[x264-devel] commit: Add AltiVec implementation of predict_8x8c_p.
gpoirier at mplayerhq.hu
Tue Jan 20 13:38:05 CET 2009
2009/1/19 Loren Merritt <lorenm at u.washington.edu>:
> On Mon, 19 Jan 2009, Guillaume POIRIER wrote:
>>+ vec_s16_t mul_b_induc0_v = vec_mladd(induc_v, b_v, zero_s16v);
>> vec_s16_t add_i0_b_0v = vec_adds(i00_v, mul_b_induc0_v);
> vec_s16_t add_i0_b_0v = vec_mladd(induc_v, b_v, i00_v);
Darn, this is so obvious! Thanks for suggesting this. This code is
faster across the board, both on G4(PPC7450) and G5(PPC970), and both
on 16x16 and 8x8 code.
Thanks a lot for the review and the suggestion!
Only a very small fraction of our DNA does anything; the rest is all
comments and ifdefs.
E. B. White - "Genius is more often found in a cracked pot than in a
More information about the x264-devel