1 /* 2 * vxeh2_vcvt: vector-enhancements facility 2 vector convert * 3 */ 4 #include <stdint.h> 5 #include "vx.h" 6 7 #define M_S 8 8 #define M4_XxC 4 9 #define M4_def M4_XxC 10 vcfps(S390Vector * v1,S390Vector * v2,const uint8_t m3,const uint8_t m4,const uint8_t m5)11 static inline void vcfps(S390Vector *v1, S390Vector *v2, 12 const uint8_t m3, const uint8_t m4, const uint8_t m5) 13 { 14 asm volatile("vcfps %[v1], %[v2], %[m3], %[m4], %[m5]\n" 15 : [v1] "=v" (v1->v) 16 : [v2] "v" (v2->v) 17 , [m3] "i" (m3) 18 , [m4] "i" (m4) 19 , [m5] "i" (m5)); 20 } 21 vcfpl(S390Vector * v1,S390Vector * v2,const uint8_t m3,const uint8_t m4,const uint8_t m5)22 static inline void vcfpl(S390Vector *v1, S390Vector *v2, 23 const uint8_t m3, const uint8_t m4, const uint8_t m5) 24 { 25 asm volatile("vcfpl %[v1], %[v2], %[m3], %[m4], %[m5]\n" 26 : [v1] "=v" (v1->v) 27 : [v2] "v" (v2->v) 28 , [m3] "i" (m3) 29 , [m4] "i" (m4) 30 , [m5] "i" (m5)); 31 } 32 vcsfp(S390Vector * v1,S390Vector * v2,const uint8_t m3,const uint8_t m4,const uint8_t m5)33 static inline void vcsfp(S390Vector *v1, S390Vector *v2, 34 const uint8_t m3, const uint8_t m4, const uint8_t m5) 35 { 36 asm volatile("vcsfp %[v1], %[v2], %[m3], %[m4], %[m5]\n" 37 : [v1] "=v" (v1->v) 38 : [v2] "v" (v2->v) 39 , [m3] "i" (m3) 40 , [m4] "i" (m4) 41 , [m5] "i" (m5)); 42 } 43 vclfp(S390Vector * v1,S390Vector * v2,const uint8_t m3,const uint8_t m4,const uint8_t m5)44 static inline void vclfp(S390Vector *v1, S390Vector *v2, 45 const uint8_t m3, const uint8_t m4, const uint8_t m5) 46 { 47 asm volatile("vclfp %[v1], %[v2], %[m3], %[m4], %[m5]\n" 48 : [v1] "=v" (v1->v) 49 : [v2] "v" (v2->v) 50 , [m3] "i" (m3) 51 , [m4] "i" (m4) 52 , [m5] "i" (m5)); 53 } 54 main(int argc,char * argv[])55 int main(int argc, char *argv[]) 56 { 57 S390Vector vd; 58 S390Vector vs_i32 = { .w[0] = 1, .w[1] = 64, .w[2] = 1024, .w[3] = -10 }; 59 S390Vector vs_u32 = { .w[0] = 2, .w[1] = 32, .w[2] = 4096, .w[3] = 8888 }; 60 S390Vector vs_f32 = { .f[0] = 3.987, .f[1] = 5.123, 61 .f[2] = 4.499, .f[3] = 0.512 }; 62 63 vd.d[0] = vd.d[1] = 0; 64 vcfps(&vd, &vs_i32, 2, M4_def, 0); 65 if (1 != vd.f[0] || 1024 != vd.f[2] || 64 != vd.f[1] || -10 != vd.f[3]) { 66 return 1; 67 } 68 69 vd.d[0] = vd.d[1] = 0; 70 vcfpl(&vd, &vs_u32, 2, M4_def, 0); 71 if (2 != vd.f[0] || 4096 != vd.f[2] || 32 != vd.f[1] || 8888 != vd.f[3]) { 72 return 1; 73 } 74 75 vd.d[0] = vd.d[1] = 0; 76 vcsfp(&vd, &vs_f32, 2, M4_def, 0); 77 if (4 != vd.w[0] || 4 != vd.w[2] || 5 != vd.w[1] || 1 != vd.w[3]) { 78 return 1; 79 } 80 81 vd.d[0] = vd.d[1] = 0; 82 vclfp(&vd, &vs_f32, 2, M4_def, 0); 83 if (4 != vd.w[0] || 4 != vd.w[2] || 5 != vd.w[1] || 1 != vd.w[3]) { 84 return 1; 85 } 86 87 return 0; 88 } 89