Author: Jon Chesterfield Date: 2021-08-05T18:46:57+01:00 New Revision: 509854b69cea0c9261ac21ceb22012a53e7a800b
URL: https://github.com/llvm/llvm-project/commit/509854b69cea0c9261ac21ceb22012a53e7a800b DIFF: https://github.com/llvm/llvm-project/commit/509854b69cea0c9261ac21ceb22012a53e7a800b.diff LOG: [clang] Replace asm with __asm__ in cuda header Asm is a gnu extension for C, so at present -fopenmp -std=c99 and similar fail to compile on nvptx, bug 51344 Changing to `__asm__` or `__asm` works for openmp, all three appear to work for cuda. Suggesting `__asm__` here as `__asm` is used by MSVC with different syntax, so this should make for better error diagnostics if the header is passed to a compiler other than clang. Reviewed By: tra, emankov Differential Revision: https://reviews.llvm.org/D107492 Added: Modified: clang/lib/Headers/__clang_cuda_device_functions.h Removed: ################################################################################ diff --git a/clang/lib/Headers/__clang_cuda_device_functions.h b/clang/lib/Headers/__clang_cuda_device_functions.h index f801e5426aa43..cc4e1a4dd96ad 100644 --- a/clang/lib/Headers/__clang_cuda_device_functions.h +++ b/clang/lib/Headers/__clang_cuda_device_functions.h @@ -34,10 +34,12 @@ __DEVICE__ unsigned long long __brevll(unsigned long long __a) { return __nv_brevll(__a); } #if defined(__cplusplus) -__DEVICE__ void __brkpt() { asm volatile("brkpt;"); } +__DEVICE__ void __brkpt() { __asm__ __volatile__("brkpt;"); } __DEVICE__ void __brkpt(int __a) { __brkpt(); } #else -__DEVICE__ void __attribute__((overloadable)) __brkpt(void) { asm volatile("brkpt;"); } +__DEVICE__ void __attribute__((overloadable)) __brkpt(void) { + __asm__ __volatile__("brkpt;"); +} __DEVICE__ void __attribute__((overloadable)) __brkpt(int __a) { __brkpt(); } #endif __DEVICE__ unsigned int __byte_perm(unsigned int __a, unsigned int __b, @@ -507,7 +509,7 @@ __DEVICE__ float __powf(float __a, float __b) { } // Parameter must have a known integer value. -#define __prof_trigger(__a) asm __volatile__("pmevent \t%0;" ::"i"(__a)) +#define __prof_trigger(__a) __asm__ __volatile__("pmevent \t%0;" ::"i"(__a)) __DEVICE__ int __rhadd(int __a, int __b) { return __nv_rhadd(__a, __b); } __DEVICE__ unsigned int __sad(int __a, int __b, unsigned int __c) { return __nv_sad(__a, __b, __c); @@ -526,7 +528,7 @@ __DEVICE__ float __tanf(float __a) { return __nv_fast_tanf(__a); } __DEVICE__ void __threadfence(void) { __nvvm_membar_gl(); } __DEVICE__ void __threadfence_block(void) { __nvvm_membar_cta(); }; __DEVICE__ void __threadfence_system(void) { __nvvm_membar_sys(); }; -__DEVICE__ void __trap(void) { asm volatile("trap;"); } +__DEVICE__ void __trap(void) { __asm__ __volatile__("trap;"); } __DEVICE__ unsigned int __uAtomicAdd(unsigned int *__p, unsigned int __v) { return __nvvm_atom_add_gen_i((int *)__p, __v); } @@ -1051,122 +1053,136 @@ __DEVICE__ unsigned int __bool2mask(unsigned int __a, int shift) { } __DEVICE__ unsigned int __vabs2(unsigned int __a) { unsigned int r; - asm("vabs diff 2.s32.s32.s32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(0), "r"(0)); + __asm__("vabs diff 2.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(0), "r"(0)); return r; } __DEVICE__ unsigned int __vabs4(unsigned int __a) { unsigned int r; - asm("vabs diff 4.s32.s32.s32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(0), "r"(0)); + __asm__("vabs diff 4.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(0), "r"(0)); return r; } __DEVICE__ unsigned int __vabs diff s2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 2.s32.s32.s32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 2.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vabs diff s4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 4.s32.s32.s32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 4.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vabs diff u2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 2.u32.u32.u32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vabs diff u4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 4.u32.u32.u32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vabsss2(unsigned int __a) { unsigned int r; - asm("vabs diff 2.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(0), "r"(0)); + __asm__("vabs diff 2.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(0), "r"(0)); return r; } __DEVICE__ unsigned int __vabsss4(unsigned int __a) { unsigned int r; - asm("vabs diff 4.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(0), "r"(0)); + __asm__("vabs diff 4.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(0), "r"(0)); return r; } __DEVICE__ unsigned int __vadd2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vadd4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vaddss2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd2.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd2.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vaddss4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd4.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd4.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vaddus2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd2.u32.u32.u32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd2.u32.u32.u32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vaddus4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vadd4.u32.u32.u32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vadd4.u32.u32.u32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vavgs2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vavrg2.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vavrg2.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vavgs4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vavrg4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vavrg4.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vavgu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vavrg2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vavrg2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vavgu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vavrg4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vavrg4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vseteq2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.eq %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.eq %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpeq2(unsigned int __a, unsigned int __b) { @@ -1174,7 +1190,9 @@ __DEVICE__ unsigned int __vcmpeq2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vseteq4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.eq %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.eq %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpeq4(unsigned int __a, unsigned int __b) { @@ -1182,7 +1200,9 @@ __DEVICE__ unsigned int __vcmpeq4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetges2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.s32.s32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.s32.s32.ge %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpges2(unsigned int __a, unsigned int __b) { @@ -1190,7 +1210,9 @@ __DEVICE__ unsigned int __vcmpges2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetges4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.s32.s32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.s32.s32.ge %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpges4(unsigned int __a, unsigned int __b) { @@ -1198,7 +1220,9 @@ __DEVICE__ unsigned int __vcmpges4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgeu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.ge %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgeu2(unsigned int __a, unsigned int __b) { @@ -1206,7 +1230,9 @@ __DEVICE__ unsigned int __vcmpgeu2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgeu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.ge %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.ge %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgeu4(unsigned int __a, unsigned int __b) { @@ -1214,7 +1240,9 @@ __DEVICE__ unsigned int __vcmpgeu4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgts2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.s32.s32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.s32.s32.gt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgts2(unsigned int __a, unsigned int __b) { @@ -1222,7 +1250,9 @@ __DEVICE__ unsigned int __vcmpgts2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgts4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.s32.s32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.s32.s32.gt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgts4(unsigned int __a, unsigned int __b) { @@ -1230,7 +1260,9 @@ __DEVICE__ unsigned int __vcmpgts4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgtu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.gt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgtu2(unsigned int __a, unsigned int __b) { @@ -1238,7 +1270,9 @@ __DEVICE__ unsigned int __vcmpgtu2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetgtu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.gt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.gt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpgtu4(unsigned int __a, unsigned int __b) { @@ -1246,7 +1280,9 @@ __DEVICE__ unsigned int __vcmpgtu4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetles2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.s32.s32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.s32.s32.le %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmples2(unsigned int __a, unsigned int __b) { @@ -1254,7 +1290,9 @@ __DEVICE__ unsigned int __vcmples2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetles4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.s32.s32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.s32.s32.le %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmples4(unsigned int __a, unsigned int __b) { @@ -1262,7 +1300,9 @@ __DEVICE__ unsigned int __vcmples4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetleu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.le %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpleu2(unsigned int __a, unsigned int __b) { @@ -1270,7 +1310,9 @@ __DEVICE__ unsigned int __vcmpleu2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetleu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.le %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.le %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpleu4(unsigned int __a, unsigned int __b) { @@ -1278,7 +1320,9 @@ __DEVICE__ unsigned int __vcmpleu4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetlts2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.s32.s32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.s32.s32.lt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmplts2(unsigned int __a, unsigned int __b) { @@ -1286,7 +1330,9 @@ __DEVICE__ unsigned int __vcmplts2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetlts4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.s32.s32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.s32.s32.lt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmplts4(unsigned int __a, unsigned int __b) { @@ -1294,7 +1340,9 @@ __DEVICE__ unsigned int __vcmplts4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetltu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.lt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpltu2(unsigned int __a, unsigned int __b) { @@ -1302,7 +1350,9 @@ __DEVICE__ unsigned int __vcmpltu2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetltu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.lt %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.lt %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpltu4(unsigned int __a, unsigned int __b) { @@ -1310,7 +1360,9 @@ __DEVICE__ unsigned int __vcmpltu4(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetne2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset2.u32.u32.ne %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset2.u32.u32.ne %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpne2(unsigned int __a, unsigned int __b) { @@ -1318,7 +1370,9 @@ __DEVICE__ unsigned int __vcmpne2(unsigned int __a, unsigned int __b) { } __DEVICE__ unsigned int __vsetne4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vset4.u32.u32.ne %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vset4.u32.u32.ne %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vcmpne4(unsigned int __a, unsigned int __b) { @@ -1345,94 +1399,112 @@ __DEVICE__ unsigned int __vmaxs2(unsigned int __a, unsigned int __b) { unsigned mask = __vcmpgts2(__a, __b); r = (__a & mask) | (__b & ~mask); } else { - asm("vmax2.s32.s32.s32 %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmax2.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); } return r; } __DEVICE__ unsigned int __vmaxs4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmax4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmax4.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vmaxu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmax2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmax2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vmaxu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmax4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmax4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vmins2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmin2.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmin2.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vmins4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmin4.s32.s32.s32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmin4.s32.s32.s32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vminu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmin2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmin2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vminu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vmin4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vmin4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsads2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 2.s32.s32.s32.add %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 2.s32.s32.s32.add %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsads4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 4.s32.s32.s32.add %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 4.s32.s32.s32.add %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsadu2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 2.u32.u32.u32.add %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 2.u32.u32.u32.add %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsadu4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vabs diff 4.u32.u32.u32.add %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vabs diff 4.u32.u32.u32.add %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsub2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub2.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub2.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vneg2(unsigned int __a) { return __vsub2(0, __a); } __DEVICE__ unsigned int __vsub4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub4.u32.u32.u32 %0,%1,%2,%3;" : "=r"(r) : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub4.u32.u32.u32 %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vneg4(unsigned int __a) { return __vsub4(0, __a); } __DEVICE__ unsigned int __vsubss2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub2.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub2.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vnegss2(unsigned int __a) { @@ -1440,9 +1512,9 @@ __DEVICE__ unsigned int __vnegss2(unsigned int __a) { } __DEVICE__ unsigned int __vsubss4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub4.s32.s32.s32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub4.s32.s32.s32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vnegss4(unsigned int __a) { @@ -1450,16 +1522,16 @@ __DEVICE__ unsigned int __vnegss4(unsigned int __a) { } __DEVICE__ unsigned int __vsubus2(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub2.u32.u32.u32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub2.u32.u32.u32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } __DEVICE__ unsigned int __vsubus4(unsigned int __a, unsigned int __b) { unsigned int r; - asm("vsub4.u32.u32.u32.sat %0,%1,%2,%3;" - : "=r"(r) - : "r"(__a), "r"(__b), "r"(0)); + __asm__("vsub4.u32.u32.u32.sat %0,%1,%2,%3;" + : "=r"(r) + : "r"(__a), "r"(__b), "r"(0)); return r; } #endif // CUDA_VERSION >= 9020 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
