fmtconvert: Add a new method, int32_to_float_fmul_array8

This is similar to int32_to_float_fmul_scalar, but
loads a new scalar multiplier every 8 input samples.
This enables the use of much larger input arrays, which
is important for pipelining on some CPUs (such as
ARMv6).

Signed-off-by: Martin Storsjö <martin@martin.st>

1 files changed, 10 insertions, 0 deletions

diff --git a/libavcodec/fmtconvert.c b/libavcodec/fmtconvert.c
index 54f70305df..63f62c396a 100644
--- a/libavcodec/fmtconvert.c
+++ b/libavcodec/fmtconvert.c
@@ -32,6 +32,15 @@ static void int32_to_float_fmul_scalar_c(float *dst, const int32_t *src,
         dst[i] = src[i] * mul;
 }
 
+static void int32_to_float_fmul_array8_c(FmtConvertContext *c, float *dst,
+                                         const int32_t *src, const float *mul,
+                                         int len)
+{
+    int i;
+    for (i = 0; i < len; i += 8)
+        c->int32_to_float_fmul_scalar(&dst[i], &src[i], *mul++, 8);
+}
+
 static av_always_inline int float_to_int16_one(const float *src){
     return av_clip_int16(lrintf(*src));
 }
@@ -81,6 +90,7 @@ void ff_float_interleave_c(float *dst, const float **src, unsigned int len,
 av_cold void ff_fmt_convert_init(FmtConvertContext *c, AVCodecContext *avctx)
 {
     c->int32_to_float_fmul_scalar = int32_to_float_fmul_scalar_c;
+    c->int32_to_float_fmul_array8 = int32_to_float_fmul_array8_c;
     c->float_to_int16             = float_to_int16_c;
     c->float_to_int16_interleave  = float_to_int16_interleave_c;
     c->float_interleave           = ff_float_interleave_c;