mirror of
https://github.com/hrydgard/ppsspp.git
synced 2026-10-17 16:30:30 +02:00
Henrik RydgårdandClaude Opus 5.5
cd1f3b0929
For where there's no vertex JIT. It skins the same way as the steps: the skin matrix is 0 + bone * weight, kept as the columns that Vec3ByMatrix43() and Norm3ByMatrix43() multiply and add unfused in the same order (by lane on ARM64, like those), and only rebuilt when the weight changes. 2.2 ns a vertex in God of War against the JIT's 1.95 and the steps' 12-18. Without the vertex JIT, God of War's instructions drop by 6.8% per 10 emulated seconds. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
23 lines
1.6 KiB
C
23 lines
1.6 KiB
C
#pragma once
|
|
|
|
// These are useful on JIT-less platforms - they don't beat the jitted vertex decoders by much, but they
|
|
// will beat the function-call-stitched ones by a lot.
|
|
|
|
void VtxDec_Tu16_C8888_Pfloat(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tu8_C5551_Ps16(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
|
|
// For the UV generation modes that don't prescale.
|
|
void VtxDec_Tu16_C8888_Pfloat_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tu8_C5551_Ps16_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
|
|
// Each with a version for the UV generation modes that don't prescale.
|
|
void VtxDec_Tu8_C565_Ns16_Pfloat(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tu8_C565_Ns16_Pfloat_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tfloat_C8888_Ns8_Ps16(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tfloat_C8888_Ns8_Ps16_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tu16_C4444_Ps16(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_Tu16_C4444_Ps16_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
// Skins in the decoder, so not for callerSkins.
|
|
void VtxDec_W1float_Tu8_C4444_Ns8_Pfloat(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|
|
void VtxDec_W1float_Tu8_C4444_Ns8_Pfloat_NoPrescale(const u8 *srcp, u8 *dstp, int count, const UVScale *uvScaleOffset);
|