github_mirrors/vectorscan
3
0
Fork 1
mirror of https://github.com/VectorCamp/vectorscan.git synced 2025-06-28 16:41:01 +03:00
Code Issues Packages Projects Releases Wiki Activity

[VSX] movemask needs to be explicitly aligned on clang for vec_ste

Browse Source
This commit is contained in:
Konstantinos Margaritis 2022-09-16 12:50:33 +03:00
parent ef66877e9e
commit 3fc6c8a532
2 changed files with 19 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

34
src/util/arch/ppc64el/simd_utils.h
View File

@ -152,7 +152,7 @@ static really_inline u32 movemask128(m128 a) {
static uint8x16_t perm = { 16, 24, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }; static uint8x16_t perm = { 16, 24, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
uint8x16_t bitmask = vec_gb((uint8x16_t) a); uint8x16_t bitmask = vec_gb((uint8x16_t) a);
bitmask = (uint8x16_t) vec_perm(vec_splat_u8(0), bitmask, perm); bitmask = (uint8x16_t) vec_perm(vec_splat_u8(0), bitmask, perm);
u32 movemask; u32 ALIGN_ATTR(16) movemask;
vec_ste((uint32x4_t) bitmask, 0, &movemask); vec_ste((uint32x4_t) bitmask, 0, &movemask);
return movemask; return movemask;
} }
@ -285,27 +285,27 @@ m128 loadbytes128(const void *ptr, unsigned int n) {
return a; return a;
} }
#define CASE_ALIGN_VECTORS(a, b, offset) case offset: return (m128)vec_sld((int8x16_t)(b), (int8x16_t)(a), (16 - offset)); break; #define CASE_ALIGN_VECTORS(a, b, offset) case offset: return (m128)vec_sld((int8x16_t)(a), (int8x16_t)(b), (16 - offset)); break;
static really_really_inline static really_really_inline
m128 palignr_imm(m128 r, m128 l, int offset) { m128 palignr_imm(m128 r, m128 l, int offset) {
switch (offset) { switch (offset) {
case 0: return l; break; case 0: return l; break;
CASE_ALIGN_VECTORS(l, r, 1); CASE_ALIGN_VECTORS(r, l, 1);
CASE_ALIGN_VECTORS(l, r, 2); CASE_ALIGN_VECTORS(r, l, 2);
CASE_ALIGN_VECTORS(l, r, 3); CASE_ALIGN_VECTORS(r, l, 3);
CASE_ALIGN_VECTORS(l, r, 4); CASE_ALIGN_VECTORS(r, l, 4);
CASE_ALIGN_VECTORS(l, r, 5); CASE_ALIGN_VECTORS(r, l, 5);
CASE_ALIGN_VECTORS(l, r, 6); CASE_ALIGN_VECTORS(r, l, 6);
CASE_ALIGN_VECTORS(l, r, 7); CASE_ALIGN_VECTORS(r, l, 7);
CASE_ALIGN_VECTORS(l, r, 8); CASE_ALIGN_VECTORS(r, l, 8);
CASE_ALIGN_VECTORS(l, r, 9); CASE_ALIGN_VECTORS(r, l, 9);
CASE_ALIGN_VECTORS(l, r, 10); CASE_ALIGN_VECTORS(r, l, 10);
CASE_ALIGN_VECTORS(l, r, 11); CASE_ALIGN_VECTORS(r, l, 11);
CASE_ALIGN_VECTORS(l, r, 12); CASE_ALIGN_VECTORS(r, l, 12);
CASE_ALIGN_VECTORS(l, r, 13); CASE_ALIGN_VECTORS(r, l, 13);
CASE_ALIGN_VECTORS(l, r, 14); CASE_ALIGN_VECTORS(r, l, 14);
CASE_ALIGN_VECTORS(l, r, 15); CASE_ALIGN_VECTORS(r, l, 15);
case 16: return r; break; case 16: return r; break;
default: return zeroes128(); break; default: return zeroes128(); break;
} }

4
src/util/supervector/arch/ppc64el/impl.cpp
View File

@ -269,10 +269,10 @@ really_inline SuperVector<16> SuperVector<16>::eq(SuperVector<16> const &b) cons
template <> template <>
really_inline typename SuperVector<16>::comparemask_type really_inline typename SuperVector<16>::comparemask_type
SuperVector<16>::comparemask(void) const { SuperVector<16>::comparemask(void) const {
uint8x16_t bitmask = vec_gb( u.u8x16[0]);
static uint8x16_t perm = { 16, 24, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }; static uint8x16_t perm = { 16, 24, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
uint8x16_t bitmask = vec_gb(u.u8x16[0]);
bitmask = (uint8x16_t) vec_perm(vec_splat_u8(0), bitmask, perm); bitmask = (uint8x16_t) vec_perm(vec_splat_u8(0), bitmask, perm);
u32 movemask; u32 ALIGN_ATTR(16) movemask;
vec_ste((uint32x4_t) bitmask, 0, &movemask); vec_ste((uint32x4_t) bitmask, 0, &movemask);
return movemask; return movemask;
} }