C / C++ / MFC

DebugBreak continue issue

3-Mar-08 3:23

Have a look at ^].
Smile | :)

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.

This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong.
-- Iain Clarke

George_George3-Mar-08 2:58

George_George

3-Mar-08 2:58

Hello everyone,

I am trying the continue function triggered by JIT debugger. As mentioned in MSDN,

[url]http://msdn2.microsoft.com/zh-cn/library/h22dk1y6(en-us).aspx[/url]

--------------------
Continue
Allows execution to continue, giving the exception handler a chance to handle the exception. This option is not available for certain types of exceptions. Continue will allow the application to continue. In a native application, it will cause the exception to be rethrown. In a managed application, it will either cause the program to terminate or the exception to be handled by a hosting application.
--------------------

I write the following program to try this feature and expect when click continue, the __except code block is executed, but actually even the JIT debugger is not displayed to let us select debugger. Do you know why and how to fix it to make it have expected function (when click continue, execute exception handler code block)?

#include "Windows.h"

int main()
{
	int a;
	int b;
	int c;

	__try 
    {
	
		a = 100;
		b = 200;
		c = 300;

		DebugBreak();
		a = 400;
	
    }
    __except(GetExceptionCode() == EXCEPTION_BREAKPOINT ? 
             EXCEPTION_EXECUTE_HANDLER : EXCEPTION_CONTINUE_SEARCH) 
    {
        // No debugger is attached, so return FALSE 
        // and continue.
        return FALSE;
    }
	
    return TRUE;
}

thanks in advance,
George

difference between CComPtr <ipin> and IPin*</ipin>

hanlei00000000093-Mar-08 2:56

hanlei0000000009

3-Mar-08 2:56

Hi all

In DirectShow sample code, have lot define like CComPtr <ipin>,
if I use IPin* superseded CComPtr <ipin>, and then build and run is no error.

So, what difference between CComPtr <ipin> and IPin* ?

Re: difference between CComPtr and IPin*

CPallini3-Mar-08 3:12

Re: difference between CComPtr and IPin*

3-Mar-08 3:12

CComPtr is a smart pointer [^].
Smile | :)

Alexander ^M.,3-Mar-08 3:27

Alexander ^M.,

3-Mar-08 3:27

And IPin is an interface used by DirectShow to connect filters together..

Don't try it, just do it! Wink | ;-)

Re: difference between CComPtr and IPin*

CPallini3-Mar-08 3:47

Problem while convesion from Float to int using VC++ 6.0

3-Mar-08 3:47

Well, the difference stands between a standard pointer and a smart one. The exact nature of the interface is irrelevant to the OP context.
Smile | :)

Kishore JP3-Mar-08 2:55

Kishore JP

3-Mar-08 2:55

Hello All,
Please go through the following piece of code which is giving problem.
#include <iostream.h>
void main()
{
int angle = 2330,angul_step=55;//change angul_step to 60
float nAngle = (float)(angle / 10) - 2;
float nAngulationStep = (float)angul_step / 100;
float nFrames = (int)(nAngle/nAngulationStep);
int nPhaseFrames = (int)nFrames + 1;
cout<<"The nFrames is "<<nFrames<<endl;
cout<<"The nPhaseFrames is "<<nPhaseFrames<<endl;
}
The out put is in Debug mode.
The nFrames is 419
The nPhaseFrames is 420
The out put is in Release mode
The nFrames is 420
The nPhaseFrames is 421

In the same way: If you change the angul_step to 60 and run the out put in Debug and Release mode are
Debug mode:
The nFrames is 384
The nPhaseFrames is 385
The out put in Release mode is
The nFrames is 385
The nPhaseFrames is 386

I have replace Float with Double:
The output are different.
Debug mode:
The nFrames is 419
The nPhaseFrames is 420
Release Mode:
The nFrames is 420
The nPhaseFrames is 421

In the same way: If you change the angul_step to 60 and run the out put in Debug and Release mode are
Debug Mode
The nFrames is 385
The nPhaseFrames is 386
Release Mode
The nFrames is 385
The nPhaseFrames is 386

Changing float to double works only for specific numbers. The problem is unknown to me.
Please let me know if any one had this experience and solution.
I need to calculate the value "ROUNDDOWN((Angle-2)/AngulationStep)+1"
where Angle will be ((angle/10) -2) and angulation Step would be angul_step/100.
Thanks much.
Kishore

Re: Problem while convesion from Float to int using VC++ 6.0

David Crow3-Mar-08 3:02

Re: Problem while convesion from Float to int using VC++ 6.0

3-Mar-08 3:02

Kishore JP wrote:
float nAngle = (float)(angle / 10) - 2;
float nAngulationStep = (float)angul_step / 100;

I would change these to:

float nAngle = (angle / 10.0) - 2.0;
float nAngulationStep = angul_step / 100.0;

"Normal is getting dressed in clothes that you buy for work and driving through traffic in a car that you are still paying for, in order to get to the job you need to pay for the clothes and the car and the house you leave vacant all day so you can afford to live in it." - Ellen Goodman

"To have a respect for ourselves guides our morals; to have deference for others governs our manners." - Laurence Sterne

CPallini3-Mar-08 3:06

Re: Problem while convesion from Float to int using VC++ 6.0

3-Mar-08 3:06

(1) please use <pre> tags to surround code snippets.
(2) the following line

float nAngle = (float)(angle / 10) - 2;

is probably wrong, replace it with

float nAngle = ((float)angle / 10) - 2;

float nAngle = (angle / 10.0) - 2;

jhwurmbach3-Mar-08 3:09

jhwurmbach

3-Mar-08 3:09

Kishore JP wrote:
float nAngle = (float)(angle / 10) - 2;

(angle / 10) is first computed as integer, and the integer-result is converted to a float. Subsequently, an integer 2 is subtracted, thereby being inplicitly converted to a 2.0 float.
The resulting float is assigned to nAngle.

Kishore JP wrote:
float nAngulationStep = (float)angul_step / 100;

A integer-division of angul_step by (integer)100 is done, and the (integer)result is convertetd to a float.

Kishore JP wrote:
float nFrames = (int)(nAngle/nAngulationStep);

A float-division is done, its decimals are thrown away (that is: rounded down) and the result is implicitly converted into a float.

Numerics in C/C++ require you to really know what you want to do. If you want to divide by a float-constant, you really must write a float into your code (as the others already showed you)!

B.T.W. Why do you use float?
Do your math in double and round when you need it (e.g. when handing coordinates over to OpenGL)
If you really must use warting (e.g. writing nAngle instead of simply Angle), which is completely obsolete in the days of IDEs and Intellisense, please do it meaningful, and distinguish integers and decimal values.

Let's think the unthinkable, let's do the undoable, let's prepare to grapple with the ineffable itself, and see if we may not eff it after all.
Douglas Adams, "Dirk Gently's Holistic Detective Agency"

Re: Problem while convesion from Float to int using VC++ 6.0

Kishore JP16-Mar-08 1:20

Kishore JP

16-Mar-08 1:20

I have resolved the issue by using the /Op compiler option. The problem is compiler is doing round up when it is getting the answer ( to any expression ) x.999xxxxx (x is any number). And it is storing the X+1 to the left side of the expression. When I do cast of this parameter i am getting the only X.
By using /Op flag for improving the consistency of floating-point tests.
You can test this by the example:
231/0.55= 420.

Re: Problem while convesion from Float to int using VC++ 6.0

jhwurmbach16-Mar-08 22:19

jhwurmbach

16-Mar-08 22:19

Fine that you solved it.
But remember that you are a bit compiler specific here: Neither ohter version of VisualStudo nor different compiler will necessary bring you the same result. Just keep that in mind or maybe create a unit-test for that.

processes

duta3-Mar-08 2:16

duta

3-Mar-08 2:16

Hi there

I have 2 questions:
- How can I know the handle ID of desktop window?
- How can I launch calc, or mspaint applications from my code and if that applications are already opened to activate them?

How to do this or where can I find such examples?
I'm working on MFC.

Re: processes

David Crow3-Mar-08 3:06

3-Mar-08 3:06

duta wrote:
- How can I know the handle ID of desktop window?

Are you referring to GetDesktopWindow()?

duta wrote:
- How can I launch calc, or mspaint applications from my code and if that applications are already opened to activate them?

It really depends on how that particular application was coded. Some applications are coded such that they can only exist in memory one time. The first thing to try would be to use CreateProcess() to run the application. The next thing to try would be to use FindWindow() to search for the application's window. If it is found, use BringWindowToTop().

Re: processes

Hamid_RT3-Mar-08 21:06

Hamid_RT

3-Mar-08 21:06

I think you said me dont use of FindWindow,right? Smile | :)

CANCEL BUTTON

Adnan Merter3-Mar-08 1:17

Adnan Merter

3-Mar-08 1:17

I need to add a button to my form which stops the process,
just like cancel button in install applications.

which way do i have to use?

please somei-one hel me!!!

CPallini3-Mar-08 2:59

3-Mar-08 2:59

Adnan Merter wrote:
I need to add a button to my form which stops the process,
just like cancel button in install applications.

What process do you need to stop?

David Crow3-Mar-08 3:07

3-Mar-08 3:07

Adnan Merter wrote:
I need to add a button to my form which stops the process,

So what do you need help with, adding a button, responding to a button click, or stopping a process?

Adnan Merter3-Mar-08 3:57

Adnan Merter

3-Mar-08 3:57

i need help about stopping the process
the prcess is driving a stepper motor

i mean the button must stop all actions while my program sending data to parallel port,
just like an emergency "stop button" or "cancel button in install application"

thanks

David Crow3-Mar-08 4:18

3-Mar-08 4:18

Ok, then you'll need at least one other thread for this. Do the work of communicating with the parallel port in the secondary thread. This will free up the primary thread to handle the UI. See here for the way to gracefully stop a thread.