diff options
| author | Christophe Gisquet <christophe.gisquet@gmail.com> | 2012-11-24 23:15:07 +0100 |
|---|---|---|
| committer | Diego Biurrun <diego@biurrun.de> | 2013-01-06 13:57:01 +0100 |
| commit | 44a0036d10579ed91e48df24859e54b08a582742 (patch) | |
| tree | 9efe5bd75b5db9a0fc2b921d2fd593129102d4ec /libavcodec | |
| parent | 5b4dfbffc258f90a7d2540d21209ac23afcf7cd0 (diff) | |
| download | ffmpeg-44a0036d10579ed91e48df24859e54b08a582742.tar.gz | |
x86: sbrdsp: Implement SSE sum64x5
698 to 174 cycles on Arrandale. Unrolling is a 6 cycles gain.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Diffstat (limited to 'libavcodec')
| -rw-r--r-- | libavcodec/x86/sbrdsp.asm | 22 | ||||
| -rw-r--r-- | libavcodec/x86/sbrdsp_init.c | 2 |
2 files changed, 24 insertions, 0 deletions
diff --git a/libavcodec/x86/sbrdsp.asm b/libavcodec/x86/sbrdsp.asm index b87da4a072..063c24f95c 100644 --- a/libavcodec/x86/sbrdsp.asm +++ b/libavcodec/x86/sbrdsp.asm @@ -181,3 +181,25 @@ cglobal sbr_hf_gen, 4,4,8, X_high, X_low, alpha0, alpha1, BW, S, E add start, 16 jnz .loop2 RET + +cglobal sbr_sum64x5, 1,2,4,z + lea r1q, [zq+ 256] +.loop: + mova m0, [zq+ 0] + mova m2, [zq+ 16] + mova m1, [zq+ 256] + mova m3, [zq+ 272] + addps m0, [zq+ 512] + addps m2, [zq+ 528] + addps m1, [zq+ 768] + addps m3, [zq+ 784] + addps m0, [zq+1024] + addps m2, [zq+1040] + addps m0, m1 + addps m2, m3 + mova [zq], m0 + mova [zq+16], m2 + add zq, 32 + cmp zq, r1q + jne .loop + REP_RET diff --git a/libavcodec/x86/sbrdsp_init.c b/libavcodec/x86/sbrdsp_init.c index 51c4bd4a16..264b455b3e 100644 --- a/libavcodec/x86/sbrdsp_init.c +++ b/libavcodec/x86/sbrdsp_init.c @@ -25,6 +25,7 @@ #include "libavcodec/sbrdsp.h" float ff_sbr_sum_square_sse(float (*x)[2], int n); +void ff_sbr_sum64x5_sse(float *z); void ff_sbr_hf_g_filt_sse(float (*Y)[2], const float (*X_high)[40][2], const float *g_filt, int m_max, intptr_t ixh); void ff_sbr_hf_gen_sse(float (*X_high)[2], const float (*X_low)[2], @@ -37,6 +38,7 @@ void ff_sbrdsp_init_x86(SBRDSPContext *s) if (EXTERNAL_SSE(mm_flags)) { s->sum_square = ff_sbr_sum_square_sse; + s->sum64x5 = ff_sbr_sum64x5_sse; s->hf_g_filt = ff_sbr_hf_g_filt_sse; s->hf_gen = ff_sbr_hf_gen_sse; } |