diff --git a/libavcodec/dsputil.c b/libavcodec/dsputil.c
index 241bad0d4f..65d884fcea 100644
--- a/libavcodec/dsputil.c
+++ b/libavcodec/dsputil.c
@@ -4471,6 +4471,11 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx)
         c->h263_v_loop_filter= h263_v_loop_filter_c;
     }
 
+    if (ENABLE_VP3_DECODER || ENABLE_THEORA_DECODER) {
+        c->vp3_h_loop_filter= ff_vp3_h_loop_filter_c;
+        c->vp3_v_loop_filter= ff_vp3_v_loop_filter_c;
+    }
+
     c->h261_loop_filter= h261_loop_filter_c;
 
     c->try_8x8basis= try_8x8basis_c;
diff --git a/libavcodec/dsputil.h b/libavcodec/dsputil.h
index b8504bc0ac..a016f2f911 100644
--- a/libavcodec/dsputil.h
+++ b/libavcodec/dsputil.h
@@ -86,6 +86,9 @@ void ff_vp3_idct_c(DCTELEM *block/* align 16*/);
 void ff_vp3_idct_put_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
 void ff_vp3_idct_add_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/);
 
+void ff_vp3_v_loop_filter_c(uint8_t *src, int stride, int *bounding_values);
+void ff_vp3_h_loop_filter_c(uint8_t *src, int stride, int *bounding_values);
+
 /* 1/2^n downscaling functions from imgconvert.c */
 void ff_img_copy_plane(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
 void ff_shrink22(uint8_t *dst, int dst_wrap, const uint8_t *src, int src_wrap, int width, int height);
@@ -359,6 +362,9 @@ typedef struct DSPContext {
     void (*x8_v_loop_filter)(uint8_t *src, int stride, int qscale);
     void (*x8_h_loop_filter)(uint8_t *src, int stride, int qscale);
 
+    void (*vp3_v_loop_filter)(uint8_t *src, int stride, int *bounding_values);
+    void (*vp3_h_loop_filter)(uint8_t *src, int stride, int *bounding_values);
+
     /* assume len is a multiple of 4, and arrays are 16-byte aligned */
     void (*vorbis_inverse_coupling)(float *mag, float *ang, int blocksize);
     void (*ac3_downmix)(float (*samples)[256], float (*matrix)[2], int out_ch, int in_ch, int len);
diff --git a/libavcodec/vp3.c b/libavcodec/vp3.c
index 558465fc8f..fec0fd77c8 100644
--- a/libavcodec/vp3.c
+++ b/libavcodec/vp3.c
@@ -1283,12 +1283,6 @@ static void reverse_dc_prediction(Vp3DecodeContext *s,
     }
 }
 
-
-static void horizontal_filter(unsigned char *first_pixel, int stride,
-    int *bounding_values);
-static void vertical_filter(unsigned char *first_pixel, int stride,
-    int *bounding_values);
-
 /*
  * Perform the final rendering for a particular slice of data.
  * The slice number ranges from 0..(macroblock_height - 1).
@@ -1495,39 +1489,6 @@ static void render_slice(Vp3DecodeContext *s, int slice)
     emms_c();
 }
 
-static void horizontal_filter(unsigned char *first_pixel, int stride,
-    int *bounding_values)
-{
-    unsigned char *end;
-    int filter_value;
-
-    for (end= first_pixel + 8*stride; first_pixel != end; first_pixel += stride) {
-        filter_value =
-            (first_pixel[-2] - first_pixel[ 1])
-         +3*(first_pixel[ 0] - first_pixel[-1]);
-        filter_value = bounding_values[(filter_value + 4) >> 3];
-        first_pixel[-1] = av_clip_uint8(first_pixel[-1] + filter_value);
-        first_pixel[ 0] = av_clip_uint8(first_pixel[ 0] - filter_value);
-    }
-}
-
-static void vertical_filter(unsigned char *first_pixel, int stride,
-    int *bounding_values)
-{
-    unsigned char *end;
-    int filter_value;
-    const int nstride= -stride;
-
-    for (end= first_pixel + 8; first_pixel < end; first_pixel++) {
-        filter_value =
-            (first_pixel[2 * nstride] - first_pixel[ stride])
-         +3*(first_pixel[0          ] - first_pixel[nstride]);
-        filter_value = bounding_values[(filter_value + 4) >> 3];
-        first_pixel[nstride] = av_clip_uint8(first_pixel[nstride] + filter_value);
-        first_pixel[0] = av_clip_uint8(first_pixel[0] - filter_value);
-    }
-}
-
 static void apply_loop_filter(Vp3DecodeContext *s)
 {
     int plane;
@@ -1569,7 +1530,7 @@ static void apply_loop_filter(Vp3DecodeContext *s)
                 /* do not perform left edge filter for left columns frags */
                 if ((x > 0) &&
                     (s->all_fragments[fragment].coding_method != MODE_COPY)) {
-                    horizontal_filter(
+                    s->dsp.vp3_h_loop_filter(
                         plane_data + s->all_fragments[fragment].first_pixel,
                         stride, bounding_values);
                 }
@@ -1577,7 +1538,7 @@ static void apply_loop_filter(Vp3DecodeContext *s)
                 /* do not perform top edge filter for top row fragments */
                 if ((y > 0) &&
                     (s->all_fragments[fragment].coding_method != MODE_COPY)) {
-                    vertical_filter(
+                    s->dsp.vp3_v_loop_filter(
                         plane_data + s->all_fragments[fragment].first_pixel,
                         stride, bounding_values);
                 }
@@ -1588,7 +1549,7 @@ static void apply_loop_filter(Vp3DecodeContext *s)
                 if ((x < width - 1) &&
                     (s->all_fragments[fragment].coding_method != MODE_COPY) &&
                     (s->all_fragments[fragment + 1].coding_method == MODE_COPY)) {
-                    horizontal_filter(
+                    s->dsp.vp3_h_loop_filter(
                         plane_data + s->all_fragments[fragment + 1].first_pixel,
                         stride, bounding_values);
                 }
@@ -1599,7 +1560,7 @@ static void apply_loop_filter(Vp3DecodeContext *s)
                 if ((y < height - 1) &&
                     (s->all_fragments[fragment].coding_method != MODE_COPY) &&
                     (s->all_fragments[fragment + width].coding_method == MODE_COPY)) {
-                    vertical_filter(
+                    s->dsp.vp3_v_loop_filter(
                         plane_data + s->all_fragments[fragment + width].first_pixel,
                         stride, bounding_values);
                 }
diff --git a/libavcodec/vp3dsp.c b/libavcodec/vp3dsp.c
index d374a85f6e..da5b108bdc 100644
--- a/libavcodec/vp3dsp.c
+++ b/libavcodec/vp3dsp.c
@@ -222,3 +222,34 @@ void ff_vp3_idct_put_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*
 void ff_vp3_idct_add_c(uint8_t *dest/*align 8*/, int line_size, DCTELEM *block/*align 16*/){
     idct(dest, line_size, block, 2);
 }
+
+void ff_vp3_v_loop_filter_c(uint8_t *first_pixel, int stride, int *bounding_values)
+{
+    unsigned char *end;
+    int filter_value;
+    const int nstride= -stride;
+
+    for (end= first_pixel + 8; first_pixel < end; first_pixel++) {
+        filter_value =
+            (first_pixel[2 * nstride] - first_pixel[ stride])
+         +3*(first_pixel[0          ] - first_pixel[nstride]);
+        filter_value = bounding_values[(filter_value + 4) >> 3];
+        first_pixel[nstride] = av_clip_uint8(first_pixel[nstride] + filter_value);
+        first_pixel[0] = av_clip_uint8(first_pixel[0] - filter_value);
+    }
+}
+
+void ff_vp3_h_loop_filter_c(uint8_t *first_pixel, int stride, int *bounding_values)
+{
+    unsigned char *end;
+    int filter_value;
+
+    for (end= first_pixel + 8*stride; first_pixel != end; first_pixel += stride) {
+        filter_value =
+            (first_pixel[-2] - first_pixel[ 1])
+         +3*(first_pixel[ 0] - first_pixel[-1]);
+        filter_value = bounding_values[(filter_value + 4) >> 3];
+        first_pixel[-1] = av_clip_uint8(first_pixel[-1] + filter_value);
+        first_pixel[ 0] = av_clip_uint8(first_pixel[ 0] - filter_value);
+    }
+}