static av_cold int cudascale_config_props(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
AVFilterLink *inlink = outlink->src->inputs[0];
CUDAScaleContext *s = ctx->priv;
AVHWFramesContext *frames_ctx = (AVHWFramesContext*)inlink->hw_frames_ctx->data;
AVCUDADeviceContext *device_hwctx = frames_ctx->device_ctx->hwctx;
CUcontext dummy, cuda_ctx = device_hwctx->cuda_ctx;
int w, h;
int ret;
extern char vf_scale_cuda_ptx[];
ret = CHECK_CU(cuCtxPushCurrent(cuda_ctx));
if (ret < 0)
goto fail;
ret = CHECK_CU(cuModuleLoadData(&s->cu_module, vf_scale_cuda_ptx));
if (ret < 0)
goto fail;
CHECK_CU(cuModuleGetFunction(&s->cu_func_uchar, s->cu_module, "Subsample_Bilinear_uchar"));
CHECK_CU(cuModuleGetFunction(&s->cu_func_uchar2, s->cu_module, "Subsample_Bilinear_uchar2"));
CHECK_CU(cuModuleGetFunction(&s->cu_func_uchar4, s->cu_module, "Subsample_Bilinear_uchar4"));
CHECK_CU(cuModuleGetFunction(&s->cu_func_ushort, s->cu_module, "Subsample_Bilinear_ushort"));
CHECK_CU(cuModuleGetFunction(&s->cu_func_ushort2, s->cu_module, "Subsample_Bilinear_ushort2"));
CHECK_CU(cuModuleGetFunction(&s->cu_func_ushort4, s->cu_module, "Subsample_Bilinear_ushort4"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_uchar, s->cu_module, "uchar_tex"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_uchar2, s->cu_module, "uchar2_tex"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_uchar4, s->cu_module, "uchar4_tex"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_ushort, s->cu_module, "ushort_tex"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_ushort2, s->cu_module, "ushort2_tex"));
CHECK_CU(cuModuleGetTexRef(&s->cu_tex_ushort4, s->cu_module, "ushort4_tex"));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_uchar, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_uchar2, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_uchar4, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_ushort, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_ushort2, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFlags(s->cu_tex_ushort4, CU_TRSF_READ_AS_INTEGER));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_uchar, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_uchar2, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_uchar4, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_ushort, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_ushort2, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuTexRefSetFilterMode(s->cu_tex_ushort4, CU_TR_FILTER_MODE_LINEAR));
CHECK_CU(cuCtxPopCurrent(&dummy));
if ((ret = ff_scale_eval_dimensions(s,
s->w_expr, s->h_expr,
inlink, outlink,
&w, &h)) < 0)
goto fail;
if (((int64_t)h * inlink->w) > INT_MAX ||
((int64_t)w * inlink->h) > INT_MAX)
av_log(ctx, AV_LOG_ERROR, "Rescaled value for width or height is too big.\n");
outlink->w = w;
outlink->h = h;
ret = init_processing_chain(ctx, inlink->w, inlink->h, w, h);
if (ret < 0)
return ret;
av_log(ctx, AV_LOG_VERBOSE, "w:%d h:%d -> w:%d h:%d\n",
inlink->w, inlink->h, outlink->w, outlink->h);
if (inlink->sample_aspect_ratio.num) {
outlink->sample_aspect_ratio = av_mul_q((AVRational){outlink->h*inlink->w,
outlink->w*inlink->h},
inlink->sample_aspect_ratio);
} else {
outlink->sample_aspect_ratio = inlink->sample_aspect_ratio;
}
return 0;
fail:
return ret;
}
static int scale_vaapi_config_output(AVFilterLink *outlink)
{
AVFilterLink *inlink = outlink->src->inputs[0];
AVFilterContext *avctx = outlink->src;
ScaleVAAPIContext *ctx = avctx->priv;
AVVAAPIHWConfig *hwconfig = NULL;
AVHWFramesConstraints *constraints = NULL;
AVVAAPIFramesContext *va_frames;
VAStatus vas;
int err, i;
scale_vaapi_pipeline_uninit(ctx);
ctx->device_ref = av_buffer_ref(ctx->input_frames->device_ref);
ctx->hwctx = ((AVHWDeviceContext*)ctx->device_ref->data)->hwctx;
av_assert0(ctx->va_config == VA_INVALID_ID);
vas = vaCreateConfig(ctx->hwctx->display, VAProfileNone,
VAEntrypointVideoProc, 0, 0, &ctx->va_config);
if (vas != VA_STATUS_SUCCESS) {
av_log(ctx, AV_LOG_ERROR, "Failed to create processing pipeline "
"config: %d (%s).\n", vas, vaErrorStr(vas));
err = AVERROR(EIO);
goto fail;
}
hwconfig = av_hwdevice_hwconfig_alloc(ctx->device_ref);
if (!hwconfig) {
err = AVERROR(ENOMEM);
goto fail;
}
hwconfig->config_id = ctx->va_config;
constraints = av_hwdevice_get_hwframe_constraints(ctx->device_ref,
hwconfig);
if (!constraints) {
err = AVERROR(ENOMEM);
goto fail;
}
if (ctx->output_format == AV_PIX_FMT_NONE)
ctx->output_format = ctx->input_frames->sw_format;
if (constraints->valid_sw_formats) {
for (i = 0; constraints->valid_sw_formats[i] != AV_PIX_FMT_NONE; i++) {
if (ctx->output_format == constraints->valid_sw_formats[i])
break;
}
if (constraints->valid_sw_formats[i] == AV_PIX_FMT_NONE) {
av_log(ctx, AV_LOG_ERROR, "Hardware does not support output "
"format %s.\n", av_get_pix_fmt_name(ctx->output_format));
err = AVERROR(EINVAL);
goto fail;
}
}
if ((err = ff_scale_eval_dimensions(ctx,
ctx->w_expr, ctx->h_expr,
inlink, outlink,
&ctx->output_width, &ctx->output_height)) < 0)
goto fail;
if (ctx->output_width < constraints->min_width ||
ctx->output_height < constraints->min_height ||
ctx->output_width > constraints->max_width ||
ctx->output_height > constraints->max_height) {
av_log(ctx, AV_LOG_ERROR, "Hardware does not support scaling to "
"size %dx%d (constraints: width %d-%d height %d-%d).\n",
ctx->output_width, ctx->output_height,
constraints->min_width, constraints->max_width,
constraints->min_height, constraints->max_height);
err = AVERROR(EINVAL);
goto fail;
}
ctx->output_frames_ref = av_hwframe_ctx_alloc(ctx->device_ref);
if (!ctx->output_frames_ref) {
av_log(ctx, AV_LOG_ERROR, "Failed to create HW frame context "
"for output.\n");
err = AVERROR(ENOMEM);
goto fail;
}
ctx->output_frames = (AVHWFramesContext*)ctx->output_frames_ref->data;
ctx->output_frames->format = AV_PIX_FMT_VAAPI;
ctx->output_frames->sw_format = ctx->output_format;
ctx->output_frames->width = ctx->output_width;
ctx->output_frames->height = ctx->output_height;
// The number of output frames we need is determined by what follows
// the filter. If it's an encoder with complex frame reference
// structures then this could be very high.
ctx->output_frames->initial_pool_size = 10;
err = av_hwframe_ctx_init(ctx->output_frames_ref);
if (err < 0) {
av_log(ctx, AV_LOG_ERROR, "Failed to initialise VAAPI frame "
"context for output: %d\n", err);
goto fail;
}
va_frames = ctx->output_frames->hwctx;
av_assert0(ctx->va_context == VA_INVALID_ID);
vas = vaCreateContext(ctx->hwctx->display, ctx->va_config,
ctx->output_width, ctx->output_height,
VA_PROGRESSIVE,
va_frames->surface_ids, va_frames->nb_surfaces,
&ctx->va_context);
if (vas != VA_STATUS_SUCCESS) {
av_log(ctx, AV_LOG_ERROR, "Failed to create processing pipeline "
"context: %d (%s).\n", vas, vaErrorStr(vas));
return AVERROR(EIO);
}
outlink->w = ctx->output_width;
outlink->h = ctx->output_height;
outlink->hw_frames_ctx = av_buffer_ref(ctx->output_frames_ref);
if (!outlink->hw_frames_ctx) {
err = AVERROR(ENOMEM);
goto fail;
}
if (inlink->sample_aspect_ratio.num)
outlink->sample_aspect_ratio = av_mul_q((AVRational){outlink->h * inlink->w, outlink->w * inlink->h}, inlink->sample_aspect_ratio);
else
outlink->sample_aspect_ratio = inlink->sample_aspect_ratio;
av_freep(&hwconfig);
av_hwframe_constraints_free(&constraints);
return 0;
fail:
av_buffer_unref(&ctx->output_frames_ref);
av_freep(&hwconfig);
av_hwframe_constraints_free(&constraints);
return err;
}
