Switch away from aligning the indexes to cacheline size by default

The users using the library on hosted platforms with multicore scenarios should pass it using the build system for extra performance.

The performance impact of false sharing doesn't justify space inefficiency of instances for embedded users if they leave the library unconfigured.

Author: DNedic

lfbb/inc/lfbb.h

@@ -62,7 +62,7 @@ extern "C" {
 /************************** DEFINE ****************************/
 
 #ifndef LFBB_MULTICORE_HOSTED
-#define LFBB_MULTICORE_HOSTED true
+#define LFBB_MULTICORE_HOSTED false
 #endif
 
 #ifndef LFBB_CACHELINE_LENGTH