Added optional row preference to ukernel config.

Details:
- Added the ability for the kernel developer to indicate the gemm micro-
  kernel as having a preference for accessing the micro-tile of C via
  contiguous rows (as opposed to contiguous columns). This property may
  be encoded in bli_kernel.h as BLIS_?GEMM_UKERNEL_PREFERS_CONTIG_ROWS,
  which may be defined or left undefined. Leaving it undefined leads to
  the default assumption of column preference.
- Changed conditionals in frame/3/*/*_front.c that induce transposition
  of the operation so that the transposition is induced only if there
  is disagreement between the storage of C and the preference of the
  micro-kernel. Previously, the only conditional that needed to be met
  was that C was row-stored, which is to say that we assumed the micro-
  kernel preferred column-contiguous access on C.
- Added a "prefers_contig_rows" property to func_t objects, and updated
  calls to bli_func_obj_create() in _cntl.c files in order to support
  the above changes.
- Removed the row-storage optimization from bli_trsm_front.c because
  it is actually ineffective. This is because the right-side case of
  trsm flips the A and B micro-panel operands (since BLIS only requires
  left-side gemmtrsm/trsm kernels), meaning any transposition done
  at the high level is then undone at the low level.
- Tweaked trmm, trmm3 _front.c files to eliminate a possible redundant
  invocation of the bli_obj_swap() macro.

frame/include/bli_kernel_macro_defs.h

@@ -36,16 +36,53 @@
 #define BLIS_KERNEL_MACRO_DEFS_H
 
 
-// -- Construct kernel function names ------------------------------------------
+// -- Define row access bools --------------------------------------------------
+
+// In this section we consider each datatype-specific "prefers contiguous rows"
+// macro. If it is defined, we re-define it to be 1 (TRUE); otherwise, we
+// define it to be 0 (FALSE).
+
+// gemm micro-kernels
+
+#ifdef  BLIS_SGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#undef  BLIS_SGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#define BLIS_SGEMM_UKERNEL_PREFERS_CONTIG_ROWS 1 
+#else
+#define BLIS_SGEMM_UKERNEL_PREFERS_CONTIG_ROWS 0 
+#endif
+
+#ifdef  BLIS_DGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#undef  BLIS_DGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#define BLIS_DGEMM_UKERNEL_PREFERS_CONTIG_ROWS 1 
+#else
+#define BLIS_DGEMM_UKERNEL_PREFERS_CONTIG_ROWS 0 
+#endif
+
+#ifdef  BLIS_CGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#undef  BLIS_CGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#define BLIS_CGEMM_UKERNEL_PREFERS_CONTIG_ROWS 1 
+#else
+#define BLIS_CGEMM_UKERNEL_PREFERS_CONTIG_ROWS 0 
+#endif
+
+#ifdef  BLIS_ZGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#undef  BLIS_ZGEMM_UKERNEL_PREFERS_CONTIG_ROWS
+#define BLIS_ZGEMM_UKERNEL_PREFERS_CONTIG_ROWS 1 
+#else
+#define BLIS_ZGEMM_UKERNEL_PREFERS_CONTIG_ROWS 0 
+#endif
+
+
+// -- Define default kernel names ----------------------------------------------
 
 // In this section we consider each datatype-specific micro-kernel macro;
 // if it is undefined, we define it to be the corresponding reference kernel.
-// In the case of complex gemm micro-kernels, we also define special _VIA_4M
-// macros so that later on we can tell whether or not to employ the 4m
-// implementations. Note that in order to properly determine whether 4m is a
-// viable option, we need to be able to test the existence of the real gemm
-// micro-kernels, which means we must consider the complex gemm micro-kernel
-// cases *BEFORE* the real cases.
+// In the case of complex gemm micro-kernels, we also define special macros so
+// that later on we can tell whether or not to employ the 4m implementations.
+// Note that in order to properly determine whether/ 4m is a viable option, we
+// need to be able to test the existence of the real gemm micro-kernels, which
+// means we must consider the complex gemm micro-kernel cases *BEFORE* the
+// real cases.
 
 //
 // Level-3