Half pixel variance further optimized for ARMv6
authorTero Rintaluoma <teror@google.com>
Mon, 28 Mar 2011 06:51:51 +0000 (09:51 +0300)
committerTero Rintaluoma <teror@google.com>
Mon, 28 Mar 2011 06:51:51 +0000 (09:51 +0300)
commitf5e433464b6a0a79978d966bd666c794415abdce
treeaf025ec4d60f343a9c9e1ddf695112c497516679
parentbeaafefcf166950aed700a63f5d40f8b87cc736a
Half pixel variance further optimized for ARMv6

Half pixel interpolations optimized in variance calculations. Separate
function calls to vp8_filter_block2d_bil_x_pass_armv6 are avoided.On
average, performance improvement is 6-7% for VGA@30fps sequences.

Change-Id: Idb5f118a9d51548e824719d2cfe5be0fa6996628
vp8/encoder/arm/armv6/vp8_variance16x16_armv6.asm
vp8/encoder/arm/armv6/vp8_variance_halfpixvar16x16_h_armv6.asm [new file with mode: 0644]
vp8/encoder/arm/armv6/vp8_variance_halfpixvar16x16_hv_armv6.asm [new file with mode: 0644]
vp8/encoder/arm/armv6/vp8_variance_halfpixvar16x16_v_armv6.asm [new file with mode: 0644]
vp8/encoder/arm/variance_arm.c
vp8/vp8cx_arm.mk