698 to 174 cycles on penrynn. Unrolling is a 6 cycles gain.
unrol 6 cycles
---
libavcodec/x86/sbrdsp.asm | 22 ++++++++++++++++++++++
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 039bf8c..11a6faf 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 51c4bd4..108a681 100644
--- a/libavcodec/x86/sbrdsp_init.c
+++ b/libavcodec/x86/sbrdsp_init.c
@@ -30,6 +30,7 @@ void ff_sbr_hf_g_filt_sse(float (*Y)[2], const float
(*X_high)[40][2],
void ff_sbr_hf_gen_sse(float (*X_high)[2], const float (*X_low)[2],
const float alpha0[2], const float alpha1[2],
float bw, int start, int end);
+void ff_sbr_sum64x5_sse(float *z);
void ff_sbrdsp_init_x86(SBRDSPContext *s)
{
@@ -39,5 +40,6 @@ void ff_sbrdsp_init_x86(SBRDSPContext *s)
s->sum_square = ff_sbr_sum_square_sse;
s->hf_g_filt = ff_sbr_hf_g_filt_sse;
s->hf_gen = ff_sbr_hf_gen_sse;
+ s->sum64x5 = ff_sbr_sum64x5_sse;
}
}
--
1.7.7.msysgit.0
_______________________________________________
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
