From 29003c3befe2159396d181ef9ac1caaadcabf382 Mon Sep 17 00:00:00 2001 From: Siarhei Siamashka Date: Wed, 9 Mar 2011 13:21:53 +0200 Subject: ARM: NEON optimization for bilinear scaled 'src_0565_x888' Benchmark on ARM Cortex-A8 r2p2 @1GHz, 32-bit LPDDR @200MHz: Microbenchmark (scaling 2000x2000 image with scale factor close to 1x): before: op=1, src=10020565, dst=20020888, speed=3.39 MPix/s after: op=1, src=10020565, dst=20020888, speed=36.82 MPix/s --- pixman/pixman-arm-neon-asm.S | 3 +++ pixman/pixman-arm-neon.c | 4 ++++ 2 files changed, 7 insertions(+) diff --git a/pixman/pixman-arm-neon-asm.S b/pixman/pixman-arm-neon-asm.S index f0b42ca..9245db9 100644 --- a/pixman/pixman-arm-neon-asm.S +++ b/pixman/pixman-arm-neon-asm.S @@ -2628,3 +2628,6 @@ generate_bilinear_scanline_func \ generate_bilinear_scanline_func \ pixman_scaled_bilinear_scanline_8888_0565_SRC_asm_neon, 8888, 0565, 2, 28 + +generate_bilinear_scanline_func \ + pixman_scaled_bilinear_scanline_0565_x888_SRC_asm_neon, 0565, 8888, 1, 28 diff --git a/pixman/pixman-arm-neon.c b/pixman/pixman-arm-neon.c index ba6de66..18e26eb 100644 --- a/pixman/pixman-arm-neon.c +++ b/pixman/pixman-arm-neon.c @@ -131,6 +131,8 @@ PIXMAN_ARM_BIND_SCALED_BILINEAR_SRC_DST (0, neon, 8888_8888, SRC, uint32_t, uint32_t) PIXMAN_ARM_BIND_SCALED_BILINEAR_SRC_DST (0, neon, 8888_0565, SRC, uint32_t, uint16_t) +PIXMAN_ARM_BIND_SCALED_BILINEAR_SRC_DST (0, neon, 0565_x888, SRC, + uint16_t, uint32_t) void pixman_composite_src_n_8_asm_neon (int32_t w, @@ -355,6 +357,8 @@ static const pixman_fast_path_t arm_neon_fast_paths[] = SIMPLE_BILINEAR_FAST_PATH (SRC, a8r8g8b8, r5g6b5, neon_8888_0565), SIMPLE_BILINEAR_FAST_PATH (SRC, x8r8g8b8, r5g6b5, neon_8888_0565), + SIMPLE_BILINEAR_FAST_PATH (SRC, r5g6b5, x8r8g8b8, neon_0565_x888), + { PIXMAN_OP_NONE }, }; -- cgit v1.2.3