mewin/glm
1
0
Fork 0
Code Releases Activity

Merge pull request #992 from amaury-ml/compfix

Browse Source 
[Aarch64] fix for glm::length #992
This commit is contained in:
Christophe 2020-02-07 14:57:13 +01:00 committed by GitHub
commit 82d500376e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
glm/detail/func_geometric_simd.inl
View File

@ -105,7 +105,7 @@ namespace detail
{ {
GLM_FUNC_QUALIFIER static float call(vec<4, float, Q> const& v) GLM_FUNC_QUALIFIER static float call(vec<4, float, Q> const& v)
{ {
return compute_dot<vec<4, float, Q>, float, true>::call(v, v); return sqrt(compute_dot<vec<4, float, Q>, float, true>::call(v, v));
} }
}; };
@ -126,9 +126,7 @@ namespace detail
{ {
#if GLM_ARCH & GLM_ARCH_ARMV8_BIT #if GLM_ARCH & GLM_ARCH_ARMV8_BIT
float32x4_t v = vmulq_f32(x.data, y.data); float32x4_t v = vmulq_f32(x.data, y.data);
v = vpaddq_f32(v, v); return vaddvq_f32(v);
v = vpaddq_f32(v, v);
return vgetq_lane_f32(v, 0);
#else // Armv7a with Neon #else // Armv7a with Neon
float32x4_t p = vmulq_f32(x.data, y.data); float32x4_t p = vmulq_f32(x.data, y.data);
float32x2_t v = vpadd_f32(vget_low_f32(p), vget_high_f32(p)); float32x2_t v = vpadd_f32(vget_low_f32(p), vget_high_f32(p));