fix(dnn/arm): use vcvtq_f32_s32 for all arm code

GitOrigin-RevId: 27effe7d24

dnn/src/arm_common/elemwise_helper/kimpl/typecvt.h

@@ -19,7 +19,6 @@ namespace arm_common {
 template <typename src_ctype, typename dst_ctype = src_ctype>
 struct TypeCvtOp;
 
-#if __ARM_ARCH >= 8
 template <>
 struct TypeCvtOp<dt_qint32, dt_qint8> : UnaryOpBase<dt_qint32, dt_qint8> {
     using UnaryOpBase::UnaryOpBase;
@@ -55,59 +54,6 @@ struct TypeCvtOp<dt_qint32, dt_qint8> : UnaryOpBase<dt_qint32, dt_qint8> {
         return QConverter::convert<int8x8_t, float32x4_t>(vitem0);
     }
 };
-#else
-template <>
-struct TypeCvtOp<dt_qint32, dt_qint8> : UnaryOpBase<dt_qint32, dt_qint8>,
-                                        FixupBase {
-    constexpr static size_t SIMD_WIDTH = 4;
-
-    TypeCvtOp(DType src_dtype, DType dst_dtype)
-            : UnaryOpBase(src_dtype, dst_dtype), FixupBase(scale) {}
-
-    TypeCvtOp(float src_scale, float dst_scale)
-            : UnaryOpBase(src_scale, dst_scale), FixupBase(scale) {}
-
-    void operator()(const int32x4x2_t& vsrc, dt_qint8* dst) const {
-        vst1_s8(reinterpret_cast<int8_t*>(dst), operator()(vsrc));
-    }
-    void operator()(const int32x4_t& vsrc, dt_qint8* dst) const {
-        vst1_lane_s32(reinterpret_cast<int32_t*>(dst),
-                      (int32x2_t)(operator()(vsrc)), 0);
-    }
-    dt_qint8 operator()(const dt_qint32& src) const {
-        float fsrc = src.as_int32() * this->scale;
-        return QConverter::convert<dt_qint8, float>(fsrc);
-    }
-    void operator()(const src_ctype& src, dst_ctype* dst) const {
-        *dst = operator()(src);
-    }
-    int8x8_t operator()(const int32x4x2_t& vsrc) const {
-        int32x4_t vitem0 = vqrdmulhq_s32(vsrc.val[0], vmultiplier);
-        int32x4_t vitem1 = vqrdmulhq_s32(vsrc.val[1], vmultiplier);
-        auto fixup0 = vshrq_n_s32(vitem0, 31);
-        auto fixup1 = vshrq_n_s32(vitem1, 31);
-        // FIXME Theoretically, we should check shift != 0 here.
-        vitem0 = vqaddq_s32(vitem0, fixup0);
-        vitem1 = vqaddq_s32(vitem1, fixup1);
-        return vqmovn_s16(vcombine_s16(vqmovn_s32(vrshlq_s32(vitem0, vshift)),
-                                       vqmovn_s32(vrshlq_s32(vitem1, vshift))));
-    }
-    int8x8_t operator()(const int32x4_t& src) const {
-        int32x4_t vitem0 = vqrdmulhq_s32(src, vmultiplier);
-        auto fixup0 = vshrq_n_s32(vitem0, 31);
-        vitem0 = vqaddq_s32(vitem0, fixup0);
-        int16x4_t vres0_int16 = vqmovn_s32(vrshlq_s32(vitem0, vshift));
-        return vqmovn_s16(vcombine_s16(vres0_int16, vres0_int16));
-    }
-    int8x8_t operator()(const float32x4_t& src) const {
-        int32x4_t vitem0 = vqrdmulhq_s32(vcvtq_s32_f32(src), vmultiplier);
-        auto fixup0 = vshrq_n_s32(vitem0, 31);
-        vitem0 = vqaddq_s32(vitem0, fixup0);
-        int16x4_t vres0_int16 = vqmovn_s32(vrshlq_s32(vitem0, vshift));
-        return vqmovn_s16(vcombine_s16(vres0_int16, vres0_int16));
-    }
-};
-#endif
 
 template <>
 struct TypeCvtOp<dt_qint32, dt_quint8> : UnaryOpBase<dt_qint32, dt_quint8> {

dnn/src/common/nchw_nchwxx_valid.h

@@ -144,7 +144,7 @@ inline bool nchw_nchwxx_valid<NCHW88>(
         const DTypeEnum dst_dtype,
         const ConvolutionBase<param::Convolution>::CanonizedFilterMeta& fm,
         const BiasMode bias_mode,
-        const param::ConvBias::NonlineMode nonline_mode) {
+        const param::ConvBias::NonlineMode ) {
     bool ok_type = ((src_dtype == DTypeEnum::Float32 &&
                      filter_dtype == DTypeEnum::Float32 &&
                      (dst_dtype == DTypeEnum::Float32))) &&

src/gopt/test/inference.cpp

@@ -32,7 +32,6 @@
 #include "./helper.h"
 #include "megbrain/comp_node_env.h"
 
-#include "cpuinfo.h"
 #include "megdnn/tensor_format.h"
 
 #include <random>