ARM: introduced 'fetch_mask_pixblock' macro to simplify code

This macro hides the implementation details of pixels fetching
for the mask image just like 'fetch_src_pixblock' does for the
source image. This provides more possibilities for reusing the
same code blocks in different compositing functions.

This patch does not introduce any functional changes and the
resulting code in the compiled object file is exactly the same.

2 files changed, 18 insertions, 13 deletions

diff --git a/pixman/pixman-arm-neon-asm.S b/pixman/pixman-arm-neon-asm.S
index a3875ee8..155a236e 100644
--- a/pixman/pixman-arm-neon-asm.S
+++ b/pixman/pixman-arm-neon-asm.S
@@ -841,7 +841,7 @@ generate_composite_function \
     pixman_composite_over_n_8_0565_process_pixblock_tail
     vst1.16     {d28, d29}, [DST_W, :128]!
     vld1.16     {d4, d5}, [DST_R, :128]!
-    vld1.8      {d24}, [MASK]!
+    fetch_mask_pixblock
     cache_preload 8, 8
     pixman_composite_over_n_8_0565_process_pixblock_head
 .endm
@@ -889,7 +889,7 @@ generate_composite_function \
     pixman_composite_over_n_8_0565_process_pixblock_tail
     fetch_src_pixblock
     cache_preload 8, 8
-    vld1.8      {d24}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_over_n_8_0565_process_pixblock_head
     vst1.16     {d28, d29}, [DST_W, :128]!
 .endm
@@ -1171,7 +1171,7 @@ generate_composite_function \
     pixman_composite_over_n_8_8888_process_pixblock_tail
     vst4.8      {d28, d29, d30, d31}, [DST_W, :128]!
     vld4.8      {d4, d5, d6, d7}, [DST_R, :128]!
-    vld1.8      {d24}, [MASK]!
+    fetch_mask_pixblock
     cache_preload 8, 8
     pixman_composite_over_n_8_8888_process_pixblock_head
 .endm
@@ -1241,7 +1241,7 @@ generate_composite_function \
 .macro pixman_composite_over_n_8_8_process_pixblock_tail_head
     vld1.8      {d4, d5, d6, d7}, [DST_R, :128]!
     pixman_composite_over_n_8_8_process_pixblock_tail
-    vld1.8      {d24, d25, d26, d27}, [MASK]!
+    fetch_mask_pixblock
     cache_preload 32, 32
     vst1.8      {d28, d29, d30, d31}, [DST_W, :128]!
     pixman_composite_over_n_8_8_process_pixblock_head
@@ -1341,7 +1341,7 @@ generate_composite_function \
         vraddhn.u16 d29, q15, q9
         vraddhn.u16 d30, q6, q10
         vraddhn.u16 d31, q7, q11
-    vld4.8      {d24, d25, d26, d27}, [MASK]!
+    fetch_mask_pixblock
         vqadd.u8    q14, q0, q14
         vqadd.u8    q15, q1, q15
     cache_preload 8, 8
@@ -1405,7 +1405,7 @@ generate_composite_function \
     pixman_composite_add_n_8_8_process_pixblock_tail
     vst1.8      {d28, d29, d30, d31}, [DST_W, :128]!
     vld1.8      {d4, d5, d6, d7}, [DST_R, :128]!
-    vld1.8      {d24, d25, d26, d27}, [MASK]!
+    fetch_mask_pixblock
     cache_preload 32, 32
     pixman_composite_add_n_8_8_process_pixblock_head
 .endm
@@ -1462,7 +1462,7 @@ generate_composite_function \
     pixman_composite_add_8_8_8_process_pixblock_tail
     vst1.8      {d28, d29, d30, d31}, [DST_W, :128]!
     vld1.8      {d4, d5, d6, d7}, [DST_R, :128]!
-    vld1.8      {d24, d25, d26, d27}, [MASK]!
+    fetch_mask_pixblock
     fetch_src_pixblock
     cache_preload 32, 32
     pixman_composite_add_8_8_8_process_pixblock_head
@@ -1515,7 +1515,7 @@ generate_composite_function \
     pixman_composite_add_8888_8888_8888_process_pixblock_tail
     vst4.8      {d28, d29, d30, d31}, [DST_W, :128]!
     vld4.8      {d4, d5, d6, d7}, [DST_R, :128]!
-    vld4.8      {d24, d25, d26, d27}, [MASK]!
+    fetch_mask_pixblock
     fetch_src_pixblock
     cache_preload 8, 8
     pixman_composite_add_8888_8888_8888_process_pixblock_head
@@ -1587,7 +1587,7 @@ generate_composite_function_single_scanline \
     pixman_composite_out_reverse_8888_n_8888_process_pixblock_tail
     fetch_src_pixblock
     cache_preload 8, 8
-    vld4.8     {d12, d13, d14, d15}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_out_reverse_8888_n_8888_process_pixblock_head
     vst4.8     {d28, d29, d30, d31}, [DST_W, :128]!
 .endm
@@ -1658,7 +1658,7 @@ generate_composite_function \
     pixman_composite_over_8888_n_8888_process_pixblock_tail
     fetch_src_pixblock
     cache_preload 8, 8
-    vld4.8     {d12, d13, d14, d15}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_over_8888_n_8888_process_pixblock_head
     vst4.8     {d28, d29, d30, d31}, [DST_W, :128]!
 .endm
@@ -1700,7 +1700,7 @@ generate_composite_function_single_scanline \
     pixman_composite_over_8888_n_8888_process_pixblock_tail
     fetch_src_pixblock
     cache_preload 8, 8
-    vld1.8     {d15}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_over_8888_n_8888_process_pixblock_head
     vst4.8     {d28, d29, d30, d31}, [DST_W, :128]!
 .endm
@@ -1917,7 +1917,7 @@ generate_composite_function \
 
 /* TODO: expand macros and do better instructions scheduling */
 .macro pixman_composite_over_0565_8_0565_process_pixblock_tail_head
-    vld1.8     {d15}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_over_0565_8_0565_process_pixblock_tail
     fetch_src_pixblock
     vld1.16    {d10, d11}, [DST_R, :128]!
@@ -1969,7 +1969,7 @@ generate_composite_function \
 
 /* TODO: expand macros and do better instructions scheduling */
 .macro pixman_composite_add_0565_8_0565_process_pixblock_tail_head
-    vld1.8     {d15}, [MASK]!
+    fetch_mask_pixblock
     pixman_composite_add_0565_8_0565_process_pixblock_tail
     fetch_src_pixblock
     vld1.16    {d10, d11}, [DST_R, :128]!
diff --git a/pixman/pixman-arm-neon-asm.h b/pixman/pixman-arm-neon-asm.h
index c75bdc37..24fa3616 100644
--- a/pixman/pixman-arm-neon-asm.h
+++ b/pixman/pixman-arm-neon-asm.h
@@ -431,6 +431,11 @@
 .endif
 .endm
 
+.macro fetch_mask_pixblock
+    pixld       pixblock_size, mask_bpp, \
+                (mask_basereg - pixblock_size * mask_bpp / 64), MASK
+.endm
+
 /*
  * Macro which is used to process leading pixels until destination
  * pointer is properly aligned (at 16 bytes boundary). When destination