• Jian Zhou's avatar
    Speed up h_predictor_16x16 · e86c7c86
    Jian Zhou authored
    Relocate the function from SSSE3 to SSE2, Unroll loop from 8 to 4,
    and reduce mem access to left.
    Speed up by >20% in ./test_intra_pred_speed.
    
    Change-Id: Ie48229c2e32404706b722442942c84983bda74cc
    e86c7c86
Name
Last commit
Last update
..
arm Loading commit data...
mips Loading commit data...
x86 Loading commit data...
bitreader.c Loading commit data...
bitreader.h Loading commit data...
bitreader_buffer.c Loading commit data...
bitreader_buffer.h Loading commit data...
bitwriter.c Loading commit data...
bitwriter.h Loading commit data...
bitwriter_buffer.c Loading commit data...
bitwriter_buffer.h Loading commit data...
fastssim.c Loading commit data...
fwd_txfm.c Loading commit data...
fwd_txfm.h Loading commit data...
intrapred.c Loading commit data...
inv_txfm.c Loading commit data...
inv_txfm.h Loading commit data...
loopfilter.c Loading commit data...
prob.c Loading commit data...
prob.h Loading commit data...
psnrhvs.c Loading commit data...
quantize.c Loading commit data...
quantize.h Loading commit data...
sad.c Loading commit data...
ssim.c Loading commit data...
ssim.h Loading commit data...
subtract.c Loading commit data...
txfm_common.h Loading commit data...
variance.c Loading commit data...
variance.h Loading commit data...
vpx_convolve.c Loading commit data...
vpx_convolve.h Loading commit data...
vpx_dsp.mk Loading commit data...
vpx_dsp_common.h Loading commit data...
vpx_dsp_rtcd.c Loading commit data...
vpx_dsp_rtcd_defs.pl Loading commit data...
vpx_filter.h Loading commit data...