Christophe GISQUET
2784d18791
SBR DSP x86: implement SSE sbr_hf_g_filt
...
Unrolling the main loop to process, instead of 4 elements:
- 8: minor gain of 2 cycles (not worth the extra object size)
- 2: loss of 8 cycles.
Assigning STEP to a register is a loss. Output address (Y) is almost always
unaligned.
Timings:
- C (32/64 bits): 117/109 cycles
- SSE: 57 cycles
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
2012-02-23 15:50:09 -08:00
..
2012-02-09 21:04:44 -05:00
2011-12-12 15:36:20 +02:00
2012-01-27 10:19:57 +08:00
2012-02-15 22:06:34 +02:00
2012-02-15 22:06:34 +02:00
2012-02-01 19:02:32 -05:00
2012-02-15 22:06:34 +02:00
2012-01-27 10:19:57 +08:00
2012-02-09 12:26:45 +01:00
2012-02-09 12:26:45 +01:00
2012-01-27 10:19:57 +08:00
2012-02-09 12:26:45 +01:00
2012-02-09 12:26:45 +01:00
2012-01-30 23:58:23 +01:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-02-19 15:31:03 -08:00
2011-12-27 09:05:14 +02:00
2012-02-07 11:37:13 -08:00
2012-02-08 10:31:14 -08:00
2012-01-27 10:19:57 +08:00
2012-02-08 10:31:14 -08:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-02-09 12:26:45 +01:00
2012-02-09 12:26:45 +01:00
2012-01-29 08:14:05 -08:00
2012-02-23 15:50:06 -08:00
2012-02-15 22:06:34 +02:00
2012-02-15 20:11:54 -08:00
2012-02-15 22:06:42 +02:00
2012-02-15 22:07:23 +02:00
2012-01-29 18:52:17 -08:00
2012-02-07 14:32:26 -08:00
2012-01-27 10:19:57 +08:00
2012-02-20 14:58:25 -08:00
2012-01-16 19:26:41 +01:00
2012-02-20 14:58:25 -08:00
2012-01-30 23:58:25 +01:00
2012-02-23 15:50:09 -08:00
2012-02-23 15:50:09 -08:00
2011-12-11 00:32:25 +01:00
2012-01-27 10:19:57 +08:00
2012-01-27 10:19:57 +08:00
2012-02-02 12:00:48 -08:00