// // ZoneMinder Ffmpeg Camera Class Implementation, $Date: 2009-01-16 12:18:50 +0000 (Fri, 16 Jan 2009) $, $Revision: 2713 $ // Copyright (C) 2001-2008 Philip Coombes // // This program is free software; you can redistribute it and/or // modify it under the terms of the GNU General Public License // as published by the Free Software Foundation; either version 2 // of the License, or (at your option) any later version. // // This program 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 General Public License for more details. // // You should have received a copy of the GNU General Public License // along with this program; if not, write to the Free Software // Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. // #include "zm.h" #include "zm_signal.h" #if HAVE_LIBAVFORMAT #include "zm_ffmpeg_camera.h" extern "C" { #include "libavutil/time.h" #if HAVE_AVUTIL_HWCONTEXT_H #include "libavutil/hwcontext.h" #include "libavutil/hwcontext_qsv.h" #endif } #ifndef AV_ERROR_MAX_STRING_SIZE #define AV_ERROR_MAX_STRING_SIZE 64 #endif #ifdef SOLARIS #include // for ESRCH #include #include #endif #if HAVE_AVUTIL_HWCONTEXT_H static AVPixelFormat get_format(AVCodecContext *avctx, const enum AVPixelFormat *pix_fmts) { while (*pix_fmts != AV_PIX_FMT_NONE) { if (*pix_fmts == AV_PIX_FMT_QSV) { DecodeContext *decode = (DecodeContext *)avctx->opaque; AVHWFramesContext *frames_ctx; AVQSVFramesContext *frames_hwctx; int ret; /* create a pool of surfaces to be used by the decoder */ avctx->hw_frames_ctx = av_hwframe_ctx_alloc(decode->hw_device_ref); if (!avctx->hw_frames_ctx) return AV_PIX_FMT_NONE; frames_ctx = (AVHWFramesContext*)avctx->hw_frames_ctx->data; frames_hwctx = (AVQSVFramesContext*)frames_ctx->hwctx; frames_ctx->format = AV_PIX_FMT_QSV; frames_ctx->sw_format = avctx->sw_pix_fmt; frames_ctx->width = FFALIGN(avctx->coded_width, 32); frames_ctx->height = FFALIGN(avctx->coded_height, 32); frames_ctx->initial_pool_size = 32; frames_hwctx->frame_type = MFX_MEMTYPE_VIDEO_MEMORY_DECODER_TARGET; ret = av_hwframe_ctx_init(avctx->hw_frames_ctx); if (ret < 0) return AV_PIX_FMT_NONE; return AV_PIX_FMT_QSV; } pix_fmts++; } Error( "The QSV pixel format not offered in get_format()"); return AV_PIX_FMT_NONE; } #endif FfmpegCamera::FfmpegCamera( int p_id, const std::string &p_path, const std::string &p_method, const std::string &p_options, int p_width, int p_height, int p_colours, int p_brightness, int p_contrast, int p_hue, int p_colour, bool p_capture, bool p_record_audio ) : Camera( p_id, FFMPEG_SRC, p_width, p_height, p_colours, ZM_SUBPIX_ORDER_DEFAULT_FOR_COLOUR(p_colours), p_brightness, p_contrast, p_hue, p_colour, p_capture, p_record_audio ), mPath( p_path ), mMethod( p_method ), mOptions( p_options ) { if ( capture ) { FFMPEGInit(); } hwaccel = false; #if HAVE_AVUTIL_HWCONTEXT_H decode = { NULL }; hwFrame = NULL; #endif mFormatContext = NULL; mVideoStreamId = -1; mAudioStreamId = -1; mVideoCodecContext = NULL; mAudioCodecContext = NULL; mVideoCodec = NULL; mAudioCodec = NULL; mRawFrame = NULL; mFrame = NULL; frameCount = 0; mCanCapture = false; } // end FFmpegCamera::FFmpegCamera FfmpegCamera::~FfmpegCamera() { Close(); avformat_network_deinit(); } int FfmpegCamera::PrimeCapture() { if ( mCanCapture ) { Info("Priming capture from %s, Closing", mPath.c_str()); Close(); } mVideoStreamId = -1; mAudioStreamId = -1; Info("Priming capture from %s", mPath.c_str()); return ! OpenFfmpeg(); } int FfmpegCamera::PreCapture() { return 0; } int FfmpegCamera::Capture(ZMPacket &zm_packet) { if ( ! mCanCapture ) { return -1; } int ret; // If the reopen thread has a value, but mCanCapture != 0, then we have just reopened the connection to the ffmpeg device, and we can clean up the thread. if ( (ret = av_read_frame(mFormatContext, &packet)) < 0 ) { if ( // Check if EOF. (ret == AVERROR_EOF || (mFormatContext->pb && mFormatContext->pb->eof_reached)) || // Check for Connection failure. (ret == -110) ) { Info("Unable to read packet from stream %d: error %d \"%s\".", packet.stream_index, ret, av_make_error_string(ret).c_str() ); } else { Error("Unable to read packet from stream %d: error %d \"%s\".", packet.stream_index, ret, av_make_error_string(ret).c_str() ); } return -1; } dumpPacket(&packet, "ffmpeg_camera in"); if ( 0 && ( packet.dts < 0 ) ) { zm_av_packet_unref(&packet); return 0; } bytes += packet.size; zm_packet.set_packet(&packet); zm_av_packet_unref(&packet); return 1; } // FfmpegCamera::Capture int FfmpegCamera::PostCapture() { // Nothing to do here return 0; } int FfmpegCamera::OpenFfmpeg() { int ret; // Open the input, not necessarily a file #if !LIBAVFORMAT_VERSION_CHECK(53, 2, 0, 4, 0) Debug(1, "Calling av_open_input_file"); if ( av_open_input_file(&mFormatContext, mPath.c_str(), NULL, 0, NULL) != 0 ) #else // Handle options AVDictionary *opts = NULL; ret = av_dict_parse_string(&opts, Options().c_str(), "=", ",", 0); if ( ret < 0 ) { Warning("Could not parse ffmpeg input options list '%s'", Options().c_str()); } else { Debug(2,"Could not parse ffmpeg input options list '%s'", Options().c_str()); } // Set transport method as specified by method field, rtpUni is default const std::string method = Method(); if ( method == "rtpMulti" ) { ret = av_dict_set(&opts, "rtsp_transport", "udp_multicast", 0); } else if ( method == "rtpRtsp" ) { ret = av_dict_set(&opts, "rtsp_transport", "tcp", 0); } else if ( method == "rtpRtspHttp" ) { ret = av_dict_set(&opts, "rtsp_transport", "http", 0); } else { Warning("Unknown method (%s)", method.c_str()); } if ( ret < 0 ) { Warning("Could not set rtsp_transport method '%s'\n", method.c_str()); } Debug(1, "Calling avformat_open_input for %s", mPath.c_str()); mFormatContext = avformat_alloc_context( ); // Speed up find_stream_info //FIXME can speed up initial analysis but need sensible parameters... //mFormatContext->probesize = 32; //mFormatContext->max_analyze_duration = 32; mFormatContext->interrupt_callback.callback = FfmpegInterruptCallback; mFormatContext->interrupt_callback.opaque = this; if ( avformat_open_input(&mFormatContext, mPath.c_str(), NULL, &opts) != 0 ) #endif { Error("Unable to open input %s due to: %s", mPath.c_str(), strerror(errno)); #if !LIBAVFORMAT_VERSION_CHECK(53, 17, 0, 25, 0) av_close_input_file(mFormatContext); #else if ( mFormatContext ) { avformat_close_input(&mFormatContext); mFormatContext = NULL; } #endif av_dict_free(&opts); return -1; } AVDictionaryEntry *e = NULL; while ( (e = av_dict_get(opts, "", e, AV_DICT_IGNORE_SUFFIX)) != NULL ) { Warning("Option %s not recognized by ffmpeg", e->key); } av_dict_free(&opts); monitor->GetLastEventId() ; Debug(1, "Opened input"); Info("Stream open %s, parsing streams...", mPath.c_str()); #if !LIBAVFORMAT_VERSION_CHECK(53, 6, 0, 6, 0) if ( av_find_stream_info(mFormatContext) < 0 ) #else if ( avformat_find_stream_info(mFormatContext, 0) < 0 ) #endif { Error("Unable to find stream info from %s due to: %s", mPath.c_str(), strerror(errno)); return -1; } Debug(4, "Got stream info"); // Find first video stream present // The one we want Might not be the first mVideoStreamId = -1; mAudioStreamId = -1; for ( unsigned int i=0; i < mFormatContext->nb_streams; i++ ) { #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) if ( mFormatContext->streams[i]->codecpar->codec_type == AVMEDIA_TYPE_VIDEO ) { #else #if (LIBAVCODEC_VERSION_CHECK(52, 64, 0, 64, 0) || LIBAVUTIL_VERSION_CHECK(50, 14, 0, 14, 0)) if ( mFormatContext->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO ) { #else if ( mFormatContext->streams[i]->codec->codec_type == CODEC_TYPE_VIDEO ) { #endif #endif if ( mVideoStreamId == -1 ) { mVideoStreamId = i; // if we break, then we won't find the audio stream continue; } else { Debug(2, "Have another video stream."); } } #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) if ( mFormatContext->streams[i]->codecpar->codec_type == AVMEDIA_TYPE_AUDIO ) { #else #if (LIBAVCODEC_VERSION_CHECK(52, 64, 0, 64, 0) || LIBAVUTIL_VERSION_CHECK(50, 14, 0, 14, 0)) if ( mFormatContext->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO ) { #else if ( mFormatContext->streams[i]->codec->codec_type == CODEC_TYPE_AUDIO ) { #endif #endif if ( mAudioStreamId == -1 ) { mAudioStreamId = i; } else { Debug(2, "Have another audio stream."); } } } // end foreach stream if ( mVideoStreamId == -1 ) Fatal("Unable to locate video stream in %s", mPath.c_str()); if ( mAudioStreamId == -1 ) Debug(3, "Unable to locate audio stream in %s", mPath.c_str()); Debug(3, "Found video stream at index %d", mVideoStreamId); Debug(3, "Found audio stream at index %d", mAudioStreamId); #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) mVideoCodecContext = avcodec_alloc_context3(NULL); avcodec_parameters_to_context( mVideoCodecContext, mFormatContext->streams[mVideoStreamId]->codecpar ); #else mVideoCodecContext = mFormatContext->streams[mVideoStreamId]->codec; #endif // STolen from ispy //this fixes issues with rtsp streams!! woot. //mVideoCodecContext->flags2 |= CODEC_FLAG2_FAST | CODEC_FLAG2_CHUNKS | CODEC_FLAG_LOW_DELAY; // Enable faster H264 decode. #ifdef CODEC_FLAG2_FAST mVideoCodecContext->flags2 |= CODEC_FLAG2_FAST | CODEC_FLAG_LOW_DELAY; #endif #if HAVE_AVUTIL_HWCONTEXT_H if ( mVideoCodecContext->codec_id == AV_CODEC_ID_H264 ) { //vaapi_decoder = new VAAPIDecoder(); //mVideoCodecContext->opaque = vaapi_decoder; //mVideoCodec = vaapi_decoder->openCodec( mVideoCodecContext ); if ( ! mVideoCodec ) { // Try to open an hwaccel codec. if ( (mVideoCodec = avcodec_find_decoder_by_name("h264_vaapi")) == NULL ) { Debug(1, "Failed to find decoder (h264_vaapi)" ); } else { Debug(1, "Success finding decoder (h264_vaapi)" ); } } if ( ! mVideoCodec ) { // Try to open an hwaccel codec. if ( (mVideoCodec = avcodec_find_decoder_by_name("h264_qsv")) == NULL ) { Debug(1, "Failed to find decoder (h264_qsv)" ); } else { Debug(1, "Success finding decoder (h264_qsv)" ); /* open the hardware device */ ret = av_hwdevice_ctx_create(&decode.hw_device_ref, AV_HWDEVICE_TYPE_QSV, "auto", NULL, 0); if (ret < 0) { Error("Failed to open the hardware device"); mVideoCodec = NULL; } else { mVideoCodecContext->opaque = &decode; mVideoCodecContext->get_format = get_format; hwaccel = true; hwFrame = zm_av_frame_alloc(); } } } } // end if h264 #endif if ( mVideoCodecContext->codec_id == AV_CODEC_ID_H264 ) { if ( (mVideoCodec = avcodec_find_decoder_by_name("h264_mmal")) == NULL ) { Debug(1, "Failed to find decoder (h264_mmal)"); } else { Debug(1, "Success finding decoder (h264_mmal)"); } } if ( (!mVideoCodec) and ( (mVideoCodec = avcodec_find_decoder(mVideoCodecContext->codec_id)) == NULL ) ) { // Try and get the codec from the codec context Error("Can't find codec for video stream from %s", mPath.c_str()); return -1; } else { Debug(1, "Video Found decoder %s", mVideoCodec->name); zm_dump_stream_format(mFormatContext, mVideoStreamId, 0, 0); // Open the codec #if !LIBAVFORMAT_VERSION_CHECK(53, 8, 0, 8, 0) Debug ( 1, "Calling avcodec_open" ); if ( avcodec_open(mVideoCodecContext, mVideoCodec) < 0 ){ #else Debug ( 1, "Calling video avcodec_open2" ); if ( avcodec_open2(mVideoCodecContext, mVideoCodec, &opts) < 0 ) { #endif AVDictionaryEntry *e = NULL; while ( (e = av_dict_get(opts, "", e, AV_DICT_IGNORE_SUFFIX)) != NULL ) { Warning("Option %s not recognized by ffmpeg", e->key); } Error("Unable to open codec for video stream from %s", mPath.c_str()); av_dict_free(&opts); return -1; } else { AVDictionaryEntry *e = NULL; if ( (e = av_dict_get(opts, "", e, AV_DICT_IGNORE_SUFFIX)) != NULL ) { Warning("Option %s not recognized by ffmpeg", e->key); } av_dict_free(&opts); } } // end if success opening codec if ( mVideoCodecContext->hwaccel != NULL ) { Debug(1, "HWACCEL in use"); } else { Debug(1, "HWACCEL not in use"); } if ( mAudioStreamId >= 0 ) { #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) mAudioCodecContext = avcodec_alloc_context3(NULL); avcodec_parameters_to_context(mAudioCodecContext, mFormatContext->streams[mAudioStreamId]->codecpar); #else mAudioCodecContext = mFormatContext->streams[mAudioStreamId]->codec; #endif if ( (mAudioCodec = avcodec_find_decoder(mAudioCodecContext->codec_id)) == NULL ) { Debug(1, "Can't find codec for audio stream from %s", mPath.c_str()); } else { Debug(1, "Audio Found decoder"); zm_dump_stream_format(mFormatContext, mAudioStreamId, 0, 0); // Open the codec #if !LIBAVFORMAT_VERSION_CHECK(53, 8, 0, 8, 0) Debug(1, "Calling avcodec_open"); if ( avcodec_open(mAudioCodecContext, mAudioCodec) < 0 ) #else Debug(1, "Calling avcodec_open2"); if ( avcodec_open2(mAudioCodecContext, mAudioCodec, 0) < 0 ) #endif Fatal("Unable to open codec for video stream from %s", mPath.c_str()); } Debug(1, "Opened audio codec"); } // end if have audio stream if ( (unsigned int)mVideoCodecContext->width != width || (unsigned int)mVideoCodecContext->height != height ) { Warning("Monitor dimensions are %dx%d but camera is sending %dx%d", width, height, mVideoCodecContext->width, mVideoCodecContext->height); } mCanCapture = true; return 1; } // int FfmpegCamera::OpenFfmpeg() int FfmpegCamera::Close() { Debug(2, "CloseFfmpeg called."); mCanCapture = false; if ( mFrame ) { av_frame_free( &mFrame ); mFrame = NULL; } if ( mRawFrame ) { av_frame_free( &mRawFrame ); mRawFrame = NULL; } if ( mVideoCodecContext ) { avcodec_close(mVideoCodecContext); #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) avcodec_free_context(&mVideoCodecContext); #endif mVideoCodecContext = NULL; // Freed by av_close_input_file } if ( mAudioCodecContext ) { avcodec_close(mAudioCodecContext); #if LIBAVCODEC_VERSION_CHECK(57, 64, 0, 64, 0) avcodec_free_context(&mAudioCodecContext); #endif mAudioCodecContext = NULL; // Freed by av_close_input_file } if ( mFormatContext ) { #if !LIBAVFORMAT_VERSION_CHECK(53, 17, 0, 25, 0) av_close_input_file(mFormatContext); #else avformat_close_input(&mFormatContext); #endif mFormatContext = NULL; } return 0; } // end FfmpegCamera::Close int FfmpegCamera::FfmpegInterruptCallback(void *ctx) { //FfmpegCamera* camera = reinterpret_cast(ctx); return zm_terminate; } #endif // HAVE_LIBAVFORMAT