video/filter: change filter API, use refcounting, remove filter DR

Change the entire filter API to use reference counted images instead
of vf_get_image().

Remove filter "direct rendering". This was useful for vf_expand and (in
rare cases) vf_sub: DR allowed these filters to pass a cropped image to
the filters before them. Then, on filtering, the image was "uncropped",
so that black bars could be added around the image without copying. This
means that in some cases, vf_expand will be slower (-vf gradfun,expand
for example).

Note that another form of DR used for in-place filters has been replaced
by simpler logic. Instead of trying to do DR, filters can check if the
image is writeable (with mp_image_is_writeable()), and do true in-place
if that's the case. This affects filters like vf_gradfun and vf_sub.

Everything has to support strides now. If something doesn't, making a
copy of the image data is required.

1 files changed, 40 insertions, 94 deletions

diff --git a/video/filter/vf_pullup.c b/video/filter/vf_pullup.c
index acac407429..84bc64b1cf 100644
--- a/video/filter/vf_pullup.c
+++ b/video/filter/vf_pullup.c
@@ -75,55 +75,25 @@ static void init_pullup(struct vf_instance *vf, mp_image_t *mpi)
 	vf->priv->qbuf = malloc(c->w[3]);
 }
 
-
-#if 0
-static void get_image(struct vf_instance *vf, mp_image_t *mpi)
-{
-	struct pullup_context *c = vf->priv->ctx;
-	struct pullup_buffer *b;
-
-	if (mpi->type == MP_IMGTYPE_STATIC) return;
-
-	if (!vf->priv->init) init_pullup(vf, mpi);
-
-	b = pullup_get_buffer(c, 2);
-	if (!b) return; /* shouldn't happen... */
-
-	mpi->priv = b;
-
-	mpi->planes[0] = b->planes[0];
-	mpi->planes[1] = b->planes[1];
-	mpi->planes[2] = b->planes[2];
-	mpi->stride[0] = c->stride[0];
-	mpi->stride[1] = c->stride[1];
-	mpi->stride[2] = c->stride[2];
-
-	mpi->flags |= MP_IMGFLAG_DIRECT;
-}
-#endif
-
-static int put_image(struct vf_instance *vf, mp_image_t *mpi, double pts)
+static struct mp_image *filter(struct vf_instance *vf, struct mp_image *mpi)
 {
 	struct pullup_context *c = vf->priv->ctx;
 	struct pullup_buffer *b;
 	struct pullup_frame *f;
-	mp_image_t *dmpi;
-	int ret;
 	int p;
 	int i;
+        double pts = mpi->pts;
+        struct mp_image *dmpi = NULL;
 
 	if (!vf->priv->init) init_pullup(vf, mpi);
 
-	if (mpi->flags & MP_IMGFLAG_DIRECT) {
-		b = mpi->priv;
-		mpi->priv = 0;
-	} else {
+	if (1) {
 		b = pullup_get_buffer(c, 2);
 		if (!b) {
 			mp_msg(MSGT_VFILTER,MSGL_ERR,"Could not get buffer from pullup!\n");
 			f = pullup_get_frame(c);
 			pullup_release_frame(f);
-			return 0;
+			goto skip;
 		}
 		memcpy_pic(b->planes[0], mpi->planes[0], mpi->w, mpi->h,
 			c->stride[0], mpi->stride[0]);
@@ -174,21 +144,23 @@ static int put_image(struct vf_instance *vf, mp_image_t *mpi, double pts)
 
 	/* Fake yes for first few frames (buffer depth) to keep from
 	 * breaking A/V sync with G1's bad architecture... */
-	if (!f) return vf->priv->fakecount ? (--vf->priv->fakecount,1) : 0;
+	//if (!f) return vf->priv->fakecount ? (--vf->priv->fakecount,1) : 0;
+	if (!f)
+            goto skip;
 
 	if (f->length < 2) {
 		pullup_release_frame(f);
 		f = pullup_get_frame(c);
-		if (!f) return 0;
+		if (!f) goto skip;
 		if (f->length < 2) {
 			pullup_release_frame(f);
 			if (!(mpi->fields & MP_IMGFIELD_REPEAT_FIRST))
-				return 0;
+				goto skip;
 			f = pullup_get_frame(c);
-			if (!f) return 0;
+			if (!f) goto skip;
 			if (f->length < 2) {
 				pullup_release_frame(f);
-				return 0;
+				goto skip;
 			}
 		}
 	}
@@ -210,66 +182,42 @@ static int put_image(struct vf_instance *vf, mp_image_t *mpi, double pts)
 	}
 #endif
 
-	/* If the frame isn't already exportable... */
-	while (!f->buffer) {
-		dmpi = vf_get_image(vf->next, mpi->imgfmt,
-			MP_IMGTYPE_TEMP, MP_IMGFLAG_ACCEPT_STRIDE,
-			mpi->width, mpi->height);
-		/* FIXME: Is it ok to discard dmpi if it's not direct? */
-		if (!(dmpi->flags & MP_IMGFLAG_DIRECT)) {
-			pullup_pack_frame(c, f);
-			break;
-		}
-		/* Direct render fields into output buffer */
-		my_memcpy_pic(dmpi->planes[0], f->ofields[0]->planes[0],
-			mpi->w, mpi->h/2, dmpi->stride[0]*2, c->stride[0]*2);
-		my_memcpy_pic(dmpi->planes[0] + dmpi->stride[0],
-			f->ofields[1]->planes[0] + c->stride[0],
-			mpi->w, mpi->h/2, dmpi->stride[0]*2, c->stride[0]*2);
-		if (mpi->flags & MP_IMGFLAG_PLANAR) {
-			my_memcpy_pic(dmpi->planes[1], f->ofields[0]->planes[1],
-				mpi->chroma_width, mpi->chroma_height/2,
-				dmpi->stride[1]*2, c->stride[1]*2);
-			my_memcpy_pic(dmpi->planes[1] + dmpi->stride[1],
-				f->ofields[1]->planes[1] + c->stride[1],
-				mpi->chroma_width, mpi->chroma_height/2,
-				dmpi->stride[1]*2, c->stride[1]*2);
-			my_memcpy_pic(dmpi->planes[2], f->ofields[0]->planes[2],
-				mpi->chroma_width, mpi->chroma_height/2,
-				dmpi->stride[2]*2, c->stride[2]*2);
-			my_memcpy_pic(dmpi->planes[2] + dmpi->stride[2],
-				f->ofields[1]->planes[2] + c->stride[2],
-				mpi->chroma_width, mpi->chroma_height/2,
-				dmpi->stride[2]*2, c->stride[2]*2);
-		}
-		pullup_release_frame(f);
-		if (mpi->qscale) {
-			dmpi->qscale = vf->priv->qbuf;
-			dmpi->qstride = mpi->qstride;
-			dmpi->qscale_type = mpi->qscale_type;
-		}
-		return vf_next_put_image(vf, dmpi, f->pts);
-	}
-	dmpi = vf_get_image(vf->next, mpi->imgfmt,
-		MP_IMGTYPE_EXPORT, MP_IMGFLAG_ACCEPT_STRIDE,
-		mpi->width, mpi->height);
+        /* If the frame isn't already exportable... */
+        if (!f->buffer)
+            pullup_pack_frame(c, f);
+
+        // NOTE: the copy could probably be avoided by changing or using the
+        //       pullup internal buffer management. But right now just do the
+        //       safe thing and always copy. Code outside the filter might
+        //       hold a buffer reference even if the filter chain is destroyed.
+        dmpi = vf_alloc_out_image(vf);
+        mp_image_copy_attributes(dmpi, mpi);
+
+        struct mp_image data = *dmpi;
 
-	dmpi->planes[0] = f->buffer->planes[0];
-	dmpi->planes[1] = f->buffer->planes[1];
-	dmpi->planes[2] = f->buffer->planes[2];
+	data.planes[0] = f->buffer->planes[0];
+	data.planes[1] = f->buffer->planes[1];
+	data.planes[2] = f->buffer->planes[2];
 
-	dmpi->stride[0] = c->stride[0];
-	dmpi->stride[1] = c->stride[1];
-	dmpi->stride[2] = c->stride[2];
+	data.stride[0] = c->stride[0];
+	data.stride[1] = c->stride[1];
+	data.stride[2] = c->stride[2];
 
+        mp_image_copy(dmpi, &data);
+
+        dmpi->pts = f->pts;
+
+        // Warning: entirely bogus memory management of qscale
 	if (mpi->qscale) {
 		dmpi->qscale = vf->priv->qbuf;
 		dmpi->qstride = mpi->qstride;
 		dmpi->qscale_type = mpi->qscale_type;
 	}
-	ret = vf_next_put_image(vf, dmpi, f->pts);
 	pullup_release_frame(f);
-	return ret;
+
+skip:
+        talloc_free(mpi);
+	return dmpi;
 }
 
 static int query_format(struct vf_instance *vf, unsigned int fmt)
@@ -302,12 +250,10 @@ static int vf_open(vf_instance_t *vf, char *args)
 {
 	struct vf_priv_s *p;
 	struct pullup_context *c;
-	//vf->get_image = get_image;
-	vf->put_image = put_image;
+	vf->filter = filter;
 	vf->config = config;
 	vf->query_format = query_format;
 	vf->uninit = uninit;
-	vf->default_reqs = VFCAP_ACCEPT_STRIDE;
 	vf->priv = p = calloc(1, sizeof(struct vf_priv_s));
 	p->ctx = c = pullup_alloc_context();
 	p->fakecount = 1;