add optimized daxpy-kernel for steamroller

kernel/x86_64/daxpy.c

@@ -31,8 +31,10 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 #if defined(NEHALEM)
 #include "daxpy_microk_nehalem-2.c"
-#elif defined(BULLDOZER) || defined(PILEDRIVER) || defined(STEAMROLLER)
+#elif defined(BULLDOZER) || defined(PILEDRIVER)
 #include "daxpy_microk_bulldozer-2.c"
+#elif defined(STEAMROLLER)
+#include "daxpy_microk_steamroller-2.c"
 #elif defined(HASWELL)
 #include "daxpy_microk_haswell-2.c"
 #elif defined(SANDYBRIDGE)

kernel/x86_64/daxpy_microk_steamroller-2.c

@@ -0,0 +1,160 @@
+/***************************************************************************
+Copyright (c) 2014, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+1. Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+2. 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.
+3. Neither the name of the OpenBLAS project nor the names of
+its contributors may be used to endorse or promote products
+derived from this software without specific prior written permission.
+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 OPENBLAS PROJECT 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.
+*****************************************************************************/
+
+#define HAVE_KERNEL_8 1
+static void daxpy_kernel_8( BLASLONG n, FLOAT *x, FLOAT *y , FLOAT *alpha) __attribute__ ((noinline));
+
+static void daxpy_kernel_8( BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *alpha)
+{
+
+
+	BLASLONG register i = 0;
+
+  if ( n < 2048 )
+  {
+
+	__asm__  __volatile__
+	(
+	"vmovddup		(%4), %%xmm0		    		\n\t"  // alpha	
+
+	".align 16				            		\n\t"
+	"1:				            			\n\t"
+
+        "vmovups                  (%3,%0,8), %%xmm8          		\n\t"  // 2  y
+        "vmovups                16(%3,%0,8), %%xmm9         		\n\t"  // 2  y
+        "vmovups                32(%3,%0,8), %%xmm10         		\n\t"  // 2  y
+        "vmovups                48(%3,%0,8), %%xmm11         		\n\t"  // 2  y
+
+        "vmovups                64(%3,%0,8), %%xmm12         		\n\t"  // 2  y
+        "vmovups                80(%3,%0,8), %%xmm13        		\n\t"  // 2  y
+        "vmovups                96(%3,%0,8), %%xmm14         		\n\t"  // 2  y
+        "vmovups               112(%3,%0,8), %%xmm15         		\n\t"  // 2  y
+
+	"vfmadd231pd       (%2,%0,8), %%xmm0 , %%xmm8  			\n\t"   // y += alpha * x
+	"vfmadd231pd     16(%2,%0,8), %%xmm0 , %%xmm9 		  	\n\t"   // y += alpha * x
+	"vfmadd231pd     32(%2,%0,8), %%xmm0 , %%xmm10 			\n\t"   // y += alpha * x
+	"vfmadd231pd     48(%2,%0,8), %%xmm0 , %%xmm11 			\n\t"   // y += alpha * x
+
+	"vfmadd231pd     64(%2,%0,8), %%xmm0 , %%xmm12 			\n\t"   // y += alpha * x
+	"vfmadd231pd     80(%2,%0,8), %%xmm0 , %%xmm13		  	\n\t"   // y += alpha * x
+	"vfmadd231pd     96(%2,%0,8), %%xmm0 , %%xmm14 			\n\t"   // y += alpha * x
+	"vfmadd231pd    112(%2,%0,8), %%xmm0 , %%xmm15 			\n\t"   // y += alpha * x
+
+	"vmovups		%%xmm8 ,   (%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm9 , 16(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm10, 32(%3,%0,8)           		\n\t"
+	"vmovups		%%xmm11, 48(%3,%0,8)                 	\n\t"
+
+	"vmovups		%%xmm12, 64(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm13, 80(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm14, 96(%3,%0,8)           		\n\t"
+	"vmovups		%%xmm15,112(%3,%0,8)                 	\n\t"
+
+	"addq		$16, %0	  	 	             		\n\t"
+	"subq	        $16, %1			             		\n\t"		
+	"jnz		1b		             			\n\t"
+
+	:
+        : 
+          "r" (i),	// 0	
+	  "r" (n),  	// 1
+          "r" (x),      // 2
+          "r" (y),      // 3
+          "r" (alpha)   // 4
+	: "cc", 
+	  "%xmm0", 
+	  "%xmm8", "%xmm9", "%xmm10", "%xmm11", 
+	  "%xmm12", "%xmm13", "%xmm14", "%xmm15",
+	  "memory"
+	);
+	return;
+  }
+
+
+	__asm__  __volatile__
+	(
+	"vmovddup		(%4), %%xmm0		    		\n\t"  // alpha	
+
+	".align 16				            		\n\t"
+	"1:				            			\n\t"
+
+	"prefetcht0	       512(%3,%0,8)				\n\t"
+        "vmovups                  (%3,%0,8), %%xmm8          		\n\t"  // 2  y
+        "vmovups                16(%3,%0,8), %%xmm9         		\n\t"  // 2  y
+        "vmovups                32(%3,%0,8), %%xmm10         		\n\t"  // 2  y
+        "vmovups                48(%3,%0,8), %%xmm11         		\n\t"  // 2  y
+
+	"prefetcht0	       576(%3,%0,8)				\n\t"
+        "vmovups                64(%3,%0,8), %%xmm12         		\n\t"  // 2  y
+        "vmovups                80(%3,%0,8), %%xmm13        		\n\t"  // 2  y
+        "vmovups                96(%3,%0,8), %%xmm14         		\n\t"  // 2  y
+        "vmovups               112(%3,%0,8), %%xmm15         		\n\t"  // 2  y
+
+	"prefetcht0	       512(%2,%0,8)				\n\t"
+	"vfmadd231pd       (%2,%0,8), %%xmm0 , %%xmm8  			\n\t"   // y += alpha * x
+	"vfmadd231pd     16(%2,%0,8), %%xmm0 , %%xmm9 		  	\n\t"   // y += alpha * x
+	"vfmadd231pd     32(%2,%0,8), %%xmm0 , %%xmm10 			\n\t"   // y += alpha * x
+	"vfmadd231pd     48(%2,%0,8), %%xmm0 , %%xmm11 			\n\t"   // y += alpha * x
+
+	"prefetcht0	       576(%2,%0,8)				\n\t"
+	"vfmadd231pd     64(%2,%0,8), %%xmm0 , %%xmm12 			\n\t"   // y += alpha * x
+	"vfmadd231pd     80(%2,%0,8), %%xmm0 , %%xmm13		  	\n\t"   // y += alpha * x
+	"vfmadd231pd     96(%2,%0,8), %%xmm0 , %%xmm14 			\n\t"   // y += alpha * x
+	"vfmadd231pd    112(%2,%0,8), %%xmm0 , %%xmm15 			\n\t"   // y += alpha * x
+
+	"vmovups		%%xmm8 ,   (%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm9 , 16(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm10, 32(%3,%0,8)           		\n\t"
+	"vmovups		%%xmm11, 48(%3,%0,8)                 	\n\t"
+
+	"vmovups		%%xmm12, 64(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm13, 80(%3,%0,8)                 	\n\t"
+	"vmovups		%%xmm14, 96(%3,%0,8)           		\n\t"
+	"vmovups		%%xmm15,112(%3,%0,8)                 	\n\t"
+
+	"addq		$16, %0	  	 	             		\n\t"
+	"subq	        $16, %1			             		\n\t"		
+	"jnz		1b		             			\n\t"
+
+	:
+        : 
+          "r" (i),	// 0	
+	  "r" (n),  	// 1
+          "r" (x),      // 2
+          "r" (y),      // 3
+          "r" (alpha)   // 4
+	: "cc", 
+	  "%xmm0", 
+	  "%xmm8", "%xmm9", "%xmm10", "%xmm11", 
+	  "%xmm12", "%xmm13", "%xmm14", "%xmm15",
+	  "memory"
+	);
+
+
+} 
+
+