kernel.getOption: Get the openCL compilation options

In Jiefei-Wang/gpuMagic: An openCL compiler with the capacity to compile R functions and run the code on GPU

Get the openCL compilation options

Description

Get the openCL compilation options

Usage

kernel.getOption()

Details

#' verbose turn the verbose mode on and off.

kernelOption$localThreadNum controls the local thread number in each group, the local thread number should be a divisor of the argument .globalThreadNum. If it is set to auto, the suggested number of local thread number will be obtained from openCL API and reduced to a divisor of .globalThreadNum.

kernelOption$localThreadNumMacro specifys whether the local thread number should be inserted into the code as an macro. If it is TRUE, the macro cl_local_thread_num will be defined. It is useful when you want to dynamically allocate the memory (Mostly local memory) according to the local thread number

kernelOption$signature This is for internal usage only, please do not change it

kernelOption$flag The openCL compiler flag.

kernelOption$autoType Determine whether the type of kernel arguments should be defined as an macro, see the .kernel document for detail

Value

A list of available options

Examples

opt=kernel.getOption()
opt

Jiefei-Wang/gpuMagic documentation built on March 27, 2022, 5:23 a.m.