Rewrite emu_edge functions to have separate src/dst_stride arguments.

This allows supporting files for which the image stride is smaller than
the max. block size + number of subpel mc taps, e.g. a 64x64 VP9 file
or a 16x16 VP8 file with -fflags +emu_edge.

1 files changed, 26 insertions, 21 deletions

diff --git a/libavcodec/vc1dec.c b/libavcodec/vc1dec.c
index 439d68eee9..399e0d2264 100644
--- a/libavcodec/vc1dec.c
+++ b/libavcodec/vc1dec.c
@@ -454,15 +454,18 @@ static void vc1_mc_1mv(VC1Context *v, int dir)
         uint8_t *uvbuf = s->edge_emu_buffer + 19 * s->linesize;
 
         srcY -= s->mspel * (1 + s->linesize);
-        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, srcY, s->linesize,
+        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, s->linesize,
+                                 srcY, s->linesize,
                                  17 + s->mspel * 2, 17 + s->mspel * 2,
                                  src_x - s->mspel, src_y - s->mspel,
                                  s->h_edge_pos, v_edge_pos);
         srcY = s->edge_emu_buffer;
-        s->vdsp.emulated_edge_mc(uvbuf     , srcU, s->uvlinesize, 8 + 1, 8 + 1,
-                                 uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, v_edge_pos >> 1);
-        s->vdsp.emulated_edge_mc(uvbuf + 16, srcV, s->uvlinesize, 8 + 1, 8 + 1,
-                                 uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, v_edge_pos >> 1);
+        s->vdsp.emulated_edge_mc(uvbuf, s->uvlinesize, srcU, s->uvlinesize,
+                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+                                 s->h_edge_pos >> 1, v_edge_pos >> 1);
+        s->vdsp.emulated_edge_mc(uvbuf + 16, s->uvlinesize, srcV, s->uvlinesize,
+                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+                                 s->h_edge_pos >> 1, v_edge_pos >> 1);
         srcU = uvbuf;
         srcV = uvbuf + 16;
         /* if we deal with range reduction we need to scale source blocks */
@@ -697,7 +700,7 @@ static void vc1_mc_4mv_luma(VC1Context *v, int n, int dir, int avg)
         || (unsigned)(src_y - (s->mspel << fieldmv)) > v_edge_pos - (my & 3) - ((8 + s->mspel * 2) << fieldmv)) {
         srcY -= s->mspel * (1 + (s->linesize << fieldmv));
         /* check emulate edge stride and offset */
-        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, srcY, s->linesize,
+        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, s->linesize, srcY, s->linesize,
                                  9 + s->mspel * 2, (9 + s->mspel * 2) << fieldmv,
                                  src_x - s->mspel, src_y - (s->mspel << fieldmv),
                                  s->h_edge_pos, v_edge_pos);
@@ -912,11 +915,11 @@ static void vc1_mc_4mv_chroma(VC1Context *v, int dir)
         || s->h_edge_pos < 18 || v_edge_pos < 18
         || (unsigned)uvsrc_x > (s->h_edge_pos >> 1) - 9
         || (unsigned)uvsrc_y > (v_edge_pos    >> 1) - 9) {
-        s->vdsp.emulated_edge_mc(s->edge_emu_buffer     , srcU, s->uvlinesize,
-                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, s->uvlinesize, srcU,
+                                 s->uvlinesize, 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
                                  s->h_edge_pos >> 1, v_edge_pos >> 1);
-        s->vdsp.emulated_edge_mc(s->edge_emu_buffer + 16, srcV, s->uvlinesize,
-                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+        s->vdsp.emulated_edge_mc(s->edge_emu_buffer + 16, s->uvlinesize, srcV,
+                                 s->uvlinesize, 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
                                  s->h_edge_pos >> 1, v_edge_pos >> 1);
         srcU = s->edge_emu_buffer;
         srcV = s->edge_emu_buffer + 16;
@@ -1033,12 +1036,12 @@ static void vc1_mc_4mv_chroma4(VC1Context *v, int dir, int dir2, int avg)
             || s->h_edge_pos < 10 || v_edge_pos < (5 << fieldmv)
             || (unsigned)uvsrc_x > (s->h_edge_pos >> 1) - 5
             || (unsigned)uvsrc_y > v_edge_pos - (5 << fieldmv)) {
-            s->vdsp.emulated_edge_mc(s->edge_emu_buffer, srcU, s->uvlinesize,
-                                     5, (5 << fieldmv), uvsrc_x, uvsrc_y,
-                                     s->h_edge_pos >> 1, v_edge_pos);
-            s->vdsp.emulated_edge_mc(s->edge_emu_buffer + 16, srcV, s->uvlinesize,
-                                     5, (5 << fieldmv), uvsrc_x, uvsrc_y,
-                                     s->h_edge_pos >> 1, v_edge_pos);
+            s->vdsp.emulated_edge_mc(s->edge_emu_buffer, s->uvlinesize, srcU,
+                                     s->uvlinesize, 5, (5 << fieldmv), uvsrc_x,
+                                     uvsrc_y, s->h_edge_pos >> 1, v_edge_pos);
+            s->vdsp.emulated_edge_mc(s->edge_emu_buffer + 16, s->uvlinesize, srcV,
+                                     s->uvlinesize, 5, (5 << fieldmv), uvsrc_x,
+                                     uvsrc_y, s->h_edge_pos >> 1, v_edge_pos);
             srcU = s->edge_emu_buffer;
             srcV = s->edge_emu_buffer + 16;
 
@@ -1966,15 +1969,17 @@ static void vc1_interp_mc(VC1Context *v)
         uint8_t *uvbuf = s->edge_emu_buffer + 19 * s->linesize;
 
         srcY -= s->mspel * (1 + s->linesize);
-        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, srcY, s->linesize,
+        s->vdsp.emulated_edge_mc(s->edge_emu_buffer, s->linesize, srcY, s->linesize,
                                  17 + s->mspel * 2, 17 + s->mspel * 2,
                                  src_x - s->mspel, src_y - s->mspel,
                                  s->h_edge_pos, v_edge_pos);
         srcY = s->edge_emu_buffer;
-        s->vdsp.emulated_edge_mc(uvbuf     , srcU, s->uvlinesize, 8 + 1, 8 + 1,
-                                 uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, v_edge_pos >> 1);
-        s->vdsp.emulated_edge_mc(uvbuf + 16, srcV, s->uvlinesize, 8 + 1, 8 + 1,
-                                 uvsrc_x, uvsrc_y, s->h_edge_pos >> 1, v_edge_pos >> 1);
+        s->vdsp.emulated_edge_mc(uvbuf, s->uvlinesize, srcU, s->uvlinesize,
+                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+                                 s->h_edge_pos >> 1, v_edge_pos >> 1);
+        s->vdsp.emulated_edge_mc(uvbuf + 16, s->uvlinesize, srcV, s->uvlinesize,
+                                 8 + 1, 8 + 1, uvsrc_x, uvsrc_y,
+                                 s->h_edge_pos >> 1, v_edge_pos >> 1);
         srcU = uvbuf;
         srcV = uvbuf + 16;
         /* if we deal with range reduction we need to scale source blocks */