Skip to content

Use SVE widening loads for vint4(uint8_t*) #514

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

Merged
merged 1 commit into from
Nov 5, 2024
Merged

Use SVE widening loads for vint4(uint8_t*) #514

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
2 changes: 2 additions & 0 deletions Docs/ChangeLog-5x.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,8 @@ The 5.1.0 release is a maintenance release.
byte-to-int index conversion.
* **Optimization:** Added improved intrinsics sequence for SSE and AVX2
`hmin()` and `hmax()`.
* **Optimization:** Added improved intrinsics sequence for `vint4(uint8_t*)`
on systems implementing Arm SVE.

<!-- ---------------------------------------------------------------------- -->
## 5.0.0
Expand Down
13 changes: 9 additions & 4 deletions Source/astcenc_vecmathlib_neon_4.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -194,10 +194,15 @@ struct vint4
*/
ASTCENC_SIMD_INLINE explicit vint4(const uint8_t *p)
{
// Cast is safe - NEON loads are allowed to be unaligned
uint32x2_t t8 = vld1_dup_u32(reinterpret_cast<const uint32_t*>(p));
uint16x4_t t16 = vget_low_u16(vmovl_u8(vreinterpret_u8_u32(t8)));
m = vreinterpretq_s32_u32(vmovl_u16(t16));
#if ASTCENC_SVE == 0
// Cast is safe - NEON loads are allowed to be unaligned
uint32x2_t t8 = vld1_dup_u32(reinterpret_cast<const uint32_t*>(p));
uint16x4_t t16 = vget_low_u16(vmovl_u8(vreinterpret_u8_u32(t8)));
m = vreinterpretq_s32_u32(vmovl_u16(t16));
#else
svint32_t data = svld1ub_s32(svptrue_pat_b32(SV_VL4), p);
m = svget_neonq(data);
#endif
}

/**
Expand Down