[FFmpeg-cvslog] ra144enc: fix use of scalarprod_int16

[Christophe Gisquet]

ffmpeg | branch: master | Christophe Gisquet <christophe.gisquet at gmail.com> | Wed Mar  5 13:00:23 2014 +0100| [93c4cd618cd2222a2f1594f19a063e2d8d955483] | committer: Michael Niedermayer

ra144enc: fix use of scalarprod_int16

c3390fd56cf55259ea7665ecea6c8aeddf56e2fc made use of the DSP function
but did not complement it with a call to emms, which is done here before
computations involving floats are performed.

Fixes ticket #3429, which affected MMX/MMXExt machines.

Signed-off-by: Michael Niedermayer <michaelni at gmx.at>

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=93c4cd618cd2222a2f1594f19a063e2d8d955483
---

 libavcodec/ra144enc.c |    1 +
 1 file changed, 1 insertion(+)

diff --git a/libavcodec/ra144enc.c b/libavcodec/ra144enc.c
index d09e281..cafedda 100644
--- a/libavcodec/ra144enc.c
+++ b/libavcodec/ra144enc.c
@@ -376,6 +376,7 @@ static void ra144_encode_subblock(RA144Context *ractx,
 
         ff_copy_and_dup(ractx->buffer_a, ractx->adapt_cb, cba_idx + BLOCKSIZE / 2 - 1);
         m[0] = (ff_irms(&ractx->dsp, ractx->buffer_a) * rms) >> 12;
+        emms_c();
     }
     fixed_cb_search(work + LPC_ORDER, coefs, data, cba_idx, &cb1_idx, &cb2_idx);
     for (i = 0; i < BLOCKSIZE; i++) {