From f0f86627fe68ede309a7ee9d7e117d618861ba88 Mon Sep 17 00:00:00 2001 From: Laurent Aimar Date: Sun, 24 Jan 2010 17:27:00 +0000 Subject: VC-1/WMV3 DXVA2 implementation It allows VLD VC-1/WMV3 decoding using DXVA2 (GPU assisted decoding API under VISTA and Windows 7). It is implemented by using AVHWAccel API. Originally committed as revision 21424 to svn://svn.ffmpeg.org/ffmpeg/trunk --- libavcodec/dxva2_vc1.c | 291 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 291 insertions(+) create mode 100644 libavcodec/dxva2_vc1.c (limited to 'libavcodec/dxva2_vc1.c') diff --git a/libavcodec/dxva2_vc1.c b/libavcodec/dxva2_vc1.c new file mode 100644 index 0000000000..95a55849be --- /dev/null +++ b/libavcodec/dxva2_vc1.c @@ -0,0 +1,291 @@ +/* + * DXVA2 WMV3/VC-1 HW acceleration. + * + * copyright (c) 2010 Laurent Aimar + * + * This file is part of FFmpeg. + * + * FFmpeg 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. + * + * FFmpeg 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 FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "dxva2_internal.h" +#include "vc1.h" +#include "vc1data.h" + +struct dxva2_picture_context { + DXVA_PictureParameters pp; + DXVA_SliceInfo si; + + const uint8_t *bitstream; + unsigned bitstream_size; +}; + +static void fill_picture_parameters(AVCodecContext *avctx, + struct dxva_context *ctx, const VC1Context *v, + DXVA_PictureParameters *pp) +{ + const MpegEncContext *s = &v->s; + const Picture *current_picture = s->current_picture_ptr; + + memset(pp, 0, sizeof(*pp)); + pp->wDecodedPictureIndex = + pp->wDeblockedPictureIndex = ff_dxva2_get_surface_index(ctx, current_picture); + if (s->pict_type != FF_I_TYPE) + pp->wForwardRefPictureIndex = ff_dxva2_get_surface_index(ctx, &s->last_picture); + else + pp->wForwardRefPictureIndex = 0xffff; + if (s->pict_type == FF_B_TYPE) + pp->wBackwardRefPictureIndex = ff_dxva2_get_surface_index(ctx, &s->next_picture); + else + pp->wBackwardRefPictureIndex = 0xffff; + if (v->profile == PROFILE_ADVANCED) { + /* It is the cropped width/height -1 of the frame */ + pp->wPicWidthInMBminus1 = avctx->width - 1; + pp->wPicHeightInMBminus1= avctx->height - 1; + } else { + /* It is the coded width/height in macroblock -1 of the frame */ + pp->wPicWidthInMBminus1 = s->mb_width - 1; + pp->wPicHeightInMBminus1= s->mb_height - 1; + } + pp->bMacroblockWidthMinus1 = 15; + pp->bMacroblockHeightMinus1 = 15; + pp->bBlockWidthMinus1 = 7; + pp->bBlockHeightMinus1 = 7; + pp->bBPPminus1 = 7; + if (s->picture_structure & PICT_TOP_FIELD) + pp->bPicStructure |= 0x01; + if (s->picture_structure & PICT_BOTTOM_FIELD) + pp->bPicStructure |= 0x02; + pp->bSecondField = v->interlace && v->fcm != 0x03 && !s->first_field; + pp->bPicIntra = s->pict_type == FF_I_TYPE; + pp->bPicBackwardPrediction = s->pict_type == FF_B_TYPE; + pp->bBidirectionalAveragingMode = (1 << 7) | + ((ctx->cfg->ConfigIntraResidUnsigned != 0) << 6) | + ((ctx->cfg->ConfigResidDiffAccelerator != 0) << 5) | + ((v->lumscale != 32 || v->lumshift != 0) << 4) | + ((v->profile == PROFILE_ADVANCED) << 3); + pp->bMVprecisionAndChromaRelation = ((v->mv_mode == MV_PMODE_1MV_HPEL_BILIN) << 3) | + (1 << 2) | + (0 << 1) | + (!s->quarter_sample ); + pp->bChromaFormat = v->chromaformat; + ctx->report_id++; + if (ctx->report_id >= (1 << 16)) + ctx->report_id = 1; + pp->bPicScanFixed = ctx->report_id >> 8; + pp->bPicScanMethod = ctx->report_id & 0xff; + pp->bPicReadbackRequests = 0; + pp->bRcontrol = v->rnd; + pp->bPicSpatialResid8 = (v->panscanflag << 7) | + (v->refdist_flag << 6) | + (s->loop_filter << 5) | + (v->fastuvmc << 4) | + (v->extended_mv << 3) | + (v->dquant << 1) | + (v->vstransform ); + pp->bPicOverflowBlocks = (v->quantizer_mode << 6) | + (v->multires << 5) | + (s->resync_marker << 4) | + (v->rangered << 3) | + (s->max_b_frames ); + pp->bPicExtrapolation = (!v->interlace || v->fcm == 0x00) ? 1 : 2; + pp->bPicDeblocked = ((v->profile != PROFILE_ADVANCED && v->rangeredfrm) << 5) | + (s->loop_filter << 1); + pp->bPicDeblockConfined = (v->postprocflag << 7) | + (v->broadcast << 6) | + (v->interlace << 5) | + (v->tfcntrflag << 4) | + (v->finterpflag << 3) | + ((s->pict_type != FF_B_TYPE) << 2) | + (v->psf << 1) | + (v->extended_dmv ); + if (s->pict_type != FF_I_TYPE) + pp->bPic4MVallowed = v->mv_mode == MV_PMODE_MIXED_MV || + (v->mv_mode == MV_PMODE_INTENSITY_COMP && + v->mv_mode2 == MV_PMODE_MIXED_MV); + if (v->profile == PROFILE_ADVANCED) + pp->bPicOBMC = (v->range_mapy_flag << 7) | + (v->range_mapy << 4) | + (v->range_mapuv_flag << 3) | + (v->range_mapuv ); + pp->bPicBinPB = 0; + pp->bMV_RPS = 0; + pp->bReservedBits = 0; + if (s->picture_structure == PICT_FRAME) { + pp->wBitstreamFcodes = v->lumscale; + pp->wBitstreamPCEelements = v->lumshift; + } else { + /* Syntax: (top_field_param << 8) | bottom_field_param */ + pp->wBitstreamFcodes = (v->lumscale << 8) | v->lumscale; + pp->wBitstreamPCEelements = (v->lumshift << 8) | v->lumshift; + } + pp->bBitstreamConcealmentNeed = 0; + pp->bBitstreamConcealmentMethod = 0; +} + +static void fill_slice(AVCodecContext *avctx, DXVA_SliceInfo *slice, + unsigned position, unsigned size) +{ + const VC1Context *v = avctx->priv_data; + const MpegEncContext *s = &v->s; + + memset(slice, 0, sizeof(*slice)); + slice->wHorizontalPosition = 0; + slice->wVerticalPosition = s->mb_y; + slice->dwSliceBitsInBuffer = 8 * size; + slice->dwSliceDataLocation = position; + slice->bStartCodeBitOffset = 0; + slice->bReservedBits = 0; + slice->wMBbitOffset = get_bits_count(&s->gb); + slice->wNumberMBsInSlice = s->mb_width * s->mb_height; /* XXX We assume 1 slice */ + slice->wQuantizerScaleCode = v->pq; + slice->wBadSliceChopping = 0; +} + +static int commit_bitstream_and_slice_buffer(AVCodecContext *avctx, + DXVA2_DecodeBufferDesc *bs, + DXVA2_DecodeBufferDesc *sc) +{ + const VC1Context *v = avctx->priv_data; + struct dxva_context *ctx = avctx->hwaccel_context; + const MpegEncContext *s = &v->s; + struct dxva2_picture_context *ctx_pic = s->current_picture_ptr->hwaccel_picture_private; + + DXVA_SliceInfo *slice = &ctx_pic->si; + + static const uint8_t start_code[] = { 0, 0, 1, 0x0d }; + const unsigned start_code_size = avctx->codec_id == CODEC_ID_VC1 ? sizeof(start_code) : 0; + const unsigned slice_size = slice->dwSliceBitsInBuffer / 8; + const unsigned padding = 128 - ((start_code_size + slice_size) & 127); + const unsigned data_size = start_code_size + slice_size + padding; + + uint8_t *dxva_data; + unsigned dxva_size; + int result; + + if (FAILED(IDirectXVideoDecoder_GetBuffer(ctx->decoder, + DXVA2_BitStreamDateBufferType, + &dxva_data, &dxva_size))) + return -1; + + result = data_size <= dxva_size ? 0 : -1; + if (!result) { + if (start_code_size > 0) + memcpy(dxva_data, start_code, start_code_size); + memcpy(dxva_data + start_code_size, + ctx_pic->bitstream + slice->dwSliceDataLocation, slice_size); + if (padding > 0) + memset(dxva_data + start_code_size + slice_size, 0, padding); + slice->dwSliceBitsInBuffer = 8 * data_size; + } + if (FAILED(IDirectXVideoDecoder_ReleaseBuffer(ctx->decoder, + DXVA2_BitStreamDateBufferType))) + return -1; + if (result) + return result; + + memset(bs, 0, sizeof(*bs)); + bs->CompressedBufferType = DXVA2_BitStreamDateBufferType; + bs->DataSize = data_size; + bs->NumMBsInBuffer = s->mb_width * s->mb_height; + assert((bs->DataSize & 127) == 0); + + return ff_dxva2_commit_buffer(avctx, ctx, sc, + DXVA2_SliceControlBufferType, + slice, sizeof(*slice), bs->NumMBsInBuffer); +} + +static int start_frame(AVCodecContext *avctx, + av_unused const uint8_t *buffer, + av_unused uint32_t size) +{ + const VC1Context *v = avctx->priv_data; + struct dxva_context *ctx = avctx->hwaccel_context; + struct dxva2_picture_context *ctx_pic = v->s.current_picture_ptr->hwaccel_picture_private; + + if (!ctx->decoder || !ctx->cfg || ctx->surface_count <= 0) + return -1; + assert(ctx_pic); + + fill_picture_parameters(avctx, ctx, v, &ctx_pic->pp); + + ctx_pic->bitstream_size = 0; + ctx_pic->bitstream = NULL; + return 0; +} + +static int decode_slice(AVCodecContext *avctx, + const uint8_t *buffer, uint32_t size) +{ + const VC1Context *v = avctx->priv_data; + const Picture *current_picture = v->s.current_picture_ptr; + struct dxva2_picture_context *ctx_pic = current_picture->hwaccel_picture_private; + + if (ctx_pic->bitstream_size > 0) + return -1; + + if (avctx->codec_id == CODEC_ID_VC1 && + size >= 4 && IS_MARKER(AV_RB32(buffer))) { + buffer += 4; + size -= 4; + } + + ctx_pic->bitstream_size = size; + ctx_pic->bitstream = buffer; + + fill_slice(avctx, &ctx_pic->si, 0, size); + return 0; +} + +static int end_frame(AVCodecContext *avctx) +{ + VC1Context *v = avctx->priv_data; + struct dxva2_picture_context *ctx_pic = v->s.current_picture_ptr->hwaccel_picture_private; + + if (ctx_pic->bitstream_size <= 0) + return -1; + + return ff_dxva2_common_end_frame(avctx, &v->s, + &ctx_pic->pp, sizeof(ctx_pic->pp), + NULL, 0, + commit_bitstream_and_slice_buffer); +} + +#if CONFIG_WMV3_DXVA2_HWACCEL +AVHWAccel wmv3_dxva2_hwaccel = { + .name = "wmv3_dxva2", + .type = CODEC_TYPE_VIDEO, + .id = CODEC_ID_WMV3, + .pix_fmt = PIX_FMT_DXVA2_VLD, + .capabilities = 0, + .start_frame = start_frame, + .decode_slice = decode_slice, + .end_frame = end_frame, + .priv_data_size = sizeof(struct dxva2_picture_context), +}; +#endif + +AVHWAccel vc1_dxva2_hwaccel = { + .name = "vc1_dxva2", + .type = CODEC_TYPE_VIDEO, + .id = CODEC_ID_VC1, + .pix_fmt = PIX_FMT_DXVA2_VLD, + .capabilities = 0, + .start_frame = start_frame, + .decode_slice = decode_slice, + .end_frame = end_frame, + .priv_data_size = sizeof(struct dxva2_picture_context), +}; + -- cgit v1.2.3