mirror of
https://github.com/hashcat/hashcat.git
synced 2024-12-22 14:48:12 +00:00
0bf4e3c34a
- Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
54 lines
4.1 KiB
C
54 lines
4.1 KiB
C
/**
|
|
* Author......: Jens Steube <jens.steube@gmail.com>
|
|
* License.....: MIT
|
|
*/
|
|
|
|
#ifndef EXT_OPENCL_H
|
|
#define EXT_OPENCL_H
|
|
|
|
#include <common.h>
|
|
|
|
#ifdef OSX
|
|
#include <OpenCL/cl.h>
|
|
#endif
|
|
|
|
#ifdef WIN
|
|
#include <CL/cl.h>
|
|
// #include <CL/cl_ext.h> // used for CL_DEVICE_TOPOLOGY_AMD but broken for dual GPU
|
|
#endif
|
|
|
|
#ifdef LINUX
|
|
#include <CL/cl.h>
|
|
// #include <CL/cl_ext.h> // used for CL_DEVICE_TOPOLOGY_AMD but broken for dual GPU
|
|
#endif
|
|
|
|
void hc_clBuildProgram (cl_program program, cl_uint num_devices, const cl_device_id *device_list, const char *options, void (CL_CALLBACK *pfn_notify) (cl_program program, void *user_data), void *user_data);
|
|
cl_mem hc_clCreateBuffer (cl_context context, cl_mem_flags flags, size_t size, void *host_ptr);
|
|
cl_command_queue hc_clCreateCommandQueue (cl_context context, cl_device_id device, cl_command_queue_properties properties);
|
|
//cl_command_queue hc_clCreateCommandQueueWithProperties (cl_context context, cl_device_id device, const cl_queue_properties *properties);
|
|
cl_context hc_clCreateContext (cl_context_properties *properties, cl_uint num_devices, const cl_device_id *devices, void (CL_CALLBACK *pfn_notify) (const char *, const void *, size_t, void *), void *user_data);
|
|
cl_kernel hc_clCreateKernel (cl_program program, const char *kernel_name);
|
|
cl_program hc_clCreateProgramWithSource (cl_context context, cl_uint count, const char **strings, const size_t *lengths);
|
|
cl_program hc_clCreateProgramWithBinary (cl_context context, cl_uint num_devices, const cl_device_id *device_list, const size_t *lengths, const unsigned char **binaries, cl_int *binary_status);
|
|
void hc_clEnqueueNDRangeKernel (cl_command_queue command_queue, cl_kernel kernel, cl_uint work_dim, const size_t *global_work_offset, const size_t *global_work_size, const size_t *local_work_size, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clEnqueueReadBuffer (cl_command_queue command_queue, cl_mem buffer, cl_bool blocking_read, size_t offset, size_t cb, void *ptr, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clEnqueueWriteBuffer (cl_command_queue command_queue, cl_mem buffer, cl_bool blocking_write, size_t offset, size_t cb, const void *ptr, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clEnqueueCopyBuffer (cl_command_queue command_queue, cl_mem src_buffer, cl_mem dst_buffer, size_t src_offset, size_t dst_offset, size_t cb, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clFlush (cl_command_queue command_queue);
|
|
void hc_clFinish (cl_command_queue command_queue);
|
|
void hc_clGetDeviceIDs (cl_platform_id platform, cl_device_type device_type, cl_uint num_entries, cl_device_id *devices, cl_uint *num_devices);
|
|
void hc_clGetDeviceInfo (cl_device_id device, cl_device_info param_name, size_t param_value_size, void *param_value, size_t *param_value_size_ret);
|
|
void hc_clGetPlatformIDs (cl_uint num_entries, cl_platform_id *platforms, cl_uint *num_platforms);
|
|
void hc_clGetPlatformInfo (cl_platform_id platform, cl_platform_info param_name, size_t param_value_size, void *param_value, size_t *param_value_size_ret);
|
|
void hc_clReleaseCommandQueue (cl_command_queue command_queue);
|
|
void hc_clReleaseContext (cl_context context);
|
|
void hc_clReleaseKernel (cl_kernel kernel);
|
|
void hc_clReleaseMemObject (cl_mem mem);
|
|
void hc_clReleaseProgram (cl_program program);
|
|
void hc_clSetKernelArg (cl_kernel kernel, cl_uint arg_index, size_t arg_size, const void *arg_value);
|
|
void *hc_clEnqueueMapBuffer (cl_command_queue command_queue, cl_mem buffer, cl_bool blocking_map, cl_map_flags map_flags, size_t offset, size_t cb, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clEnqueueUnmapMemObject (cl_command_queue command_queue, cl_mem memobj, void *mapped_ptr, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
void hc_clEnqueueFillBuffer (cl_command_queue command_queue, cl_mem buffer, const void *pattern, size_t pattern_size, size_t offset, size_t size, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event);
|
|
|
|
#endif
|