Merge pull request #30 from xinhuang/sqr

implementation of v?Sqr

include/openvml.h

@@ -46,6 +46,9 @@ OPENVML_EXPORT void OpenVML_FUNCNAME(vdSub)(const VML_INT n, const double * a, c
 OPENVML_EXPORT void OpenVML_FUNCNAME(vcSub)(const VML_INT n, const float * a, const float * b, float * y);
 OPENVML_EXPORT void OpenVML_FUNCNAME(vzSub)(const VML_INT n, const double * a, const double * b, double * y);
 
+OPENVML_EXPORT void OpenVML_FUNCNAME(vsSqr)(const VML_INT n, const float * a, float * y);
+OPENVML_EXPORT void OpenVML_FUNCNAME(vdSqr)(const VML_INT n, const double * a, double * y);
+
 OPENVML_EXPORT void OpenVML_FUNCNAME(vsPow)(const VML_INT n, const float * a, const float * b, float * y);
 OPENVML_EXPORT void OpenVML_FUNCNAME(vdPow)(const VML_INT n, const double * a, const double * b, double * y);
 

include/openvml_kernel.h

@@ -41,6 +41,8 @@ void OpenVML_FUNCNAME(dsub_k)(VMLLONG n, double * a, double * b, double * y, dou
 void OpenVML_FUNCNAME(csub_k)(VMLLONG n, float * a, float * b, float * y, float * z, float * other_params);
 void OpenVML_FUNCNAME(zsub_k)(VMLLONG n, double * a, double * b, double * y, double * z, double * other_params);
 
+void OpenVML_FUNCNAME(ssqr_k)(VMLLONG n, float * a, float * b, float * y, float * z, float * other_params);
+void OpenVML_FUNCNAME(sdqr_k)(VMLLONG n, double * a, double * b, double * y, double * z, double * other_params);
 
 void OpenVML_FUNCNAME(spow_k)(VMLLONG n, float * a, float * b, float * y, float * z, float * other_params);
 void OpenVML_FUNCNAME(dpow_k)(VMLLONG n, double * a, double * b, double * y, double * z, double * other_params);

include/openvml_macros.h

@@ -43,6 +43,9 @@
 #define CSUB_K OpenVML_FUNCNAME(csub_k)
 #define ZSUB_K OpenVML_FUNCNAME(zsub_k)
 
+#define SSQR_K OpenVML_FUNCNAME(ssqr_k)
+#define DSQR_K OpenVML_FUNCNAME(dsqr_k)
+
 #define SPOW_K OpenVML_FUNCNAME(spow_k)
 #define DPOW_K OpenVML_FUNCNAME(dpow_k)
 #define CPOW_K OpenVML_FUNCNAME(cpow_k)
@@ -114,6 +117,7 @@
 #ifndef DOUBLE
 #define ADD_K SADD_K
 #define SUB_K SSUB_K
+#define SQR_K SSQR_K
 #define POW_K SPOW_K
 #define POWX_K SPOWX_K
 #define EXP_K SEXP_K
@@ -134,6 +138,7 @@
 #else
 #define ADD_K DADD_K
 #define SUB_K DSUB_K
+#define SQR_K DSQR_K
 #define POW_K DPOW_K
 #define POWX_K DPOWX_K
 #define EXP_K DEXP_K

include/openvml_reference.h

@@ -46,6 +46,9 @@ OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vdSub)(const VML_INT n, const double *
 OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vcSub)(const VML_INT n, const float * a, const float * b, float * y);
 OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vzSub)(const VML_INT n, const double * a, const double * b, double * y);
 
+OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vsSqr)(const VML_INT n, const float * a, float * y);
+OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vdSqr)(const VML_INT n, const double * a, double * y);
+
 OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vsPow)(const VML_INT n, const float * a, const float * b, float * y);
 OPENVML_EXPORT void OpenVML_FUNCNAME_REF(vdPow)(const VML_INT n, const double * a, const double * b, double * y);
 

interface/CMakeLists.txt

@@ -6,7 +6,7 @@ set(OpenVML_LIBSRC_C "")
 set(OpenVML_LIBSRC_Z "")
 
 set(REAL_INTERFACE_LIST 
-  add sub 
+  add sub sqr
   pow powx pow2o3 pow3o2 exp expm1 
   tanh 
   log10 ln log1p 
@@ -108,4 +108,4 @@ Endforeach(INTERFACE)
 
 add_library(openvml_interface_core OBJECT ${OpenVML_LIBSRC_S} ${OpenVML_LIBSRC_D} ${OpenVML_LIBSRC_C} ${OpenVML_LIBSRC_Z} ${OpenVML_LIBSRC_OTHER})
 
-target_compile_definitions(openvml_interface_core PUBLIC openvml_EXPORTS)
+target_compile_definitions(openvml_interface_core PUBLIC openvml_EXPORTS)

interface/sqr.c

@@ -0,0 +1,39 @@
+/* * Copyright (c) 2014, 2015                                          Zhang Xianyi
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice, this
+ *   list of conditions and the following disclaimer in the documentation and/or
+ *   other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <openvml.h>
+#include <openvml_driver.h>
+#include <openvml_kernel.h>
+
+
+void CNAME(const VML_INT n, const VML_FLOAT * a, VML_FLOAT * y) {
+
+  if (n<=0) return;
+  if (a==NULL || y==NULL) return;
+
+
+  EXEC_VML(0, SQR_K, n, (VML_FLOAT*)a, NULL, y, NULL, NULL);
+
+}

kernel/CMakeLists.txt

@@ -7,7 +7,7 @@ set(OpenVML_LIBSRC_C "")
 set(OpenVML_LIBSRC_Z "")
 
 #s,d
-set(KERNEL_LIST add sub pow powx exp expm1 tanh log10 ln log1p floor
+set(KERNEL_LIST add sub sqr pow powx exp expm1 tanh log10 ln log1p floor
   sin cos sincos tan asin acos atan atan2)
 
 #c,z

kernel/aarch64/Kernel_generic.txt

@@ -8,6 +8,9 @@ set(sub_D_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_C_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_Z_KERNEL_SOURCE generic/sub_kernel.c)
 
+set(sqr_S_KERNEL_SOURCE generic/sqr_kernel.c)
+set(sqr_D_KERNEL_SOURCE generic/sqr_kernel.c)
+
 set(pow_S_KERNEL_SOURCE generic/pow_kernel.c)
 set(pow_D_KERNEL_SOURCE generic/pow_kernel.c)
 

kernel/arm/Kernel_generic.txt

@@ -8,6 +8,9 @@ set(sub_D_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_C_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_Z_KERNEL_SOURCE generic/sub_kernel.c)
 
+set(sqr_S_KERNEL_SOURCE generic/sqr_kernel.c)
+set(sqr_D_KERNEL_SOURCE generic/sqr_kernel.c)
+
 set(pow_S_KERNEL_SOURCE generic/pow_kernel.c)
 set(pow_D_KERNEL_SOURCE generic/pow_kernel.c)
 

kernel/generic/Kernel_generic.txt

@@ -8,6 +8,9 @@ set(sub_D_KERNEL_SOURCE ${OpenVML_ARCH}/sub_kernel.c)
 set(sub_C_KERNEL_SOURCE ${OpenVML_ARCH}/sub_kernel.c)
 set(sub_Z_KERNEL_SOURCE ${OpenVML_ARCH}/sub_kernel.c)
 
+set(sqr_S_KERNEL_SOURCE ${OpenVML_ARCH}/sqr_kernel.c)
+set(sqr_D_KERNEL_SOURCE ${OpenVML_ARCH}/sqr_kernel.c)
+
 set(pow_S_KERNEL_SOURCE ${OpenVML_ARCH}/pow_kernel.c)
 set(pow_D_KERNEL_SOURCE ${OpenVML_ARCH}/pow_kernel.c)
 #set(pow_C_KERNEL_SOURCE ${OpenVML_ARCH}/pow_kernel.c)

kernel/generic/sqr_kernel.c

@@ -0,0 +1,33 @@
+/* * Copyright (c) 2014, 2015                                          Zhang Xianyi
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice, this
+ *   list of conditions and the following disclaimer in the documentation and/or
+ *   other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "openvml_kernel.h"
+
+void KERNEL_NAME(VMLLONG n, VML_FLOAT * a, VML_FLOAT * b, VML_FLOAT * y, VML_FLOAT * z, VML_FLOAT * other_params) {
+  VMLLONG i=0;
+  for(i=0; i<COMPSIZE*n; i++){
+    y[i]=a[i]*a[i];
+  }
+}

kernel/x86_64/Kernel_generic.txt

@@ -8,6 +8,9 @@ set(sub_D_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_C_KERNEL_SOURCE generic/sub_kernel.c)
 set(sub_Z_KERNEL_SOURCE generic/sub_kernel.c)
 
+set(sqr_S_KERNEL_SOURCE generic/sqr_kernel.c)
+set(sqr_D_KERNEL_SOURCE generic/sqr_kernel.c)
+
 set(pow_S_KERNEL_SOURCE generic/pow_kernel.c)
 set(pow_D_KERNEL_SOURCE generic/pow_kernel.c)
 

kernel/x86_64/Kernel_haswell.txt

@@ -16,6 +16,9 @@ set(sub_D_KERNEL_SOURCE ${OpenVML_ARCH}/dsub_kernel_avx.c)
 set(sub_C_KERNEL_SOURCE ${OpenVML_ARCH}/ssub_kernel_avx.c)
 set(sub_Z_KERNEL_SOURCE ${OpenVML_ARCH}/dsub_kernel_avx.c)
 
+set(sqr_S_KERNEL_SOURCE ${OpenVML_ARCH}/ssqr_kernel_avx.c)
+set(sqr_D_KERNEL_SOURCE ${OpenVML_ARCH}/dsqr_kernel_avx.c)
+
 set(pow_S_KERNEL_SOURCE ${OpenVML_ARCH}/spow_kernel_avx.c)
 set(pow_D_KERNEL_SOURCE ${OpenVML_ARCH}/dpow_kernel_avx.c)
 

kernel/x86_64/Kernel_sandybridge.txt

@@ -16,6 +16,9 @@ set(sub_D_KERNEL_SOURCE ${OpenVML_ARCH}/dsub_kernel_avx.c)
 set(sub_C_KERNEL_SOURCE ${OpenVML_ARCH}/ssub_kernel_avx.c)
 set(sub_Z_KERNEL_SOURCE ${OpenVML_ARCH}/dsub_kernel_avx.c)
 
+set(sqr_S_KERNEL_SOURCE ${OpenVML_ARCH}/ssqr_kernel_avx.c)
+set(sqr_D_KERNEL_SOURCE ${OpenVML_ARCH}/dsqr_kernel_avx.c)
+
 set(pow_S_KERNEL_SOURCE ${OpenVML_ARCH}/spow_kernel_avx.c)
 set(pow_D_KERNEL_SOURCE ${OpenVML_ARCH}/dpow_kernel_avx.c)
 

kernel/x86_64/dsqr_kernel_avx.c

@@ -0,0 +1,78 @@
+/* * Copyright (c) 2014, 2015                                          Zhang Xianyi
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice, this
+ *   list of conditions and the following disclaimer in the documentation and/or
+ *   other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "openvml_kernel.h"
+
+#include <immintrin.h>
+
+void KERNEL_NAME(VMLLONG n, VML_FLOAT * a, VML_FLOAT * b, VML_FLOAT * y, VML_FLOAT * z, VML_FLOAT * other_params) {
+  VMLLONG loop_count=(COMPSIZE*n) >> 5;
+  VMLLONG remain_count=(COMPSIZE*n) & 0x1f;
+
+  int i=0;
+
+  while(loop_count>0){
+
+    __m256d av0=_mm256_loadu_pd(a);
+    __m256d av1=_mm256_loadu_pd(a+4);
+    __m256d av2=_mm256_loadu_pd(a+8);
+    __m256d av3=_mm256_loadu_pd(a+12);
+
+    __m256d av4=_mm256_loadu_pd(a+16);
+    __m256d av5=_mm256_loadu_pd(a+20);
+    __m256d av6=_mm256_loadu_pd(a+24);
+    __m256d av7=_mm256_loadu_pd(a+28);
+
+
+    __m256d yv0=_mm256_mul_pd(av0, av0);
+    __m256d yv1=_mm256_mul_pd(av1, av1);
+    __m256d yv2=_mm256_mul_pd(av2, av2);
+    __m256d yv3=_mm256_mul_pd(av3, av3);
+
+    __m256d yv4=_mm256_mul_pd(av4, av4);
+    __m256d yv5=_mm256_mul_pd(av5, av5);
+    __m256d yv6=_mm256_mul_pd(av6, av6);
+    __m256d yv7=_mm256_mul_pd(av7, av7);
+
+    _mm256_storeu_pd(y, yv0);
+    _mm256_storeu_pd(y+4, yv1);
+    _mm256_storeu_pd(y+8, yv2);
+    _mm256_storeu_pd(y+12, yv3);
+
+    _mm256_storeu_pd(y+16, yv4);
+    _mm256_storeu_pd(y+20, yv5);
+    _mm256_storeu_pd(y+24, yv6);
+    _mm256_storeu_pd(y+28, yv7);
+
+    a+=32;
+    y+=32;
+    loop_count--;
+  }
+
+  for(i=0; i<remain_count; i++){
+    y[i]=a[i]*a[i];
+  }
+}
+

kernel/x86_64/sqr_kernel.c

@@ -0,0 +1,33 @@
+/* * Copyright (c) 2014, 2015                                          Zhang Xianyi
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice, this
+ *   list of conditions and the following disclaimer in the documentation and/or
+ *   other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "openvml_kernel.h"
+
+void KERNEL_NAME(VMLLONG n, VML_FLOAT * a, VML_FLOAT * b, VML_FLOAT * y, VML_FLOAT * z, VML_FLOAT * other_params) {
+  VMLLONG i=0;
+  for(i=0; i<COMPSIZE*n; i++){
+    y[i]=a[i]*a[i];
+  }
+}

kernel/x86_64/ssqr_kernel_avx.c

@@ -0,0 +1,64 @@
+/* * Copyright (c) 2014, 2015                                          Zhang Xianyi
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice, this
+ *   list of conditions and the following disclaimer in the documentation and/or
+ *   other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "openvml_kernel.h"
+
+#include <immintrin.h>
+
+void KERNEL_NAME(VMLLONG n, VML_FLOAT * a, VML_FLOAT * b, VML_FLOAT * y, VML_FLOAT * z, VML_FLOAT * other_params) {
+  VMLLONG loop_count=(COMPSIZE*n) >> 5;
+  VMLLONG remain_count=(COMPSIZE*n) & 0x1f;
+
+  int i=0;
+
+  while(loop_count>0){
+
+    __m256 av0=_mm256_loadu_ps(a);
+    __m256 av1=_mm256_loadu_ps(a+8);
+    __m256 av2=_mm256_loadu_ps(a+16);
+    __m256 av3=_mm256_loadu_ps(a+24);
+
+
+    __m256 yv0=_mm256_mul_ps(av0, av0);
+    __m256 yv1=_mm256_mul_ps(av1, av1);
+    __m256 yv2=_mm256_mul_ps(av2, av2);
+    __m256 yv3=_mm256_mul_ps(av3, av3);
+
+
+    _mm256_storeu_ps(y, yv0);
+    _mm256_storeu_ps(y+8, yv1);
+    _mm256_storeu_ps(y+16, yv2);
+    _mm256_storeu_ps(y+24, yv3);
+
+    a+=32;
+    y+=32;
+    loop_count--;
+  }
+
+  for(i=0; i<remain_count; i++){
+    y[i]=a[i]*a[i];
+  }
+}
+