#include "hwcontext.h"
#include "hwcontext_internal.h"
#include "hwcontext_cuda_internal.h"
+#include "cuda_check.h"
#include "mem.h"
#include "pixdesc.h"
#include "pixfmt.h"
AV_PIX_FMT_0BGR32,
};
+#define CHECK_CU(x) FF_CUDA_CHECK_DL(device_ctx, cu, x)
+
static int cuda_frames_get_constraints(AVHWDeviceContext *ctx,
const void *hwconfig,
AVHWFramesConstraints *constraints)
static void cuda_buffer_free(void *opaque, uint8_t *data)
{
- AVHWFramesContext *ctx = opaque;
- AVCUDADeviceContext *hwctx = ctx->device_ctx->hwctx;
- CudaFunctions *cu = hwctx->internal->cuda_dl;
+ AVHWFramesContext *ctx = opaque;
+ AVHWDeviceContext *device_ctx = ctx->device_ctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
+ CudaFunctions *cu = hwctx->internal->cuda_dl;
CUcontext dummy;
- cu->cuCtxPushCurrent(hwctx->cuda_ctx);
+ CHECK_CU(cu->cuCtxPushCurrent(hwctx->cuda_ctx));
- cu->cuMemFree((CUdeviceptr)data);
+ CHECK_CU(cu->cuMemFree((CUdeviceptr)data));
- cu->cuCtxPopCurrent(&dummy);
+ CHECK_CU(cu->cuCtxPopCurrent(&dummy));
}
static AVBufferRef *cuda_pool_alloc(void *opaque, int size)
{
- AVHWFramesContext *ctx = opaque;
- AVCUDADeviceContext *hwctx = ctx->device_ctx->hwctx;
- CudaFunctions *cu = hwctx->internal->cuda_dl;
+ AVHWFramesContext *ctx = opaque;
+ AVHWDeviceContext *device_ctx = ctx->device_ctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
+ CudaFunctions *cu = hwctx->internal->cuda_dl;
AVBufferRef *ret = NULL;
CUcontext dummy = NULL;
CUdeviceptr data;
- CUresult err;
+ int err;
- err = cu->cuCtxPushCurrent(hwctx->cuda_ctx);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error setting current CUDA context\n");
+ err = CHECK_CU(cu->cuCtxPushCurrent(hwctx->cuda_ctx));
+ if (err < 0)
return NULL;
- }
- err = cu->cuMemAlloc(&data, size);
- if (err != CUDA_SUCCESS)
+ err = CHECK_CU(cu->cuMemAlloc(&data, size));
+ if (err < 0)
goto fail;
ret = av_buffer_create((uint8_t*)data, size, cuda_buffer_free, ctx, 0);
if (!ret) {
- cu->cuMemFree(data);
+ CHECK_CU(cu->cuMemFree(data));
goto fail;
}
fail:
- cu->cuCtxPopCurrent(&dummy);
+ CHECK_CU(cu->cuCtxPopCurrent(&dummy));
return ret;
}
static int cuda_transfer_data_from(AVHWFramesContext *ctx, AVFrame *dst,
const AVFrame *src)
{
- CUDAFramesContext *priv = ctx->internal->priv;
- AVCUDADeviceContext *device_hwctx = ctx->device_ctx->hwctx;
- CudaFunctions *cu = device_hwctx->internal->cuda_dl;
+ CUDAFramesContext *priv = ctx->internal->priv;
+ AVHWDeviceContext *device_ctx = ctx->device_ctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
+ CudaFunctions *cu = hwctx->internal->cuda_dl;
CUcontext dummy;
- CUresult err;
- int i;
+ int i, ret;
- err = cu->cuCtxPushCurrent(device_hwctx->cuda_ctx);
- if (err != CUDA_SUCCESS)
- return AVERROR_UNKNOWN;
+ ret = CHECK_CU(cu->cuCtxPushCurrent(hwctx->cuda_ctx));
+ if (ret < 0)
+ return ret;
for (i = 0; i < FF_ARRAY_ELEMS(src->data) && src->data[i]; i++) {
CUDA_MEMCPY2D cpy = {
.Height = src->height >> (i ? priv->shift_height : 0),
};
- err = cu->cuMemcpy2DAsync(&cpy, device_hwctx->stream);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error transferring the data from the CUDA frame\n");
- return AVERROR_UNKNOWN;
- }
+ ret = CHECK_CU(cu->cuMemcpy2DAsync(&cpy, hwctx->stream));
+ if (ret < 0)
+ goto exit;
}
- err = cu->cuStreamSynchronize(device_hwctx->stream);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error synchronizing CUDA stream\n");
- return AVERROR_UNKNOWN;
- }
+ ret = CHECK_CU(cu->cuStreamSynchronize(hwctx->stream));
+ if (ret < 0)
+ goto exit;
- cu->cuCtxPopCurrent(&dummy);
+exit:
+ CHECK_CU(cu->cuCtxPopCurrent(&dummy));
return 0;
}
static int cuda_transfer_data_to(AVHWFramesContext *ctx, AVFrame *dst,
const AVFrame *src)
{
- CUDAFramesContext *priv = ctx->internal->priv;
- AVCUDADeviceContext *device_hwctx = ctx->device_ctx->hwctx;
- CudaFunctions *cu = device_hwctx->internal->cuda_dl;
+ CUDAFramesContext *priv = ctx->internal->priv;
+ AVHWDeviceContext *device_ctx = ctx->device_ctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
+ CudaFunctions *cu = hwctx->internal->cuda_dl;
CUcontext dummy;
- CUresult err;
- int i;
+ int i, ret;
- err = cu->cuCtxPushCurrent(device_hwctx->cuda_ctx);
- if (err != CUDA_SUCCESS)
- return AVERROR_UNKNOWN;
+ ret = CHECK_CU(cu->cuCtxPushCurrent(hwctx->cuda_ctx));
+ if (ret < 0)
+ return ret;
for (i = 0; i < FF_ARRAY_ELEMS(src->data) && src->data[i]; i++) {
CUDA_MEMCPY2D cpy = {
.Height = src->height >> (i ? priv->shift_height : 0),
};
- err = cu->cuMemcpy2DAsync(&cpy, device_hwctx->stream);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error transferring the data to the CUDA frame\n");
- return AVERROR_UNKNOWN;
- }
- }
-
- err = cu->cuStreamSynchronize(device_hwctx->stream);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error synchronizing CUDA stream\n");
- return AVERROR_UNKNOWN;
+ ret = CHECK_CU(cu->cuMemcpy2DAsync(&cpy, hwctx->stream));
+ if (ret < 0)
+ goto exit;
}
- cu->cuCtxPopCurrent(&dummy);
+exit:
+ CHECK_CU(cu->cuCtxPopCurrent(&dummy));
return 0;
}
-static void cuda_device_uninit(AVHWDeviceContext *ctx)
+static void cuda_device_uninit(AVHWDeviceContext *device_ctx)
{
- AVCUDADeviceContext *hwctx = ctx->hwctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
if (hwctx->internal) {
+ CudaFunctions *cu = hwctx->internal->cuda_dl;
+
if (hwctx->internal->is_allocated && hwctx->cuda_ctx) {
- hwctx->internal->cuda_dl->cuCtxDestroy(hwctx->cuda_ctx);
+ if (hwctx->internal->flags & AV_CUDA_USE_PRIMARY_CONTEXT)
+ CHECK_CU(cu->cuDevicePrimaryCtxRelease(hwctx->internal->cuda_device));
+ else
+ CHECK_CU(cu->cuCtxDestroy(hwctx->cuda_ctx));
+
hwctx->cuda_ctx = NULL;
}
+
cuda_free_functions(&hwctx->internal->cuda_dl);
}
return ret;
}
-static int cuda_device_create(AVHWDeviceContext *ctx, const char *device,
+static int cuda_device_create(AVHWDeviceContext *device_ctx,
+ const char *device,
AVDictionary *opts, int flags)
{
- AVCUDADeviceContext *hwctx = ctx->hwctx;
+ AVCUDADeviceContext *hwctx = device_ctx->hwctx;
CudaFunctions *cu;
- CUdevice cu_device;
CUcontext dummy;
- CUresult err;
- int device_idx = 0;
+ int ret, dev_active = 0, device_idx = 0;
+ unsigned int dev_flags = 0;
+
+ const unsigned int desired_flags = CU_CTX_SCHED_BLOCKING_SYNC;
if (device)
device_idx = strtol(device, NULL, 0);
- if (cuda_device_init(ctx) < 0)
+ if (cuda_device_init(device_ctx) < 0)
goto error;
cu = hwctx->internal->cuda_dl;
- err = cu->cuInit(0);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Could not initialize the CUDA driver API\n");
+ ret = CHECK_CU(cu->cuInit(0));
+ if (ret < 0)
goto error;
- }
- err = cu->cuDeviceGet(&cu_device, device_idx);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Could not get the device number %d\n", device_idx);
+ ret = CHECK_CU(cu->cuDeviceGet(&hwctx->internal->cuda_device, device_idx));
+ if (ret < 0)
goto error;
- }
- err = cu->cuCtxCreate(&hwctx->cuda_ctx, CU_CTX_SCHED_BLOCKING_SYNC, cu_device);
- if (err != CUDA_SUCCESS) {
- av_log(ctx, AV_LOG_ERROR, "Error creating a CUDA context\n");
- goto error;
- }
+ hwctx->internal->flags = flags;
- // Setting stream to NULL will make functions automatically use the default CUstream
- hwctx->stream = NULL;
+ if (flags & AV_CUDA_USE_PRIMARY_CONTEXT) {
+ ret = CHECK_CU(cu->cuDevicePrimaryCtxGetState(hwctx->internal->cuda_device, &dev_flags, &dev_active));
+ if (ret < 0)
+ goto error;
+
+ if (dev_active && dev_flags != desired_flags) {
+ av_log(device_ctx, AV_LOG_ERROR, "Primary context already active with incompatible flags.\n");
+ goto error;
+ } else if (dev_flags != desired_flags) {
+ ret = CHECK_CU(cu->cuDevicePrimaryCtxSetFlags(hwctx->internal->cuda_device, desired_flags));
+ if (ret < 0)
+ goto error;
+ }
- cu->cuCtxPopCurrent(&dummy);
+ ret = CHECK_CU(cu->cuDevicePrimaryCtxRetain(&hwctx->cuda_ctx, hwctx->internal->cuda_device));
+ if (ret < 0)
+ goto error;
+ } else {
+ ret = CHECK_CU(cu->cuCtxCreate(&hwctx->cuda_ctx, desired_flags, hwctx->internal->cuda_device));
+ if (ret < 0)
+ goto error;
+
+ CHECK_CU(cu->cuCtxPopCurrent(&dummy));
+ }
hwctx->internal->is_allocated = 1;
+ // Setting stream to NULL will make functions automatically use the default CUstream
+ hwctx->stream = NULL;
+
return 0;
error:
- cuda_device_uninit(ctx);
+ cuda_device_uninit(device_ctx);
return AVERROR_UNKNOWN;
}
projects / ffmpeg / blobdiff