vp3: move idct and loop filter pointers to new vp3dsp context

This moves all VP3-specific function pointers from dsputil to a
new vp3dsp context.  There is no reason to ever use the VP3 IDCT
where an MPEG2 IDCT is expected or vice versa.

Signed-off-by: Mans Rullgard <mans@mansr.com>

4 files changed, 46 insertions, 47 deletions

diff --git a/libavcodec/arm/Makefile b/libavcodec/arm/Makefile
index 60cb2072ae..a8e531cf18 100644
--- a/libavcodec/arm/Makefile
+++ b/libavcodec/arm/Makefile
@@ -11,6 +11,7 @@ ARMV6-OBJS-$(CONFIG_AC3DSP)            += arm/ac3dsp_armv6.o
 OBJS-$(CONFIG_MPEGAUDIODSP)            += arm/mpegaudiodsp_init_arm.o
 ARMV6-OBJS-$(CONFIG_MPEGAUDIODSP)      += arm/mpegaudiodsp_fixed_armv6.o
 
+OBJS-$(CONFIG_VP3DSP)                  += arm/vp3dsp_init_arm.o
 OBJS-$(CONFIG_VP5_DECODER)             += arm/vp56dsp_init_arm.o
 OBJS-$(CONFIG_VP6_DECODER)             += arm/vp56dsp_init_arm.o
 OBJS-$(CONFIG_VP8_DECODER)             += arm/vp8dsp_init_arm.o
diff --git a/libavcodec/arm/dsputil_init_neon.c b/libavcodec/arm/dsputil_init_neon.c
index 65db20d2b3..a132f6f993 100644
--- a/libavcodec/arm/dsputil_init_neon.c
+++ b/libavcodec/arm/dsputil_init_neon.c
@@ -29,11 +29,6 @@ void ff_simple_idct_neon(DCTELEM *data);
 void ff_simple_idct_put_neon(uint8_t *dest, int line_size, DCTELEM *data);
 void ff_simple_idct_add_neon(uint8_t *dest, int line_size, DCTELEM *data);
 
-void ff_vp3_idct_neon(DCTELEM *data);
-void ff_vp3_idct_put_neon(uint8_t *dest, int line_size, DCTELEM *data);
-void ff_vp3_idct_add_neon(uint8_t *dest, int line_size, DCTELEM *data);
-void ff_vp3_idct_dc_add_neon(uint8_t *dest, int line_size, const DCTELEM *data);
-
 void ff_clear_block_neon(DCTELEM *block);
 void ff_clear_blocks_neon(DCTELEM *blocks);
 
@@ -147,9 +142,6 @@ void ff_avg_h264_chroma_mc8_neon(uint8_t *, uint8_t *, int, int, int, int);
 void ff_avg_h264_chroma_mc4_neon(uint8_t *, uint8_t *, int, int, int, int);
 void ff_avg_h264_chroma_mc2_neon(uint8_t *, uint8_t *, int, int, int, int);
 
-void ff_vp3_v_loop_filter_neon(uint8_t *, int, int *);
-void ff_vp3_h_loop_filter_neon(uint8_t *, int, int *);
-
 void ff_vector_fmul_window_neon(float *dst, const float *src0,
                                 const float *src1, const float *win, int len);
 void ff_vector_fmul_scalar_neon(float *dst, const float *src, float mul,
@@ -186,13 +178,6 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
             c->idct_add              = ff_simple_idct_add_neon;
             c->idct                  = ff_simple_idct_neon;
             c->idct_permutation_type = FF_PARTTRANS_IDCT_PERM;
-        } else if ((CONFIG_VP3_DECODER || CONFIG_VP5_DECODER ||
-                    CONFIG_VP6_DECODER) &&
-                   avctx->idct_algo == FF_IDCT_VP3) {
-            c->idct_put              = ff_vp3_idct_put_neon;
-            c->idct_add              = ff_vp3_idct_add_neon;
-            c->idct                  = ff_vp3_idct_neon;
-            c->idct_permutation_type = FF_TRANSPOSE_IDCT_PERM;
         }
     }
 
@@ -319,12 +304,6 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
         c->avg_h264_qpel_pixels_tab[1][15] = ff_avg_h264_qpel8_mc33_neon;
     }
 
-    if (CONFIG_VP3_DECODER) {
-        c->vp3_v_loop_filter = ff_vp3_v_loop_filter_neon;
-        c->vp3_h_loop_filter = ff_vp3_h_loop_filter_neon;
-        c->vp3_idct_dc_add   = ff_vp3_idct_dc_add_neon;
-    }
-
     c->vector_fmul_window         = ff_vector_fmul_window_neon;
     c->vector_fmul_scalar         = ff_vector_fmul_scalar_neon;
     c->butterflies_float          = ff_butterflies_float_neon;
diff --git a/libavcodec/arm/vp3dsp_init_arm.c b/libavcodec/arm/vp3dsp_init_arm.c
new file mode 100644
index 0000000000..ea99bfd2b3
--- /dev/null
+++ b/libavcodec/arm/vp3dsp_init_arm.c
@@ -0,0 +1,45 @@
+/*
+ * This file is part of Libav.
+ *
+ * Libav is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * Libav is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with Libav; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include <stdint.h>
+
+#include "libavutil/attributes.h"
+#include "libavutil/cpu.h"
+#include "libavutil/arm/cpu.h"
+#include "libavcodec/vp3dsp.h"
+
+void ff_vp3_idct_put_neon(uint8_t *dest, int line_size, DCTELEM *data);
+void ff_vp3_idct_add_neon(uint8_t *dest, int line_size, DCTELEM *data);
+void ff_vp3_idct_dc_add_neon(uint8_t *dest, int line_size, const DCTELEM *data);
+
+void ff_vp3_v_loop_filter_neon(uint8_t *, int, int *);
+void ff_vp3_h_loop_filter_neon(uint8_t *, int, int *);
+
+av_cold void ff_vp3dsp_init_arm(VP3DSPContext *c, int flags)
+{
+    int cpu_flags = av_get_cpu_flags();
+
+    if (have_neon(cpu_flags)) {
+        c->idct_put      = ff_vp3_idct_put_neon;
+        c->idct_add      = ff_vp3_idct_add_neon;
+        c->idct_dc_add   = ff_vp3_idct_dc_add_neon;
+        c->v_loop_filter = ff_vp3_v_loop_filter_neon;
+        c->h_loop_filter = ff_vp3_h_loop_filter_neon;
+        c->idct_perm     = FF_TRANSPOSE_IDCT_PERM;
+    }
+}
diff --git a/libavcodec/arm/vp3dsp_neon.S b/libavcodec/arm/vp3dsp_neon.S
index 2a9b25f34e..e09de57281 100644
--- a/libavcodec/arm/vp3dsp_neon.S
+++ b/libavcodec/arm/vp3dsp_neon.S
@@ -260,32 +260,6 @@ endfunc
 VP3_IDCT_END row
 VP3_IDCT_END col
 
-function ff_vp3_idct_neon, export=1
-    mov             ip,  lr
-    mov             r2,  r0
-    bl              vp3_idct_start_neon
-    bl              vp3_idct_end_row_neon
-    mov             r3,  #8
-    bl              vp3_idct_core_neon
-    bl              vp3_idct_end_col_neon
-    mov             lr,  ip
-    vpop            {d8-d15}
-
-    vshr.s16        q8,  q8,  #4
-    vshr.s16        q9,  q9,  #4
-    vshr.s16        q10, q10, #4
-    vshr.s16        q11, q11, #4
-    vshr.s16        q12, q12, #4
-    vst1.64         {d16-d19}, [r0,:128]!
-    vshr.s16        q13, q13, #4
-    vshr.s16        q14, q14, #4
-    vst1.64         {d20-d23}, [r0,:128]!
-    vshr.s16        q15, q15, #4
-    vst1.64         {d24-d27}, [r0,:128]!
-    vst1.64         {d28-d31}, [r0,:128]!
-    bx              lr
-endfunc
-
 function ff_vp3_idct_put_neon, export=1
     mov             ip,  lr
     bl              vp3_idct_start_neon