly8x16x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.poly8x16x4_t [[TMP4]] // CHECK-A64: ret %struct.poly8x16x4_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
poly8x16x4_t test_vld1q_p8_x4(poly8_t const *a) { poly8x16x4_t test_vld1q_p8_x4(poly8_t const *a) {
return vld1q_p8_x4(a); return vld1q_p8_x4(a);
} }
// CHECK-LABEL: @test_vld1q_s16_x2( // CHECK-LABEL: @test_vld1q_s16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16
// CHECK-A32: %struct.int16x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x8x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int16x8x2_t [[TMP6]] // CHECK-A64: ret %struct.int16x8x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x8x2_t test_vld1q_s16_x2(int16_t const *a) { int16x8x2_t test_vld1q_s16_x2(int16_t const *a) {
return vld1q_s16_x2(a); return vld1q_s16_x2(a);
} }
// CHECK-LABEL: @test_vld1q_s16_x3( // CHECK-LABEL: @test_vld1q_s16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x3_t, align 16
// CHECK-A32: %struct.int16x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x8x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x3_t, %struct.int16x8x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x3_t, %struct.int16x8x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int16x8x3_t [[TMP6]] // CHECK-A64: ret %struct.int16x8x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x8x3_t test_vld1q_s16_x3(int16_t const *a) { int16x8x3_t test_vld1q_s16_x3(int16_t const *a) {
return vld1q_s16_x3(a); return vld1q_s16_x3(a);
} }
// CHECK-LABEL: @test_vld1q_s16_x4( // CHECK-LABEL: @test_vld1q_s16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int16x8x4_t, align 16
// CHECK-A32: %struct.int16x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int16x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int16x8x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int16x8x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int16x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x4_t, %struct.int16x8x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int16x8x4_t, %struct.int16x8x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int16x8x4_t [[TMP6]] // CHECK-A64: ret %struct.int16x8x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int16x8x4_t test_vld1q_s16_x4(int16_t const *a) { int16x8x4_t test_vld1q_s16_x4(int16_t const *a) {
return vld1q_s16_x4(a); return vld1q_s16_x4(a);
} }
// CHECK-LABEL: @test_vld1q_s32_x2( // CHECK-LABEL: @test_vld1q_s32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16
// CHECK-A32: %struct.int32x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x4x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int32x4x2_t [[TMP6]] // CHECK-A64: ret %struct.int32x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x4x2_t test_vld1q_s32_x2(int32_t const *a) { int32x4x2_t test_vld1q_s32_x2(int32_t const *a) {
return vld1q_s32_x2(a); return vld1q_s32_x2(a);
} }
// CHECK-LABEL: @test_vld1q_s32_x3( // CHECK-LABEL: @test_vld1q_s32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x3_t, align 16
// CHECK-A32: %struct.int32x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x4x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x3_t, %struct.int32x4x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x3_t, %struct.int32x4x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int32x4x3_t [[TMP6]] // CHECK-A64: ret %struct.int32x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x4x3_t test_vld1q_s32_x3(int32_t const *a) { int32x4x3_t test_vld1q_s32_x3(int32_t const *a) {
return vld1q_s32_x3(a); return vld1q_s32_x3(a);
} }
// CHECK-LABEL: @test_vld1q_s32_x4( // CHECK-LABEL: @test_vld1q_s32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int32x4x4_t, align 16
// CHECK-A32: %struct.int32x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int32x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int32x4x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int32x4x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int32x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x4_t, %struct.int32x4x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int32x4x4_t, %struct.int32x4x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int32x4x4_t [[TMP6]] // CHECK-A64: ret %struct.int32x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int32x4x4_t test_vld1q_s32_x4(int32_t const *a) { int32x4x4_t test_vld1q_s32_x4(int32_t const *a) {
return vld1q_s32_x4(a); return vld1q_s32_x4(a);
} }
// CHECK-LABEL: @test_vld1q_s64_x2( // CHECK-LABEL: @test_vld1q_s64_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x2_t, align 16
// CHECK-A32: %struct.int64x2x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x2x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x2x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x2_t, %struct.int64x2x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x2_t, %struct.int64x2x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int64x2x2_t [[TMP6]] // CHECK-A64: ret %struct.int64x2x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x2x2_t test_vld1q_s64_x2(int64_t const *a) { int64x2x2_t test_vld1q_s64_x2(int64_t const *a) {
return vld1q_s64_x2(a); return vld1q_s64_x2(a);
} }
// CHECK-LABEL: @test_vld1q_s64_x3( // CHECK-LABEL: @test_vld1q_s64_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x3_t, align 16
// CHECK-A32: %struct.int64x2x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x2x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x2x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x3_t, %struct.int64x2x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x3_t, %struct.int64x2x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int64x2x3_t [[TMP6]] // CHECK-A64: ret %struct.int64x2x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x2x3_t test_vld1q_s64_x3(int64_t const *a) { int64x2x3_t test_vld1q_s64_x3(int64_t const *a) {
return vld1q_s64_x3(a); return vld1q_s64_x3(a);
} }
// CHECK-LABEL: @test_vld1q_s64_x4( // CHECK-LABEL: @test_vld1q_s64_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int64x2x4_t, align 16
// CHECK-A32: %struct.int64x2x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int64x2x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int64x2x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int64x2x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int64x2x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.int64x2x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.int64x2x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x4_t, %struct.int64x2x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.int64x2x4_t, %struct.int64x2x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int64x2x4_t [[TMP6]] // CHECK-A64: ret %struct.int64x2x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
int64x2x4_t test_vld1q_s64_x4(int64_t const *a) { int64x2x4_t test_vld1q_s64_x4(int64_t const *a) {
return vld1q_s64_x4(a); return vld1q_s64_x4(a);
} }
// CHECK-LABEL: @test_vld1q_s8_x2( // CHECK-LABEL: @test_vld1q_s8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16
// CHECK-A32: %struct.int8x16x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x16x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x16x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int8x16x2_t [[TMP4]] // CHECK-A64: ret %struct.int8x16x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x16x2_t test_vld1q_s8_x2(int8_t const *a) { int8x16x2_t test_vld1q_s8_x2(int8_t const *a) {
return vld1q_s8_x2(a); return vld1q_s8_x2(a);
} }
// CHECK-LABEL: @test_vld1q_s8_x3( // CHECK-LABEL: @test_vld1q_s8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x3_t, align 16
// CHECK-A32: %struct.int8x16x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x16x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x16x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x3_t, %struct.int8x16x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x3_t, %struct.int8x16x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int8x16x3_t [[TMP4]] // CHECK-A64: ret %struct.int8x16x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x16x3_t test_vld1q_s8_x3(int8_t const *a) { int8x16x3_t test_vld1q_s8_x3(int8_t const *a) {
return vld1q_s8_x3(a); return vld1q_s8_x3(a);
} }
// CHECK-LABEL: @test_vld1q_s8_x4( // CHECK-LABEL: @test_vld1q_s8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.int8x16x4_t, align 16
// CHECK-A32: %struct.int8x16x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.int8x16x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.int8x16x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.int8x16x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.int8x16x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x4_t, %struct.int8x16x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.int8x16x4_t, %struct.int8x16x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.int8x16x4_t [[TMP4]] // CHECK-A64: ret %struct.int8x16x4_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
int8x16x4_t test_vld1q_s8_x4(int8_t const *a) { int8x16x4_t test_vld1q_s8_x4(int8_t const *a) {
return vld1q_s8_x4(a); return vld1q_s8_x4(a);
} }
// CHECK-LABEL: @test_vld1q_u16_x2( // CHECK-LABEL: @test_vld1q_u16_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16
// CHECK-A32: %struct.uint16x8x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x8x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint16x8x2_t [[TMP6]] // CHECK-A64: ret %struct.uint16x8x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x8x2_t test_vld1q_u16_x2(uint16_t const *a) { uint16x8x2_t test_vld1q_u16_x2(uint16_t const *a) {
return vld1q_u16_x2(a); return vld1q_u16_x2(a);
} }
// CHECK-LABEL: @test_vld1q_u16_x3( // CHECK-LABEL: @test_vld1q_u16_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x3_t, align 16
// CHECK-A32: %struct.uint16x8x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x8x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x3_t, %struct.uint16x8x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x3_t, %struct.uint16x8x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint16x8x3_t [[TMP6]] // CHECK-A64: ret %struct.uint16x8x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x8x3_t test_vld1q_u16_x3(uint16_t const *a) { uint16x8x3_t test_vld1q_u16_x3(uint16_t const *a) {
return vld1q_u16_x3(a); return vld1q_u16_x3(a);
} }
// CHECK-LABEL: @test_vld1q_u16_x4( // CHECK-LABEL: @test_vld1q_u16_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint16x8x4_t, align 16
// CHECK-A32: %struct.uint16x8x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint16x8x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint16x8x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i16* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i16*
// CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v8i16.p0i16(i16* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }*
// CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]] // CHECK: store { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> } [[VLD1XN]], { <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint16x8x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x4_t, %struct.uint16x8x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint16x8x4_t, %struct.uint16x8x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint16x8x4_t [[TMP6]] // CHECK-A64: ret %struct.uint16x8x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint16x8x4_t test_vld1q_u16_x4(uint16_t const *a) { uint16x8x4_t test_vld1q_u16_x4(uint16_t const *a) {
return vld1q_u16_x4(a); return vld1q_u16_x4(a);
} }
// CHECK-LABEL: @test_vld1q_u32_x2( // CHECK-LABEL: @test_vld1q_u32_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16
// CHECK-A32: %struct.uint32x4x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x4x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint32x4x2_t [[TMP6]] // CHECK-A64: ret %struct.uint32x4x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x4x2_t test_vld1q_u32_x2(uint32_t const *a) { uint32x4x2_t test_vld1q_u32_x2(uint32_t const *a) {
return vld1q_u32_x2(a); return vld1q_u32_x2(a);
} }
// CHECK-LABEL: @test_vld1q_u32_x3( // CHECK-LABEL: @test_vld1q_u32_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x3_t, align 16
// CHECK-A32: %struct.uint32x4x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x4x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x3_t, %struct.uint32x4x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x3_t, %struct.uint32x4x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint32x4x3_t [[TMP6]] // CHECK-A64: ret %struct.uint32x4x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x4x3_t test_vld1q_u32_x3(uint32_t const *a) { uint32x4x3_t test_vld1q_u32_x3(uint32_t const *a) {
return vld1q_u32_x3(a); return vld1q_u32_x3(a);
} }
// CHECK-LABEL: @test_vld1q_u32_x4( // CHECK-LABEL: @test_vld1q_u32_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint32x4x4_t, align 16
// CHECK-A32: %struct.uint32x4x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint32x4x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint32x4x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i32* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i32*
// CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0i32(i32* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v4i32.p0i32(i32* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }*
// CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]] // CHECK: store { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> } [[VLD1XN]], { <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint32x4x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x4_t, %struct.uint32x4x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint32x4x4_t, %struct.uint32x4x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint32x4x4_t [[TMP6]] // CHECK-A64: ret %struct.uint32x4x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint32x4x4_t test_vld1q_u32_x4(uint32_t const *a) { uint32x4x4_t test_vld1q_u32_x4(uint32_t const *a) {
return vld1q_u32_x4(a); return vld1q_u32_x4(a);
} }
// CHECK-LABEL: @test_vld1q_u64_x2( // CHECK-LABEL: @test_vld1q_u64_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x2_t, align 16
// CHECK-A32: %struct.uint64x2x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x2x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x2_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x2_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x2_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x2_t, %struct.uint64x2x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x2_t, %struct.uint64x2x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint64x2x2_t [[TMP6]] // CHECK-A64: ret %struct.uint64x2x2_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x2x2_t test_vld1q_u64_x2(uint64_t const *a) { uint64x2x2_t test_vld1q_u64_x2(uint64_t const *a) {
return vld1q_u64_x2(a); return vld1q_u64_x2(a);
} }
// CHECK-LABEL: @test_vld1q_u64_x3( // CHECK-LABEL: @test_vld1q_u64_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x3_t, align 16
// CHECK-A32: %struct.uint64x2x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x2x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x3_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x3_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x3_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x3_t, %struct.uint64x2x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x3_t, %struct.uint64x2x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint64x2x3_t [[TMP6]] // CHECK-A64: ret %struct.uint64x2x3_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x2x3_t test_vld1q_u64_x3(uint64_t const *a) { uint64x2x3_t test_vld1q_u64_x3(uint64_t const *a) {
return vld1q_u64_x3(a); return vld1q_u64_x3(a);
} }
// CHECK-LABEL: @test_vld1q_u64_x4( // CHECK-LABEL: @test_vld1q_u64_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint64x2x4_t, align 16
// CHECK-A32: %struct.uint64x2x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint64x2x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint64x2x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint64x2x4_t* [[__RET]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8* // CHECK: [[TMP1:%.*]] = bitcast i64* %a to i8*
// CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64* // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to i64*
// CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0i64(i64* [[TMP2]]) // CHECK: [[VLD1XN:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v2i64.p0i64(i64* [[TMP2]])
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }*
// CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD1XN]], { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]]
// CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x4_t* [[RETVAL]] to i8* // CHECK: [[TMP4:%.*]] = bitcast %struct.uint64x2x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x4_t* [[__RET]] to i8* // CHECK: [[TMP5:%.*]] = bitcast %struct.uint64x2x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP4]], i8* align {{16|8}} [[TMP5]], {{i64|i32}} 64, i1 false)
// CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x4_t, %struct.uint64x2x4_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP6:%.*]] = load %struct.uint64x2x4_t, %struct.uint64x2x4_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint64x2x4_t [[TMP6]] // CHECK-A64: ret %struct.uint64x2x4_t [[TMP6]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint64x2x4_t test_vld1q_u64_x4(uint64_t const *a) { uint64x2x4_t test_vld1q_u64_x4(uint64_t const *a) {
return vld1q_u64_x4(a); return vld1q_u64_x4(a);
} }
// CHECK-LABEL: @test_vld1q_u8_x2( // CHECK-LABEL: @test_vld1q_u8_x2(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16
// CHECK-A32: %struct.uint8x16x2_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x16x2_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x2_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x2_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x2|arm.neon.vld1x2}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 32, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint8x16x2_t [[TMP4]] // CHECK-A64: ret %struct.uint8x16x2_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint8x16x2_t test_vld1q_u8_x2(uint8_t const *a) { uint8x16x2_t test_vld1q_u8_x2(uint8_t const *a) {
return vld1q_u8_x2(a); return vld1q_u8_x2(a);
} }
// CHECK-LABEL: @test_vld1q_u8_x3( // CHECK-LABEL: @test_vld1q_u8_x3(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x3_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x3_t, align 16
// CHECK-A32: %struct.uint8x16x3_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x16x3_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x3_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x3_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x3_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x3|arm.neon.vld1x3}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x3_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x3_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x3_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x3_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 48, i1 false)
// CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x16x3_t, %struct.uint8x16x3_t* [[RETVAL]], align 16 // CHECK-A64: [[TMP4:%.*]] = load %struct.uint8x16x3_t, %struct.uint8x16x3_t* [[RETVAL]], align 16
// CHECK-A64: ret %struct.uint8x16x3_t [[TMP4]] // CHECK-A64: ret %struct.uint8x16x3_t [[TMP4]]
// CHECK-A32: ret void // CHECK-A32: ret void
uint8x16x3_t test_vld1q_u8_x3(uint8_t const *a) { uint8x16x3_t test_vld1q_u8_x3(uint8_t const *a) {
return vld1q_u8_x3(a); return vld1q_u8_x3(a);
} }
// CHECK-LABEL: @test_vld1q_u8_x4( // CHECK-LABEL: @test_vld1q_u8_x4(
// CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x4_t, align 16 // CHECK-A64: [[RETVAL:%.*]] = alloca %struct.uint8x16x4_t, align 16
// CHECK-A32: %struct.uint8x16x4_t* noalias sret [[RETVAL:%.*]], // CHECK-A32: %struct.uint8x16x4_t* noalias sret align 8 [[RETVAL:%.*]],
// CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x4_t, align {{16|8}} // CHECK: [[__RET:%.*]] = alloca %struct.uint8x16x4_t, align {{16|8}}
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x4_t* [[__RET]] to i8*
// CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a) // CHECK: [[VLD1XN:%.*]] = call { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } @llvm.{{aarch64.neon.ld1x4|arm.neon.vld1x4}}.v16i8.p0i8(i8* %a)
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }*
// CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]] // CHECK: store { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> } [[VLD1XN]], { <16 x i8>, <16 x i8>, <16 x i8>, <16 x i8> }* [[TMP1]]
// CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x4_t* [[RETVAL]] to i8* // CHECK: [[TMP2:%.*]] = bitcast %struct.uint8x16x4_t* [[RETVAL]] to i8*
// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x4_t* [[__RET]] to i8* // CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x4_t* [[__RET]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.{{i64|i32}}(i8* align {{16|8}} [[TMP2]], i8* align {{16|8}} [[TMP3]], {{i64|i32}} 64, i1 false)
▲ Show 20 Lines • Show All 492 Lines • Show Last 20 Lines

clang/test/CodeGen/arm-varargs.c

Show All 18 Lines
// CHECK: ret i32 [[RESULT]] // CHECK: ret i32 [[RESULT]]
} }
struct bigstruct { struct bigstruct {
int a[10]; int a[10];
}; };
struct bigstruct simple_struct(void) { struct bigstruct simple_struct(void) {
// CHECK-LABEL: define void @simple_struct(%struct.bigstruct* noalias sret %agg.result) // CHECK-LABEL: define void @simple_struct(%struct.bigstruct* noalias sret align 4 %agg.result)
return va_arg(the_list, struct bigstruct); return va_arg(the_list, struct bigstruct);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 40 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 40
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.bigstruct* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.bigstruct*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.bigstruct* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.bigstruct* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.bigstruct* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.bigstruct* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 40, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 40, i1 false)
// CHECK: ret void // CHECK: ret void
} }
struct aligned_bigstruct { struct aligned_bigstruct {
float a; float a;
long double b; long double b;
}; };
struct aligned_bigstruct simple_aligned_struct(void) { struct aligned_bigstruct simple_aligned_struct(void) {
// CHECK-LABEL: define void @simple_aligned_struct(%struct.aligned_bigstruct* noalias sret %agg.result) // CHECK-LABEL: define void @simple_aligned_struct(%struct.aligned_bigstruct* noalias sret align 8 %agg.result)
return va_arg(the_list, struct aligned_bigstruct); return va_arg(the_list, struct aligned_bigstruct);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32 // CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32
// CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7
// CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8 // CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8
// CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8* // CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8*
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
Show All 19 Lines
// CHECK: ret double [[RESULT]] // CHECK: ret double [[RESULT]]
} }
struct hfa { struct hfa {
float a, b; float a, b;
}; };
struct hfa simple_hfa(void) { struct hfa simple_hfa(void) {
// CHECK-LABEL: define void @simple_hfa(%struct.hfa* noalias sret %agg.result) // CHECK-LABEL: define void @simple_hfa(%struct.hfa* noalias sret align 4 %agg.result)
return va_arg(the_list, struct hfa); return va_arg(the_list, struct hfa);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.hfa* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.hfa*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.hfa* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.hfa* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.hfa* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.hfa* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false)
▲ Show 20 Lines • Show All 90 Lines • ▼ Show 20 Lines
// CHECK: [[RESULT:%[a-z0-9._]+]] = load i32, i32* [[COERCE]] // CHECK: [[RESULT:%[a-z0-9._]+]] = load i32, i32* [[COERCE]]
// CHECK: ret i32 [[RESULT]] // CHECK: ret i32 [[RESULT]]
} }
typedef struct __attribute__((aligned(16))) { typedef struct __attribute__((aligned(16))) {
int val; int val;
} overaligned_int_struct; } overaligned_int_struct;
overaligned_int_struct overaligned_int_struct_test() { overaligned_int_struct overaligned_int_struct_test() {
// CHECK-LABEL: define void @overaligned_int_struct_test(%struct.overaligned_int_struct* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_int_struct_test(%struct.overaligned_int_struct* noalias sret align 16 %agg.result)
return va_arg(the_list, overaligned_int_struct); return va_arg(the_list, overaligned_int_struct);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 16 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 16
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.overaligned_int_struct* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.overaligned_int_struct*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 16, i1 false)
// CHECK: ret void // CHECK: ret void
} }
typedef struct __attribute__((packed,aligned(2))) { typedef struct __attribute__((packed,aligned(2))) {
long long val; long long val;
} underaligned_long_long_struct; } underaligned_long_long_struct;
underaligned_long_long_struct underaligned_long_long_struct_test() { underaligned_long_long_struct underaligned_long_long_struct_test() {
// CHECK-LABEL: define void @underaligned_long_long_struct_test(%struct.underaligned_long_long_struct* noalias sret %agg.result) // CHECK-LABEL: define void @underaligned_long_long_struct_test(%struct.underaligned_long_long_struct* noalias sret align 2 %agg.result)
return va_arg(the_list, underaligned_long_long_struct); return va_arg(the_list, underaligned_long_long_struct);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.underaligned_long_long_struct* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.underaligned_long_long_struct*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false)
// CHECK: ret void // CHECK: ret void
} }
typedef struct __attribute__((aligned(16))) { typedef struct __attribute__((aligned(16))) {
long long val; long long val;
} overaligned_long_long_struct; } overaligned_long_long_struct;
overaligned_long_long_struct overaligned_long_long_struct_test() { overaligned_long_long_struct overaligned_long_long_struct_test() {
// CHECK-LABEL: define void @overaligned_long_long_struct_test(%struct.overaligned_long_long_struct* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_long_long_struct_test(%struct.overaligned_long_long_struct* noalias sret align 16 %agg.result)
return va_arg(the_list, overaligned_long_long_struct); return va_arg(the_list, overaligned_long_long_struct);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32 // CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32
// CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7
// CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8 // CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8
// CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8* // CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8*
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
Show All 25 Lines
// CHECK: [[RESULT:%[a-z0-9._]+]] = load i32, i32* [[COERCE]] // CHECK: [[RESULT:%[a-z0-9._]+]] = load i32, i32* [[COERCE]]
// CHECK: ret i32 [[RESULT]] // CHECK: ret i32 [[RESULT]]
} }
typedef struct { typedef struct {
int val __attribute__((aligned(16))); int val __attribute__((aligned(16)));
} overaligned_int_struct_member; } overaligned_int_struct_member;
overaligned_int_struct_member overaligned_int_struct_member_test() { overaligned_int_struct_member overaligned_int_struct_member_test() {
// CHECK-LABEL: define void @overaligned_int_struct_member_test(%struct.overaligned_int_struct_member* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_int_struct_member_test(%struct.overaligned_int_struct_member* noalias sret align 16 %agg.result)
return va_arg(the_list, overaligned_int_struct_member); return va_arg(the_list, overaligned_int_struct_member);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32 // CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32
// CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7
// CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8 // CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8
// CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8* // CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8*
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR_ALIGNED]] to %struct.overaligned_int_struct_member* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR_ALIGNED]] to %struct.overaligned_int_struct_member*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct_member* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct_member* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct_member* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.overaligned_int_struct_member* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 [[DEST_ADDR]], i8* align 8 [[SRC_ADDR]], i32 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 [[DEST_ADDR]], i8* align 8 [[SRC_ADDR]], i32 16, i1 false)
// CHECK: ret void // CHECK: ret void
} }
typedef struct { typedef struct {
long long val __attribute__((packed,aligned(2))); long long val __attribute__((packed,aligned(2)));
} underaligned_long_long_struct_member; } underaligned_long_long_struct_member;
underaligned_long_long_struct_member underaligned_long_long_struct_member_test() { underaligned_long_long_struct_member underaligned_long_long_struct_member_test() {
// CHECK-LABEL: define void @underaligned_long_long_struct_member_test(%struct.underaligned_long_long_struct_member* noalias sret %agg.result) // CHECK-LABEL: define void @underaligned_long_long_struct_member_test(%struct.underaligned_long_long_struct_member* noalias sret align 2 %agg.result)
return va_arg(the_list, underaligned_long_long_struct_member); return va_arg(the_list, underaligned_long_long_struct_member);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR]], i32 8
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.underaligned_long_long_struct_member* // CHECK: [[ADDR:%[a-z0-9._]+]] = bitcast i8* [[CUR]] to %struct.underaligned_long_long_struct_member*
// CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct_member* %agg.result to i8* // CHECK: [[DEST_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct_member* %agg.result to i8*
// CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct_member* [[ADDR]] to i8* // CHECK: [[SRC_ADDR:%[a-z0-9._]+]] = bitcast %struct.underaligned_long_long_struct_member* [[ADDR]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[DEST_ADDR]], i8* align 4 [[SRC_ADDR]], i32 8, i1 false)
// CHECK: ret void // CHECK: ret void
} }
typedef struct { typedef struct {
long long val __attribute__((aligned(16))); long long val __attribute__((aligned(16)));
} overaligned_long_long_struct_member; } overaligned_long_long_struct_member;
overaligned_long_long_struct_member overaligned_long_long_struct_member_test() { overaligned_long_long_struct_member overaligned_long_long_struct_member_test() {
// CHECK-LABEL: define void @overaligned_long_long_struct_member_test(%struct.overaligned_long_long_struct_member* noalias sret %agg.result) // CHECK-LABEL: define void @overaligned_long_long_struct_member_test(%struct.overaligned_long_long_struct_member* noalias sret align 16 %agg.result)
return va_arg(the_list, overaligned_long_long_struct_member); return va_arg(the_list, overaligned_long_long_struct_member);
// CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: [[CUR:%[a-z0-9._]+]] = load i8*, i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
// CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32 // CHECK: [[CUR_INT:%[a-z0-9._]+]] = ptrtoint i8* [[CUR]] to i32
// CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7 // CHECK: [[CUR_INT_ADD:%[a-z0-9._]+]] = add i32 [[CUR_INT]], 7
// CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8 // CHECK: [[CUR_INT_ALIGNED:%[a-z0-9._]+]] = and i32 [[CUR_INT_ADD]], -8
// CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8* // CHECK: [[CUR_ALIGNED:%[a-z0-9._]+]] = inttoptr i32 [[CUR_INT_ALIGNED]] to i8*
// CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16 // CHECK: [[NEXT:%[a-z0-9._]+]] = getelementptr inbounds i8, i8* [[CUR_ALIGNED]], i32 16
// CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4 // CHECK: store i8* [[NEXT]], i8** getelementptr inbounds (%struct.__va_list, %struct.__va_list* @the_list, i32 0, i32 0), align 4
Show All 16 Lines

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

// REQUIRES: arm-registered-target // REQUIRES: arm-registered-target
// RUN: %clang_cc1 -triple thumbv7-apple-darwin9 \ // RUN: %clang_cc1 -triple thumbv7-apple-darwin9 \
// RUN: -target-abi apcs-gnu \ // RUN: -target-abi apcs-gnu \
// RUN: -target-cpu cortex-a8 \ // RUN: -target-cpu cortex-a8 \
// RUN: -mfloat-abi soft \ // RUN: -mfloat-abi soft \
// RUN: -target-feature +soft-float-abi \ // RUN: -target-feature +soft-float-abi \
// RUN: -ffreestanding \ // RUN: -ffreestanding \
// RUN: -emit-llvm -w -o - %s | FileCheck %s // RUN: -emit-llvm -w -o - %s | FileCheck %s
#include <arm_neon.h> #include <arm_neon.h>
// CHECK: define void @f0(%struct.int8x16x2_t* noalias sret %agg.result, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}) // CHECK: define void @f0(%struct.int8x16x2_t* noalias sret align 16 %agg.result, <16 x i8> %{{.*}}, <16 x i8> %{{.*}})
int8x16x2_t f0(int8x16_t a0, int8x16_t a1) { int8x16x2_t f0(int8x16_t a0, int8x16_t a1) {
return vzipq_s8(a0, a1); return vzipq_s8(a0, a1);
} }
// Test direct vector passing. // Test direct vector passing.
typedef float T_float32x2 __attribute__ ((__vector_size__ (8))); typedef float T_float32x2 __attribute__ ((__vector_size__ (8)));
typedef float T_float32x4 __attribute__ ((__vector_size__ (16))); typedef float T_float32x4 __attribute__ ((__vector_size__ (16)));
typedef float T_float32x8 __attribute__ ((__vector_size__ (32))); typedef float T_float32x8 __attribute__ ((__vector_size__ (32)));
typedef float T_float32x16 __attribute__ ((__vector_size__ (64))); typedef float T_float32x16 __attribute__ ((__vector_size__ (64)));
// CHECK: define <2 x float> @f1_0(<2 x float> %{{.*}}) // CHECK: define <2 x float> @f1_0(<2 x float> %{{.*}})
T_float32x2 f1_0(T_float32x2 a0) { return a0; } T_float32x2 f1_0(T_float32x2 a0) { return a0; }
// CHECK: define <4 x float> @f1_1(<4 x float> %{{.*}}) // CHECK: define <4 x float> @f1_1(<4 x float> %{{.*}})
T_float32x4 f1_1(T_float32x4 a0) { return a0; } T_float32x4 f1_1(T_float32x4 a0) { return a0; }
// CHECK: define void @f1_2(<8 x float>* noalias sret %{{.*}}, <8 x float> %{{.*}}) // CHECK: define void @f1_2(<8 x float>* noalias sret align 32 %{{.*}}, <8 x float> %{{.*}})
T_float32x8 f1_2(T_float32x8 a0) { return a0; } T_float32x8 f1_2(T_float32x8 a0) { return a0; }
// CHECK: define void @f1_3(<16 x float>* noalias sret %{{.*}}, <16 x float> %{{.*}}) // CHECK: define void @f1_3(<16 x float>* noalias sret align 64 %{{.*}}, <16 x float> %{{.*}})
T_float32x16 f1_3(T_float32x16 a0) { return a0; } T_float32x16 f1_3(T_float32x16 a0) { return a0; }

clang/test/CodeGen/arm-vfp16-arguments.c

Show First 20 Lines • Show All 65 Lines • ▼ Show 20 Lines
void test_hfa(hfa_t a) {} void test_hfa(hfa_t a) {}
// CHECK-SOFT: define void @test_hfa([2 x i64] %a.coerce) // CHECK-SOFT: define void @test_hfa([2 x i64] %a.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc void @test_hfa([2 x <2 x i32>] %a.coerce) // CHECK-HARD: define arm_aapcs_vfpcc void @test_hfa([2 x <2 x i32>] %a.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc void @test_hfa(%struct.hfa_t %a.coerce) // CHECK-FULL: define arm_aapcs_vfpcc void @test_hfa(%struct.hfa_t %a.coerce)
hfa_t ghfa; hfa_t ghfa;
hfa_t test_ret_hfa(void) { return ghfa; } hfa_t test_ret_hfa(void) { return ghfa; }
// CHECK-SOFT: define void @test_ret_hfa(%struct.hfa_t* noalias nocapture sret %agg.result) // CHECK-SOFT: define void @test_ret_hfa(%struct.hfa_t* noalias nocapture sret align 8 %agg.result)
// CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @test_ret_hfa() // CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @test_ret_hfa()
// CHECK-FULL: define arm_aapcs_vfpcc %struct.hfa_t @test_ret_hfa() // CHECK-FULL: define arm_aapcs_vfpcc %struct.hfa_t @test_ret_hfa()

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

Show All 31 Linesstruct S4 : B1 {
B2 M[1]; B2 M[1];
}; };
// S5 does not contain any FP16 vectors // S5 does not contain any FP16 vectors
struct S5 : B1 { struct S5 : B1 {
B1 M[1]; B1 M[1];
}; };
// CHECK-SOFT: define void @_Z2f12S1(%struct.S1* noalias nocapture sret %agg.result, [2 x i64] %s1.coerce) // CHECK-SOFT: define void @_Z2f12S1(%struct.S1* noalias nocapture sret align 8 %agg.result, [2 x i64] %s1.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f12S1([2 x <2 x i32>] returned %s1.coerce) // CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f12S1([2 x <2 x i32>] returned %s1.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc %struct.S1 @_Z2f12S1(%struct.S1 returned %s1.coerce) // CHECK-FULL: define arm_aapcs_vfpcc %struct.S1 @_Z2f12S1(%struct.S1 returned %s1.coerce)
struct S1 f1(struct S1 s1) { return s1; } struct S1 f1(struct S1 s1) { return s1; }
// CHECK-SOFT: define void @_Z2f22S2(%struct.S2* noalias nocapture sret %agg.result, [4 x i32] %s2.coerce) // CHECK-SOFT: define void @_Z2f22S2(%struct.S2* noalias nocapture sret align 8 %agg.result, [4 x i32] %s2.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f22S2([2 x <2 x i32>] returned %s2.coerce) // CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f22S2([2 x <2 x i32>] returned %s2.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc %struct.S2 @_Z2f22S2(%struct.S2 returned %s2.coerce) // CHECK-FULL: define arm_aapcs_vfpcc %struct.S2 @_Z2f22S2(%struct.S2 returned %s2.coerce)
struct S2 f2(struct S2 s2) { return s2; } struct S2 f2(struct S2 s2) { return s2; }
// CHECK-SOFT: define void @_Z2f32S3(%struct.S3* noalias nocapture sret %agg.result, [2 x i64] %s3.coerce) // CHECK-SOFT: define void @_Z2f32S3(%struct.S3* noalias nocapture sret align 8 %agg.result, [2 x i64] %s3.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f32S3([2 x <2 x i32>] returned %s3.coerce) // CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f32S3([2 x <2 x i32>] returned %s3.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc %struct.S3 @_Z2f32S3(%struct.S3 returned %s3.coerce) // CHECK-FULL: define arm_aapcs_vfpcc %struct.S3 @_Z2f32S3(%struct.S3 returned %s3.coerce)
struct S3 f3(struct S3 s3) { return s3; } struct S3 f3(struct S3 s3) { return s3; }
// CHECK-SOFT: define void @_Z2f42S4(%struct.S4* noalias nocapture sret %agg.result, [2 x i64] %s4.coerce) // CHECK-SOFT: define void @_Z2f42S4(%struct.S4* noalias nocapture sret align 8 %agg.result, [2 x i64] %s4.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f42S4([2 x <2 x i32>] returned %s4.coerce) // CHECK-HARD: define arm_aapcs_vfpcc [2 x <2 x i32>] @_Z2f42S4([2 x <2 x i32>] returned %s4.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc %struct.S4 @_Z2f42S4(%struct.S4 returned %s4.coerce) // CHECK-FULL: define arm_aapcs_vfpcc %struct.S4 @_Z2f42S4(%struct.S4 returned %s4.coerce)
struct S4 f4(struct S4 s4) { return s4; } struct S4 f4(struct S4 s4) { return s4; }
// CHECK-SOFT: define void @_Z2f52S5(%struct.S5* noalias nocapture sret %agg.result, [2 x i64] %s5.coerce) // CHECK-SOFT: define void @_Z2f52S5(%struct.S5* noalias nocapture sret align 8 %agg.result, [2 x i64] %s5.coerce)
// CHECK-HARD: define arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce) // CHECK-HARD: define arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce)
// CHECK-FULL: define arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce) // CHECK-FULL: define arm_aapcs_vfpcc %struct.S5 @_Z2f52S5(%struct.S5 returned %s5.coerce)
struct S5 f5(struct S5 s5) { return s5; } struct S5 f5(struct S5 s5) { return s5; }

clang/test/CodeGen/arm64-arguments.c

Show First 20 Lines • Show All 175 Lines • ▼ Show 20 Lines
typedef float T_float32x8 __attribute__ ((__vector_size__ (32))); typedef float T_float32x8 __attribute__ ((__vector_size__ (32)));
typedef float T_float32x16 __attribute__ ((__vector_size__ (64))); typedef float T_float32x16 __attribute__ ((__vector_size__ (64)));
// CHECK: define <2 x float> @f1_0(<2 x float> %{{.*}}) // CHECK: define <2 x float> @f1_0(<2 x float> %{{.*}})
T_float32x2 f1_0(T_float32x2 a0) { return a0; } T_float32x2 f1_0(T_float32x2 a0) { return a0; }
// CHECK: define <4 x float> @f1_1(<4 x float> %{{.*}}) // CHECK: define <4 x float> @f1_1(<4 x float> %{{.*}})
T_float32x4 f1_1(T_float32x4 a0) { return a0; } T_float32x4 f1_1(T_float32x4 a0) { return a0; }
// Vector with length bigger than 16-byte is illegal and is passed indirectly. // Vector with length bigger than 16-byte is illegal and is passed indirectly.
// CHECK: define void @f1_2(<8 x float>* noalias sret %{{.*}}, <8 x float>* %0) // CHECK: define void @f1_2(<8 x float>* noalias sret align 16 %{{.*}}, <8 x float>* %0)
T_float32x8 f1_2(T_float32x8 a0) { return a0; } T_float32x8 f1_2(T_float32x8 a0) { return a0; }
// CHECK: define void @f1_3(<16 x float>* noalias sret %{{.*}}, <16 x float>* %0) // CHECK: define void @f1_3(<16 x float>* noalias sret align 16 %{{.*}}, <16 x float>* %0)
T_float32x16 f1_3(T_float32x16 a0) { return a0; } T_float32x16 f1_3(T_float32x16 a0) { return a0; }
// Testing alignment with aggregates: HFA, aggregates with size <= 16 bytes and // Testing alignment with aggregates: HFA, aggregates with size <= 16 bytes and
// aggregates with size > 16 bytes. // aggregates with size > 16 bytes.
struct s35 struct s35
{ {
float v[4]; //Testing HFA. float v[4]; //Testing HFA.
} __attribute__((aligned(16))); } __attribute__((aligned(16)));
▲ Show 20 Lines • Show All 492 Lines • Show Last 20 Lines

clang/test/CodeGen/arm64-microsoft-arguments.cpp

Show All 22 Lines
S2 func2(S2 x); S2 func2(S2 x);
S2 f2() { S2 f2() {
S2 x; S2 x;
return func2(x); return func2(x);
} }
// Pass and return for type size > 16 bytes. // Pass and return for type size > 16 bytes.
// CHECK: define {{.*}} void @{{.*}}f3{{.*}}(%struct.S3* noalias sret %agg.result) // CHECK: define {{.*}} void @{{.*}}f3{{.*}}(%struct.S3* noalias sret align 4 %agg.result)
// CHECK: call void {{.*}}func3{{.*}}(%struct.S3* sret %agg.result, %struct.S3* %agg.tmp) // CHECK: call void {{.*}}func3{{.*}}(%struct.S3* sret align 4 %agg.result, %struct.S3* %agg.tmp)
struct S3 { struct S3 {
int a[5]; int a[5];
}; };
S3 func3(S3 x); S3 func3(S3 x);
S3 f3() { S3 f3() {
S3 x; S3 x;
return func3(x); return func3(x);
} }
// Pass and return aggregate (of size < 16 bytes) with non-trivial destructor. // Pass and return aggregate (of size < 16 bytes) with non-trivial destructor.
// Passed directly but returned indirectly. // Passed directly but returned indirectly.
// CHECK: define {{.*}} void {{.*}}f4{{.*}}(%struct.S4* inreg noalias sret %agg.result) // CHECK: define {{.*}} void {{.*}}f4{{.*}}(%struct.S4* inreg noalias sret align 4 %agg.result)
// CHECK: call void {{.*}}func4{{.*}}(%struct.S4* inreg sret %agg.result, [2 x i64] %5) // CHECK: call void {{.*}}func4{{.*}}(%struct.S4* inreg sret align 4 %agg.result, [2 x i64] %5)
struct S4 { struct S4 {
int a[3]; int a[3];
~S4(); ~S4();
}; };
S4 func4(S4 x); S4 func4(S4 x);
S4 f4() { S4 f4() {
S4 x; S4 x;
return func4(x); return func4(x);
} }
// Pass and return from instance method called from instance method. // Pass and return from instance method called from instance method.
// CHECK: define {{.*}} void @{{.*}}bar@Q1{{.*}}(%class.Q1* %this, %class.P1* inreg noalias sret %agg.result) // CHECK: define {{.*}} void @{{.*}}bar@Q1{{.*}}(%class.Q1* %this, %class.P1* inreg noalias sret align 1 %agg.result)
// CHECK: call void {{.*}}foo@P1{{.*}}(%class.P1* %ref.tmp, %class.P1* inreg sret %agg.result, i8 %1) // CHECK: call void {{.*}}foo@P1{{.*}}(%class.P1* %ref.tmp, %class.P1* inreg sret align 1 %agg.result, i8 %1)
class P1 { class P1 {
public: public:
P1 foo(P1 x); P1 foo(P1 x);
}; };
class Q1 { class Q1 {
public: public:
P1 bar(); P1 bar();
}; };
P1 Q1::bar() { P1 Q1::bar() {
P1 p1; P1 p1;
return P1().foo(p1); return P1().foo(p1);
} }
// Pass and return from instance method called from free function. // Pass and return from instance method called from free function.
// CHECK: define {{.*}} void {{.*}}bar{{.*}}() // CHECK: define {{.*}} void {{.*}}bar{{.*}}()
// CHECK: call void {{.*}}foo@P2{{.*}}(%class.P2* %ref.tmp, %class.P2* inreg sret %retval, i8 %0) // CHECK: call void {{.*}}foo@P2{{.*}}(%class.P2* %ref.tmp, %class.P2* inreg sret align 1 %retval, i8 %0)
class P2 { class P2 {
public: public:
P2 foo(P2 x); P2 foo(P2 x);
}; };
P2 bar() { P2 bar() {
P2 p2; P2 p2;
return P2().foo(p2); return P2().foo(p2);
} }
// Pass and return an object with a user-provided constructor (passed directly, // Pass and return an object with a user-provided constructor (passed directly,
// returned indirectly) // returned indirectly)
// CHECK: define {{.*}} void @{{.*}}f5{{.*}}(%struct.S5* inreg noalias sret %agg.result) // CHECK: define {{.*}} void @{{.*}}f5{{.*}}(%struct.S5* inreg noalias sret align 4 %agg.result)
// CHECK: call void {{.*}}func5{{.*}}(%struct.S5* inreg sret %agg.result, i64 {{.*}}) // CHECK: call void {{.*}}func5{{.*}}(%struct.S5* inreg sret align 4 %agg.result, i64 {{.*}})
struct S5 { struct S5 {
S5(); S5();
int x; int x;
}; };
S5 func5(S5 x); S5 func5(S5 x);
S5 f5() { S5 f5() {
S5 x; S5 x;
Show All 39 Lines
// Pass and return an object with a non-trivial default destructor (passed // Pass and return an object with a non-trivial default destructor (passed
// directly, returne indirectly) // directly, returne indirectly)
struct S8 { struct S8 {
S8a x; S8a x;
int y; int y;
}; };
// CHECK: define {{.*}} void {{.*}}?f8{{.*}}(%struct.S8* inreg noalias sret {{.*}}) // CHECK: define {{.*}} void {{.*}}?f8{{.*}}(%struct.S8* inreg noalias sret align 4 {{.*}})
// CHECK: call void {{.*}}func8{{.*}}(%struct.S8* inreg sret {{.*}}, i64 {{.*}}) // CHECK: call void {{.*}}func8{{.*}}(%struct.S8* inreg sret align 4 {{.*}}, i64 {{.*}})
S8 func8(S8 x); S8 func8(S8 x);
S8 f8() { S8 f8() {
S8 x; S8 x;
return func8(x); return func8(x);
} }
// Pass and return an object with a non-trivial copy-assignment operator and // Pass and return an object with a non-trivial copy-assignment operator and
// a trivial copy constructor (passed directly, returned indirectly) // a trivial copy constructor (passed directly, returned indirectly)
// CHECK: define {{.*}} void @"?f9@@YA?AUS9@@XZ"(%struct.S9* inreg noalias sret {{.*}}) // CHECK: define {{.*}} void @"?f9@@YA?AUS9@@XZ"(%struct.S9* inreg noalias sret align 4 {{.*}})
// CHECK: call void {{.*}}func9{{.*}}(%struct.S9* inreg sret {{.*}}, i64 {{.*}}) // CHECK: call void {{.*}}func9{{.*}}(%struct.S9* inreg sret align 4 {{.*}}, i64 {{.*}})
struct S9 { struct S9 {
S9& operator=(const S9&); S9& operator=(const S9&);
int x; int x;
}; };
S9 func9(S9 x); S9 func9(S9 x);
S9 f9() { S9 f9() {
S9 x; S9 x;
S9 y = x; S9 y = x;
x = y; x = y;
return func9(x); return func9(x);
} }
// Pass and return an object with a base class (passed directly, returned // Pass and return an object with a base class (passed directly, returned
// indirectly). // indirectly).
// CHECK: define dso_local void {{.*}}f10{{.*}}(%struct.S10* inreg noalias sret {{.*}}) // CHECK: define dso_local void {{.*}}f10{{.*}}(%struct.S10* inreg noalias sret align 4 {{.*}})
// CHECK: call void {{.*}}func10{{.*}}(%struct.S10* inreg sret {{.*}}, [2 x i64] {{.*}}) // CHECK: call void {{.*}}func10{{.*}}(%struct.S10* inreg sret align 4 {{.*}}, [2 x i64] {{.*}})
struct S10 : public S1 { struct S10 : public S1 {
int x; int x;
}; };
S10 func10(S10 x); S10 func10(S10 x);
S10 f10() { S10 f10() {
S10 x; S10 x;
return func10(x); return func10(x);
} }
// Pass and return a non aggregate object exceeding > 128 bits (passed // Pass and return a non aggregate object exceeding > 128 bits (passed
// indirectly, returned indirectly) // indirectly, returned indirectly)
// CHECK: define dso_local void {{.*}}f11{{.*}}(%struct.S11* inreg noalias sret {{.*}}) // CHECK: define dso_local void {{.*}}f11{{.*}}(%struct.S11* inreg noalias sret align 8 {{.*}})
// CHECK: call void {{.*}}func11{{.*}}(%struct.S11* inreg sret {{.*}}, %struct.S11* {{.*}}) // CHECK: call void {{.*}}func11{{.*}}(%struct.S11* inreg sret align 8 {{.*}}, %struct.S11* {{.*}})
struct S11 { struct S11 {
virtual void f(); virtual void f();
int a[5]; int a[5];
}; };
S11 func11(S11 x); S11 func11(S11 x);
S11 f11() { S11 f11() {
S11 x; S11 x;
return func11(x); return func11(x);
} }

clang/test/CodeGen/arm64_32.c

Show All 21 Lines
// Not expected to change, but it's a difference between AAPCS and DarwinPCS // Not expected to change, but it's a difference between AAPCS and DarwinPCS
// that we need to be preserved for compatibility with ARMv7k. // that we need to be preserved for compatibility with ARMv7k.
long double LongDoubleVar = 0.0; long double LongDoubleVar = 0.0;
// CHECK: @LongDoubleVar = global double // CHECK: @LongDoubleVar = global double
typedef float __attribute__((ext_vector_type(16))) v16f32; typedef float __attribute__((ext_vector_type(16))) v16f32;
v16f32 func(v16f32 in) { return in; } v16f32 func(v16f32 in) { return in; }
// CHECK: define void @func(<16 x float>* noalias sret {{%.*}}, <16 x float> {{%.*}}) // CHECK: define void @func(<16 x float>* noalias sret align 16 {{%.*}}, <16 x float> {{%.*}})

clang/test/CodeGen/arm_neon_intrinsics.c

Show First 20 Lines • Show All 492 Lines • ▼ Show 20 Lines
// CHECK: [[ARRAYIDX6_I:%.*]] = getelementptr inbounds [4 x <8 x i8>], [4 x <8 x i8>]* [[VAL5_I]], i32 0, i32 3 // CHECK: [[ARRAYIDX6_I:%.*]] = getelementptr inbounds [4 x <8 x i8>], [4 x <8 x i8>]* [[VAL5_I]], i32 0, i32 3
// CHECK: [[TMP7:%.*]] = load <8 x i8>, <8 x i8>* [[ARRAYIDX6_I]], align 8 // CHECK: [[TMP7:%.*]] = load <8 x i8>, <8 x i8>* [[ARRAYIDX6_I]], align 8
// CHECK: [[VTBX4_I:%.*]] = call <8 x i8> @llvm.arm.neon.vtbx4(<8 x i8> %a, <8 x i8> [[TMP4]], <8 x i8> [[TMP5]], <8 x i8> [[TMP6]], <8 x i8> [[TMP7]], <8 x i8> %c) // CHECK: [[VTBX4_I:%.*]] = call <8 x i8> @llvm.arm.neon.vtbx4(<8 x i8> %a, <8 x i8> [[TMP4]], <8 x i8> [[TMP5]], <8 x i8> [[TMP6]], <8 x i8> [[TMP7]], <8 x i8> %c)
// CHECK: ret <8 x i8> [[VTBX4_I]] // CHECK: ret <8 x i8> [[VTBX4_I]]
poly8x8_t test_vtbx4_p8(poly8x8_t a, poly8x8x4_t b, uint8x8_t c) { poly8x8_t test_vtbx4_p8(poly8x8_t a, poly8x8x4_t b, uint8x8_t c) {
return vtbx4_p8(a, b, c); return vtbx4_p8(a, b, c);
} }
// CHECK: @test_vtrn_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_s8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !3 // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !3
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
// CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !3 // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !3
// CHECK: ret void // CHECK: ret void
int8x8x2_t test_vtrn_s8(int8x8_t a, int8x8_t b) { int8x8x2_t test_vtrn_s8(int8x8_t a, int8x8_t b) {
return vtrn_s8(a, b); return vtrn_s8(a, b);
} }
// CHECK: @test_vtrn_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_s16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !6 // CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !6
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !6 // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !6
// CHECK: ret void // CHECK: ret void
int16x4x2_t test_vtrn_s16(int16x4_t a, int16x4_t b) { int16x4x2_t test_vtrn_s16(int16x4_t a, int16x4_t b) {
return vtrn_s16(a, b); return vtrn_s16(a, b);
} }
// CHECK: @test_vtrn_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_s32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VTRN_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VTRN_I]], <2 x i32>* [[TMP3]], !alias.scope !9 // CHECK: store <2 x i32> [[VTRN_I]], <2 x i32>* [[TMP3]], !alias.scope !9
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]], !alias.scope !9 // CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]], !alias.scope !9
// CHECK: ret void // CHECK: ret void
int32x2x2_t test_vtrn_s32(int32x2_t a, int32x2_t b) { int32x2x2_t test_vtrn_s32(int32x2_t a, int32x2_t b) {
return vtrn_s32(a, b); return vtrn_s32(a, b);
} }
// CHECK: @test_vtrn_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_u8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !12 // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !12
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
// CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !12 // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !12
// CHECK: ret void // CHECK: ret void
uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) { uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) {
return vtrn_u8(a, b); return vtrn_u8(a, b);
} }
// CHECK: @test_vtrn_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_u16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !15 // CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !15
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !15 // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !15
// CHECK: ret void // CHECK: ret void
uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) { uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) {
return vtrn_u16(a, b); return vtrn_u16(a, b);
} }
// CHECK: @test_vtrn_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_u32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VTRN_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VTRN_I]], <2 x i32>* [[TMP3]], !alias.scope !18 // CHECK: store <2 x i32> [[VTRN_I]], <2 x i32>* [[TMP3]], !alias.scope !18
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]], !alias.scope !18 // CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]], !alias.scope !18
// CHECK: ret void // CHECK: ret void
uint32x2x2_t test_vtrn_u32(uint32x2_t a, uint32x2_t b) { uint32x2x2_t test_vtrn_u32(uint32x2_t a, uint32x2_t b) {
return vtrn_u32(a, b); return vtrn_u32(a, b);
} }
// CHECK: @test_vtrn_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_f32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VTRN_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x float> [[VTRN_I]], <2 x float>* [[TMP3]], !alias.scope !21 // CHECK: store <2 x float> [[VTRN_I]], <2 x float>* [[TMP3]], !alias.scope !21
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x float> [[VTRN1_I]], <2 x float>* [[TMP4]], !alias.scope !21 // CHECK: store <2 x float> [[VTRN1_I]], <2 x float>* [[TMP4]], !alias.scope !21
// CHECK: ret void // CHECK: ret void
float32x2x2_t test_vtrn_f32(float32x2_t a, float32x2_t b) { float32x2x2_t test_vtrn_f32(float32x2_t a, float32x2_t b) {
return vtrn_f32(a, b); return vtrn_f32(a, b);
} }
// CHECK: @test_vtrn_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_p8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !24 // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]], !alias.scope !24
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
// CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !24 // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]], !alias.scope !24
// CHECK: ret void // CHECK: ret void
poly8x8x2_t test_vtrn_p8(poly8x8_t a, poly8x8_t b) { poly8x8x2_t test_vtrn_p8(poly8x8_t a, poly8x8_t b) {
return vtrn_p8(a, b); return vtrn_p8(a, b);
} }
// CHECK: @test_vtrn_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrn_p16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !27 // CHECK: store <4 x i16> [[VTRN_I]], <4 x i16>* [[TMP3]], !alias.scope !27
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !27 // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]], !alias.scope !27
// CHECK: ret void // CHECK: ret void
poly16x4x2_t test_vtrn_p16(poly16x4_t a, poly16x4_t b) { poly16x4x2_t test_vtrn_p16(poly16x4_t a, poly16x4_t b) {
return vtrn_p16(a, b); return vtrn_p16(a, b);
} }
// CHECK: @test_vtrnq_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_s8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30>
// CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !30 // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !30
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31>
// CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !30 // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !30
// CHECK: ret void // CHECK: ret void
int8x16x2_t test_vtrnq_s8(int8x16_t a, int8x16_t b) { int8x16x2_t test_vtrnq_s8(int8x16_t a, int8x16_t b) {
return vtrnq_s8(a, b); return vtrnq_s8(a, b);
} }
// CHECK: @test_vtrnq_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_s16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !33 // CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !33
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
// CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]], !alias.scope !33 // CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]], !alias.scope !33
// CHECK: ret void // CHECK: ret void
int16x8x2_t test_vtrnq_s16(int16x8_t a, int16x8_t b) { int16x8x2_t test_vtrnq_s16(int16x8_t a, int16x8_t b) {
return vtrnq_s16(a, b); return vtrnq_s16(a, b);
} }
// CHECK: @test_vtrnq_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_s32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x i32> [[VTRN_I]], <4 x i32>* [[TMP3]], !alias.scope !36 // CHECK: store <4 x i32> [[VTRN_I]], <4 x i32>* [[TMP3]], !alias.scope !36
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]], !alias.scope !36 // CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]], !alias.scope !36
// CHECK: ret void // CHECK: ret void
int32x4x2_t test_vtrnq_s32(int32x4_t a, int32x4_t b) { int32x4x2_t test_vtrnq_s32(int32x4_t a, int32x4_t b) {
return vtrnq_s32(a, b); return vtrnq_s32(a, b);
} }
// CHECK: @test_vtrnq_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_u8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30>
// CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !39 // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !39
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31>
// CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !39 // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !39
// CHECK: ret void // CHECK: ret void
uint8x16x2_t test_vtrnq_u8(uint8x16_t a, uint8x16_t b) { uint8x16x2_t test_vtrnq_u8(uint8x16_t a, uint8x16_t b) {
return vtrnq_u8(a, b); return vtrnq_u8(a, b);
} }
// CHECK: @test_vtrnq_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_u16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !42 // CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !42
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
// CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]], !alias.scope !42 // CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]], !alias.scope !42
// CHECK: ret void // CHECK: ret void
uint16x8x2_t test_vtrnq_u16(uint16x8_t a, uint16x8_t b) { uint16x8x2_t test_vtrnq_u16(uint16x8_t a, uint16x8_t b) {
return vtrnq_u16(a, b); return vtrnq_u16(a, b);
} }
// CHECK: @test_vtrnq_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_u32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x i32> [[VTRN_I]], <4 x i32>* [[TMP3]], !alias.scope !45 // CHECK: store <4 x i32> [[VTRN_I]], <4 x i32>* [[TMP3]], !alias.scope !45
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]], !alias.scope !45 // CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]], !alias.scope !45
// CHECK: ret void // CHECK: ret void
uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) { uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) {
return vtrnq_u32(a, b); return vtrnq_u32(a, b);
} }
// CHECK: @test_vtrnq_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_f32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6> // CHECK: [[VTRN_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 2, i32 6>
// CHECK: store <4 x float> [[VTRN_I]], <4 x float>* [[TMP3]], !alias.scope !48 // CHECK: store <4 x float> [[VTRN_I]], <4 x float>* [[TMP3]], !alias.scope !48
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7>
// CHECK: store <4 x float> [[VTRN1_I]], <4 x float>* [[TMP4]], !alias.scope !48 // CHECK: store <4 x float> [[VTRN1_I]], <4 x float>* [[TMP4]], !alias.scope !48
// CHECK: ret void // CHECK: ret void
float32x4x2_t test_vtrnq_f32(float32x4_t a, float32x4_t b) { float32x4x2_t test_vtrnq_f32(float32x4_t a, float32x4_t b) {
return vtrnq_f32(a, b); return vtrnq_f32(a, b);
} }
// CHECK: @test_vtrnq_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_p8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30>
// CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !51 // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]], !alias.scope !51
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31>
// CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !51 // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]], !alias.scope !51
// CHECK: ret void // CHECK: ret void
poly8x16x2_t test_vtrnq_p8(poly8x16_t a, poly8x16_t b) { poly8x16x2_t test_vtrnq_p8(poly8x16_t a, poly8x16_t b) {
return vtrnq_p8(a, b); return vtrnq_p8(a, b);
} }
// CHECK: @test_vtrnq_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vtrnq_p16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14>
// CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !54 // CHECK: store <8 x i16> [[VTRN_I]], <8 x i16>* [[TMP3]], !alias.scope !54
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15>
▲ Show 20 Lines • Show All 162 Lines • ▼ Show 20 Lines
// CHECK: [[TMP2:%.*]] = and <8 x i16> %a, %b // CHECK: [[TMP2:%.*]] = and <8 x i16> %a, %b
// CHECK: [[TMP3:%.*]] = icmp ne <8 x i16> [[TMP2]], zeroinitializer // CHECK: [[TMP3:%.*]] = icmp ne <8 x i16> [[TMP2]], zeroinitializer
// CHECK: [[VTST_I:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i16> // CHECK: [[VTST_I:%.*]] = sext <8 x i1> [[TMP3]] to <8 x i16>
// CHECK: ret <8 x i16> [[VTST_I]] // CHECK: ret <8 x i16> [[VTST_I]]
uint16x8_t test_vtstq_p16(poly16x8_t a, poly16x8_t b) { uint16x8_t test_vtstq_p16(poly16x8_t a, poly16x8_t b) {
return vtstq_p16(a, b); return vtstq_p16(a, b);
} }
// CHECK: @test_vuzp_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_s8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !57 // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !57
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !57 // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !57
// CHECK: ret void // CHECK: ret void
int8x8x2_t test_vuzp_s8(int8x8_t a, int8x8_t b) { int8x8x2_t test_vuzp_s8(int8x8_t a, int8x8_t b) {
return vuzp_s8(a, b); return vuzp_s8(a, b);
} }
// CHECK: @test_vuzp_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_s16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !60 // CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !60
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !60 // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !60
// CHECK: ret void // CHECK: ret void
int16x4x2_t test_vuzp_s16(int16x4_t a, int16x4_t b) { int16x4x2_t test_vuzp_s16(int16x4_t a, int16x4_t b) {
return vuzp_s16(a, b); return vuzp_s16(a, b);
} }
// CHECK: @test_vuzp_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_s32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VUZP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VUZP_I]], <2 x i32>* [[TMP3]], !alias.scope !63 // CHECK: store <2 x i32> [[VUZP_I]], <2 x i32>* [[TMP3]], !alias.scope !63
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]], !alias.scope !63 // CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]], !alias.scope !63
// CHECK: ret void // CHECK: ret void
int32x2x2_t test_vuzp_s32(int32x2_t a, int32x2_t b) { int32x2x2_t test_vuzp_s32(int32x2_t a, int32x2_t b) {
return vuzp_s32(a, b); return vuzp_s32(a, b);
} }
// CHECK: @test_vuzp_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_u8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !66 // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !66
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !66 // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !66
// CHECK: ret void // CHECK: ret void
uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) { uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) {
return vuzp_u8(a, b); return vuzp_u8(a, b);
} }
// CHECK: @test_vuzp_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_u16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !69 // CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !69
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !69 // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !69
// CHECK: ret void // CHECK: ret void
uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) { uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) {
return vuzp_u16(a, b); return vuzp_u16(a, b);
} }
// CHECK: @test_vuzp_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_u32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VUZP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VUZP_I]], <2 x i32>* [[TMP3]], !alias.scope !72 // CHECK: store <2 x i32> [[VUZP_I]], <2 x i32>* [[TMP3]], !alias.scope !72
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]], !alias.scope !72 // CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]], !alias.scope !72
// CHECK: ret void // CHECK: ret void
uint32x2x2_t test_vuzp_u32(uint32x2_t a, uint32x2_t b) { uint32x2x2_t test_vuzp_u32(uint32x2_t a, uint32x2_t b) {
return vuzp_u32(a, b); return vuzp_u32(a, b);
} }
// CHECK: @test_vuzp_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_f32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VUZP_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x float> [[VUZP_I]], <2 x float>* [[TMP3]], !alias.scope !75 // CHECK: store <2 x float> [[VUZP_I]], <2 x float>* [[TMP3]], !alias.scope !75
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x float> [[VUZP1_I]], <2 x float>* [[TMP4]], !alias.scope !75 // CHECK: store <2 x float> [[VUZP1_I]], <2 x float>* [[TMP4]], !alias.scope !75
// CHECK: ret void // CHECK: ret void
float32x2x2_t test_vuzp_f32(float32x2_t a, float32x2_t b) { float32x2x2_t test_vuzp_f32(float32x2_t a, float32x2_t b) {
return vuzp_f32(a, b); return vuzp_f32(a, b);
} }
// CHECK: @test_vuzp_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_p8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !78 // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]], !alias.scope !78
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !78 // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]], !alias.scope !78
// CHECK: ret void // CHECK: ret void
poly8x8x2_t test_vuzp_p8(poly8x8_t a, poly8x8_t b) { poly8x8x2_t test_vuzp_p8(poly8x8_t a, poly8x8_t b) {
return vuzp_p8(a, b); return vuzp_p8(a, b);
} }
// CHECK: @test_vuzp_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzp_p16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !81 // CHECK: store <4 x i16> [[VUZP_I]], <4 x i16>* [[TMP3]], !alias.scope !81
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !81 // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]], !alias.scope !81
// CHECK: ret void // CHECK: ret void
poly16x4x2_t test_vuzp_p16(poly16x4_t a, poly16x4_t b) { poly16x4x2_t test_vuzp_p16(poly16x4_t a, poly16x4_t b) {
return vuzp_p16(a, b); return vuzp_p16(a, b);
} }
// CHECK: @test_vuzpq_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_s8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30>
// CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !84 // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !84
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31>
// CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !84 // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !84
// CHECK: ret void // CHECK: ret void
int8x16x2_t test_vuzpq_s8(int8x16_t a, int8x16_t b) { int8x16x2_t test_vuzpq_s8(int8x16_t a, int8x16_t b) {
return vuzpq_s8(a, b); return vuzpq_s8(a, b);
} }
// CHECK: @test_vuzpq_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_s16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !87 // CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !87
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !87 // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !87
// CHECK: ret void // CHECK: ret void
int16x8x2_t test_vuzpq_s16(int16x8_t a, int16x8_t b) { int16x8x2_t test_vuzpq_s16(int16x8_t a, int16x8_t b) {
return vuzpq_s16(a, b); return vuzpq_s16(a, b);
} }
// CHECK: @test_vuzpq_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_s32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x i32> [[VUZP_I]], <4 x i32>* [[TMP3]], !alias.scope !90 // CHECK: store <4 x i32> [[VUZP_I]], <4 x i32>* [[TMP3]], !alias.scope !90
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]], !alias.scope !90 // CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]], !alias.scope !90
// CHECK: ret void // CHECK: ret void
int32x4x2_t test_vuzpq_s32(int32x4_t a, int32x4_t b) { int32x4x2_t test_vuzpq_s32(int32x4_t a, int32x4_t b) {
return vuzpq_s32(a, b); return vuzpq_s32(a, b);
} }
// CHECK: @test_vuzpq_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_u8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30>
// CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !93 // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !93
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31>
// CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !93 // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !93
// CHECK: ret void // CHECK: ret void
uint8x16x2_t test_vuzpq_u8(uint8x16_t a, uint8x16_t b) { uint8x16x2_t test_vuzpq_u8(uint8x16_t a, uint8x16_t b) {
return vuzpq_u8(a, b); return vuzpq_u8(a, b);
} }
// CHECK: @test_vuzpq_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_u16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !96 // CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !96
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !96 // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !96
// CHECK: ret void // CHECK: ret void
uint16x8x2_t test_vuzpq_u16(uint16x8_t a, uint16x8_t b) { uint16x8x2_t test_vuzpq_u16(uint16x8_t a, uint16x8_t b) {
return vuzpq_u16(a, b); return vuzpq_u16(a, b);
} }
// CHECK: @test_vuzpq_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_u32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x i32> [[VUZP_I]], <4 x i32>* [[TMP3]], !alias.scope !99 // CHECK: store <4 x i32> [[VUZP_I]], <4 x i32>* [[TMP3]], !alias.scope !99
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]], !alias.scope !99 // CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]], !alias.scope !99
// CHECK: ret void // CHECK: ret void
uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) { uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) {
return vuzpq_u32(a, b); return vuzpq_u32(a, b);
} }
// CHECK: @test_vuzpq_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_f32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6> // CHECK: [[VUZP_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
// CHECK: store <4 x float> [[VUZP_I]], <4 x float>* [[TMP3]], !alias.scope !102 // CHECK: store <4 x float> [[VUZP_I]], <4 x float>* [[TMP3]], !alias.scope !102
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
// CHECK: store <4 x float> [[VUZP1_I]], <4 x float>* [[TMP4]], !alias.scope !102 // CHECK: store <4 x float> [[VUZP1_I]], <4 x float>* [[TMP4]], !alias.scope !102
// CHECK: ret void // CHECK: ret void
float32x4x2_t test_vuzpq_f32(float32x4_t a, float32x4_t b) { float32x4x2_t test_vuzpq_f32(float32x4_t a, float32x4_t b) {
return vuzpq_f32(a, b); return vuzpq_f32(a, b);
} }
// CHECK: @test_vuzpq_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_p8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30>
// CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !105 // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]], !alias.scope !105
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31>
// CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !105 // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]], !alias.scope !105
// CHECK: ret void // CHECK: ret void
poly8x16x2_t test_vuzpq_p8(poly8x16_t a, poly8x16_t b) { poly8x16x2_t test_vuzpq_p8(poly8x16_t a, poly8x16_t b) {
return vuzpq_p8(a, b); return vuzpq_p8(a, b);
} }
// CHECK: @test_vuzpq_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vuzpq_p16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
// CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !108 // CHECK: store <8 x i16> [[VUZP_I]], <8 x i16>* [[TMP3]], !alias.scope !108
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
// CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !108 // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]], !alias.scope !108
// CHECK: ret void // CHECK: ret void
poly16x8x2_t test_vuzpq_p16(poly16x8_t a, poly16x8_t b) { poly16x8x2_t test_vuzpq_p16(poly16x8_t a, poly16x8_t b) {
return vuzpq_p16(a, b); return vuzpq_p16(a, b);
} }
// CHECK: @test_vzip_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_s8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !111 // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !111
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !111 // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !111
// CHECK: ret void // CHECK: ret void
int8x8x2_t test_vzip_s8(int8x8_t a, int8x8_t b) { int8x8x2_t test_vzip_s8(int8x8_t a, int8x8_t b) {
return vzip_s8(a, b); return vzip_s8(a, b);
} }
// CHECK: @test_vzip_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_s16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !114 // CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !114
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !114 // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !114
// CHECK: ret void // CHECK: ret void
int16x4x2_t test_vzip_s16(int16x4_t a, int16x4_t b) { int16x4x2_t test_vzip_s16(int16x4_t a, int16x4_t b) {
return vzip_s16(a, b); return vzip_s16(a, b);
} }
// CHECK: @test_vzip_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_s32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VZIP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VZIP_I]], <2 x i32>* [[TMP3]], !alias.scope !117 // CHECK: store <2 x i32> [[VZIP_I]], <2 x i32>* [[TMP3]], !alias.scope !117
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]], !alias.scope !117 // CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]], !alias.scope !117
// CHECK: ret void // CHECK: ret void
int32x2x2_t test_vzip_s32(int32x2_t a, int32x2_t b) { int32x2x2_t test_vzip_s32(int32x2_t a, int32x2_t b) {
return vzip_s32(a, b); return vzip_s32(a, b);
} }
// CHECK: @test_vzip_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_u8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !120 // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !120
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !120 // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !120
// CHECK: ret void // CHECK: ret void
uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) { uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) {
return vzip_u8(a, b); return vzip_u8(a, b);
} }
// CHECK: @test_vzip_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_u16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !123 // CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !123
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !123 // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !123
// CHECK: ret void // CHECK: ret void
uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) { uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) {
return vzip_u16(a, b); return vzip_u16(a, b);
} }
// CHECK: @test_vzip_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_u32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VZIP_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x i32> [[VZIP_I]], <2 x i32>* [[TMP3]], !alias.scope !126 // CHECK: store <2 x i32> [[VZIP_I]], <2 x i32>* [[TMP3]], !alias.scope !126
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]], !alias.scope !126 // CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]], !alias.scope !126
// CHECK: ret void // CHECK: ret void
uint32x2x2_t test_vzip_u32(uint32x2_t a, uint32x2_t b) { uint32x2x2_t test_vzip_u32(uint32x2_t a, uint32x2_t b) {
return vzip_u32(a, b); return vzip_u32(a, b);
} }
// CHECK: @test_vzip_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_f32({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2> // CHECK: [[VZIP_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 0, i32 2>
// CHECK: store <2 x float> [[VZIP_I]], <2 x float>* [[TMP3]], !alias.scope !129 // CHECK: store <2 x float> [[VZIP_I]], <2 x float>* [[TMP3]], !alias.scope !129
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3>
// CHECK: store <2 x float> [[VZIP1_I]], <2 x float>* [[TMP4]], !alias.scope !129 // CHECK: store <2 x float> [[VZIP1_I]], <2 x float>* [[TMP4]], !alias.scope !129
// CHECK: ret void // CHECK: ret void
float32x2x2_t test_vzip_f32(float32x2_t a, float32x2_t b) { float32x2x2_t test_vzip_f32(float32x2_t a, float32x2_t b) {
return vzip_f32(a, b); return vzip_f32(a, b);
} }
// CHECK: @test_vzip_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_p8({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !132 // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]], !alias.scope !132
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !132 // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]], !alias.scope !132
// CHECK: ret void // CHECK: ret void
poly8x8x2_t test_vzip_p8(poly8x8_t a, poly8x8_t b) { poly8x8x2_t test_vzip_p8(poly8x8_t a, poly8x8_t b) {
return vzip_p8(a, b); return vzip_p8(a, b);
} }
// CHECK: @test_vzip_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzip_p16({{.*}} sret align 8 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !135 // CHECK: store <4 x i16> [[VZIP_I]], <4 x i16>* [[TMP3]], !alias.scope !135
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !135 // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]], !alias.scope !135
// CHECK: ret void // CHECK: ret void
poly16x4x2_t test_vzip_p16(poly16x4_t a, poly16x4_t b) { poly16x4x2_t test_vzip_p16(poly16x4_t a, poly16x4_t b) {
return vzip_p16(a, b); return vzip_p16(a, b);
} }
// CHECK: @test_vzipq_s8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_s8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23>
// CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !138 // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !138
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
// CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !138 // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !138
// CHECK: ret void // CHECK: ret void
int8x16x2_t test_vzipq_s8(int8x16_t a, int8x16_t b) { int8x16x2_t test_vzipq_s8(int8x16_t a, int8x16_t b) {
return vzipq_s8(a, b); return vzipq_s8(a, b);
} }
// CHECK: @test_vzipq_s16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_s16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !141 // CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !141
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !141 // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !141
// CHECK: ret void // CHECK: ret void
int16x8x2_t test_vzipq_s16(int16x8_t a, int16x8_t b) { int16x8x2_t test_vzipq_s16(int16x8_t a, int16x8_t b) {
return vzipq_s16(a, b); return vzipq_s16(a, b);
} }
// CHECK: @test_vzipq_s32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_s32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x i32> [[VZIP_I]], <4 x i32>* [[TMP3]], !alias.scope !144 // CHECK: store <4 x i32> [[VZIP_I]], <4 x i32>* [[TMP3]], !alias.scope !144
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]], !alias.scope !144 // CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]], !alias.scope !144
// CHECK: ret void // CHECK: ret void
int32x4x2_t test_vzipq_s32(int32x4_t a, int32x4_t b) { int32x4x2_t test_vzipq_s32(int32x4_t a, int32x4_t b) {
return vzipq_s32(a, b); return vzipq_s32(a, b);
} }
// CHECK: @test_vzipq_u8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_u8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23>
// CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !147 // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !147
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
// CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !147 // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !147
// CHECK: ret void // CHECK: ret void
uint8x16x2_t test_vzipq_u8(uint8x16_t a, uint8x16_t b) { uint8x16x2_t test_vzipq_u8(uint8x16_t a, uint8x16_t b) {
return vzipq_u8(a, b); return vzipq_u8(a, b);
} }
// CHECK: @test_vzipq_u16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_u16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !150 // CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !150
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !150 // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !150
// CHECK: ret void // CHECK: ret void
uint16x8x2_t test_vzipq_u16(uint16x8_t a, uint16x8_t b) { uint16x8x2_t test_vzipq_u16(uint16x8_t a, uint16x8_t b) {
return vzipq_u16(a, b); return vzipq_u16(a, b);
} }
// CHECK: @test_vzipq_u32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_u32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x i32> [[VZIP_I]], <4 x i32>* [[TMP3]], !alias.scope !153 // CHECK: store <4 x i32> [[VZIP_I]], <4 x i32>* [[TMP3]], !alias.scope !153
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]], !alias.scope !153 // CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]], !alias.scope !153
// CHECK: ret void // CHECK: ret void
uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) { uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) {
return vzipq_u32(a, b); return vzipq_u32(a, b);
} }
// CHECK: @test_vzipq_f32({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_f32({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5> // CHECK: [[VZIP_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 0, i32 4, i32 1, i32 5>
// CHECK: store <4 x float> [[VZIP_I]], <4 x float>* [[TMP3]], !alias.scope !156 // CHECK: store <4 x float> [[VZIP_I]], <4 x float>* [[TMP3]], !alias.scope !156
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7>
// CHECK: store <4 x float> [[VZIP1_I]], <4 x float>* [[TMP4]], !alias.scope !156 // CHECK: store <4 x float> [[VZIP1_I]], <4 x float>* [[TMP4]], !alias.scope !156
// CHECK: ret void // CHECK: ret void
float32x4x2_t test_vzipq_f32(float32x4_t a, float32x4_t b) { float32x4x2_t test_vzipq_f32(float32x4_t a, float32x4_t b) {
return vzipq_f32(a, b); return vzipq_f32(a, b);
} }
// CHECK: @test_vzipq_p8({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_p8({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23>
// CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !159 // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]], !alias.scope !159
// CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31>
// CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !159 // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]], !alias.scope !159
// CHECK: ret void // CHECK: ret void
poly8x16x2_t test_vzipq_p8(poly8x16_t a, poly8x16_t b) { poly8x16x2_t test_vzipq_p8(poly8x16_t a, poly8x16_t b) {
return vzipq_p8(a, b); return vzipq_p8(a, b);
} }
// CHECK: @test_vzipq_p16({{.*}} sret [[AGG_RESULT:%[0-9a-zA-Z.]+]], // CHECK: @test_vzipq_p16({{.*}} sret align 16 [[AGG_RESULT:%[0-9a-zA-Z.]+]],
// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8* // CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[AGG_RESULT]] to i8*
// CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8>
// CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8>
// CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>*
// CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11>
// CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !162 // CHECK: store <8 x i16> [[VZIP_I]], <8 x i16>* [[TMP3]], !alias.scope !162
// CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1
// CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
// CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !162 // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]], !alias.scope !162
// CHECK: ret void // CHECK: ret void
poly16x8x2_t test_vzipq_p16(poly16x8_t a, poly16x8_t b) { poly16x8x2_t test_vzipq_p16(poly16x8_t a, poly16x8_t b) {
return vzipq_p16(a, b); return vzipq_p16(a, b);
} }
Context not available.

clang/test/CodeGen/blocks.c

Show All 12 Lines
} }
// Verify that attributes on blocks are set correctly. // Verify that attributes on blocks are set correctly.
typedef struct s0 T; typedef struct s0 T;
struct s0 { struct s0 {
int a[64]; int a[64];
}; };
// CHECK: define internal void @__f2_block_invoke(%struct.s0* noalias sret {{%.*}}, i8* {{%.*}}, %struct.s0* byval(%struct.s0) align 4 {{.*}}) // CHECK: define internal void @__f2_block_invoke(%struct.s0* noalias sret align 4 {{%.*}}, i8* {{%.*}}, %struct.s0* byval(%struct.s0) align 4 {{.*}})
struct s0 f2(struct s0 a0) { struct s0 f2(struct s0 a0) {
return ^(struct s0 a1){ return a1; }(a0); return ^(struct s0 a1){ return a1; }(a0);
} }
// This should not crash: rdar://6808051 // This should not crash: rdar://6808051
void *P = ^{ void *P = ^{
void *Q = __func__; void *Q = __func__;
}; };
▲ Show 20 Lines • Show All 95 Lines • Show Last 20 Lines

clang/test/CodeGen/c11atomics-ios.c

Show First 20 Lines • Show All 197 Lines • ▼ Show 20 Lines
// CHECK-NEXT: [[T6:%.*]] = bitcast [[APS]]* [[T0]] to i64* // CHECK-NEXT: [[T6:%.*]] = bitcast [[APS]]* [[T0]] to i64*
// CHECK-NEXT: store atomic i64 [[T5]], i64* [[T6]] seq_cst, align 8 // CHECK-NEXT: store atomic i64 [[T5]], i64* [[T6]] seq_cst, align 8
*fp = f; *fp = f;
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
} }
PS test_promoted_load(_Atomic(PS) *addr) { PS test_promoted_load(_Atomic(PS) *addr) {
// CHECK-LABEL: @test_promoted_load(%struct.PS* noalias sret %agg.result, { %struct.PS, [2 x i8] }* %addr) // CHECK-LABEL: @test_promoted_load(%struct.PS* noalias sret align 2 %agg.result, { %struct.PS, [2 x i8] }* %addr)
// CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4 // CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4
// CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64* // CHECK: [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
// CHECK: [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64* // CHECK: [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64*
// CHECK: [[VAL:%.*]] = load atomic i64, i64* [[ADDR64]] seq_cst, align 8 // CHECK: [[VAL:%.*]] = load atomic i64, i64* [[ADDR64]] seq_cst, align 8
// CHECK: store i64 [[VAL]], i64* [[ATOMIC_RES64]], align 8 // CHECK: store i64 [[VAL]], i64* [[ATOMIC_RES64]], align 8
Show All 25 Linesvoid test_promoted_store(_Atomic(PS) *addr, PS *val) {
// CHECK: [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64* // CHECK: [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
// CHECK: [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 8 // CHECK: [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 8
// CHECK: store atomic i64 [[VAL64]], i64* [[ADDR64]] seq_cst, align 8 // CHECK: store atomic i64 [[VAL64]], i64* [[ADDR64]] seq_cst, align 8
__c11_atomic_store(addr, *val, 5); __c11_atomic_store(addr, *val, 5);
} }
PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) { PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
// CHECK-LABEL: @test_promoted_exchange(%struct.PS* noalias sret %agg.result, { %struct.PS, [2 x i8] }* %addr, %struct.PS* %val) // CHECK-LABEL: @test_promoted_exchange(%struct.PS* noalias sret align 2 %agg.result, { %struct.PS, [2 x i8] }* %addr, %struct.PS* %val)
// CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4 // CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4
// CHECK: [[VAL_ARG:%.*]] = alloca %struct.PS*, align 4 // CHECK: [[VAL_ARG:%.*]] = alloca %struct.PS*, align 4
// CHECK: [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2 // CHECK: [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2
// CHECK: [[ATOMIC_VAL:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_VAL:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: store %struct.PS* %val, %struct.PS** [[VAL_ARG]], align 4 // CHECK: store %struct.PS* %val, %struct.PS** [[VAL_ARG]], align 4
// CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
▲ Show 20 Lines • Show All 81 Lines • Show Last 20 Lines

clang/test/CodeGen/c11atomics.c

Show First 20 Lines • Show All 362 Lines • ▼ Show 20 Lines
// CHECK-NEXT: [[T2:%.*]] = sext i16 [[T1]] to i32 // CHECK-NEXT: [[T2:%.*]] = sext i16 [[T1]] to i32
// CHECK-NEXT: store i32 [[T2]], i32* [[A]], align 4 // CHECK-NEXT: store i32 [[T2]], i32* [[A]], align 4
int a = ((PS)*fp).x; int a = ((PS)*fp).x;
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
} }
PS test_promoted_load(_Atomic(PS) *addr) { PS test_promoted_load(_Atomic(PS) *addr) {
// CHECK-LABEL: @test_promoted_load(%struct.PS* noalias sret %agg.result, { %struct.PS, [2 x i8] }* %addr) // CHECK-LABEL: @test_promoted_load(%struct.PS* noalias sret align 2 %agg.result, { %struct.PS, [2 x i8] }* %addr)
// CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4 // CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4
// CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64* // CHECK: [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
// CHECK: [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64* // CHECK: [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64*
// CHECK: [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8* // CHECK: [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8*
// CHECK: [[RES:%.*]] = call arm_aapcscc i64 @__atomic_load_8(i8* [[ADDR8]], i32 5) // CHECK: [[RES:%.*]] = call arm_aapcscc i64 @__atomic_load_8(i8* [[ADDR8]], i32 5)
Show All 26 Linesvoid test_promoted_store(_Atomic(PS) *addr, PS *val) {
// CHECK: [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64* // CHECK: [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
// CHECK: [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8* // CHECK: [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8*
// CHECK: [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 2 // CHECK: [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 2
// CHECK: call arm_aapcscc void @__atomic_store_8(i8* [[ADDR8]], i64 [[VAL64]], i32 5) // CHECK: call arm_aapcscc void @__atomic_store_8(i8* [[ADDR8]], i64 [[VAL64]], i32 5)
__c11_atomic_store(addr, *val, 5); __c11_atomic_store(addr, *val, 5);
} }
PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) { PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
// CHECK-LABEL: @test_promoted_exchange(%struct.PS* noalias sret %agg.result, { %struct.PS, [2 x i8] }* %addr, %struct.PS* %val) // CHECK-LABEL: @test_promoted_exchange(%struct.PS* noalias sret align 2 %agg.result, { %struct.PS, [2 x i8] }* %addr, %struct.PS* %val)
// CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4 // CHECK: [[ADDR_ARG:%.*]] = alloca { %struct.PS, [2 x i8] }*, align 4
// CHECK: [[VAL_ARG:%.*]] = alloca %struct.PS*, align 4 // CHECK: [[VAL_ARG:%.*]] = alloca %struct.PS*, align 4
// CHECK: [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2 // CHECK: [[NONATOMIC_TMP:%.*]] = alloca %struct.PS, align 2
// CHECK: [[ATOMIC_VAL:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_VAL:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_RES:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8
// CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: store { %struct.PS, [2 x i8] }* %addr, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
// CHECK: store %struct.PS* %val, %struct.PS** [[VAL_ARG]], align 4 // CHECK: store %struct.PS* %val, %struct.PS** [[VAL_ARG]], align 4
// CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4 // CHECK: [[ADDR:%.*]] = load { %struct.PS, [2 x i8] }*, { %struct.PS, [2 x i8] }** [[ADDR_ARG]], align 4
▲ Show 20 Lines • Show All 68 Lines • Show Last 20 Lines

clang/test/CodeGen/lanai-arguments.c

Show All 10 Linestypedef struct {
int bb; int bb;
} s1; } s1;
// CHECK: define void @f1(i32 inreg %i.coerce0, i32 inreg %i.coerce1) // CHECK: define void @f1(i32 inreg %i.coerce0, i32 inreg %i.coerce1)
void f1(s1 i) {} void f1(s1 i) {}
typedef struct { typedef struct {
int cc; int cc;
} s2; } s2;
// CHECK: define void @f2(%struct.s2* noalias sret %agg.result) // CHECK: define void @f2(%struct.s2* noalias sret align 4 %agg.result)
s2 f2() { s2 f2() {
s2 foo; s2 foo;
return foo; return foo;
} }
typedef struct { typedef struct {
int cc; int cc;
int dd; int dd;
} s3; } s3;
// CHECK: define void @f3(%struct.s3* noalias sret %agg.result) // CHECK: define void @f3(%struct.s3* noalias sret align 4 %agg.result)
s3 f3() { s3 f3() {
s3 foo; s3 foo;
return foo; return foo;
} }
// CHECK: define void @f4(i64 inreg %i) // CHECK: define void @f4(i64 inreg %i)
void f4(long long i) {} void f4(long long i) {}
Show All 38 Lines

clang/test/CodeGen/le32-arguments.c

Show All 11 Lines
// Structs should be passed byval and not split up // Structs should be passed byval and not split up
// CHECK-LABEL: define void @f1(%struct.s1* byval(%struct.s1) align 4 %i) // CHECK-LABEL: define void @f1(%struct.s1* byval(%struct.s1) align 4 %i)
void f1(s1 i) {} void f1(s1 i) {}
typedef struct { typedef struct {
int cc; int cc;
} s2; } s2;
// Structs should be returned sret and not simplified by the frontend // Structs should be returned sret and not simplified by the frontend
// CHECK-LABEL: define void @f2(%struct.s2* noalias sret %agg.result) // CHECK-LABEL: define void @f2(%struct.s2* noalias sret align 4 %agg.result)
s2 f2() { s2 f2() {
s2 foo; s2 foo;
return foo; return foo;
} }
// CHECK-LABEL: define void @f3(i64 %i) // CHECK-LABEL: define void @f3(i64 %i)
void f3(long long i) {} void f3(long long i) {}
Show All 33 Lines

clang/test/CodeGen/mcu-struct-return.c

Show All 36 Lines
union U1 foo1() { return u1; } union U1 foo1() { return u1; }
union U2 foo2() { return u2; } union U2 foo2() { return u2; }
union U3 foo3() { return u3; } union U3 foo3() { return u3; }
struct S1 bar1() { return s1; } struct S1 bar1() { return s1; }
struct S2 bar2() { return s2; } struct S2 bar2() { return s2; }
struct S1 bar3(union U1 u) { return s1; } struct S1 bar3(union U1 u) { return s1; }
// CHECK: define void @foo1() // CHECK: define void @foo1()
// CHECK: define void @foo2([[UNION2_TYPE]]* noalias sret %{{.+}}) // CHECK: define void @foo2([[UNION2_TYPE]]* noalias sret align 4 %{{.+}})
// CHECK: define i32 @foo3() // CHECK: define i32 @foo3()
// CHECK: define void @bar1() // CHECK: define void @bar1()
// CHECK: define i32 @bar2() // CHECK: define i32 @bar2()
// CHECK: define void @bar3() // CHECK: define void @bar3()
void run() { void run() {
union U1 x1 = foo1(); union U1 x1 = foo1();
union U2 x2 = foo2(); union U2 x2 = foo2();
union U3 x3 = foo3(); union U3 x3 = foo3();
struct S1 y1 = bar1(); struct S1 y1 = bar1();
struct S2 y2 = bar2(); struct S2 y2 = bar2();
struct S1 y3 = bar3(x1); struct S1 y3 = bar3(x1);
// CHECK: [[X1:%.+]] = alloca [[UNION1_TYPE]] // CHECK: [[X1:%.+]] = alloca [[UNION1_TYPE]]
// CHECK: [[X2:%.+]] = alloca [[UNION2_TYPE]] // CHECK: [[X2:%.+]] = alloca [[UNION2_TYPE]]
// CHECK: [[X3:%.+]] = alloca [[UNION3_TYPE]] // CHECK: [[X3:%.+]] = alloca [[UNION3_TYPE]]
// CHECK: [[Y1:%.+]] = alloca [[STRUCT1_TYPE]] // CHECK: [[Y1:%.+]] = alloca [[STRUCT1_TYPE]]
// CHECK: [[Y2:%.+]] = alloca [[STRUCT2_TYPE]] // CHECK: [[Y2:%.+]] = alloca [[STRUCT2_TYPE]]
// CHECK: call void @foo1() // CHECK: call void @foo1()
// CHECK: call void @foo2([[UNION2_TYPE]]* sret [[X2]]) // CHECK: call void @foo2([[UNION2_TYPE]]* sret align 4 [[X2]])
// CHECK: {{.+}} = call i32 @foo3() // CHECK: {{.+}} = call i32 @foo3()
// CHECK: call void @bar1() // CHECK: call void @bar1()
// CHECK: {{.+}} = call i32 @bar2() // CHECK: {{.+}} = call i32 @bar2()
// CHECK: call void @bar3() // CHECK: call void @bar3()
} }

clang/test/CodeGen/mingw-long-double.c

Show All 26 Lines
// GNU32: @dataLDC = dso_local global { x86_fp80, x86_fp80 } { x86_fp80 0xK3FFF8000000000000000, x86_fp80 0xK3FFF8000000000000000 }, align 4 // GNU32: @dataLDC = dso_local global { x86_fp80, x86_fp80 } { x86_fp80 0xK3FFF8000000000000000, x86_fp80 0xK3FFF8000000000000000 }, align 4
// GNU64: @dataLDC = dso_local global { x86_fp80, x86_fp80 } { x86_fp80 0xK3FFF8000000000000000, x86_fp80 0xK3FFF8000000000000000 }, align 16 // GNU64: @dataLDC = dso_local global { x86_fp80, x86_fp80 } { x86_fp80 0xK3FFF8000000000000000, x86_fp80 0xK3FFF8000000000000000 }, align 16
// MSC64: @dataLDC = dso_local global { double, double } { double 1.000000e+00, double 1.000000e+00 }, align 8 // MSC64: @dataLDC = dso_local global { double, double } { double 1.000000e+00, double 1.000000e+00 }, align 8
long double TestLD(long double x) { long double TestLD(long double x) {
return x * x; return x * x;
} }
// GNU32: define dso_local x86_fp80 @TestLD(x86_fp80 %x) // GNU32: define dso_local x86_fp80 @TestLD(x86_fp80 %x)
// GNU64: define dso_local void @TestLD(x86_fp80* noalias sret %agg.result, x86_fp80* %0) // GNU64: define dso_local void @TestLD(x86_fp80* noalias sret align 16 %agg.result, x86_fp80* %0)
// MSC64: define dso_local double @TestLD(double %x) // MSC64: define dso_local double @TestLD(double %x)
long double _Complex TestLDC(long double _Complex x) { long double _Complex TestLDC(long double _Complex x) {
return x * x; return x * x;
} }
// GNU32: define dso_local void @TestLDC({ x86_fp80, x86_fp80 }* noalias sret %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 4 %x) // GNU32: define dso_local void @TestLDC({ x86_fp80, x86_fp80 }* noalias sret align 4 %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 4 %x)
// GNU64: define dso_local void @TestLDC({ x86_fp80, x86_fp80 }* noalias sret %agg.result, { x86_fp80, x86_fp80 }* %x) // GNU64: define dso_local void @TestLDC({ x86_fp80, x86_fp80 }* noalias sret align 16 %agg.result, { x86_fp80, x86_fp80 }* %x)
// MSC64: define dso_local void @TestLDC({ double, double }* noalias sret %agg.result, { double, double }* %x) // MSC64: define dso_local void @TestLDC({ double, double }* noalias sret align 8 %agg.result, { double, double }* %x)
// GNU32: declare dso_local void @__mulxc3 // GNU32: declare dso_local void @__mulxc3
// GNU64: declare dso_local void @__mulxc3 // GNU64: declare dso_local void @__mulxc3
// MSC64: declare dso_local void @__muldc3 // MSC64: declare dso_local void @__muldc3

clang/test/CodeGen/mips-zero-sized-struct.c

Show All 13 Lines
// RUN: %clang_cc1 -triple mips64el-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mips64el-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mipsisa64r6-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mipsisa64r6-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mipsisa64r6el-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mipsisa64r6el-unknown-linux-gnu -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mips64-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mips64-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mips64el-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mips64el-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mipsisa64r6-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mipsisa64r6-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// RUN: %clang_cc1 -triple mipsisa64r6el-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s // RUN: %clang_cc1 -triple mipsisa64r6el-unknown-linux-gnuabi64 -S -emit-llvm -o - %s | FileCheck -check-prefix=N64 %s
// O32: define void @fn28(%struct.T2* noalias sret %agg.result, i8 signext %arg0) // O32: define void @fn28(%struct.T2* noalias sret align 1 %agg.result, i8 signext %arg0)
// N32: define void @fn28(i8 signext %arg0) // N32: define void @fn28(i8 signext %arg0)
// N64: define void @fn28(i8 signext %arg0) // N64: define void @fn28(i8 signext %arg0)
typedef struct T2 { } T2; typedef struct T2 { } T2;
T2 T2_retval; T2 T2_retval;
T2 fn28(char arg0) { T2 fn28(char arg0) {
return T2_retval; return T2_retval;
} }

clang/test/CodeGen/mips64-padding-arg.c

Show All 27 Lines
extern void foo4(int, int, int, long double); extern void foo4(int, int, int, long double);
void foo3(int a0, long double a1) { void foo3(int a0, long double a1) {
foo4(1, 2, a0, a1); foo4(1, 2, a0, a1);
} }
// Insert padding after hidden argument. // Insert padding after hidden argument.
// //
// N64-LABEL: define void @foo5(%struct.S0* noalias sret %agg.result, i64 %0, fp128 %a0) // N64-LABEL: define void @foo5(%struct.S0* noalias sret align 16 %agg.result, i64 %0, fp128 %a0)
// N64: call void @foo6(%struct.S0* sret %agg.result, i32 signext 1, i32 signext 2, i64 undef, fp128 %a0) // N64: call void @foo6(%struct.S0* sret align 16 %agg.result, i32 signext 1, i32 signext 2, i64 undef, fp128 %a0)
// N64: declare void @foo6(%struct.S0* sret, i32 signext, i32 signext, i64, fp128) // N64: declare void @foo6(%struct.S0* sret align 16, i32 signext, i32 signext, i64, fp128)
extern S0 foo6(int, int, long double); extern S0 foo6(int, int, long double);
S0 foo5(long double a0) { S0 foo5(long double a0) {
return foo6(1, 2, a0); return foo6(1, 2, a0);
} }
// Do not insert padding if ABI is O32. // Do not insert padding if ABI is O32.
Show All 25 Lines

clang/test/CodeGen/ms_abi.c

Show First 20 Lines • Show All 149 Lines • ▼ Show 20 Lines
// This test checks if structs are passed according to Win64 calling convention // This test checks if structs are passed according to Win64 calling convention
// when it's enforced by __attribute((ms_abi)). // when it's enforced by __attribute((ms_abi)).
struct i128 { struct i128 {
unsigned long long a; unsigned long long a;
unsigned long long b; unsigned long long b;
}; };
__attribute__((ms_abi)) struct i128 f7(struct i128 a) { __attribute__((ms_abi)) struct i128 f7(struct i128 a) {
// WIN64: define dso_local void @f7(%struct.i128* noalias sret %agg.result, %struct.i128* %a) // WIN64: define dso_local void @f7(%struct.i128* noalias sret align 8 %agg.result, %struct.i128* %a)
// FREEBSD: define win64cc void @f7(%struct.i128* noalias sret %agg.result, %struct.i128* %a) // FREEBSD: define win64cc void @f7(%struct.i128* noalias sret align 8 %agg.result, %struct.i128* %a)
return a; return a;
} }

clang/test/CodeGen/ppc64-align-struct.c

Show First 20 Lines • Show All 42 Lines • ▼ Show 20 Lines
// This case requires run-time realignment of the incoming struct // This case requires run-time realignment of the incoming struct
// CHECK-LABEL: define void @test7(i32 signext %x, %struct.test7* byval(%struct.test7) align 16 %0) // CHECK-LABEL: define void @test7(i32 signext %x, %struct.test7* byval(%struct.test7) align 16 %0)
// CHECK: %y = alloca %struct.test7, align 32 // CHECK: %y = alloca %struct.test7, align 32
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64
void test7 (int x, struct test7 y) void test7 (int x, struct test7 y)
{ {
} }
// CHECK: define void @test1va(%struct.test1* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @test1va(%struct.test1* noalias sret align 4 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 8 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 8
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test1* // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test1*
// CHECK: [[DEST:%.*]] = bitcast %struct.test1* %[[AGG_RESULT]] to i8* // CHECK: [[DEST:%.*]] = bitcast %struct.test1* %[[AGG_RESULT]] to i8*
// CHECK: [[SRC:%.*]] = bitcast %struct.test1* [[T0]] to i8* // CHECK: [[SRC:%.*]] = bitcast %struct.test1* [[T0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 8, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 8, i1 false)
struct test1 test1va (int x, ...) struct test1 test1va (int x, ...)
{ {
struct test1 y; struct test1 y;
va_list ap; va_list ap;
va_start(ap, x); va_start(ap, x);
y = va_arg (ap, struct test1); y = va_arg (ap, struct test1);
va_end(ap); va_end(ap);
return y; return y;
} }
// CHECK: define void @test2va(%struct.test2* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @test2va(%struct.test2* noalias sret align 16 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64 // CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64
// CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15 // CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15
// CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16 // CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16
// CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8* // CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8*
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 16 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 16
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test2* // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test2*
// CHECK: [[DEST:%.*]] = bitcast %struct.test2* %[[AGG_RESULT]] to i8* // CHECK: [[DEST:%.*]] = bitcast %struct.test2* %[[AGG_RESULT]] to i8*
// CHECK: [[SRC:%.*]] = bitcast %struct.test2* [[T0]] to i8* // CHECK: [[SRC:%.*]] = bitcast %struct.test2* [[T0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 16 [[SRC]], i64 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 16 [[SRC]], i64 16, i1 false)
struct test2 test2va (int x, ...) struct test2 test2va (int x, ...)
{ {
struct test2 y; struct test2 y;
va_list ap; va_list ap;
va_start(ap, x); va_start(ap, x);
y = va_arg (ap, struct test2); y = va_arg (ap, struct test2);
va_end(ap); va_end(ap);
return y; return y;
} }
// CHECK: define void @test3va(%struct.test3* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @test3va(%struct.test3* noalias sret align 32 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64 // CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64
// CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15 // CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15
// CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16 // CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16
// CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8* // CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8*
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 32 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 32
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test3* // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test3*
// CHECK: [[DEST:%.*]] = bitcast %struct.test3* %[[AGG_RESULT]] to i8* // CHECK: [[DEST:%.*]] = bitcast %struct.test3* %[[AGG_RESULT]] to i8*
// CHECK: [[SRC:%.*]] = bitcast %struct.test3* [[T0]] to i8* // CHECK: [[SRC:%.*]] = bitcast %struct.test3* [[T0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 32 [[DEST]], i8* align 16 [[SRC]], i64 32, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 32 [[DEST]], i8* align 16 [[SRC]], i64 32, i1 false)
struct test3 test3va (int x, ...) struct test3 test3va (int x, ...)
{ {
struct test3 y; struct test3 y;
va_list ap; va_list ap;
va_start(ap, x); va_start(ap, x);
y = va_arg (ap, struct test3); y = va_arg (ap, struct test3);
va_end(ap); va_end(ap);
return y; return y;
} }
// CHECK: define void @test4va(%struct.test4* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @test4va(%struct.test4* noalias sret align 4 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 16 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 16
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test4* // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test4*
// CHECK: [[DEST:%.*]] = bitcast %struct.test4* %[[AGG_RESULT]] to i8* // CHECK: [[DEST:%.*]] = bitcast %struct.test4* %[[AGG_RESULT]] to i8*
// CHECK: [[SRC:%.*]] = bitcast %struct.test4* [[T0]] to i8* // CHECK: [[SRC:%.*]] = bitcast %struct.test4* [[T0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 12, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 12, i1 false)
struct test4 test4va (int x, ...) struct test4 test4va (int x, ...)
{ {
struct test4 y; struct test4 y;
va_list ap; va_list ap;
va_start(ap, x); va_start(ap, x);
y = va_arg (ap, struct test4); y = va_arg (ap, struct test4);
va_end(ap); va_end(ap);
return y; return y;
} }
// CHECK: define void @testva_longdouble(%struct.test_longdouble* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @testva_longdouble(%struct.test_longdouble* noalias sret align 16 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 16 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[CUR]], i64 16
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test_longdouble* // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test_longdouble*
// CHECK: [[DEST:%.*]] = bitcast %struct.test_longdouble* %[[AGG_RESULT]] to i8* // CHECK: [[DEST:%.*]] = bitcast %struct.test_longdouble* %[[AGG_RESULT]] to i8*
// CHECK: [[SRC:%.*]] = bitcast %struct.test_longdouble* [[T0]] to i8* // CHECK: [[SRC:%.*]] = bitcast %struct.test_longdouble* [[T0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 8 [[SRC]], i64 16, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 8 [[SRC]], i64 16, i1 false)
struct test_longdouble { long double x; }; struct test_longdouble { long double x; };
struct test_longdouble testva_longdouble (int x, ...) struct test_longdouble testva_longdouble (int x, ...)
{ {
struct test_longdouble y; struct test_longdouble y;
va_list ap; va_list ap;
va_start(ap, x); va_start(ap, x);
y = va_arg (ap, struct test_longdouble); y = va_arg (ap, struct test_longdouble);
va_end(ap); va_end(ap);
return y; return y;
} }
// CHECK: define void @testva_vector(%struct.test_vector* noalias sret %[[AGG_RESULT:.*]], i32 signext %x, ...) // CHECK: define void @testva_vector(%struct.test_vector* noalias sret align 16 %[[AGG_RESULT:.*]], i32 signext %x, ...)
// CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap // CHECK: %[[CUR:[^ ]+]] = load i8*, i8** %ap
// CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64 // CHECK: %[[TMP0:[^ ]+]] = ptrtoint i8* %[[CUR]] to i64
// CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15 // CHECK: %[[TMP1:[^ ]+]] = add i64 %[[TMP0]], 15
// CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16 // CHECK: %[[TMP2:[^ ]+]] = and i64 %[[TMP1]], -16
// CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8* // CHECK: %[[ALIGN:[^ ]+]] = inttoptr i64 %[[TMP2]] to i8*
// CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 16 // CHECK: %[[NEXT:[^ ]+]] = getelementptr inbounds i8, i8* %[[ALIGN]], i64 16
// CHECK: store i8* %[[NEXT]], i8** %ap // CHECK: store i8* %[[NEXT]], i8** %ap
// CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test_vector* // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test_vector*
Show All 14 Lines

clang/test/CodeGen/ppc64-elf-abi.c

Show All 11 Lines
// RUN: -target-abi elfv2 | FileCheck %s --check-prefix=CHECK-ELFv2 // RUN: -target-abi elfv2 | FileCheck %s --check-prefix=CHECK-ELFv2
// RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \ // RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \
// RUN: | FileCheck %s --check-prefix=CHECK-ELFv2 // RUN: | FileCheck %s --check-prefix=CHECK-ELFv2
// RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \ // RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \
// RUN: -target-abi elfv1 | FileCheck %s --check-prefix=CHECK-ELFv1 // RUN: -target-abi elfv1 | FileCheck %s --check-prefix=CHECK-ELFv1
// RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \ // RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm -o - %s \
// RUN: -target-abi elfv2 | FileCheck %s --check-prefix=CHECK-ELFv2 // RUN: -target-abi elfv2 | FileCheck %s --check-prefix=CHECK-ELFv2
// CHECK-ELFv1: define void @func_fab(%struct.fab* noalias sret %agg.result, i64 %x.coerce) // CHECK-ELFv1: define void @func_fab(%struct.fab* noalias sret align 4 %agg.result, i64 %x.coerce)
// CHECK-ELFv2: define [2 x float] @func_fab([2 x float] %x.coerce) // CHECK-ELFv2: define [2 x float] @func_fab([2 x float] %x.coerce)
struct fab { float a; float b; }; struct fab { float a; float b; };
struct fab func_fab(struct fab x) { return x; } struct fab func_fab(struct fab x) { return x; }

clang/test/CodeGen/ppc64-qpx-vector.c

Show All 18 Lines
v4df foo2(struct sdf a, v4df b, struct sdf2 c) { v4df foo2(struct sdf a, v4df b, struct sdf2 c) {
return a.v + b; return a.v + b;
} }
// QPX-LABEL: define <4 x double> @foo2(<4 x double> inreg %a.coerce, <4 x double> %b, [2 x i256] %c.coerce) // QPX-LABEL: define <4 x double> @foo2(<4 x double> inreg %a.coerce, <4 x double> %b, [2 x i256] %c.coerce)
// QPX: ret <4 x double> // QPX: ret <4 x double>
// NORMAL-LABEL: define void @foo2(<4 x double>* noalias sret %agg.result, [2 x i128] %a.coerce, <4 x double>* %0, [4 x i128] %c.coerce) // NORMAL-LABEL: define void @foo2(<4 x double>* noalias sret align 32 %agg.result, [2 x i128] %a.coerce, <4 x double>* %0, [4 x i128] %c.coerce)
// NORMAL: ret void // NORMAL: ret void

clang/test/CodeGen/ppc64-soft-float.c

Show All 24 Lines
struct f9 { float f[9]; }; struct f9 { float f[9]; };
struct fab { float a; float b; }; struct fab { float a; float b; };
struct fabc { float a; float b; float c; }; struct fabc { float a; float b; float c; };
struct f2a2b { float a[2]; float b[2]; }; struct f2a2b { float a[2]; float b[2]; };
// CHECK-LE: define i32 @func_f1(float inreg %x.coerce) // CHECK-LE: define i32 @func_f1(float inreg %x.coerce)
// CHECK-BE: define void @func_f1(%struct.f1* noalias sret %agg.result, float inreg %x.coerce) // CHECK-BE: define void @func_f1(%struct.f1* noalias sret align 4 %agg.result, float inreg %x.coerce)
struct f1 func_f1(struct f1 x) { return x; } struct f1 func_f1(struct f1 x) { return x; }
// CHECK-LE: define i64 @func_f2(i64 %x.coerce) // CHECK-LE: define i64 @func_f2(i64 %x.coerce)
// CHECK-BE: define void @func_f2(%struct.f2* noalias sret %agg.result, i64 %x.coerce) // CHECK-BE: define void @func_f2(%struct.f2* noalias sret align 4 %agg.result, i64 %x.coerce)
struct f2 func_f2(struct f2 x) { return x; } struct f2 func_f2(struct f2 x) { return x; }
// CHECK-LE: define { i64, i64 } @func_f3([2 x i64] %x.coerce) // CHECK-LE: define { i64, i64 } @func_f3([2 x i64] %x.coerce)
// CHECK-BE: define void @func_f3(%struct.f3* noalias sret %agg.result, [2 x i64] %x.coerce) // CHECK-BE: define void @func_f3(%struct.f3* noalias sret align 4 %agg.result, [2 x i64] %x.coerce)
struct f3 func_f3(struct f3 x) { return x; } struct f3 func_f3(struct f3 x) { return x; }
// CHECK-LE: define { i64, i64 } @func_f4([2 x i64] %x.coerce) // CHECK-LE: define { i64, i64 } @func_f4([2 x i64] %x.coerce)
// CHECK-BE: define void @func_f4(%struct.f4* noalias sret %agg.result, [2 x i64] %x.coerce) // CHECK-BE: define void @func_f4(%struct.f4* noalias sret align 4 %agg.result, [2 x i64] %x.coerce)
struct f4 func_f4(struct f4 x) { return x; } struct f4 func_f4(struct f4 x) { return x; }
// CHECK: define void @func_f5(%struct.f5* noalias sret %agg.result, [3 x i64] %x.coerce) // CHECK: define void @func_f5(%struct.f5* noalias sret align 4 %agg.result, [3 x i64] %x.coerce)
struct f5 func_f5(struct f5 x) { return x; } struct f5 func_f5(struct f5 x) { return x; }
// CHECK: define void @func_f6(%struct.f6* noalias sret %agg.result, [3 x i64] %x.coerce) // CHECK: define void @func_f6(%struct.f6* noalias sret align 4 %agg.result, [3 x i64] %x.coerce)
struct f6 func_f6(struct f6 x) { return x; } struct f6 func_f6(struct f6 x) { return x; }
// CHECK: define void @func_f7(%struct.f7* noalias sret %agg.result, [4 x i64] %x.coerce) // CHECK: define void @func_f7(%struct.f7* noalias sret align 4 %agg.result, [4 x i64] %x.coerce)
struct f7 func_f7(struct f7 x) { return x; } struct f7 func_f7(struct f7 x) { return x; }
// CHECK: define void @func_f8(%struct.f8* noalias sret %agg.result, [4 x i64] %x.coerce) // CHECK: define void @func_f8(%struct.f8* noalias sret align 4 %agg.result, [4 x i64] %x.coerce)
struct f8 func_f8(struct f8 x) { return x; } struct f8 func_f8(struct f8 x) { return x; }
// CHECK: define void @func_f9(%struct.f9* noalias sret %agg.result, [5 x i64] %x.coerce) // CHECK: define void @func_f9(%struct.f9* noalias sret align 4 %agg.result, [5 x i64] %x.coerce)
struct f9 func_f9(struct f9 x) { return x; } struct f9 func_f9(struct f9 x) { return x; }
// CHECK-LE: define i64 @func_fab(i64 %x.coerce) // CHECK-LE: define i64 @func_fab(i64 %x.coerce)
// CHECK-BE: define void @func_fab(%struct.fab* noalias sret %agg.result, i64 %x.coerce) // CHECK-BE: define void @func_fab(%struct.fab* noalias sret align 4 %agg.result, i64 %x.coerce)
struct fab func_fab(struct fab x) { return x; } struct fab func_fab(struct fab x) { return x; }
// CHECK-LE: define { i64, i64 } @func_fabc([2 x i64] %x.coerce) // CHECK-LE: define { i64, i64 } @func_fabc([2 x i64] %x.coerce)
// CHECK-BE: define void @func_fabc(%struct.fabc* noalias sret %agg.result, [2 x i64] %x.coerce) // CHECK-BE: define void @func_fabc(%struct.fabc* noalias sret align 4 %agg.result, [2 x i64] %x.coerce)
struct fabc func_fabc(struct fabc x) { return x; } struct fabc func_fabc(struct fabc x) { return x; }
// CHECK-LE: define { i64, i64 } @func_f2a2b([2 x i64] %x.coerce) // CHECK-LE: define { i64, i64 } @func_f2a2b([2 x i64] %x.coerce)
// CHECK-BE: define void @func_f2a2b(%struct.f2a2b* noalias sret %agg.result, [2 x i64] %x.coerce) // CHECK-BE: define void @func_f2a2b(%struct.f2a2b* noalias sret align 4 %agg.result, [2 x i64] %x.coerce)
struct f2a2b func_f2a2b(struct f2a2b x) { return x; } struct f2a2b func_f2a2b(struct f2a2b x) { return x; }
// CHECK-LABEL: @call_f1 // CHECK-LABEL: @call_f1
// CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f1, align 4 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f1, align 4
// CHECK: %[[TMP:[^ ]+]] = load float, float* getelementptr inbounds (%struct.f1, %struct.f1* @global_f1, i32 0, i32 0, i32 0), align 4 // CHECK: %[[TMP:[^ ]+]] = load float, float* getelementptr inbounds (%struct.f1, %struct.f1* @global_f1, i32 0, i32 0, i32 0), align 4
// CHECK-LE: call i32 @func_f1(float inreg %[[TMP]]) // CHECK-LE: call i32 @func_f1(float inreg %[[TMP]])
// CHECK-BE: call void @func_f1(%struct.f1* sret %[[TMP0]], float inreg %[[TMP]]) // CHECK-BE: call void @func_f1(%struct.f1* sret align 4 %[[TMP0]], float inreg %[[TMP]])
struct f1 global_f1; struct f1 global_f1;
void call_f1(void) { global_f1 = func_f1(global_f1); } void call_f1(void) { global_f1 = func_f1(global_f1); }
// CHECK-LABEL: @call_f2 // CHECK-LABEL: @call_f2
// CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f2, align 4 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f2, align 4
// CHECK: %[[TMP:[^ ]+]] = load i64, i64* bitcast (%struct.f2* @global_f2 to i64*), align 4 // CHECK: %[[TMP:[^ ]+]] = load i64, i64* bitcast (%struct.f2* @global_f2 to i64*), align 4
// CHECK-LE: call i64 @func_f2(i64 %[[TMP]]) // CHECK-LE: call i64 @func_f2(i64 %[[TMP]])
// CHECK-BE: call void @func_f2(%struct.f2* sret %[[TMP0]], i64 %[[TMP]]) // CHECK-BE: call void @func_f2(%struct.f2* sret align 4 %[[TMP0]], i64 %[[TMP]])
struct f2 global_f2; struct f2 global_f2;
void call_f2(void) { global_f2 = func_f2(global_f2); } void call_f2(void) { global_f2 = func_f2(global_f2); }
// CHECK-LABEL: @call_f3 // CHECK-LABEL: @call_f3
// CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f3, align 4 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f3, align 4
// CHECK: %[[TMP1:[^ ]+]] = alloca [2 x i64] // CHECK: %[[TMP1:[^ ]+]] = alloca [2 x i64]
// CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP1]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP1]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f3* @global_f3 to i8*), i64 12, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f3* @global_f3 to i8*), i64 12, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP1]] // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP1]]
// CHECK-LE: call { i64, i64 } @func_f3([2 x i64] %[[TMP3]]) // CHECK-LE: call { i64, i64 } @func_f3([2 x i64] %[[TMP3]])
// CHECK-BE: call void @func_f3(%struct.f3* sret %[[TMP0]], [2 x i64] %[[TMP3]]) // CHECK-BE: call void @func_f3(%struct.f3* sret align 4 %[[TMP0]], [2 x i64] %[[TMP3]])
struct f3 global_f3; struct f3 global_f3;
void call_f3(void) { global_f3 = func_f3(global_f3); } void call_f3(void) { global_f3 = func_f3(global_f3); }
// CHECK-LABEL: @call_f4 // CHECK-LABEL: @call_f4
// CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f4, align 4 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f4, align 4
// CHECK: %[[TMP:[^ ]+]] = load [2 x i64], [2 x i64]* bitcast (%struct.f4* @global_f4 to [2 x i64]*), align 4 // CHECK: %[[TMP:[^ ]+]] = load [2 x i64], [2 x i64]* bitcast (%struct.f4* @global_f4 to [2 x i64]*), align 4
// CHECK-LE: call { i64, i64 } @func_f4([2 x i64] %[[TMP]]) // CHECK-LE: call { i64, i64 } @func_f4([2 x i64] %[[TMP]])
// CHECK-BE: call void @func_f4(%struct.f4* sret %[[TMP0]], [2 x i64] %[[TMP]]) // CHECK-BE: call void @func_f4(%struct.f4* sret align 4 %[[TMP0]], [2 x i64] %[[TMP]])
struct f4 global_f4; struct f4 global_f4;
void call_f4(void) { global_f4 = func_f4(global_f4); } void call_f4(void) { global_f4 = func_f4(global_f4); }
// CHECK-LABEL: @call_f5 // CHECK-LABEL: @call_f5
// CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f5, align 4 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f5, align 4
// CHECK: %[[TMP1:[^ ]+]] = alloca [3 x i64] // CHECK: %[[TMP1:[^ ]+]] = alloca [3 x i64]
// CHECK: %[[TMP2:[^ ]+]] = bitcast [3 x i64]* %[[TMP1]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [3 x i64]* %[[TMP1]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f5* @global_f5 to i8*), i64 20, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f5* @global_f5 to i8*), i64 20, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [3 x i64], [3 x i64]* %[[TMP1]] // CHECK: %[[TMP3:[^ ]+]] = load [3 x i64], [3 x i64]* %[[TMP1]]
// CHECK: call void @func_f5(%struct.f5* sret %[[TMP0]], [3 x i64] %[[TMP3]]) // CHECK: call void @func_f5(%struct.f5* sret align 4 %[[TMP0]], [3 x i64] %[[TMP3]])
struct f5 global_f5; struct f5 global_f5;
void call_f5(void) { global_f5 = func_f5(global_f5); } void call_f5(void) { global_f5 = func_f5(global_f5); }
// CHECK-LABEL: @call_f6 // CHECK-LABEL: @call_f6
// CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f6, align 4 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f6, align 4
// CHECK: %[[TMP:[^ ]+]] = load [3 x i64], [3 x i64]* bitcast (%struct.f6* @global_f6 to [3 x i64]*), align 4 // CHECK: %[[TMP:[^ ]+]] = load [3 x i64], [3 x i64]* bitcast (%struct.f6* @global_f6 to [3 x i64]*), align 4
// CHECK: call void @func_f6(%struct.f6* sret %[[TMP0]], [3 x i64] %[[TMP]]) // CHECK: call void @func_f6(%struct.f6* sret align 4 %[[TMP0]], [3 x i64] %[[TMP]])
struct f6 global_f6; struct f6 global_f6;
void call_f6(void) { global_f6 = func_f6(global_f6); } void call_f6(void) { global_f6 = func_f6(global_f6); }
// CHECK-LABEL: @call_f7 // CHECK-LABEL: @call_f7
// CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f7, align 4 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f7, align 4
// CHECK: %[[TMP1:[^ ]+]] = alloca [4 x i64], align 8 // CHECK: %[[TMP1:[^ ]+]] = alloca [4 x i64], align 8
// CHECK: %[[TMP2:[^ ]+]] = bitcast [4 x i64]* %[[TMP1]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [4 x i64]* %[[TMP1]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f7* @global_f7 to i8*), i64 28, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f7* @global_f7 to i8*), i64 28, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [4 x i64], [4 x i64]* %[[TMP1]], align 8 // CHECK: %[[TMP3:[^ ]+]] = load [4 x i64], [4 x i64]* %[[TMP1]], align 8
// CHECK: call void @func_f7(%struct.f7* sret %[[TMP0]], [4 x i64] %[[TMP3]]) // CHECK: call void @func_f7(%struct.f7* sret align 4 %[[TMP0]], [4 x i64] %[[TMP3]])
struct f7 global_f7; struct f7 global_f7;
void call_f7(void) { global_f7 = func_f7(global_f7); } void call_f7(void) { global_f7 = func_f7(global_f7); }
// CHECK-LABEL: @call_f8 // CHECK-LABEL: @call_f8
// CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f8, align 4 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f8, align 4
// CHECK: %[[TMP:[^ ]+]] = load [4 x i64], [4 x i64]* bitcast (%struct.f8* @global_f8 to [4 x i64]*), align 4 // CHECK: %[[TMP:[^ ]+]] = load [4 x i64], [4 x i64]* bitcast (%struct.f8* @global_f8 to [4 x i64]*), align 4
// CHECK: call void @func_f8(%struct.f8* sret %[[TMP0]], [4 x i64] %[[TMP]]) // CHECK: call void @func_f8(%struct.f8* sret align 4 %[[TMP0]], [4 x i64] %[[TMP]])
struct f8 global_f8; struct f8 global_f8;
void call_f8(void) { global_f8 = func_f8(global_f8); } void call_f8(void) { global_f8 = func_f8(global_f8); }
// CHECK-LABEL: @call_f9 // CHECK-LABEL: @call_f9
// CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64] // CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64]
// CHECK: %[[TMP2:[^ ]+]] = bitcast [5 x i64]* %[[TMP1]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [5 x i64]* %[[TMP1]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], [5 x i64]* %[[TMP1]] // CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], [5 x i64]* %[[TMP1]]
// CHECK: call void @func_f9(%struct.f9* sret %{{[^ ]+}}, [5 x i64] %[[TMP3]]) // CHECK: call void @func_f9(%struct.f9* sret align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
struct f9 global_f9; struct f9 global_f9;
void call_f9(void) { global_f9 = func_f9(global_f9); } void call_f9(void) { global_f9 = func_f9(global_f9); }
// CHECK-LABEL: @call_fab // CHECK-LABEL: @call_fab
// CHECK: %[[TMP0:[^ ]+]] = alloca %struct.fab, align 4 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.fab, align 4
// CHECK: %[[TMP:[^ ]+]] = load i64, i64* bitcast (%struct.fab* @global_fab to i64*), align 4 // CHECK: %[[TMP:[^ ]+]] = load i64, i64* bitcast (%struct.fab* @global_fab to i64*), align 4
// CHECK-LE: %call = call i64 @func_fab(i64 %[[TMP]]) // CHECK-LE: %call = call i64 @func_fab(i64 %[[TMP]])
// CHECK-BE: call void @func_fab(%struct.fab* sret %[[TMP0]], i64 %[[TMP]]) // CHECK-BE: call void @func_fab(%struct.fab* sret align 4 %[[TMP0]], i64 %[[TMP]])
struct fab global_fab; struct fab global_fab;
void call_fab(void) { global_fab = func_fab(global_fab); } void call_fab(void) { global_fab = func_fab(global_fab); }
// CHECK-LABEL: @call_fabc // CHECK-LABEL: @call_fabc
// CHECK-BE: %[[TMPX:[^ ]+]] = alloca %struct.fabc, align 4 // CHECK-BE: %[[TMPX:[^ ]+]] = alloca %struct.fabc, align 4
// CHECK: %[[TMP0:[^ ]+]] = alloca [2 x i64], align 8 // CHECK: %[[TMP0:[^ ]+]] = alloca [2 x i64], align 8
// CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP0]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP0]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.fabc* @global_fabc to i8*), i64 12, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.fabc* @global_fabc to i8*), i64 12, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP0]], align 8 // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP0]], align 8
// CHECK-LE: %call = call { i64, i64 } @func_fabc([2 x i64] %[[TMP3]]) // CHECK-LE: %call = call { i64, i64 } @func_fabc([2 x i64] %[[TMP3]])
// CHECK-BE: call void @func_fabc(%struct.fabc* sret %[[TMPX]], [2 x i64] %[[TMP3]]) // CHECK-BE: call void @func_fabc(%struct.fabc* sret align 4 %[[TMPX]], [2 x i64] %[[TMP3]])
struct fabc global_fabc; struct fabc global_fabc;
void call_fabc(void) { global_fabc = func_fabc(global_fabc); } void call_fabc(void) { global_fabc = func_fabc(global_fabc); }

clang/test/CodeGen/ppc64-vector.c

Show All 33 Lines
} }
// CHECK: define <8 x i16> @test_v8i16(<8 x i16> %x) // CHECK: define <8 x i16> @test_v8i16(<8 x i16> %x)
v8i16 test_v8i16(v8i16 x) v8i16 test_v8i16(v8i16 x)
{ {
return x; return x;
} }
// CHECK: define void @test_v16i16(<16 x i16>* noalias sret %agg.result, <16 x i16>* %0) // CHECK: define void @test_v16i16(<16 x i16>* noalias sret align 32 %agg.result, <16 x i16>* %0)
v16i16 test_v16i16(v16i16 x) v16i16 test_v16i16(v16i16 x)
{ {
return x; return x;
} }
// CHECK: define void @test_struct_v16i16(%struct.v16i16* noalias sret %agg.result, [2 x i128] %x.coerce) // CHECK: define void @test_struct_v16i16(%struct.v16i16* noalias sret align 32 %agg.result, [2 x i128] %x.coerce)
struct v16i16 test_struct_v16i16(struct v16i16 x) struct v16i16 test_struct_v16i16(struct v16i16 x)
{ {
return x; return x;
} }

clang/test/CodeGen/ppc64le-aggregates.c

Show All 35 Lines
struct f6 func_f6(struct f6 x) { return x; } struct f6 func_f6(struct f6 x) { return x; }
// CHECK: define [7 x float] @func_f7([7 x float] %x.coerce) // CHECK: define [7 x float] @func_f7([7 x float] %x.coerce)
struct f7 func_f7(struct f7 x) { return x; } struct f7 func_f7(struct f7 x) { return x; }
// CHECK: define [8 x float] @func_f8([8 x float] %x.coerce) // CHECK: define [8 x float] @func_f8([8 x float] %x.coerce)
struct f8 func_f8(struct f8 x) { return x; } struct f8 func_f8(struct f8 x) { return x; }
// CHECK: define void @func_f9(%struct.f9* noalias sret %agg.result, [5 x i64] %x.coerce) // CHECK: define void @func_f9(%struct.f9* noalias sret align 4 %agg.result, [5 x i64] %x.coerce)
struct f9 func_f9(struct f9 x) { return x; } struct f9 func_f9(struct f9 x) { return x; }
// CHECK: define [2 x float] @func_fab([2 x float] %x.coerce) // CHECK: define [2 x float] @func_fab([2 x float] %x.coerce)
struct fab func_fab(struct fab x) { return x; } struct fab func_fab(struct fab x) { return x; }
// CHECK: define [3 x float] @func_fabc([3 x float] %x.coerce) // CHECK: define [3 x float] @func_fabc([3 x float] %x.coerce)
struct fabc func_fabc(struct fabc x) { return x; } struct fabc func_fabc(struct fabc x) { return x; }
▲ Show 20 Lines • Show All 48 Lines • ▼ Show 20 Lines
struct f8 global_f8; struct f8 global_f8;
void call_f8(void) { global_f8 = func_f8(global_f8); } void call_f8(void) { global_f8 = func_f8(global_f8); }
// CHECK-LABEL: @call_f9 // CHECK-LABEL: @call_f9
// CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64] // CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64]
// CHECK: %[[TMP2:[^ ]+]] = bitcast [5 x i64]* %[[TMP1]] to i8* // CHECK: %[[TMP2:[^ ]+]] = bitcast [5 x i64]* %[[TMP1]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, i1 false) // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, i1 false)
// CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], [5 x i64]* %[[TMP1]] // CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], [5 x i64]* %[[TMP1]]
// CHECK: call void @func_f9(%struct.f9* sret %{{[^ ]+}}, [5 x i64] %[[TMP3]]) // CHECK: call void @func_f9(%struct.f9* sret align 4 %{{[^ ]+}}, [5 x i64] %[[TMP3]])
struct f9 global_f9; struct f9 global_f9;
void call_f9(void) { global_f9 = func_f9(global_f9); } void call_f9(void) { global_f9 = func_f9(global_f9); }
// CHECK-LABEL: @call_fab // CHECK-LABEL: @call_fab
// CHECK: %[[TMP:[^ ]+]] = load [2 x float], [2 x float]* bitcast (%struct.fab* @global_fab to [2 x float]*) // CHECK: %[[TMP:[^ ]+]] = load [2 x float], [2 x float]* bitcast (%struct.fab* @global_fab to [2 x float]*)
// CHECK: call [2 x float] @func_fab([2 x float] %[[TMP]]) // CHECK: call [2 x float] @func_fab([2 x float] %[[TMP]])
struct fab global_fab; struct fab global_fab;
void call_fab(void) { global_fab = func_fab(global_fab); } void call_fab(void) { global_fab = func_fab(global_fab); }
Show All 39 Lines
struct v6 func_v6(struct v6 x) { return x; } struct v6 func_v6(struct v6 x) { return x; }
// CHECK: define [7 x <4 x i32>] @func_v7([7 x <4 x i32>] %x.coerce) // CHECK: define [7 x <4 x i32>] @func_v7([7 x <4 x i32>] %x.coerce)
struct v7 func_v7(struct v7 x) { return x; } struct v7 func_v7(struct v7 x) { return x; }
// CHECK: define [8 x <4 x i32>] @func_v8([8 x <4 x i32>] %x.coerce) // CHECK: define [8 x <4 x i32>] @func_v8([8 x <4 x i32>] %x.coerce)
struct v8 func_v8(struct v8 x) { return x; } struct v8 func_v8(struct v8 x) { return x; }
// CHECK: define void @func_v9(%struct.v9* noalias sret %agg.result, %struct.v9* byval(%struct.v9) align 16 %x) // CHECK: define void @func_v9(%struct.v9* noalias sret align 16 %agg.result, %struct.v9* byval(%struct.v9) align 16 %x)
struct v9 func_v9(struct v9 x) { return x; } struct v9 func_v9(struct v9 x) { return x; }
// CHECK: define [2 x <4 x i32>] @func_vab([2 x <4 x i32>] %x.coerce) // CHECK: define [2 x <4 x i32>] @func_vab([2 x <4 x i32>] %x.coerce)
struct vab func_vab(struct vab x) { return x; } struct vab func_vab(struct vab x) { return x; }
// CHECK: define [3 x <4 x i32>] @func_vabc([3 x <4 x i32>] %x.coerce) // CHECK: define [3 x <4 x i32>] @func_vabc([3 x <4 x i32>] %x.coerce)
struct vabc func_vabc(struct vabc x) { return x; } struct vabc func_vabc(struct vabc x) { return x; }
▲ Show 20 Lines • Show All 41 Lines • ▼ Show 20 Lines
// CHECK-LABEL: @call_v8 // CHECK-LABEL: @call_v8
// CHECK: %[[TMP:[^ ]+]] = load [8 x <4 x i32>], [8 x <4 x i32>]* getelementptr inbounds (%struct.v8, %struct.v8* @global_v8, i32 0, i32 0), align 1 // CHECK: %[[TMP:[^ ]+]] = load [8 x <4 x i32>], [8 x <4 x i32>]* getelementptr inbounds (%struct.v8, %struct.v8* @global_v8, i32 0, i32 0), align 1
// CHECK: call [8 x <4 x i32>] @func_v8([8 x <4 x i32>] %[[TMP]]) // CHECK: call [8 x <4 x i32>] @func_v8([8 x <4 x i32>] %[[TMP]])
struct v8 global_v8; struct v8 global_v8;
void call_v8(void) { global_v8 = func_v8(global_v8); } void call_v8(void) { global_v8 = func_v8(global_v8); }
// CHECK-LABEL: @call_v9 // CHECK-LABEL: @call_v9
// CHECK: call void @func_v9(%struct.v9* sret %{{[^ ]+}}, %struct.v9* byval(%struct.v9) align 16 @global_v9) // CHECK: call void @func_v9(%struct.v9* sret align 16 %{{[^ ]+}}, %struct.v9* byval(%struct.v9) align 16 @global_v9)
struct v9 global_v9; struct v9 global_v9;
void call_v9(void) { global_v9 = func_v9(global_v9); } void call_v9(void) { global_v9 = func_v9(global_v9); }
// CHECK-LABEL: @call_vab // CHECK-LABEL: @call_vab
// CHECK: %[[TMP:[^ ]+]] = load [2 x <4 x i32>], [2 x <4 x i32>]* bitcast (%struct.vab* @global_vab to [2 x <4 x i32>]*) // CHECK: %[[TMP:[^ ]+]] = load [2 x <4 x i32>], [2 x <4 x i32>]* bitcast (%struct.vab* @global_vab to [2 x <4 x i32>]*)
// CHECK: call [2 x <4 x i32>] @func_vab([2 x <4 x i32>] %[[TMP]]) // CHECK: call [2 x <4 x i32>] @func_vab([2 x <4 x i32>] %[[TMP]])
struct vab global_vab; struct vab global_vab;
void call_vab(void) { global_vab = func_vab(global_vab); } void call_vab(void) { global_vab = func_vab(global_vab); }
▲ Show 20 Lines • Show All 42 Lines • ▼ Show 20 Lines
struct v3f6 func_v3f6(struct v3f6 x) { return x; } struct v3f6 func_v3f6(struct v3f6 x) { return x; }
// CHECK: define [7 x <4 x float>] @func_v3f7([7 x <4 x float>] %x.coerce) // CHECK: define [7 x <4 x float>] @func_v3f7([7 x <4 x float>] %x.coerce)
struct v3f7 func_v3f7(struct v3f7 x) { return x; } struct v3f7 func_v3f7(struct v3f7 x) { return x; }
// CHECK: define [8 x <4 x float>] @func_v3f8([8 x <4 x float>] %x.coerce) // CHECK: define [8 x <4 x float>] @func_v3f8([8 x <4 x float>] %x.coerce)
struct v3f8 func_v3f8(struct v3f8 x) { return x; } struct v3f8 func_v3f8(struct v3f8 x) { return x; }
// CHECK: define void @func_v3f9(%struct.v3f9* noalias sret %agg.result, %struct.v3f9* byval(%struct.v3f9) align 16 %x) // CHECK: define void @func_v3f9(%struct.v3f9* noalias sret align 16 %agg.result, %struct.v3f9* byval(%struct.v3f9) align 16 %x)
struct v3f9 func_v3f9(struct v3f9 x) { return x; } struct v3f9 func_v3f9(struct v3f9 x) { return x; }
// CHECK: define [2 x <4 x float>] @func_v3fab([2 x <4 x float>] %x.coerce) // CHECK: define [2 x <4 x float>] @func_v3fab([2 x <4 x float>] %x.coerce)
struct v3fab func_v3fab(struct v3fab x) { return x; } struct v3fab func_v3fab(struct v3fab x) { return x; }
// CHECK: define [3 x <4 x float>] @func_v3fabc([3 x <4 x float>] %x.coerce) // CHECK: define [3 x <4 x float>] @func_v3fabc([3 x <4 x float>] %x.coerce)
struct v3fabc func_v3fabc(struct v3fabc x) { return x; } struct v3fabc func_v3fabc(struct v3fabc x) { return x; }
▲ Show 20 Lines • Show All 41 Lines • ▼ Show 20 Lines
// CHECK-LABEL: @call_v3f8 // CHECK-LABEL: @call_v3f8
// CHECK: %[[TMP:[^ ]+]] = load [8 x <4 x float>], [8 x <4 x float>]* bitcast (%struct.v3f8* @global_v3f8 to [8 x <4 x float>]*), align 16 // CHECK: %[[TMP:[^ ]+]] = load [8 x <4 x float>], [8 x <4 x float>]* bitcast (%struct.v3f8* @global_v3f8 to [8 x <4 x float>]*), align 16
// CHECK: call [8 x <4 x float>] @func_v3f8([8 x <4 x float>] %[[TMP]]) // CHECK: call [8 x <4 x float>] @func_v3f8([8 x <4 x float>] %[[TMP]])
struct v3f8 global_v3f8; struct v3f8 global_v3f8;
void call_v3f8(void) { global_v3f8 = func_v3f8(global_v3f8); } void call_v3f8(void) { global_v3f8 = func_v3f8(global_v3f8); }
// CHECK-LABEL: @call_v3f9 // CHECK-LABEL: @call_v3f9
// CHECK: call void @func_v3f9(%struct.v3f9* sret %{{[^ ]+}}, %struct.v3f9* byval(%struct.v3f9) align 16 @global_v3f9) // CHECK: call void @func_v3f9(%struct.v3f9* sret align 16 %{{[^ ]+}}, %struct.v3f9* byval(%struct.v3f9) align 16 @global_v3f9)
struct v3f9 global_v3f9; struct v3f9 global_v3f9;
void call_v3f9(void) { global_v3f9 = func_v3f9(global_v3f9); } void call_v3f9(void) { global_v3f9 = func_v3f9(global_v3f9); }
// CHECK-LABEL: @call_v3fab // CHECK-LABEL: @call_v3fab
// CHECK: %[[TMP:[^ ]+]] = load [2 x <4 x float>], [2 x <4 x float>]* bitcast (%struct.v3fab* @global_v3fab to [2 x <4 x float>]*), align 16 // CHECK: %[[TMP:[^ ]+]] = load [2 x <4 x float>], [2 x <4 x float>]* bitcast (%struct.v3fab* @global_v3fab to [2 x <4 x float>]*), align 16
// CHECK: call [2 x <4 x float>] @func_v3fab([2 x <4 x float>] %[[TMP]]) // CHECK: call [2 x <4 x float>] @func_v3fab([2 x <4 x float>] %[[TMP]])
struct v3fab global_v3fab; struct v3fab global_v3fab;
void call_v3fab(void) { global_v3fab = func_v3fab(global_v3fab); } void call_v3fab(void) { global_v3fab = func_v3fab(global_v3fab); }
▲ Show 20 Lines • Show All 79 Lines • Show Last 20 Lines

clang/test/CodeGen/ppc64le-f128Aggregates.c

Show All 36 Lines
struct fp6 func_f6(struct fp6 x) { return x; } struct fp6 func_f6(struct fp6 x) { return x; }
// CHECK: define [7 x fp128] @func_f7([7 x fp128] %x.coerce) // CHECK: define [7 x fp128] @func_f7([7 x fp128] %x.coerce)
struct fp7 func_f7(struct fp7 x) { return x; } struct fp7 func_f7(struct fp7 x) { return x; }
// CHECK: define [8 x fp128] @func_f8([8 x fp128] %x.coerce) // CHECK: define [8 x fp128] @func_f8([8 x fp128] %x.coerce)
struct fp8 func_f8(struct fp8 x) { return x; } struct fp8 func_f8(struct fp8 x) { return x; }
// CHECK: define void @func_f9(%struct.fp9* noalias sret %agg.result, %struct.fp9* byval(%struct.fp9) align 16 %x) // CHECK: define void @func_f9(%struct.fp9* noalias sret align 16 %agg.result, %struct.fp9* byval(%struct.fp9) align 16 %x)
struct fp9 func_f9(struct fp9 x) { return x; } struct fp9 func_f9(struct fp9 x) { return x; }
// CHECK: define [2 x fp128] @func_fab([2 x fp128] %x.coerce) // CHECK: define [2 x fp128] @func_fab([2 x fp128] %x.coerce)
struct fpab func_fab(struct fpab x) { return x; } struct fpab func_fab(struct fpab x) { return x; }
// CHECK: define [3 x fp128] @func_fabc([3 x fp128] %x.coerce) // CHECK: define [3 x fp128] @func_fabc([3 x fp128] %x.coerce)
struct fpabc func_fabc(struct fpabc x) { return x; } struct fpabc func_fabc(struct fpabc x) { return x; }
▲ Show 20 Lines • Show All 45 Lines • ▼ Show 20 Lines
// CHECK-LABEL: @call_fp8 // CHECK-LABEL: @call_fp8
// CHECK: %[[TMP:[^ ]+]] = load [8 x fp128], [8 x fp128]* getelementptr inbounds (%struct.fp8, %struct.fp8* @global_f8, i32 0, i32 0), align 16 // CHECK: %[[TMP:[^ ]+]] = load [8 x fp128], [8 x fp128]* getelementptr inbounds (%struct.fp8, %struct.fp8* @global_f8, i32 0, i32 0), align 16
// CHECK: call [8 x fp128] @func_f8([8 x fp128] %[[TMP]]) // CHECK: call [8 x fp128] @func_f8([8 x fp128] %[[TMP]])
struct fp8 global_f8; struct fp8 global_f8;
void call_fp8(void) { global_f8 = func_f8(global_f8); } void call_fp8(void) { global_f8 = func_f8(global_f8); }
// CHECK-LABEL: @call_fp9 // CHECK-LABEL: @call_fp9
// CHECK: %[[TMP1:[^ ]+]] = alloca %struct.fp9, align 16 // CHECK: %[[TMP1:[^ ]+]] = alloca %struct.fp9, align 16
// CHECK: call void @func_f9(%struct.fp9* sret %[[TMP2:[^ ]+]], %struct.fp9* byval(%struct.fp9) align 16 @global_f9 // CHECK: call void @func_f9(%struct.fp9* sret align 16 %[[TMP2:[^ ]+]], %struct.fp9* byval(%struct.fp9) align 16 @global_f9
// CHECK: %[[TMP3:[^ ]+]] = bitcast %struct.fp9* %[[TMP2]] to i8* // CHECK: %[[TMP3:[^ ]+]] = bitcast %struct.fp9* %[[TMP2]] to i8*
// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 bitcast (%struct.fp9* @global_f9 to i8*), i8* align 16 %[[TMP3]], i64 144, i1 false // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 bitcast (%struct.fp9* @global_f9 to i8*), i8* align 16 %[[TMP3]], i64 144, i1 false
// CHECK: ret void // CHECK: ret void
struct fp9 global_f9; struct fp9 global_f9;
void call_fp9(void) { global_f9 = func_f9(global_f9); } void call_fp9(void) { global_f9 = func_f9(global_f9); }
// CHECK-LABEL: @call_fpab // CHECK-LABEL: @call_fpab
// CHECK: %[[TMP:[^ ]+]] = load [2 x fp128], [2 x fp128]* bitcast (%struct.fpab* @global_fab to [2 x fp128]*) // CHECK: %[[TMP:[^ ]+]] = load [2 x fp128], [2 x fp128]* bitcast (%struct.fpab* @global_fab to [2 x fp128]*)
Show All 9 Lines

clang/test/CodeGen/regparm-struct.c

Show First 20 Lines • Show All 153 Lines • ▼ Show 20 Lines
__attribute__((regparm(3))) void f16(struct s12 a, int b); __attribute__((regparm(3))) void f16(struct s12 a, int b);
// CHECK: declare void @f16(i32 inreg, i32 inreg, i32 inreg, i32) // CHECK: declare void @f16(i32 inreg, i32 inreg, i32 inreg, i32)
void g16(void) { void g16(void) {
struct s12 x = { 41, 42 }; struct s12 x = { 41, 42 };
f16(x, 43); f16(x, 43);
} }
__attribute__((regparm(3))) struct s12 f17(int a, int b, int c); __attribute__((regparm(3))) struct s12 f17(int a, int b, int c);
// CHECK: declare void @f17(%struct.s12* inreg sret, i32 inreg, i32 inreg, i32) // CHECK: declare void @f17(%struct.s12* inreg sret align 4, i32 inreg, i32 inreg, i32)
void g17(void) { void g17(void) {
f17(41, 42, 43); f17(41, 42, 43);
} }
struct s13 { struct s13 {
struct inner { struct inner {
float x; float x;
} y; } y;
}; };
__attribute__((regparm(3))) void f18(struct s13 a, int b, int c, int d); __attribute__((regparm(3))) void f18(struct s13 a, int b, int c, int d);
// CHECK: declare void @f18(%struct.s13* byval(%struct.s13) align 4, i32 inreg, i32 inreg, i32 inreg) // CHECK: declare void @f18(%struct.s13* byval(%struct.s13) align 4, i32 inreg, i32 inreg, i32 inreg)
void g18(void) { void g18(void) {
struct s13 x = {{41}}; struct s13 x = {{41}};
f18(x, 42, 43, 44); f18(x, 42, 43, 44);
} }

clang/test/CodeGen/renderscript.c

Show First 20 Lines • Show All 77 Lines • ▼ Show 20 Lines
// CHECK-RS64: void @argLongInt([2 x i64] %s.coerce) // CHECK-RS64: void @argLongInt([2 x i64] %s.coerce)
void argLongInt(sLongInt s) {} void argLongInt(sLongInt s) {}
// ============================================================================= // =============================================================================
// aggregate return value <= 16 bytes: returned on stack for 32-bit RenderScript // aggregate return value <= 16 bytes: returned on stack for 32-bit RenderScript
// and coerced to [a x iNN] for 64-bit RenderScript // and coerced to [a x iNN] for 64-bit RenderScript
// ============================================================================= // =============================================================================
// CHECK-RS32: void @retShortCharShort(%struct.sShortCharShort* noalias sret %agg.result) // CHECK-RS32: void @retShortCharShort(%struct.sShortCharShort* noalias sret align 2 %agg.result)
// CHECK-RS64: [3 x i16] @retShortCharShort() // CHECK-RS64: [3 x i16] @retShortCharShort()
sShortCharShort retShortCharShort() { sShortCharShort r; return r; } sShortCharShort retShortCharShort() { sShortCharShort r; return r; }
// CHECK-RS32: void @retIntShortChar(%struct.sIntShortChar* noalias sret %agg.result) // CHECK-RS32: void @retIntShortChar(%struct.sIntShortChar* noalias sret align 4 %agg.result)
// CHECK-RS64: [2 x i32] @retIntShortChar() // CHECK-RS64: [2 x i32] @retIntShortChar()
sIntShortChar retIntShortChar() { sIntShortChar r; return r; } sIntShortChar retIntShortChar() { sIntShortChar r; return r; }
// CHECK-RS32: void @retLongInt(%struct.sLongInt* noalias sret %agg.result) // CHECK-RS32: void @retLongInt(%struct.sLongInt* noalias sret align 8 %agg.result)
// CHECK-RS64: [2 x i64] @retLongInt() // CHECK-RS64: [2 x i64] @retLongInt()
sLongInt retLongInt() { sLongInt r; return r; } sLongInt retLongInt() { sLongInt r; return r; }
// ============================================================================= // =============================================================================
// aggregate parameter <= 64 bytes: coerced to [a x iNN] for 32-bit RenderScript // aggregate parameter <= 64 bytes: coerced to [a x iNN] for 32-bit RenderScript
// and passed on the stack for 64-bit RenderScript // and passed on the stack for 64-bit RenderScript
// ============================================================================= // =============================================================================
typedef struct {int i1, i2, i3, i4, i5; } sInt5; typedef struct {int i1, i2, i3, i4, i5; } sInt5;
typedef struct {long l1, l2; char c; } sLong2Char; typedef struct {long l1, l2; char c; } sLong2Char;
// CHECK-RS32: void @argInt5([5 x i32] %s.coerce) // CHECK-RS32: void @argInt5([5 x i32] %s.coerce)
// CHECK-RS64: void @argInt5(%struct.sInt5* %s) // CHECK-RS64: void @argInt5(%struct.sInt5* %s)
void argInt5(sInt5 s) {} void argInt5(sInt5 s) {}
// CHECK-RS32: void @argLong2Char([3 x i64] %s.coerce) // CHECK-RS32: void @argLong2Char([3 x i64] %s.coerce)
// CHECK-RS64: void @argLong2Char(%struct.sLong2Char* %s) // CHECK-RS64: void @argLong2Char(%struct.sLong2Char* %s)
void argLong2Char(sLong2Char s) {} void argLong2Char(sLong2Char s) {}
// ============================================================================= // =============================================================================
// aggregate return value <= 64 bytes: returned on stack for both 32-bit and // aggregate return value <= 64 bytes: returned on stack for both 32-bit and
// 64-bit RenderScript // 64-bit RenderScript
// ============================================================================= // =============================================================================
// CHECK-RS32: void @retInt5(%struct.sInt5* noalias sret %agg.result) // CHECK-RS32: void @retInt5(%struct.sInt5* noalias sret align 4 %agg.result)
// CHECK-RS64: void @retInt5(%struct.sInt5* noalias sret %agg.result) // CHECK-RS64: void @retInt5(%struct.sInt5* noalias sret align 4 %agg.result)
sInt5 retInt5() { sInt5 r; return r;} sInt5 retInt5() { sInt5 r; return r;}
// CHECK-RS32: void @retLong2Char(%struct.sLong2Char* noalias sret %agg.result) // CHECK-RS32: void @retLong2Char(%struct.sLong2Char* noalias sret align 8 %agg.result)
// CHECK-RS64: void @retLong2Char(%struct.sLong2Char* noalias sret %agg.result) // CHECK-RS64: void @retLong2Char(%struct.sLong2Char* noalias sret align 8 %agg.result)
sLong2Char retLong2Char() { sLong2Char r; return r;} sLong2Char retLong2Char() { sLong2Char r; return r;}
// ============================================================================= // =============================================================================
// aggregate parameters and return values > 64 bytes: passed and returned on the // aggregate parameters and return values > 64 bytes: passed and returned on the
// stack for both 32-bit and 64-bit RenderScript // stack for both 32-bit and 64-bit RenderScript
// ============================================================================= // =============================================================================
typedef struct {long l1, l2, l3, l4, l5, l6, l7, l8, l9; } sLong9; typedef struct {long l1, l2, l3, l4, l5, l6, l7, l8, l9; } sLong9;
// CHECK-RS32: void @argLong9(%struct.sLong9* byval(%struct.sLong9) align 8 %s) // CHECK-RS32: void @argLong9(%struct.sLong9* byval(%struct.sLong9) align 8 %s)
// CHECK-RS64: void @argLong9(%struct.sLong9* %s) // CHECK-RS64: void @argLong9(%struct.sLong9* %s)
void argLong9(sLong9 s) {} void argLong9(sLong9 s) {}
// CHECK-RS32: void @retLong9(%struct.sLong9* noalias sret %agg.result) // CHECK-RS32: void @retLong9(%struct.sLong9* noalias sret align 8 %agg.result)
// CHECK-RS64: void @retLong9(%struct.sLong9* noalias sret %agg.result) // CHECK-RS64: void @retLong9(%struct.sLong9* noalias sret align 8 %agg.result)
sLong9 retLong9() { sLong9 r; return r; } sLong9 retLong9() { sLong9 r; return r; }

clang/test/CodeGen/riscv32-ilp32-abi.c

Show All 29 Linesint f_scalar_stack_1(int32_t a, int64_t b, float c, double d, long double e,
uint8_t f, int8_t g, uint8_t h) { uint8_t f, int8_t g, uint8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret %agg.result, float %a, i64 %b, double %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret align 4 %agg.result, float %a, i64 %b, double %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_2(float a, int64_t b, double c, long double d, struct large f_scalar_stack_2(float a, int64_t b, double c, long double d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }
// Aggregates and >=XLen scalars passed on the stack should be lowered just as // Aggregates and >=XLen scalars passed on the stack should be lowered just as
// they would be if passed via registers. // they would be if passed via registers.
// CHECK-LABEL: define void @f_scalar_stack_3(double %a, i64 %b, double %c, i64 %d, i32 %e, i64 %f, float %g, double %h, fp128 %i) // CHECK-LABEL: define void @f_scalar_stack_3(double %a, i64 %b, double %c, i64 %d, i32 %e, i64 %f, float %g, double %h, fp128 %i)
void f_scalar_stack_3(double a, int64_t b, double c, int64_t d, int e, void f_scalar_stack_3(double a, int64_t b, double c, int64_t d, int e,
int64_t f, float g, double h, long double i) {} int64_t f, float g, double h, long double i) {}
// CHECK-LABEL: define void @f_agg_stack(double %a, i64 %b, double %c, i64 %d, i32 %e.coerce, [2 x i32] %f.coerce, i64 %g.coerce, %struct.large* %h) // CHECK-LABEL: define void @f_agg_stack(double %a, i64 %b, double %c, i64 %d, i32 %e.coerce, [2 x i32] %f.coerce, i64 %g.coerce, %struct.large* %h)
void f_agg_stack(double a, int64_t b, double c, int64_t d, struct tiny e, void f_agg_stack(double a, int64_t b, double c, int64_t d, struct tiny e,
struct small f, struct small_aligned g, struct large h) {} struct small f, struct small_aligned g, struct large h) {}

clang/test/CodeGen/riscv32-ilp32-ilp32f-abi.c

Show All 31 Linesint f_scalar_stack_1(int32_t a, int64_t b, int32_t c, double d, long double e,
uint8_t f, int8_t g, uint8_t h) { uint8_t f, int8_t g, uint8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret %agg.result, i32 %a, i64 %b, double %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret align 4 %agg.result, i32 %a, i64 %b, double %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_2(int32_t a, int64_t b, double c, long double d, struct large f_scalar_stack_2(int32_t a, int64_t b, double c, long double d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }
// Aggregates and >=XLen scalars passed on the stack should be lowered just as // Aggregates and >=XLen scalars passed on the stack should be lowered just as
// they would be if passed via registers. // they would be if passed via registers.
// CHECK-LABEL: define void @f_scalar_stack_3(double %a, i64 %b, double %c, i64 %d, i32 %e, i64 %f, i32 %g, double %h, fp128 %i) // CHECK-LABEL: define void @f_scalar_stack_3(double %a, i64 %b, double %c, i64 %d, i32 %e, i64 %f, i32 %g, double %h, fp128 %i)
void f_scalar_stack_3(double a, int64_t b, double c, int64_t d, int e, void f_scalar_stack_3(double a, int64_t b, double c, int64_t d, int e,
int64_t f, int32_t g, double h, long double i) {} int64_t f, int32_t g, double h, long double i) {}
// CHECK-LABEL: define void @f_agg_stack(double %a, i64 %b, double %c, i64 %d, i32 %e.coerce, [2 x i32] %f.coerce, i64 %g.coerce, %struct.large* %h) // CHECK-LABEL: define void @f_agg_stack(double %a, i64 %b, double %c, i64 %d, i32 %e.coerce, [2 x i32] %f.coerce, i64 %g.coerce, %struct.large* %h)
void f_agg_stack(double a, int64_t b, double c, int64_t d, struct tiny e, void f_agg_stack(double a, int64_t b, double c, int64_t d, struct tiny e,
struct small f, struct small_aligned g, struct large h) {} struct small f, struct small_aligned g, struct large h) {}

clang/test/CodeGen/riscv32-ilp32-ilp32f-ilp32d-abi.c

Show First 20 Lines • Show All 171 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define void @f_agg_large(%struct.large* %x) // CHECK-LABEL: define void @f_agg_large(%struct.large* %x)
void f_agg_large(struct large x) { void f_agg_large(struct large x) {
x.a = x.b + x.c + x.d; x.a = x.b + x.c + x.d;
} }
// The address where the struct should be written to will be the first // The address where the struct should be written to will be the first
// argument // argument
// CHECK-LABEL: define void @f_agg_large_ret(%struct.large* noalias sret %agg.result, i32 %i, i8 signext %j) // CHECK-LABEL: define void @f_agg_large_ret(%struct.large* noalias sret align 4 %agg.result, i32 %i, i8 signext %j)
struct large f_agg_large_ret(int32_t i, int8_t j) { struct large f_agg_large_ret(int32_t i, int8_t j) {
return (struct large){1, 2, 3, 4}; return (struct large){1, 2, 3, 4};
} }
typedef unsigned char v16i8 __attribute__((vector_size(16))); typedef unsigned char v16i8 __attribute__((vector_size(16)));
// CHECK-LABEL: define void @f_vec_large_v16i8(<16 x i8>* %0) // CHECK-LABEL: define void @f_vec_large_v16i8(<16 x i8>* %0)
void f_vec_large_v16i8(v16i8 x) { void f_vec_large_v16i8(v16i8 x) {
x[0] = x[7]; x[0] = x[7];
} }
// CHECK-LABEL: define void @f_vec_large_v16i8_ret(<16 x i8>* noalias sret %agg.result) // CHECK-LABEL: define void @f_vec_large_v16i8_ret(<16 x i8>* noalias sret align 16 %agg.result)
v16i8 f_vec_large_v16i8_ret() { v16i8 f_vec_large_v16i8_ret() {
return (v16i8){1, 2, 3, 4, 5, 6, 7, 8}; return (v16i8){1, 2, 3, 4, 5, 6, 7, 8};
} }
// Scalars passed on the stack should not have signext/zeroext attributes // Scalars passed on the stack should not have signext/zeroext attributes
// (they are anyext). // (they are anyext).
// CHECK-LABEL: define i32 @f_scalar_stack_1(i32 %a.coerce, [2 x i32] %b.coerce, i64 %c.coerce, %struct.large* %d, i8 zeroext %e, i8 signext %f, i8 %g, i8 %h) // CHECK-LABEL: define i32 @f_scalar_stack_1(i32 %a.coerce, [2 x i32] %b.coerce, i64 %c.coerce, %struct.large* %d, i8 zeroext %e, i8 signext %f, i8 %g, i8 %h)
int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c, int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c,
struct large d, uint8_t e, int8_t f, uint8_t g, int8_t h) { struct large d, uint8_t e, int8_t f, uint8_t g, int8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret %agg.result, i32 %a, i64 %b, i64 %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret align 4 %agg.result, i32 %a, i64 %b, i64 %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_2(int32_t a, int64_t b, int64_t c, long double d, struct large f_scalar_stack_2(int32_t a, int64_t b, int64_t c, long double d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }
// CHECK-LABEL: define fp128 @f_scalar_stack_4(i32 %a, i64 %b, i64 %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define fp128 @f_scalar_stack_4(i32 %a, i64 %b, i64 %c, fp128 %d, i8 zeroext %e, i8 %f, i8 %g)
long double f_scalar_stack_4(int32_t a, int64_t b, int64_t c, long double d, long double f_scalar_stack_4(int32_t a, int64_t b, int64_t c, long double d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
▲ Show 20 Lines • Show All 213 Lines • Show Last 20 Lines

clang/test/CodeGen/riscv32-ilp32d-abi.c

Show First 20 Lines • Show All 113 Lines • ▼ Show 20 Lines
// CHECK: define { double, i32 } @f_ret_double_int32_s() // CHECK: define { double, i32 } @f_ret_double_int32_s()
struct double_int32_s f_ret_double_int32_s() { struct double_int32_s f_ret_double_int32_s() {
return (struct double_int32_s){1.0, 2}; return (struct double_int32_s){1.0, 2};
} }
// CHECK: define void @f_double_int64_s_arg(%struct.double_int64_s* %a) // CHECK: define void @f_double_int64_s_arg(%struct.double_int64_s* %a)
void f_double_int64_s_arg(struct double_int64_s a) {} void f_double_int64_s_arg(struct double_int64_s a) {}
// CHECK: define void @f_ret_double_int64_s(%struct.double_int64_s* noalias sret %agg.result) // CHECK: define void @f_ret_double_int64_s(%struct.double_int64_s* noalias sret align 8 %agg.result)
struct double_int64_s f_ret_double_int64_s() { struct double_int64_s f_ret_double_int64_s() {
return (struct double_int64_s){1.0, 2}; return (struct double_int64_s){1.0, 2};
} }
// CHECK: define void @f_double_int64bf_s_arg(double %0, i32 %1) // CHECK: define void @f_double_int64bf_s_arg(double %0, i32 %1)
void f_double_int64bf_s_arg(struct double_int64bf_s a) {} void f_double_int64bf_s_arg(struct double_int64bf_s a) {}
// CHECK: define { double, i32 } @f_ret_double_int64bf_s() // CHECK: define { double, i32 } @f_ret_double_int64bf_s()
▲ Show 20 Lines • Show All 107 Lines • ▼ Show 20 Lines
// Test structs that should be passed according to the normal integer calling // Test structs that should be passed according to the normal integer calling
// convention. // convention.
struct int_double_int_s { int a; double b; int c; }; struct int_double_int_s { int a; double b; int c; };
// CHECK: define void @f_int_double_int_s_arg(%struct.int_double_int_s* %a) // CHECK: define void @f_int_double_int_s_arg(%struct.int_double_int_s* %a)
void f_int_double_int_s_arg(struct int_double_int_s a) {} void f_int_double_int_s_arg(struct int_double_int_s a) {}
// CHECK: define void @f_ret_int_double_int_s(%struct.int_double_int_s* noalias sret %agg.result) // CHECK: define void @f_ret_int_double_int_s(%struct.int_double_int_s* noalias sret align 8 %agg.result)
struct int_double_int_s f_ret_int_double_int_s() { struct int_double_int_s f_ret_int_double_int_s() {
return (struct int_double_int_s){1, 2.0, 3}; return (struct int_double_int_s){1, 2.0, 3};
} }
struct int64_double_s { int64_t a; double b; }; struct int64_double_s { int64_t a; double b; };
// CHECK: define void @f_int64_double_s_arg(%struct.int64_double_s* %a) // CHECK: define void @f_int64_double_s_arg(%struct.int64_double_s* %a)
void f_int64_double_s_arg(struct int64_double_s a) {} void f_int64_double_s_arg(struct int64_double_s a) {}
// CHECK: define void @f_ret_int64_double_s(%struct.int64_double_s* noalias sret %agg.result) // CHECK: define void @f_ret_int64_double_s(%struct.int64_double_s* noalias sret align 8 %agg.result)
struct int64_double_s f_ret_int64_double_s() { struct int64_double_s f_ret_int64_double_s() {
return (struct int64_double_s){1, 2.0}; return (struct int64_double_s){1, 2.0};
} }
struct char_char_double_s { char a; char b; double c; }; struct char_char_double_s { char a; char b; double c; };
// CHECK-LABEL: define void @f_char_char_double_s_arg(%struct.char_char_double_s* %a) // CHECK-LABEL: define void @f_char_char_double_s_arg(%struct.char_char_double_s* %a)
void f_char_char_double_s_arg(struct char_char_double_s a) {} void f_char_char_double_s_arg(struct char_char_double_s a) {}
// CHECK: define void @f_ret_char_char_double_s(%struct.char_char_double_s* noalias sret %agg.result) // CHECK: define void @f_ret_char_char_double_s(%struct.char_char_double_s* noalias sret align 8 %agg.result)
struct char_char_double_s f_ret_char_char_double_s() { struct char_char_double_s f_ret_char_char_double_s() {
return (struct char_char_double_s){1, 2, 3.0}; return (struct char_char_double_s){1, 2, 3.0};
} }
// Unions are always passed according to the integer calling convention, even // Unions are always passed according to the integer calling convention, even
// if they can only contain a double. // if they can only contain a double.
union double_u { double a; }; union double_u { double a; };
Show All 32 Lines

clang/test/CodeGen/riscv32-ilp32f-abi.c

Show All 20 Linesstruct double_s f_ret_double_s() {
return (struct double_s){1.0}; return (struct double_s){1.0};
} }
struct double_double_s { double d; double e; }; struct double_double_s { double d; double e; };
// CHECK: define void @f_double_double_s_arg(%struct.double_double_s* %a) // CHECK: define void @f_double_double_s_arg(%struct.double_double_s* %a)
void f_double_double_s_arg(struct double_double_s a) {} void f_double_double_s_arg(struct double_double_s a) {}
// CHECK: define void @f_ret_double_double_s(%struct.double_double_s* noalias sret %agg.result) // CHECK: define void @f_ret_double_double_s(%struct.double_double_s* noalias sret align 8 %agg.result)
struct double_double_s f_ret_double_double_s() { struct double_double_s f_ret_double_double_s() {
return (struct double_double_s){1.0, 2.0}; return (struct double_double_s){1.0, 2.0};
} }
struct double_int8_s { double d; int64_t i; }; struct double_int8_s { double d; int64_t i; };
struct int_double_s { int a; double b; }; struct int_double_s { int a; double b; };
// CHECK: define void @f_int_double_s_arg(%struct.int_double_s* %a) // CHECK: define void @f_int_double_s_arg(%struct.int_double_s* %a)
void f_int_double_s_arg(struct int_double_s a) {} void f_int_double_s_arg(struct int_double_s a) {}
// CHECK: define void @f_ret_int_double_s(%struct.int_double_s* noalias sret %agg.result) // CHECK: define void @f_ret_int_double_s(%struct.int_double_s* noalias sret align 8 %agg.result)
struct int_double_s f_ret_int_double_s() { struct int_double_s f_ret_int_double_s() {
return (struct int_double_s){1, 2.0}; return (struct int_double_s){1, 2.0};
} }

clang/test/CodeGen/riscv32-ilp32f-ilp32d-abi.c

Show First 20 Lines • Show All 106 Lines • ▼ Show 20 Lines
// CHECK: define { float, i32 } @f_ret_float_int32_s() // CHECK: define { float, i32 } @f_ret_float_int32_s()
struct float_int32_s f_ret_float_int32_s() { struct float_int32_s f_ret_float_int32_s() {
return (struct float_int32_s){1.0, 2}; return (struct float_int32_s){1.0, 2};
} }
// CHECK: define void @f_float_int64_s_arg(%struct.float_int64_s* %a) // CHECK: define void @f_float_int64_s_arg(%struct.float_int64_s* %a)
void f_float_int64_s_arg(struct float_int64_s a) {} void f_float_int64_s_arg(struct float_int64_s a) {}
// CHECK: define void @f_ret_float_int64_s(%struct.float_int64_s* noalias sret %agg.result) // CHECK: define void @f_ret_float_int64_s(%struct.float_int64_s* noalias sret align 8 %agg.result)
struct float_int64_s f_ret_float_int64_s() { struct float_int64_s f_ret_float_int64_s() {
return (struct float_int64_s){1.0, 2}; return (struct float_int64_s){1.0, 2};
} }
// CHECK: define void @f_float_int64bf_s_arg(float %0, i32 %1) // CHECK: define void @f_float_int64bf_s_arg(float %0, i32 %1)
void f_float_int64bf_s_arg(struct float_int64bf_s a) {} void f_float_int64bf_s_arg(struct float_int64bf_s a) {}
// CHECK: define { float, i32 } @f_ret_float_int64bf_s() // CHECK: define { float, i32 } @f_ret_float_int64bf_s()
▲ Show 20 Lines • Show All 107 Lines • ▼ Show 20 Lines
// Test structs that should be passed according to the normal integer calling // Test structs that should be passed according to the normal integer calling
// convention. // convention.
struct int_float_int_s { int a; float b; int c; }; struct int_float_int_s { int a; float b; int c; };
// CHECK: define void @f_int_float_int_s_arg(%struct.int_float_int_s* %a) // CHECK: define void @f_int_float_int_s_arg(%struct.int_float_int_s* %a)
void f_int_float_int_s_arg(struct int_float_int_s a) {} void f_int_float_int_s_arg(struct int_float_int_s a) {}
// CHECK: define void @f_ret_int_float_int_s(%struct.int_float_int_s* noalias sret %agg.result) // CHECK: define void @f_ret_int_float_int_s(%struct.int_float_int_s* noalias sret align 4 %agg.result)
struct int_float_int_s f_ret_int_float_int_s() { struct int_float_int_s f_ret_int_float_int_s() {
return (struct int_float_int_s){1, 2.0, 3}; return (struct int_float_int_s){1, 2.0, 3};
} }
struct int64_float_s { int64_t a; float b; }; struct int64_float_s { int64_t a; float b; };
// CHECK: define void @f_int64_float_s_arg(%struct.int64_float_s* %a) // CHECK: define void @f_int64_float_s_arg(%struct.int64_float_s* %a)
void f_int64_float_s_arg(struct int64_float_s a) {} void f_int64_float_s_arg(struct int64_float_s a) {}
// CHECK: define void @f_ret_int64_float_s(%struct.int64_float_s* noalias sret %agg.result) // CHECK: define void @f_ret_int64_float_s(%struct.int64_float_s* noalias sret align 8 %agg.result)
struct int64_float_s f_ret_int64_float_s() { struct int64_float_s f_ret_int64_float_s() {
return (struct int64_float_s){1, 2.0}; return (struct int64_float_s){1, 2.0};
} }
struct char_char_float_s { char a; char b; float c; }; struct char_char_float_s { char a; char b; float c; };
// CHECK-LABEL: define void @f_char_char_float_s_arg([2 x i32] %a.coerce) // CHECK-LABEL: define void @f_char_char_float_s_arg([2 x i32] %a.coerce)
void f_char_char_float_s_arg(struct char_char_float_s a) {} void f_char_char_float_s_arg(struct char_char_float_s a) {}
Show All 18 Lines

clang/test/CodeGen/riscv64-lp64-abi.c

Show All 19 Linesint f_scalar_stack_1(int32_t a, __int128_t b, float c, long double d, v32i8 e,
uint8_t f, int8_t g, uint8_t h) { uint8_t f, int8_t g, uint8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret %agg.result, double %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret align 8 %agg.result, double %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_2(double a, __int128_t b, long double c, v32i8 d, struct large f_scalar_stack_2(double a, __int128_t b, long double c, v32i8 d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }

clang/test/CodeGen/riscv64-lp64-lp64f-abi.c

Show All 21 Linesint f_scalar_stack_1(int32_t a, __int128_t b, double c, long double d, v32i8 e,
uint8_t f, int8_t g, uint8_t h) { uint8_t f, int8_t g, uint8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret %agg.result, double %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_2(%struct.large* noalias sret align 8 %agg.result, double %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_2(double a, __int128_t b, long double c, v32i8 d, struct large f_scalar_stack_2(double a, __int128_t b, long double c, v32i8 d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }

clang/test/CodeGen/riscv64-lp64-lp64f-lp64d-abi.c

Show First 20 Lines • Show All 160 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define void @f_agg_large(%struct.large* %x) // CHECK-LABEL: define void @f_agg_large(%struct.large* %x)
void f_agg_large(struct large x) { void f_agg_large(struct large x) {
x.a = x.b + x.c + x.d; x.a = x.b + x.c + x.d;
} }
// The address where the struct should be written to will be the first // The address where the struct should be written to will be the first
// argument // argument
// CHECK-LABEL: define void @f_agg_large_ret(%struct.large* noalias sret %agg.result, i32 signext %i, i8 signext %j) // CHECK-LABEL: define void @f_agg_large_ret(%struct.large* noalias sret align 8 %agg.result, i32 signext %i, i8 signext %j)
struct large f_agg_large_ret(int32_t i, int8_t j) { struct large f_agg_large_ret(int32_t i, int8_t j) {
return (struct large){1, 2, 3, 4}; return (struct large){1, 2, 3, 4};
} }
typedef unsigned char v32i8 __attribute__((vector_size(32))); typedef unsigned char v32i8 __attribute__((vector_size(32)));
// CHECK-LABEL: define void @f_vec_large_v32i8(<32 x i8>* %0) // CHECK-LABEL: define void @f_vec_large_v32i8(<32 x i8>* %0)
void f_vec_large_v32i8(v32i8 x) { void f_vec_large_v32i8(v32i8 x) {
x[0] = x[7]; x[0] = x[7];
} }
// CHECK-LABEL: define void @f_vec_large_v32i8_ret(<32 x i8>* noalias sret %agg.result) // CHECK-LABEL: define void @f_vec_large_v32i8_ret(<32 x i8>* noalias sret align 32 %agg.result)
v32i8 f_vec_large_v32i8_ret() { v32i8 f_vec_large_v32i8_ret() {
return (v32i8){1, 2, 3, 4, 5, 6, 7, 8}; return (v32i8){1, 2, 3, 4, 5, 6, 7, 8};
} }
// Scalars passed on the stack should not have signext/zeroext attributes // Scalars passed on the stack should not have signext/zeroext attributes
// (they are anyext). // (they are anyext).
// CHECK-LABEL: define signext i32 @f_scalar_stack_1(i64 %a.coerce, [2 x i64] %b.coerce, i128 %c.coerce, %struct.large* %d, i8 zeroext %e, i8 signext %f, i8 %g, i8 %h) // CHECK-LABEL: define signext i32 @f_scalar_stack_1(i64 %a.coerce, [2 x i64] %b.coerce, i128 %c.coerce, %struct.large* %d, i8 zeroext %e, i8 signext %f, i8 %g, i8 %h)
int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c, int f_scalar_stack_1(struct tiny a, struct small b, struct small_aligned c,
struct large d, uint8_t e, int8_t f, uint8_t g, int8_t h) { struct large d, uint8_t e, int8_t f, uint8_t g, int8_t h) {
return g + h; return g + h;
} }
// CHECK-LABEL: define signext i32 @f_scalar_stack_2(i32 signext %a, i128 %b, i64 %c, fp128 %d, <32 x i8>* %0, i8 zeroext %f, i8 %g, i8 %h) // CHECK-LABEL: define signext i32 @f_scalar_stack_2(i32 signext %a, i128 %b, i64 %c, fp128 %d, <32 x i8>* %0, i8 zeroext %f, i8 %g, i8 %h)
int f_scalar_stack_2(int32_t a, __int128_t b, int64_t c, long double d, v32i8 e, int f_scalar_stack_2(int32_t a, __int128_t b, int64_t c, long double d, v32i8 e,
uint8_t f, int8_t g, uint8_t h) { uint8_t f, int8_t g, uint8_t h) {
return g + h; return g + h;
} }
// Ensure that scalars passed on the stack are still determined correctly in // Ensure that scalars passed on the stack are still determined correctly in
// the presence of large return values that consume a register due to the need // the presence of large return values that consume a register due to the need
// to pass a pointer. // to pass a pointer.
// CHECK-LABEL: define void @f_scalar_stack_3(%struct.large* noalias sret %agg.result, i32 signext %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g) // CHECK-LABEL: define void @f_scalar_stack_3(%struct.large* noalias sret align 8 %agg.result, i32 signext %a, i128 %b, fp128 %c, <32 x i8>* %0, i8 zeroext %e, i8 %f, i8 %g)
struct large f_scalar_stack_3(uint32_t a, __int128_t b, long double c, v32i8 d, struct large f_scalar_stack_3(uint32_t a, __int128_t b, long double c, v32i8 d,
uint8_t e, int8_t f, uint8_t g) { uint8_t e, int8_t f, uint8_t g) {
return (struct large){a, e, f, g}; return (struct large){a, e, f, g};
} }
// Ensure that ABI lowering happens as expected for vararg calls. // Ensure that ABI lowering happens as expected for vararg calls.
// Specifically, ensure that signext is emitted for varargs that will be // Specifically, ensure that signext is emitted for varargs that will be
// passed in registers but not on the stack. Ensure this takes into account // passed in registers but not on the stack. Ensure this takes into account
▲ Show 20 Lines • Show All 216 Lines • Show Last 20 Lines

clang/test/CodeGen/riscv64-lp64d-abi.c

Show First 20 Lines • Show All 237 Lines • ▼ Show 20 Lines
// Test structs that should be passed according to the normal integer calling // Test structs that should be passed according to the normal integer calling
// convention. // convention.
struct int_double_int_s { int a; double b; int c; }; struct int_double_int_s { int a; double b; int c; };
// CHECK: define void @f_int_double_int_s_arg(%struct.int_double_int_s* %a) // CHECK: define void @f_int_double_int_s_arg(%struct.int_double_int_s* %a)
void f_int_double_int_s_arg(struct int_double_int_s a) {} void f_int_double_int_s_arg(struct int_double_int_s a) {}
// CHECK: define void @f_ret_int_double_int_s(%struct.int_double_int_s* noalias sret %agg.result) // CHECK: define void @f_ret_int_double_int_s(%struct.int_double_int_s* noalias sret align 8 %agg.result)
struct int_double_int_s f_ret_int_double_int_s() { struct int_double_int_s f_ret_int_double_int_s() {
return (struct int_double_int_s){1, 2.0, 3}; return (struct int_double_int_s){1, 2.0, 3};
} }
struct char_char_double_s { char a; char b; double c; }; struct char_char_double_s { char a; char b; double c; };
// CHECK-LABEL: define void @f_char_char_double_s_arg([2 x i64] %a.coerce) // CHECK-LABEL: define void @f_char_char_double_s_arg([2 x i64] %a.coerce)
void f_char_char_double_s_arg(struct char_char_double_s a) {} void f_char_char_double_s_arg(struct char_char_double_s a) {}
Show All 18 Lines

clang/test/CodeGen/sparcv9-abi.c

Show First 20 Lines • Show All 47 Lines • ▼ Show 20 Lines
// Large structs are also returned indirectly. // Large structs are also returned indirectly.
struct large { struct large {
int *a, *b; int *a, *b;
int *c, *d; int *c, *d;
int x; int x;
}; };
// CHECK-LABEL: define void @f_large(%struct.large* noalias sret %agg.result, %struct.large* %x) // CHECK-LABEL: define void @f_large(%struct.large* noalias sret align 8 %agg.result, %struct.large* %x)
struct large f_large(struct large x) { struct large f_large(struct large x) {
x.a += *x.b; x.a += *x.b;
x.b = 0; x.b = 0;
return x; return x;
} }
// A 64-bit struct fits in a register. // A 64-bit struct fits in a register.
struct reg { struct reg {
▲ Show 20 Lines • Show All 120 Lines • Show Last 20 Lines

clang/test/CodeGen/struct-passing.c

Show All 12 Lines
T1 __attribute__((pure)) f3(void); T1 __attribute__((pure)) f3(void);
void __attribute__((const)) f4(T1 a); void __attribute__((const)) f4(T1 a);
void __attribute__((pure)) f5(T1 a); void __attribute__((pure)) f5(T1 a);
void *ps[] = { f0, f1, f2, f3, f4, f5 }; void *ps[] = { f0, f1, f2, f3, f4, f5 };
// CHECK: declare i32 @f0() [[RN:#[0-9]+]] // CHECK: declare i32 @f0() [[RN:#[0-9]+]]
// CHECK: declare i32 @f1() [[RO:#[0-9]+]] // CHECK: declare i32 @f1() [[RO:#[0-9]+]]
// CHECK: declare void @f2({{.*}} sret) // CHECK: declare void @f2({{.*}} sret align 4)
// CHECK: declare void @f3({{.*}} sret) // CHECK: declare void @f3({{.*}} sret align 4)
// CHECK: declare void @f4({{.*}} byval({{.*}}) align 4) // CHECK: declare void @f4({{.*}} byval({{.*}}) align 4)
// CHECK: declare void @f5({{.*}} byval({{.*}}) align 4) // CHECK: declare void @f5({{.*}} byval({{.*}}) align 4)
// CHECK: attributes [[RN]] = { nounwind readnone{{.*}} } // CHECK: attributes [[RN]] = { nounwind readnone{{.*}} }
// CHECK: attributes [[RO]] = { nounwind readonly{{.*}} } // CHECK: attributes [[RO]] = { nounwind readonly{{.*}} }

clang/test/CodeGen/systemz-abi-vector.c

Show First 20 Lines • Show All 44 Lines • ▼ Show 20 Lines
typedef __attribute__((vector_size(32))) char v32i8; typedef __attribute__((vector_size(32))) char v32i8;
unsigned int align = __alignof__ (v16i8); unsigned int align = __alignof__ (v16i8);
// CHECK: @align = global i32 16 // CHECK: @align = global i32 16
// CHECK-VECTOR: @align = global i32 8 // CHECK-VECTOR: @align = global i32 8
v1i8 pass_v1i8(v1i8 arg) { return arg; } v1i8 pass_v1i8(v1i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1i8(<1 x i8>* noalias sret %{{.*}}, <1 x i8>* %0) // CHECK-LABEL: define void @pass_v1i8(<1 x i8>* noalias sret align 1 %{{.*}}, <1 x i8>* %0)
// CHECK-VECTOR-LABEL: define <1 x i8> @pass_v1i8(<1 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x i8> @pass_v1i8(<1 x i8> %{{.*}})
v2i8 pass_v2i8(v2i8 arg) { return arg; } v2i8 pass_v2i8(v2i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2i8(<2 x i8>* noalias sret %{{.*}}, <2 x i8>* %0) // CHECK-LABEL: define void @pass_v2i8(<2 x i8>* noalias sret align 2 %{{.*}}, <2 x i8>* %0)
// CHECK-VECTOR-LABEL: define <2 x i8> @pass_v2i8(<2 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x i8> @pass_v2i8(<2 x i8> %{{.*}})
v4i8 pass_v4i8(v4i8 arg) { return arg; } v4i8 pass_v4i8(v4i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v4i8(<4 x i8>* noalias sret %{{.*}}, <4 x i8>* %0) // CHECK-LABEL: define void @pass_v4i8(<4 x i8>* noalias sret align 4 %{{.*}}, <4 x i8>* %0)
// CHECK-VECTOR-LABEL: define <4 x i8> @pass_v4i8(<4 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define <4 x i8> @pass_v4i8(<4 x i8> %{{.*}})
v8i8 pass_v8i8(v8i8 arg) { return arg; } v8i8 pass_v8i8(v8i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v8i8(<8 x i8>* noalias sret %{{.*}}, <8 x i8>* %0) // CHECK-LABEL: define void @pass_v8i8(<8 x i8>* noalias sret align 8 %{{.*}}, <8 x i8>* %0)
// CHECK-VECTOR-LABEL: define <8 x i8> @pass_v8i8(<8 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define <8 x i8> @pass_v8i8(<8 x i8> %{{.*}})
v16i8 pass_v16i8(v16i8 arg) { return arg; } v16i8 pass_v16i8(v16i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v16i8(<16 x i8>* noalias sret %{{.*}}, <16 x i8>* %0) // CHECK-LABEL: define void @pass_v16i8(<16 x i8>* noalias sret align 16 %{{.*}}, <16 x i8>* %0)
// CHECK-VECTOR-LABEL: define <16 x i8> @pass_v16i8(<16 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define <16 x i8> @pass_v16i8(<16 x i8> %{{.*}})
v32i8 pass_v32i8(v32i8 arg) { return arg; } v32i8 pass_v32i8(v32i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_v32i8(<32 x i8>* noalias sret %{{.*}}, <32 x i8>* %0) // CHECK-LABEL: define void @pass_v32i8(<32 x i8>* noalias sret align 32 %{{.*}}, <32 x i8>* %0)
// CHECK-VECTOR-LABEL: define void @pass_v32i8(<32 x i8>* noalias sret %{{.*}}, <32 x i8>* %0) // CHECK-VECTOR-LABEL: define void @pass_v32i8(<32 x i8>* noalias sret align 8 %{{.*}}, <32 x i8>* %0)
v1i16 pass_v1i16(v1i16 arg) { return arg; } v1i16 pass_v1i16(v1i16 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1i16(<1 x i16>* noalias sret %{{.*}}, <1 x i16>* %0) // CHECK-LABEL: define void @pass_v1i16(<1 x i16>* noalias sret align 2 %{{.*}}, <1 x i16>* %0)
// CHECK-VECTOR-LABEL: define <1 x i16> @pass_v1i16(<1 x i16> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x i16> @pass_v1i16(<1 x i16> %{{.*}})
v2i16 pass_v2i16(v2i16 arg) { return arg; } v2i16 pass_v2i16(v2i16 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2i16(<2 x i16>* noalias sret %{{.*}}, <2 x i16>* %0) // CHECK-LABEL: define void @pass_v2i16(<2 x i16>* noalias sret align 4 %{{.*}}, <2 x i16>* %0)
// CHECK-VECTOR-LABEL: define <2 x i16> @pass_v2i16(<2 x i16> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x i16> @pass_v2i16(<2 x i16> %{{.*}})
v4i16 pass_v4i16(v4i16 arg) { return arg; } v4i16 pass_v4i16(v4i16 arg) { return arg; }
// CHECK-LABEL: define void @pass_v4i16(<4 x i16>* noalias sret %{{.*}}, <4 x i16>* %0) // CHECK-LABEL: define void @pass_v4i16(<4 x i16>* noalias sret align 8 %{{.*}}, <4 x i16>* %0)
// CHECK-VECTOR-LABEL: define <4 x i16> @pass_v4i16(<4 x i16> %{{.*}}) // CHECK-VECTOR-LABEL: define <4 x i16> @pass_v4i16(<4 x i16> %{{.*}})
v8i16 pass_v8i16(v8i16 arg) { return arg; } v8i16 pass_v8i16(v8i16 arg) { return arg; }
// CHECK-LABEL: define void @pass_v8i16(<8 x i16>* noalias sret %{{.*}}, <8 x i16>* %0) // CHECK-LABEL: define void @pass_v8i16(<8 x i16>* noalias sret align 16 %{{.*}}, <8 x i16>* %0)
// CHECK-VECTOR-LABEL: define <8 x i16> @pass_v8i16(<8 x i16> %{{.*}}) // CHECK-VECTOR-LABEL: define <8 x i16> @pass_v8i16(<8 x i16> %{{.*}})
v1i32 pass_v1i32(v1i32 arg) { return arg; } v1i32 pass_v1i32(v1i32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1i32(<1 x i32>* noalias sret %{{.*}}, <1 x i32>* %0) // CHECK-LABEL: define void @pass_v1i32(<1 x i32>* noalias sret align 4 %{{.*}}, <1 x i32>* %0)
// CHECK-VECTOR-LABEL: define <1 x i32> @pass_v1i32(<1 x i32> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x i32> @pass_v1i32(<1 x i32> %{{.*}})
v2i32 pass_v2i32(v2i32 arg) { return arg; } v2i32 pass_v2i32(v2i32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2i32(<2 x i32>* noalias sret %{{.*}}, <2 x i32>* %0) // CHECK-LABEL: define void @pass_v2i32(<2 x i32>* noalias sret align 8 %{{.*}}, <2 x i32>* %0)
// CHECK-VECTOR-LABEL: define <2 x i32> @pass_v2i32(<2 x i32> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x i32> @pass_v2i32(<2 x i32> %{{.*}})
v4i32 pass_v4i32(v4i32 arg) { return arg; } v4i32 pass_v4i32(v4i32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v4i32(<4 x i32>* noalias sret %{{.*}}, <4 x i32>* %0) // CHECK-LABEL: define void @pass_v4i32(<4 x i32>* noalias sret align 16 %{{.*}}, <4 x i32>* %0)
// CHECK-VECTOR-LABEL: define <4 x i32> @pass_v4i32(<4 x i32> %{{.*}}) // CHECK-VECTOR-LABEL: define <4 x i32> @pass_v4i32(<4 x i32> %{{.*}})
v1i64 pass_v1i64(v1i64 arg) { return arg; } v1i64 pass_v1i64(v1i64 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1i64(<1 x i64>* noalias sret %{{.*}}, <1 x i64>* %0) // CHECK-LABEL: define void @pass_v1i64(<1 x i64>* noalias sret align 8 %{{.*}}, <1 x i64>* %0)
// CHECK-VECTOR-LABEL: define <1 x i64> @pass_v1i64(<1 x i64> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x i64> @pass_v1i64(<1 x i64> %{{.*}})
v2i64 pass_v2i64(v2i64 arg) { return arg; } v2i64 pass_v2i64(v2i64 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2i64(<2 x i64>* noalias sret %{{.*}}, <2 x i64>* %0) // CHECK-LABEL: define void @pass_v2i64(<2 x i64>* noalias sret align 16 %{{.*}}, <2 x i64>* %0)
// CHECK-VECTOR-LABEL: define <2 x i64> @pass_v2i64(<2 x i64> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x i64> @pass_v2i64(<2 x i64> %{{.*}})
v1i128 pass_v1i128(v1i128 arg) { return arg; } v1i128 pass_v1i128(v1i128 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1i128(<1 x i128>* noalias sret %{{.*}}, <1 x i128>* %0) // CHECK-LABEL: define void @pass_v1i128(<1 x i128>* noalias sret align 16 %{{.*}}, <1 x i128>* %0)
// CHECK-VECTOR-LABEL: define <1 x i128> @pass_v1i128(<1 x i128> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x i128> @pass_v1i128(<1 x i128> %{{.*}})
v1f32 pass_v1f32(v1f32 arg) { return arg; } v1f32 pass_v1f32(v1f32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1f32(<1 x float>* noalias sret %{{.*}}, <1 x float>* %0) // CHECK-LABEL: define void @pass_v1f32(<1 x float>* noalias sret align 4 %{{.*}}, <1 x float>* %0)
// CHECK-VECTOR-LABEL: define <1 x float> @pass_v1f32(<1 x float> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x float> @pass_v1f32(<1 x float> %{{.*}})
v2f32 pass_v2f32(v2f32 arg) { return arg; } v2f32 pass_v2f32(v2f32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2f32(<2 x float>* noalias sret %{{.*}}, <2 x float>* %0) // CHECK-LABEL: define void @pass_v2f32(<2 x float>* noalias sret align 8 %{{.*}}, <2 x float>* %0)
// CHECK-VECTOR-LABEL: define <2 x float> @pass_v2f32(<2 x float> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x float> @pass_v2f32(<2 x float> %{{.*}})
v4f32 pass_v4f32(v4f32 arg) { return arg; } v4f32 pass_v4f32(v4f32 arg) { return arg; }
// CHECK-LABEL: define void @pass_v4f32(<4 x float>* noalias sret %{{.*}}, <4 x float>* %0) // CHECK-LABEL: define void @pass_v4f32(<4 x float>* noalias sret align 16 %{{.*}}, <4 x float>* %0)
// CHECK-VECTOR-LABEL: define <4 x float> @pass_v4f32(<4 x float> %{{.*}}) // CHECK-VECTOR-LABEL: define <4 x float> @pass_v4f32(<4 x float> %{{.*}})
v1f64 pass_v1f64(v1f64 arg) { return arg; } v1f64 pass_v1f64(v1f64 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1f64(<1 x double>* noalias sret %{{.*}}, <1 x double>* %0) // CHECK-LABEL: define void @pass_v1f64(<1 x double>* noalias sret align 8 %{{.*}}, <1 x double>* %0)
// CHECK-VECTOR-LABEL: define <1 x double> @pass_v1f64(<1 x double> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x double> @pass_v1f64(<1 x double> %{{.*}})
v2f64 pass_v2f64(v2f64 arg) { return arg; } v2f64 pass_v2f64(v2f64 arg) { return arg; }
// CHECK-LABEL: define void @pass_v2f64(<2 x double>* noalias sret %{{.*}}, <2 x double>* %0) // CHECK-LABEL: define void @pass_v2f64(<2 x double>* noalias sret align 16 %{{.*}}, <2 x double>* %0)
// CHECK-VECTOR-LABEL: define <2 x double> @pass_v2f64(<2 x double> %{{.*}}) // CHECK-VECTOR-LABEL: define <2 x double> @pass_v2f64(<2 x double> %{{.*}})
v1f128 pass_v1f128(v1f128 arg) { return arg; } v1f128 pass_v1f128(v1f128 arg) { return arg; }
// CHECK-LABEL: define void @pass_v1f128(<1 x fp128>* noalias sret %{{.*}}, <1 x fp128>* %0) // CHECK-LABEL: define void @pass_v1f128(<1 x fp128>* noalias sret align 16 %{{.*}}, <1 x fp128>* %0)
// CHECK-VECTOR-LABEL: define <1 x fp128> @pass_v1f128(<1 x fp128> %{{.*}}) // CHECK-VECTOR-LABEL: define <1 x fp128> @pass_v1f128(<1 x fp128> %{{.*}})
// Vector-like aggregate types // Vector-like aggregate types
struct agg_v1i8 { v1i8 a; }; struct agg_v1i8 { v1i8 a; };
struct agg_v1i8 pass_agg_v1i8(struct agg_v1i8 arg) { return arg; } struct agg_v1i8 pass_agg_v1i8(struct agg_v1i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v1i8(%struct.agg_v1i8* noalias sret %{{.*}}, i8 %{{.*}}) // CHECK-LABEL: define void @pass_agg_v1i8(%struct.agg_v1i8* noalias sret align 1 %{{.*}}, i8 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v1i8(%struct.agg_v1i8* noalias sret %{{.*}}, <1 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v1i8(%struct.agg_v1i8* noalias sret align 1 %{{.*}}, <1 x i8> %{{.*}})
struct agg_v2i8 { v2i8 a; }; struct agg_v2i8 { v2i8 a; };
struct agg_v2i8 pass_agg_v2i8(struct agg_v2i8 arg) { return arg; } struct agg_v2i8 pass_agg_v2i8(struct agg_v2i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v2i8(%struct.agg_v2i8* noalias sret %{{.*}}, i16 %{{.*}}) // CHECK-LABEL: define void @pass_agg_v2i8(%struct.agg_v2i8* noalias sret align 2 %{{.*}}, i16 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v2i8(%struct.agg_v2i8* noalias sret %{{.*}}, <2 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v2i8(%struct.agg_v2i8* noalias sret align 2 %{{.*}}, <2 x i8> %{{.*}})
struct agg_v4i8 { v4i8 a; }; struct agg_v4i8 { v4i8 a; };
struct agg_v4i8 pass_agg_v4i8(struct agg_v4i8 arg) { return arg; } struct agg_v4i8 pass_agg_v4i8(struct agg_v4i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v4i8(%struct.agg_v4i8* noalias sret %{{.*}}, i32 %{{.*}}) // CHECK-LABEL: define void @pass_agg_v4i8(%struct.agg_v4i8* noalias sret align 4 %{{.*}}, i32 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v4i8(%struct.agg_v4i8* noalias sret %{{.*}}, <4 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v4i8(%struct.agg_v4i8* noalias sret align 4 %{{.*}}, <4 x i8> %{{.*}})
struct agg_v8i8 { v8i8 a; }; struct agg_v8i8 { v8i8 a; };
struct agg_v8i8 pass_agg_v8i8(struct agg_v8i8 arg) { return arg; } struct agg_v8i8 pass_agg_v8i8(struct agg_v8i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v8i8(%struct.agg_v8i8* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_v8i8(%struct.agg_v8i8* noalias sret align 8 %{{.*}}, i64 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v8i8(%struct.agg_v8i8* noalias sret %{{.*}}, <8 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v8i8(%struct.agg_v8i8* noalias sret align 8 %{{.*}}, <8 x i8> %{{.*}})
struct agg_v16i8 { v16i8 a; }; struct agg_v16i8 { v16i8 a; };
struct agg_v16i8 pass_agg_v16i8(struct agg_v16i8 arg) { return arg; } struct agg_v16i8 pass_agg_v16i8(struct agg_v16i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v16i8(%struct.agg_v16i8* noalias sret %{{.*}}, %struct.agg_v16i8* %{{.*}}) // CHECK-LABEL: define void @pass_agg_v16i8(%struct.agg_v16i8* noalias sret align 16 %{{.*}}, %struct.agg_v16i8* %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v16i8(%struct.agg_v16i8* noalias sret %{{.*}}, <16 x i8> %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v16i8(%struct.agg_v16i8* noalias sret align 8 %{{.*}}, <16 x i8> %{{.*}})
struct agg_v32i8 { v32i8 a; }; struct agg_v32i8 { v32i8 a; };
struct agg_v32i8 pass_agg_v32i8(struct agg_v32i8 arg) { return arg; } struct agg_v32i8 pass_agg_v32i8(struct agg_v32i8 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_v32i8(%struct.agg_v32i8* noalias sret %{{.*}}, %struct.agg_v32i8* %{{.*}}) // CHECK-LABEL: define void @pass_agg_v32i8(%struct.agg_v32i8* noalias sret align 32 %{{.*}}, %struct.agg_v32i8* %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_v32i8(%struct.agg_v32i8* noalias sret %{{.*}}, %struct.agg_v32i8* %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_v32i8(%struct.agg_v32i8* noalias sret align 8 %{{.*}}, %struct.agg_v32i8* %{{.*}})
// Verify that the following are *not* vector-like aggregate types // Verify that the following are *not* vector-like aggregate types
struct agg_novector1 { v4i8 a; v4i8 b; }; struct agg_novector1 { v4i8 a; v4i8 b; };
struct agg_novector1 pass_agg_novector1(struct agg_novector1 arg) { return arg; } struct agg_novector1 pass_agg_novector1(struct agg_novector1 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_novector1(%struct.agg_novector1* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_novector1(%struct.agg_novector1* noalias sret align 4 %{{.*}}, i64 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_novector1(%struct.agg_novector1* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_novector1(%struct.agg_novector1* noalias sret align 4 %{{.*}}, i64 %{{.*}})
struct agg_novector2 { v4i8 a; float b; }; struct agg_novector2 { v4i8 a; float b; };
struct agg_novector2 pass_agg_novector2(struct agg_novector2 arg) { return arg; } struct agg_novector2 pass_agg_novector2(struct agg_novector2 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_novector2(%struct.agg_novector2* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_novector2(%struct.agg_novector2* noalias sret align 4 %{{.*}}, i64 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_novector2(%struct.agg_novector2* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_novector2(%struct.agg_novector2* noalias sret align 4 %{{.*}}, i64 %{{.*}})
struct agg_novector3 { v4i8 a; int : 0; }; struct agg_novector3 { v4i8 a; int : 0; };
struct agg_novector3 pass_agg_novector3(struct agg_novector3 arg) { return arg; } struct agg_novector3 pass_agg_novector3(struct agg_novector3 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_novector3(%struct.agg_novector3* noalias sret %{{.*}}, i32 %{{.*}}) // CHECK-LABEL: define void @pass_agg_novector3(%struct.agg_novector3* noalias sret align 4 %{{.*}}, i32 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_novector3(%struct.agg_novector3* noalias sret %{{.*}}, i32 %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_novector3(%struct.agg_novector3* noalias sret align 4 %{{.*}}, i32 %{{.*}})
struct agg_novector4 { v4i8 a __attribute__((aligned (8))); }; struct agg_novector4 { v4i8 a __attribute__((aligned (8))); };
struct agg_novector4 pass_agg_novector4(struct agg_novector4 arg) { return arg; } struct agg_novector4 pass_agg_novector4(struct agg_novector4 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_novector4(%struct.agg_novector4* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_novector4(%struct.agg_novector4* noalias sret align 8 %{{.*}}, i64 %{{.*}})
// CHECK-VECTOR-LABEL: define void @pass_agg_novector4(%struct.agg_novector4* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-VECTOR-LABEL: define void @pass_agg_novector4(%struct.agg_novector4* noalias sret align 8 %{{.*}}, i64 %{{.*}})
// Accessing variable argument lists // Accessing variable argument lists
v1i8 va_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, v1i8); } v1i8 va_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, v1i8); }
// CHECK-LABEL: define void @va_v1i8(<1 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v1i8(<1 x i8>* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 15 Lines
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <1 x i8>* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <1 x i8>*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RET:%[^ ]+]] = load <1 x i8>, <1 x i8>* [[MEM_ADDR]] // CHECK-VECTOR: [[RET:%[^ ]+]] = load <1 x i8>, <1 x i8>* [[MEM_ADDR]]
// CHECK-VECTOR: ret <1 x i8> [[RET]] // CHECK-VECTOR: ret <1 x i8> [[RET]]
v2i8 va_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, v2i8); } v2i8 va_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, v2i8); }
// CHECK-LABEL: define void @va_v2i8(<2 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v2i8(<2 x i8>* noalias sret align 2 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 15 Lines
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <2 x i8>* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <2 x i8>*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RET:%[^ ]+]] = load <2 x i8>, <2 x i8>* [[MEM_ADDR]] // CHECK-VECTOR: [[RET:%[^ ]+]] = load <2 x i8>, <2 x i8>* [[MEM_ADDR]]
// CHECK-VECTOR: ret <2 x i8> [[RET]] // CHECK-VECTOR: ret <2 x i8> [[RET]]
v4i8 va_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, v4i8); } v4i8 va_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, v4i8); }
// CHECK-LABEL: define void @va_v4i8(<4 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v4i8(<4 x i8>* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 15 Lines
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <4 x i8>* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <4 x i8>*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RET:%[^ ]+]] = load <4 x i8>, <4 x i8>* [[MEM_ADDR]] // CHECK-VECTOR: [[RET:%[^ ]+]] = load <4 x i8>, <4 x i8>* [[MEM_ADDR]]
// CHECK-VECTOR: ret <4 x i8> [[RET]] // CHECK-VECTOR: ret <4 x i8> [[RET]]
v8i8 va_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, v8i8); } v8i8 va_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, v8i8); }
// CHECK-LABEL: define void @va_v8i8(<8 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v8i8(<8 x i8>* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 15 Lines
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <8 x i8>* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <8 x i8>*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RET:%[^ ]+]] = load <8 x i8>, <8 x i8>* [[MEM_ADDR]] // CHECK-VECTOR: [[RET:%[^ ]+]] = load <8 x i8>, <8 x i8>* [[MEM_ADDR]]
// CHECK-VECTOR: ret <8 x i8> [[RET]] // CHECK-VECTOR: ret <8 x i8> [[RET]]
v16i8 va_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, v16i8); } v16i8 va_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, v16i8); }
// CHECK-LABEL: define void @va_v16i8(<16 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v16i8(<16 x i8>* noalias sret align 16 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 15 Lines
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <16 x i8>* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to <16 x i8>*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 16 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 16
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RET:%[^ ]+]] = load <16 x i8>, <16 x i8>* [[MEM_ADDR]] // CHECK-VECTOR: [[RET:%[^ ]+]] = load <16 x i8>, <16 x i8>* [[MEM_ADDR]]
// CHECK-VECTOR: ret <16 x i8> [[RET]] // CHECK-VECTOR: ret <16 x i8> [[RET]]
v32i8 va_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, v32i8); } v32i8 va_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, v32i8); }
// CHECK-LABEL: define void @va_v32i8(<32 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_v32i8(<32 x i8>* noalias sret align 32 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to <32 x i8>** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to <32 x i8>**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to <32 x i8>** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to <32 x i8>**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi <32 x i8>** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi <32 x i8>** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load <32 x i8>*, <32 x i8>** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load <32 x i8>*, <32 x i8>** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_v32i8(<32 x i8>* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_v32i8(<32 x i8>* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK-VECTOR: br i1 [[FITS_IN_REGS]], // CHECK-VECTOR: br i1 [[FITS_IN_REGS]],
// CHECK-VECTOR: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK-VECTOR: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK-VECTOR: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK-VECTOR: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK-VECTOR: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK-VECTOR: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK-VECTOR: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK-VECTOR: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK-VECTOR: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK-VECTOR: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK-VECTOR: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to <32 x i8>** // CHECK-VECTOR: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to <32 x i8>**
// CHECK-VECTOR: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK-VECTOR: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK-VECTOR: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK-VECTOR: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK-VECTOR: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to <32 x i8>** // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to <32 x i8>**
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[VA_ARG_ADDR:%[^ ]+]] = phi <32 x i8>** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK-VECTOR: [[VA_ARG_ADDR:%[^ ]+]] = phi <32 x i8>** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK-VECTOR: [[INDIRECT_ARG:%[^ ]+]] = load <32 x i8>*, <32 x i8>** [[VA_ARG_ADDR]] // CHECK-VECTOR: [[INDIRECT_ARG:%[^ ]+]] = load <32 x i8>*, <32 x i8>** [[VA_ARG_ADDR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v1i8 va_agg_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v1i8); } struct agg_v1i8 va_agg_v1i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v1i8); }
// CHECK-LABEL: define void @va_agg_v1i8(%struct.agg_v1i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v1i8(%struct.agg_v1i8* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 23 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 23
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v1i8* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v1i8*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 7 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 7
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v1i8* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v1i8*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v1i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v1i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v1i8(%struct.agg_v1i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v1i8(%struct.agg_v1i8* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v1i8* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v1i8*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v2i8 va_agg_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v2i8); } struct agg_v2i8 va_agg_v2i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v2i8); }
// CHECK-LABEL: define void @va_agg_v2i8(%struct.agg_v2i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v2i8(%struct.agg_v2i8* noalias sret align 2 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 22 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 22
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v2i8* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v2i8*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 6 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 6
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v2i8* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v2i8*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v2i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v2i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v2i8(%struct.agg_v2i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v2i8(%struct.agg_v2i8* noalias sret align 2 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v2i8* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v2i8*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v4i8 va_agg_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v4i8); } struct agg_v4i8 va_agg_v4i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v4i8); }
// CHECK-LABEL: define void @va_agg_v4i8(%struct.agg_v4i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v4i8(%struct.agg_v4i8* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v4i8* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v4i8*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v4i8* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v4i8*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v4i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v4i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v4i8(%struct.agg_v4i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v4i8(%struct.agg_v4i8* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v4i8* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v4i8*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v8i8 va_agg_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v8i8); } struct agg_v8i8 va_agg_v8i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v8i8); }
// CHECK-LABEL: define void @va_agg_v8i8(%struct.agg_v8i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v8i8(%struct.agg_v8i8* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v8i8* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v8i8*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v8i8* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v8i8*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v8i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v8i8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v8i8(%struct.agg_v8i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v8i8(%struct.agg_v8i8* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v8i8* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v8i8*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v16i8 va_agg_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v16i8); } struct agg_v16i8 va_agg_v16i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v16i8); }
// CHECK-LABEL: define void @va_agg_v16i8(%struct.agg_v16i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v16i8(%struct.agg_v16i8* noalias sret align 16 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v16i8** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v16i8**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v16i8** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v16i8**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v16i8** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v16i8** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_v16i8*, %struct.agg_v16i8** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_v16i8*, %struct.agg_v16i8** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v16i8(%struct.agg_v16i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v16i8(%struct.agg_v16i8* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v16i8* // CHECK-VECTOR: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[OVERFLOW_ARG_AREA]] to %struct.agg_v16i8*
// CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 16 // CHECK-VECTOR: [[OVERFLOW_ARG_AREA1:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 16
// CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK-VECTOR: store i8* [[OVERFLOW_ARG_AREA1]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK-VECTOR: ret void // CHECK-VECTOR: ret void
struct agg_v32i8 va_agg_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v32i8); } struct agg_v32i8 va_agg_v32i8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_v32i8); }
// CHECK-LABEL: define void @va_agg_v32i8(%struct.agg_v32i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_agg_v32i8(%struct.agg_v32i8* noalias sret align 32 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v32i8** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_v32i8**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v32i8** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_v32i8**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v32i8** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_v32i8** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_v32i8*, %struct.agg_v32i8** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_v32i8*, %struct.agg_v32i8** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
// CHECK-VECTOR-LABEL: define void @va_agg_v32i8(%struct.agg_v32i8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-VECTOR-LABEL: define void @va_agg_v32i8(%struct.agg_v32i8* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK-VECTOR: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK-VECTOR: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK-VECTOR: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK-VECTOR: br i1 [[FITS_IN_REGS]], // CHECK-VECTOR: br i1 [[FITS_IN_REGS]],
// CHECK-VECTOR: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK-VECTOR: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK-VECTOR: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK-VECTOR: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK-VECTOR: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK-VECTOR: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK-VECTOR: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK-VECTOR: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 13 Lines

clang/test/CodeGen/systemz-abi.c

Show All 30 Lines
long pass_long(long arg) { return arg; } long pass_long(long arg) { return arg; }
// CHECK-LABEL: define i64 @pass_long(i64 %{{.*}}) // CHECK-LABEL: define i64 @pass_long(i64 %{{.*}})
long long pass_longlong(long long arg) { return arg; } long long pass_longlong(long long arg) { return arg; }
// CHECK-LABEL: define i64 @pass_longlong(i64 %{{.*}}) // CHECK-LABEL: define i64 @pass_longlong(i64 %{{.*}})
__int128 pass_int128(__int128 arg) { return arg; } __int128 pass_int128(__int128 arg) { return arg; }
// CHECK-LABEL: define void @pass_int128(i128* noalias sret %{{.*}}, i128* %0) // CHECK-LABEL: define void @pass_int128(i128* noalias sret align 16 %{{.*}}, i128* %0)
float pass_float(float arg) { return arg; } float pass_float(float arg) { return arg; }
// CHECK-LABEL: define float @pass_float(float %{{.*}}) // CHECK-LABEL: define float @pass_float(float %{{.*}})
double pass_double(double arg) { return arg; } double pass_double(double arg) { return arg; }
// CHECK-LABEL: define double @pass_double(double %{{.*}}) // CHECK-LABEL: define double @pass_double(double %{{.*}})
long double pass_longdouble(long double arg) { return arg; } long double pass_longdouble(long double arg) { return arg; }
// CHECK-LABEL: define void @pass_longdouble(fp128* noalias sret %{{.*}}, fp128* %0) // CHECK-LABEL: define void @pass_longdouble(fp128* noalias sret align 8 %{{.*}}, fp128* %0)
// Complex types // Complex types
_Complex char pass_complex_char(_Complex char arg) { return arg; } _Complex char pass_complex_char(_Complex char arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_char({ i8, i8 }* noalias sret %{{.*}}, { i8, i8 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_char({ i8, i8 }* noalias sret align 1 %{{.*}}, { i8, i8 }* %{{.*}}arg)
_Complex short pass_complex_short(_Complex short arg) { return arg; } _Complex short pass_complex_short(_Complex short arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_short({ i16, i16 }* noalias sret %{{.*}}, { i16, i16 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_short({ i16, i16 }* noalias sret align 2 %{{.*}}, { i16, i16 }* %{{.*}}arg)
_Complex int pass_complex_int(_Complex int arg) { return arg; } _Complex int pass_complex_int(_Complex int arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_int({ i32, i32 }* noalias sret %{{.*}}, { i32, i32 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_int({ i32, i32 }* noalias sret align 4 %{{.*}}, { i32, i32 }* %{{.*}}arg)
_Complex long pass_complex_long(_Complex long arg) { return arg; } _Complex long pass_complex_long(_Complex long arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_long({ i64, i64 }* noalias sret %{{.*}}, { i64, i64 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_long({ i64, i64 }* noalias sret align 8 %{{.*}}, { i64, i64 }* %{{.*}}arg)
_Complex long long pass_complex_longlong(_Complex long long arg) { return arg; } _Complex long long pass_complex_longlong(_Complex long long arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_longlong({ i64, i64 }* noalias sret %{{.*}}, { i64, i64 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_longlong({ i64, i64 }* noalias sret align 8 %{{.*}}, { i64, i64 }* %{{.*}}arg)
_Complex float pass_complex_float(_Complex float arg) { return arg; } _Complex float pass_complex_float(_Complex float arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_float({ float, float }* noalias sret %{{.*}}, { float, float }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_float({ float, float }* noalias sret align 4 %{{.*}}, { float, float }* %{{.*}}arg)
_Complex double pass_complex_double(_Complex double arg) { return arg; } _Complex double pass_complex_double(_Complex double arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_double({ double, double }* noalias sret %{{.*}}, { double, double }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_double({ double, double }* noalias sret align 8 %{{.*}}, { double, double }* %{{.*}}arg)
_Complex long double pass_complex_longdouble(_Complex long double arg) { return arg; } _Complex long double pass_complex_longdouble(_Complex long double arg) { return arg; }
// CHECK-LABEL: define void @pass_complex_longdouble({ fp128, fp128 }* noalias sret %{{.*}}, { fp128, fp128 }* %{{.*}}arg) // CHECK-LABEL: define void @pass_complex_longdouble({ fp128, fp128 }* noalias sret align 8 %{{.*}}, { fp128, fp128 }* %{{.*}}arg)
// Aggregate types // Aggregate types
struct agg_1byte { char a[1]; }; struct agg_1byte { char a[1]; };
struct agg_1byte pass_agg_1byte(struct agg_1byte arg) { return arg; } struct agg_1byte pass_agg_1byte(struct agg_1byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_1byte(%struct.agg_1byte* noalias sret %{{.*}}, i8 %{{.*}}) // CHECK-LABEL: define void @pass_agg_1byte(%struct.agg_1byte* noalias sret align 1 %{{.*}}, i8 %{{.*}})
struct agg_2byte { char a[2]; }; struct agg_2byte { char a[2]; };
struct agg_2byte pass_agg_2byte(struct agg_2byte arg) { return arg; } struct agg_2byte pass_agg_2byte(struct agg_2byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_2byte(%struct.agg_2byte* noalias sret %{{.*}}, i16 %{{.*}}) // CHECK-LABEL: define void @pass_agg_2byte(%struct.agg_2byte* noalias sret align 1 %{{.*}}, i16 %{{.*}})
struct agg_3byte { char a[3]; }; struct agg_3byte { char a[3]; };
struct agg_3byte pass_agg_3byte(struct agg_3byte arg) { return arg; } struct agg_3byte pass_agg_3byte(struct agg_3byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_3byte(%struct.agg_3byte* noalias sret %{{.*}}, %struct.agg_3byte* %{{.*}}) // CHECK-LABEL: define void @pass_agg_3byte(%struct.agg_3byte* noalias sret align 1 %{{.*}}, %struct.agg_3byte* %{{.*}})
struct agg_4byte { char a[4]; }; struct agg_4byte { char a[4]; };
struct agg_4byte pass_agg_4byte(struct agg_4byte arg) { return arg; } struct agg_4byte pass_agg_4byte(struct agg_4byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_4byte(%struct.agg_4byte* noalias sret %{{.*}}, i32 %{{.*}}) // CHECK-LABEL: define void @pass_agg_4byte(%struct.agg_4byte* noalias sret align 1 %{{.*}}, i32 %{{.*}})
struct agg_5byte { char a[5]; }; struct agg_5byte { char a[5]; };
struct agg_5byte pass_agg_5byte(struct agg_5byte arg) { return arg; } struct agg_5byte pass_agg_5byte(struct agg_5byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_5byte(%struct.agg_5byte* noalias sret %{{.*}}, %struct.agg_5byte* %{{.*}}) // CHECK-LABEL: define void @pass_agg_5byte(%struct.agg_5byte* noalias sret align 1 %{{.*}}, %struct.agg_5byte* %{{.*}})
struct agg_6byte { char a[6]; }; struct agg_6byte { char a[6]; };
struct agg_6byte pass_agg_6byte(struct agg_6byte arg) { return arg; } struct agg_6byte pass_agg_6byte(struct agg_6byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_6byte(%struct.agg_6byte* noalias sret %{{.*}}, %struct.agg_6byte* %{{.*}}) // CHECK-LABEL: define void @pass_agg_6byte(%struct.agg_6byte* noalias sret align 1 %{{.*}}, %struct.agg_6byte* %{{.*}})
struct agg_7byte { char a[7]; }; struct agg_7byte { char a[7]; };
struct agg_7byte pass_agg_7byte(struct agg_7byte arg) { return arg; } struct agg_7byte pass_agg_7byte(struct agg_7byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_7byte(%struct.agg_7byte* noalias sret %{{.*}}, %struct.agg_7byte* %{{.*}}) // CHECK-LABEL: define void @pass_agg_7byte(%struct.agg_7byte* noalias sret align 1 %{{.*}}, %struct.agg_7byte* %{{.*}})
struct agg_8byte { char a[8]; }; struct agg_8byte { char a[8]; };
struct agg_8byte pass_agg_8byte(struct agg_8byte arg) { return arg; } struct agg_8byte pass_agg_8byte(struct agg_8byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_8byte(%struct.agg_8byte* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_8byte(%struct.agg_8byte* noalias sret align 1 %{{.*}}, i64 %{{.*}})
struct agg_16byte { char a[16]; }; struct agg_16byte { char a[16]; };
struct agg_16byte pass_agg_16byte(struct agg_16byte arg) { return arg; } struct agg_16byte pass_agg_16byte(struct agg_16byte arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_16byte(%struct.agg_16byte* noalias sret %{{.*}}, %struct.agg_16byte* %{{.*}}) // CHECK-LABEL: define void @pass_agg_16byte(%struct.agg_16byte* noalias sret align 1 %{{.*}}, %struct.agg_16byte* %{{.*}})
// Float-like aggregate types // Float-like aggregate types
struct agg_float { float a; }; struct agg_float { float a; };
struct agg_float pass_agg_float(struct agg_float arg) { return arg; } struct agg_float pass_agg_float(struct agg_float arg) { return arg; }
// HARD-FLOAT-LABEL: define void @pass_agg_float(%struct.agg_float* noalias sret %{{.*}}, float %{{.*}}) // HARD-FLOAT-LABEL: define void @pass_agg_float(%struct.agg_float* noalias sret align 4 %{{.*}}, float %{{.*}})
// SOFT-FLOAT-LABEL: define void @pass_agg_float(%struct.agg_float* noalias sret %{{.*}}, i32 %{{.*}}) // SOFT-FLOAT-LABEL: define void @pass_agg_float(%struct.agg_float* noalias sret align 4 %{{.*}}, i32 %{{.*}})
struct agg_double { double a; }; struct agg_double { double a; };
struct agg_double pass_agg_double(struct agg_double arg) { return arg; } struct agg_double pass_agg_double(struct agg_double arg) { return arg; }
// HARD-FLOAT-LABEL: define void @pass_agg_double(%struct.agg_double* noalias sret %{{.*}}, double %{{.*}}) // HARD-FLOAT-LABEL: define void @pass_agg_double(%struct.agg_double* noalias sret align 8 %{{.*}}, double %{{.*}})
// SOFT-FLOAT-LABEL: define void @pass_agg_double(%struct.agg_double* noalias sret %{{.*}}, i64 %{{.*}}) // SOFT-FLOAT-LABEL: define void @pass_agg_double(%struct.agg_double* noalias sret align 8 %{{.*}}, i64 %{{.*}})
struct agg_longdouble { long double a; }; struct agg_longdouble { long double a; };
struct agg_longdouble pass_agg_longdouble(struct agg_longdouble arg) { return arg; } struct agg_longdouble pass_agg_longdouble(struct agg_longdouble arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_longdouble(%struct.agg_longdouble* noalias sret %{{.*}}, %struct.agg_longdouble* %{{.*}}) // CHECK-LABEL: define void @pass_agg_longdouble(%struct.agg_longdouble* noalias sret align 8 %{{.*}}, %struct.agg_longdouble* %{{.*}})
struct agg_float_a8 { float a __attribute__((aligned (8))); }; struct agg_float_a8 { float a __attribute__((aligned (8))); };
struct agg_float_a8 pass_agg_float_a8(struct agg_float_a8 arg) { return arg; } struct agg_float_a8 pass_agg_float_a8(struct agg_float_a8 arg) { return arg; }
// HARD-FLOAT-LABEL: define void @pass_agg_float_a8(%struct.agg_float_a8* noalias sret %{{.*}}, double %{{.*}}) // HARD-FLOAT-LABEL: define void @pass_agg_float_a8(%struct.agg_float_a8* noalias sret align 8 %{{.*}}, double %{{.*}})
// SOFT-FLOAT-LABEL: define void @pass_agg_float_a8(%struct.agg_float_a8* noalias sret %{{.*}}, i64 %{{.*}}) // SOFT-FLOAT-LABEL: define void @pass_agg_float_a8(%struct.agg_float_a8* noalias sret align 8 %{{.*}}, i64 %{{.*}})
struct agg_float_a16 { float a __attribute__((aligned (16))); }; struct agg_float_a16 { float a __attribute__((aligned (16))); };
struct agg_float_a16 pass_agg_float_a16(struct agg_float_a16 arg) { return arg; } struct agg_float_a16 pass_agg_float_a16(struct agg_float_a16 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_float_a16(%struct.agg_float_a16* noalias sret %{{.*}}, %struct.agg_float_a16* %{{.*}}) // CHECK-LABEL: define void @pass_agg_float_a16(%struct.agg_float_a16* noalias sret align 16 %{{.*}}, %struct.agg_float_a16* %{{.*}})
// Verify that the following are *not* float-like aggregate types // Verify that the following are *not* float-like aggregate types
struct agg_nofloat1 { float a; float b; }; struct agg_nofloat1 { float a; float b; };
struct agg_nofloat1 pass_agg_nofloat1(struct agg_nofloat1 arg) { return arg; } struct agg_nofloat1 pass_agg_nofloat1(struct agg_nofloat1 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_nofloat1(%struct.agg_nofloat1* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_nofloat1(%struct.agg_nofloat1* noalias sret align 4 %{{.*}}, i64 %{{.*}})
struct agg_nofloat2 { float a; int b; }; struct agg_nofloat2 { float a; int b; };
struct agg_nofloat2 pass_agg_nofloat2(struct agg_nofloat2 arg) { return arg; } struct agg_nofloat2 pass_agg_nofloat2(struct agg_nofloat2 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_nofloat2(%struct.agg_nofloat2* noalias sret %{{.*}}, i64 %{{.*}}) // CHECK-LABEL: define void @pass_agg_nofloat2(%struct.agg_nofloat2* noalias sret align 4 %{{.*}}, i64 %{{.*}})
struct agg_nofloat3 { float a; int : 0; }; struct agg_nofloat3 { float a; int : 0; };
struct agg_nofloat3 pass_agg_nofloat3(struct agg_nofloat3 arg) { return arg; } struct agg_nofloat3 pass_agg_nofloat3(struct agg_nofloat3 arg) { return arg; }
// CHECK-LABEL: define void @pass_agg_nofloat3(%struct.agg_nofloat3* noalias sret %{{.*}}, i32 %{{.*}}) // CHECK-LABEL: define void @pass_agg_nofloat3(%struct.agg_nofloat3* noalias sret align 4 %{{.*}}, i32 %{{.*}})
// Accessing variable argument lists // Accessing variable argument lists
int va_int(__builtin_va_list l) { return __builtin_va_arg(l, int); } int va_int(__builtin_va_list l) { return __builtin_va_arg(l, int); }
// CHECK-LABEL: define signext i32 @va_int(%struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define signext i32 @va_int(%struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
▲ Show 20 Lines • Show All 88 Lines • ▼ Show 20 Lines
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to double* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to double*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi double* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi double* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[RET:%[^ ]+]] = load double, double* [[VA_ARG_ADDR]] // CHECK: [[RET:%[^ ]+]] = load double, double* [[VA_ARG_ADDR]]
// CHECK: ret double [[RET]] // CHECK: ret double [[RET]]
long double va_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, long double); } long double va_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, long double); }
// CHECK-LABEL: define void @va_longdouble(fp128* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}}) // CHECK-LABEL: define void @va_longdouble(fp128* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}})
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 9 Lines
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi fp128** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi fp128** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load fp128*, fp128** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load fp128*, fp128** [[VA_ARG_ADDR]]
// CHECK: [[RET:%[^ ]+]] = load fp128, fp128* [[INDIRECT_ARG]] // CHECK: [[RET:%[^ ]+]] = load fp128, fp128* [[INDIRECT_ARG]]
// CHECK: store fp128 [[RET]], fp128* %{{.*}} // CHECK: store fp128 [[RET]], fp128* %{{.*}}
// CHECK: ret void // CHECK: ret void
_Complex char va_complex_char(__builtin_va_list l) { return __builtin_va_arg(l, _Complex char); } _Complex char va_complex_char(__builtin_va_list l) { return __builtin_va_arg(l, _Complex char); }
// CHECK-LABEL: define void @va_complex_char({ i8, i8 }* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_complex_char({ i8, i8 }* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to { i8, i8 }** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to { i8, i8 }**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to { i8, i8 }** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to { i8, i8 }**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi { i8, i8 }** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi { i8, i8 }** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load { i8, i8 }*, { i8, i8 }** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load { i8, i8 }*, { i8, i8 }** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
struct agg_1byte va_agg_1byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_1byte); } struct agg_1byte va_agg_1byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_1byte); }
// CHECK-LABEL: define void @va_agg_1byte(%struct.agg_1byte* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_1byte(%struct.agg_1byte* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 23 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 23
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_1byte* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_1byte*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 7 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 7
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_1byte* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_1byte*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_1byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_1byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_2byte va_agg_2byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_2byte); } struct agg_2byte va_agg_2byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_2byte); }
// CHECK-LABEL: define void @va_agg_2byte(%struct.agg_2byte* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_2byte(%struct.agg_2byte* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 22 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 22
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_2byte* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_2byte*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 6 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 6
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_2byte* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_2byte*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_2byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_2byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_3byte va_agg_3byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_3byte); } struct agg_3byte va_agg_3byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_3byte); }
// CHECK-LABEL: define void @va_agg_3byte(%struct.agg_3byte* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_3byte(%struct.agg_3byte* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_3byte** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_3byte**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_3byte** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_3byte**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_3byte** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_3byte** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_3byte*, %struct.agg_3byte** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_3byte*, %struct.agg_3byte** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
struct agg_4byte va_agg_4byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_4byte); } struct agg_4byte va_agg_4byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_4byte); }
// CHECK-LABEL: define void @va_agg_4byte(%struct.agg_4byte* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_4byte(%struct.agg_4byte* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_4byte* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_4byte*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_4byte* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_4byte*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_4byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_4byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_8byte va_agg_8byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_8byte); } struct agg_8byte va_agg_8byte(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_8byte); }
// CHECK-LABEL: define void @va_agg_8byte(%struct.agg_8byte* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_8byte(%struct.agg_8byte* noalias sret align 1 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_8byte* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_8byte*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_8byte* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_8byte*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_8byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_8byte* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_float va_agg_float(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float); } struct agg_float va_agg_float(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float); }
// CHECK-LABEL: define void @va_agg_float(%struct.agg_float* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_float(%struct.agg_float* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1
// SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4 // HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4
// SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128 // HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128
Show All 9 Lines
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 4
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_double va_agg_double(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_double); } struct agg_double va_agg_double(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_double); }
// CHECK-LABEL: define void @va_agg_double(%struct.agg_double* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_double(%struct.agg_double* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1
// SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4 // HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4
// SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128 // HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128
Show All 9 Lines
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_double* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_double*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_double* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_double* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_longdouble va_agg_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_longdouble); } struct agg_longdouble va_agg_longdouble(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_longdouble); }
// CHECK-LABEL: define void @va_agg_longdouble(%struct.agg_longdouble* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_longdouble(%struct.agg_longdouble* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_longdouble** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_longdouble**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_longdouble** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_longdouble**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_longdouble** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_longdouble** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_longdouble*, %struct.agg_longdouble** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_longdouble*, %struct.agg_longdouble** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
struct agg_float_a8 va_agg_float_a8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a8); } struct agg_float_a8 va_agg_float_a8(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a8); }
// CHECK-LABEL: define void @va_agg_float_a8(%struct.agg_float_a8* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_float_a8(%struct.agg_float_a8* noalias sret align 8 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1 // HARD-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 1
// SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // SOFT-FLOAT: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4 // HARD-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 4
// SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // SOFT-FLOAT: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128 // HARD-FLOAT: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 128
Show All 9 Lines
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float_a8* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float_a8*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float_a8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float_a8* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_float_a16 va_agg_float_a16(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a16); } struct agg_float_a16 va_agg_float_a16(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_float_a16); }
// CHECK-LABEL: define void @va_agg_float_a16(%struct.agg_float_a16* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_float_a16(%struct.agg_float_a16* noalias sret align 16 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_float_a16** // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_float_a16**
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float_a16** // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_float_a16**
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float_a16** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_float_a16** [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_float_a16*, %struct.agg_float_a16** [[VA_ARG_ADDR]] // CHECK: [[INDIRECT_ARG:%[^ ]+]] = load %struct.agg_float_a16*, %struct.agg_float_a16** [[VA_ARG_ADDR]]
// CHECK: ret void // CHECK: ret void
struct agg_nofloat1 va_agg_nofloat1(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat1); } struct agg_nofloat1 va_agg_nofloat1(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat1); }
// CHECK-LABEL: define void @va_agg_nofloat1(%struct.agg_nofloat1* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_nofloat1(%struct.agg_nofloat1* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_nofloat1* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_nofloat1*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_nofloat1* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_nofloat1*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_nofloat1* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_nofloat1* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_nofloat2 va_agg_nofloat2(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat2); } struct agg_nofloat2 va_agg_nofloat2(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat2); }
// CHECK-LABEL: define void @va_agg_nofloat2(%struct.agg_nofloat2* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_nofloat2(%struct.agg_nofloat2* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 16
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
// CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]] // CHECK: [[RAW_REG_ADDR:%[^ ]+]] = getelementptr i8, i8* [[REG_SAVE_AREA]], i64 [[REG_OFFSET]]
// CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_nofloat2* // CHECK: [[REG_ADDR:%[^ ]+]] = bitcast i8* [[RAW_REG_ADDR]] to %struct.agg_nofloat2*
// CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1 // CHECK: [[REG_COUNT1:%[^ ]+]] = add i64 [[REG_COUNT]], 1
// CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]] // CHECK: store i64 [[REG_COUNT1]], i64* [[REG_COUNT_PTR]]
// CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2 // CHECK: [[OVERFLOW_ARG_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 2
// CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: [[OVERFLOW_ARG_AREA:%[^ ]+]] = load i8*, i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0 // CHECK: [[RAW_MEM_ADDR:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 0
// CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_nofloat2* // CHECK: [[MEM_ADDR:%[^ ]+]] = bitcast i8* [[RAW_MEM_ADDR]] to %struct.agg_nofloat2*
// CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8 // CHECK: [[OVERFLOW_ARG_AREA2:%[^ ]+]] = getelementptr i8, i8* [[OVERFLOW_ARG_AREA]], i64 8
// CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]] // CHECK: store i8* [[OVERFLOW_ARG_AREA2]], i8** [[OVERFLOW_ARG_AREA_PTR]]
// CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_nofloat2* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ] // CHECK: [[VA_ARG_ADDR:%[^ ]+]] = phi %struct.agg_nofloat2* [ [[REG_ADDR]], %{{.*}} ], [ [[MEM_ADDR]], %{{.*}} ]
// CHECK: ret void // CHECK: ret void
struct agg_nofloat3 va_agg_nofloat3(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat3); } struct agg_nofloat3 va_agg_nofloat3(__builtin_va_list l) { return __builtin_va_arg(l, struct agg_nofloat3); }
// CHECK-LABEL: define void @va_agg_nofloat3(%struct.agg_nofloat3* noalias sret %{{.*}}, %struct.__va_list_tag* %{{.*}} // CHECK-LABEL: define void @va_agg_nofloat3(%struct.agg_nofloat3* noalias sret align 4 %{{.*}}, %struct.__va_list_tag* %{{.*}}
// CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0 // CHECK: [[REG_COUNT_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 0
// CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]] // CHECK: [[REG_COUNT:%[^ ]+]] = load i64, i64* [[REG_COUNT_PTR]]
// CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5 // CHECK: [[FITS_IN_REGS:%[^ ]+]] = icmp ult i64 [[REG_COUNT]], 5
// CHECK: br i1 [[FITS_IN_REGS]], // CHECK: br i1 [[FITS_IN_REGS]],
// CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8 // CHECK: [[SCALED_REG_COUNT:%[^ ]+]] = mul i64 [[REG_COUNT]], 8
// CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20 // CHECK: [[REG_OFFSET:%[^ ]+]] = add i64 [[SCALED_REG_COUNT]], 20
// CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3 // CHECK: [[REG_SAVE_AREA_PTR:%[^ ]+]] = getelementptr inbounds %struct.__va_list_tag, %struct.__va_list_tag* %{{.*}}, i32 0, i32 3
// CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]] // CHECK: [[REG_SAVE_AREA:%[^ ]+]] = load i8*, i8** [[REG_SAVE_AREA_PTR:[^ ]+]]
Show All 13 Lines

clang/test/CodeGen/systemz-abi.cpp

// RUN: %clang_cc1 -triple s390x-linux-gnu -emit-llvm -x c++ -o - %s | FileCheck %s // RUN: %clang_cc1 -triple s390x-linux-gnu -emit-llvm -x c++ -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple s390x-linux-gnu -emit-llvm -x c++ -o - %s -mfloat-abi soft \ // RUN: %clang_cc1 -triple s390x-linux-gnu -emit-llvm -x c++ -o - %s -mfloat-abi soft \
// RUN: | FileCheck %s --check-prefix=SOFT-FLOAT // RUN: | FileCheck %s --check-prefix=SOFT-FLOAT
// For compatibility with GCC, this structure is passed in an FPR in C++, // For compatibility with GCC, this structure is passed in an FPR in C++,
// but passed in a GPR in C (checked in systemz-abi.c). // but passed in a GPR in C (checked in systemz-abi.c).
struct agg_float_cpp { float a; int : 0; }; struct agg_float_cpp { float a; int : 0; };
struct agg_float_cpp pass_agg_float_cpp(struct agg_float_cpp arg) { return arg; } struct agg_float_cpp pass_agg_float_cpp(struct agg_float_cpp arg) { return arg; }
// CHECK-LABEL: define void @_Z18pass_agg_float_cpp13agg_float_cpp(%struct.agg_float_cpp* noalias sret %{{.*}}, float %{{.*}}) // CHECK-LABEL: define void @_Z18pass_agg_float_cpp13agg_float_cpp(%struct.agg_float_cpp* noalias sret align 4 %{{.*}}, float %{{.*}})
// SOFT-FLOAT: define void @_Z18pass_agg_float_cpp13agg_float_cpp(%struct.agg_float_cpp* noalias sret %{{.*}}, i32 %{{.*}}) // SOFT-FLOAT: define void @_Z18pass_agg_float_cpp13agg_float_cpp(%struct.agg_float_cpp* noalias sret align 4 %{{.*}}, i32 %{{.*}})

clang/test/CodeGen/systemz-inline-asm.c

Show First 20 Lines • Show All 117 Lines • ▼ Show 20 Linesdouble test_f64(double f, double g) {
return f; return f;
// CHECK-LABEL: define double @test_f64(double %f, double %g) // CHECK-LABEL: define double @test_f64(double %f, double %g)
// CHECK: call double asm "adbr $0, $2", "=f,0,f"(double %f, double %g) // CHECK: call double asm "adbr $0, $2", "=f,0,f"(double %f, double %g)
} }
long double test_f128(long double f, long double g) { long double test_f128(long double f, long double g) {
asm("axbr %0, %2" : "=f" (f) : "0" (f), "f" (g)); asm("axbr %0, %2" : "=f" (f) : "0" (f), "f" (g));
return f; return f;
// CHECK: define void @test_f128(fp128* noalias nocapture sret [[DEST:%.*]], fp128* nocapture readonly %0, fp128* nocapture readonly %1) // CHECK: define void @test_f128(fp128* noalias nocapture sret align 8 [[DEST:%.*]], fp128* nocapture readonly %0, fp128* nocapture readonly %1)
// CHECK: %f = load fp128, fp128* %0 // CHECK: %f = load fp128, fp128* %0
// CHECK: %g = load fp128, fp128* %1 // CHECK: %g = load fp128, fp128* %1
// CHECK: [[RESULT:%.*]] = tail call fp128 asm "axbr $0, $2", "=f,0,f"(fp128 %f, fp128 %g) // CHECK: [[RESULT:%.*]] = tail call fp128 asm "axbr $0, $2", "=f,0,f"(fp128 %f, fp128 %g)
// CHECK: store fp128 [[RESULT]], fp128* [[DEST]] // CHECK: store fp128 [[RESULT]], fp128* [[DEST]]
} }

clang/test/CodeGen/vectorcall.c

Show First 20 Lines • Show All 80 Lines • ▼ Show 20 Lines
// X32: define dso_local x86_vectorcallcc <4 x float> @"\01hva5@@144"(%struct.HVA3 inreg %a.coerce, %struct.HVA3* inreg %b, <4 x float> inreg %c, %struct.HVA2 inreg %d.coerce) // X32: define dso_local x86_vectorcallcc <4 x float> @"\01hva5@@144"(%struct.HVA3 inreg %a.coerce, %struct.HVA3* inreg %b, <4 x float> inreg %c, %struct.HVA2 inreg %d.coerce)
// X64: define dso_local x86_vectorcallcc <4 x float> @"\01hva5@@144"(%struct.HVA3 inreg %a.coerce, %struct.HVA3* %b, <4 x float> %c, %struct.HVA2 inreg %d.coerce) // X64: define dso_local x86_vectorcallcc <4 x float> @"\01hva5@@144"(%struct.HVA3 inreg %a.coerce, %struct.HVA3* %b, <4 x float> %c, %struct.HVA2 inreg %d.coerce)
struct HVA4 __vectorcall hva6(struct HVA4 a, struct HVA4 b) { return b;} struct HVA4 __vectorcall hva6(struct HVA4 a, struct HVA4 b) { return b;}
// X32: define dso_local x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* inreg %b) // X32: define dso_local x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* inreg %b)
// X64: define dso_local x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* %b) // X64: define dso_local x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* %b)
struct HVA5 __vectorcall hva7() {struct HVA5 a = {}; return a;} struct HVA5 __vectorcall hva7() {struct HVA5 a = {}; return a;}
// X32: define dso_local x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* inreg noalias sret %agg.result) // X32: define dso_local x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* inreg noalias sret align 16 %agg.result)
// X64: define dso_local x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* noalias sret %agg.result) // X64: define dso_local x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* noalias sret align 16 %agg.result)
v4f32 __vectorcall hva8(v4f32 a, v4f32 b, v4f32 c, v4f32 d, int e, v4f32 f) {return f;} v4f32 __vectorcall hva8(v4f32 a, v4f32 b, v4f32 c, v4f32 d, int e, v4f32 f) {return f;}
// X32: define dso_local x86_vectorcallcc <4 x float> @"\01hva8@@84"(<4 x float> inreg %a, <4 x float> inreg %b, <4 x float> inreg %c, <4 x float> inreg %d, i32 inreg %e, <4 x float> inreg %f) // X32: define dso_local x86_vectorcallcc <4 x float> @"\01hva8@@84"(<4 x float> inreg %a, <4 x float> inreg %b, <4 x float> inreg %c, <4 x float> inreg %d, i32 inreg %e, <4 x float> inreg %f)
// X64: define dso_local x86_vectorcallcc <4 x float> @"\01hva8@@88"(<4 x float> %a, <4 x float> %b, <4 x float> %c, <4 x float> %d, i32 %e, <4 x float> %f) // X64: define dso_local x86_vectorcallcc <4 x float> @"\01hva8@@88"(<4 x float> %a, <4 x float> %b, <4 x float> %c, <4 x float> %d, i32 %e, <4 x float> %f)
typedef float __attribute__((ext_vector_type(3))) v3f32; typedef float __attribute__((ext_vector_type(3))) v3f32;
struct OddSizeHVA { v3f32 x, y; }; struct OddSizeHVA { v3f32 x, y; };
▲ Show 20 Lines • Show All 41 Lines • Show Last 20 Lines

clang/test/CodeGen/wasm-arguments.c

Show All 19 Lines
// WEBASSEMBLY32: define void @struct_arg(%struct.s1* byval(%struct.s1) align 4 %i) // WEBASSEMBLY32: define void @struct_arg(%struct.s1* byval(%struct.s1) align 4 %i)
// WEBASSEMBLY64: define void @struct_arg(%struct.s1* byval(%struct.s1) align 4 %i) // WEBASSEMBLY64: define void @struct_arg(%struct.s1* byval(%struct.s1) align 4 %i)
// Except in the experimental multivalue ABI, where structs are passed in args // Except in the experimental multivalue ABI, where structs are passed in args
// EXPERIMENTAL-MV: define void @struct_arg(i32 %i.0, i32 %i.1) // EXPERIMENTAL-MV: define void @struct_arg(i32 %i.0, i32 %i.1)
void struct_arg(s1 i) {} void struct_arg(s1 i) {}
// Structs should be returned sret and not simplified by the frontend. // Structs should be returned sret and not simplified by the frontend.
// WEBASSEMBLY32: define void @struct_ret(%struct.s1* noalias sret %agg.result) // WEBASSEMBLY32: define void @struct_ret(%struct.s1* noalias sret align 4 %agg.result)
// WEBASSEMBLY32: ret void // WEBASSEMBLY32: ret void
// WEBASSEMBLY64: define void @struct_ret(%struct.s1* noalias sret %agg.result) // WEBASSEMBLY64: define void @struct_ret(%struct.s1* noalias sret align 4 %agg.result)
// WEBASSEMBLY64: ret void // WEBASSEMBLY64: ret void
// Except with the experimental multivalue ABI, which returns structs by value // Except with the experimental multivalue ABI, which returns structs by value
// EXPERIMENTAL-MV: define %struct.s1 @struct_ret() // EXPERIMENTAL-MV: define %struct.s1 @struct_ret()
// EXPERIMENTAL-MV: ret %struct.s1 %0 // EXPERIMENTAL-MV: ret %struct.s1 %0
s1 struct_ret() { s1 struct_ret() {
s1 foo; s1 foo;
return foo; return foo;
▲ Show 20 Lines • Show All 59 Lines • ▼ Show 20 Lines
// Unions should be passed as byval structs. // Unions should be passed as byval structs.
// WEBASSEMBLY32: define void @union_arg(%union.simple_union* byval(%union.simple_union) align 4 %s) // WEBASSEMBLY32: define void @union_arg(%union.simple_union* byval(%union.simple_union) align 4 %s)
// WEBASSEMBLY64: define void @union_arg(%union.simple_union* byval(%union.simple_union) align 4 %s) // WEBASSEMBLY64: define void @union_arg(%union.simple_union* byval(%union.simple_union) align 4 %s)
// EXPERIMENTAL-MV: define void @union_arg(i32 %s.0) // EXPERIMENTAL-MV: define void @union_arg(i32 %s.0)
void union_arg(union simple_union s) {} void union_arg(union simple_union s) {}
// Unions should be returned sret and not simplified by the frontend. // Unions should be returned sret and not simplified by the frontend.
// WEBASSEMBLY32: define void @union_ret(%union.simple_union* noalias sret %agg.result) // WEBASSEMBLY32: define void @union_ret(%union.simple_union* noalias sret align 4 %agg.result)
// WEBASSEMBLY32: ret void // WEBASSEMBLY32: ret void
// WEBASSEMBLY64: define void @union_ret(%union.simple_union* noalias sret %agg.result) // WEBASSEMBLY64: define void @union_ret(%union.simple_union* noalias sret align 4 %agg.result)
// WEBASSEMBLY64: ret void // WEBASSEMBLY64: ret void
// The experimental multivalue ABI returns them by value, though. // The experimental multivalue ABI returns them by value, though.
// EXPERIMENTAL-MV: define %union.simple_union @union_ret() // EXPERIMENTAL-MV: define %union.simple_union @union_ret()
// EXPERIMENTAL-MV: ret %union.simple_union %0 // EXPERIMENTAL-MV: ret %union.simple_union %0
union simple_union union_ret() { union simple_union union_ret() {
union simple_union bar; union simple_union bar;
return bar; return bar;
} }
typedef struct { typedef struct {
int b4 : 4; int b4 : 4;
int b3 : 3; int b3 : 3;
int b8 : 8; int b8 : 8;
} bitfield1; } bitfield1;
// Bitfields should be passed as byval structs. // Bitfields should be passed as byval structs.
// WEBASSEMBLY32: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1) // WEBASSEMBLY32: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1)
// WEBASSEMBLY64: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1) // WEBASSEMBLY64: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1)
// EXPERIMENTAL-MV: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1) // EXPERIMENTAL-MV: define void @bitfield_arg(%struct.bitfield1* byval(%struct.bitfield1) align 4 %bf1)
void bitfield_arg(bitfield1 bf1) {} void bitfield_arg(bitfield1 bf1) {}
// And returned via sret pointers. // And returned via sret pointers.
// WEBASSEMBLY32: define void @bitfield_ret(%struct.bitfield1* noalias sret %agg.result) // WEBASSEMBLY32: define void @bitfield_ret(%struct.bitfield1* noalias sret align 4 %agg.result)
// WEBASSEMBLY64: define void @bitfield_ret(%struct.bitfield1* noalias sret %agg.result) // WEBASSEMBLY64: define void @bitfield_ret(%struct.bitfield1* noalias sret align 4 %agg.result)
// Except, of course, in the experimental multivalue ABI // Except, of course, in the experimental multivalue ABI
// EXPERIMENTAL-MV: define %struct.bitfield1 @bitfield_ret() // EXPERIMENTAL-MV: define %struct.bitfield1 @bitfield_ret()
bitfield1 bitfield_ret() { bitfield1 bitfield_ret() {
bitfield1 baz; bitfield1 baz;
return baz; return baz;
} }

clang/test/CodeGen/wasm-varargs.c

Show First 20 Lines • Show All 74 Lines • ▼ Show 20 Linesstruct S test_struct(char *fmt, ...) {
va_start(va, fmt); va_start(va, fmt);
struct S v = va_arg(va, struct S); struct S v = va_arg(va, struct S);
va_end(va); va_end(va);
return v; return v;
} }
// CHECK: define void @test_struct([[STRUCT_S:%[^,=]+]]*{{.*}} noalias sret [[AGG_RESULT:%.*]], i8*{{.*}} %fmt, ...) {{.*}} { // CHECK: define void @test_struct([[STRUCT_S:%[^,=]+]]*{{.*}} noalias sret align 4 [[AGG_RESULT:%.*]], i8*{{.*}} %fmt, ...) {{.*}} {
// CHECK: [[FMT_ADDR:%[^,=]+]] = alloca i8*, align 4 // CHECK: [[FMT_ADDR:%[^,=]+]] = alloca i8*, align 4
// CHECK-NEXT: [[VA:%[^,=]+]] = alloca i8*, align 4 // CHECK-NEXT: [[VA:%[^,=]+]] = alloca i8*, align 4
// CHECK-NEXT: store i8* %fmt, i8** [[FMT_ADDR]], align 4 // CHECK-NEXT: store i8* %fmt, i8** [[FMT_ADDR]], align 4
// CHECK-NEXT: [[VA1:%[^,=]+]] = bitcast i8** [[VA]] to i8* // CHECK-NEXT: [[VA1:%[^,=]+]] = bitcast i8** [[VA]] to i8*
// CHECK-NEXT: call void @llvm.va_start(i8* [[VA1]]) // CHECK-NEXT: call void @llvm.va_start(i8* [[VA1]])
// CHECK-NEXT: [[ARGP_CUR:%[^,=]+]] = load i8*, i8** [[VA]], align 4 // CHECK-NEXT: [[ARGP_CUR:%[^,=]+]] = load i8*, i8** [[VA]], align 4
// CHECK-NEXT: [[ARGP_NEXT:%[^,=]+]] = getelementptr inbounds i8, i8* [[ARGP_CUR]], i32 4 // CHECK-NEXT: [[ARGP_NEXT:%[^,=]+]] = getelementptr inbounds i8, i8* [[ARGP_CUR]], i32 4
// CHECK-NEXT: store i8* [[ARGP_NEXT]], i8** [[VA]], align 4 // CHECK-NEXT: store i8* [[ARGP_NEXT]], i8** [[VA]], align 4
Show All 15 Linesstruct S test_empty_struct(char *fmt, ...) {
va_start(va, fmt); va_start(va, fmt);
struct Z u = va_arg(va, struct Z); struct Z u = va_arg(va, struct Z);
struct S v = va_arg(va, struct S); struct S v = va_arg(va, struct S);
va_end(va); va_end(va);
return v; return v;
} }
// CHECK: define void @test_empty_struct([[STRUCT_S:%[^,=]+]]*{{.*}} noalias sret [[AGG_RESULT:%.*]], i8*{{.*}} %fmt, ...) {{.*}} { // CHECK: define void @test_empty_struct([[STRUCT_S:%[^,=]+]]*{{.*}} noalias sret align 4 [[AGG_RESULT:%.*]], i8*{{.*}} %fmt, ...) {{.*}} {
// CHECK: [[FMT_ADDR:%[^,=]+]] = alloca i8*, align 4 // CHECK: [[FMT_ADDR:%[^,=]+]] = alloca i8*, align 4
// CHECK-NEXT: [[VA:%[^,=]+]] = alloca i8*, align 4 // CHECK-NEXT: [[VA:%[^,=]+]] = alloca i8*, align 4
// CHECK-NEXT: [[U:%[^,=]+]] = alloca [[STRUCT_Z:%[^,=]+]], align 1 // CHECK-NEXT: [[U:%[^,=]+]] = alloca [[STRUCT_Z:%[^,=]+]], align 1
// CHECK-NEXT: store i8* %fmt, i8** [[FMT_ADDR]], align 4 // CHECK-NEXT: store i8* %fmt, i8** [[FMT_ADDR]], align 4
// CHECK-NEXT: [[VA1:%[^,=]+]] = bitcast i8** [[VA]] to i8* // CHECK-NEXT: [[VA1:%[^,=]+]] = bitcast i8** [[VA]] to i8*
// CHECK-NEXT: call void @llvm.va_start(i8* [[VA1]]) // CHECK-NEXT: call void @llvm.va_start(i8* [[VA1]])
// CHECK-NEXT: [[ARGP_CUR:%[^,=]+]] = load i8*, i8** [[VA]], align 4 // CHECK-NEXT: [[ARGP_CUR:%[^,=]+]] = load i8*, i8** [[VA]], align 4
// CHECK-NEXT: [[ARGP_NEXT:%[^,=]+]] = getelementptr inbounds i8, i8* [[ARGP_CUR]], i32 0 // CHECK-NEXT: [[ARGP_NEXT:%[^,=]+]] = getelementptr inbounds i8, i8* [[ARGP_CUR]], i32 0
Show All 17 Lines

clang/test/CodeGen/windows-struct-abi.c

Show All 28 Linesstruct f4 {
float f; float f;
float g; float g;
float h; float h;
float i; float i;
}; };
struct f4 return_f4(void) { while (1); } struct f4 return_f4(void) { while (1); }
// CHECK: define dso_local void @return_f4(%struct.f4* noalias sret %agg.result) // CHECK: define dso_local void @return_f4(%struct.f4* noalias sret align 4 %agg.result)
void receive_f4(struct f4 a0) { } void receive_f4(struct f4 a0) { }
// CHECK: define dso_local void @receive_f4(float %a0.0, float %a0.1, float %a0.2, float %a0.3) // CHECK: define dso_local void @receive_f4(float %a0.0, float %a0.1, float %a0.2, float %a0.3)

clang/test/CodeGen/x86_32-arguments-darwin.c

Show First 20 Lines • Show All 65 Lines • ▼ Show 20 Lines
struct s10 { struct s10 {
union { }; union { };
float f; float f;
} f10(void) { while (1) {} } } f10(void) { while (1) {} }
// Small vectors and 1 x {i64,double} are returned in registers // Small vectors and 1 x {i64,double} are returned in registers
// CHECK: i32 @f11() // CHECK: i32 @f11()
// CHECK: void @f12(<2 x i32>* noalias sret %agg.result) // CHECK: void @f12(<2 x i32>* noalias sret align 8 %agg.result)
// CHECK: i64 @f13() // CHECK: i64 @f13()
// CHECK: i64 @f14() // CHECK: i64 @f14()
// CHECK: <2 x i64> @f15() // CHECK: <2 x i64> @f15()
// CHECK: <2 x i64> @f16() // CHECK: <2 x i64> @f16()
typedef short T11 __attribute__ ((vector_size (4))); typedef short T11 __attribute__ ((vector_size (4)));
T11 f11(void) { while (1) {} } T11 f11(void) { while (1) {} }
typedef int T12 __attribute__ ((vector_size (8))); typedef int T12 __attribute__ ((vector_size (8)));
T12 f12(void) { while (1) {} } T12 f12(void) { while (1) {} }
typedef long long T13 __attribute__ ((vector_size (8))); typedef long long T13 __attribute__ ((vector_size (8)));
T13 f13(void) { while (1) {} } T13 f13(void) { while (1) {} }
typedef double T14 __attribute__ ((vector_size (8))); typedef double T14 __attribute__ ((vector_size (8)));
T14 f14(void) { while (1) {} } T14 f14(void) { while (1) {} }
typedef long long T15 __attribute__ ((vector_size (16))); typedef long long T15 __attribute__ ((vector_size (16)));
T15 f15(void) { while (1) {} } T15 f15(void) { while (1) {} }
typedef double T16 __attribute__ ((vector_size (16))); typedef double T16 __attribute__ ((vector_size (16)));
T16 f16(void) { while (1) {} } T16 f16(void) { while (1) {} }
// And when the single element in a struct (but not for 64 and // And when the single element in a struct (but not for 64 and
// 128-bits). // 128-bits).
// CHECK: i32 @f17() // CHECK: i32 @f17()
// CHECK: void @f18(%{{.*}}* noalias sret %agg.result) // CHECK: void @f18(%{{.*}}* noalias sret align 8 %agg.result)
// CHECK: void @f19(%{{.*}}* noalias sret %agg.result) // CHECK: void @f19(%{{.*}}* noalias sret align 8 %agg.result)
// CHECK: void @f20(%{{.*}}* noalias sret %agg.result) // CHECK: void @f20(%{{.*}}* noalias sret align 8 %agg.result)
// CHECK: void @f21(%{{.*}}* noalias sret %agg.result) // CHECK: void @f21(%{{.*}}* noalias sret align 16 %agg.result)
// CHECK: void @f22(%{{.*}}* noalias sret %agg.result) // CHECK: void @f22(%{{.*}}* noalias sret align 16 %agg.result)
struct { T11 a; } f17(void) { while (1) {} } struct { T11 a; } f17(void) { while (1) {} }
struct { T12 a; } f18(void) { while (1) {} } struct { T12 a; } f18(void) { while (1) {} }
struct { T13 a; } f19(void) { while (1) {} } struct { T13 a; } f19(void) { while (1) {} }
struct { T14 a; } f20(void) { while (1) {} } struct { T14 a; } f20(void) { while (1) {} }
struct { T15 a; } f21(void) { while (1) {} } struct { T15 a; } f21(void) { while (1) {} }
struct { T16 a; } f22(void) { while (1) {} } struct { T16 a; } f22(void) { while (1) {} }
// Single element structures are handled specially // Single element structures are handled specially
// CHECK: float @f23() // CHECK: float @f23()
// CHECK: float @f24() // CHECK: float @f24()
// CHECK: float @f25() // CHECK: float @f25()
struct { float a; } f23(void) { while (1) {} } struct { float a; } f23(void) { while (1) {} }
struct { float a[1]; } f24(void) { while (1) {} } struct { float a[1]; } f24(void) { while (1) {} }
struct { struct {} a; struct { float a[1]; } b; } f25(void) { while (1) {} } struct { struct {} a; struct { float a[1]; } b; } f25(void) { while (1) {} }
// Small structures are handled recursively // Small structures are handled recursively
// CHECK: i32 @f26() // CHECK: i32 @f26()
// CHECK: void @f27(%struct.s27* noalias sret %agg.result) // CHECK: void @f27(%struct.s27* noalias sret align 1 %agg.result)
struct s26 { struct { char a, b; } a; struct { char a, b; } b; } f26(void) { while (1) {} } struct s26 { struct { char a, b; } a; struct { char a, b; } b; } f26(void) { while (1) {} }
struct s27 { struct { char a, b, c; } a; struct { char a; } b; } f27(void) { while (1) {} } struct s27 { struct { char a, b, c; } a; struct { char a; } b; } f27(void) { while (1) {} }
// CHECK: void @f28(%struct.s28* noalias sret %agg.result) // CHECK: void @f28(%struct.s28* noalias sret align 4 %agg.result)
struct s28 { int a; int b[]; } f28(void) { while (1) {} } struct s28 { int a; int b[]; } f28(void) { while (1) {} }
// CHECK-LABEL: define i16 @f29() // CHECK-LABEL: define i16 @f29()
struct s29 { struct { } a[1]; char b; char c; } f29(void) { while (1) {} } struct s29 { struct { } a[1]; char b; char c; } f29(void) { while (1) {} }
// CHECK-LABEL: define i16 @f30() // CHECK-LABEL: define i16 @f30()
struct s30 { char a; char b : 4; } f30(void) { while (1) {} } struct s30 { char a; char b : 4; } f30(void) { while (1) {} }
Show All 13 Lines
struct s35 { struct { int : 0; } a; char b; char c; } f35(void) { while (1) {} } struct s35 { struct { int : 0; } a; char b; char c; } f35(void) { while (1) {} }
// CHECK-LABEL: define i16 @f36() // CHECK-LABEL: define i16 @f36()
struct s36 { struct { int : 0; } a[2][10]; char b; char c; } f36(void) { while (1) {} } struct s36 { struct { int : 0; } a[2][10]; char b; char c; } f36(void) { while (1) {} }
// CHECK-LABEL: define float @f37() // CHECK-LABEL: define float @f37()
struct s37 { float c[1][1]; } f37(void) { while (1) {} } struct s37 { float c[1][1]; } f37(void) { while (1) {} }
// CHECK-LABEL: define void @f38(%struct.s38* noalias sret %agg.result) // CHECK-LABEL: define void @f38(%struct.s38* noalias sret align 2 %agg.result)
struct s38 { char a[3]; short b; } f38(void) { while (1) {} } struct s38 { char a[3]; short b; } f38(void) { while (1) {} }
// CHECK-LABEL: define void @f39(%struct.s39* byval(%struct.s39) align 16 %x) // CHECK-LABEL: define void @f39(%struct.s39* byval(%struct.s39) align 16 %x)
typedef int v39 __attribute((vector_size(16))); typedef int v39 __attribute((vector_size(16)));
struct s39 { v39 x; }; struct s39 { v39 x; };
void f39(struct s39 x) {} void f39(struct s39 x) {}
// <rdar://problem/7247671> // <rdar://problem/7247671>
▲ Show 20 Lines • Show All 183 Lines • Show Last 20 Lines

clang/test/CodeGen/x86_32-arguments-iamcu.c

Show First 20 Lines • Show All 52 Lines • ▼ Show 20 Lines
st3_t retNonPow2Struct(st3_t r) { return r; } st3_t retNonPow2Struct(st3_t r) { return r; }
// CHECK-LABEL: define i32 @retSmallStruct(i32 %r.coerce) // CHECK-LABEL: define i32 @retSmallStruct(i32 %r.coerce)
st4_t retSmallStruct(st4_t r) { return r; } st4_t retSmallStruct(st4_t r) { return r; }
// CHECK-LABEL: define i64 @retPaddedStruct(i32 %r.coerce0, i32 %r.coerce1) // CHECK-LABEL: define i64 @retPaddedStruct(i32 %r.coerce0, i32 %r.coerce1)
st5_t retPaddedStruct(st5_t r) { return r; } st5_t retPaddedStruct(st5_t r) { return r; }
// CHECK-LABEL: define void @retLargeStruct(%struct.st12_t* noalias sret %agg.result, i32 %i1, %struct.st12_t* byval(%struct.st12_t) align 4 %r) // CHECK-LABEL: define void @retLargeStruct(%struct.st12_t* noalias sret align 4 %agg.result, i32 %i1, %struct.st12_t* byval(%struct.st12_t) align 4 %r)
st12_t retLargeStruct(int i1, st12_t r) { return r; } st12_t retLargeStruct(int i1, st12_t r) { return r; }
// CHECK-LABEL: define i32 @varArgs(i32 %i1, ...) // CHECK-LABEL: define i32 @varArgs(i32 %i1, ...)
int varArgs(int i1, ...) { return i1; } int varArgs(int i1, ...) { return i1; }
// CHECK-LABEL: define double @longDoubleArg(double %ld1) // CHECK-LABEL: define double @longDoubleArg(double %ld1)
long double longDoubleArg(long double ld1) { return ld1; } long double longDoubleArg(long double ld1) { return ld1; }

clang/test/CodeGen/x86_64-arguments-nacl.c

Show First 20 Lines • Show All 55 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define i32 @f12_0() // CHECK-LABEL: define i32 @f12_0()
// CHECK-LABEL: define void @f12_1(i32 %a0.coerce) // CHECK-LABEL: define void @f12_1(i32 %a0.coerce)
struct s12 { int a __attribute__((aligned(16))); }; struct s12 { int a __attribute__((aligned(16))); };
struct s12 f12_0(void) { while (1) {} } struct s12 f12_0(void) { while (1) {} }
void f12_1(struct s12 a0) {} void f12_1(struct s12 a0) {}
// Check that sret parameter is accounted for when checking available integer // Check that sret parameter is accounted for when checking available integer
// registers. // registers.
// CHECK: define void @f13(%struct.s13_0* noalias sret %agg.result, i32 %a, i32 %b, i32 %c, i32 %d, {{.*}}* byval({{.*}}) align 8 %e, i32 %f) // CHECK: define void @f13(%struct.s13_0* noalias sret align 8 %agg.result, i32 %a, i32 %b, i32 %c, i32 %d, {{.*}}* byval({{.*}}) align 8 %e, i32 %f)
struct s13_0 { long long f0[3]; }; struct s13_0 { long long f0[3]; };
struct s13_1 { long long f0[2]; }; struct s13_1 { long long f0[2]; };
struct s13_0 f13(int a, int b, int c, int d, struct s13_0 f13(int a, int b, int c, int d,
struct s13_1 e, int f) { while (1) {} } struct s13_1 e, int f) { while (1) {} }
// CHECK-LABEL: define void @f20(%struct.s20* byval(%struct.s20) align 32 %x) // CHECK-LABEL: define void @f20(%struct.s20* byval(%struct.s20) align 32 %x)
struct __attribute__((aligned(32))) s20 { struct __attribute__((aligned(32))) s20 {
Show All 20 Lines

clang/test/CodeGen/x86_64-arguments-win32.c

Show All 21 Lines
void f5(_Complex float a) {} void f5(_Complex float a) {}
// CHECK-LABEL: define dso_local void @f6({ double, double }* %a) // CHECK-LABEL: define dso_local void @f6({ double, double }* %a)
void f6(_Complex double a) {} void f6(_Complex double a) {}
// CHECK-LABEL: define dso_local i64 @f7() // CHECK-LABEL: define dso_local i64 @f7()
_Complex float f7() { return 1.0; } _Complex float f7() { return 1.0; }
// CHECK-LABEL: define dso_local void @f8({ double, double }* noalias sret %agg.result) // CHECK-LABEL: define dso_local void @f8({ double, double }* noalias sret align 8 %agg.result)
_Complex double f8() { return 1.0; } _Complex double f8() { return 1.0; }

clang/test/CodeGen/x86_64-arguments.c

Show First 20 Lines • Show All 41 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define void @f7(i32 %a0) // CHECK-LABEL: define void @f7(i32 %a0)
typedef enum { A, B, C } e7; typedef enum { A, B, C } e7;
void f7(e7 a0) { void f7(e7 a0) {
} }
// Test merging/passing of upper eightbyte with X87 class. // Test merging/passing of upper eightbyte with X87 class.
// //
// CHECK-LABEL: define void @f8_1(%union.u8* noalias sret %agg.result) // CHECK-LABEL: define void @f8_1(%union.u8* noalias sret align 16 %agg.result)
// CHECK-LABEL: define void @f8_2(%union.u8* byval(%union.u8) align 16 %a0) // CHECK-LABEL: define void @f8_2(%union.u8* byval(%union.u8) align 16 %a0)
union u8 { union u8 {
long double a; long double a;
int b; int b;
}; };
union u8 f8_1() { while (1) {} } union u8 f8_1() { while (1) {} }
void f8_2(union u8 a0) {} void f8_2(union u8 a0) {}
// CHECK-LABEL: define i64 @f9() // CHECK-LABEL: define i64 @f9()
struct s9 { int a; int b; int : 0; } f9(void) { while (1) {} } struct s9 { int a; int b; int : 0; } f9(void) { while (1) {} }
// CHECK-LABEL: define void @f10(i64 %a0.coerce) // CHECK-LABEL: define void @f10(i64 %a0.coerce)
struct s10 { int a; int b; int : 0; }; struct s10 { int a; int b; int : 0; };
void f10(struct s10 a0) {} void f10(struct s10 a0) {}
// CHECK-LABEL: define void @f11(%union.anon* noalias sret %agg.result) // CHECK-LABEL: define void @f11(%union.anon* noalias sret align 16 %agg.result)
union { long double a; float b; } f11() { while (1) {} } union { long double a; float b; } f11() { while (1) {} }
// CHECK-LABEL: define i32 @f12_0() // CHECK-LABEL: define i32 @f12_0()
// CHECK-LABEL: define void @f12_1(i32 %a0.coerce) // CHECK-LABEL: define void @f12_1(i32 %a0.coerce)
struct s12 { int a __attribute__((aligned(16))); }; struct s12 { int a __attribute__((aligned(16))); };
struct s12 f12_0(void) { while (1) {} } struct s12 f12_0(void) { while (1) {} }
void f12_1(struct s12 a0) {} void f12_1(struct s12 a0) {}
// Check that sret parameter is accounted for when checking available integer // Check that sret parameter is accounted for when checking available integer
// registers. // registers.
// CHECK: define void @f13(%struct.s13_0* noalias sret %agg.result, i32 %a, i32 %b, i32 %c, i32 %d, {{.*}}* byval({{.*}}) align 8 %e, i32 %f) // CHECK: define void @f13(%struct.s13_0* noalias sret align 8 %agg.result, i32 %a, i32 %b, i32 %c, i32 %d, {{.*}}* byval({{.*}}) align 8 %e, i32 %f)
struct s13_0 { long long f0[3]; }; struct s13_0 { long long f0[3]; };
struct s13_1 { long long f0[2]; }; struct s13_1 { long long f0[2]; };
struct s13_0 f13(int a, int b, int c, int d, struct s13_0 f13(int a, int b, int c, int d,
struct s13_1 e, int f) { while (1) {} } struct s13_1 e, int f) { while (1) {} }
// CHECK: define void @f14({{.*}}, i8 signext %X) // CHECK: define void @f14({{.*}}, i8 signext %X)
void f14(int a, int b, int c, int d, int e, int f, char X) {} void f14(int a, int b, int c, int d, int e, int f, char X) {}
▲ Show 20 Lines • Show All 462 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/arm-cc.cpp

Show All 10 Lines
foo(x); foo(x);
} }
void zed(SMLoc x); void zed(SMLoc x);
void baz() { void baz() {
SMLoc a; SMLoc a;
zed(a); zed(a);
} }
// CHECK: declare void @_Z3fooPv(%class.SMLoc* sret, i8*) // CHECK: declare void @_Z3fooPv(%class.SMLoc* sret align 4, i8*)
// CHECK: declare void @_Z3zed5SMLoc(%class.SMLoc*) // CHECK: declare void @_Z3zed5SMLoc(%class.SMLoc*)

clang/test/CodeGenCXX/builtin-source-location.cpp

Show First 20 Lines • Show All 59 Lines • ▼ Show 20 Lines
// //
// CHECK-GLOBAL-TWO-DAG: @runtime_init_global = global %struct.source_location zeroinitializer, align 8 // CHECK-GLOBAL-TWO-DAG: @runtime_init_global = global %struct.source_location zeroinitializer, align 8
// //
// CHECK-GLOBAL-TWO-DAG: @[[FILE:.*]] = {{.*}}c"test_runtime_init.cpp\00" // CHECK-GLOBAL-TWO-DAG: @[[FILE:.*]] = {{.*}}c"test_runtime_init.cpp\00"
// CHECK-GLOBAL-TWO-DAG: @[[FUNC:.*]] = private unnamed_addr constant [1 x i8] zeroinitializer, align 1 // CHECK-GLOBAL-TWO-DAG: @[[FUNC:.*]] = private unnamed_addr constant [1 x i8] zeroinitializer, align 1
// //
// CHECK-GLOBAL-TWO: define internal void @__cxx_global_var_init() // CHECK-GLOBAL-TWO: define internal void @__cxx_global_var_init()
// CHECK-GLOBAL-TWO-NOT: ret // CHECK-GLOBAL-TWO-NOT: ret
// CHECK-GLOBAL-TWO: call void @_ZN15source_location11bad_currentEjjPKcS1_(%struct.source_location* sret @runtime_init_global, // CHECK-GLOBAL-TWO: call void @_ZN15source_location11bad_currentEjjPKcS1_(%struct.source_location* sret align 8 @runtime_init_global,
// CHECK-GLOBAL-TWO-SAME: i32 1100, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]], // CHECK-GLOBAL-TWO-SAME: i32 1100, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]],
#line 1100 "test_runtime_init.cpp" #line 1100 "test_runtime_init.cpp"
SL runtime_init_global = SL::bad_current(); SL runtime_init_global = SL::bad_current();
#line 2000 "test_function.cpp" #line 2000 "test_function.cpp"
extern "C" void test_function() { extern "C" void test_function() {
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-LOCAL-ONE // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-LOCAL-ONE
// //
// CHECK-LOCAL-ONE-DAG: @[[FILE:.*]] = {{.*}}c"test_current.cpp\00" // CHECK-LOCAL-ONE-DAG: @[[FILE:.*]] = {{.*}}c"test_current.cpp\00"
// CHECK-LOCAL-ONE-DAG: @[[FUNC:.*]] = {{.*}}c"test_function\00" // CHECK-LOCAL-ONE-DAG: @[[FUNC:.*]] = {{.*}}c"test_function\00"
// //
// CHECK-LOCAL-ONE: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %local, // CHECK-LOCAL-ONE: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %local,
// CHECK-LOCAL-ONE-SAME: i32 2100, i32 {{[0-9]+}}, // CHECK-LOCAL-ONE-SAME: i32 2100, i32 {{[0-9]+}},
// CHECK-LOCAL-ONE-SAME: {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]], // CHECK-LOCAL-ONE-SAME: {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]],
#line 2100 "test_current.cpp" #line 2100 "test_current.cpp"
SL local = SL::current(); SL local = SL::current();
} }
#line 3000 "TestInitClass.cpp" #line 3000 "TestInitClass.cpp"
struct TestInit { struct TestInit {
SL info = SL::current(); SL info = SL::current();
SL arg_info; SL arg_info;
#line 3100 "TestInitCtor.cpp" #line 3100 "TestInitCtor.cpp"
TestInit(SL arg_info = SL::current()) : arg_info(arg_info) {} TestInit(SL arg_info = SL::current()) : arg_info(arg_info) {}
}; };
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CTOR-GLOBAL // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CTOR-GLOBAL
// //
// CHECK-CTOR-GLOBAL-DAG: @GlobalInitVal = global %struct.TestInit zeroinitializer, align 8 // CHECK-CTOR-GLOBAL-DAG: @GlobalInitVal = global %struct.TestInit zeroinitializer, align 8
// CHECK-CTOR-GLOBAL-DAG: @[[FILE:.*]] = {{.*}}c"GlobalInitVal.cpp\00" // CHECK-CTOR-GLOBAL-DAG: @[[FILE:.*]] = {{.*}}c"GlobalInitVal.cpp\00"
// CHECK-CTOR-GLOBAL-DAG: @[[FUNC:.*]] = private unnamed_addr constant [1 x i8] zeroinitializer, align 1 // CHECK-CTOR-GLOBAL-DAG: @[[FUNC:.*]] = private unnamed_addr constant [1 x i8] zeroinitializer, align 1
// //
// CHECK-CTOR-GLOBAL: define internal void @__cxx_global_var_init.{{[0-9]+}}() // CHECK-CTOR-GLOBAL: define internal void @__cxx_global_var_init.{{[0-9]+}}()
// CHECK-CTOR-GLOBAL-NOT: ret // CHECK-CTOR-GLOBAL-NOT: ret
// //
// CHECK-CTOR-GLOBAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[TMP_ONE:[^,]*]], // CHECK-CTOR-GLOBAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[TMP_ONE:[^,]*]],
// CHECK-CTOR-GLOBAL-SAME: i32 3400, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]], // CHECK-CTOR-GLOBAL-SAME: i32 3400, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]],
// CHECK-CTOR-GLOBAL-NEXT: call void @_ZN8TestInitC1E15source_location(%struct.TestInit* @GlobalInitVal, %struct.source_location* {{.*}}%[[TMP_ONE]]) // CHECK-CTOR-GLOBAL-NEXT: call void @_ZN8TestInitC1E15source_location(%struct.TestInit* @GlobalInitVal, %struct.source_location* {{.*}}%[[TMP_ONE]])
#line 3400 "GlobalInitVal.cpp" #line 3400 "GlobalInitVal.cpp"
TestInit GlobalInitVal; TestInit GlobalInitVal;
extern "C" void test_init_function() { extern "C" void test_init_function() {
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CTOR-LOCAL // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CTOR-LOCAL
// //
// CHECK-CTOR-LOCAL-DAG: @[[FILE:.*]] = {{.*}}c"LocalInitVal.cpp\00" // CHECK-CTOR-LOCAL-DAG: @[[FILE:.*]] = {{.*}}c"LocalInitVal.cpp\00"
// CHECK-CTOR-LOCAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_init_function\00" // CHECK-CTOR-LOCAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_init_function\00"
// //
// CHECK-CTOR-LOCAL: define void @test_init_function() // CHECK-CTOR-LOCAL: define void @test_init_function()
// CHECK-CTOR-LOCAL-NOT: ret // CHECK-CTOR-LOCAL-NOT: ret
// //
// CHECK-CTOR-LOCAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[TMP:[^,]*]], // CHECK-CTOR-LOCAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[TMP:[^,]*]],
// CHECK-CTOR-LOCAL-SAME: i32 3500, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]], // CHECK-CTOR-LOCAL-SAME: i32 3500, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]],
// CHECK-CTOR-LOCAL-NEXT: call void @_ZN8TestInitC1E15source_location(%struct.TestInit* %init_local, %struct.source_location* {{.*}}%[[TMP]]) // CHECK-CTOR-LOCAL-NEXT: call void @_ZN8TestInitC1E15source_location(%struct.TestInit* %init_local, %struct.source_location* {{.*}}%[[TMP]])
#line 3500 "LocalInitVal.cpp" #line 3500 "LocalInitVal.cpp"
TestInit init_local; TestInit init_local;
sink(init_local); sink(init_local);
} }
#line 4000 "ConstexprClass.cpp" #line 4000 "ConstexprClass.cpp"
Show All 19 Lines
extern "C" void test_init_function_constexpr() { extern "C" void test_init_function_constexpr() {
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CONSTEXPR-LOCAL // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-CONSTEXPR-LOCAL
// //
// CHECK-CONSTEXPR-LOCAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_init_function_constexpr\00" // CHECK-CONSTEXPR-LOCAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_init_function_constexpr\00"
// CHECK-CONSTEXPR-LOCAL-DAG: @[[FILE:.*]] = {{.*}}c"ConstexprLocal.cpp\00" // CHECK-CONSTEXPR-LOCAL-DAG: @[[FILE:.*]] = {{.*}}c"ConstexprLocal.cpp\00"
// //
// CHECK-CONSTEXPR-LOCAL: define void @test_init_function_constexpr() // CHECK-CONSTEXPR-LOCAL: define void @test_init_function_constexpr()
// CHECK-CONSTEXPR-LOCAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[TMP:[^,]*]], // CHECK-CONSTEXPR-LOCAL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[TMP:[^,]*]],
// CHECK-CONSTEXPR-LOCAL-SAME: i32 4600, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]] // CHECK-CONSTEXPR-LOCAL-SAME: i32 4600, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]]
// CHECK-CONSTEXPR-LOCAL: call void @_ZN17TestInitConstexprC1E15source_location(%struct.TestInitConstexpr* %local_val, {{.*}}%[[TMP]]) // CHECK-CONSTEXPR-LOCAL: call void @_ZN17TestInitConstexprC1E15source_location(%struct.TestInitConstexpr* %local_val, {{.*}}%[[TMP]])
#line 4600 "ConstexprLocal.cpp" #line 4600 "ConstexprLocal.cpp"
TestInitConstexpr local_val; TestInitConstexpr local_val;
} }
#line 5000 "TestInitAgg.cpp" #line 5000 "TestInitAgg.cpp"
struct TestInitAgg { struct TestInitAgg {
Show All 19 Lines
extern "C" void test_agg_init() { extern "C" void test_agg_init() {
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-BRACE // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-BRACE
// //
// CHECK-AGG-BRACE-DAG: @[[FILE:.*]] = {{.*}}c"BraceInitEnd.cpp\00" // CHECK-AGG-BRACE-DAG: @[[FILE:.*]] = {{.*}}c"BraceInitEnd.cpp\00"
// CHECK-AGG-BRACE-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00" // CHECK-AGG-BRACE-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00"
// //
// CHECK-AGG-BRACE: define void @test_agg_init() // CHECK-AGG-BRACE: define void @test_agg_init()
// CHECK-AGG-BRACE: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_brace_init, i32 0, i32 1 // CHECK-AGG-BRACE: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_brace_init, i32 0, i32 1
// CHECK-AGG-BRACE-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[I2]], // CHECK-AGG-BRACE-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[I2]],
// CHECK-AGG-BRACE-SAME: i32 5700, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]] // CHECK-AGG-BRACE-SAME: i32 5700, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]]
#line 5600 "BraceInitStart.cpp" #line 5600 "BraceInitStart.cpp"
TestInitAgg local_brace_init{ TestInitAgg local_brace_init{
#line 5700 "BraceInitEnd.cpp" #line 5700 "BraceInitEnd.cpp"
}; };
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-EQUAL // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-EQUAL
// //
// CHECK-AGG-EQUAL-DAG: @[[FILE:.*]] = {{.*}}c"EqualInitEnd.cpp\00" // CHECK-AGG-EQUAL-DAG: @[[FILE:.*]] = {{.*}}c"EqualInitEnd.cpp\00"
// CHECK-AGG-EQUAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00" // CHECK-AGG-EQUAL-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00"
// //
// CHECK-AGG-EQUAL: define void @test_agg_init() // CHECK-AGG-EQUAL: define void @test_agg_init()
// CHECK-AGG-EQUAL: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_equal_init, i32 0, i32 1 // CHECK-AGG-EQUAL: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_equal_init, i32 0, i32 1
// CHECK-AGG-EQUAL-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[I2]], // CHECK-AGG-EQUAL-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[I2]],
// CHECK-AGG-EQUAL-SAME: i32 5900, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]] // CHECK-AGG-EQUAL-SAME: i32 5900, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]]
#line 5800 "EqualInitStart.cpp" #line 5800 "EqualInitStart.cpp"
TestInitAgg local_equal_init = TestInitAgg local_equal_init =
{ {
#line 5900 "EqualInitEnd.cpp" #line 5900 "EqualInitEnd.cpp"
}; };
// RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-LIST // RUN: FileCheck --input-file %t.ll %s --check-prefix=CHECK-AGG-LIST
// //
// CHECK-AGG-LIST-DAG: @[[FILE_DEFAULT:.*]] = {{.*}}c"InitListEnd.cpp\00" // CHECK-AGG-LIST-DAG: @[[FILE_DEFAULT:.*]] = {{.*}}c"InitListEnd.cpp\00"
// CHECK-AGG-LIST-DAG: @[[FILE_ELEM:.*]] = {{.*}}c"ListElem.cpp\00" // CHECK-AGG-LIST-DAG: @[[FILE_ELEM:.*]] = {{.*}}c"ListElem.cpp\00"
// CHECK-AGG-LIST-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00" // CHECK-AGG-LIST-DAG: @[[FUNC:.*]] = {{.*}}c"test_agg_init\00"
// //
// CHECK-AGG-LIST: define void @test_agg_init() // CHECK-AGG-LIST: define void @test_agg_init()
// //
// CHECK-AGG-LIST: %[[I1:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_list_init, i32 0, i32 0 // CHECK-AGG-LIST: %[[I1:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_list_init, i32 0, i32 0
// CHECK-AGG-LIST-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[I1]], // CHECK-AGG-LIST-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[I1]],
// CHECK-AGG-LIST-SAME: i32 6100, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE_ELEM]], {{[^@]*}}@[[FUNC]] // CHECK-AGG-LIST-SAME: i32 6100, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE_ELEM]], {{[^@]*}}@[[FUNC]]
// //
// CHECK-AGG-LIST: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_list_init, i32 0, i32 1 // CHECK-AGG-LIST: %[[I2:.*]] = getelementptr inbounds %struct.TestInitAgg, %struct.TestInitAgg* %local_list_init, i32 0, i32 1
// CHECK-AGG-LIST-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[I2]], // CHECK-AGG-LIST-NEXT: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[I2]],
// CHECK-AGG-LIST-SAME: i32 6200, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE_DEFAULT]], {{[^@]*}}@[[FUNC]] // CHECK-AGG-LIST-SAME: i32 6200, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE_DEFAULT]], {{[^@]*}}@[[FUNC]]
#line 6000 "InitListStart.cpp" #line 6000 "InitListStart.cpp"
TestInitAgg local_list_init = TestInitAgg local_list_init =
{ {
#line 6100 "ListElem.cpp" #line 6100 "ListElem.cpp"
{SL::current()} {SL::current()}
#line 6200 "InitListEnd.cpp" #line 6200 "InitListEnd.cpp"
}; };
Show All 17 Lines
// //
// CHECK-TEMPL-DAG: @[[FILE:.*]] = {{.*}}c"local_templ.cpp\00" // CHECK-TEMPL-DAG: @[[FILE:.*]] = {{.*}}c"local_templ.cpp\00"
// CHECK-TEMPL-DAG: @[[FUNC:.*]] = {{.*}}c"test_template\00" // CHECK-TEMPL-DAG: @[[FUNC:.*]] = {{.*}}c"test_template\00"
// //
// CHECK-TEMPL: define weak_odr void @_Z13test_templateI15source_locationLi[[INT_ID]]EEvv() // CHECK-TEMPL: define weak_odr void @_Z13test_templateI15source_locationLi[[INT_ID]]EEvv()
// CHECK-TEMPL-NEXT: entry: // CHECK-TEMPL-NEXT: entry:
// CHECK-TEMPL-NOT: ret // CHECK-TEMPL-NOT: ret
// //
// CHECK-TEMPL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret %[[TMP:[^,]*]], // CHECK-TEMPL: call void @_ZN15source_location7currentEjjPKcS1_(%struct.source_location* sret align 8 %[[TMP:[^,]*]],
// CHECK-TEMPL-SAME: i32 7300, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]] // CHECK-TEMPL-SAME: i32 7300, i32 {{[0-9]+}}, {{[^@]*}}@[[FILE]], {{[^@]*}}@[[FUNC]]
#line 7300 "local_templ.cpp" #line 7300 "local_templ.cpp"
TestTemplate<T, V> local_templ; TestTemplate<T, V> local_templ;
} }
#line 7400 "EndTestTemplate.cpp" #line 7400 "EndTestTemplate.cpp"
template void test_template<SL, 0>(); template void test_template<SL, 0>();
template void test_template<SL, 1>(); template void test_template<SL, 1>();

clang/test/CodeGenCXX/call-with-static-chain.cpp

Show All 19 Lines
void test() { void test() {
A a; A a;
// CHECK32: call i32 bitcast (i32 (i32, i32, i32, i32, i32, i32, i32, i32)* @f1 to i32 (i8*, i32, i32, i32, i32, i32, i32, i32, i32)*)(i8* nest bitcast (i32 (i32, i32, i32, i32, i32, i32, i32, i32)* @f1 to i8*) // CHECK32: call i32 bitcast (i32 (i32, i32, i32, i32, i32, i32, i32, i32)* @f1 to i32 (i8*, i32, i32, i32, i32, i32, i32, i32, i32)*)(i8* nest bitcast (i32 (i32, i32, i32, i32, i32, i32, i32, i32)* @f1 to i8*)
// CHECK64: call i32 bitcast (i32 (i64, i64, i64, i64, i64, i64, %struct.A*)* @f1 to i32 (i8*, i64, i64, i64, i64, i64, i64, %struct.A*)*)(i8* nest bitcast (i32 (i64, i64, i64, i64, i64, i64, %struct.A*)* @f1 to i8*) // CHECK64: call i32 bitcast (i32 (i64, i64, i64, i64, i64, i64, %struct.A*)* @f1 to i32 (i8*, i64, i64, i64, i64, i64, i64, %struct.A*)*)(i8* nest bitcast (i32 (i64, i64, i64, i64, i64, i64, %struct.A*)* @f1 to i8*)
__builtin_call_with_static_chain(f1(a, a, a, a), f1); __builtin_call_with_static_chain(f1(a, a, a, a), f1);
// CHECK32: call void bitcast (void (%struct.B*)* @f2 to void (%struct.B*, i8*)*)(%struct.B* sret %{{[0-9a-z]+}}, i8* nest bitcast (void (%struct.B*)* @f2 to i8*)) // CHECK32: call void bitcast (void (%struct.B*)* @f2 to void (%struct.B*, i8*)*)(%struct.B* sret align 4 %{{[0-9a-z]+}}, i8* nest bitcast (void (%struct.B*)* @f2 to i8*))
// CHECK64: call void bitcast (void (%struct.B*)* @f2 to void (%struct.B*, i8*)*)(%struct.B* sret %{{[0-9a-z]+}}, i8* nest bitcast (void (%struct.B*)* @f2 to i8*)) // CHECK64: call void bitcast (void (%struct.B*)* @f2 to void (%struct.B*, i8*)*)(%struct.B* sret align 8 %{{[0-9a-z]+}}, i8* nest bitcast (void (%struct.B*)* @f2 to i8*))
__builtin_call_with_static_chain(f2(), f2); __builtin_call_with_static_chain(f2(), f2);
// CHECK32: call i64 bitcast (i64 ()* @f3 to i64 (i8*)*)(i8* nest bitcast (i64 ()* @f3 to i8*)) // CHECK32: call i64 bitcast (i64 ()* @f3 to i64 (i8*)*)(i8* nest bitcast (i64 ()* @f3 to i8*))
// CHECK64: call <2 x float> bitcast (<2 x float> ()* @f3 to <2 x float> (i8*)*)(i8* nest bitcast (<2 x float> ()* @f3 to i8*)) // CHECK64: call <2 x float> bitcast (<2 x float> ()* @f3 to <2 x float> (i8*)*)(i8* nest bitcast (<2 x float> ()* @f3 to i8*))
__builtin_call_with_static_chain(f3(), f3); __builtin_call_with_static_chain(f3(), f3);
// CHECK32: call dereferenceable(8) %struct.A* bitcast (%struct.A* ()* @f4 to %struct.A* (i8*)*)(i8* nest bitcast (%struct.A* ()* @f4 to i8*)) // CHECK32: call dereferenceable(8) %struct.A* bitcast (%struct.A* ()* @f4 to %struct.A* (i8*)*)(i8* nest bitcast (%struct.A* ()* @f4 to i8*))
// CHECK64: call dereferenceable(16) %struct.A* bitcast (%struct.A* ()* @f4 to %struct.A* (i8*)*)(i8* nest bitcast (%struct.A* ()* @f4 to i8*)) // CHECK64: call dereferenceable(16) %struct.A* bitcast (%struct.A* ()* @f4 to %struct.A* (i8*)*)(i8* nest bitcast (%struct.A* ()* @f4 to i8*))
__builtin_call_with_static_chain(f4(), f4); __builtin_call_with_static_chain(f4(), f4);
} }

clang/test/CodeGenCXX/conditional-gnu-ext.cpp

Show First 20 Lines • Show All 88 Lines • ▼ Show 20 Lines B test0(B &x) {
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: ret void // CHECK: ret void
return x ?: B(); return x ?: B();
} }
B test1() { B test1() {
// CHECK-LABEL: define void @_ZN5test35test1Ev( // CHECK-LABEL: define void @_ZN5test35test1Ev(
// CHECK: [[TEMP:%.*]] = alloca [[B]], // CHECK: [[TEMP:%.*]] = alloca [[B]],
// CHECK: call void @_ZN5test312test1_helperEv([[B]]* sret [[TEMP]]) // CHECK: call void @_ZN5test312test1_helperEv([[B]]* sret align 1 [[TEMP]])
// CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[TEMP]]) // CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[TEMP]])
// CHECK-NEXT: br i1 [[BOOL]] // CHECK-NEXT: br i1 [[BOOL]]
// CHECK: call void @_ZN5test31BC1ERKS0_([[B]]* [[RESULT:%.*]], [[B]]* dereferenceable({{[0-9]+}}) [[TEMP]]) // CHECK: call void @_ZN5test31BC1ERKS0_([[B]]* [[RESULT:%.*]], [[B]]* dereferenceable({{[0-9]+}}) [[TEMP]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: call void @_ZN5test31BC1Ev([[B]]* [[RESULT]]) // CHECK: call void @_ZN5test31BC1Ev([[B]]* [[RESULT]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: call void @_ZN5test31BD1Ev([[B]]* [[TEMP]]) // CHECK: call void @_ZN5test31BD1Ev([[B]]* [[TEMP]])
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
extern B test1_helper(); extern B test1_helper();
return test1_helper() ?: B(); return test1_helper() ?: B();
} }
A test2(B &x) { A test2(B &x) {
// CHECK-LABEL: define void @_ZN5test35test2ERNS_1BE( // CHECK-LABEL: define void @_ZN5test35test2ERNS_1BE(
// CHECK: [[X:%.*]] = alloca [[B]]*, // CHECK: [[X:%.*]] = alloca [[B]]*,
// CHECK: store [[B]]* {{%.*}}, [[B]]** [[X]] // CHECK: store [[B]]* {{%.*}}, [[B]]** [[X]]
// CHECK-NEXT: [[T0:%.*]] = load [[B]]*, [[B]]** [[X]] // CHECK-NEXT: [[T0:%.*]] = load [[B]]*, [[B]]** [[X]]
// CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[T0]]) // CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[T0]])
// CHECK-NEXT: br i1 [[BOOL]] // CHECK-NEXT: br i1 [[BOOL]]
// CHECK: call void @_ZN5test31BcvNS_1AEEv([[A:%.*]]* sret [[RESULT:%.*]], [[B]]* [[T0]]) // CHECK: call void @_ZN5test31BcvNS_1AEEv([[A:%.*]]* sret align 1 [[RESULT:%.*]], [[B]]* [[T0]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: call void @_ZN5test31AC1Ev([[A]]* [[RESULT]]) // CHECK: call void @_ZN5test31AC1Ev([[A]]* [[RESULT]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: ret void // CHECK: ret void
return x ?: A(); return x ?: A();
} }
A test3() { A test3() {
// CHECK-LABEL: define void @_ZN5test35test3Ev( // CHECK-LABEL: define void @_ZN5test35test3Ev(
// CHECK: [[TEMP:%.*]] = alloca [[B]], // CHECK: [[TEMP:%.*]] = alloca [[B]],
// CHECK: call void @_ZN5test312test3_helperEv([[B]]* sret [[TEMP]]) // CHECK: call void @_ZN5test312test3_helperEv([[B]]* sret align 1 [[TEMP]])
// CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[TEMP]]) // CHECK-NEXT: [[BOOL:%.*]] = call zeroext i1 @_ZN5test31BcvbEv([[B]]* [[TEMP]])
// CHECK-NEXT: br i1 [[BOOL]] // CHECK-NEXT: br i1 [[BOOL]]
// CHECK: call void @_ZN5test31BcvNS_1AEEv([[A]]* sret [[RESULT:%.*]], [[B]]* [[TEMP]]) // CHECK: call void @_ZN5test31BcvNS_1AEEv([[A]]* sret align 1 [[RESULT:%.*]], [[B]]* [[TEMP]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: call void @_ZN5test31AC1Ev([[A]]* [[RESULT]]) // CHECK: call void @_ZN5test31AC1Ev([[A]]* [[RESULT]])
// CHECK-NEXT: br label // CHECK-NEXT: br label
// CHECK: call void @_ZN5test31BD1Ev([[B]]* [[TEMP]]) // CHECK: call void @_ZN5test31BD1Ev([[B]]* [[TEMP]])
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
extern B test3_helper(); extern B test3_helper();
return test3_helper() ?: A(); return test3_helper() ?: A();
} }
Show All 10 Lines

clang/test/CodeGenCXX/cxx1z-copy-omission.cpp

Show All 13 Lines
A f(); A f();
void h(); void h();
// CHECK-LABEL: define {{.*}} @_Z1gv( // CHECK-LABEL: define {{.*}} @_Z1gv(
void g() { void g() {
// CHECK: %[[A:.*]] = alloca // CHECK: %[[A:.*]] = alloca
// CHECK-NOT: alloca // CHECK-NOT: alloca
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call {{.*}} @_Z1fv({{.*}}* sret %[[A]]) // CHECK: call {{.*}} @_Z1fv({{.*}}* sret align 4 %[[A]])
A a = A( A{ f() } ); A a = A( A{ f() } );
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call void @_Z1hv( // CHECK: call void @_Z1hv(
h(); h();
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call void @_ZN1AD1Ev({{.*}}* %[[A]]) // CHECK: call void @_ZN1AD1Ev({{.*}}* %[[A]])
// CHECK-NOT: call // CHECK-NOT: call
// CHECK-LABEL: } // CHECK-LABEL: }
} }
void f(A); void f(A);
// CHECK-LABEL: define {{.*}} @_Z1hv( // CHECK-LABEL: define {{.*}} @_Z1hv(
void h() { void h() {
// CHECK: %[[A:.*]] = alloca // CHECK: %[[A:.*]] = alloca
// CHECK-NOT: alloca // CHECK-NOT: alloca
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call {{.*}} @_Z1fv({{.*}}* sret %[[A]]) // CHECK: call {{.*}} @_Z1fv({{.*}}* sret align 4 %[[A]])
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call {{.*}} @_Z1f1A({{.*}}* %[[A]]) // CHECK: call {{.*}} @_Z1f1A({{.*}}* %[[A]])
f(f()); f(f());
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: call void @_ZN1AD1Ev({{.*}}* %[[A]]) // CHECK: call void @_ZN1AD1Ev({{.*}}* %[[A]])
// CHECK: call void @_Z1hv( // CHECK: call void @_Z1hv(
h(); h();
▲ Show 20 Lines • Show All 56 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/cxx1z-lambda-star-this.cpp

// RUN: %clang_cc1 -std=c++1y -triple i686-pc-windows-msvc -emit-llvm %s -o - | FileCheck %s // RUN: %clang_cc1 -std=c++1y -triple i686-pc-windows-msvc -emit-llvm %s -o - | FileCheck %s
//CHECK: %[[A_LAMBDA:.*]] = type { %struct.A } //CHECK: %[[A_LAMBDA:.*]] = type { %struct.A }
//CHECK: %[[B_LAMBDA:.*]] = type { %struct.B* } //CHECK: %[[B_LAMBDA:.*]] = type { %struct.B* }
struct A { struct A {
double a = 111; double a = 111;
auto foo() { return [*this] { return a; }; } auto foo() { return [*this] { return a; }; }
}; };
namespace ns1 { namespace ns1 {
int X = A{}.foo()(); int X = A{}.foo()();
} //end ns1 } //end ns1
//CHECK: @"?foo@A@@QAE?A?<auto>@@XZ"(%struct.A* %this, %class.anon* noalias sret %[[A_LAMBDA_RETVAL:.*]]) //CHECK: @"?foo@A@@QAE?A?<auto>@@XZ"(%struct.A* %this, %class.anon* noalias sret align 8 %[[A_LAMBDA_RETVAL:.*]])
// get the first object with the closure type, which is of type 'struct.A' // get the first object with the closure type, which is of type 'struct.A'
//CHECK: %[[I0:.+]] = getelementptr inbounds %[[A_LAMBDA]], %[[A_LAMBDA]]* %[[A_LAMBDA_RETVAL]], i32 0, i32 0 //CHECK: %[[I0:.+]] = getelementptr inbounds %[[A_LAMBDA]], %[[A_LAMBDA]]* %[[A_LAMBDA_RETVAL]], i32 0, i32 0
//CHECK: %[[I1:.+]] = bitcast %struct.A* %[[I0]] to i8* //CHECK: %[[I1:.+]] = bitcast %struct.A* %[[I0]] to i8*
//CHECK: %[[I2:.+]] = bitcast %struct.A* %this1 to i8* //CHECK: %[[I2:.+]] = bitcast %struct.A* %this1 to i8*
// copy the contents ... // copy the contents ...
//CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 %[[I1]], i8* align 8 %[[I2]], i32 8, i1 false) //CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 %[[I1]], i8* align 8 %[[I2]], i32 8, i1 false)
struct B { struct B {
double b = 222; double b = 222;
auto bar() { return [this] { return b; }; }; auto bar() { return [this] { return b; }; };
}; };
namespace ns2 { namespace ns2 {
int X = B{}.bar()(); int X = B{}.bar()();
} }
//CHECK: @"?bar@B@@QAE?A?<auto>@@XZ"(%struct.B* %this, %class.anon.0* noalias sret %agg.result) //CHECK: @"?bar@B@@QAE?A?<auto>@@XZ"(%struct.B* %this, %class.anon.0* noalias sret align 4 %agg.result)
//CHECK: %[[I20:.+]] = getelementptr inbounds %class.anon.0, %class.anon.0* %agg.result, i32 0, i32 0 //CHECK: %[[I20:.+]] = getelementptr inbounds %class.anon.0, %class.anon.0* %agg.result, i32 0, i32 0
//CHECK: store %struct.B* %this1, %struct.B** %[[I20]], align 4 //CHECK: store %struct.B* %this1, %struct.B** %[[I20]], align 4

clang/test/CodeGenCXX/exceptions.cpp

Show First 20 Lines • Show All 140 Lines • ▼ Show 20 Linesnamespace test1 {
A *i() { A *i() {
// CHECK: define [[A:%.*]]* @_ZN5test11iEv() // CHECK: define [[A:%.*]]* @_ZN5test11iEv()
// CHECK: [[X:%.*]] = alloca [[A]]*, align 8 // CHECK: [[X:%.*]] = alloca [[A]]*, align 8
// CHECK: [[ACTIVE:%.*]] = alloca i1 // CHECK: [[ACTIVE:%.*]] = alloca i1
// CHECK: [[NEW:%.*]] = call noalias nonnull i8* @_Znwm(i64 8) // CHECK: [[NEW:%.*]] = call noalias nonnull i8* @_Znwm(i64 8)
// CHECK-NEXT: store i1 true, i1* [[ACTIVE]] // CHECK-NEXT: store i1 true, i1* [[ACTIVE]]
// CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[NEW]] to [[A]]* // CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[NEW]] to [[A]]*
// CHECK-NEXT: invoke void @_ZN5test15makeBEv([[B:%.*]]* sret [[T0:%.*]]) // CHECK-NEXT: invoke void @_ZN5test15makeBEv([[B:%.*]]* sret align 4 [[T0:%.*]])
// CHECK: [[T1:%.*]] = invoke i32 @_ZN5test11BcviEv([[B]]* [[T0]]) // CHECK: [[T1:%.*]] = invoke i32 @_ZN5test11BcviEv([[B]]* [[T0]])
// CHECK: invoke void @_ZN5test11AC1Ei([[A]]* [[CAST]], i32 [[T1]]) // CHECK: invoke void @_ZN5test11AC1Ei([[A]]* [[CAST]], i32 [[T1]])
// CHECK: store i1 false, i1* [[ACTIVE]] // CHECK: store i1 false, i1* [[ACTIVE]]
// CHECK-NEXT: store [[A]]* [[CAST]], [[A]]** [[X]], align 8 // CHECK-NEXT: store [[A]]* [[CAST]], [[A]]** [[X]], align 8
// CHECK: invoke void @_ZN5test15makeBEv([[B:%.*]]* sret [[T2:%.*]]) // CHECK: invoke void @_ZN5test15makeBEv([[B:%.*]]* sret align 4 [[T2:%.*]])
// CHECK: [[RET:%.*]] = load [[A]]*, [[A]]** [[X]], align 8 // CHECK: [[RET:%.*]] = load [[A]]*, [[A]]** [[X]], align 8
// CHECK98: invoke void @_ZN5test11BD1Ev([[B]]* [[T2]]) // CHECK98: invoke void @_ZN5test11BD1Ev([[B]]* [[T2]])
// CHECK11: call void @_ZN5test11BD1Ev([[B]]* [[T2]]) // CHECK11: call void @_ZN5test11BD1Ev([[B]]* [[T2]])
// CHECK98: invoke void @_ZN5test11BD1Ev([[B]]* [[T0]]) // CHECK98: invoke void @_ZN5test11BD1Ev([[B]]* [[T0]])
// CHECK11: call void @_ZN5test11BD1Ev([[B]]* [[T0]]) // CHECK11: call void @_ZN5test11BD1Ev([[B]]* [[T0]])
▲ Show 20 Lines • Show All 71 Lines • ▼ Show 20 Lines A *b(bool cond) {
return (cond ? return (cond ?
// CHECK: [[FOO:%.*]] = call i8* @_ZN5test33fooEv() // CHECK: [[FOO:%.*]] = call i8* @_ZN5test33fooEv()
// CHECK-NEXT: [[NEW:%.*]] = call i8* @_ZN5test31AnwEmPvd(i64 8, i8* [[FOO]], double [[CONST:.*]]) // CHECK-NEXT: [[NEW:%.*]] = call i8* @_ZN5test31AnwEmPvd(i64 8, i8* [[FOO]], double [[CONST:.*]])
// CHECK-NEXT: store i8* [[NEW]], i8** [[SAVED0]] // CHECK-NEXT: store i8* [[NEW]], i8** [[SAVED0]]
// CHECK-NEXT: store i8* [[FOO]], i8** [[SAVED1]] // CHECK-NEXT: store i8* [[FOO]], i8** [[SAVED1]]
// CHECK-NEXT: store i1 true, i1* [[CLEANUPACTIVE]] // CHECK-NEXT: store i1 true, i1* [[CLEANUPACTIVE]]
// CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[NEW]] to [[A]]* // CHECK-NEXT: [[CAST:%.*]] = bitcast i8* [[NEW]] to [[A]]*
// CHECK-NEXT: invoke void @_ZN5test35makeAEv([[A]]* sret [[CAST]]) // CHECK-NEXT: invoke void @_ZN5test35makeAEv([[A]]* sret align 8 [[CAST]])
// CHECK: br label // CHECK: br label
// -> cond.end // -> cond.end
new(foo(),10.0) A(makeA()) : new(foo(),10.0) A(makeA()) :
// CHECK: [[MAKE:%.*]] = call [[A]]* @_ZN5test38makeAPtrEv() // CHECK: [[MAKE:%.*]] = call [[A]]* @_ZN5test38makeAPtrEv()
// CHECK: br label // CHECK: br label
// -> cond.end // -> cond.end
makeAPtr()); makeAPtr());
▲ Show 20 Lines • Show All 347 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/homogeneous-aggregates.cpp

Show All 32 Linesstruct D4 : Base2, Base3 { // homogeneous aggregate
double y, z; double y, z;
}; };
struct I1 : Base2 {}; struct I1 : Base2 {};
struct I2 : Base2 {}; struct I2 : Base2 {};
struct I3 : Base2 {}; struct I3 : Base2 {};
struct D5 : I1, I2, I3 {}; // homogeneous aggregate struct D5 : I1, I2, I3 {}; // homogeneous aggregate
// PPC: define void @_Z7func_D12D1(%struct.D1* noalias sret %agg.result, [3 x i64] %x.coerce) // PPC: define void @_Z7func_D12D1(%struct.D1* noalias sret align 8 %agg.result, [3 x i64] %x.coerce)
// ARM32: define arm_aapcs_vfpcc void @_Z7func_D12D1(%struct.D1* noalias sret %agg.result, [3 x i64] %x.coerce) // ARM32: define arm_aapcs_vfpcc void @_Z7func_D12D1(%struct.D1* noalias sret align 8 %agg.result, [3 x i64] %x.coerce)
// ARM64: define void @_Z7func_D12D1(%struct.D1* noalias sret %agg.result, %struct.D1* %x) // ARM64: define void @_Z7func_D12D1(%struct.D1* noalias sret align 8 %agg.result, %struct.D1* %x)
// X64: define dso_local x86_vectorcallcc void @"\01_Z7func_D12D1@@24"(%struct.D1* noalias sret %agg.result, %struct.D1* %x) // X64: define dso_local x86_vectorcallcc void @"\01_Z7func_D12D1@@24"(%struct.D1* noalias sret align 8 %agg.result, %struct.D1* %x)
D1 CC func_D1(D1 x) { return x; } D1 CC func_D1(D1 x) { return x; }
// PPC: define [3 x double] @_Z7func_D22D2([3 x double] %x.coerce) // PPC: define [3 x double] @_Z7func_D22D2([3 x double] %x.coerce)
// ARM32: define arm_aapcs_vfpcc %struct.D2 @_Z7func_D22D2(%struct.D2 %x.coerce) // ARM32: define arm_aapcs_vfpcc %struct.D2 @_Z7func_D22D2(%struct.D2 %x.coerce)
// ARM64: define %struct.D2 @_Z7func_D22D2([3 x double] %x.coerce) // ARM64: define %struct.D2 @_Z7func_D22D2([3 x double] %x.coerce)
// X64: define dso_local x86_vectorcallcc %struct.D2 @"\01_Z7func_D22D2@@24"(%struct.D2 inreg %x.coerce) // X64: define dso_local x86_vectorcallcc %struct.D2 @"\01_Z7func_D22D2@@24"(%struct.D2 inreg %x.coerce)
D2 CC func_D2(D2 x) { return x; } D2 CC func_D2(D2 x) { return x; }
// PPC: define void @_Z7func_D32D3(%struct.D3* noalias sret %agg.result, [4 x i64] %x.coerce) // PPC: define void @_Z7func_D32D3(%struct.D3* noalias sret align 8 %agg.result, [4 x i64] %x.coerce)
// ARM32: define arm_aapcs_vfpcc void @_Z7func_D32D3(%struct.D3* noalias sret %agg.result, [4 x i64] %x.coerce) // ARM32: define arm_aapcs_vfpcc void @_Z7func_D32D3(%struct.D3* noalias sret align 8 %agg.result, [4 x i64] %x.coerce)
// ARM64: define void @_Z7func_D32D3(%struct.D3* noalias sret %agg.result, %struct.D3* %x) // ARM64: define void @_Z7func_D32D3(%struct.D3* noalias sret align 8 %agg.result, %struct.D3* %x)
D3 CC func_D3(D3 x) { return x; } D3 CC func_D3(D3 x) { return x; }
// PPC: define [4 x double] @_Z7func_D42D4([4 x double] %x.coerce) // PPC: define [4 x double] @_Z7func_D42D4([4 x double] %x.coerce)
// ARM32: define arm_aapcs_vfpcc %struct.D4 @_Z7func_D42D4(%struct.D4 %x.coerce) // ARM32: define arm_aapcs_vfpcc %struct.D4 @_Z7func_D42D4(%struct.D4 %x.coerce)
// ARM64: define %struct.D4 @_Z7func_D42D4([4 x double] %x.coerce) // ARM64: define %struct.D4 @_Z7func_D42D4([4 x double] %x.coerce)
D4 CC func_D4(D4 x) { return x; } D4 CC func_D4(D4 x) { return x; }
D5 CC func_D5(D5 x) { return x; } D5 CC func_D5(D5 x) { return x; }
▲ Show 20 Lines • Show All 43 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/lambda-expressions.cpp

Show First 20 Lines • Show All 188 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define internal i32 @"_ZZ1fvEN3$_68__invokeEii" // CHECK-LABEL: define internal i32 @"_ZZ1fvEN3$_68__invokeEii"
// CHECK: store i32 // CHECK: store i32
// CHECK-NEXT: store i32 // CHECK-NEXT: store i32
// CHECK-NEXT: load i32, i32* // CHECK-NEXT: load i32, i32*
// CHECK-NEXT: load i32, i32* // CHECK-NEXT: load i32, i32*
// CHECK-NEXT: call i32 @"_ZZ1fvENK3$_6clEii" // CHECK-NEXT: call i32 @"_ZZ1fvENK3$_6clEii"
// CHECK-NEXT: ret i32 // CHECK-NEXT: ret i32
// CHECK-LABEL: define internal void @"_ZZ1hvEN4$_118__invokeEv"(%struct.A* noalias sret %agg.result) {{.*}} { // CHECK-LABEL: define internal void @"_ZZ1hvEN4$_118__invokeEv"(%struct.A* noalias sret align 1 %agg.result) {{.*}} {
// CHECK: call void @"_ZZ1hvENK4$_11clEv"(%struct.A* sret %agg.result, // CHECK: call void @"_ZZ1hvENK4$_11clEv"(%struct.A* sret align 1 %agg.result,
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
struct A { ~A(); }; struct A { ~A(); };
void h() { void h() {
A (*h)() = [] { return A(); }; A (*h)() = [] { return A(); };
} }
// <rdar://problem/12778708> // <rdar://problem/12778708>
struct XXX {}; struct XXX {};
Show All 17 Lines

clang/test/CodeGenCXX/microsoft-abi-byval-sret.cpp

Show First 20 Lines • Show All 43 Lines • ▼ Show 20 Lines
// CHECK: load %struct.A*, %struct.A** // CHECK: load %struct.A*, %struct.A**
// CHECK: ret %struct.A* // CHECK: ret %struct.A*
A B::qux(A x) { A B::qux(A x) {
return x; return x;
} }
// CHECK-LABEL: define dso_local x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z" // CHECK-LABEL: define dso_local x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z"
// CHECK: (%struct.B* inreg %this, %struct.A* inreg noalias sret %agg.result, <{ %struct.A }>* inalloca %0) // CHECK: (%struct.B* inreg %this, %struct.A* inreg noalias sret align 4 %agg.result, <{ %struct.A }>* inalloca %0)
// CHECK: ret void // CHECK: ret void
int main() { int main() {
B b; B b;
A a = b.foo(A()); A a = b.foo(A());
a = b.bar(a); a = b.bar(a);
a = b.baz(a); a = b.baz(a);
a = b.qux(a); a = b.qux(a);
} }
// CHECK: call x86_thiscallcc %struct.A* @"?foo@B@@QAE?AUA@@U2@@Z" // CHECK: call x86_thiscallcc %struct.A* @"?foo@B@@QAE?AUA@@U2@@Z"
// CHECK: (%struct.B* %{{[^,]*}}, <{ %struct.A*, %struct.A }>* inalloca %{{[^,]*}}) // CHECK: (%struct.B* %{{[^,]*}}, <{ %struct.A*, %struct.A }>* inalloca %{{[^,]*}})
// CHECK: call %struct.A* @"?bar@B@@QAA?AUA@@U2@@Z" // CHECK: call %struct.A* @"?bar@B@@QAA?AUA@@U2@@Z"
// CHECK: (<{ %struct.B*, %struct.A*, %struct.A }>* inalloca %{{[^,]*}}) // CHECK: (<{ %struct.B*, %struct.A*, %struct.A }>* inalloca %{{[^,]*}})
// CHECK: call x86_stdcallcc %struct.A* @"?baz@B@@QAG?AUA@@U2@@Z" // CHECK: call x86_stdcallcc %struct.A* @"?baz@B@@QAG?AUA@@U2@@Z"
// CHECK: (<{ %struct.B*, %struct.A*, %struct.A }>* inalloca %{{[^,]*}}) // CHECK: (<{ %struct.B*, %struct.A*, %struct.A }>* inalloca %{{[^,]*}})
// CHECK: call x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z" // CHECK: call x86_fastcallcc void @"?qux@B@@QAI?AUA@@U2@@Z"
// CHECK: (%struct.B* inreg %{{[^,]*}}, %struct.A* inreg sret %{{.*}}, <{ %struct.A }>* inalloca %{{[^,]*}}) // CHECK: (%struct.B* inreg %{{[^,]*}}, %struct.A* inreg sret align 4 %{{.*}}, <{ %struct.A }>* inalloca %{{[^,]*}})

clang/test/CodeGenCXX/microsoft-abi-byval-thunks.cpp

Show First 20 Lines • Show All 80 Lines • ▼ Show 20 Lines
// CHECK32: load %"struct.sret_thunk::C"*, %"struct.sret_thunk::C"** %[[this_slot]] // CHECK32: load %"struct.sret_thunk::C"*, %"struct.sret_thunk::C"** %[[this_slot]]
// CHECK32: getelementptr i8, i8* %{{.*}}, i32 -4 // CHECK32: getelementptr i8, i8* %{{.*}}, i32 -4
// CHECK32: store %"struct.sret_thunk::C"* %{{.*}}, %"struct.sret_thunk::C"** %[[this_slot]] // CHECK32: store %"struct.sret_thunk::C"* %{{.*}}, %"struct.sret_thunk::C"** %[[this_slot]]
// CHECK32: %[[rv:[^ ]*]] = musttail call %"struct.sret_thunk::Agg"* @"?foo@C@sret_thunk@@UAA?AUAgg@2@U32@@Z" // CHECK32: %[[rv:[^ ]*]] = musttail call %"struct.sret_thunk::Agg"* @"?foo@C@sret_thunk@@UAA?AUAgg@2@U32@@Z"
// CHECK32: (<{ %"struct.sret_thunk::C"*, %"struct.sret_thunk::Agg"*, %"struct.sret_thunk::Agg" }>* inalloca %0) // CHECK32: (<{ %"struct.sret_thunk::C"*, %"struct.sret_thunk::Agg"*, %"struct.sret_thunk::Agg" }>* inalloca %0)
// CHECK32-NEXT: ret %"struct.sret_thunk::Agg"* %[[rv]] // CHECK32-NEXT: ret %"struct.sret_thunk::Agg"* %[[rv]]
// CHECK64-LABEL: define linkonce_odr dso_local void @"?foo@C@sret_thunk@@W7EAA?AUAgg@2@U32@@Z" // CHECK64-LABEL: define linkonce_odr dso_local void @"?foo@C@sret_thunk@@W7EAA?AUAgg@2@U32@@Z"
// CHECK64: (%"struct.sret_thunk::C"* %this, %"struct.sret_thunk::Agg"* noalias sret %agg.result, %"struct.sret_thunk::Agg"* %x) // CHECK64: (%"struct.sret_thunk::C"* %this, %"struct.sret_thunk::Agg"* noalias sret align 4 %agg.result, %"struct.sret_thunk::Agg"* %x)
// CHECK64: getelementptr i8, i8* %{{.*}}, i32 -8 // CHECK64: getelementptr i8, i8* %{{.*}}, i32 -8
// CHECK64: call void @"?foo@C@sret_thunk@@UEAA?AUAgg@2@U32@@Z" // CHECK64: call void @"?foo@C@sret_thunk@@UEAA?AUAgg@2@U32@@Z"
// CHECK64: (%"struct.sret_thunk::C"* %{{.*}}, %"struct.sret_thunk::Agg"* sret %agg.result, %"struct.sret_thunk::Agg"* %x) // CHECK64: (%"struct.sret_thunk::C"* %{{.*}}, %"struct.sret_thunk::Agg"* sret align 4 %agg.result, %"struct.sret_thunk::Agg"* %x)
// CHECK64-NOT: call // CHECK64-NOT: call
// CHECK64: ret void // CHECK64: ret void
} }
#if 0 #if 0
// FIXME: When we extend LLVM IR to allow forwarding of varargs through musttail // FIXME: When we extend LLVM IR to allow forwarding of varargs through musttail
// calls, use this test. // calls, use this test.
namespace variadic_thunk { namespace variadic_thunk {
Show All 13 Lines

clang/test/CodeGenCXX/microsoft-abi-cdecl-method-sret.cpp

Show All 13 Linesstruct C {
S __cdecl byval_and_sret(S a); S __cdecl byval_and_sret(S a);
int c; int c;
}; };
S C::variadic_sret(const char *f, ...) { return S(); } S C::variadic_sret(const char *f, ...) { return S(); }
S C::cdecl_sret() { return S(); } S C::cdecl_sret() { return S(); }
S C::byval_and_sret(S a) { return S(); } S C::byval_and_sret(S a) { return S(); }
// CHECK: define dso_local void @"?variadic_sret@C@@QAA?AUS@@PBDZZ"(%struct.C* %this, %struct.S* noalias sret %agg.result, i8* %f, ...) // CHECK: define dso_local void @"?variadic_sret@C@@QAA?AUS@@PBDZZ"(%struct.C* %this, %struct.S* noalias sret align 4 %agg.result, i8* %f, ...)
// CHECK: define dso_local void @"?cdecl_sret@C@@QAA?AUS@@XZ"(%struct.C* %this, %struct.S* noalias sret %agg.result) // CHECK: define dso_local void @"?cdecl_sret@C@@QAA?AUS@@XZ"(%struct.C* %this, %struct.S* noalias sret align 4 %agg.result)
// CHECK: define dso_local void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z"(%struct.C* %this, %struct.S* noalias sret %agg.result, %struct.S* byval(%struct.S) align 4 %a) // CHECK: define dso_local void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z"(%struct.C* %this, %struct.S* noalias sret align 4 %agg.result, %struct.S* byval(%struct.S) align 4 %a)
int main() { int main() {
C c; C c;
c.variadic_sret("asdf"); c.variadic_sret("asdf");
c.cdecl_sret(); c.cdecl_sret();
c.byval_and_sret(S()); c.byval_and_sret(S());
} }
// CHECK-LABEL: define dso_local i32 @main() // CHECK-LABEL: define dso_local i32 @main()
// CHECK: call void {{.*}} @"?variadic_sret@C@@QAA?AUS@@PBDZZ" // CHECK: call void {{.*}} @"?variadic_sret@C@@QAA?AUS@@PBDZZ"
// CHECK: call void @"?cdecl_sret@C@@QAA?AUS@@XZ" // CHECK: call void @"?cdecl_sret@C@@QAA?AUS@@XZ"
// CHECK: call void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z" // CHECK: call void @"?byval_and_sret@C@@QAA?AUS@@U2@@Z"
// __fastcall has similar issues. // __fastcall has similar issues.
struct A { struct A {
S __fastcall f(int x); S __fastcall f(int x);
}; };
S A::f(int x) { S A::f(int x) {
return S(); return S();
} }
// CHECK-LABEL: define dso_local x86_fastcallcc void @"?f@A@@QAI?AUS@@H@Z"(%struct.A* inreg %this, %struct.S* inreg noalias sret %agg.result, i32 %x) // CHECK-LABEL: define dso_local x86_fastcallcc void @"?f@A@@QAI?AUS@@H@Z"(%struct.A* inreg %this, %struct.S* inreg noalias sret align 4 %agg.result, i32 %x)

clang/test/CodeGenCXX/microsoft-abi-eh-cleanups.cpp

Show All 12 Lines
void HasEHCleanup() { void HasEHCleanup() {
TakesTwo(getA(), getA()); TakesTwo(getA(), getA());
} }
// With exceptions, we need to clean up at least one of these temporaries. // With exceptions, we need to clean up at least one of these temporaries.
// WIN32-LABEL: define dso_local void @"?HasEHCleanup@@YAXXZ"() {{.*}} { // WIN32-LABEL: define dso_local void @"?HasEHCleanup@@YAXXZ"() {{.*}} {
// WIN32: %[[base:.*]] = call i8* @llvm.stacksave() // WIN32: %[[base:.*]] = call i8* @llvm.stacksave()
// If this call throws, we have to restore the stack. // If this call throws, we have to restore the stack.
// WIN32: call void @"?getA@@YA?AUA@@XZ"(%struct.A* sret %{{.*}}) // WIN32: call void @"?getA@@YA?AUA@@XZ"(%struct.A* sret align 4 %{{.*}})
// If this call throws, we have to cleanup the first temporary. // If this call throws, we have to cleanup the first temporary.
// WIN32: invoke void @"?getA@@YA?AUA@@XZ"(%struct.A* sret %{{.*}}) // WIN32: invoke void @"?getA@@YA?AUA@@XZ"(%struct.A* sret align 4 %{{.*}})
// If this call throws, we have to cleanup the stacksave. // If this call throws, we have to cleanup the stacksave.
// WIN32: call i32 @"?TakesTwo@@YAHUA@@0@Z" // WIN32: call i32 @"?TakesTwo@@YAHUA@@0@Z"
// WIN32: call void @llvm.stackrestore // WIN32: call void @llvm.stackrestore
// WIN32: ret void // WIN32: ret void
// //
// There should be one dtor call for unwinding from the second getA. // There should be one dtor call for unwinding from the second getA.
// WIN32: cleanuppad // WIN32: cleanuppad
// WIN32: call x86_thiscallcc void @"??1A@@QAE@XZ"({{.*}}) // WIN32: call x86_thiscallcc void @"??1A@@QAE@XZ"({{.*}})
▲ Show 20 Lines • Show All 308 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/microsoft-abi-sret-and-byval.cpp

Show First 20 Lines • Show All 78 Lines • ▼ Show 20 Lines
// WIN32: i8, [3 x i8], i8, [3 x i8], i32, i8, [3 x i8] }>* inalloca) // WIN32: i8, [3 x i8], i8, [3 x i8], i32, i8, [3 x i8] }>* inalloca)
void take_bools_and_chars(char a, char b, SmallWithDtor c, char d, bool e, int f, bool g); void take_bools_and_chars(char a, char b, SmallWithDtor c, char d, bool e, int f, bool g);
void call_bools_and_chars() { void call_bools_and_chars() {
take_bools_and_chars('A', 'B', SmallWithDtor(), 'D', true, 13, false); take_bools_and_chars('A', 'B', SmallWithDtor(), 'D', true, 13, false);
} }
// Returning structs that fit into a register. // Returning structs that fit into a register.
Small small_return() { return Small(); } Small small_return() { return Small(); }
// LINUX-LABEL: define void @_Z12small_returnv(%struct.Small* noalias sret %agg.result) // LINUX-LABEL: define void @_Z12small_returnv(%struct.Small* noalias sret align 4 %agg.result)
// WIN32: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"() // WIN32: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
// WIN64: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"() // WIN64: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
Medium medium_return() { return Medium(); } Medium medium_return() { return Medium(); }
// LINUX-LABEL: define void @_Z13medium_returnv(%struct.Medium* noalias sret %agg.result) // LINUX-LABEL: define void @_Z13medium_returnv(%struct.Medium* noalias sret align 4 %agg.result)
// WIN32: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"() // WIN32: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
// WIN64: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"() // WIN64: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
// Returning structs that fit into a register but are not POD. // Returning structs that fit into a register but are not POD.
SmallCpp11NotCpp03Pod small_non_pod_return() { return SmallCpp11NotCpp03Pod(); } SmallCpp11NotCpp03Pod small_non_pod_return() { return SmallCpp11NotCpp03Pod(); }
// LINUX-LABEL: define void @_Z20small_non_pod_returnv(%struct.SmallCpp11NotCpp03Pod* noalias sret %agg.result) // LINUX-LABEL: define void @_Z20small_non_pod_returnv(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret %agg.result) // WIN32: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret %agg.result) // WIN64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
SmallWithCtor small_with_ctor_return() { return SmallWithCtor(); } SmallWithCtor small_with_ctor_return() { return SmallWithCtor(); }
// LINUX-LABEL: define void @_Z22small_with_ctor_returnv(%struct.SmallWithCtor* noalias sret %agg.result) // LINUX-LABEL: define void @_Z22small_with_ctor_returnv(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret %agg.result) // WIN32: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret %agg.result) // WIN64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
// FIXME: The 'sret' mark here doesn't seem to be enough to convince LLVM to // FIXME: The 'sret' mark here doesn't seem to be enough to convince LLVM to
// preserve the hidden sret pointer in R0 across the function. // preserve the hidden sret pointer in R0 across the function.
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret %agg.result) // WOA: define dso_local arm_aapcs_vfpcc void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
SmallWithVftable small_with_vftable_return() { return SmallWithVftable(); } SmallWithVftable small_with_vftable_return() { return SmallWithVftable(); }
// LINUX-LABEL: define void @_Z25small_with_vftable_returnv(%struct.SmallWithVftable* noalias sret %agg.result) // LINUX-LABEL: define void @_Z25small_with_vftable_returnv(%struct.SmallWithVftable* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret %agg.result) // WIN32: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret %agg.result) // WIN64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret align 8 %agg.result)
MediumWithCopyCtor medium_with_copy_ctor_return() { return MediumWithCopyCtor(); } MediumWithCopyCtor medium_with_copy_ctor_return() { return MediumWithCopyCtor(); }
// LINUX-LABEL: define void @_Z28medium_with_copy_ctor_returnv(%struct.MediumWithCopyCtor* noalias sret %agg.result) // LINUX-LABEL: define void @_Z28medium_with_copy_ctor_returnv(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret %agg.result) // WIN32: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret %agg.result) // WIN64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret %agg.result) // WOA: define dso_local arm_aapcs_vfpcc void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// Returning a large struct that doesn't fit into a register. // Returning a large struct that doesn't fit into a register.
Big big_return() { return Big(); } Big big_return() { return Big(); }
// LINUX-LABEL: define void @_Z10big_returnv(%struct.Big* noalias sret %agg.result) // LINUX-LABEL: define void @_Z10big_returnv(%struct.Big* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret %agg.result) // WIN32: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret %agg.result) // WIN64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret align 4 %agg.result)
void small_arg(Small s) {} void small_arg(Small s) {}
// LINUX-LABEL: define void @_Z9small_arg5Small(i32 %s.0) // LINUX-LABEL: define void @_Z9small_arg5Small(i32 %s.0)
// WIN32: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.0) // WIN32: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.0)
// WIN64: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.coerce) // WIN64: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg@@YAXUSmall@@@Z"([1 x i32] %s.coerce) // WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg@@YAXUSmall@@@Z"([1 x i32] %s.coerce)
▲ Show 20 Lines • Show All 42 Lines • ▼ Show 20 Lines
// FIXME: MSVC incompatible! // FIXME: MSVC incompatible!
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(%struct.SmallWithDtor* %s) {{.*}} { // WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(%struct.SmallWithDtor* %s) {{.*}} {
// WOA: call arm_aapcs_vfpcc void @"??1SmallWithDtor@@QAA@XZ"(%struct.SmallWithDtor* %s) // WOA: call arm_aapcs_vfpcc void @"??1SmallWithDtor@@QAA@XZ"(%struct.SmallWithDtor* %s)
// WOA: } // WOA: }
// Test that the eligible non-aggregate is passed directly, but returned // Test that the eligible non-aggregate is passed directly, but returned
// indirectly on ARM64 Windows. // indirectly on ARM64 Windows.
// WOA64: define dso_local void @"?small_arg_with_private_member@@YA?AUSmallWithPrivate@@U1@@Z"(%struct.SmallWithPrivate* inreg noalias sret %agg.result, i64 %s.coerce) {{.*}} { // WOA64: define dso_local void @"?small_arg_with_private_member@@YA?AUSmallWithPrivate@@U1@@Z"(%struct.SmallWithPrivate* inreg noalias sret align 4 %agg.result, i64 %s.coerce) {{.*}} {
SmallWithPrivate small_arg_with_private_member(SmallWithPrivate s) { return s; } SmallWithPrivate small_arg_with_private_member(SmallWithPrivate s) { return s; }
void call_small_arg_with_dtor() { void call_small_arg_with_dtor() {
small_arg_with_dtor(SmallWithDtor()); small_arg_with_dtor(SmallWithDtor());
} }
// WIN64-LABEL: define dso_local void @"?call_small_arg_with_dtor@@YAXXZ"() // WIN64-LABEL: define dso_local void @"?call_small_arg_with_dtor@@YAXXZ"()
// WIN64: call %struct.SmallWithDtor* @"??0SmallWithDtor@@QEAA@XZ" // WIN64: call %struct.SmallWithDtor* @"??0SmallWithDtor@@QEAA@XZ"
// WIN64: call void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(i32 %{{.*}}) // WIN64: call void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(i32 %{{.*}})
▲ Show 20 Lines • Show All 83 Lines • ▼ Show 20 Lines
// WIN32-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"() // WIN32-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"()
// WIN32: call void @"?takes_ref_field@@YAXURefField@@@Z"(i32* %{{.*}}) // WIN32: call void @"?takes_ref_field@@YAXURefField@@@Z"(i32* %{{.*}})
// WIN64-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"() // WIN64-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"()
// WIN64: call void @"?takes_ref_field@@YAXURefField@@@Z"(i64 %{{.*}}) // WIN64: call void @"?takes_ref_field@@YAXURefField@@@Z"(i64 %{{.*}})
class Class { class Class {
public: public:
Small thiscall_method_small() { return Small(); } Small thiscall_method_small() { return Small(); }
// LINUX: define {{.*}} void @_ZN5Class21thiscall_method_smallEv(%struct.Small* noalias sret %agg.result, %class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class21thiscall_method_smallEv(%struct.Small* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small@Class@@QAE?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret %agg.result) // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small@Class@@QAE?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret %agg.result) // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
SmallWithCtor thiscall_method_small_with_ctor() { return SmallWithCtor(); } SmallWithCtor thiscall_method_small_with_ctor() { return SmallWithCtor(); }
// LINUX: define {{.*}} void @_ZN5Class31thiscall_method_small_with_ctorEv(%struct.SmallWithCtor* noalias sret %agg.result, %class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class31thiscall_method_small_with_ctorEv(%struct.SmallWithCtor* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small_with_ctor@Class@@QAE?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret %agg.result) // WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small_with_ctor@Class@@QAE?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret %agg.result) // WIN64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret align 4 %agg.result)
Small __cdecl cdecl_method_small() { return Small(); } Small __cdecl cdecl_method_small() { return Small(); }
// LINUX: define {{.*}} void @_ZN5Class18cdecl_method_smallEv(%struct.Small* noalias sret %agg.result, %class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class18cdecl_method_smallEv(%struct.Small* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} void @"?cdecl_method_small@Class@@QAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret %agg.result) // WIN32: define {{.*}} void @"?cdecl_method_small@Class@@QAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?cdecl_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret %agg.result) // WIN64: define linkonce_odr dso_local void @"?cdecl_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
Big __cdecl cdecl_method_big() { return Big(); } Big __cdecl cdecl_method_big() { return Big(); }
// LINUX: define {{.*}} void @_ZN5Class16cdecl_method_bigEv(%struct.Big* noalias sret %agg.result, %class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class16cdecl_method_bigEv(%struct.Big* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} void @"?cdecl_method_big@Class@@QAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret %agg.result) // WIN32: define {{.*}} void @"?cdecl_method_big@Class@@QAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret %agg.result) // WIN64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret align 4 %agg.result)
void thiscall_method_arg(Empty s) {} void thiscall_method_arg(Empty s) {}
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE5Empty(%class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE5Empty(%class.Class* %this)
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUEmpty@@@Z"(%class.Class* %this, %struct.Empty* byval(%struct.Empty) align 4 %s) // WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUEmpty@@@Z"(%class.Class* %this, %struct.Empty* byval(%struct.Empty) align 4 %s)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUEmpty@@@Z"(%class.Class* %this, i8 %s.coerce) // WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUEmpty@@@Z"(%class.Class* %this, i8 %s.coerce)
void thiscall_method_arg(EmptyWithCtor s) {} void thiscall_method_arg(EmptyWithCtor s) {}
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE13EmptyWithCtor(%class.Class* %this) // LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE13EmptyWithCtor(%class.Class* %this)
▲ Show 20 Lines • Show All 145 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/microsoft-abi-vmemptr-conflicts.cpp

Show First 20 Lines • Show All 59 Lines • ▼ Show 20 Lines
void f(C *c) { void f(C *c) {
(c->*(&C::a))(); (c->*(&C::a))();
Big b((c->*(&C::b))()); Big b((c->*(&C::b))());
} }
} }
// CHECK-LABEL: define dso_local void @"?f@sret@@YAXPAUC@1@@Z"(%"struct.sret::C"* %c) // CHECK-LABEL: define dso_local void @"?f@sret@@YAXPAUC@1@@Z"(%"struct.sret::C"* %c)
// CHECK: call x86_thiscallcc i32 bitcast (void (%"struct.sret::C"*, ...)* @"??_9C@sret@@$BA@AE" to i32 (%"struct.sret::C"*)*)(%"struct.sret::C"* %{{.*}}) // CHECK: call x86_thiscallcc i32 bitcast (void (%"struct.sret::C"*, ...)* @"??_9C@sret@@$BA@AE" to i32 (%"struct.sret::C"*)*)(%"struct.sret::C"* %{{.*}})
// CHECK: call x86_thiscallcc void bitcast (void (%"struct.sret::C"*, ...)* @"??_9C@sret@@$BA@AE" to void (%"struct.sret::C"*, %"struct.sret::Big"*)*)(%"struct.sret::C"* %{{.*}}, %"struct.sret::Big"* sret %{{.*}}) // CHECK: call x86_thiscallcc void bitcast (void (%"struct.sret::C"*, ...)* @"??_9C@sret@@$BA@AE" to void (%"struct.sret::C"*, %"struct.sret::Big"*)*)(%"struct.sret::C"* %{{.*}}, %"struct.sret::Big"* sret align 4 %{{.*}})
// CHECK-LABEL: define linkonce_odr x86_thiscallcc void @"??_9C@sret@@$BA@AE"(%"struct.sret::C"* %this, ...) {{.*}} comdat // CHECK-LABEL: define linkonce_odr x86_thiscallcc void @"??_9C@sret@@$BA@AE"(%"struct.sret::C"* %this, ...) {{.*}} comdat
// CHECK: musttail call x86_thiscallcc void (%"struct.sret::C"*, ...) %{{.*}}(%"struct.sret::C"* %{{.*}}, ...) // CHECK: musttail call x86_thiscallcc void (%"struct.sret::C"*, ...) %{{.*}}(%"struct.sret::C"* %{{.*}}, ...)
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
namespace cdecl_inalloca { namespace cdecl_inalloca {
// Fairly evil, since now we end up doing an inalloca-style call through a // Fairly evil, since now we end up doing an inalloca-style call through a
// thunk that doesn't use inalloca. Hopefully the stacks line up? // thunk that doesn't use inalloca. Hopefully the stacks line up?
Show All 25 Lines

clang/test/CodeGenCXX/regcall.cpp

Show First 20 Lines • Show All 68 Lines • ▼ Show 20 Lines
}; };
bool __regcall operator ==(const test_class&, const test_class&){ --x; return false;} bool __regcall operator ==(const test_class&, const test_class&){ --x; return false;}
// CHECK-LIN-DAG: define x86_regcallcc zeroext i1 @_ZeqRK10test_classS1_ // CHECK-LIN-DAG: define x86_regcallcc zeroext i1 @_ZeqRK10test_classS1_
// CHECK-WIN64-DAG: define dso_local x86_regcallcc zeroext i1 @"??8@Yw_NAEBVtest_class@@0@Z" // CHECK-WIN64-DAG: define dso_local x86_regcallcc zeroext i1 @"??8@Yw_NAEBVtest_class@@0@Z"
// CHECK-WIN32-DAG: define dso_local x86_regcallcc zeroext i1 @"??8@Yw_NABVtest_class@@0@Z" // CHECK-WIN32-DAG: define dso_local x86_regcallcc zeroext i1 @"??8@Yw_NABVtest_class@@0@Z"
test_class __regcall operator""_test_class (unsigned long long) { ++x; return test_class{};} test_class __regcall operator""_test_class (unsigned long long) { ++x; return test_class{};}
// CHECK-LIN64-DAG: define x86_regcallcc void @_Zli11_test_classy(%class.test_class* noalias sret %agg.result, i64 %0) // CHECK-LIN64-DAG: define x86_regcallcc void @_Zli11_test_classy(%class.test_class* noalias sret align 4 %agg.result, i64 %0)
// CHECK-LIN32-DAG: define x86_regcallcc void @_Zli11_test_classy(%class.test_class* inreg noalias sret %agg.result, i64 %0) // CHECK-LIN32-DAG: define x86_regcallcc void @_Zli11_test_classy(%class.test_class* inreg noalias sret align 4 %agg.result, i64 %0)
// CHECK-WIN64-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z" // CHECK-WIN64-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z"
// CHECK-WIN32-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z" // CHECK-WIN32-DAG: ??__K_test_class@@Yw?AVtest_class@@_K@Z"
template<typename T> template<typename T>
void __regcall freeTempFunc(T i){} void __regcall freeTempFunc(T i){}
// CHECK-LIN-DAG: define linkonce_odr x86_regcallcc void @_Z24__regcall3__freeTempFuncIiEvT_ // CHECK-LIN-DAG: define linkonce_odr x86_regcallcc void @_Z24__regcall3__freeTempFuncIiEvT_
// CHECK-WIN64-DAG: define linkonce_odr dso_local x86_regcallcc void @"??$freeTempFunc@H@@YwXH@Z" // CHECK-WIN64-DAG: define linkonce_odr dso_local x86_regcallcc void @"??$freeTempFunc@H@@YwXH@Z"
// CHECK-WIN32-DAG: define linkonce_odr dso_local x86_regcallcc void @"??$freeTempFunc@H@@YwXH@Z" // CHECK-WIN32-DAG: define linkonce_odr dso_local x86_regcallcc void @"??$freeTempFunc@H@@YwXH@Z"
// class to force generation of functions // class to force generation of functions
void force_gen() { void force_gen() {
test_class t; test_class t;
test_class t2 = 12_test_class; test_class t2 = 12_test_class;
t += t2; t += t2;
auto t3 = 100_test_class; auto t3 = 100_test_class;
t3.tempFunc(1); t3.tempFunc(1);
freeTempFunc(1); freeTempFunc(1);
t3.do_thing(); t3.do_thing();
} }
long double _Complex __regcall foo(long double _Complex f) { long double _Complex __regcall foo(long double _Complex f) {
return f; return f;
} }
// CHECK-LIN64-DAG: define x86_regcallcc void @_Z15__regcall3__fooCe({ x86_fp80, x86_fp80 }* noalias sret %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 16 %f) // CHECK-LIN64-DAG: define x86_regcallcc void @_Z15__regcall3__fooCe({ x86_fp80, x86_fp80 }* noalias sret align 16 %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 16 %f)
// CHECK-LIN32-DAG: define x86_regcallcc void @_Z15__regcall3__fooCe({ x86_fp80, x86_fp80 }* inreg noalias sret %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 4 %f) // CHECK-LIN32-DAG: define x86_regcallcc void @_Z15__regcall3__fooCe({ x86_fp80, x86_fp80 }* inreg noalias sret align 4 %agg.result, { x86_fp80, x86_fp80 }* byval({ x86_fp80, x86_fp80 }) align 4 %f)
// CHECK-WIN64-DAG: define dso_local x86_regcallcc { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double %f.0, double %f.1) // CHECK-WIN64-DAG: define dso_local x86_regcallcc { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double %f.0, double %f.1)
// CHECK-WIN32-DAG: define dso_local x86_regcallcc { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double %f.0, double %f.1) // CHECK-WIN32-DAG: define dso_local x86_regcallcc { double, double } @"?foo@@YwU?$_Complex@O@__clang@@U12@@Z"(double %f.0, double %f.1)

clang/test/CodeGenCXX/stack-reuse-miscompile.cpp

Show All 33 Lines
// CHECK: [[T2i8:%.*]] = bitcast %class.T* [[T2]] to i8* // CHECK: [[T2i8:%.*]] = bitcast %class.T* [[T2]] to i8*
// CHECK: call void @llvm.lifetime.start.p0i8(i64 16, i8* [[T2i8]]) // CHECK: call void @llvm.lifetime.start.p0i8(i64 16, i8* [[T2i8]])
// CHECK: [[T4:%.*]] = call %class.T* @_ZN1TC1EPKc(%class.T* [[T2]], i8* getelementptr inbounds ([2 x i8], [2 x i8]* @.str, i32 0, i32 0)) // CHECK: [[T4:%.*]] = call %class.T* @_ZN1TC1EPKc(%class.T* [[T2]], i8* getelementptr inbounds ([2 x i8], [2 x i8]* @.str, i32 0, i32 0))
// //
// CHECK: [[T3i8:%.*]] = bitcast %class.T* [[T3]] to i8* // CHECK: [[T3i8:%.*]] = bitcast %class.T* [[T3]] to i8*
// CHECK: call void @llvm.lifetime.start.p0i8(i64 16, i8* [[T3i8]]) // CHECK: call void @llvm.lifetime.start.p0i8(i64 16, i8* [[T3i8]])
// CHECK: [[T5:%.*]] = call %class.T* @_ZN1TC1E1S(%class.T* [[T3]], [2 x i32] %{{.*}}) // CHECK: [[T5:%.*]] = call %class.T* @_ZN1TC1E1S(%class.T* [[T3]], [2 x i32] %{{.*}})
// //
// CHECK: call void @_ZNK1T6concatERKS_(%class.T* sret [[T1]], %class.T* [[T2]], %class.T* dereferenceable(16) [[T3]]) // CHECK: call void @_ZNK1T6concatERKS_(%class.T* sret align 4 [[T1]], %class.T* [[T2]], %class.T* dereferenceable(16) [[T3]])
// CHECK: [[T6:%.*]] = call i8* @_ZNK1T3strEv(%class.T* [[T1]]) // CHECK: [[T6:%.*]] = call i8* @_ZNK1T3strEv(%class.T* [[T1]])
// //
// CHECK: call void @llvm.lifetime.end.p0i8( // CHECK: call void @llvm.lifetime.end.p0i8(
// CHECK: call void @llvm.lifetime.end.p0i8( // CHECK: call void @llvm.lifetime.end.p0i8(
// CHECK: call void @llvm.lifetime.end.p0i8( // CHECK: call void @llvm.lifetime.end.p0i8(
// CHECK: ret i8* [[T6]] // CHECK: ret i8* [[T6]]
return T("[").concat(T(s)).str(); return T("[").concat(T(s)).str();
} }
// CHECK: declare {{.*}}llvm.lifetime.start // CHECK: declare {{.*}}llvm.lifetime.start
// CHECK: declare {{.*}}llvm.lifetime.end // CHECK: declare {{.*}}llvm.lifetime.end

clang/test/CodeGenCXX/stack-reuse.cpp

Show First 20 Lines • Show All 129 Lines • ▼ Show 20 Lines// CHECK: call void @llvm.lifetime.end
} }
} }
int large_combiner_test(S_large s) { int large_combiner_test(S_large s) {
// CHECK-LABEL: define i32 @large_combiner_test // CHECK-LABEL: define i32 @large_combiner_test
// CHECK: [[T2:%.*]] = alloca %struct.Combiner // CHECK: [[T2:%.*]] = alloca %struct.Combiner
// CHECK: [[T1:%.*]] = alloca %struct.Combiner // CHECK: [[T1:%.*]] = alloca %struct.Combiner
// CHECK: [[T3:%.*]] = call %struct.Combiner* @_ZN8CombinerC1E7S_large(%struct.Combiner* nonnull [[T1]], [9 x i32] %s.coerce) // CHECK: [[T3:%.*]] = call %struct.Combiner* @_ZN8CombinerC1E7S_large(%struct.Combiner* nonnull [[T1]], [9 x i32] %s.coerce)
// CHECK: call void @_ZN8Combiner1fEv(%struct.Combiner* nonnull sret [[T2]], %struct.Combiner* nonnull [[T1]]) // CHECK: call void @_ZN8Combiner1fEv(%struct.Combiner* nonnull sret align 4 [[T2]], %struct.Combiner* nonnull [[T1]])
// CHECK: [[T4:%.*]] = getelementptr inbounds %struct.Combiner, %struct.Combiner* [[T2]], i32 0, i32 0, i32 0, i32 0 // CHECK: [[T4:%.*]] = getelementptr inbounds %struct.Combiner, %struct.Combiner* [[T2]], i32 0, i32 0, i32 0, i32 0
// CHECK: [[T5:%.*]] = load i32, i32* [[T4]] // CHECK: [[T5:%.*]] = load i32, i32* [[T4]]
// CHECK: ret i32 [[T5]] // CHECK: ret i32 [[T5]]
return Combiner(s).f().a.a[0]; return Combiner(s).f().a.a[0];
} }
} }

clang/test/CodeGenCXX/temporaries.cpp

Show First 20 Lines • Show All 397 Lines • ▼ Show 20 Lines void test0() {
// CHECK-NEXT: [[T0:%.*]] = alloca [[A]], align 8 // CHECK-NEXT: [[T0:%.*]] = alloca [[A]], align 8
// CHECK-NEXT: [[K:%.*]] = alloca [[A]], align 8 // CHECK-NEXT: [[K:%.*]] = alloca [[A]], align 8
// CHECK-NEXT: [[T1:%.*]] = alloca [[A]], align 8 // CHECK-NEXT: [[T1:%.*]] = alloca [[A]], align 8
// CHECK-NEXT: call void @_ZN7Elision3fooEv() // CHECK-NEXT: call void @_ZN7Elision3fooEv()
// CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[I]]) // CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[I]])
A i = (foo(), A()); A i = (foo(), A());
// CHECK-NEXT: call void @_ZN7Elision4fooAEv([[A]]* sret [[T0]]) // CHECK-NEXT: call void @_ZN7Elision4fooAEv([[A]]* sret align 8 [[T0]])
// CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[J]]) // CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[J]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T0]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T0]])
A j = (fooA(), A()); A j = (fooA(), A());
// CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[T1]]) // CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[T1]])
// CHECK-NEXT: call void @_ZN7Elision4fooAEv([[A]]* sret [[K]]) // CHECK-NEXT: call void @_ZN7Elision4fooAEv([[A]]* sret align 8 [[K]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T1]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T1]])
A k = (A(), fooA()); A k = (A(), fooA());
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[K]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[K]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[J]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[J]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[I]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[I]])
} }
Show All 10 Lines void test1(bool c, A x) {
// CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[J]], [[A]]* dereferenceable({{[0-9]+}}) [[X]]) // CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[J]], [[A]]* dereferenceable({{[0-9]+}}) [[X]])
// CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[J]]) // CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[J]])
A j = (c ? x : A()); A j = (c ? x : A());
// CHECK: call void @_ZN7Elision1AD1Ev([[A]]* [[J]]) // CHECK: call void @_ZN7Elision1AD1Ev([[A]]* [[J]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[I]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[I]])
} }
// CHECK: define void @_ZN7Elision5test2Ev([[A]]* noalias sret // CHECK: define void @_ZN7Elision5test2Ev([[A]]* noalias sret align 8
A test2() { A test2() {
// CHECK: call void @_ZN7Elision3fooEv() // CHECK: call void @_ZN7Elision3fooEv()
// CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[RET:%.*]]) // CHECK-NEXT: call void @_ZN7Elision1AC1Ev([[A]]* [[RET:%.*]])
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
return (foo(), A()); return (foo(), A());
} }
// CHECK: define void @_ZN7Elision5test3EiNS_1AE([[A]]* noalias sret // CHECK: define void @_ZN7Elision5test3EiNS_1AE([[A]]* noalias sret align 8
A test3(int v, A x) { A test3(int v, A x) {
if (v < 5) if (v < 5)
// CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[RET:%.*]]) // CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[RET:%.*]])
// CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[RET]], [[A]]* dereferenceable({{[0-9]+}}) [[X:%.*]]) // CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[RET]], [[A]]* dereferenceable({{[0-9]+}}) [[X:%.*]])
return (v < 0 ? A() : x); return (v < 0 ? A() : x);
else else
// CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[RET]], [[A]]* dereferenceable({{[0-9]+}}) [[X]]) // CHECK: call void @_ZN7Elision1AC1ERKS0_([[A]]* [[RET]], [[A]]* dereferenceable({{[0-9]+}}) [[X]])
// CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[RET]]) // CHECK: call void @_ZN7Elision1AC1Ev([[A]]* [[RET]])
Show All 24 Lines void test4() {
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[CUR]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[CUR]])
// CHECK-NEXT: [[T0:%.*]] = icmp eq [[A]]* [[CUR]], [[BEGIN]] // CHECK-NEXT: [[T0:%.*]] = icmp eq [[A]]* [[CUR]], [[BEGIN]]
// CHECK-NEXT: br i1 [[T0]], // CHECK-NEXT: br i1 [[T0]],
// CHECK: call void @_ZN7Elision1AD1Ev([[A]]* [[X]]) // CHECK: call void @_ZN7Elision1AD1Ev([[A]]* [[X]])
} }
// rdar://problem/8433352 // rdar://problem/8433352
// CHECK: define void @_ZN7Elision5test5Ev([[A]]* noalias sret // CHECK: define void @_ZN7Elision5test5Ev([[A]]* noalias sret align 8
struct B { A a; B(); }; struct B { A a; B(); };
A test5() { A test5() {
// CHECK: [[AT0:%.*]] = alloca [[A]], align 8 // CHECK: [[AT0:%.*]] = alloca [[A]], align 8
// CHECK-NEXT: [[BT0:%.*]] = alloca [[B:%.*]], align 8 // CHECK-NEXT: [[BT0:%.*]] = alloca [[B:%.*]], align 8
// CHECK-NEXT: [[X:%.*]] = alloca [[A]], align 8 // CHECK-NEXT: [[X:%.*]] = alloca [[A]], align 8
// CHECK-NEXT: [[BT1:%.*]] = alloca [[B]], align 8 // CHECK-NEXT: [[BT1:%.*]] = alloca [[B]], align 8
// CHECK-NEXT: [[BT2:%.*]] = alloca [[B]], align 8 // CHECK-NEXT: [[BT2:%.*]] = alloca [[B]], align 8
Show All 21 Linesnamespace Elision {
} }
// Reduced from webkit. // Reduced from webkit.
// CHECK: define void @_ZN7Elision5test6EPKNS_1CE([[C:%.*]]* // CHECK: define void @_ZN7Elision5test6EPKNS_1CE([[C:%.*]]*
struct C { operator A() const; }; struct C { operator A() const; };
void test6(const C *x) { void test6(const C *x) {
// CHECK: [[T0:%.*]] = alloca [[A]], align 8 // CHECK: [[T0:%.*]] = alloca [[A]], align 8
// CHECK: [[X:%.*]] = load [[C]]*, [[C]]** {{%.*}}, align 8 // CHECK: [[X:%.*]] = load [[C]]*, [[C]]** {{%.*}}, align 8
// CHECK-NEXT: call void @_ZNK7Elision1CcvNS_1AEEv([[A]]* sret [[T0]], [[C]]* [[X]]) // CHECK-NEXT: call void @_ZNK7Elision1CcvNS_1AEEv([[A]]* sret align 8 [[T0]], [[C]]* [[X]])
// CHECK-NEXT: call void @_ZNK7Elision1A3fooEv([[A]]* [[T0]]) // CHECK-NEXT: call void @_ZNK7Elision1A3fooEv([[A]]* [[T0]])
// CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T0]]) // CHECK-NEXT: call void @_ZN7Elision1AD1Ev([[A]]* [[T0]])
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
A(*x).foo(); A(*x).foo();
} }
} }
namespace PR8623 { namespace PR8623 {
▲ Show 20 Lines • Show All 382 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/thiscall-struct-return.cpp

Show All 28 Linespublic:
} }
}; };
// CHECK-LABEL: define void @_Z4testv() // CHECK-LABEL: define void @_Z4testv()
void test( void ) { void test( void ) {
// CHECK: call void @_ZN1CC1Ev(%class.C* [[C:%.+]]) // CHECK: call void @_ZN1CC1Ev(%class.C* [[C:%.+]])
C c; C c;
// CHECK: call x86_thiscallcc void @_ZNK1C5SmallEv(%struct.S* sret %{{.+}}, %class.C* [[C]]) // CHECK: call x86_thiscallcc void @_ZNK1C5SmallEv(%struct.S* sret align 4 %{{.+}}, %class.C* [[C]])
(void)c.Small(); (void)c.Small();
// CHECK: call x86_thiscallcc void @_ZNK1C6MediumEv(%struct.M* sret %{{.+}}, %class.C* [[C]]) // CHECK: call x86_thiscallcc void @_ZNK1C6MediumEv(%struct.M* sret align 4 %{{.+}}, %class.C* [[C]])
(void)c.Medium(); (void)c.Medium();
} }

clang/test/CodeGenCXX/thunk-returning-memptr.cpp

Show All 17 Lines
}; };
C::C() {} C::C() {}
// Make sure the member pointer is returned from the thunk via the return slot. // Make sure the member pointer is returned from the thunk via the return slot.
// Because of the tail call, the return value cannot be copied into a local // Because of the tail call, the return value cannot be copied into a local
// alloca. (PR39901) // alloca. (PR39901)
// CHECK-LABEL: define linkonce_odr void @_ZThn4_N1C1fEv({ i32, i32 }* noalias sret %agg.result, %struct.C* %this) // CHECK-LABEL: define linkonce_odr void @_ZThn4_N1C1fEv({ i32, i32 }* noalias sret align 4 %agg.result, %struct.C* %this)
// CHECK: tail call void @_ZN1C1fEv({ i32, i32 }* sret %agg.result // CHECK: tail call void @_ZN1C1fEv({ i32, i32 }* sret align 4 %agg.result

clang/test/CodeGenCXX/thunks.cpp

Show First 20 Lines • Show All 200 Lines • ▼ Show 20 Lines struct Thunks : Base1, Base2 {
long l; long l;
virtual X f(); virtual X f();
}; };
// CHECK-LABEL: define void @_ZThn16_N5Test66Thunks1fEv // CHECK-LABEL: define void @_ZThn16_N5Test66Thunks1fEv
// CHECK-DBG-NOT: dbg.declare // CHECK-DBG-NOT: dbg.declare
// CHECK-NOT: memcpy // CHECK-NOT: memcpy
// CHECK: {{call void @_ZN5Test66Thunks1fEv.*sret}} // CHECK: {{call void @_ZN5Test66Thunks1fEv.*sret align 1}}
// CHECK: ret void // CHECK: ret void
X Thunks::f() { return X(); } X Thunks::f() { return X(); }
// WIN64-LABEL: define linkonce_odr dso_local void @"?f@Thunks@Test6@@WBA@EAA?AUX@2@XZ"({{.*}} sret %{{.*}}) // WIN64-LABEL: define linkonce_odr dso_local void @"?f@Thunks@Test6@@WBA@EAA?AUX@2@XZ"({{.*}} sret align 1 %{{.*}})
// WIN64-NOT: memcpy // WIN64-NOT: memcpy
// WIN64: tail call void @"?f@Thunks@Test6@@UEAA?AUX@2@XZ"({{.*}} sret %{{.*}}) // WIN64: tail call void @"?f@Thunks@Test6@@UEAA?AUX@2@XZ"({{.*}} sret align 1 %{{.*}})
} }
namespace Test7 { namespace Test7 {
// PR7188 // PR7188
struct X { struct X {
X(); X();
X(const X&); X(const X&);
X &operator=(const X&); X &operator=(const X&);
▲ Show 20 Lines • Show All 323 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/trivial_abi.cpp

Show First 20 Lines • Show All 120 Lines • ▼ Show 20 Lines
// CHECK: define void @_Z14testParamLarge5Large(%[[STRUCT_LARGE:.*]]* %[[A:.*]]) // CHECK: define void @_Z14testParamLarge5Large(%[[STRUCT_LARGE:.*]]* %[[A:.*]])
// CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeD1Ev(%[[STRUCT_LARGE]]* %[[A]]) // CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeD1Ev(%[[STRUCT_LARGE]]* %[[A]])
// CHECK: ret void // CHECK: ret void
// CHECK: } // CHECK: }
void testParamLarge(Large a) noexcept { void testParamLarge(Large a) noexcept {
} }
// CHECK: define void @_Z15testReturnLargev(%[[STRUCT_LARGE:.*]]* noalias sret %[[AGG_RESULT:.*]]) // CHECK: define void @_Z15testReturnLargev(%[[STRUCT_LARGE:.*]]* noalias sret align 8 %[[AGG_RESULT:.*]])
// CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeC1Ev(%[[STRUCT_LARGE]]* %[[AGG_RESULT]]) // CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeC1Ev(%[[STRUCT_LARGE]]* %[[AGG_RESULT]])
// CHECK: ret void // CHECK: ret void
// CHECK: } // CHECK: }
Large testReturnLarge() { Large testReturnLarge() {
Large t; Large t;
return t; return t;
} }
Show All 10 Lines
void testCallLarge0() { void testCallLarge0() {
Large t; Large t;
testParamLarge(t); testParamLarge(t);
} }
// CHECK: define void @_Z14testCallLarge1v() // CHECK: define void @_Z14testCallLarge1v()
// CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8 // CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8
// CHECK: call void @_Z15testReturnLargev(%[[STRUCT_LARGE]]* sret %[[AGG_TMP]]) // CHECK: call void @_Z15testReturnLargev(%[[STRUCT_LARGE]]* sret align 8 %[[AGG_TMP]])
// CHECK: call void @_Z14testParamLarge5Large(%[[STRUCT_LARGE]]* %[[AGG_TMP]]) // CHECK: call void @_Z14testParamLarge5Large(%[[STRUCT_LARGE]]* %[[AGG_TMP]])
// CHECK: ret void // CHECK: ret void
// CHECK: } // CHECK: }
void testCallLarge1() { void testCallLarge1() {
testParamLarge(testReturnLarge()); testParamLarge(testReturnLarge());
} }
// CHECK: define void @_Z16testIgnoredLargev() // CHECK: define void @_Z16testIgnoredLargev()
// CHECK: %[[AGG_TMP_ENSURED:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8 // CHECK: %[[AGG_TMP_ENSURED:.*]] = alloca %[[STRUCT_LARGE:.*]], align 8
// CHECK: call void @_Z15testReturnLargev(%[[STRUCT_LARGE]]* sret %[[AGG_TMP_ENSURED]]) // CHECK: call void @_Z15testReturnLargev(%[[STRUCT_LARGE]]* sret align 8 %[[AGG_TMP_ENSURED]])
// CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeD1Ev(%[[STRUCT_LARGE]]* %[[AGG_TMP_ENSURED]]) // CHECK: %[[CALL:.*]] = call %[[STRUCT_LARGE]]* @_ZN5LargeD1Ev(%[[STRUCT_LARGE]]* %[[AGG_TMP_ENSURED]])
// CHECK: ret void // CHECK: ret void
// CHECK: } // CHECK: }
void testIgnoredLarge() { void testIgnoredLarge() {
testReturnLarge(); testReturnLarge();
} }
// CHECK: define i64 @_Z20testReturnHasTrivialv() // CHECK: define i64 @_Z20testReturnHasTrivialv()
// CHECK: %[[RETVAL:.*]] = alloca %[[STRUCT_TRIVIAL:.*]], align 4 // CHECK: %[[RETVAL:.*]] = alloca %[[STRUCT_TRIVIAL:.*]], align 4
// CHECK: %[[COERCE_DIVE:.*]] = getelementptr inbounds %[[STRUCT_TRIVIAL]], %[[STRUCT_TRIVIAL]]* %[[RETVAL]], i32 0, i32 0 // CHECK: %[[COERCE_DIVE:.*]] = getelementptr inbounds %[[STRUCT_TRIVIAL]], %[[STRUCT_TRIVIAL]]* %[[RETVAL]], i32 0, i32 0
// CHECK: %[[V0:.*]] = load i32, i32* %[[COERCE_DIVE]], align 4 // CHECK: %[[V0:.*]] = load i32, i32* %[[COERCE_DIVE]], align 4
// CHECK: %[[COERCE_VAL_II:.*]] = zext i32 %[[V0]] to i64 // CHECK: %[[COERCE_VAL_II:.*]] = zext i32 %[[V0]] to i64
// CHECK: ret i64 %[[COERCE_VAL_II]] // CHECK: ret i64 %[[COERCE_VAL_II]]
// CHECK: } // CHECK: }
Trivial testReturnHasTrivial() { Trivial testReturnHasTrivial() {
Trivial t; Trivial t;
return t; return t;
} }
// CHECK: define void @_Z23testReturnHasNonTrivialv(%[[STRUCT_NONTRIVIAL:.*]]* noalias sret %[[AGG_RESULT:.*]]) // CHECK: define void @_Z23testReturnHasNonTrivialv(%[[STRUCT_NONTRIVIAL:.*]]* noalias sret align 4 %[[AGG_RESULT:.*]])
// CHECK: %[[CALL:.*]] = call %[[STRUCT_NONTRIVIAL]]* @_ZN10NonTrivialC1Ev(%[[STRUCT_NONTRIVIAL]]* %[[AGG_RESULT]]) // CHECK: %[[CALL:.*]] = call %[[STRUCT_NONTRIVIAL]]* @_ZN10NonTrivialC1Ev(%[[STRUCT_NONTRIVIAL]]* %[[AGG_RESULT]])
// CHECK: ret void // CHECK: ret void
// CHECK: } // CHECK: }
NonTrivial testReturnHasNonTrivial() { NonTrivial testReturnHasNonTrivial() {
NonTrivial t; NonTrivial t;
return t; return t;
} }
▲ Show 20 Lines • Show All 42 Lines • Show Last 20 Lines

clang/test/CodeGenCXX/unknown-anytype.cpp

Show First 20 Lines • Show All 65 Lines • ▼ Show 20 Lineslong test6() {
return (long long) test6_any(0); return (long long) test6_any(0);
} }
struct Test7 { struct Test7 {
~Test7(); ~Test7();
}; };
extern "C" __unknown_anytype test7_any(int); extern "C" __unknown_anytype test7_any(int);
Test7 test7() { Test7 test7() {
// COMMON: call void @test7_any({{%.*}}* sret {{%.*}}, i32 5) // COMMON: call void @test7_any({{%.*}}* sret align 1 {{%.*}}, i32 5)
return (Test7) test7_any(5); return (Test7) test7_any(5);
} }
struct Test8 { struct Test8 {
__unknown_anytype foo(); __unknown_anytype foo();
__unknown_anytype foo(int); __unknown_anytype foo(int);
void test(); void test();
▲ Show 20 Lines • Show All 43 Lines • Show Last 20 Lines

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

Show All 24 Lines
// //
// CHECK: declare void @_Z3use9one_field(double) // CHECK: declare void @_Z3use9one_field(double)
// CHECK: declare double @_Z13def_one_fieldv() // CHECK: declare double @_Z13def_one_fieldv()
struct two_fields { struct two_fields {
double d, e; double d, e;
}; };
test(two_fields); test(two_fields);
// CHECK: define void @_Z7forward10two_fields(%struct.two_fields* noalias nocapture sret %{{.*}}, %struct.two_fields* nocapture readonly byval(%struct.two_fields) align 8 %{{.*}}) // CHECK: define void @_Z7forward10two_fields(%struct.two_fields* noalias nocapture sret align 8 %{{.*}}, %struct.two_fields* nocapture readonly byval(%struct.two_fields) align 8 %{{.*}})
// //
// CHECK: define void @_Z15test_two_fieldsv() // CHECK: define void @_Z15test_two_fieldsv()
// CHECK: %[[tmp:.*]] = alloca %struct.two_fields, align 8 // CHECK: %[[tmp:.*]] = alloca %struct.two_fields, align 8
// CHECK: call void @_Z14def_two_fieldsv(%struct.two_fields* nonnull sret %[[tmp]]) // CHECK: call void @_Z14def_two_fieldsv(%struct.two_fields* nonnull sret align 8 %[[tmp]])
// CHECK: call void @_Z3use10two_fields(%struct.two_fields* nonnull byval(%struct.two_fields) align 8 %[[tmp]]) // CHECK: call void @_Z3use10two_fields(%struct.two_fields* nonnull byval(%struct.two_fields) align 8 %[[tmp]])
// CHECK: ret void // CHECK: ret void
// //
// CHECK: declare void @_Z3use10two_fields(%struct.two_fields* byval(%struct.two_fields) align 8) // CHECK: declare void @_Z3use10two_fields(%struct.two_fields* byval(%struct.two_fields) align 8)
// CHECK: declare void @_Z14def_two_fieldsv(%struct.two_fields* sret) // CHECK: declare void @_Z14def_two_fieldsv(%struct.two_fields* sret align 8)
struct copy_ctor { struct copy_ctor {
double d; double d;
copy_ctor(copy_ctor const &); copy_ctor(copy_ctor const &);
}; };
test(copy_ctor); test(copy_ctor);
// CHECK: define void @_Z7forward9copy_ctor(%struct.copy_ctor* noalias sret %{{.*}}, %struct.copy_ctor* nonnull %{{.*}}) // CHECK: define void @_Z7forward9copy_ctor(%struct.copy_ctor* noalias sret align 8 %{{.*}}, %struct.copy_ctor* nonnull %{{.*}})
// //
// CHECK: declare %struct.copy_ctor* @_ZN9copy_ctorC1ERKS_(%struct.copy_ctor* returned, %struct.copy_ctor* dereferenceable(8)) // CHECK: declare %struct.copy_ctor* @_ZN9copy_ctorC1ERKS_(%struct.copy_ctor* returned, %struct.copy_ctor* dereferenceable(8))
// //
// CHECK: define void @_Z14test_copy_ctorv() // CHECK: define void @_Z14test_copy_ctorv()
// CHECK: %[[tmp:.*]] = alloca %struct.copy_ctor, align 8 // CHECK: %[[tmp:.*]] = alloca %struct.copy_ctor, align 8
// CHECK: call void @_Z13def_copy_ctorv(%struct.copy_ctor* nonnull sret %[[tmp]]) // CHECK: call void @_Z13def_copy_ctorv(%struct.copy_ctor* nonnull sret align 8 %[[tmp]])
// CHECK: call void @_Z3use9copy_ctor(%struct.copy_ctor* nonnull %[[tmp]]) // CHECK: call void @_Z3use9copy_ctor(%struct.copy_ctor* nonnull %[[tmp]])
// CHECK: ret void // CHECK: ret void
// //
// CHECK: declare void @_Z3use9copy_ctor(%struct.copy_ctor*) // CHECK: declare void @_Z3use9copy_ctor(%struct.copy_ctor*)
// CHECK: declare void @_Z13def_copy_ctorv(%struct.copy_ctor* sret) // CHECK: declare void @_Z13def_copy_ctorv(%struct.copy_ctor* sret align 8)
struct __attribute__((aligned(16))) aligned_copy_ctor { struct __attribute__((aligned(16))) aligned_copy_ctor {
double d, e; double d, e;
aligned_copy_ctor(aligned_copy_ctor const &); aligned_copy_ctor(aligned_copy_ctor const &);
}; };
test(aligned_copy_ctor); test(aligned_copy_ctor);
// CHECK: define void @_Z7forward17aligned_copy_ctor(%struct.aligned_copy_ctor* noalias sret %{{.*}}, %struct.aligned_copy_ctor* nonnull %{{.*}}) // CHECK: define void @_Z7forward17aligned_copy_ctor(%struct.aligned_copy_ctor* noalias sret align 16 %{{.*}}, %struct.aligned_copy_ctor* nonnull %{{.*}})
// //
// CHECK: declare %struct.aligned_copy_ctor* @_ZN17aligned_copy_ctorC1ERKS_(%struct.aligned_copy_ctor* returned, %struct.aligned_copy_ctor* dereferenceable(16)) // CHECK: declare %struct.aligned_copy_ctor* @_ZN17aligned_copy_ctorC1ERKS_(%struct.aligned_copy_ctor* returned, %struct.aligned_copy_ctor* dereferenceable(16))
// //
// CHECK: define void @_Z22test_aligned_copy_ctorv() // CHECK: define void @_Z22test_aligned_copy_ctorv()
// CHECK: %[[tmp:.*]] = alloca %struct.aligned_copy_ctor, align 16 // CHECK: %[[tmp:.*]] = alloca %struct.aligned_copy_ctor, align 16
// CHECK: call void @_Z21def_aligned_copy_ctorv(%struct.aligned_copy_ctor* nonnull sret %[[tmp]]) // CHECK: call void @_Z21def_aligned_copy_ctorv(%struct.aligned_copy_ctor* nonnull sret align 16 %[[tmp]])
// CHECK: call void @_Z3use17aligned_copy_ctor(%struct.aligned_copy_ctor* nonnull %[[tmp]]) // CHECK: call void @_Z3use17aligned_copy_ctor(%struct.aligned_copy_ctor* nonnull %[[tmp]])
// CHECK: ret void // CHECK: ret void
// //
// CHECK: declare void @_Z3use17aligned_copy_ctor(%struct.aligned_copy_ctor*) // CHECK: declare void @_Z3use17aligned_copy_ctor(%struct.aligned_copy_ctor*)
// CHECK: declare void @_Z21def_aligned_copy_ctorv(%struct.aligned_copy_ctor* sret) // CHECK: declare void @_Z21def_aligned_copy_ctorv(%struct.aligned_copy_ctor* sret align 16)
struct empty {}; struct empty {};
test(empty); test(empty);
// CHECK: define void @_Z7forward5empty() // CHECK: define void @_Z7forward5empty()
// //
// CHECK: define void @_Z10test_emptyv() // CHECK: define void @_Z10test_emptyv()
// CHECK: call void @_Z9def_emptyv() // CHECK: call void @_Z9def_emptyv()
// CHECK: call void @_Z3use5empty() // CHECK: call void @_Z3use5empty()
Show All 18 Lines

clang/test/CodeGenCXX/x86_32-arguments.cpp

// RUN: %clang_cc1 -triple i386-apple-darwin9 -emit-llvm -o - %s | FileCheck %s // RUN: %clang_cc1 -triple i386-apple-darwin9 -emit-llvm -o - %s | FileCheck %s
// Non-trivial dtors, should both be passed indirectly. // Non-trivial dtors, should both be passed indirectly.
struct S { struct S {
~S(); ~S();
short s; short s;
}; };
// CHECK-LABEL: define void @_Z1fv(%struct.S* noalias sret % // CHECK-LABEL: define void @_Z1fv(%struct.S* noalias sret align 2 %
S f() { return S(); } S f() { return S(); }
// CHECK-LABEL: define void @_Z1f1S(%struct.S* %0) // CHECK-LABEL: define void @_Z1f1S(%struct.S* %0)
void f(S) { } void f(S) { }
// Non-trivial dtors, should both be passed indirectly. // Non-trivial dtors, should both be passed indirectly.
class C { class C {
public: public:
~C(); ~C();
double c; double c;
}; };
// CHECK-LABEL: define void @_Z1gv(%class.C* noalias sret % // CHECK-LABEL: define void @_Z1gv(%class.C* noalias sret align 4 %
C g() { return C(); } C g() { return C(); }
// CHECK-LABEL: define void @_Z1f1C(%class.C* %0) // CHECK-LABEL: define void @_Z1f1C(%class.C* %0)
void f(C) { } void f(C) { }
▲ Show 20 Lines • Show All 68 Lines • ▼ Show 20 Lines
s6_mfp f6_1(s6_mfp a) { return a; } s6_mfp f6_1(s6_mfp a) { return a; }
// CHECK-LABEL: define double @_Z2f7v() // CHECK-LABEL: define double @_Z2f7v()
struct s7_0 { unsigned : 0; }; struct s7_0 { unsigned : 0; };
struct s7_1 { double x; }; struct s7_1 { double x; };
struct s7 : s7_0, s7_1 { }; struct s7 : s7_0, s7_1 { };
s7 f7() { return s7(); } s7 f7() { return s7(); }
// CHECK-LABEL: define void @_Z2f8v(%struct.s8* noalias sret %agg.result) // CHECK-LABEL: define void @_Z2f8v(%struct.s8* noalias sret align 4 %agg.result)
struct s8_0 { }; struct s8_0 { };
struct s8_1 { double x; }; struct s8_1 { double x; };
struct s8 { s8_0 a; s8_1 b; }; struct s8 { s8_0 a; s8_1 b; };
s8 f8() { return s8(); } s8 f8() { return s8(); }
// CHECK-LABEL: define void @_Z2f9v(%struct.s9* noalias sret %agg.result) // CHECK-LABEL: define void @_Z2f9v(%struct.s9* noalias sret align 4 %agg.result)
struct s9_0 { unsigned : 0; }; struct s9_0 { unsigned : 0; };
struct s9_1 { double x; }; struct s9_1 { double x; };
struct s9 { s9_0 a; s9_1 b; }; struct s9 { s9_0 a; s9_1 b; };
s9 f9() { return s9(); } s9 f9() { return s9(); }

clang/test/CodeGenCXX/x86_64-arguments.cpp

Show First 20 Lines • Show All 170 Lines • ▼ Show 20 Linesnamespace test9 {
// Large enough to be passed indirectly. // Large enough to be passed indirectly.
struct S { void *data[3]; }; struct S { void *data[3]; };
struct T { void *data[2]; }; struct T { void *data[2]; };
// CHECK: define void @_ZN5test93fooEPNS_1SEPNS_1TE([[S:%.*]]* %0, [[T:%.*]]* %1) // CHECK: define void @_ZN5test93fooEPNS_1SEPNS_1TE([[S:%.*]]* %0, [[T:%.*]]* %1)
void foo(S*, T*) {} void foo(S*, T*) {}
// CHECK: define void @_ZN5test91aEiiiiNS_1TEPv([[S]]* noalias sret {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, [[T]]* byval([[T]]) align 8 %4, i8* %5) // CHECK: define void @_ZN5test91aEiiiiNS_1TEPv([[S]]* noalias sret align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, [[T]]* byval([[T]]) align 8 %4, i8* %5)
S a(int, int, int, int, T, void*) { S a(int, int, int, int, T, void*) {
return S(); return S();
} }
// CHECK: define [[S]]* @_ZN5test91bEPNS_1SEiiiiNS_1TEPv([[S]]* {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, [[T:%.*]]* byval([[T]]) align 8 %4, i8* %5) // CHECK: define [[S]]* @_ZN5test91bEPNS_1SEiiiiNS_1TEPv([[S]]* {{%.*}}, i32 %0, i32 %1, i32 %2, i32 %3, [[T:%.*]]* byval([[T]]) align 8 %4, i8* %5)
S* b(S* sret, int, int, int, int, T, void*) { S* b(S* sret, int, int, int, int, T, void*) {
return sret; return sret;
} }
// CHECK: define void @_ZN5test91cEiiiNS_1TEPv([[S]]* noalias sret {{%.*}}, i32 %0, i32 %1, i32 %2, i8* {{%.*}}, i8* {{%.*}}, i8* %3) // CHECK: define void @_ZN5test91cEiiiNS_1TEPv([[S]]* noalias sret align 8 {{%.*}}, i32 %0, i32 %1, i32 %2, i8* {{%.*}}, i8* {{%.*}}, i8* %3)
S c(int, int, int, T, void*) { S c(int, int, int, T, void*) {
return S(); return S();
} }
// CHECK: define [[S]]* @_ZN5test91dEPNS_1SEiiiNS_1TEPv([[S]]* {{%.*}}, i32 %0, i32 %1, i32 %2, i8* {{%.*}}, i8* {{%.*}}, i8* %3) // CHECK: define [[S]]* @_ZN5test91dEPNS_1SEiiiNS_1TEPv([[S]]* {{%.*}}, i32 %0, i32 %1, i32 %2, i8* {{%.*}}, i8* {{%.*}}, i8* %3)
S* d(S* sret, int, int, int, T, void*) { S* d(S* sret, int, int, int, T, void*) {
return sret; return sret;
} }
Show All 26 Lines

clang/test/CodeGenCoroutines/coro-await.cpp

Show First 20 Lines • Show All 124 Lines • ▼ Show 20 Linesstruct std::experimental::coroutine_traits<void,int> {
}; };
}; };
// CHECK-LABEL: f1( // CHECK-LABEL: f1(
extern "C" void f1(int) { extern "C" void f1(int) {
// CHECK: %[[PROMISE:.+]] = alloca %"struct.std::experimental::coroutine_traits<void, int>::promise_type" // CHECK: %[[PROMISE:.+]] = alloca %"struct.std::experimental::coroutine_traits<void, int>::promise_type"
// CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin( // CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin(
co_yield 42; co_yield 42;
// CHECK: call void @_ZNSt12experimental16coroutine_traitsIJviEE12promise_type11yield_valueEi(%struct.suspend_maybe* sret %[[AWAITER:.+]], %"struct.std::experimental::coroutine_traits<void, int>::promise_type"* %[[PROMISE]], i32 42) // CHECK: call void @_ZNSt12experimental16coroutine_traitsIJviEE12promise_type11yield_valueEi(%struct.suspend_maybe* sret align 4 %[[AWAITER:.+]], %"struct.std::experimental::coroutine_traits<void, int>::promise_type"* %[[PROMISE]], i32 42)
// See if we need to suspend: // See if we need to suspend:
// -------------------------- // --------------------------
// CHECK: %[[READY:.+]] = call zeroext i1 @_ZN13suspend_maybe11await_readyEv(%struct.suspend_maybe* %[[AWAITABLE]]) // CHECK: %[[READY:.+]] = call zeroext i1 @_ZN13suspend_maybe11await_readyEv(%struct.suspend_maybe* %[[AWAITABLE]])
// CHECK: br i1 %[[READY]], label %[[READY_BB:.+]], label %[[SUSPEND_BB:.+]] // CHECK: br i1 %[[READY]], label %[[READY_BB:.+]], label %[[SUSPEND_BB:.+]]
// If we are suspending: // If we are suspending:
// --------------------- // ---------------------
▲ Show 20 Lines • Show All 50 Lines • ▼ Show 20 Lines
// FIXME: Once the cleanup code is in, add testing that destructors for Aggr // FIXME: Once the cleanup code is in, add testing that destructors for Aggr
// are invoked properly on the cleanup branches. // are invoked properly on the cleanup branches.
// CHECK-LABEL: @TestAggr( // CHECK-LABEL: @TestAggr(
extern "C" void TestAggr() { extern "C" void TestAggr() {
UseAggr(co_await AggrAwaiter{}); UseAggr(co_await AggrAwaiter{});
Whatever(); Whatever();
// CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret %[[AwaitResume:.+]], // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret align 4 %[[AwaitResume:.+]],
// CHECK: call void @UseAggr(%struct.Aggr* dereferenceable(12) %[[AwaitResume]]) // CHECK: call void @UseAggr(%struct.Aggr* dereferenceable(12) %[[AwaitResume]])
// CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume]]) // CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume]])
// CHECK: call void @Whatever() // CHECK: call void @Whatever()
co_await AggrAwaiter{}; co_await AggrAwaiter{};
Whatever(); Whatever();
// CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret %[[AwaitResume2:.+]], // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret align 4 %[[AwaitResume2:.+]],
// CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume2]]) // CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume2]])
// CHECK: call void @Whatever() // CHECK: call void @Whatever()
Aggr Val = co_await AggrAwaiter{}; Aggr Val = co_await AggrAwaiter{};
Whatever(); Whatever();
// CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret %[[AwaitResume3:.+]], // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret align 4 %[[AwaitResume3:.+]],
// CHECK: call void @Whatever() // CHECK: call void @Whatever()
// CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume3]]) // CHECK: call void @_ZN4AggrD1Ev(%struct.Aggr* %[[AwaitResume3]])
} }
struct ScalarAwaiter { struct ScalarAwaiter {
template <typename F> void await_suspend(F); template <typename F> void await_suspend(F);
bool await_ready(); bool await_ready();
int await_resume(); int await_resume();
Show All 26 Lines
// CHECK-LABEL: @TestOpAwait( // CHECK-LABEL: @TestOpAwait(
extern "C" void TestOpAwait() { extern "C" void TestOpAwait() {
co_await MyInt(42); co_await MyInt(42);
// CHECK: call void @_Zaw5MyInt(i32 42) // CHECK: call void @_Zaw5MyInt(i32 42)
// CHECK: call i32 @_ZN13ScalarAwaiter12await_resumeEv(%struct.ScalarAwaiter* % // CHECK: call i32 @_ZN13ScalarAwaiter12await_resumeEv(%struct.ScalarAwaiter* %
co_await MyAgg{}; co_await MyAgg{};
// CHECK: call void @_ZN5MyAggawEv(%struct.MyAgg* % // CHECK: call void @_ZN5MyAggawEv(%struct.MyAgg* %
// CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret % // CHECK: call void @_ZN11AggrAwaiter12await_resumeEv(%struct.Aggr* sret align 4 %
} }
// CHECK-LABEL: EndlessLoop( // CHECK-LABEL: EndlessLoop(
extern "C" void EndlessLoop() { extern "C" void EndlessLoop() {
// CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin( // CHECK: %[[FRAME:.+]] = call i8* @llvm.coro.begin(
// See if initial_suspend was issued: // See if initial_suspend was issued:
// ---------------------------------- // ----------------------------------
▲ Show 20 Lines • Show All 82 Lines • Show Last 20 Lines

clang/test/CodeGenCoroutines/coro-gro-nrvo.cpp

Show All 28 Lines
struct coro { struct coro {
using promise_type = promise_type<coro>; using promise_type = promise_type<coro>;
coro(coro const&); coro(coro const&);
struct Impl; struct Impl;
Impl *impl; Impl *impl;
}; };
// Verify that the NRVO is applied to the Gro object. // Verify that the NRVO is applied to the Gro object.
// CHECK-LABEL: define void @_Z1fi(%struct.coro* noalias sret %agg.result, i32 %0) // CHECK-LABEL: define void @_Z1fi(%struct.coro* noalias sret align 8 %agg.result, i32 %0)
coro f(int) { coro f(int) {
// CHECK: %call = call noalias nonnull i8* @_Znwm( // CHECK: %call = call noalias nonnull i8* @_Znwm(
// CHECK-NEXT: br label %[[CoroInit:.*]] // CHECK-NEXT: br label %[[CoroInit:.*]]
// CHECK: {{.*}}[[CoroInit]]: // CHECK: {{.*}}[[CoroInit]]:
// CHECK: store i1 false, i1* %gro.active // CHECK: store i1 false, i1* %gro.active
// CHECK: call void @{{.*get_return_objectEv}}(%struct.coro* sret %agg.result // CHECK: call void @{{.*get_return_objectEv}}(%struct.coro* sret align 8 %agg.result
// CHECK-NEXT: store i1 true, i1* %gro.active // CHECK-NEXT: store i1 true, i1* %gro.active
co_return; co_return;
} }
template <class RetObject> template <class RetObject>
struct promise_type_with_on_alloc_failure { struct promise_type_with_on_alloc_failure {
static RetObject get_return_object_on_allocation_failure(); static RetObject get_return_object_on_allocation_failure();
RetObject get_return_object(); RetObject get_return_object();
suspend_always initial_suspend(); suspend_always initial_suspend();
suspend_never final_suspend(); suspend_never final_suspend();
void return_void(); void return_void();
static void unhandled_exception(); static void unhandled_exception();
}; };
struct coro_two { struct coro_two {
using promise_type = promise_type_with_on_alloc_failure<coro_two>; using promise_type = promise_type_with_on_alloc_failure<coro_two>;
coro_two(coro_two const&); coro_two(coro_two const&);
struct Impl; struct Impl;
Impl *impl; Impl *impl;
}; };
// Verify that the NRVO is applied to the Gro object. // Verify that the NRVO is applied to the Gro object.
// CHECK-LABEL: define void @_Z1hi(%struct.coro_two* noalias sret %agg.result, i32 %0) // CHECK-LABEL: define void @_Z1hi(%struct.coro_two* noalias sret align 8 %agg.result, i32 %0)
coro_two h(int) { coro_two h(int) {
// CHECK: %call = call noalias i8* @_ZnwmRKSt9nothrow_t // CHECK: %call = call noalias i8* @_ZnwmRKSt9nothrow_t
// CHECK-NEXT: %[[CheckNull:.*]] = icmp ne i8* %call, null // CHECK-NEXT: %[[CheckNull:.*]] = icmp ne i8* %call, null
// CHECK-NEXT: br i1 %[[CheckNull]], label %[[InitOnSuccess:.*]], label %[[InitOnFailure:.*]] // CHECK-NEXT: br i1 %[[CheckNull]], label %[[InitOnSuccess:.*]], label %[[InitOnFailure:.*]]
// CHECK: {{.*}}[[InitOnFailure]]: // CHECK: {{.*}}[[InitOnFailure]]:
// CHECK-NEXT: call void @{{.*get_return_object_on_allocation_failureEv}}(%struct.coro_two* sret %agg.result // CHECK-NEXT: call void @{{.*get_return_object_on_allocation_failureEv}}(%struct.coro_two* sret align 8 %agg.result
// CHECK-NEXT: br label %[[RetLabel:.*]] // CHECK-NEXT: br label %[[RetLabel:.*]]
// CHECK: {{.*}}[[InitOnSuccess]]: // CHECK: {{.*}}[[InitOnSuccess]]:
// CHECK: store i1 false, i1* %gro.active // CHECK: store i1 false, i1* %gro.active
// CHECK: call void @{{.*get_return_objectEv}}(%struct.coro_two* sret %agg.result // CHECK: call void @{{.*get_return_objectEv}}(%struct.coro_two* sret align 8 %agg.result
// CHECK-NEXT: store i1 true, i1* %gro.active // CHECK-NEXT: store i1 true, i1* %gro.active
// CHECK: [[RetLabel]]: // CHECK: [[RetLabel]]:
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
co_return; co_return;
} }

clang/test/CodeGenObjC/arc.m

Show First 20 Lines • Show All 492 Lines • ▼ Show 20 Lines
// CHECK-LABEL: define void @test71 // CHECK-LABEL: define void @test71
void test71(void) { void test71(void) {
// FIXME: It would be nice if the __destructor_8_s40 for the first call (and // FIXME: It would be nice if the __destructor_8_s40 for the first call (and
// the following lifetime.end) came before the second call. // the following lifetime.end) came before the second call.
// //
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1:[^ ]+]] to i8* // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1:[^ ]+]] to i8*
// CHECK: call void @llvm.lifetime.start.p0i8({{[^,]+}}, i8* %[[T]]) // CHECK: call void @llvm.lifetime.start.p0i8({{[^,]+}}, i8* %[[T]])
// CHECK: call void @getAggDtor(%struct.AggDtor* sret %[[TMP1]]) // CHECK: call void @getAggDtor(%struct.AggDtor* sret align 8 %[[TMP1]])
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2:[^ ]+]] to i8* // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2:[^ ]+]] to i8*
// CHECK: call void @llvm.lifetime.start.p0i8({{[^,]+}}, i8* %[[T]]) // CHECK: call void @llvm.lifetime.start.p0i8({{[^,]+}}, i8* %[[T]])
// CHECK: call void @getAggDtor(%struct.AggDtor* sret %[[TMP2]]) // CHECK: call void @getAggDtor(%struct.AggDtor* sret align 8 %[[TMP2]])
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2]] to i8** // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2]] to i8**
// CHECK: call void @__destructor_8_s40(i8** %[[T]]) // CHECK: call void @__destructor_8_s40(i8** %[[T]])
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2:[^ ]+]] to i8* // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP2:[^ ]+]] to i8*
// CHECK: call void @llvm.lifetime.end.p0i8({{[^,]+}}, i8* %[[T]]) // CHECK: call void @llvm.lifetime.end.p0i8({{[^,]+}}, i8* %[[T]])
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1]] to i8** // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1]] to i8**
// CHECK: call void @__destructor_8_s40(i8** %[[T]]) // CHECK: call void @__destructor_8_s40(i8** %[[T]])
// CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1:[^ ]+]] to i8* // CHECK: %[[T:[^ ]+]] = bitcast %struct.AggDtor* %[[TMP1:[^ ]+]] to i8*
// CHECK: call void @llvm.lifetime.end.p0i8({{[^,]+}}, i8* %[[T]]) // CHECK: call void @llvm.lifetime.end.p0i8({{[^,]+}}, i8* %[[T]])
▲ Show 20 Lines • Show All 44 Lines • Show Last 20 Lines

clang/test/CodeGenObjC/direct-method.m

Show First 20 Lines • Show All 114 Lines • ▼ Show 20 Lines
+ (struct my_complex_struct)classGetComplex __attribute__((objc_direct)) { + (struct my_complex_struct)classGetComplex __attribute__((objc_direct)) {
struct my_complex_struct st = {.a = 42}; struct my_complex_struct st = {.a = 42};
return st; return st;
// CHECK: ret i64 // CHECK: ret i64
} }
// CHECK-LABEL: define hidden void @"\01-[Root getAggregate]"( // CHECK-LABEL: define hidden void @"\01-[Root getAggregate]"(
- (struct my_aggregate_struct)getAggregate __attribute__((objc_direct)) { - (struct my_aggregate_struct)getAggregate __attribute__((objc_direct)) {
// CHECK: %struct.my_aggregate_struct* noalias sret [[RETVAL:%[^,]*]], // CHECK: %struct.my_aggregate_struct* noalias sret align 4 [[RETVAL:%[^,]*]],
// loading parameters // loading parameters
// CHECK-LABEL: entry: // CHECK-LABEL: entry:
// CHECK-NEXT: [[SELFADDR:%.*]] = alloca %0*, // CHECK-NEXT: [[SELFADDR:%.*]] = alloca %0*,
// CHECK-NEXT: [[_CMDADDR:%.*]] = alloca i8*, // CHECK-NEXT: [[_CMDADDR:%.*]] = alloca i8*,
// CHECK-NEXT: store %0* %{{.*}}, %0** [[SELFADDR]], // CHECK-NEXT: store %0* %{{.*}}, %0** [[SELFADDR]],
// CHECK-NEXT: store i8* %{{.*}}, i8** [[_CMDADDR]], // CHECK-NEXT: store i8* %{{.*}}, i8** [[_CMDADDR]],
▲ Show 20 Lines • Show All 106 Lines • Show Last 20 Lines

clang/test/CodeGenObjC/nontrivial-c-struct-exception.m

Show All 35 Lines
void testStrongException(void) { void testStrongException(void) {
calleeStrong(genStrong(), genStrong()); calleeStrong(genStrong(), genStrong());
} }
// CHECK: define void @testWeakException() // CHECK: define void @testWeakException()
// CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_WEAK]], align 8 // CHECK: %[[AGG_TMP:.*]] = alloca %[[STRUCT_WEAK]], align 8
// CHECK: %[[AGG_TMP1:.*]] = alloca %[[STRUCT_WEAK]], align 8 // CHECK: %[[AGG_TMP1:.*]] = alloca %[[STRUCT_WEAK]], align 8
// CHECK: call void @genWeak(%[[STRUCT_WEAK]]* sret %[[AGG_TMP]]) // CHECK: call void @genWeak(%[[STRUCT_WEAK]]* sret align 8 %[[AGG_TMP]])
// CHECK: invoke void @genWeak(%[[STRUCT_WEAK]]* sret %[[AGG_TMP1]]) // CHECK: invoke void @genWeak(%[[STRUCT_WEAK]]* sret align 8 %[[AGG_TMP1]])
// CHECK: call void @calleeWeak(%[[STRUCT_WEAK]]* %[[AGG_TMP]], %[[STRUCT_WEAK]]* %[[AGG_TMP1]]) // CHECK: call void @calleeWeak(%[[STRUCT_WEAK]]* %[[AGG_TMP]], %[[STRUCT_WEAK]]* %[[AGG_TMP1]])
// CHECK: ret void // CHECK: ret void
// CHECK: landingpad { i8*, i32 } // CHECK: landingpad { i8*, i32 }
// CHECK: %[[V3:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_TMP]] to i8**{{.*}}, !dbg [[ARTIFICIAL_LOC_2:![0-9]+]] // CHECK: %[[V3:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_TMP]] to i8**{{.*}}, !dbg [[ARTIFICIAL_LOC_2:![0-9]+]]
// CHECK: call void @__destructor_8_w8(i8** %[[V3]]){{.*}}, !dbg [[ARTIFICIAL_LOC_2]] // CHECK: call void @__destructor_8_w8(i8** %[[V3]]){{.*}}, !dbg [[ARTIFICIAL_LOC_2]]
// CHECK: br label // CHECK: br label
Show All 12 Lines

clang/test/CodeGenObjC/objc-non-trivial-struct-nrvo.m

Show All 31 Lines
Trivial testTrivial(void) { Trivial testTrivial(void) {
Trivial a; Trivial a;
func0(&a); func0(&a);
return a; return a;
} }
void func1(TrivialBig *); void func1(TrivialBig *);
// CHECK: define void @testTrivialBig(%[[STRUCT_TRIVIALBIG]]* noalias sret %[[AGG_RESULT:.*]]) // CHECK: define void @testTrivialBig(%[[STRUCT_TRIVIALBIG]]* noalias sret align 4 %[[AGG_RESULT:.*]])
// CHECK: call void @func1(%[[STRUCT_TRIVIALBIG]]* %[[AGG_RESULT]]) // CHECK: call void @func1(%[[STRUCT_TRIVIALBIG]]* %[[AGG_RESULT]])
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
TrivialBig testTrivialBig(void) { TrivialBig testTrivialBig(void) {
TrivialBig a; TrivialBig a;
func1(&a); func1(&a);
return a; return a;
} }
Show All 15 Lines
// CHECK: %[[V2:.*]] = load i8*, i8** %[[COERCE_DIVE]], align 8 // CHECK: %[[V2:.*]] = load i8*, i8** %[[COERCE_DIVE]], align 8
// CHECK: ret i8* %[[V2]] // CHECK: ret i8* %[[V2]]
Strong testStrong(void) { Strong testStrong(void) {
Strong a; Strong a;
return a; return a;
} }
// CHECK: define void @testWeak(%[[STRUCT_WEAK]]* noalias sret %[[AGG_RESULT:.*]]) // CHECK: define void @testWeak(%[[STRUCT_WEAK]]* noalias sret align 8 %[[AGG_RESULT:.*]])
// CHECK: %[[NRVO:.*]] = alloca i1, align 1 // CHECK: %[[NRVO:.*]] = alloca i1, align 1
// CHECK: %[[V0:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8** // CHECK: %[[V0:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8**
// CHECK: call void @__default_constructor_8_w0(i8** %[[V0]]) // CHECK: call void @__default_constructor_8_w0(i8** %[[V0]])
// CHECK: store i1 true, i1* %[[NRVO]], align 1 // CHECK: store i1 true, i1* %[[NRVO]], align 1
// CHECK: %[[NRVO_VAL:.*]] = load i1, i1* %[[NRVO]], align 1 // CHECK: %[[NRVO_VAL:.*]] = load i1, i1* %[[NRVO]], align 1
// CHECK: br i1 %[[NRVO_VAL]], // CHECK: br i1 %[[NRVO_VAL]],
// CHECK: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8** // CHECK: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8**
Show All 19 Lines
Weak testWeak2(int c) { Weak testWeak2(int c) {
Weak a, b; Weak a, b;
if (c) if (c)
return a; return a;
else else
return b; return b;
} }
// CHECK: define internal void @"\01-[C1 foo1]"(%[[STRUCT_WEAK]]* noalias sret %[[AGG_RESULT:.*]], %{{.*}}* %{{.*}}, i8* %{{.*}}) // CHECK: define internal void @"\01-[C1 foo1]"(%[[STRUCT_WEAK]]* noalias sret align 8 %[[AGG_RESULT:.*]], %{{.*}}* %{{.*}}, i8* %{{.*}})
// CHECK: %[[NRVO:.*]] = alloca i1, align 1 // CHECK: %[[NRVO:.*]] = alloca i1, align 1
// CHECK: %[[V0:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8** // CHECK: %[[V0:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8**
// CHECK: call void @__default_constructor_8_w0(i8** %[[V0]]) // CHECK: call void @__default_constructor_8_w0(i8** %[[V0]])
// CHECK: store i1 true, i1* %[[NRVO]], align 1 // CHECK: store i1 true, i1* %[[NRVO]], align 1
// CHECK: %[[NRVO_VAL:.*]] = load i1, i1* %[[NRVO]], align 1 // CHECK: %[[NRVO_VAL:.*]] = load i1, i1* %[[NRVO]], align 1
// CHECK: br i1 %[[NRVO_VAL]], // CHECK: br i1 %[[NRVO_VAL]],
// CHECK: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8** // CHECK: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8**
Show All 17 Lines

clang/test/CodeGenObjC/stret-1.m

// RUN: %clang_cc1 -fblocks -triple arm64-apple-darwin %s -emit-llvm -o - | FileCheck %s // RUN: %clang_cc1 -fblocks -triple arm64-apple-darwin %s -emit-llvm -o - | FileCheck %s
// rdar://12416433 // rdar://12416433
struct stret { int x[100]; }; struct stret { int x[100]; };
struct stret one = {{1}}; struct stret one = {{1}};
@interface Test @end @interface Test @end
@implementation Test @implementation Test
+(struct stret) method { return one; } +(struct stret) method { return one; }
@end @end
int main(int argc, const char **argv) int main(int argc, const char **argv)
{ {
struct stret s; struct stret s;
s = [(id)(argc&~255) method]; s = [(id)(argc&~255) method];
// CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret [[T0:%[^,]+]] // CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret align 4 [[T0:%[^,]+]]
// CHECK: [[T0P:%.*]] = bitcast %struct.stret* [[T0]] to i8* // CHECK: [[T0P:%.*]] = bitcast %struct.stret* [[T0]] to i8*
// CHECK: call void @llvm.memset.p0i8.i64(i8* align 4 [[T0P]], i8 0, i64 400, i1 false) // CHECK: call void @llvm.memset.p0i8.i64(i8* align 4 [[T0P]], i8 0, i64 400, i1 false)
s = [Test method]; s = [Test method];
// CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret [[T1:%[^,]+]] // CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret align 4 [[T1:%[^,]+]]
// CHECK-NOT: call void @llvm.memset.p0i8.i64( // CHECK-NOT: call void @llvm.memset.p0i8.i64(
[(id)(argc&~255) method]; [(id)(argc&~255) method];
// CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret [[T1:%[^,]+]] // CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret align 4 [[T1:%[^,]+]]
// CHECK-NOT: call void @llvm.memset.p0i8.i64( // CHECK-NOT: call void @llvm.memset.p0i8.i64(
[Test method]; [Test method];
// CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret [[T1:%[^,]+]] // CHECK: call void bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to void (%struct.stret*, i8*, i8*)*)(%struct.stret* sret align 4 [[T1:%[^,]+]]
// CHECK-NOT: call void @llvm.memset.p0i8.i64( // CHECK-NOT: call void @llvm.memset.p0i8.i64(
} }

clang/test/CodeGenObjC/weak-in-c-struct.m

Show First 20 Lines • Show All 173 Lines • ▼ Show 20 Lines
// COMMON: call void @__copy_constructor_{{.*}}(i8** %[[V1]], i8** %[[V2]]) // COMMON: call void @__copy_constructor_{{.*}}(i8** %[[V1]], i8** %[[V2]])
// COMMON: call void @calleeWeak(%[[STRUCT_WEAK]]* %[[AGG_TMP]]) // COMMON: call void @calleeWeak(%[[STRUCT_WEAK]]* %[[AGG_TMP]])
// COMMON-NEXT: ret // COMMON-NEXT: ret
void test_argument_Weak(Weak *a) { void test_argument_Weak(Weak *a) {
calleeWeak(*a); calleeWeak(*a);
} }
// COMMON: define void @test_return_Weak(%[[STRUCT_WEAK]]* noalias sret %[[AGG_RESULT:.*]], %[[STRUCT_WEAK]]* %[[A:.*]]) // COMMON: define void @test_return_Weak(%[[STRUCT_WEAK]]* noalias sret align {{.*}} %[[AGG_RESULT:.*]], %[[STRUCT_WEAK]]* %[[A:.*]])
// COMMON: %[[A_ADDR:.*]] = alloca %[[STRUCT_WEAK]]* // COMMON: %[[A_ADDR:.*]] = alloca %[[STRUCT_WEAK]]*
// COMMON: store %[[STRUCT_WEAK]]* %[[A]], %[[STRUCT_WEAK]]** %[[A_ADDR]] // COMMON: store %[[STRUCT_WEAK]]* %[[A]], %[[STRUCT_WEAK]]** %[[A_ADDR]]
// COMMON: %[[V0:.*]] = load %[[STRUCT_WEAK]]*, %[[STRUCT_WEAK]]** %[[A_ADDR]] // COMMON: %[[V0:.*]] = load %[[STRUCT_WEAK]]*, %[[STRUCT_WEAK]]** %[[A_ADDR]]
// COMMON: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8** // COMMON: %[[V1:.*]] = bitcast %[[STRUCT_WEAK]]* %[[AGG_RESULT]] to i8**
// COMMON: %[[V2:.*]] = bitcast %[[STRUCT_WEAK]]* %[[V0]] to i8** // COMMON: %[[V2:.*]] = bitcast %[[STRUCT_WEAK]]* %[[V0]] to i8**
// COMMON: call void @__copy_constructor_{{.*}}(i8** %[[V1]], i8** %[[V2]]) // COMMON: call void @__copy_constructor_{{.*}}(i8** %[[V1]], i8** %[[V2]])
// COMMON: ret void // COMMON: ret void
Weak test_return_Weak(Weak *a) { Weak test_return_Weak(Weak *a) {
return *a; return *a;
} }

clang/test/CodeGenObjCXX/objc-struct-cxx-abi.mm

Show First 20 Lines • Show All 84 Lines • ▼ Show 20 Lines
// CHECK: call void @_Z19testParamStrongWeak10StrongWeak(%[[STRUCT_STRONGWEAK]]* %[[AGG_TMP]]) // CHECK: call void @_Z19testParamStrongWeak10StrongWeak(%[[STRUCT_STRONGWEAK]]* %[[AGG_TMP]])
// CHECK-NOT: call // CHECK-NOT: call
// CHECK: ret void // CHECK: ret void
void testCallStrongWeak(StrongWeak *a) { void testCallStrongWeak(StrongWeak *a) {
testParamStrongWeak(*a); testParamStrongWeak(*a);
} }
// CHECK: define void @_Z20testReturnStrongWeakP10StrongWeak(%[[STRUCT_STRONGWEAK:.*]]* noalias sret %[[AGG_RESULT:.*]], %[[STRUCT_STRONGWEAK]]* %[[A:.*]]) // CHECK: define void @_Z20testReturnStrongWeakP10StrongWeak(%[[STRUCT_STRONGWEAK:.*]]* noalias sret align 8 %[[AGG_RESULT:.*]], %[[STRUCT_STRONGWEAK]]* %[[A:.*]])
// CHECK: %[[A_ADDR:.*]] = alloca %[[STRUCT_STRONGWEAK]]*, align 8 // CHECK: %[[A_ADDR:.*]] = alloca %[[STRUCT_STRONGWEAK]]*, align 8
// CHECK: store %[[STRUCT_STRONGWEAK]]* %[[A]], %[[STRUCT_STRONGWEAK]]** %[[A_ADDR]], align 8 // CHECK: store %[[STRUCT_STRONGWEAK]]* %[[A]], %[[STRUCT_STRONGWEAK]]** %[[A_ADDR]], align 8
// CHECK: %[[V0:.*]] = load %[[STRUCT_STRONGWEAK]]*, %[[STRUCT_STRONGWEAK]]** %[[A_ADDR]], align 8 // CHECK: %[[V0:.*]] = load %[[STRUCT_STRONGWEAK]]*, %[[STRUCT_STRONGWEAK]]** %[[A_ADDR]], align 8
// CHECK: %[[CALL:.*]] = call %[[STRUCT_STRONGWEAK]]* @_ZN10StrongWeakC1ERKS_(%[[STRUCT_STRONGWEAK]]* %[[AGG_RESULT]], %[[STRUCT_STRONGWEAK]]* dereferenceable(16) %[[V0]]) // CHECK: %[[CALL:.*]] = call %[[STRUCT_STRONGWEAK]]* @_ZN10StrongWeakC1ERKS_(%[[STRUCT_STRONGWEAK]]* %[[AGG_RESULT]], %[[STRUCT_STRONGWEAK]]* dereferenceable(16) %[[V0]])
// CHECK: ret void // CHECK: ret void
StrongWeak testReturnStrongWeak(StrongWeak *a) { StrongWeak testReturnStrongWeak(StrongWeak *a) {
return *a; return *a;
▲ Show 20 Lines • Show All 79 Lines • Show Last 20 Lines

clang/test/CodeGenOpenCL/addr-space-struct-arg.cl

Show All 37 Linesstruct LargeStructTwoMember {
int2 x[40]; int2 x[40];
int2 y[20]; int2 y[20];
}; };
#if __OPENCL_C_VERSION__ >= 200 #if __OPENCL_C_VERSION__ >= 200
struct LargeStructOneMember g_s; struct LargeStructOneMember g_s;
#endif #endif
// X86-LABEL: define void @foo(%struct.Mat4X4* noalias sret %agg.result, %struct.Mat3X3* byval(%struct.Mat3X3) align 4 %in) // X86-LABEL: define void @foo(%struct.Mat4X4* noalias sret align 4 %agg.result, %struct.Mat3X3* byval(%struct.Mat3X3) align 4 %in)
// AMDGCN-LABEL: define %struct.Mat4X4 @foo([9 x i32] %in.coerce) // AMDGCN-LABEL: define %struct.Mat4X4 @foo([9 x i32] %in.coerce)
Mat4X4 __attribute__((noinline)) foo(Mat3X3 in) { Mat4X4 __attribute__((noinline)) foo(Mat3X3 in) {
Mat4X4 out; Mat4X4 out;
return out; return out;
} }
// COM-LABEL: define {{.*}} void @ker // COM-LABEL: define {{.*}} void @ker
// Expect two mem copies: one for the argument "in", and one for // Expect two mem copies: one for the argument "in", and one for
// the return value. // the return value.
// X86: call void @llvm.memcpy.p0i8.p1i8.i32(i8* // X86: call void @llvm.memcpy.p0i8.p1i8.i32(i8*
// X86: call void @llvm.memcpy.p1i8.p0i8.i32(i8 addrspace(1)* // X86: call void @llvm.memcpy.p1i8.p0i8.i32(i8 addrspace(1)*
// AMDGCN: load [9 x i32], [9 x i32] addrspace(1)* // AMDGCN: load [9 x i32], [9 x i32] addrspace(1)*
// AMDGCN: call %struct.Mat4X4 @foo([9 x i32] // AMDGCN: call %struct.Mat4X4 @foo([9 x i32]
// AMDGCN: call void @llvm.memcpy.p1i8.p5i8.i64(i8 addrspace(1)* // AMDGCN: call void @llvm.memcpy.p1i8.p5i8.i64(i8 addrspace(1)*
kernel void ker(global Mat3X3 *in, global Mat4X4 *out) { kernel void ker(global Mat3X3 *in, global Mat4X4 *out) {
out[0] = foo(in[1]); out[0] = foo(in[1]);
} }
// X86-LABEL: define void @foo_large(%struct.Mat64X64* noalias sret %agg.result, %struct.Mat32X32* byval(%struct.Mat32X32) align 4 %in) // X86-LABEL: define void @foo_large(%struct.Mat64X64* noalias sret align 4 %agg.result, %struct.Mat32X32* byval(%struct.Mat32X32) align 4 %in)
// AMDGCN-LABEL: define void @foo_large(%struct.Mat64X64 addrspace(5)* noalias sret %agg.result, %struct.Mat32X32 addrspace(5)* byval(%struct.Mat32X32) align 4 %in) // AMDGCN-LABEL: define void @foo_large(%struct.Mat64X64 addrspace(5)* noalias sret align 4 %agg.result, %struct.Mat32X32 addrspace(5)* byval(%struct.Mat32X32) align 4 %in)
Mat64X64 __attribute__((noinline)) foo_large(Mat32X32 in) { Mat64X64 __attribute__((noinline)) foo_large(Mat32X32 in) {
Mat64X64 out; Mat64X64 out;
return out; return out;
} }
// COM-LABEL: define {{.*}} void @ker_large // COM-LABEL: define {{.*}} void @ker_large
// Expect two mem copies: one for the argument "in", and one for // Expect two mem copies: one for the argument "in", and one for
// the return value. // the return value.
▲ Show 20 Lines • Show All 102 Lines • Show Last 20 Lines

clang/test/CodeGenOpenCL/amdgpu-abi-struct-coerce.cl

Show First 20 Lines • Show All 398 Lines • ▼ Show 20 Lines
// CHECK: define %struct.struct_arr16 @func_ret_struct_arr16() // CHECK: define %struct.struct_arr16 @func_ret_struct_arr16()
// CHECK: ret %struct.struct_arr16 zeroinitializer // CHECK: ret %struct.struct_arr16 zeroinitializer
struct_arr16 func_ret_struct_arr16() struct_arr16 func_ret_struct_arr16()
{ {
struct_arr16 s = { 0 }; struct_arr16 s = { 0 };
return s; return s;
} }
// CHECK: define void @func_ret_struct_arr32(%struct.struct_arr32 addrspace(5)* noalias nocapture sret %agg.result) // CHECK: define void @func_ret_struct_arr32(%struct.struct_arr32 addrspace(5)* noalias nocapture sret align 4 %agg.result)
struct_arr32 func_ret_struct_arr32() struct_arr32 func_ret_struct_arr32()
{ {
struct_arr32 s = { 0 }; struct_arr32 s = { 0 };
return s; return s;
} }
// CHECK: define void @func_ret_struct_arr33(%struct.struct_arr33 addrspace(5)* noalias nocapture sret %agg.result) // CHECK: define void @func_ret_struct_arr33(%struct.struct_arr33 addrspace(5)* noalias nocapture sret align 4 %agg.result)
struct_arr33 func_ret_struct_arr33() struct_arr33 func_ret_struct_arr33()
{ {
struct_arr33 s = { 0 }; struct_arr33 s = { 0 };
return s; return s;
} }
// CHECK: define %struct.struct_char_arr32 @func_ret_struct_char_arr32() // CHECK: define %struct.struct_char_arr32 @func_ret_struct_char_arr32()
struct_char_arr32 func_ret_struct_char_arr32() struct_char_arr32 func_ret_struct_char_arr32()
Show All 12 Lines
// CHECK: define %struct.different_size_type_pair @func_different_size_type_pair_ret() // CHECK: define %struct.different_size_type_pair @func_different_size_type_pair_ret()
different_size_type_pair func_different_size_type_pair_ret() different_size_type_pair func_different_size_type_pair_ret()
{ {
different_size_type_pair s = { 0 }; different_size_type_pair s = { 0 };
return s; return s;
} }
// CHECK: define void @func_flexible_array_ret(%struct.flexible_array addrspace(5)* noalias nocapture sret %agg.result) // CHECK: define void @func_flexible_array_ret(%struct.flexible_array addrspace(5)* noalias nocapture sret align 4 %agg.result)
flexible_array func_flexible_array_ret() flexible_array func_flexible_array_ret()
{ {
flexible_array s = { 0 }; flexible_array s = { 0 };
return s; return s;
} }
// CHECK: define void @func_reg_state_lo(<4 x i32> %arg0, <4 x i32> %arg1, <4 x i32> %arg2, i32 %arg3, i32 %s.coerce0, float %s.coerce1, i32 %s.coerce2) // CHECK: define void @func_reg_state_lo(<4 x i32> %arg0, <4 x i32> %arg1, <4 x i32> %arg2, i32 %arg3, i32 %s.coerce0, float %s.coerce1, i32 %s.coerce2)
void func_reg_state_lo(int4 arg0, int4 arg1, int4 arg2, int arg3, struct_arg_t s) { } void func_reg_state_lo(int4 arg0, int4 arg1, int4 arg2, int arg3, struct_arg_t s) { }
▲ Show 20 Lines • Show All 72 Lines • Show Last 20 Lines

clang/test/CodeGenOpenCLCXX/addrspace-of-this.cl

Show First 20 Lines • Show All 108 Lines • ▼ Show 20 Lines
// EXPL: call spir_func dereferenceable(4) %class.C addrspace(4)* @_ZNU3AS41CaSERU3AS4KS_(%class.C addrspace(4)* [[C2GEN]], %class.C addrspace(4)* dereferenceable(4) [[C1GEN]]) // EXPL: call spir_func dereferenceable(4) %class.C addrspace(4)* @_ZNU3AS41CaSERU3AS4KS_(%class.C addrspace(4)* [[C2GEN]], %class.C addrspace(4)* dereferenceable(4) [[C1GEN]])
// IMPL: [[C2GENVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[C2GEN]] to i8 addrspace(4)* // IMPL: [[C2GENVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[C2GEN]] to i8 addrspace(4)*
// IMPL: [[C1GENVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[C1GEN]] to i8 addrspace(4)* // IMPL: [[C1GENVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[C1GEN]] to i8 addrspace(4)*
// IMPL: call void @llvm.memcpy.p4i8.p4i8.i32(i8 addrspace(4)* {{.*}}[[C2GENVOID]], i8 addrspace(4)* {{.*}}[[C1GENVOID]] // IMPL: call void @llvm.memcpy.p4i8.p4i8.i32(i8 addrspace(4)* {{.*}}[[C2GENVOID]], i8 addrspace(4)* {{.*}}[[C1GENVOID]]
// Test the address space of 'this' when invoking the operator+ // Test the address space of 'this' when invoking the operator+
// COMMON: [[C1GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c1 to %class.C addrspace(4)* // COMMON: [[C1GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c1 to %class.C addrspace(4)*
// COMMON: [[C2GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c2 to %class.C addrspace(4)* // COMMON: [[C2GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c2 to %class.C addrspace(4)*
// COMMON: call spir_func void @_ZNU3AS41CplERU3AS4KS_(%class.C* sret %c3, %class.C addrspace(4)* [[C1GEN]], %class.C addrspace(4)* dereferenceable(4) [[C2GEN]]) // COMMON: call spir_func void @_ZNU3AS41CplERU3AS4KS_(%class.C* sret align 4 %c3, %class.C addrspace(4)* [[C1GEN]], %class.C addrspace(4)* dereferenceable(4) [[C2GEN]])
// Test the address space of 'this' when invoking the move constructor // Test the address space of 'this' when invoking the move constructor
// COMMON: [[C4GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c4 to %class.C addrspace(4)* // COMMON: [[C4GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c4 to %class.C addrspace(4)*
// COMMON: [[CALL:%call[0-9]+]] = call spir_func dereferenceable(4) %class.C addrspace(4)* @_Z3foov() // COMMON: [[CALL:%call[0-9]+]] = call spir_func dereferenceable(4) %class.C addrspace(4)* @_Z3foov()
// EXPL: call spir_func void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C4GEN]], %class.C addrspace(4)* dereferenceable(4) [[CALL]]) // EXPL: call spir_func void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C4GEN]], %class.C addrspace(4)* dereferenceable(4) [[CALL]])
// IMPL: [[C4VOID:%[0-9]+]] = bitcast %class.C* %c4 to i8* // IMPL: [[C4VOID:%[0-9]+]] = bitcast %class.C* %c4 to i8*
// IMPL: [[CALLVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[CALL]] to i8 addrspace(4)* // IMPL: [[CALLVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[CALL]] to i8 addrspace(4)*
// IMPL: call void @llvm.memcpy.p0i8.p4i8.i32(i8* {{.*}}[[C4VOID]], i8 addrspace(4)* {{.*}}[[CALLVOID]] // IMPL: call void @llvm.memcpy.p0i8.p4i8.i32(i8* {{.*}}[[C4VOID]], i8 addrspace(4)* {{.*}}[[CALLVOID]]
// Test the address space of 'this' when invoking the move assignment // Test the address space of 'this' when invoking the move assignment
// COMMON: [[C5GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c5 to %class.C addrspace(4)* // COMMON: [[C5GEN:%[.a-z0-9]+]] = addrspacecast %class.C* %c5 to %class.C addrspace(4)*
// COMMON: [[CALL:%call[0-9]+]] = call spir_func dereferenceable(4) %class.C addrspace(4)* @_Z3foov() // COMMON: [[CALL:%call[0-9]+]] = call spir_func dereferenceable(4) %class.C addrspace(4)* @_Z3foov()
// EXPL: call spir_func void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C5GEN]], %class.C addrspace(4)* dereferenceable(4) [[CALL]]) // EXPL: call spir_func void @_ZNU3AS41CC1EOU3AS4S_(%class.C addrspace(4)* [[C5GEN]], %class.C addrspace(4)* dereferenceable(4) [[CALL]])
// IMPL: [[C5VOID:%[0-9]+]] = bitcast %class.C* %c5 to i8* // IMPL: [[C5VOID:%[0-9]+]] = bitcast %class.C* %c5 to i8*
// IMPL: [[CALLVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[CALL]] to i8 addrspace(4)* // IMPL: [[CALLVOID:%[0-9]+]] = bitcast %class.C addrspace(4)* [[CALL]] to i8 addrspace(4)*
// IMPL: call void @llvm.memcpy.p0i8.p4i8.i32(i8* {{.*}}[[C5VOID]], i8 addrspace(4)* {{.*}}[[CALLVOID]] // IMPL: call void @llvm.memcpy.p0i8.p4i8.i32(i8* {{.*}}[[C5VOID]], i8 addrspace(4)* {{.*}}[[CALLVOID]]
// Tests address space of inline members // Tests address space of inline members
//COMMON: @_ZNU3AS41C3getEv(%class.C addrspace(4)* %this) //COMMON: @_ZNU3AS41C3getEv(%class.C addrspace(4)* %this)
//COMMON: @_ZNU3AS41CplERU3AS4KS_(%class.C* noalias sret %agg.result, %class.C addrspace(4)* %this //COMMON: @_ZNU3AS41CplERU3AS4KS_(%class.C* noalias sret align 4 %agg.result, %class.C addrspace(4)* %this
#define TEST(AS) \ #define TEST(AS) \
__kernel void test##AS() { \ __kernel void test##AS() { \
AS C c; \ AS C c; \
int i = c.get(); \ int i = c.get(); \
C c1(c); \ C c1(c); \
C c2; \ C c2; \
c2 = c1; \ c2 = c1; \
} }
▲ Show 20 Lines • Show All 71 Lines • Show Last 20 Lines

clang/test/Modules/templates.mm

Show First 20 Lines • Show All 119 Lines • ▼ Show 20 Linesvoid testWithAttributes() {
auto a = make_with_attributes_left(); auto a = make_with_attributes_left();
auto b = make_with_attributes_right(); auto b = make_with_attributes_right();
static_assert(alignof(decltype(a)) == 2, ""); static_assert(alignof(decltype(a)) == 2, "");
static_assert(alignof(decltype(b)) == 2, ""); static_assert(alignof(decltype(b)) == 2, "");
} }
// Check that returnNonTrivial doesn't return Class0<S0> directly in registers. // Check that returnNonTrivial doesn't return Class0<S0> directly in registers.
// CHECK: declare void @_Z16returnNonTrivialv(%struct.Class0* sret) // CHECK: declare void @_Z16returnNonTrivialv(%struct.Class0* sret align 8)
@import template_nontrivial0; @import template_nontrivial0;
@import template_nontrivial1; @import template_nontrivial1;
S1::S1() : a(returnNonTrivial()) { S1::S1() : a(returnNonTrivial()) {
} }