[FFmpeg-devel] [PATCH 1/2] libavutil/libavfilter: add opencl wrapper to ffmpeg
Stefano Sabatini
stefasab at gmail.com
Fri Mar 22 11:39:43 CET 2013
On date Thursday 2013-03-21 16:17:12 +0800, highgod0401 encoded:
>
>
>
>
>
> highgod0401
> From 6acf8825090a722c634dba9049b32553f9daef42 Mon Sep 17 00:00:00 2001
> From: highgod0401 <highgod0401 at gmail.com>
> Date: Thu, 21 Mar 2013 15:56:29 +0800
> Subject: [PATCH 1/2] add opencl wrapper to ffmpeg
>
> ---
> configure | 4 +
> libavutil/Makefile | 3 +
> libavutil/opencl.c | 988 +++++++++++++++++++++++++++++++++++++++++++++++++++++
> libavutil/opencl.h | 219 ++++++++++++
> 4 files changed, 1214 insertions(+)
> create mode 100644 libavutil/opencl.c
> create mode 100644 libavutil/opencl.h
[...]
> diff --git a/libavutil/opencl.c b/libavutil/opencl.c
> new file mode 100644
> index 0000000..c577280
> --- /dev/null
> +++ b/libavutil/opencl.c
> @@ -0,0 +1,988 @@
> +/*
> + * Copyright (C) 2012 Peng Gao <peng at multicorewareinc.com>
> + * Copyright (C) 2012 Li Cao <li at multicorewareinc.com>
> + * Copyright (C) 2012 Wei Gao <weigao at multicorewareinc.com>
> + *
> + * 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 "opencl.h"
> +#include "avstring.h"
> +#include "log.h"
> +
> +
> +#define MAX_KERNEL_STRING_LEN 64
> +#define MAX_CLFILE_NUM 50
> +#define MAX_CLKERNEL_NUM 200
> +#define MAX_CLFILE_PATH 255
> +#define MAX_KERNEL_NUM 50
What's the difference between MAX_CLKERNEL_NUM and MAX_KERNEL_NUM?
> +#define MAX_FILTER_NAME_LEN 64
> +#define MAX_FILTER_NUM 200
> +#define MAX_KERNEL_SRCFILE_LEN 256
> +
> +
> +typedef struct OpenCLEnv {
> + cl_platform_id platform;
> + cl_context context;
> + cl_device_id devices;
> + cl_command_queue command_queue;
> +} OpenCLEnv;
> +
> +typedef struct GPUEnv {
> + //share vb in all modules in hb library
it is not obvious to me what "vb" and "hb" stand for
> + cl_platform_id platform;
> + cl_device_type device_type;
> + cl_context context;
> + cl_device_id *device_ids;
> + cl_device_id device_id;
> + cl_command_queue command_queue;
> + cl_kernel kernels[MAX_CLFILE_NUM];
> + cl_program programs[MAX_CLFILE_NUM]; //one program object maps one kernel source file
> + char kernel_srcfile[MAX_CLFILE_NUM][MAX_KERNEL_SRCFILE_LEN]; //the max len of kernel file name is 256
> + int file_count; // only one kernel file
> +
> + char kernel_names[MAX_CLKERNEL_NUM][MAX_KERNEL_STRING_LEN+1];
> + av_opencl_kernel_function kernel_functions[MAX_CLKERNEL_NUM];
> + const char *kernel_code[MAX_CLKERNEL_NUM];
> + int kernel_count;
> + int reg_kernel_count;
more meaningful name (e.g. kernel_function_count?)
> + int is_user_created; // 1: created , 0:no create and needed to create by opencl wrapper
> + uint8_t *temp_buffer;
> + int temp_buffer_size;
> +} GPUEnv;
> +
> +typedef struct FilterBufferNode {
> + char filter_name[MAX_FILTER_NAME_LEN+1];
> + void *cl_inbuf;
> + int buf_size;
> +} FilterBufferNode;
> +
> +typedef struct OpenclUtils {
> + const AVClass *class;
> + int log_offset;
> + void *log_ctx;
> +} OpenclUtils;
> +
> +typedef struct OpenclErrorMsg {
> + int err_code;
> + const char *err_str;
> +} OpenclErrorMsg;
> +
> +static OpenclErrorMsg opencl_err_msg[] = {
> + {CL_DEVICE_NOT_FOUND, "DEVICE NOT FOUND"},
> + {CL_DEVICE_NOT_AVAILABLE, "DEVICE NOT AVAILABLE"},
> + {CL_COMPILER_NOT_AVAILABLE, "COMPILER NOT AVAILABLE"},
> + {CL_MEM_OBJECT_ALLOCATION_FAILURE, "MEM OBJECT ALLOCATION FAILURE"},
> + {CL_OUT_OF_RESOURCES, "OUT OF RESOURCES"},
> + {CL_OUT_OF_HOST_MEMORY, "OUT OF HOST MEMORY"},
> + {CL_PROFILING_INFO_NOT_AVAILABLE, "PROFILING INFO NOT AVAILABLE"},
> + {CL_MEM_COPY_OVERLAP, "MEM COPY OVERLAP"},
> + {CL_IMAGE_FORMAT_MISMATCH, "IMAGE FORMAT MISMATCH"},
> + {CL_IMAGE_FORMAT_NOT_SUPPORTED, "IMAGE FORMAT NOT_SUPPORTED"},
> + {CL_BUILD_PROGRAM_FAILURE, "BUILD PROGRAM FAILURE"},
> + {CL_MAP_FAILURE, "MAP FAILURE"},
> + {CL_MISALIGNED_SUB_BUFFER_OFFSET, "MISALIGNED SUB BUFFER OFFSET"},
> + {CL_EXEC_STATUS_ERROR_FOR_EVENTS_IN_WAIT_LIST, "EXEC STATUS ERROR FOR EVENTS IN WAIT LIST"},
> + {CL_COMPILE_PROGRAM_FAILURE, "COMPILE PROGRAM FAILURE"},
> + {CL_LINKER_NOT_AVAILABLE, "LINKER NOT AVAILABLE"},
> + {CL_LINK_PROGRAM_FAILURE, "LINK PROGRAM FAILURE"},
> + {CL_DEVICE_PARTITION_FAILED, "DEVICE PARTITION FAILED"},
> + {CL_KERNEL_ARG_INFO_NOT_AVAILABLE, "KERNEL ARG INFO NOT AVAILABLE"},
> + {CL_INVALID_VALUE, "INVALID VALUE"},
> + {CL_INVALID_DEVICE_TYPE, "INVALID DEVICE TYPE"},
> + {CL_INVALID_PLATFORM, "INVALID PLATFORM"},
> + {CL_INVALID_DEVICE, "INVALID DEVICE"},
> + {CL_INVALID_CONTEXT, "INVALID CONTEXT"},
> + {CL_INVALID_QUEUE_PROPERTIES, "INVALID QUEUE PROPERTIES"},
> + {CL_INVALID_COMMAND_QUEUE, "INVALID COMMAND QUEUE"},
> + {CL_INVALID_HOST_PTR, "INVALID HOST PTR"},
> + {CL_INVALID_MEM_OBJECT, "INVALID MEM OBJECT"},
> + {CL_INVALID_IMAGE_FORMAT_DESCRIPTOR, "INVALID IMAGE FORMAT DESCRIPTOR"},
> + {CL_INVALID_IMAGE_SIZE, "INVALID IMAGE SIZE"},
> + {CL_INVALID_SAMPLER, "INVALID SAMPLER"},
> + {CL_INVALID_BINARY, "INVALID BINARY"},
> + {CL_INVALID_BUILD_OPTIONS, "INVALID BUILD OPTIONS"},
> + {CL_INVALID_PROGRAM, "INVALID PROGRAM"},
> + {CL_INVALID_PROGRAM_EXECUTABLE, "INVALID PROGRAM EXECUTABLE"},
> + {CL_INVALID_KERNEL_NAME, "INVALID KERNEL NAME"},
> + {CL_INVALID_KERNEL_DEFINITION, "INVALID KERNEL DEFINITION"},
> + {CL_INVALID_KERNEL, "INVALID KERNEL"},
> + {CL_INVALID_ARG_INDEX, "INVALID ARG INDEX"},
> + {CL_INVALID_ARG_VALUE, "INVALID ARG VALUE"},
> + {CL_INVALID_ARG_SIZE, "INVALID ARG_SIZE"},
> + {CL_INVALID_KERNEL_ARGS, "INVALID KERNEL ARGS"},
> + {CL_INVALID_WORK_DIMENSION, "INVALID WORK DIMENSION"},
> + {CL_INVALID_WORK_GROUP_SIZE, "INVALID WORK GROUP SIZE"},
> + {CL_INVALID_WORK_ITEM_SIZE, "INVALID WORK ITEM SIZE"},
> + {CL_INVALID_GLOBAL_OFFSET, "INVALID GLOBAL OFFSET"},
> + {CL_INVALID_EVENT_WAIT_LIST, "INVALID EVENT WAIT LIST"},
> + {CL_INVALID_EVENT, "INVALID EVENT"},
> + {CL_INVALID_OPERATION, "INVALID OPERATION"},
> + {CL_INVALID_GL_OBJECT, "INVALID GL OBJECT"},
> + {CL_INVALID_BUFFER_SIZE, "INVALID BUFFER SIZE"},
> + {CL_INVALID_MIP_LEVEL, "INVALID MIP LEVEL"},
> + {CL_INVALID_GLOBAL_WORK_SIZE, "INVALID GLOBAL WORK SIZE"},
> + {CL_INVALID_PROPERTY, "INVALID PROPERTY"},
> + {CL_INVALID_IMAGE_DESCRIPTOR, "INVALID IMAGE DESCRIPTOR"},
> + {CL_INVALID_COMPILER_OPTIONS, "INVALID COMPILER OPTIONS"},
> + {CL_INVALID_LINKER_OPTIONS, "INVALID LINKER OPTIONS"},
> + {CL_INVALID_DEVICE_PARTITION_COUNT, "INVALID DEVICE PARTITION COUNT"},
> +};
> +
> +static const AVClass openclutils_class = {"OPENCLUTILS", av_default_item_name,
> + NULL, LIBAVUTIL_VERSION_INT,
> + offsetof(OpenclUtils, log_offset),
> + offsetof(OpenclUtils, log_ctx)};
> +static OpenclUtils openclutils = {&openclutils_class,0,NULL};
static structs are initialized to 0, thus I don't think you need to set
the other fields to 0, same below.
> +static GPUEnv gpu_env = {0};
> +static FilterBufferNode filter_buffer[MAX_FILTER_NUM] = {{"", NULL,0}};
> +static int isinited = 0;
> +
> +void av_opencl_register_kernel(const char *kernel_name,const char *kernel_code)
> +{
> + if (gpu_env.kernel_count < MAX_CLKERNEL_NUM) {
> + gpu_env.kernel_code[gpu_env.kernel_count] = kernel_code;
> + av_strlcpy(gpu_env.kernel_names[gpu_env.kernel_count],kernel_name,MAX_KERNEL_STRING_LEN+1);
nit: ..., kernel_name, MAX_KERNEL_STRING_LEN+1
that is add missing spaces after ",", here and in the rest of the code
Also, I suspect that it is better to just complain to the user and
abort in case the registered name is too long, than silently clipping
it.
> + gpu_env.kernel_count++;
> + } else {
> + av_log(&openclutils,AV_LOG_ERROR,"av_opencl_register_kernel,kernel number is more than %d\n",MAX_CLKERNEL_NUM);
"Could not register kernel with name '%s', maximum number of registered kernels %d already reached\n"
Also the user wants to know when the function failed, so you should
return an error code.
> + }
> +}
> +static const char* opencl_errstr(int status)
nit: static const chat *opencl_errstr(int status)
for consistency
> +{
> + for (int i = 0;i < sizeof(opencl_err_msg);i++) {
> + if(opencl_err_msg[i].err_code == status)
> + return opencl_err_msg[i].err_str;
> + }
> + return "unknown error";
> +}
> +
> +static int access_binaries(cl_device_id *devices, int numdevices, const char *cl_file_name, FILE **fhandle,
> + size_t *binarysizes, char **binaries, int *binary_existed,int write)
> +{
> + FILE *fd = NULL;
> + int status;
> + char filename[1024] = {0};
> + char cl_name[1024] = {0};
> + for (int i = 0; i < numdevices; i++) {
> + if (devices[i] != 0) {
> + char devicename[1024];
> + status = clGetDeviceInfo(devices[i],
> + CL_DEVICE_NAME,
> + sizeof(devicename),
> + devicename,
> + NULL);
> + if (status == CL_SUCCESS) {
> + av_strlcpy(cl_name,cl_file_name,1024);
"1024" can be changed to "sizeof(cl_name)", so that you avoid
inconsistencies in case the size is changed (and improved readability)
> + snprintf(filename, sizeof(filename),"./%s-%s.bin", cl_name, devicename);
> + if (!write) {
> + fd = fopen(filename,"rb");
> + if (binary_existed) {
> + *binary_existed = (fd != NULL) ? 1 : 0;
*binary_existed = !!fd;
> + if (*binary_existed) {
> + if (fhandle)
> + *fhandle = fd;
> + break;
I suspect you could remove the param binary_existed, and just keep
fhandle
> + }
> + }
> + } else {
> + FILE *output = NULL;
> + if (binarysizes[i] != 0) {
> + output = fopen(filename, "wb");
> + if (!output) {
> + av_log(&openclutils,AV_LOG_ERROR,"access_binaries error,write file error\n");
Could not write file '%s'\n
> + return AVERROR_EXTERNAL;
you can return errno here, something like
if (!output) {
int err = errno;
av_log(...);
return AVERROR(err);
}
you need to save errno since the call to av_log() may overwrite it
> + }
> + fwrite(binaries[i], sizeof(char), binarysizes[i], output);
> + fclose(output);
> + }
> + }
> + } else {
> + av_log(&openclutils,AV_LOG_ERROR,"access_binaries error,clGetDeviceInfo:%s\n",opencl_errstr(status));
Could not get OpenCL device information: %s\n
> + return AVERROR_EXTERNAL;
> + break;
unneeded
> + }
> + }
> + }
> + return 0;
> +}
> +static int check_generated_binary(cl_context context, const char * cl_file_name, FILE **fhandle, int *binary_existed)
nit: add empty line after the previous function body
> +{
> + int ret = 0;
> + cl_int status;
> + size_t numdevices;
> + cl_device_id *devices;
> + status = clGetContextInfo(context,
> + CL_CONTEXT_NUM_DEVICES,
> + sizeof(numdevices),
> + &numdevices,
> + NULL);
> + if (status != CL_SUCCESS){
> + av_log(&openclutils,AV_LOG_ERROR,"check_generated_binary error,clGetContextInfo:%s\n",opencl_errstr(status));
Could not get OpenCL context number of devices: %s\n
> + return AVERROR_EXTERNAL;
> + }
> +
> + devices = av_malloc(sizeof(cl_device_id) * numdevices);
> + if (!devices)
> + return AVERROR(ENOMEM);
> +
> + /* grab the handles to all of the devices in the context. */
> + status = clGetContextInfo(context,
> + CL_CONTEXT_DEVICES,
> + sizeof(cl_device_id) * numdevices,
> + devices,
> + NULL);
> + if (status != CL_SUCCESS){
> + av_log(&openclutils,AV_LOG_ERROR,"check_generated_binary error,clGetContextInfo:%s\n",opencl_errstr(status));
Could not get OpenCL context devices: %s\n
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + ret = access_binaries(devices, numdevices,cl_file_name, fhandle, NULL, NULL, binary_existed, 0);
> +end:
> + av_free(devices);
> + return ret;
> +}
> +
> +static int generate_bin_from_kernel_source(cl_program program, const char * cl_file_name)
> +{
> + int i = 0;
> + cl_int status;
> + size_t *binarysizes = NULL;
> + size_t numdevices;
> + cl_device_id *devices = NULL;
> + char **binaries = NULL;
> + int ret = 0;
> + status = clGetProgramInfo(program,
> + CL_PROGRAM_NUM_DEVICES,
> + sizeof(numdevices),
> + &numdevices,
> + NULL);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"generate_bin_from_kernel_source error,clGetProgramInfo:%s\n",opencl_errstr(status));
> + return AVERROR_EXTERNAL;
> + }
> + devices = av_mallocz(sizeof(cl_device_id) * numdevices);
> + if (!devices)
> + return AVERROR(ENOMEM);
> + /* grab the handles to all of the devices in the program. */
> + status = clGetProgramInfo(program,
> + CL_PROGRAM_DEVICES,
> + sizeof(cl_device_id) * numdevices,
> + devices,
> + NULL);
this code could be possibly be factorized
> +
> + /* figure out the sizes of each of the binaries. */
> + binarysizes = av_mallocz(sizeof(size_t) * numdevices);
> + if (!binarysizes) {
> + ret = AVERROR(ENOMEM);
> + goto end;
> + }
> +
> + status = clGetProgramInfo(program,
> + CL_PROGRAM_BINARY_SIZES,
> + sizeof(size_t) * numdevices,
> + binarysizes, NULL);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"generate_bin_from_kernel_source error,clGetProgramInfo:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + /* copy over all of the generated binaries. */
> + binaries = av_mallocz(sizeof(char *) * numdevices);
> + if (!binaries) {
> + ret = AVERROR(ENOMEM);
> + goto end;
> + }
> + for (i = 0; i < numdevices; i++) {
> + if (binarysizes[i] != 0) {
> + binaries[i] = av_mallocz(sizeof(char) * binarysizes[i]);
> + if (!binaries[i]) {
> + ret = AVERROR(ENOMEM);
> + goto end;
> + }
> + }
> + }
> +
> + status = clGetProgramInfo(program,
> + CL_PROGRAM_BINARIES,
> + sizeof(char *) * numdevices,
> + binaries,
> + NULL);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"generate_bin_from_kernel_source error,clGetProgramInfo:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + ret = access_binaries(devices, numdevices, cl_file_name, NULL,binarysizes, binaries,NULL, 1);
> +end:
> + // Release all resouces and memory
> + if (binaries) {
> + for (i = 0;i < numdevices;i++ ) {
> + av_free(binaries[i]);
> + }
> + }
> + av_free(binaries);
> + av_free(binarysizes);
> + av_free(devices);
> + return ret;
> +}
> +
> +int av_opencl_create_kernel(const char * kernelname, AVOpenCLKernelEnv * env)
nit: ...(const char *kernel_name, AVOpenCLKernelEnv *env)
> +{
> + int status;
> + env->kernel = clCreateKernel(gpu_env.programs[0], kernelname, &status);
> + env->context = gpu_env.context;
> + env->command_queue = gpu_env.command_queue;
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"av_opencl_create_kernel error,clCreateKernel:%s\n",opencl_errstr(status));
In general it's better to provide more user readable messages and
avoid to provide too much detailed information (such as function name).
I suggest:
"Could not create OpenCL kernel: %s\n"
> + return AVERROR_EXTERNAL;
> + }
> + return 0;
> +}
> +
> +void av_opencl_release_kernel(AVOpenCLKernelEnv * env)
> +{
> + int status = clReleaseKernel(env->kernel);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"av_opencl_release_kernel error,clReleaseKernel:%s\n",opencl_errstr(status));
> + }
> + return;
useless
> +}
> +
> +static int init_opencl_env(GPUEnv *gpu_info
GPUEnv *gpu_info
for consistency it could be: *gpu_env (less confusing)
> ... ,void *ext_opencl_info)
Also why void *...
> +{
> + size_t length;
a more specific name could be useful
> + cl_int status;
> + cl_uint numplatforms, numdevices;
> + cl_platform_id *platforms = NULL;
> + cl_context_properties cps[3];
> + char platform_name[100];
> + unsigned int i;
> + int ret = 0;
> + AVOpenCLExternalInfo *opencl_info = ext_opencl_info;
and then the cast?
> + if (opencl_info) {
> + if (gpu_info->is_user_created)
> + return 0;
> + gpu_info->platform = opencl_info->platform;
> + gpu_info->is_user_created = 1;
> + gpu_info->command_queue = opencl_info->command_queue;
> + gpu_info->context = opencl_info->context;
> + gpu_info->device_ids = opencl_info->device_ids;
> + gpu_info->device_id = opencl_info->device_id;
> + gpu_info->device_type = opencl_info->device_type;
> + } else {
> + if (!gpu_info->is_user_created) {
> + status = clGetPlatformIDs(0,NULL,&numplatforms);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetPlatformIDs:%s\n",opencl_errstr(status));
> + return AVERROR_EXTERNAL;
> + }
> + if (0 < numplatforms) {
> + platforms = av_mallocz(
platform_ids may be a better name
> + numplatforms * sizeof(cl_platform_id));
> + if (!platforms) {
> + ret = AVERROR(ENOMEM);
> + goto end;
> + }
> + status = clGetPlatformIDs(numplatforms, platforms, NULL);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetPlatformIDs:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + for (i = 0; i < numplatforms; i++) {
> + status = clGetPlatformInfo(platforms[i], CL_PLATFORM_VENDOR,
> + sizeof(platform_name), platform_name,
> + NULL);
> +
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetPlatformInfo:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + gpu_info->platform = platforms[i];
> + status = clGetDeviceIDs(gpu_info->platform /* platform */,
> + CL_DEVICE_TYPE_GPU /* device_type */,
> + 0 /* num_entries */,
> + NULL /* devices */,
> + &numdevices);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetDeviceIDs:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + if (numdevices == 0) {
> + //find CPU device
> + status = clGetDeviceIDs( gpu_info->platform /* platform */,
> + CL_DEVICE_TYPE_CPU /* device_type */,
> + 0 /* num_entries */,
> + NULL /* devices */,
> + &numdevices );
> + }
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetDeviceIDs:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + if(numdevices)
> + break;
> +
> + }
> + }
> + if (!gpu_info->platform) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,get platform error\n");
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> +
> + /*
> + * Use available platform.
> + */
> + av_log(&openclutils,AV_LOG_VERBOSE,"Platform Name: %s\n",platform_name);
> + cps[0] = CL_CONTEXT_PLATFORM;
> + cps[1] = (cl_context_properties)gpu_info->platform;
> + cps[2] = 0;
> + /* Check for GPU. */
> + gpu_info->device_type = CL_DEVICE_TYPE_GPU;
> + gpu_info->context = clCreateContextFromType(
> + cps, gpu_info->device_type, NULL, NULL, &status );
> + if ((!gpu_info->context) || (status != CL_SUCCESS)) {
> + gpu_info->device_type = CL_DEVICE_TYPE_CPU;
> + gpu_info->context = clCreateContextFromType(
> + cps, gpu_info->device_type, NULL, NULL, &status );
> + }
> + if ((!gpu_info->context) || (status != CL_SUCCESS)) {
> + gpu_info->device_type = CL_DEVICE_TYPE_DEFAULT;
> + gpu_info->context = clCreateContextFromType(
> + cps, gpu_info->device_type, NULL, NULL, &status );
> + }
> + if ((!gpu_info->context) || (status != CL_SUCCESS)) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clCreateContextFromType:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
probably a loop could be more readable
> + /* Detect OpenCL devices. */
> + /* First, get the size of device list data */
> + status = clGetContextInfo(gpu_info->context, CL_CONTEXT_DEVICES,
> + 0, NULL, &length);
> + if ((status != CL_SUCCESS) || (length == 0)) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetContextInfo:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + /* Now allocate memory for device list based on the size we got earlier */
> + gpu_info->device_ids = av_mallocz(length);
> + if (!gpu_info->device_ids) {
> + ret = AVERROR(ENOMEM);
> + goto end;
> + }
> + /* Now, get the device list data */
> + status = clGetContextInfo(gpu_info->context, CL_CONTEXT_DEVICES, length,
> + gpu_info->device_ids, NULL);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clGetContextInfo:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + /* Create OpenCL command queue. */
> + gpu_info->command_queue = clCreateCommandQueue(gpu_info->context,
> + gpu_info->device_ids[0],
> + 0, &status);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"init_opencl_env error,clCreateCommandQueue:%s\n",opencl_errstr(status));
> + ret = AVERROR_EXTERNAL;
> + goto end;
> + }
> + }
> + }
> +end:
> + av_free(platforms);
> + return ret;
> +}
> +
> +static void release_opencl_env( GPUEnv *gpu_info )
gpu info/env inconsistency again
> +{
> + int i, status;
> + if (!isinited)
> + return;
> + if (gpu_info->is_user_created)
> + return;
> + gpu_info->reg_kernel_count--;
> + if (!gpu_info->reg_kernel_count) {
> + for (i = 0; i<gpu_env.file_count; i++) {
> + if (gpu_env.programs[i]) {
> + status = clReleaseProgram(gpu_env.programs[i]);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"release_opencl_env error,clReleaseProgram:%s\n",opencl_errstr(status));
> + }
> + gpu_env.programs[i] = NULL;
> + }
> + }
> + if (gpu_env.command_queue) {
> + status = clReleaseCommandQueue(gpu_env.command_queue);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"release_opencl_env error,clReleaseCommandQueue:%s\n",opencl_errstr(status));
> + }
> + gpu_env.command_queue = NULL;
> + }
> + if (gpu_env.context) {
> + status = clReleaseContext(gpu_env.context);
> + if (status != CL_SUCCESS) {
> + av_log(&openclutils,AV_LOG_ERROR,"release_opencl_env error,clReleaseContext:%s\n",opencl_errstr(status));
> + }
> + gpu_env.context = NULL;
> + }
> + isinited = 0;
> + }
you also need probably to free the allocated device_ids (?)
> + return;
> +}
> +
> +int av_opencl_register_kernel_function(const char *kernel_name, av_opencl_kernel_function function)
> +{
> + for (int i = 0; i < gpu_env.kernel_count; i++) {
This syntax is not supported by some compilers.
Use:
int i;
for (i = 0; ...)
> + if (av_strcasecmp(kernel_name, gpu_env.kernel_names[i])==0) {
> + gpu_env.kernel_functions[i] = function;
> + gpu_env.reg_kernel_count++;
> + return 0;
> + }
> + }
> + return AVERROR(EIO);
> +}
> +
> +static int detect_kerner_program(const GPUEnv *gpu_env, const char * cl_file_name)
kerner -> typo?
> +{
> + for (int i = 0; i < gpu_env->file_count; i++) {
ditto
[...]
> diff --git a/libavutil/opencl.h b/libavutil/opencl.h
> new file mode 100644
> index 0000000..e703cbc
> --- /dev/null
> +++ b/libavutil/opencl.h
> @@ -0,0 +1,219 @@
> +/*
> + * Copyright (C) 2012 Peng Gao <peng at multicorewareinc.com>
> + * Copyright (C) 2012 Li Cao <li at multicorewareinc.com>
> + * Copyright (C) 2012 Wei Gao <weigao at multicorewareinc.com>
> + *
> + * 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 "config.h"
> +
> +#ifndef LIBAVUTIL_OPENCLWRAPPER_H
> +#define LIBAVUTIL_OPENCLWRAPPER_H
> +
> +#include <CL/cl.h>
> +
> +#define AV_OPENCL_KERNEL( ... )# __VA_ARGS__
is this useful?
> +
> +#define MAX_KERNEL_NAME_LEN 150
Since this is public header you need a prefix to avoid conflicts.
#define AV_OPENCL_MAX_KERNEL_NAME_SIZE 150
"SIZE" should be better since "LEN" is usually associated to the
length of a string, which is not the same as the corresponding buffer
size.
> +
> +typedef struct AVOpenCLKernelEnv {
> + cl_context context;
> + cl_command_queue command_queue;
> + cl_program program;
> + cl_kernel kernel;
> + char kernel_name[MAX_KERNEL_NAME_LEN];
> +} AVOpenCLKernelEnv;
> +
> +typedef struct AVOpenCLExternalInfo {
> + cl_platform_id platform;
> + cl_device_type device_type;
> + cl_context context;
> + cl_device_id *device_ids;
> + cl_device_id device_id;
> + cl_command_queue command_queue;
> + char *platform_name;
> +} AVOpenCLExternalInfo;
> +
> +/**
> + * User defined function, used to set the input parameter in the kernel
> + *environment. This function launches kernel and copies data from GPU to
> + *CPU, or from CPU to GPU.
> + */
> +
> +typedef int (*av_opencl_kernel_function)(void **userdata, AVOpenCLKernelEnv *kenv);
nit here and below:
space after the leading "*"
/**
* User defined function, used to set the input parameter in the kernel
* environment. This function launches kernel and copies data from GPU to
* CPU, or from CPU to GPU.
*/
typedef int (*av_opencl_kernel_function)(void **userdata, AVOpenCLKernelEnv *kenv);
> +
> +/**
> + * Register a function for running the kernel specified by the kernel name.The function is user defined to run the kernel.
> + *
> + */
> +
> +int av_opencl_register_kernel_function(const char *kernel_name, av_opencl_kernel_function function);
/**
* Register a user defined function for running the kernel specified by the kernel name.
*/
int av_opencl_register_kernel_function(const char *kernel_name, av_opencl_kernel_function function);
Avoid the empty line after the doxy, same below.
> +
> +/**
> + *Load OpenCL kernel.
> + *
> + *@param kernel_name this kernel name is used to find the kernel in OpenCL runtime environment.
> + *@param userdata this userdata is the all parameters for running the kernel specified by kernel name
> + *@return 0 on success, others on failure
> + */
> +
> +int av_opencl_run_kernel(const char *kernel_name, void **userdata);
/**
* Load OpenCL kernel.
*
* @param kernel_name kernel name used to find the kernel in the OpenCL runtime environment
* @param userdata parameters for running the kernel
* @return 0 on success, a negative error code in case of failure
*/
int av_opencl_run_kernel(const char *kernel_name, void **userdata);
also it is a good idea to say that the function returns >= 0 in case of
success, so that it is possible to extend the function later in case
you want to provide more information through the return code.
> +
> +/**
> + * Init the run time OpenCL environment.
> + *
> + *This function must be called befor calling any function related to OpenCL.
> + *
> + *
> + *@param build_option option of compile the kernel in OpenCL runtime environment,reference "OpenCL Specification Version: 1.2 chapter 5.6.4"
> + *@param ext_opencl_info this is the extern OpenCL environment which the application program has created
> + *@return 0 on success, a negative value on error
> + */
> +
> +int av_opencl_init_run_env(const char *build_option,void *ext_opencl_info);
> +
> +/**
> + * Release OpenCL resources , this function must be called after calling any functions related to OpenCL.
> + */
> +
> +void av_opencl_release_run_env(void);
> +/**
> + * Get the OpenCL status, this function is used the check whether or not the OpenCL run time has been created.
> + *
> + *@return 0 not init, 1, inited;
> + *
> + */
> +int av_opencl_is_inited(void);
> +
> +/**
> + * Create kernel object by a kernel name on the specified OpenCL run time indicated by env parameter.
> + *
> + *@param kernelname kernel name.
> + *@param env The kernel environment which has been created by av_opencl_init_run_env
> + *@return 0 on success, a negative value on error
> + *
> + */
> +
> +int av_opencl_create_kernel(const char *kernelname, AVOpenCLKernelEnv *env);
> +
> +/**
> + * Release kernel object.
> + *
> + *@param env The kernel environment which has been created by av_opencl_init_run_env.
> + */
> +void av_opencl_release_kernel(AVOpenCLKernelEnv * env);
> +
> +/**
> + * Get the kernel environment.
> + *
> + *@param env The kernel environment which has been created by av_opencl_init_run_env.
> + *
> + */
> +
> +void av_opencl_get_kernel_env(AVOpenCLKernelEnv *env);
> +
> +/**
> + * Create OpenCL buffer, the buffer is used to save the data which is used by OpenCL kernel.
> + *
> + *@param cl_buf The pointer of OpenCL buffer.
> + *@param flags The flags which used to control buffer attribute
> + *@param size The size of OpenCL buffer
> + *@param host_ptr The host pointer of OpenCL buffer
> + *@return 0 on success, a negative value on error
> + */
> +
> +int av_opencl_create_buffer(void **cl_buf, int flags, size_t size,void *host_ptr);
> +
> +/**
> + * Read data form OpenCL buffer to memory. Src is OpenCL buffer, dst is CPU memory
> + *
> + *@param cl_buf The pointer of OpenCL buffer.
> + *@param outbuf CPU memory
> + *@param size The size of OpenCL buffer
> + *@return 0 on success, a negative value on error
> + */
> +int av_opencl_read_buffer(void *cl_inbuf, uint8_t *outbuf, size_t size);
> +
> +/**
> + * Write data from memory to OpenCL buffer. Src is frames buffer(data[0],data[1]...), dst is OpenCL buffer.
> + *
> + *@param cl_buf The pointer of OpenCL buffer.
> + *@param cl_buffer_size OpenCL buffer size
> + *@param data Picture or audio data for each plane
> + *@plane_size Size of each plane
> + *@param plane_num The input plane number
> + *@param offset The offset of OpenCL buffer start position
> + *@return 0 on success, other values on error
> + */
> +
> +int av_opencl_write_buffer(void *cl_inbuf, size_t cl_buffer_size, uint8_t **data, int *plane_size, int plane_num, int offset);
I suggest to skip this API for this round of review, especially if it
is not yet used by the deshake filter.
> +
> +/**
> + * Get OpenCL device id.
> + *
> + */
> +
> +cl_device_id av_opencl_get_device_id(void);
> +
> +/**
> + * Get OpenCL context.
> + *
> + */
> +
> +cl_context av_opencl_get_context(void);
> +
> +/**
> + * Get OpenCL command queue.
> + *
> + */
> +
> +cl_command_queue av_opencl_get_command_queue(void);
> +
> +/**
> + * Release OpenCL buffer.
> + *
> + */
> +
> +void av_opencl_release_buffer(void *cl_buf);
> +
> +/**
> + * Read frame data from OpenCL buffer to frame buffer, src buffer is OpenCL buffer, dst buffer is frame buffer(data[0],data[1]....).
> + *
> + *@param cl_buf The pointer of OpenCL buffer.
> + *@param cl_buffer_size OpenCL buffer size
> + *@param data Picture or audio data for each plane
> + *@plane_size Size of each plane
> + *@param plane_num The input plane number
> + *@return 0 on success, other values on error
> + */
> +
> +int av_opencl_read_to_frame_buffer(void *cl_inbuf, size_t cl_buffer_size, uint8_t **data,
> + int *plane_size, int plane_num);
should be put together with the other buffer-related functions.
Also "av_opencl_copy_buffer" is probably a better name.
> +
> +/**
> + * Register kernel.This function is use to register kernel and kernel code.OpenCL wrapper will use the kernel name
> + * to find the kernel code and compile it in the runtime. The kernel name should no longer than 64 and the max kernel
> + * number is 200.
> + *
> + *@param kernel_name Regist kernel name
> + *@param kernel_code Kernel code
> + */
> +
> +void av_opencl_register_kernel(const char *kernel_name,const char *kernel_code);
/**
* Register kernel code with name kernel name.
*
* The OpenCL wrapper will use the kernel name to find the kernel code
* and compile it in the runtime.
*
* @param kernel_name kernel name, should be no longer than AV_OPENCL_MAX_KERNEL_NAME_SIZE-1
* @param kernel_code kernel code to register
*/
void av_opencl_register_kernel(const char *kernel_name, const char *kernel_code);
and this function should probably return an error code.
> +
> +#endif
Mention the corresponding guard, like it is done in the other files:
#endif /* AVUTIL_OPENCL_... */
--
FFmpeg = Fabulous Fabulous Multimedia Philosofic Extravagant Gem
More information about the ffmpeg-devel
mailing list