WAsm SIMD versions of Vector Elementwise Binary microkernels

PiperOrigin-RevId: 318522846
diff --git a/test/f32-vsqrdiff.cc b/test/f32-vsqrdiff.cc
index 767ee9d..ddb3c34 100644
--- a/test/f32-vsqrdiff.cc
+++ b/test/f32-vsqrdiff.cc
@@ -697,6 +697,128 @@
 #endif  // !XNN_ARCH_ASMJS && !XNN_ARCH_WASM && !XNN_COMPILER_MSVC && !XNN_COMPILER_ICC
 
 
+#if XNN_ARCH_WASMSIMD
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, batch_eq_4) {
+    VBinOpMicrokernelTester()
+      .batch_size(4)
+      .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, batch_div_4) {
+    for (size_t batch_size = 8; batch_size < 40; batch_size += 4) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, batch_lt_4) {
+    for (size_t batch_size = 1; batch_size < 4; batch_size++) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, batch_gt_4) {
+    for (size_t batch_size = 5; batch_size < 8; batch_size++) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, inplace_a) {
+    for (size_t batch_size = 1; batch_size <= 20; batch_size += 3) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_a(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, inplace_b) {
+    for (size_t batch_size = 1; batch_size <= 20; batch_size += 3) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_b(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X4, inplace_a_and_b) {
+    for (size_t batch_size = 1; batch_size <= 20; batch_size += 3) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_a(true)
+        .inplace_b(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x4, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+#endif  // XNN_ARCH_WASMSIMD
+
+
+#if XNN_ARCH_WASMSIMD
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, batch_eq_8) {
+    VBinOpMicrokernelTester()
+      .batch_size(8)
+      .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, batch_div_8) {
+    for (size_t batch_size = 16; batch_size < 80; batch_size += 8) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, batch_lt_8) {
+    for (size_t batch_size = 1; batch_size < 8; batch_size++) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, batch_gt_8) {
+    for (size_t batch_size = 9; batch_size < 16; batch_size++) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, inplace_a) {
+    for (size_t batch_size = 1; batch_size <= 40; batch_size += 7) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_a(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, inplace_b) {
+    for (size_t batch_size = 1; batch_size <= 40; batch_size += 7) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_b(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+
+  TEST(F32_VSQRDIFF__WASMSIMD_X8, inplace_a_and_b) {
+    for (size_t batch_size = 1; batch_size <= 40; batch_size += 7) {
+      VBinOpMicrokernelTester()
+        .batch_size(batch_size)
+        .inplace_a(true)
+        .inplace_b(true)
+        .Test(xnn_f32_vsqrdiff_ukernel__wasmsimd_x8, VBinOpMicrokernelTester::OpType::SqrDiff, VBinOpMicrokernelTester::Variant::Scalar);
+    }
+  }
+#endif  // XNN_ARCH_WASMSIMD
+
+
 TEST(F32_VSQRDIFF__SCALAR_X1, batch_eq_1) {
   VBinOpMicrokernelTester()
     .batch_size(1)