// RUN: %clang -emit-llvm -S -o - %s | FileCheck %s
typedef float float4 __attribute__((ext_vector_type(4)));
typedef unsigned int uint4 __attribute__((ext_vector_type(4)));
// CHECK-LABEL: define void @clang_shufflevector_v_v(
void clang_shufflevector_v_v( float4* A, float4 x, uint4 mask ) {
// CHECK: [[MASK:%.*]] = and <4 x i32> {{%.*}}, <i32 3, i32 3, i32 3, i32 3>
// CHECK: [[I:%.*]] = extractelement <4 x i32> [[MASK]], i32 0
// CHECK: [[E:%.*]] = extractelement <4 x float> [[X:%.*]], i32 [[I]]
//
// Here is where ToT Clang code generation makes a mistake.
// It uses [[I]] as the insertion index instead of 0.
// Similarly on the remaining insertelement.
// CHECK: [[V:%[a-zA-Z0-9._]+]] = insertelement <4 x float> undef, float [[E]], i32 0
// CHECK: [[I:%.*]] = extractelement <4 x i32> [[MASK]], i32 1
// CHECK: [[E:%.*]] = extractelement <4 x float> [[X]], i32 [[I]]
// CHECK: [[V2:%.*]] = insertelement <4 x float> [[V]], float [[E]], i32 1
// CHECK: [[I:%.*]] = extractelement <4 x i32> [[MASK]], i32 2
// CHECK: [[E:%.*]] = extractelement <4 x float> [[X]], i32 [[I]]
// CHECK: [[V3:%.*]] = insertelement <4 x float> [[V2]], float [[E]], i32 2
// CHECK: [[I:%.*]] = extractelement <4 x i32> [[MASK]], i32 3
// CHECK: [[E:%.*]] = extractelement <4 x float> [[X]], i32 [[I]]
// CHECK: [[V4:%.*]] = insertelement <4 x float> [[V3]], float [[E]], i32 3
// CHECK: store <4 x float> [[V4]], <4 x float>* {{%.*}},
*A = __builtin_shufflevector( x, mask );
}
// CHECK-LABEL: define void @clang_shufflevector_v_v_c(
void clang_shufflevector_v_v_c( float4* A, float4 x, float4 y) {
void test(int i) {
char c = (char __attribute__((aligned(8)))) i; // expected-warning {{'aligned' attribute ignored when parsing type}}
}
// http://llvm.org/PR11082
//
// FIXME: This may or may not be the correct approach (no warning or error),
// but large amounts of Linux and FreeBSD code need this attribute to not be
// a hard error in order to work correctly.
void test2(int i) {
char c = (char __attribute__((may_alias))) i;
}
// vector size too large
int __attribute__ ((vector_size(8192))) x1; // expected-error {{vector size too large}}
typedef int __attribute__ ((ext_vector_type(8192))) x2; // expected-error {{vector size too large}}
// no support for vector enum type
enum { e_2 } x3 __attribute__((vector_size(64))); // expected-error {{invalid vector element type}}
int x4 __attribute__((ext_vector_type(64))); // expected-error {{'ext_vector_type' attribute only applies to types}}
// rdar://16492792
typedef __attribute__ ((ext_vector_type(32),__aligned__(32))) unsigned char uchar32;
void convert() {
uchar32 r = 0;
r.s[ 1234 ] = 1; // expected-error {{illegal vector component name 's'}}
}
// RUN: %clang_cc1 -fsyntax-only -verify %s
typedef __attribute__(( ext_vector_type(2) )) float float2;
typedef __attribute__(( ext_vector_type(3) )) float float3;
typedef __attribute__(( ext_vector_type(4) )) float float4;
typedef __attribute__(( ext_vector_type(16) )) float float16;
static float4 vec4_0 = (float4)0.5f;
static void test() {
float2 vec2, vec2_2;
float3 vec3;
float4 vec4, vec4_2, *vec4p;
float16 vec16;
float f;
vec2.z; // expected-error {{vector component access exceeds type 'float2'}}
vec2.xyzw; // expected-error {{vector component access exceeds type 'float2'}}
vec4.xyzw; // expected-warning {{expression result unused}}
vec4.xyzc; // expected-error {{illegal vector component name 'c'}}
vec4.s01z; // expected-error {{illegal vector component name 'z'}}
vec2 = vec4.s01; // legal, shorten
vec2 = vec4.S01; // legal, shorten
vec3 = vec4.xyz; // legal, shorten
f = vec2.x; // legal, shorten
f = vec4.xy.x; // legal, shorten
vec4_2.xyzx = vec4.xyzw; // expected-error {{vector is not assignable (contains duplicate components)}}
vec4_2.xyzz = vec4.xyzw; // expected-error {{vector is not assignable (contains duplicate components)}}
vec4_2.xyyw = vec4.xyzw; // expected-error {{vector is not assignable (contains duplicate components)}}
enum E8559831a a3 = value_d;
a3 = value_d;
}
void test26(int si, long sl) {
si = sl % sl; // expected-warning {{implicit conversion loses integer precision: 'long' to 'int'}}
si = sl % si;
si = si % sl;
si = si / sl;
si = sl / si; // expected-warning {{implicit conversion loses integer precision: 'long' to 'int'}}
}
// rdar://16502418
typedef unsigned short uint16_t;
typedef unsigned int uint32_t;
typedef __attribute__ ((ext_vector_type(16),__aligned__(32))) uint16_t ushort16;
typedef __attribute__ ((ext_vector_type( 8),__aligned__( 32))) uint32_t uint8;
void test27(ushort16 constants) {
uint8 pairedConstants = (uint8) constants;
ushort16 crCbScale = pairedConstants.s4; // expected-warning {{implicit conversion loses integer precision: 'uint32_t' (aka 'unsigned int') to 'ushort16'}}
ushort16 brBias = pairedConstants.s6; // expected-warning {{implicit conversion loses integer precision: 'uint32_t' (aka 'unsigned int') to 'ushort16'}}
}
float double2float_test1(double a) {
return a; // expected-warning {{implicit conversion loses floating-point precision: 'double' to 'float'}}
}
void double2float_test2(double a, float *b) {
*b += a; // expected-warning {{implicit conversion when assigning computation result loses floating-point precision: 'double' to 'float'}}
struct { int x; } F2;
volatile struct { int x; } vF2;
volatile struct { int x; } *vpF2;
struct { struct { int y; } x; } F3;
volatile struct { struct { int y; } x; } vF3;
struct { int x:3; } BF;
struct { volatile int x:3; } vBF;
typedef int v4si __attribute__ ((vector_size (16)));
v4si V;
volatile v4si vV;
typedef __attribute__(( ext_vector_type(4) )) int extv4;
extv4 VE;
volatile extv4 vVE;
volatile struct {int x;} aggFct(void);
typedef volatile int volatile_int;
volatile_int vtS;
int main() {
int i;
// load
i=S;
i=vS;
i=*pS;
void *ptrs[];
};
struct foo bar;
// PR4143
struct foo2 {
enum bar *bar;
};
struct foo2 foo2;
// Radar 7325611
// CHECK: !"0x16\00barfoo\00{{.*}}"
typedef int barfoo;
barfoo foo() {
}
// CHECK: __uint128_t
__uint128_t foo128 ()
{
__uint128_t int128 = 44;
return int128;
}
// CHECK: uint64x2_t
typedef unsigned long long uint64_t;
typedef uint64_t uint64x2_t __attribute__((ext_vector_type(2)));
uint64x2_t extvectbar[4];
int a : 14;
} __attribute__ ((__packed__)) bfx3_packed;
struct BitFields4 {
short : 16;
int a: 1;
long b : 7;
} bfx4;
struct BitFields4_packed {
short : 16;
int a: 1;
long b : 7;
} __attribute__ ((__packed__)) bfx4_packed;
typedef float float2 __attribute__((ext_vector_type(2)));
float2 float2x;
// Register "0" is currently an invalid register for global register variables.
// Use "esp" instead of "0".
// register int rix __asm__("0");
register int rix __asm__("esp");
int main() {
// CHECK: store atomic i32 1, i32* getelementptr inbounds ({ i32, i32 }, { i32, i32 }* @civ, i32 0, i32 1) monotonic,
#pragma omp atomic write
__imag(civ) = 1;
// CHECK: load i8, i8*
// CHECK: store atomic i8
#pragma omp atomic write
bx = bv;
typedef __attribute__((ext_vector_type(5))) __short5;
fn1() {
__short5 a;
a.y;
}
//===-- main.cpp ------------------------------------------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
typedef float float4 __attribute__((ext_vector_type(4)));
typedef unsigned char vec __attribute__((ext_vector_type(16)));
int main() {
float4 f4 = {1.25, 1.25, 2.50, 2.50};
vec v = (vec)f4;
return 0; // break here
}
// RUN: %clang_cc1 -triple armv7-apple-darwin -target-abi aapcs -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple armv7-apple-darwin -target-abi apcs-gnu -emit-llvm -o - %s | FileCheck -check-prefix=APCS-GNU %s
#include <stdarg.h>
typedef __attribute__(( ext_vector_type(2) )) int __int2;
typedef __attribute__(( ext_vector_type(3) )) char __char3;
typedef __attribute__(( ext_vector_type(5) )) char __char5;
typedef __attribute__(( ext_vector_type(9) )) char __char9;
typedef __attribute__(( ext_vector_type(19) )) char __char19;
typedef __attribute__(( ext_vector_type(3) )) short __short3;
typedef __attribute__(( ext_vector_type(5) )) short __short5;
// Passing legal vector types as varargs.
double varargs_vec_2i(int fixed, ...) {
// CHECK: varargs_vec_2i
// CHECK: [[VAR:%.*]] = alloca <2 x i32>, align 8
// CHECK: [[ALIGN:%.*]] = and i32 {{%.*}}, -8
// CHECK: [[AP_ALIGN:%.*]] = inttoptr i32 [[ALIGN]] to i8*
// CHECK: [[AP_NEXT:%.*]] = getelementptr inbounds i8, i8* [[AP_ALIGN]], i32 8
// CHECK: [[AP_CAST:%.*]] = bitcast i8* [[AP_ALIGN]] to <2 x i32>*
// CHECK: [[VEC:%.*]] = load <2 x i32>, <2 x i32>* [[AP_CAST]], align 8
// CHECK: store <2 x i32> [[VEC]], <2 x i32>* [[VAR]], align 8
// APCS-GNU: varargs_vec_2i
// APCS-GNU: [[VAR:%.*]] = alloca <2 x i32>, align 8
// APCS-GNU: [[AP:%.*]] = load i8*,
// APCS-GNU: [[AP_NEXT:%.*]] = getelementptr inbounds i8, i8* [[AP]], i32 8
// APCS-GNU: [[AP_CAST:%.*]] = bitcast i8* [[AP]] to <2 x i32>*
// APCS-GNU: [[VEC:%.*]] = load <2 x i32>, <2 x i32>* [[AP_CAST]], align 4
// APCS-GNU: store <2 x i32> [[VEC]], <2 x i32>* [[VAR]], align 8
va_list ap;
}
// As a Clang extension, _Complex can be decomposed.
float decompose_complex(_Complex float cf) {
static _Complex float scf;
auto &[sre, sim] = scf;
// ok, this is references initialized by constant expressions all the way down
static_assert(&sre == &__real scf);
static_assert(&sim == &__imag scf);
auto [re, im] = cf;
return re*re + im*im;
}
// As a Clang extension, vector types can be decomposed.
typedef float vf3 __attribute__((ext_vector_type(3)));
float decompose_vector(vf3 v) {
auto [x, y, z] = v;
auto *p = &x; // expected-error {{address of vector element requested}}
return x + y + z;
}
struct S { int a, b; };
constexpr int f(S s) {
auto &[a, b] = s;
return a * 10 + b;
}
static_assert(f({1, 2}) == 12);
constexpr bool g(S &&s) {
auto &[a, b] = s;
// RUN: %clang_cc1 -triple x86_64-apple-macos10.7.0 -fsyntax-only -verify -fno-lax-vector-conversions -Wconversion %s
typedef __attribute__(( ext_vector_type(2) )) float float2;
typedef __attribute__(( ext_vector_type(3) )) float float3;
typedef __attribute__(( ext_vector_type(4) )) int int4;
typedef __attribute__(( ext_vector_type(8) )) short short8;
typedef __attribute__(( ext_vector_type(4) )) float float4;
typedef float t3 __attribute__ ((vector_size (16)));
typedef __typeof__(sizeof(int)) size_t;
typedef unsigned long ulong2 __attribute__ ((ext_vector_type(2)));
typedef size_t stride4 __attribute__((ext_vector_type(4)));
static void test() {
float2 vec2;
float3 vec3;
float4 vec4, vec4_2;
int4 ivec4;
short8 ish8;
t3 vec4_3;
int *ptr;
int i;
vec3 += vec2; // expected-error {{can't convert between vector values of different size}}
vec4 += vec3; // expected-error {{can't convert between vector values of different size}}
vec4 = 5.0f;
vec4 = (float4)5.0f;
vec4 = (float4)5;
vec4 = (float4)vec4_3;
ivec4 = (int4)5.0f;
int j[42]; // expected-error {{redefinition of 'j' with a different type: 'int [42]' vs 'int [4]'}}
// rdar://6880104
_Decimal32 x; // expected-error {{GNU decimal type extension not supported}}
// rdar://6880951
int __attribute__ ((vector_size (8), vector_size (8))) v; // expected-error {{invalid vector element type}}
void test(int i) {
char c = (char __attribute__((aligned(8)))) i; // expected-warning {{'aligned' attribute ignored when parsing type}}
}
// http://llvm.org/PR11082
//
// FIXME: This may or may not be the correct approach (no warning or error),
// but large amounts of Linux and FreeBSD code need this attribute to not be
// a hard error in order to work correctly.
void test2(int i) {
char c = (char __attribute__((may_alias))) i;
}
// vector size too large
int __attribute__ ((vector_size(8192))) x1; // expected-error {{vector size too large}}
typedef int __attribute__ ((ext_vector_type(8192))) x2; // expected-error {{vector size too large}}
// no support for vector enum type
enum { e_2 } x3 __attribute__((vector_size(64))); // expected-error {{invalid vector element type}}
int x4 __attribute__((ext_vector_type(64))); // expected-error {{'ext_vector_type' attribute only applies to types}}
// X64: define x86_vectorcallcc <4 x float> @"\01hva4@@112"(%struct.HVA4 inreg %a.coerce, %struct.HVA2* %b, <4 x float> %c)
v4f32 __vectorcall hva5(struct HVA3 a, struct HVA3 b, v4f32 c, struct HVA2 d) {return d.y;}
// X32: define x86_vectorcallcc <4 x float> @"\01hva5@@144"(%struct.HVA3 inreg %a.coerce, %struct.HVA3* inreg %b, <4 x float> %c, %struct.HVA2 inreg %d.coerce)
// X64: define 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;}
// X32: define x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* inreg %b)
// X64: define x86_vectorcallcc %struct.HVA4 @"\01hva6@@128"(%struct.HVA4 inreg %a.coerce, %struct.HVA4* %b)
struct HVA5 __vectorcall hva7() {struct HVA5 a = {}; return a;}
// X32: define x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* inreg noalias sret %agg.result)
// X64: define x86_vectorcallcc void @"\01hva7@@0"(%struct.HVA5* noalias sret %agg.result)
v4f32 __vectorcall hva8(v4f32 a, v4f32 b, v4f32 c, v4f32 d, int e, v4f32 f) {return f;}
// X32: define x86_vectorcallcc <4 x float> @"\01hva8@@84"(<4 x float> %a, <4 x float> %b, <4 x float> %c, <4 x float> %d, i32 inreg %e, <4 x float> %f)
// X64: define 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;
struct OddSizeHVA { v3f32 x, y; };
void __vectorcall odd_size_hva(struct OddSizeHVA a) {}
// X32: define x86_vectorcallcc void @"\01odd_size_hva@@32"(%struct.OddSizeHVA inreg %a.coerce)
// X64: define x86_vectorcallcc void @"\01odd_size_hva@@32"(%struct.OddSizeHVA inreg %a.coerce)
// The Vectorcall ABI only allows passing the first 6 items in registers, so this shouldn't
// consider 'p7' as a register. Instead p5 gets put into the register on the second pass.
struct HFA2 __vectorcall AddParticles(struct HFA2 p1, float p2, struct HFA4 p3, int p4, struct HFA2 p5, float p6, float p7){ return p1;}
// X32: define x86_vectorcallcc %struct.HFA2 @"\01AddParticles@@80"(%struct.HFA2 inreg %p1.coerce, float %p2, %struct.HFA4* inreg %p3, i32 inreg %p4, %struct.HFA2 inreg %p5.coerce, float %p6, float %p7)
// X64: define x86_vectorcallcc %struct.HFA2 @"\01AddParticles@@96"(%struct.HFA2 inreg %p1.coerce, float %p2, %struct.HFA4* %p3, i32 %p4, %struct.HFA2 inreg %p5.coerce, float %p6, float %p7)
// CHECK: [[ERROR:%.*]] = alloca float*, align 4
// CHECK: [[TEMP:%.*]] = alloca swifterror float*, align 4
// CHECK: [[T0:%.*]] = load float*, float** [[ERROR]], align 4
// CHECK: store float* [[T0]], float** [[TEMP]], align 4
// CHECK: call [[SWIFTCC:swiftcc]] void @context_error_1(i32* swiftself [[X]], float** swifterror [[TEMP]])
// CHECK: [[T0:%.*]] = load float*, float** [[TEMP]], align 4
// CHECK: store float* [[T0]], float** [[ERROR]], align 4
SWIFTCALL void context_error_2(short s, CONTEXT int *self, ERROR float **error) {}
// CHECK-LABEL: define {{.*}} void @context_error_2(i16{{.*}}, i32* swiftself{{.*}}, float** swifterror)
/*****************************************************************************/
/********************************** LOWERING *********************************/
/*****************************************************************************/
typedef float float4 __attribute__((ext_vector_type(4)));
typedef float float8 __attribute__((ext_vector_type(8)));
typedef double double2 __attribute__((ext_vector_type(2)));
typedef double double4 __attribute__((ext_vector_type(4)));
typedef int int3 __attribute__((ext_vector_type(3)));
typedef int int4 __attribute__((ext_vector_type(4)));
typedef int int5 __attribute__((ext_vector_type(5)));
typedef int int8 __attribute__((ext_vector_type(8)));
typedef char char16 __attribute__((ext_vector_type(16)));
typedef short short8 __attribute__((ext_vector_type(8)));
typedef long long long2 __attribute__((ext_vector_type(2)));
#define TEST(TYPE) \
SWIFTCALL TYPE return_##TYPE(void) { \
TYPE result = {}; \
return result; \
// RUN: %clang_cc1 -triple x86_64-apple-macos10.7.0 -fsyntax-only -verify -fno-lax-vector-conversions %s
typedef __attribute__(( ext_vector_type(2) )) float float2;
typedef __attribute__(( ext_vector_type(3) )) float float3;
typedef __attribute__(( ext_vector_type(4) )) int int4;
typedef __attribute__(( ext_vector_type(8) )) short short8;
typedef __attribute__(( ext_vector_type(4) )) float float4;
typedef float t3 __attribute__ ((vector_size (16)));
typedef __typeof__(sizeof(int)) size_t;
typedef unsigned long ulong2 __attribute__ ((ext_vector_type(2)));
typedef size_t stride4 __attribute__((ext_vector_type(4)));
static void test() {
float2 vec2;
float3 vec3;
float4 vec4, vec4_2;
int4 ivec4;
short8 ish8;
t3 vec4_3;
int *ptr;
int i;
vec3 += vec2; // expected-error {{can't convert between vector values of different size}}
vec4 += vec3; // expected-error {{can't convert between vector values of different size}}
vec4 = 5.0f;
vec4 = (float4)5.0f;
vec4 = (float4)5;
vec4 = (float4)vec4_3;
ivec4 = (int4)5.0f;
// RUN: %clang_cc1 -fsyntax-only -verify -std=c++11 %s
static_assert(__is_literal(int), "fail");
static_assert(__is_literal_type(int), "fail"); // alternate spelling for GCC
static_assert(__is_literal(void*), "fail");
enum E { E1 };
static_assert(__is_literal(E), "fail");
static_assert(__is_literal(decltype(E1)), "fail");
typedef int IAR[10];
static_assert(__is_literal(IAR), "fail");
typedef int Vector __attribute__((vector_size(16)));
typedef int VectorExt __attribute__((ext_vector_type(4)));
static_assert(__is_literal(Vector), "fail");
static_assert(__is_literal(VectorExt), "fail");
// C++0x [basic.types]p10:
// A type is a literal type if it is:
// [...]
// -- a class type that has all of the following properties:
// -- it has a trivial destructor
// -- every constructor call and full-expression in the
// brace-or-equal-initializers for non-static data members (if an) is
// a constant expression,
// -- it is an aggregate type or has at least one constexpr constructor
// or constructor template that is not a copy or move constructor, and
// -- it has all non-static data members and base classes of literal
// types
struct Empty {};
struct LiteralType {
int x;
E e;
// RUN: %clang_cc1 -triple arm64-apple-ios7 -target-abi darwinpcs -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -triple aarch64-linux-android -emit-llvm -o - %s | FileCheck -check-prefix=ANDROID %s
#include <stdarg.h>
typedef __attribute__(( ext_vector_type(2) )) char __char2;
typedef __attribute__(( ext_vector_type(3) )) char __char3;
typedef __attribute__(( ext_vector_type(4) )) char __char4;
typedef __attribute__(( ext_vector_type(5) )) char __char5;
typedef __attribute__(( ext_vector_type(9) )) char __char9;
typedef __attribute__(( ext_vector_type(19) )) char __char19;
typedef __attribute__(( ext_vector_type(3) )) short __short3;
typedef __attribute__(( ext_vector_type(5) )) short __short5;
typedef __attribute__(( ext_vector_type(3) )) int __int3;
typedef __attribute__(( ext_vector_type(5) )) int __int5;
typedef __attribute__(( ext_vector_type(3) )) double __double3;
// Passing legal vector types as varargs. Check that we've allocated the appropriate size
double varargs_vec_2c(int fixed, ...) {
// ANDROID: varargs_vec_2c
// ANDROID: [[VAR:%.*]] = alloca <2 x i8>, align 2
// ANDROID: [[AP_NEXT:%.*]] = getelementptr inbounds i8, i8* [[AP_CUR:%.*]], i64 8
// ANDROID: bitcast i8* [[AP_CUR]] to <2 x i8>*
va_list ap;
double sum = fixed;
va_start(ap, fixed);
__char2 c3 = va_arg(ap, __char2);
sum = sum + c3.x + c3.y;
va_end(ap);
return sum;
}
extern void takes_struct_of_four_doubles_variadic(double a, struct_of_four_doubles b, struct_of_four_doubles c, double d, ...);
void test_struct_of_four_doubles_variadic(void) {
// CHECK: test_struct_of_four_doubles_variadic
// CHECK: call arm_aapcs_vfpcc void (double, [4 x i64], [4 x i64], double, ...)* @takes_struct_of_four_doubles_variadic(double {{.*}}, [4 x i64] {{.*}}, [4 x i64] {{.*}}, double {{.*}})
takes_struct_of_four_doubles_variadic(3.0, g_s4d, g_s4d, 4.0);
}
extern void takes_struct_with_backfill(float f1, double a, float f2, struct_of_four_doubles b, struct_of_four_doubles c, double d);
void test_struct_with_backfill(void) {
// CHECK: test_struct_with_backfill
// CHECK: call arm_aapcs_vfpcc void @takes_struct_with_backfill(float {{.*}}, double {{.*}}, float {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, [4 x float] undef, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}}, double {{.*}})
takes_struct_with_backfill(3.0, 3.1, 3.2, g_s4d, g_s4d, 4.0);
}
typedef __attribute__(( ext_vector_type(8) )) char __char8;
typedef __attribute__(( ext_vector_type(4) )) short __short4;
typedef struct {
__char8 a1;
__short4 a2;
__char8 a3;
__short4 a4;
} struct_of_vecs;
extern void takes_struct_of_vecs(double a, struct_of_vecs b, struct_of_vecs c, double d);
struct_of_vecs g_vec;
void test_struct_of_vecs(void) {
// CHECK: test_struct_of_vecs
// CHECK: call arm_aapcs_vfpcc void @takes_struct_of_vecs(double {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, [6 x float] undef, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, double {{.*}})
// CHECK64: test_struct_of_vecs
// CHECK64: call void @takes_struct_of_vecs(double {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, [3 x float] undef, <8 x i8> {{.*}}, <4 x i16> {{.*}}, <8 x i8> {{.*}}, <4 x i16> {{.*}}, double {{.*}})
// RUN: %clang_cc1 -fsyntax-only -verify %s
// Test template instantiation for Clang-specific features.
// ---------------------------------------------------------------------
// Vector types
// ---------------------------------------------------------------------
typedef __attribute__(( ext_vector_type(2) )) double double2;
typedef __attribute__(( ext_vector_type(4) )) double double4;
template<typename T>
struct ExtVectorAccess0 {
void f(T v1, double4 v2) {
v1.xy = v2.yx;
}
};
template struct ExtVectorAccess0<double2>;
template struct ExtVectorAccess0<double4>;
typedef __attribute__(( ext_vector_type(2) )) double double2;
template<typename T, typename U, int N, int M>
struct ShuffleVector0 {
void f(T t, U u, double2 a, double2 b) {
(void)__builtin_shufflevector(t, u, N, M); // expected-error{{index}}
(void)__builtin_shufflevector(a, b, N, M);
(void)__builtin_shufflevector(a, b, 2, 1);
}
};
*
* 3. Neither the name of the copyright holder nor the names of its contributors may be used to
* endorse or promote products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS
* OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
* COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
* GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
* NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
* OF THE POSSIBILITY OF SUCH DAMAGE.
*/
typedef double vec __attribute__((ext_vector_type(2)));
void scale_vector(double *arr, int len, double scale) {
vec *vector = (vec*) arr;
vec vscale = scale;
int vlen = len / 2;
int i;
for (i = 0; i < vlen; i++) {
vector[i] *= vscale;
}
i *= 2;
if (i < len) {
arr[i] *= scale;
}
}
// RUN: clang-cc -emit-llvm %s -o %t
typedef __attribute__(( ext_vector_type(4) )) float float4;
typedef __attribute__(( ext_vector_type(2) )) float float2;
typedef __attribute__(( ext_vector_type(4) )) int int4;
float4 foo = (float4){ 1.0, 2.0, 3.0, 4.0 };
const float4 bar = (float4){ 1.0, 2.0, 3.0, __builtin_inff() };
float4 test1(float4 V) {
return V.wzyx+V;
}
float2 vec2, vec2_2;
float4 vec4, vec4_2;
float f;
void test2() {
vec2 = vec4.xy; // shorten
f = vec2.x; // extract elt
vec4 = vec4.yyyy; // splat
vec2.x = f; // insert one.
vec2.yx = vec2; // reverse
}
void test3(float4 *out) {
*out = ((float4) {1.0f, 2.0f, 3.0f, 4.0f });
}
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "rs_core.rsh"
typedef unsigned long long ull;
typedef unsigned long long ull2 __attribute__((ext_vector_type(2)));
typedef unsigned long long ull3 __attribute__((ext_vector_type(3)));
typedef unsigned long long ull4 __attribute__((ext_vector_type(4)));
#define S_CLAMP(T) \
extern T __attribute__((overloadable)) clamp(T amount, T low, T high) { \
return amount < low ? low : (amount > high ? high : amount); \
}
//_CLAMP(float); implemented in .ll
S_CLAMP(double);
S_CLAMP(char);
S_CLAMP(uchar);
S_CLAMP(short);
S_CLAMP(ushort);
S_CLAMP(int);
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
//
#include "Barrier.h"
#include "Matrix2x2.h"
typedef int veci16 __attribute__((ext_vector_type(16)));
typedef float vecf16 __attribute__((ext_vector_type(16)));
veci16* const kFrameBufferAddress = (veci16*) 0x10000000;
const vecf16 kXOffsets = { 0.0f, 1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f,
8.0f, 9.0f, 10.0f, 11.0f, 12.0f, 13.0f, 14.0f, 15.0f };
extern unsigned int kImage[];
const int kImageWidth = 16;
const int kImageHeight = 16;
const int kBytesPerPixel = 4;
const int kScreenWidth = 640;
const int kScreenHeight = 480;
void dflush(void *ptr)
{
__asm("dflush %0" : : "r" (ptr));
// RUN: %clang_cc1 -fsyntax-only -verify %s
typedef __attribute__(( ext_vector_type(2) )) float float2;
typedef __attribute__(( ext_vector_type(4) )) int int4;
typedef __attribute__(( ext_vector_type(8) )) short short8;
typedef __attribute__(( ext_vector_type(4) )) float float4;
typedef float t3 __attribute__ ((vector_size (16)));
static void test() {
float2 vec2;
float4 vec4, vec4_2;
int4 ivec4;
short8 ish8;
t3 vec4_3;
int *ptr;
int i;
vec4 = 5.0f;
vec4 = (float4)5.0f;
vec4 = (float4)5;
vec4 = (float4)vec4_3;
ivec4 = (int4)5.0f;
ivec4 = (int4)5;
ivec4 = (int4)vec4_3;
i = (int)ivec4; // expected-error {{invalid conversion between vector type 'int4' and integer type 'int' of different size}}
i = ivec4; // expected-error {{assigning to 'int' from incompatible type 'int4'}}
ivec4 = (int4)ptr; // expected-error {{invalid conversion between vector type 'int4' and scalar type 'int *'}}
// CHECK: [[ERROR:%.*]] = alloca float*, align 4
// CHECK: [[TEMP:%.*]] = alloca swifterror float*, align 4
// CHECK: [[T0:%.*]] = load float*, float** [[ERROR]], align 4
// CHECK: store float* [[T0]], float** [[TEMP]], align 4
// CHECK: call [[SWIFTCC:swiftcc]] void @context_error_1(i32* swiftself [[X]], float** swifterror [[TEMP]])
// CHECK: [[T0:%.*]] = load float*, float** [[TEMP]], align 4
// CHECK: store float* [[T0]], float** [[ERROR]], align 4
SWIFTCALL void context_error_2(short s, CONTEXT int *self, ERROR float **error) {}
// CHECK-LABEL: define {{.*}} void @context_error_2(i16{{.*}}, i32* swiftself{{.*}}, float** swifterror)
/*****************************************************************************/
/********************************** LOWERING *********************************/
/*****************************************************************************/
typedef float float4 __attribute__((ext_vector_type(4)));
typedef float float8 __attribute__((ext_vector_type(8)));
typedef double double2 __attribute__((ext_vector_type(2)));
typedef double double4 __attribute__((ext_vector_type(4)));
typedef int int3 __attribute__((ext_vector_type(3)));
typedef int int4 __attribute__((ext_vector_type(4)));
typedef int int5 __attribute__((ext_vector_type(5)));
typedef int int8 __attribute__((ext_vector_type(8)));
#define TEST(TYPE) \
SWIFTCALL TYPE return_##TYPE(void) { \
TYPE result = {}; \
return result; \
} \
SWIFTCALL void take_##TYPE(TYPE v) { \
} \
// RUN: %clang_cc1 -triple x86_64-unknown-unknown -emit-llvm %s -o - | FileCheck %s
typedef __attribute__(( ext_vector_type(4) )) float float4;
// CHECK: @test
void test()
{
float4 va;
va.hi[0] = 3.0;
// CHECK: [[VA:%.*]] = alloca <4 x float>
// CHECK: [[CONV:%.*]] = bitcast <4 x float>* [[VA]] to float*
// CHECK: [[ADD:%.*]] = getelementptr inbounds float, float* [[CONV]], i64 2
// CHECK: [[ARRIDX:%.*]] = getelementptr inbounds float, float* [[ADD]], i64 0
// CHECK: store float 3.000000e+00, float* [[ARRIDX]]
}
// CHECK-LABEL: nvvm_ldg
__device__ void nvvm_ldg(const void *p) {
// CHECK: call i8 @llvm.nvvm.ldg.global.i.i8.p0i8(i8* {{%[0-9]+}}, i32 1)
// CHECK: call i8 @llvm.nvvm.ldg.global.i.i8.p0i8(i8* {{%[0-9]+}}, i32 1)
__nvvm_ldg_c((const char *)p);
__nvvm_ldg_uc((const unsigned char *)p);
// CHECK: call i16 @llvm.nvvm.ldg.global.i.i16.p0i16(i16* {{%[0-9]+}}, i32 2)
// CHECK: call i16 @llvm.nvvm.ldg.global.i.i16.p0i16(i16* {{%[0-9]+}}, i32 2)
__nvvm_ldg_s((const short *)p);
__nvvm_ldg_us((const unsigned short *)p);
// CHECK: call i32 @llvm.nvvm.ldg.global.i.i32.p0i32(i32* {{%[0-9]+}}, i32 4)
// CHECK: call i32 @llvm.nvvm.ldg.global.i.i32.p0i32(i32* {{%[0-9]+}}, i32 4)
__nvvm_ldg_i((const int *)p);
__nvvm_ldg_ui((const unsigned int *)p);
// LP32: call i32 @llvm.nvvm.ldg.global.i.i32.p0i32(i32* {{%[0-9]+}}, i32 4)
// LP32: call i32 @llvm.nvvm.ldg.global.i.i32.p0i32(i32* {{%[0-9]+}}, i32 4)
// LP64: call i64 @llvm.nvvm.ldg.global.i.i64.p0i64(i64* {{%[0-9]+}}, i32 8)
// LP64: call i64 @llvm.nvvm.ldg.global.i.i64.p0i64(i64* {{%[0-9]+}}, i32 8)
__nvvm_ldg_l((const long *)p);
__nvvm_ldg_ul((const unsigned long *)p);
// CHECK: call float @llvm.nvvm.ldg.global.f.f32.p0f32(float* {{%[0-9]+}}, i32 4)
__nvvm_ldg_f((const float *)p);
// CHECK: call double @llvm.nvvm.ldg.global.f.f64.p0f64(double* {{%[0-9]+}}, i32 8)
__nvvm_ldg_d((const double *)p);
// In practice, the pointers we pass to __ldg will be aligned as appropriate
// for the CUDA <type>N vector types (e.g. short4), which are not the same as
// the LLVM vector types. However, each LLVM vector type has an alignment
// less than or equal to its corresponding CUDA type, so we're OK.
//
// PTX Interoperability section 2.2: "For a vector with an even number of
// elements, its alignment is set to number of elements times the alignment of
// its member: n*alignof(t)."
// CHECK: call <2 x i8> @llvm.nvvm.ldg.global.i.v2i8.p0v2i8(<2 x i8>* {{%[0-9]+}}, i32 2)
// CHECK: call <2 x i8> @llvm.nvvm.ldg.global.i.v2i8.p0v2i8(<2 x i8>* {{%[0-9]+}}, i32 2)
typedef char char2 __attribute__((ext_vector_type(2)));
typedef unsigned char uchar2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_c2((const char2 *)p);
__nvvm_ldg_uc2((const uchar2 *)p);
// CHECK: call <4 x i8> @llvm.nvvm.ldg.global.i.v4i8.p0v4i8(<4 x i8>* {{%[0-9]+}}, i32 4)
// CHECK: call <4 x i8> @llvm.nvvm.ldg.global.i.v4i8.p0v4i8(<4 x i8>* {{%[0-9]+}}, i32 4)
typedef char char4 __attribute__((ext_vector_type(4)));
typedef unsigned char uchar4 __attribute__((ext_vector_type(4)));
__nvvm_ldg_c4((const char4 *)p);
__nvvm_ldg_uc4((const uchar4 *)p);
// CHECK: call <2 x i16> @llvm.nvvm.ldg.global.i.v2i16.p0v2i16(<2 x i16>* {{%[0-9]+}}, i32 4)
// CHECK: call <2 x i16> @llvm.nvvm.ldg.global.i.v2i16.p0v2i16(<2 x i16>* {{%[0-9]+}}, i32 4)
typedef short short2 __attribute__((ext_vector_type(2)));
typedef unsigned short ushort2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_s2((const short2 *)p);
__nvvm_ldg_us2((const ushort2 *)p);
// CHECK: call <4 x i16> @llvm.nvvm.ldg.global.i.v4i16.p0v4i16(<4 x i16>* {{%[0-9]+}}, i32 8)
// CHECK: call <4 x i16> @llvm.nvvm.ldg.global.i.v4i16.p0v4i16(<4 x i16>* {{%[0-9]+}}, i32 8)
typedef short short4 __attribute__((ext_vector_type(4)));
typedef unsigned short ushort4 __attribute__((ext_vector_type(4)));
__nvvm_ldg_s4((const short4 *)p);
__nvvm_ldg_us4((const ushort4 *)p);
// CHECK: call <2 x i32> @llvm.nvvm.ldg.global.i.v2i32.p0v2i32(<2 x i32>* {{%[0-9]+}}, i32 8)
// CHECK: call <2 x i32> @llvm.nvvm.ldg.global.i.v2i32.p0v2i32(<2 x i32>* {{%[0-9]+}}, i32 8)
typedef int int2 __attribute__((ext_vector_type(2)));
typedef unsigned int uint2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_i2((const int2 *)p);
__nvvm_ldg_ui2((const uint2 *)p);
// CHECK: call <4 x i32> @llvm.nvvm.ldg.global.i.v4i32.p0v4i32(<4 x i32>* {{%[0-9]+}}, i32 16)
// CHECK: call <4 x i32> @llvm.nvvm.ldg.global.i.v4i32.p0v4i32(<4 x i32>* {{%[0-9]+}}, i32 16)
typedef int int4 __attribute__((ext_vector_type(4)));
typedef unsigned int uint4 __attribute__((ext_vector_type(4)));
__nvvm_ldg_i4((const int4 *)p);
__nvvm_ldg_ui4((const uint4 *)p);
// CHECK: call <2 x i64> @llvm.nvvm.ldg.global.i.v2i64.p0v2i64(<2 x i64>* {{%[0-9]+}}, i32 16)
// CHECK: call <2 x i64> @llvm.nvvm.ldg.global.i.v2i64.p0v2i64(<2 x i64>* {{%[0-9]+}}, i32 16)
typedef long long longlong2 __attribute__((ext_vector_type(2)));
typedef unsigned long long ulonglong2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_ll2((const longlong2 *)p);
__nvvm_ldg_ull2((const ulonglong2 *)p);
// CHECK: call <2 x float> @llvm.nvvm.ldg.global.f.v2f32.p0v2f32(<2 x float>* {{%[0-9]+}}, i32 8)
typedef float float2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_f2((const float2 *)p);
// CHECK: call <4 x float> @llvm.nvvm.ldg.global.f.v4f32.p0v4f32(<4 x float>* {{%[0-9]+}}, i32 16)
typedef float float4 __attribute__((ext_vector_type(4)));
__nvvm_ldg_f4((const float4 *)p);
// CHECK: call <2 x double> @llvm.nvvm.ldg.global.f.v2f64.p0v2f64(<2 x double>* {{%[0-9]+}}, i32 16)
typedef double double2 __attribute__((ext_vector_type(2)));
__nvvm_ldg_d2((const double2 *)p);
}
