[Performance][Hardware][RISC-V] Reduce LMUL pressure in INT4 LUT dequant (#47538)

Signed-off-by: liutong <liutong@iscas.ac.cn>
Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
LiuTong
2026-07-06 05:58:56 +00:00
committed by GitHub
co-authored by Claude
parent 16f8110935
commit e433634c78
+23 -10
View File
@@ -623,17 +623,30 @@ struct FP32Vec16 : public Vec<FP32Vec16> {
data.reg, data.reg)) {};
explicit FP32Vec16(const FP32Vec16& data) : reg(data.reg) {};
explicit FP32Vec16(int64_t value, const FP32Vec16& lut) {
const uint64_t q_values = static_cast<uint64_t>(value);
auto packed = RVVI(__riscv_vmv_v_x_u64, LMUL_1024)(q_values, VEC_ELEM_NUM);
auto lane_ids = RVVI(__riscv_vid_v_u64, LMUL_1024)(VEC_ELEM_NUM);
// Split into two 32-bit halves to avoid u64 @ LMUL_1024 (m8 on
// VLEN=128 / m4 on VLEN=256), which causes heavy register spilling.
constexpr int HALF = VEC_ELEM_NUM / 2;
const auto q = static_cast<uint64_t>(value);
const uint32_t lo = static_cast<uint32_t>(q);
const uint32_t hi = static_cast<uint32_t>(q >> 32);
auto lane_ids = RVVI(__riscv_vid_v_u32, LMUL_256)(HALF);
auto shifts =
RVVI(__riscv_vsll_vx_u64, LMUL_1024)(lane_ids, 2, VEC_ELEM_NUM);
auto shifted =
RVVI(__riscv_vsrl_vv_u64, LMUL_1024)(packed, shifts, VEC_ELEM_NUM);
auto idx64 =
RVVI(__riscv_vand_vx_u64, LMUL_1024)(shifted, 0xF, VEC_ELEM_NUM);
auto idx32 = RVVI(__riscv_vnsrl_wx_u32, LMUL_512)(idx64, 0, VEC_ELEM_NUM);
reg = RVVI(__riscv_vrgather_vv_f32, LMUL_512)(lut.reg, idx32, VEC_ELEM_NUM);
RVVI(__riscv_vsll_vx_u32, LMUL_256)(lane_ids, 2, HALF);
auto packed_lo = RVVI(__riscv_vmv_v_x_u32, LMUL_256)(lo, HALF);
auto idx_lo = RVVI(__riscv_vand_vx_u32, LMUL_256)(
RVVI(__riscv_vsrl_vv_u32, LMUL_256)(packed_lo, shifts, HALF),
0xF, HALF);
auto packed_hi = RVVI(__riscv_vmv_v_x_u32, LMUL_256)(hi, HALF);
auto idx_hi = RVVI(__riscv_vand_vx_u32, LMUL_256)(
RVVI(__riscv_vsrl_vv_u32, LMUL_256)(packed_hi, shifts, HALF),
0xF, HALF);
auto idx = RVVI4(__riscv_vcreate_v_u32, LMUL_256, _u32,
LMUL_512)(idx_lo, idx_hi);
reg = RVVI(__riscv_vrgather_vv_f32, LMUL_512)(lut.reg, idx, VEC_ELEM_NUM);
}
explicit FP32Vec16(const FP16Vec16& v);