FFmpeg
hwcontext_videotoolbox.c
Go to the documentation of this file.
1 /*
2  * This file is part of FFmpeg.
3  *
4  * FFmpeg is free software; you can redistribute it and/or
5  * modify it under the terms of the GNU Lesser General Public
6  * License as published by the Free Software Foundation; either
7  * version 2.1 of the License, or (at your option) any later version.
8  *
9  * FFmpeg is distributed in the hope that it will be useful,
10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12  * Lesser General Public License for more details.
13  *
14  * You should have received a copy of the GNU Lesser General Public
15  * License along with FFmpeg; if not, write to the Free Software
16  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17  */
18 
19 #include "config.h"
20 
21 #include <stdint.h>
22 #include <string.h>
23 
24 #include <VideoToolbox/VideoToolbox.h>
25 
26 #include "buffer.h"
27 #include "common.h"
28 #include "hwcontext.h"
29 #include "hwcontext_internal.h"
30 #include "hwcontext_videotoolbox.h"
31 #include "mem.h"
32 #include "pixfmt.h"
33 #include "pixdesc.h"
34 
35 static const struct {
36  uint32_t cv_fmt;
38 } cv_pix_fmts[] = {
39  { kCVPixelFormatType_420YpCbCr8Planar, AV_PIX_FMT_YUV420P },
40  { kCVPixelFormatType_422YpCbCr8, AV_PIX_FMT_UYVY422 },
41  { kCVPixelFormatType_32BGRA, AV_PIX_FMT_BGRA },
42 #ifdef kCFCoreFoundationVersionNumber10_7
43  { kCVPixelFormatType_420YpCbCr8BiPlanarVideoRange, AV_PIX_FMT_NV12 },
44 #endif
45 #if HAVE_KCVPIXELFORMATTYPE_420YPCBCR10BIPLANARVIDEORANGE
47 #endif
48 };
49 
51 {
52  int i;
53  for (i = 0; i < FF_ARRAY_ELEMS(cv_pix_fmts); i++) {
54  if (cv_pix_fmts[i].cv_fmt == cv_fmt)
55  return cv_pix_fmts[i].pix_fmt;
56  }
57  return AV_PIX_FMT_NONE;
58 }
59 
61 {
62  int i;
63  for (i = 0; i < FF_ARRAY_ELEMS(cv_pix_fmts); i++) {
64  if (cv_pix_fmts[i].pix_fmt == pix_fmt)
65  return cv_pix_fmts[i].cv_fmt;
66  }
67  return 0;
68 }
69 
71 {
72  frame->buf[0] = av_buffer_pool_get(ctx->pool);
73  if (!frame->buf[0])
74  return AVERROR(ENOMEM);
75 
76  frame->data[3] = frame->buf[0]->data;
78  frame->width = ctx->width;
79  frame->height = ctx->height;
80 
81  return 0;
82 }
83 
86  enum AVPixelFormat **formats)
87 {
88  enum AVPixelFormat *fmts = av_malloc_array(2, sizeof(*fmts));
89  if (!fmts)
90  return AVERROR(ENOMEM);
91 
92  fmts[0] = ctx->sw_format;
93  fmts[1] = AV_PIX_FMT_NONE;
94 
95  *formats = fmts;
96  return 0;
97 }
98 
100 {
101  CVPixelBufferRef pixbuf = (CVPixelBufferRef)hwmap->source->data[3];
102 
103  CVPixelBufferUnlockBaseAddress(pixbuf, (uintptr_t)hwmap->priv);
104 }
105 
107  int flags)
108 {
109  CVPixelBufferRef pixbuf = (CVPixelBufferRef)src->data[3];
110  OSType pixel_format = CVPixelBufferGetPixelFormatType(pixbuf);
111  CVReturn err;
112  uint32_t map_flags = 0;
113  int ret;
114  int i;
115  enum AVPixelFormat format;
116 
117  format = av_map_videotoolbox_format_to_pixfmt(pixel_format);
118  if (dst->format != format) {
119  av_log(ctx, AV_LOG_ERROR, "Unsupported or mismatching pixel format: %s\n",
120  av_fourcc2str(pixel_format));
121  return AVERROR_UNKNOWN;
122  }
123 
124  if (CVPixelBufferGetWidth(pixbuf) != ctx->width ||
125  CVPixelBufferGetHeight(pixbuf) != ctx->height) {
126  av_log(ctx, AV_LOG_ERROR, "Inconsistent frame dimensions.\n");
127  return AVERROR_UNKNOWN;
128  }
129 
130  if (flags == AV_HWFRAME_MAP_READ)
131  map_flags = kCVPixelBufferLock_ReadOnly;
132 
133  err = CVPixelBufferLockBaseAddress(pixbuf, map_flags);
134  if (err != kCVReturnSuccess) {
135  av_log(ctx, AV_LOG_ERROR, "Error locking the pixel buffer.\n");
136  return AVERROR_UNKNOWN;
137  }
138 
139  if (CVPixelBufferIsPlanar(pixbuf)) {
140  int planes = CVPixelBufferGetPlaneCount(pixbuf);
141  for (i = 0; i < planes; i++) {
142  dst->data[i] = CVPixelBufferGetBaseAddressOfPlane(pixbuf, i);
143  dst->linesize[i] = CVPixelBufferGetBytesPerRowOfPlane(pixbuf, i);
144  }
145  } else {
146  dst->data[0] = CVPixelBufferGetBaseAddress(pixbuf);
147  dst->linesize[0] = CVPixelBufferGetBytesPerRow(pixbuf);
148  }
149 
150  ret = ff_hwframe_map_create(src->hw_frames_ctx, dst, src, vt_unmap,
151  (void *)(uintptr_t)map_flags);
152  if (ret < 0)
153  goto unlock;
154 
155  return 0;
156 
157 unlock:
158  CVPixelBufferUnlockBaseAddress(pixbuf, map_flags);
159  return ret;
160 }
161 
163  AVFrame *dst, const AVFrame *src)
164 {
165  AVFrame *map;
166  int err;
167 
168  if (dst->width > hwfc->width || dst->height > hwfc->height)
169  return AVERROR(EINVAL);
170 
171  map = av_frame_alloc();
172  if (!map)
173  return AVERROR(ENOMEM);
174  map->format = dst->format;
175 
176  err = vt_map_frame(hwfc, map, src, AV_HWFRAME_MAP_READ);
177  if (err)
178  goto fail;
179 
180  map->width = dst->width;
181  map->height = dst->height;
182 
183  err = av_frame_copy(dst, map);
184  if (err)
185  goto fail;
186 
187  err = 0;
188 fail:
189  av_frame_free(&map);
190  return err;
191 }
192 
194  AVFrame *dst, const AVFrame *src)
195 {
196  AVFrame *map;
197  int err;
198 
199  if (src->width > hwfc->width || src->height > hwfc->height)
200  return AVERROR(EINVAL);
201 
202  map = av_frame_alloc();
203  if (!map)
204  return AVERROR(ENOMEM);
205  map->format = src->format;
206 
208  if (err)
209  goto fail;
210 
211  map->width = src->width;
212  map->height = src->height;
213 
214  err = av_frame_copy(map, src);
215  if (err)
216  goto fail;
217 
218  err = 0;
219 fail:
220  av_frame_free(&map);
221  return err;
222 }
223 
224 static int vt_device_create(AVHWDeviceContext *ctx, const char *device,
225  AVDictionary *opts, int flags)
226 {
227  if (device && device[0]) {
228  av_log(ctx, AV_LOG_ERROR, "Device selection unsupported.\n");
229  return AVERROR_UNKNOWN;
230  }
231 
232  return 0;
233 }
234 
237  .name = "videotoolbox",
238 
239  .device_create = vt_device_create,
240  .frames_get_buffer = vt_get_buffer,
241  .transfer_get_formats = vt_transfer_get_formats,
242  .transfer_data_to = vt_transfer_data_to,
243  .transfer_data_from = vt_transfer_data_from,
244 
245  .pix_fmts = (const enum AVPixelFormat[]){ AV_PIX_FMT_VIDEOTOOLBOX, AV_PIX_FMT_NONE },
246 };
This struct aggregates all the (hardware/vendor-specific) "high-level" state, i.e.
Definition: hwcontext.h:60
packed YUV 4:2:2, 16bpp, Cb Y0 Cr Y1
Definition: pixfmt.h:81
This structure describes decoded (raw) audio or video data.
Definition: frame.h:295
static const struct @305 cv_pix_fmts[]
Memory handling functions.
static int vt_map_frame(AVHWFramesContext *ctx, AVFrame *dst, const AVFrame *src, int flags)
AVBufferRef * buf[AV_NUM_DATA_POINTERS]
AVBuffer references backing the data for this frame.
Definition: frame.h:486
hardware decoding through Videotoolbox
Definition: pixfmt.h:282
static int vt_transfer_data_to(AVHWFramesContext *hwfc, AVFrame *dst, const AVFrame *src)
uint32_t av_map_videotoolbox_format_from_pixfmt(enum AVPixelFormat pix_fmt)
Convert an AVPixelFormat to a VideoToolbox (actually CoreVideo) format.
int width
The allocated dimensions of the frames in this pool.
Definition: hwcontext.h:228
static int vt_transfer_get_formats(AVHWFramesContext *ctx, enum AVHWFrameTransferDirection dir, enum AVPixelFormat **formats)
#define src
Definition: vp8dsp.c:254
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample format(the sample packing is implied by the sample format) and sample rate.The lists are not just lists
enum AVPixelFormat pix_fmt
AVBufferRef * hw_frames_ctx
For hwaccel-format frames, this should be a reference to the AVHWFramesContext describing the frame...
Definition: frame.h:634
#define AV_PIX_FMT_P010
Definition: pixfmt.h:436
enum AVHWDeviceType type
AVFrame * av_frame_alloc(void)
Allocate an AVFrame and set its fields to default values.
Definition: frame.c:189
An API-specific header for AV_HWDEVICE_TYPE_VIDEOTOOLBOX.
The mapping must be writeable.
Definition: hwcontext.h:507
static int vt_transfer_data_from(AVHWFramesContext *hwfc, AVFrame *dst, const AVFrame *src)
static int vt_get_buffer(AVHWFramesContext *ctx, AVFrame *frame)
static void vt_unmap(AVHWFramesContext *ctx, HWMapDescriptor *hwmap)
#define av_log(a,...)
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:259
int width
Definition: frame.h:353
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:176
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:202
packed BGRA 8:8:8:8, 32bpp, BGRABGRA...
Definition: pixfmt.h:95
#define av_fourcc2str(fourcc)
Definition: avutil.h:348
planar YUV 4:2:0, 12bpp, 1 plane for Y and 1 plane for the UV components, which are interleaved (firs...
Definition: pixfmt.h:89
#define fail()
Definition: checkasm.h:120
int av_frame_copy(AVFrame *dst, const AVFrame *src)
Copy the frame data from src to dst.
Definition: frame.c:792
AVDictionary * opts
Definition: movenc.c:50
AVFrame * source
A reference to the original source of the mapping.
AVFormatContext * ctx
Definition: movenc.c:48
these buffered frames must be flushed immediately if a new input produces new the filter must not call request_frame to get more It must just process the frame or queue it The task of requesting more frames is left to the filter s request_frame method or the application If a filter has several the filter must be ready for frames arriving randomly on any input any filter with several inputs will most likely require some kind of queuing mechanism It is perfectly acceptable to have a limited queue and to drop frames when the inputs are too unbalanced request_frame For filters that do not use the this method is called when a frame is wanted on an output For a it should directly call filter_frame on the corresponding output For a if there are queued frames already one of these frames should be pushed If the filter should request a frame on one of its repeatedly until at least one frame has been pushed Return or at least make progress towards producing a frame
#define FF_ARRAY_ELEMS(a)
static const struct @316 planes[]
int format
format of the frame, -1 if unknown or unset Values correspond to enum AVPixelFormat for video frames...
Definition: frame.h:368
static int vt_device_create(AVHWDeviceContext *ctx, const char *device, AVDictionary *opts, int flags)
void * priv
Hardware-specific private data associated with the mapping.
int linesize[AV_NUM_DATA_POINTERS]
For video, size in bytes of each picture line.
Definition: frame.h:326
uint8_t * data
The data buffer.
Definition: buffer.h:89
const HWContextType ff_hwcontext_type_videotoolbox
int ff_hwframe_map_create(AVBufferRef *hwframe_ref, AVFrame *dst, const AVFrame *src, void(*unmap)(AVHWFramesContext *ctx, HWMapDescriptor *hwmap), void *priv)
Definition: hwcontext.c:688
The mapping must be readable.
Definition: hwcontext.h:503
This struct describes a set or pool of "hardware" frames (i.e.
Definition: hwcontext.h:123
refcounted data buffer API
const VDPAUPixFmtMap * map
#define flags(name, subs,...)
Definition: cbs_av1.c:561
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:309
planar YUV 4:2:0, 12bpp, (1 Cr & Cb sample per 2x2 Y samples)
Definition: pixfmt.h:66
common internal and external API header
uint32_t cv_fmt
#define AVERROR_UNKNOWN
Unknown error, typically from an external library.
Definition: error.h:71
AVHWFrameTransferDirection
Definition: hwcontext.h:394
pixel format definitions
AVBufferPool * pool
A pool from which the frames are allocated by av_hwframe_get_buffer().
Definition: hwcontext.h:189
enum AVPixelFormat av_map_videotoolbox_format_to_pixfmt(uint32_t cv_fmt)
Convert a VideoToolbox (actually CoreVideo) format to AVPixelFormat.
int height
Definition: frame.h:353
AVBufferRef * av_buffer_pool_get(AVBufferPool *pool)
Allocate a new AVBuffer, reusing an old buffer from the pool when available.
Definition: buffer.c:334
#define av_malloc_array(a, b)
formats
Definition: signature.h:48
The mapped frame will be overwritten completely in subsequent operations, so the current frame data n...
Definition: hwcontext.h:513
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later.That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another.Frame references ownership and permissions
enum AVPixelFormat sw_format
The pixel format identifying the actual data layout of the hardware frames.
Definition: hwcontext.h:221
AVPixelFormat
Pixel format.
Definition: pixfmt.h:64