[8/9,v2] armv7: More use of fast paths with localized destination alpha

Submitted by Ben Avison on April 11, 2016, 12:26 p.m.

Details

Message ID 1460377590-23285-9-git-send-email-bavison@riscosopen.org
State New
Series "Changes to existing ARMv7 routines"
Headers show

Commit Message

Ben Avison April 11, 2016, 12:26 p.m.
There are a group of combiner types - SRC, OVER, IN_REVERSE, OUT_REVERSE
and ADD - where the destination alpha component is only used (if at all) to
determine the destination alpha component. This means that any such fast
paths with an a8r8g8b8 destination can also be applied to an x8r8g8b8
destination just by updating the fast path table, and likewise with
a8b8g8r8 and x8b8g8r8. The following operations are affected:

over_8888_8888_x888
add_n_8_x888
add_8888_8_x888
add_8888_8888_x888
add_8888_n_x888
add_8888_x888
out_reverse_8_x888

v2: Changed summary line to make it distinct from similar patch relating
to localized source alpha

Signed-off-by: Ben Avison <bavison@riscosopen.org>
---
 pixman/pixman-arm-neon.c |   12 ++++++++++++
 1 files changed, 12 insertions(+), 0 deletions(-)

Patch hide | download patch | download mbox

diff --git a/pixman/pixman-arm-neon.c b/pixman/pixman-arm-neon.c
index be761c9..5f0561a 100644
--- a/pixman/pixman-arm-neon.c
+++ b/pixman/pixman-arm-neon.c
@@ -331,6 +331,7 @@  static const pixman_fast_path_t arm_neon_fast_paths[] =
     PIXMAN_STD_FAST_PATH (OVER, a8b8g8r8, a8,       b5g6r5,   neon_composite_over_8888_8_0565),
     PIXMAN_STD_FAST_PATH (OVER, r5g6b5,   a8,       r5g6b5,   neon_composite_over_0565_8_0565),
     PIXMAN_STD_FAST_PATH (OVER, b5g6r5,   a8,       b5g6r5,   neon_composite_over_0565_8_0565),
+    PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, a8r8g8b8, x8r8g8b8, neon_composite_over_8888_8888_8888),
     PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, a8r8g8b8, a8r8g8b8, neon_composite_over_8888_8888_8888),
     PIXMAN_STD_FAST_PATH (OVER, a8r8g8b8, null,     r5g6b5,   neon_composite_over_8888_0565),
     PIXMAN_STD_FAST_PATH (OVER, a8b8g8r8, null,     b5g6r5,   neon_composite_over_8888_0565),
@@ -341,17 +342,26 @@  static const pixman_fast_path_t arm_neon_fast_paths[] =
     PIXMAN_STD_FAST_PATH (OVER, x8r8g8b8, null,     a8r8g8b8, neon_composite_src_x888_8888),
     PIXMAN_STD_FAST_PATH (OVER, x8b8g8r8, null,     a8b8g8r8, neon_composite_src_x888_8888),
     PIXMAN_STD_FAST_PATH (ADD,  solid,    a8,       a8,       neon_composite_add_n_8_8),
+    PIXMAN_STD_FAST_PATH (ADD,  solid,    a8,       x8r8g8b8, neon_composite_add_n_8_8888),
     PIXMAN_STD_FAST_PATH (ADD,  solid,    a8,       a8r8g8b8, neon_composite_add_n_8_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  solid,    a8,       x8b8g8r8, neon_composite_add_n_8_8888),
     PIXMAN_STD_FAST_PATH (ADD,  solid,    a8,       a8b8g8r8, neon_composite_add_n_8_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8,       a8,       a8,       neon_composite_add_8_8_8),
     PIXMAN_STD_FAST_PATH (ADD,  r5g6b5,   a8,       r5g6b5,   neon_composite_add_0565_8_0565),
     PIXMAN_STD_FAST_PATH (ADD,  b5g6r5,   a8,       b5g6r5,   neon_composite_add_0565_8_0565),
+    PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, a8,       x8r8g8b8, neon_composite_add_8888_8_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, a8,       x8b8g8r8, neon_composite_add_8888_8_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, a8,       a8r8g8b8, neon_composite_add_8888_8_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, a8,       a8b8g8r8, neon_composite_add_8888_8_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, a8r8g8b8, x8r8g8b8, neon_composite_add_8888_8888_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, a8r8g8b8, a8r8g8b8, neon_composite_add_8888_8888_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, solid,    x8r8g8b8, neon_composite_add_8888_n_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, solid,    x8b8g8r8, neon_composite_add_8888_n_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, solid,    a8r8g8b8, neon_composite_add_8888_n_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, solid,    a8b8g8r8, neon_composite_add_8888_n_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8,       null,     a8,       neon_composite_add_8_8),
+    PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, null,     x8r8g8b8, neon_composite_add_8888_8888),
+    PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, null,     x8b8g8r8, neon_composite_add_8888_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8r8g8b8, null,     a8r8g8b8, neon_composite_add_8888_8888),
     PIXMAN_STD_FAST_PATH (ADD,  a8b8g8r8, null,     a8b8g8r8, neon_composite_add_8888_8888),
     PIXMAN_STD_FAST_PATH (IN,   solid,    null,     a8,       neon_composite_in_n_8),
@@ -359,7 +369,9 @@  static const pixman_fast_path_t arm_neon_fast_paths[] =
     PIXMAN_STD_FAST_PATH (OVER_REVERSE, solid, null, a8b8g8r8, neon_composite_over_reverse_n_8888),
     PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, r5g6b5,   neon_composite_out_reverse_8_0565),
     PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, b5g6r5,   neon_composite_out_reverse_8_0565),
+    PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, x8r8g8b8, neon_composite_out_reverse_8_8888),
     PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, a8r8g8b8, neon_composite_out_reverse_8_8888),
+    PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, x8b8g8r8, neon_composite_out_reverse_8_8888),
     PIXMAN_STD_FAST_PATH (OUT_REVERSE,  a8,    null, a8b8g8r8, neon_composite_out_reverse_8_8888),
 
     SIMPLE_NEAREST_FAST_PATH (OVER, a8r8g8b8, a8r8g8b8, neon_8888_8888),