Janne Grunau <[email protected]> writes: > Overall almost 4% faster, idct_add down from 350 to 85 cycles, idct_dc_add > down from 83 to 30 cycles. > > squash: rv34 idct rearrange partial register loads > --- > libavcodec/arm/rv34dsp_init_neon.c | 6 ++++ > libavcodec/arm/rv34dsp_neon.S | 59 > ++++++++++++++++++++++++++++++++++-- > 2 files changed, 62 insertions(+), 3 deletions(-)
OK, I see nothing more to tweak. -- Måns Rullgård [email protected] _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
