error in using calllib

2012 12 14
2 个回答
7 次查看（30 天）

0 个投票

I am using Microsoft visual studio 2008 and MATLAB 7.11.0(R2010b) i want to use callib function but it shows me an error?
>> loadlibrary('t1.dll', 't1header.h')
>> calllib('t1','Add',2,3) ??? Error using ==> calllib Method was not found.
please help me

1 个评论
显示 -1更早的评论 隐藏 -1更早的评论

Philip Borghesani
Philip Borghesani 2012-12-17
Do you recive any warnings when you load the library? What is the output of libfunctions t1? If you issue the command:
NotFound=loadlibrary('t1.dll', 't1header.h');
What is the content of NotFound?

请先登录，再进行评论。

请先登录，再回答此问题。

回答（2 个）

Walter Roberson
Walter Roberson 2012-12-14

0 个投票

That could mean that Add was not found in the library.
In the code that defines Add, is that code C or C++ ? If it is C++, is it in a
extern "C"
section? If it is C++ and is not in an extern C section, then internally it would not be known by the name Add . (This is a deliberate design feature of C++ in order to prevent accidental calling of a C++ routine without a proper "this" object in place.)

3 个评论
显示 1更早的评论 隐藏 1更早的评论

Bency Baby
Bency Baby 2012-12-17
编辑：Walter Roberson 2012-12-17
T1Header.h ----------
#ifndef _T1_HEADER_H_
#define _T1_HEADER_H_
/*check if the compiler is of C++*/
#ifdef __cplusplus
extern "C" {
int Add( int a, int b );
}
#endif
#endif
T1Header.cpp
------------
#include "stdafx.h"
#include <stdio.h>
#include "T1Header.h"
extern "C" int Add( int a, int b )
{
return( a + b );
}
These are the files.
still this error is coming
calllib('T1Header','Add',2,3)
??? Error using ==> calllib Method was not found.
Walter Roberson
Walter Roberson 2012-12-17
I'm sure I answered this...
You are attempting to call upon a function named Add, passing to it two double numbers. Your routine defines a function named Add that accepts int numbers. You must match the type of parameter you are passing with the type of parameter the routine accepts.
Reminder, in MATLAB,
A = 2
defines A to be double, not integer. It is exactly the same as if you had used
A = 2.0000000;
or
A = 0.02E+2
Whereas
A = uint8(2)
would assign an 8 bit unsigned integer to A.
The number of bits in an "int" depends upon your C++ compiler.
Philip Borghesani
Philip Borghesani 2012-12-17
Calllib will automatically convert all simple data types for you. The types are not the problem. Matlab matrices are also automatically converted to native pointer types to a vector/matrix.

请先登录，再进行评论。

Philip Borghesani
Philip Borghesani 2012-12-17

0 个投票

Given the header and code you posted it looks like you are not exporting your function from the library. Read up on building a DLL with exported functions and the declspec compiler extension. Dependency walker can be used to verify that the function was correctly exported without name mangling.

类别

在 帮助中心 和 File Exchange 中查找有关 Call C from MATLAB 的更多信息

产品

标签

提问:

Bency Baby
Bency Baby
2012-12-14

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by