Count cpu cores on ARMV8 and use that to pick the GEMM_PQ parameters (#2267)

There is currently no simple way to query cache sizes on ARMV8, so this takes the number of cores as a trivial indication if the target is a server-class device with a big cache, or just a single-board toy or smartphone.

cpuid_arm64.c

@@ -206,6 +206,33 @@ void get_subdirname(void)
 	printf("arm64");
 }
 
+void get_cpucount(void)
+{
+int n=0;
+
+#ifdef linux
+	FILE *infile;
+  	char buffer[2048], *p,*t;
+  	p = (char *) NULL ;
+
+  	infile = fopen("/proc/cpuinfo", "r");
+
+	while (fgets(buffer, sizeof(buffer), infile))
+	{
+
+		if (!strncmp("processor", buffer, 9))
+		n++;
+  	}
+
+  	fclose(infile);
+
+	printf("#define NUM_CORES %d\n",n);
+#endif
+
+}
+
+
+
 void get_cpuconfig(void)
 {
 
@@ -309,6 +336,7 @@ void get_cpuconfig(void)
 			printf("#define DTB_SIZE             4096     \n");
 			break;	
 	}
+	get_cpucount();
 }
 
 
@@ -351,5 +379,3 @@ void get_features(void)
 #endif
 	return;
 }
-
-

param.h

@@ -2636,15 +2636,30 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 #define ZGEMM_DEFAULT_UNROLL_M  4
 #define ZGEMM_DEFAULT_UNROLL_N  4
 
-#define SGEMM_DEFAULT_P	512
-#define DGEMM_DEFAULT_P	256
-#define CGEMM_DEFAULT_P 256
-#define ZGEMM_DEFAULT_P 128
+/*FIXME: this should be using the cache size, but there is currently no easy way to
+query that on ARM. So if getarch counted more than 8 cores we simply assume the host
+is a big desktop or server with abundant cache rather than a phone or embedded device */ 
+#if NUM_CORES > 8
+  #define SGEMM_DEFAULT_P 512
+  #define DGEMM_DEFAULT_P 256
+  #define CGEMM_DEFAULT_P 256
+  #define ZGEMM_DEFAULT_P 128
+
+  #define SGEMM_DEFAULT_Q 1024
+  #define DGEMM_DEFAULT_Q 512
+  #define CGEMM_DEFAULT_Q 512
+  #define ZGEMM_DEFAULT_Q 512
+#else
+  #define SGEMM_DEFAULT_P 128
+  #define DGEMM_DEFAULT_P 160
+  #define CGEMM_DEFAULT_P 128
+  #define ZGEMM_DEFAULT_P 128
 
-#define SGEMM_DEFAULT_Q 1024
-#define DGEMM_DEFAULT_Q 512
-#define CGEMM_DEFAULT_Q 512
-#define ZGEMM_DEFAULT_Q 512
+  #define SGEMM_DEFAULT_Q 352
+  #define DGEMM_DEFAULT_Q 128
+  #define CGEMM_DEFAULT_Q 224
+  #define ZGEMM_DEFAULT_Q 112
+#endif
 
 #define SGEMM_DEFAULT_R 4096
 #define DGEMM_DEFAULT_R 4096