[FFmpeg-cvslog] avcodec/arm/int_neon: fix handling sizes % 16 != 0

Michael Niedermayer git at videolan.org
Thu Feb 13 02:53:24 CET 2014


ffmpeg | branch: master | Michael Niedermayer <michaelni at gmx.at> | Thu Feb 13 02:20:08 2014 +0100| [ccc48b318b56e91ef5756a30d2d1c145c58aca71] | committer: Michael Niedermayer

avcodec/arm/int_neon: fix handling sizes % 16 != 0

This assumes the array is sufficiently padded with 0

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

> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=ccc48b318b56e91ef5756a30d2d1c145c58aca71
---

 libavcodec/arm/int_neon.S |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/libavcodec/arm/int_neon.S b/libavcodec/arm/int_neon.S
index 7751a55..ba84065 100644
--- a/libavcodec/arm/int_neon.S
+++ b/libavcodec/arm/int_neon.S
@@ -35,7 +35,7 @@ function ff_scalarproduct_int16_neon, export=1
         vmlal.s16       q2,  d18,  d22
         vmlal.s16       q3,  d19,  d23
         subs            r2,  r2,   #16
-        bne             1b
+        bgt             1b
 
         vpadd.s32       d16, d0,   d1
         vpadd.s32       d17, d2,   d3
@@ -75,7 +75,7 @@ function ff_scalarproduct_and_madd_int16_neon, export=1
         vst1.16         {q10},     [r12,:128]!
         subs            r3,  r3,   #16
         vst1.16         {q13},     [r12,:128]!
-        bne             1b
+        bgt             1b
 
         vpadd.s32       d16, d0,   d1
         vpadd.s32       d17, d2,   d3



More information about the ffmpeg-cvslog mailing list