The xmcImportFunctionSettings
command sets
options for the import function feature in Vitis
Model Composer. Options are specified in the form of name/value pairs. The current
options include:
-
build
: Specifies the build environment for Model Composer. Supported values include:-
release
: The default build mode. Generates the specified output, and lets you perform simulation. -
debug
: Provides integration into a third-party compiler to let you step through and observe the imported C/C++ function using standard debug features.
-
-
compiler
: Specifies the third-party compiler to use for debugging purposes. The supported values are:-
default
: Compile with GCC or MinGW compiler for debugging using GDB debugger.Note: GCC or MinGW, and GDB are included with the standard installation of Model Composer.
-
-
blocks
: Specifies the specific block or blocks that you want to observe during the debugging process. If the option is not specified, all imported function blocks in the current design are included for debugging. You can specify one or more blocks using the following command form:
You can also unset the selection of the blocks using the following command:xmcImportFunctionSettings('blocks', {'block1','block2', ...})
xmcImportFunctionSettings('blocks', {})
You can use the MATLABgcb
command to get the block path for a specific block. The blocks must be specified as a list of blocks, even if only one block is specified. For example:xmcImportFunctionSettings('blocks', {'xmc_optical_flow/Lucas-Kanade'})
Tip: The settings specified by
xmcImportFunctionSettings
remain set until you exit
Model Composer, or change the options using xmcImportFunctionSettings
.Note: In earlier versions of Model
Composer, there existed a set of Computer Vision libraries which have been removed from
2020.2 and subsequent releases. You can import these functions into Model Composer as a
block using the
xmcImportFunction
feature. To do this,
you need to write a wrapper file around the library function. Refer to the Model
Composer HLS examples, Sobel Edge detection and Color Detection (from GitHub), to understand how to write a wrapper file and
import it as a block into the Model Composer design.For example:
Type, xmcOpenExample('color_detection')
in the
MATLAB command window. You can get the
latest version of Vision libraries from this GitHub repository.