Modified some review comments

Author: sunjiweiswift

src/ATen/native/xpu/LinearInt4.cpp

@@ -8,7 +8,7 @@
 #include <comm/xpu_aten.h>
 
 namespace at::native {
-Tensor& linear_int4_xpu(
+Tensor& _weight_int4pack_mm_with_scales_and_zeros_xpu(
     const Tensor& input,
     const Tensor& weight,
     int qGroupSize,

src/ATen/native/xpu/sycl/LinearInt4.cpp

@@ -47,6 +47,11 @@ struct LinearInt4KernelFunctor : public __SYCL_KER_CONFIG_CONVENTION__ {
     auto bptr = B + g_n * k / 2;
     auto aptr = A;
     auto cptr = C + g_n;
+
+    // AT_DISPATCH_ALL_TYPES_AND_COMPLEX_AND2
+    // (kBFloat16, kBFloat16, scalar_t, "LinearInt4KernelFunctor", [&] {
+ 
+          
     if constexpr (std::is_same_v<scalar_t, sycl::half>) {
       sycl::half2 tmpAcc = {0.f, 0.f};
       uint8_t tmps8[TileK / 2];

yaml/native/native_functions.yaml

@@ -8549,7 +8549,7 @@
 
 - func: _weight_int4pack_mm_with_scales_and_zeros(Tensor self, Tensor mat2, int qGroupSize, Tensor qScaleAndZeros) -> Tensor
   dispatch:
-    XPU: linear_int4_xpu
+    XPU: _weight_int4pack_mm_with_scales_and_zeros_xpu
   autogen: _weight_int4pack_mm_with_scales_and_zeros.out
   tags: core