maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/CodeGen/altivec.c

41 lines
2.2 KiB
C
Raw Normal View History

Remove the -faltivec alias option and replace it with -maltivec everywhere. The alias was only ever used on darwin and had some issues there, and isn't used in practice much. Also fixes a problem with -mno-altivec not turning off -maltivec. Also add a diagnostic for faltivec/fno-altivec that directs users to use maltivec options and include the altivec.h file explicitly. llvm-svn: 298449
2017-03-22 06:06:18 +08:00
// RUN: %clang_cc1 -target-feature +altivec -triple powerpc-unknown-unknown -emit-llvm %s -o - | FileCheck %s
Fix the constant evaluator for AltiVec-style vector literals so that the vector is filled with the given constant; we were just initializing the first element. llvm-svn: 105824
2010-06-12 01:54:15 +08:00
supported: AltiVec vector initialization with a single literal according to PIM section 2.5.1 - after initialization all elements have the value specified by the literal llvm-svn: 128375
2011-03-27 17:32:40 +08:00
// Check initialization
vector int test0 = (vector int)(1); // CHECK: @test0 = global <4 x i32> <i32 1, i32 1, i32 1, i32 1>
vector float test1 = (vector float)(1.0); // CHECK: @test1 = global <4 x float> <float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}>
Add support for bitcasts to vector type in Evaluate. llvm-svn: 147137
2011-12-22 11:51:45 +08:00
// CHECK: @v1 = global <16 x i8> <i8 0, i8 0, i8 0, i8 1, i8 0, i8 0, i8 0, i8 2, i8 0, i8 0, i8 0, i8 3, i8 0, i8 0, i8 0, i8 4>
Move vector bitcast handling in constant expressions from the expression evaluator into constant initializer handling / IRGen. The practical consequence of this is that the bitcast now lives in the constant's definition, rather than in its uses. The code in the constant expression evaluator was producing vectors of the wrong type and size (and possibly of the wrong value for a big-endian int-to-vector bitcast). We were getting away with this only because we don't yet support constant-folding of any expressions which inspect vector values. llvm-svn: 145981
2011-12-07 06:44:34 +08:00
vector char v1 = (vector char)((vector int)(1, 2, 3, 4));
Add support for bitcasts to vector type in Evaluate. llvm-svn: 147137
2011-12-22 11:51:45 +08:00
// CHECK: @v2 = global <16 x i8> <i8 63, i8 -128, i8 0, i8 0, i8 64, i8 0, i8 0, i8 0, i8 64, i8 64, i8 0, i8 0, i8 64, i8 -128, i8 0, i8 0>
Move vector bitcast handling in constant expressions from the expression evaluator into constant initializer handling / IRGen. The practical consequence of this is that the bitcast now lives in the constant's definition, rather than in its uses. The code in the constant expression evaluator was producing vectors of the wrong type and size (and possibly of the wrong value for a big-endian int-to-vector bitcast). We were getting away with this only because we don't yet support constant-folding of any expressions which inspect vector values. llvm-svn: 145981
2011-12-07 06:44:34 +08:00
vector char v2 = (vector char)((vector float)(1.0f, 2.0f, 3.0f, 4.0f));
Add support for bitcasts to vector type in Evaluate. llvm-svn: 147137
2011-12-22 11:51:45 +08:00
// CHECK: @v3 = global <16 x i8> <i8 0, i8 0, i8 0, i8 97, i8 0, i8 0, i8 0, i8 98, i8 0, i8 0, i8 0, i8 99, i8 0, i8 0, i8 0, i8 100>
Move vector bitcast handling in constant expressions from the expression evaluator into constant initializer handling / IRGen. The practical consequence of this is that the bitcast now lives in the constant's definition, rather than in its uses. The code in the constant expression evaluator was producing vectors of the wrong type and size (and possibly of the wrong value for a big-endian int-to-vector bitcast). We were getting away with this only because we don't yet support constant-folding of any expressions which inspect vector values. llvm-svn: 145981
2011-12-07 06:44:34 +08:00
vector char v3 = (vector char)((vector int)('a', 'b', 'c', 'd'));
Add support for bitcasts to vector type in Evaluate. llvm-svn: 147137
2011-12-22 11:51:45 +08:00
// CHECK: @v4 = global <4 x i32> <i32 16909060, i32 0, i32 0, i32 0>
Move vector bitcast handling in constant expressions from the expression evaluator into constant initializer handling / IRGen. The practical consequence of this is that the bitcast now lives in the constant's definition, rather than in its uses. The code in the constant expression evaluator was producing vectors of the wrong type and size (and possibly of the wrong value for a big-endian int-to-vector bitcast). We were getting away with this only because we don't yet support constant-folding of any expressions which inspect vector values. llvm-svn: 145981
2011-12-07 06:44:34 +08:00
vector int v4 = (vector char){1, 2, 3, 4};
supported: AltiVec vector initialization with a single literal according to PIM section 2.5.1 - after initialization all elements have the value specified by the literal llvm-svn: 128375
2011-03-27 17:32:40 +08:00
void test2()
{
vector int vi;
vector float vf;
vi = (vector int)(1); // CHECK: <i32 1, i32 1, i32 1, i32 1>
vf = (vector float)(1.0); // CHECK: <float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}>
vi = (vector int)(1, 2, 3, 4); // CHECK: <i32 1, i32 2, i32 3, i32 4>
vi = (vector int)(1, 2, 3, 4, 5); // CHECK: <i32 1, i32 2, i32 3, i32 4>
vi = (vector int){1}; // CHECK: <i32 1, i32 0, i32 0, i32 0>
vi = (vector int){1, 2}; // CHECK: <i32 1, i32 2, i32 0, i32 0>
vi = (vector int){1, 2, 3, 4}; // CHECK: <i32 1, i32 2, i32 3, i32 4>
}
// Check pre/post increment/decrement
void test3() {
vector int vi;
Don't emit nsw flags for vector operations; there's basically no benefit, and a lot of downside (like PR9850, which is about clang's xmmintrin.h making an unexpected transformation on an expression involving _mm_add_epi32). llvm-svn: 131000
2011-05-07 02:04:18 +08:00
vi++; // CHECK: add <4 x i32> {{.*}} <i32 1, i32 1, i32 1, i32 1>
supported: AltiVec vector initialization with a single literal according to PIM section 2.5.1 - after initialization all elements have the value specified by the literal llvm-svn: 128375
2011-03-27 17:32:40 +08:00
vector unsigned int vui;
--vui; // CHECK: add <4 x i32> {{.*}} <i32 -1, i32 -1, i32 -1, i32 -1>
vector float vf;
vf++; // CHECK: fadd <4 x float> {{.*}} <float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}, float 1.000000e+{{0+}}>
}