[Pixman] [PATCH 2/5] vmx: adjust macros when loading vectors on ppc64le

[Oded Gabbay]

From: Fernando Seiti Furusato <ferseiti at linux.vnet.ibm.com>

Replaced usage of vec_lvsl to direct unaligned assignment
operation (=). That is because, according to Power ABI Specification,
the usage of lvsl is deprecated on ppc64le.

Changed COMPUTE_SHIFT_{MASK,MASKS,MASKC} macro usage to no-op for powerpc
little endian since unaligned access is supported on ppc64le.

Signed-off-by: Fernando Seiti Furusato <ferseiti at linux.vnet.ibm.com>
---
 pixman/pixman-vmx.c | 54 ++++++++++++++++++++++++++++++++++++++---------------
 1 file changed, 39 insertions(+), 15 deletions(-)

diff --git a/pixman/pixman-vmx.c b/pixman/pixman-vmx.c
index 23537f6..7e5eca7 100644
--- a/pixman/pixman-vmx.c
+++ b/pixman/pixman-vmx.c
@@ -136,6 +136,7 @@ over (vector unsigned int src,
     over (pix_multiply (src, mask),					\
           pix_multiply (srca, mask), dest)
 
+#ifndef _LITTLE_ENDIAN
 
 #define COMPUTE_SHIFT_MASK(source)					\
     source ## _mask = vec_lvsl (0, source);
@@ -151,23 +152,46 @@ over (vector unsigned int src,
  * Note: tmp3 and tmp4 must remain untouched!
  */
 
-#define LOAD_VECTORS(dest, source)			  \
-    tmp1 = (typeof(tmp1))vec_ld (0, source);		  \
-    tmp2 = (typeof(tmp2))vec_ld (15, source);		  \
-    v ## source = (typeof(v ## source))			  \
-	vec_perm (tmp1, tmp2, source ## _mask);		  \
+#define LOAD_VECTORS(dest, source)            \
+    tmp1 = (typeof(tmp1))vec_ld (0, source);          \
+    tmp2 = (typeof(tmp2))vec_ld (15, source);         \
+    v ## source = (typeof(v ## source))           \
+    vec_perm (tmp1, tmp2, source ## _mask);       \
     v ## dest = (typeof(v ## dest))vec_ld (0, dest);
 
-#define LOAD_VECTORSC(dest, source, mask)		  \
-    tmp1 = (typeof(tmp1))vec_ld (0, source);		  \
-    tmp2 = (typeof(tmp2))vec_ld (15, source);		  \
-    v ## source = (typeof(v ## source))			  \
-	vec_perm (tmp1, tmp2, source ## _mask);		  \
-    tmp1 = (typeof(tmp1))vec_ld (0, mask);		  \
-    v ## dest = (typeof(v ## dest))vec_ld (0, dest);	  \
-    tmp2 = (typeof(tmp2))vec_ld (15, mask);		  \
-    v ## mask = (typeof(v ## mask))			  \
-	vec_perm (tmp1, tmp2, mask ## _mask);
+#define LOAD_VECTORSC(dest, source, mask)         \
+    tmp1 = (typeof(tmp1))vec_ld (0, source);          \
+    tmp2 = (typeof(tmp2))vec_ld (15, source);         \
+    v ## source = (typeof(v ## source))           \
+    vec_perm (tmp1, tmp2, source ## _mask);       \
+    tmp1 = (typeof(tmp1))vec_ld (0, mask);        \
+    v ## dest = (typeof(v ## dest))vec_ld (0, dest);      \
+    tmp2 = (typeof(tmp2))vec_ld (15, mask);       \
+    v ## mask = (typeof(v ## mask))           \
+    vec_perm (tmp1, tmp2, mask ## _mask);
+#else
+
+/* Now the COMPUTE_SHIFT_{MASK, MASKS, MASKC} below are just no-op.
+ * They are defined that way because little endian altivec can do unaligned
+ * reads natively and have no need for constructing the permutation pattern
+ * variables.
+ */
+#define COMPUTE_SHIFT_MASK(source)
+
+#define COMPUTE_SHIFT_MASKS(dest, source)
+
+#define COMPUTE_SHIFT_MASKC(dest, source, mask)
+
+# define LOAD_VECTORS(dest, source)                        \
+    v ## source = *((typeof(v ## source)*)source);        \
+    v ## dest = *((typeof(v ## dest)*)dest);
+
+# define LOAD_VECTORSC(dest, source, mask)                 \
+    v ## source = *((typeof(v ## source)*)source);        \
+    v ## dest = *((typeof(v ## dest)*)dest);              \
+    v ## mask = *((typeof(v ## mask)*)mask);
+
+#endif /* _LITTLE_ENDIAN */
 
 #define LOAD_VECTORSM(dest, source, mask)				\
     LOAD_VECTORSC (dest, source, mask)					\
-- 
2.4.3