Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

This patch updates helpers for ppc64el, s390x and purec_scalar. #591

Merged
merged 1 commit into from
Oct 14, 2024
Merged

This patch updates helpers for ppc64el, s390x and purec_scalar. #591

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 11 additions & 11 deletions src/arch/helperpower_128.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,16 +103,16 @@ typedef vquad vargquad;
#define vset__s64(...) ((v__i64) {__VA_ARGS__})
#define vset__u64(...) ((v__u64) {__VA_ARGS__})

#define vsetall__vi(v) vset__vi(v, v)
#define vsetall__vi2(v) vset__vi2(v, v, v, v)
#define vsetall__vi(v) vset__vi((int)v, (int)v)
#define vsetall__vi2(v) vset__vi2((int)v, (int)v, (int)v, (int)v)
#define vsetall__vm(v) vset__vm(v, v, v, v)
#define vsetall__vo(v) vset__vo(v, v, v, v)
#define vsetall__vf(v) vset__vf(v, v, v, v)
#define vsetall__vd(v) vset__vd(v, v)
#define vsetall__u8(v) vset__u8(v, v, v, v, v, v, v, v, v, v, v, v, v, v, v, v)
#define vsetall__u32(v) vset__u32(v, v, v, v)
#define vsetall__s64(v) vset__s64(v, v)
#define vsetall__u64(v) vset__u64(v, v)
#define vsetall__vf(v) vset__vf((float)v, (float)v, (float)v, (float)v)
#define vsetall__vd(v) vset__vd((double)v, (double)v)
#define vsetall__u8(v) vset__u8((uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v, (uint8_t)v)
#define vsetall__u32(v) vset__u32((uint32_t)v, (uint32_t)v, (uint32_t)v, (uint32_t)v)
#define vsetall__s64(v) vset__s64((int64_t)v, (int64_t)v)
#define vsetall__u64(v) vset__u64((uint64_t)v, (uint64_t)v)

#define vzero__vi() vsetall__vi(0)
#define vzero__vi2() vsetall__vi2(0)
Expand Down Expand Up @@ -351,7 +351,7 @@ static INLINE vmask vcastu_vm_vi(vint vi)

static INLINE vopmask vcast_vo_i(int i) {
i = i ? -1 : 0;
return (vopmask) { i, i, i, i };
return (vopmask) { (unsigned int)i, (unsigned int)i, (unsigned int)i, (unsigned int)i };
}

// signed int to single-precision
Expand Down Expand Up @@ -860,11 +860,11 @@ static INLINE vopmask vgt64_vo_vm_vm(vmask x, vmask y) {
#define vsrl64_vm_vm_i(x, c) ((vmask)vec_sr((__vector signed long long)x, (__vector unsigned long long)vsetall__vm(c)))

static INLINE vint vcast_vi_vm(vmask vm) {
return (vint) { vm[0], vm[2] };
return (vint) { (int)vm[0], (int)vm[2] };
}

static INLINE vmask vcast_vm_vi(vint vi) {
return (vmask) (__vector signed long long) { vi[0], vi[1] };
return (vmask) (__vector signed long long) { (signed long long)vi[0], (signed long long)vi[1] };
}

static INLINE vmask vreinterpret_vm_vi64(vint64 v) { return (vmask)v; }
Expand Down
2 changes: 1 addition & 1 deletion src/arch/helperpurec_scalar.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -426,7 +426,7 @@ static INLINE void vstream_v_p_vf(float *ptr, vfloat v) { *ptr = v; }
static vquad loadu_vq_p(void *p) {
vquad vq;
memcpy(8 + (char *)&vq, p, 8);
memcpy((char *)&vq, 8 + p, 8);
memcpy((char *)&vq, 8 + (char *)p, 8);
return vq;
}

Expand Down
32 changes: 16 additions & 16 deletions src/arch/helpers390x_128.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,7 +78,7 @@ typedef vquad vargquad;

static INLINE int vavailability_i(int n) {
if (n == 1 || n == 2) {
return vec_max((vdouble) {n, n}, (vdouble) {n, n})[0] != 0;
return vec_max((vdouble) {(double)n, (double)n}, (vdouble) {(double)n, (double)n})[0] != 0;
}
return 0;
}
Expand Down Expand Up @@ -127,23 +127,23 @@ static INLINE vfloat vgather_vf_p_vi2(const float *p, vint2 vi2) {
return ((vfloat) { p[vi2[0]], p[vi2[1]], p[vi2[2]], p[vi2[3]] });
}

static INLINE vopmask vcast_vo_i(int i) { return (vopmask) { i ? (long long)-1 : 0, i ? (long long)-1 : 0 }; }
static INLINE vopmask vcast_vo_i(int i) { return (vopmask) { i ? (unsigned long long)-1 : 0, i ? (unsigned long long)-1 : 0 }; }
static INLINE vint vcast_vi_i(int i) { return (vint) { i, i }; }
static INLINE vint2 vcast_vi2_i(int i) { return (vint2) { i, i, i, i }; }
static INLINE vfloat vcast_vf_f(float f) { return (vfloat) { f, f, f, f }; }
static INLINE vdouble vcast_vd_d(double d) { return (vdouble) { d, d }; }

static INLINE vdouble vcast_vd_vi(vint vi) { return (vdouble) { vi[0], vi[1] }; }
static INLINE vfloat vcast_vf_vi2(vint2 vi) { return (vfloat) { vi[0], vi[1], vi[2], vi[3] }; }
static INLINE vdouble vcast_vd_vi(vint vi) { return (vdouble) { (double)vi[0], (double)vi[1] }; }
static INLINE vfloat vcast_vf_vi2(vint2 vi) { return (vfloat) { (float)vi[0], (float)vi[1], (float)vi[2], (float)vi[3] }; }
static INLINE vdouble vtruncate_vd_vd(vdouble vd) { return __builtin_s390_vfidb(vd, 4, 5); }
static INLINE vdouble vrint_vd_vd(vdouble vd) { return __builtin_s390_vfidb(vd, 4, 4); }

static INLINE vint vrint_vi_vd(vdouble vd) {
vd = vrint_vd_vd(vd);
return (vint) { vd[0], vd[1] };
return (vint) { (int)vd[0], (int)vd[1] };
}
static INLINE vint vtruncate_vi_vd(vdouble vd) { return (vint) { vd[0], vd[1] }; }
static INLINE vint2 vtruncate_vi2_vf(vfloat vf) { return (vint) { vf[0], vf[1], vf[2], vf[3] }; }
static INLINE vint vtruncate_vi_vd(vdouble vd) { return (vint) { (int)vd[0], (int)vd[1] }; }
static INLINE vint2 vtruncate_vi2_vf(vfloat vf) { return (vint) { (int)vf[0], (int)vf[1], (int)vf[2], (int)vf[3] }; }

static INLINE vmask vreinterpret_vm_vd(vdouble vd) { return (vmask)vd; }
static INLINE vdouble vreinterpret_vd_vm(vmask vm) { return (vdouble)vm; }
Expand Down Expand Up @@ -202,7 +202,7 @@ static INLINE vmask vcast_vm_i64(int64_t i) { return (vmask)(vint64){ i, i }; }
static INLINE vmask vcast_vm_u64(uint64_t i) { return (vmask)(vuint64){ i, i }; }

static INLINE vmask vcastu_vm_vi(vint vi) { return (vmask)(vint2){ vi[0], 0, vi[1], 0 }; }
static INLINE vint vcastu_vi_vm(vmask vi2) { return (vint){ vi2[0] >> 32, vi2[1] >> 32 }; }
static INLINE vint vcastu_vi_vm(vmask vi2) { return (vint){ (int)(vi2[0] >> 32), (int)(vi2[1] >> 32) }; }

static INLINE vint vreinterpretFirstHalf_vi_vi2(vint2 vi2) { return (vint){ vi2[0], vi2[1] }; }
static INLINE vint2 vreinterpretFirstHalf_vi2_vi(vint vi) { return (vint2){ vi[0], vi[1], 0, 0 }; }
Expand Down Expand Up @@ -309,8 +309,8 @@ static INLINE vint vxor_vi_vi_vi(vint x, vint y) { return x ^ y; }
static INLINE vint vand_vi_vo_vi(vopmask x, vint y) { return vreinterpretFirstHalf_vi_vi2((vint2)x) & y; }
static INLINE vint vandnot_vi_vo_vi(vopmask x, vint y) { return vec_andc(y, vreinterpretFirstHalf_vi_vi2((vint2)x)); }

static INLINE vint vsll_vi_vi_i(vint x, int c) { return (vint)(((__vector unsigned int)x) << (__vector unsigned int){c, c, c, c}); }
static INLINE vint vsrl_vi_vi_i(vint x, int c) { return (vint)(((__vector unsigned int)x) >> (__vector unsigned int){c, c, c, c}); }
static INLINE vint vsll_vi_vi_i(vint x, int c) { return (vint)(((__vector unsigned int)x) << (__vector unsigned int){(unsigned int)c, (unsigned int)c, (unsigned int)c, (unsigned int)c}); }
static INLINE vint vsrl_vi_vi_i(vint x, int c) { return (vint)(((__vector unsigned int)x) >> (__vector unsigned int){(unsigned int)c, (unsigned int)c, (unsigned int)c, (unsigned int)c}); }
static INLINE vint vsra_vi_vi_i(vint x, int c) { return x >> (__vector int){c, c, c, c}; }

static INLINE vint veq_vi_vi_vi(vint x, vint y) { return vec_cmpeq(x, y); }
Expand Down Expand Up @@ -364,8 +364,8 @@ static INLINE vint2 vxor_vi2_vi2_vi2(vint2 x, vint2 y) { return x ^ y; }
static INLINE vint2 vand_vi2_vo_vi2(vopmask x, vint2 y) { return (vint2)x & y; }
static INLINE vint2 vandnot_vi2_vo_vi2(vopmask x, vint2 y) { return y & ~(vint2)x; }

static INLINE vint2 vsll_vi2_vi2_i(vint2 x, int c) { return (vint2)(((__vector unsigned int)x) << (__vector unsigned int){c, c, c, c}); }
static INLINE vint2 vsrl_vi2_vi2_i(vint2 x, int c) { return (vint2)(((__vector unsigned int)x) >> (__vector unsigned int){c, c, c, c}); }
static INLINE vint2 vsll_vi2_vi2_i(vint2 x, int c) { return (vint2)(((__vector unsigned int)x) << (__vector unsigned int){(unsigned int)c, (unsigned int)c, (unsigned int)c, (unsigned int)c}); }
static INLINE vint2 vsrl_vi2_vi2_i(vint2 x, int c) { return (vint2)(((__vector unsigned int)x) >> (__vector unsigned int){(unsigned int)c, (unsigned int)c, (unsigned int)c, (unsigned int)c}); }
static INLINE vint2 vsra_vi2_vi2_i(vint2 x, int c) { return x >> (__vector int){c, c, c, c}; }

static INLINE vopmask veq_vo_vi2_vi2(vint2 x, vint2 y) { return (vopmask)vec_cmpeq(x, y); }
Expand Down Expand Up @@ -405,7 +405,7 @@ static INLINE vopmask visnan_vo_vf (vfloat d) { return vneq_vo_vf_vf(d, d); }

static INLINE vint2 vrint_vi2_vf(vfloat vf) {
vf = vrint_vf_vf(vf);
return (vint) { vf[0], vf[1], vf[2], vf[3] };
return (vint) { (int)vf[0], (int)vf[1], (int)vf[2], (int)vf[3] };
}

//
Expand Down Expand Up @@ -445,11 +445,11 @@ static INLINE vopmask vgt64_vo_vm_vm(vmask x, vmask y) {
return (vopmask)vec_cmpgt((__vector signed long long)x, (__vector signed long long)y);
}

#define vsll64_vm_vm_i(x, c) ((vmask)((__vector unsigned long long)x << (__vector unsigned long long) { c, c }))
#define vsrl64_vm_vm_i(x, c) ((vmask)((__vector unsigned long long)x >> (__vector unsigned long long) { c, c }))
#define vsll64_vm_vm_i(x, c) ((vmask)((__vector unsigned long long)x << (__vector unsigned long long) { (unsigned long long)c, (unsigned long long)c }))
#define vsrl64_vm_vm_i(x, c) ((vmask)((__vector unsigned long long)x >> (__vector unsigned long long) { (unsigned long long)c, (unsigned long long)c }))

static INLINE vint vcast_vi_vm(vmask vm) {
return (vint) { vm[0], vm[1] };
return (vint) { (int)vm[0], (int)vm[1] };
}

static INLINE vmask vcast_vm_vi(vint vi) {
Expand Down
Loading