If you encounter run-time errors in your MATLAB® functions, the run-time stack appears in the MATLAB command window. Use the error message and stack information to learn more about the source of the error, and then either fix the issue or add error-handling code. For more information, see Viewing Errors in the Run-Time StackHandling Run-Time Errors.
The run-time stack is enabled by default for MEX code generation from MATLAB. To learn more about the source of the error, use the error message and the following stack information:
The name of the function that generated the error
The line number of the attempted operation
The sequence of function calls that led up to the execution of the function and the line at which each of these function calls occurred
Example Run-Time Stack Trace. This example shows the run-time stack trace for MEX function
Index exceeds matrix dimensions. Index value -1 exceeds valid range [1-4] of array x. Error in mlstack>mayfail (line 31) y = x(u); Error in mlstack>subfcn1 (line 5) switch (mayfail(u)) Error in mlstack (line 2) y = subfcn1(u);
The stack trace provides the following information:
The type of error.
??? Index exceeds matrix dimensions. Index value -1 exceeds valid range [1-4] of array x.
Where the error occurred.
Error in ==>mlstack>mayfail at 31 y = x(u);
The function call sequence prior to the failure.
Error in ==> mlstack>subfcn1 at 5 switch (mayfail(u)) Error in ==> mlstack at 2 y = subfcn1(u);
To help you find the source of run-time errors, the run-time stack is useful during debugging. However, when the stack is enabled, the generated code contains instructions for maintaining the run-time stack, which might slow the run time. Consider disabling the run-time stack for faster run time.
Disable the Run-Time Stack. You can disable the run-time stack by disabling the memory integrity checks as described in How to Disable Run-Time Checks.
Before disabling the memory integrity checks, verify that all array bounds and dimension checking is unnecessary.
The code generator propagates error IDs. If you throw an error
or warning in your MATLAB code, use the
in your test bench code to examine the error information and attempt
to recover, or clean up and abort. For example, for the function in Example Run-Time Stack Trace, create
a test script containing:
try mlstack_mex(u) catch % Add your error handling code here end