Block positions have been restricted to be less than 32768. You can probably only reach this limit by using ADD_BLOCK to automatically generate extremely large models. Workarounds include shrinking the size of your blocks, or rearranging the blocks to fit the available space.
The Simulink data class designer prevents users from modifying classes if they have already been instantiated during the current MATLAB session. Reason (should probably not be published): - When modifying classes we need to regenerate the P-code for the class. However, if the class has already been instantiated then we are unable to delete the old P-file. As a result, we prevent users from doing this.
Changing the number of an input or output port number in a subsystem can cause an extra port to be added to the subsystem block in the parent system.To fix the problem:
- Copy the contents of the subsystem to the clipboard.
- Delete the old subsystem block.
- Create a new subsystem block in its place.
- Copy the contents from the clipboard to the subsystem.
- Reconnect the subsystem.
In simulation, the Deadzone block outputs NaN when input is NaN, whereas it outputs 0 in Real-Time Workshop code generation. This bug is planned to be fixed in a future release.
In R14, debugging support can be turned off for Embedded MATLAB blocks from the Stateflow Simulation field of the Configuration set in the Model Explorer. A limitation in this release is that debugging can only be disabled on a per-model basis, not a per-block basis. Disabling debugging can make some models run significantly faster.
The Blocksets submenu sometimes fails to appear when selected from the model window's Help menu. If this happens, click anywhere in the model window and then select "Blocksets" from the Help menu.
This release fails to report an error for an unsupported combination of settings for this block: you cannot set the interpolation method to benone-flatand the extrapolation method to belinear. This combination of settings causes a code generation failure.
The first time you save a new model that hasPostSaveFcnfunctions, thePostSaveFcnfunctions cannot find the model.
On Linux, the Saturation block outputsNaNif its input isNaN; on Windows, the block outputs its lower limit if its input isNaN.
When discretizing a model in the S domain, you cannot specify a sample time of 0 if the model contains a Transfer Function, State Space, or Zero Pole block. If you do specify a sample time of 0, MATLAB generates an error.