|
or, if you ALWAYS init pointers to NULL, and ALWAYS reset them to NULL after delete, like I do, then:
if (pointer)
delete [] pointer;
pointer = NULL;
Sorry to dissapoint you all with my lack of a witty or poignant signature.
|
|
|
|
|
If I were you, I would review my design. By this I mean that an object shouldn't deallocate memory it didn't allocate (which is obviously the case here since your object can't figure out if it is stack or heap memory), unless you have a good reason.
If I am wrong or said something stupid, I apologize in advance 
Michel
|
|
|
|
|
Thanks for your thoughts. I didn't design or write this stuff. I'm only trying to wrap an audio interface I got pre-cooked. I watch my own new's and delete's like a hawk and always have the leakage detector on.
The bit about nulling the pointer is very good (simple and obvious) and will be standard in my code from now on.
I do however find it strange that delete doesn't throw a catchable exception. There must be a good reason for it. I will try it on Linux and see if it tells me why.
|
|
|
|
|
Hi!
Try using:
__try
{
}
__except(1)
{
}
Mukkie
|
|
|
|
|
Hello,
First of all let me thank you for reading my question.
QUESTION:
Hello, I have a program that has one WM_TIMER missage (SetTimer...), This message is redistributed to some of my classes via SendMessage(...).
I know that postmessage differs from sendmessage in the fact that postmessage waits for the message to return, and sendmessage don't waits.
Do the messages that I redistribute work at the same time?
The problem is that all of them call functions in a non-thread safe DLL. (that I have not linked and I don't have the code).
As always thank you for your help.
|
|
|
|
|
No, timers aren't like threads. They are just regular messages processed in the same way as the rest. Think of the sending of WM_TIMER as if an external user makes the timing by himself and then clicks on a button to generate the message. No threading problems, right?
You can apply a simple sanity check: If you don't have explicitly created any thread with AfxBeginThread, CreateThread or _beginthread(ex), then your program is single-threaded and you just cannot suffer any threading problem.
In a single-threaded Windows app, all of your messages are driven by a unique message pump, which in fact serializes message processing. Two situations can arise:
- If you broadcast your message in your
OnTimer handler with SendMessages, then each SendMessage executes directly (rather than thru the message pump) and completely. So, no concurrency problems here. - If you do the broadcasting with
PostMessages, all the messages get queued and are serially executed by the message pump. No problem either.
Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo
|
|
|
|
|
Although Joaquin has already answered your question I just want to emphasized that you have gotten the SendMessage/PostMessage difference all backwards.
SendMessage is synchronous, meaning that the function calling SendMessage will be blocked until SendMessage returns (as if SendMessage was just another C++ function)
PostMessage is asynchronous, and basically it just puts you message on the receiving window's message queue. However, the handler of the posted message will not be called until the function calling PostMessage returns control to Windows (this is usually done in the message pump through calls to GetMessage/PeekMessage).
Cheers
Steen.
"To claim that computer games influence children is rediculous. If Pacman had influenced children born in the 80'ies we would see a lot of youngsters running around in dark rooms eating pills while listening to monotonous music"
|
|
|
|
|
I have a question then... (sorry for that).
The program that I'm talking about is a Numerical Control, then I need to ask for the coordinates of the machine and to look for possible errors that could have happened, and so on... (also I have to be able to send information to the DLL that allow me to retrieve and to send data to the Numerical Control Interface DLL).
I want this process to be as fastest as it can, and I have written a WM_TIMER message that is sent to everywhere in wich requiring data is needed: getting coordinates, errors, status...
The WM_TIMER handler is called each 10 Ms and re-sent to all those places that need the notification.
In this situation I believe that is the same that I use the SendMessage or the PostMessage because at the end I will be asking a lot of times the same information.
Wich one do you think that's the best choice? (SEND or POST)message?
As always thank you very much for your help and attention.
|
|
|
|
|
In a single threaded app, the only difference it makes is in the order that the code is executed.
somecode
SendMessage(somewhere...
Somemore code
Some more code, as you know, does not execute until the message has been processed.
somecode
PostMessage(somewhere...
Somemore code
In this case Somemore code will execute immediately, while the message sits in a queue.
If your main program doesn't run out of things to do (e.g. go to an idle state), the message handlers will never execute.
e.g.
void doit(
{
while (true)
{
somecode
PostMessage(somewhere...
Somemore code
}
}
in this case the message will never be handled.
versus something like
void doit()
{
somecode
PostMessage(somewhere...
Somemore code
}
in this case the message may be handled (depending on what other messages are already in the queue) immeidately after the doit procedure ends.
Therefore, SendMessage lets the sender control WHEN the message is processed, while PostMessage leaves it up to the vagueries of your message handler.
Hope this helps,
Bill
Thanks for the help,
Bill
|
|
|
|
|
I'm not quite sure I fully understand your program design. What are you Send/PostMessaging to? Windows in your app, other apps on the same PC, other apps on different PCs? If the targets of the message is in another process I'd go for your current design, using PostMessage. However, if the targets are in-process I'd let each target be represented by class. On creation each of these classes registeres itself to your OnTimer-handling class (you could keep an array of CTargetClass pointers). On each WM_TIMER I would simply call a method on each class. A simple example could look like this:
class CTargetClass : public CObject_derivative
{
public:
CTargetClass (CHandlingClass* pHandler)
....
public:
void ReturnInfo (CInfo* pInfo);
}
CTargetClass::CTargetClass (CHandlingClass* pHandler)
{
pHandler->RegisterTarget(this);
}
class CHandlingClass : public CSomeBaseClass
{
....
public:
void RegisterTarget (CTargetClass* pTarget);
afx_msg void OnTimer(UINT nIDEvent);
protected:
int m_iMaxTargets;
CTargetClass* m_pTargets;
}
void CHandlingClass::CRegisterTarget (CTargerClass* pTarget)
{
m_pTargets[m_iMaxTargets] = pTarget;
m_iMaxTargets++;
}
void CHandlingClass::OnTimer (UINT nIDEvent)
{
if (nIDEvent == ID_OF_YOUR_TIMER) {
int i;
CInfo info;
for (i=0; i<m_iMaxTargets; i++) {
m_pTargets[i].ReturnInfo(&info);
DoWhateverYouNeedWithTheInfo(info);
}
}
}
Lots of stuff omitted, but I hope you get the idea.
Cheers
Steen.
"To claim that computer games influence children is ridiculous. If Pacman had influenced children born in the 80'ies we would see a lot of youngsters running around in dark rooms eating pills while listening to monotonous music"
|
|
|
|
|
Hi,
How to access network shared resource(with UNC name) from a service
program?
for example: A service program MyService will try to open a file
MyFile.dat which resides at shared folder SharedFolder on another
machine named AnotherMachine.
\\AnotherMachine\SharedFolder\MyFile.dat
If the service MyService is configured as LocalSystem, error message
"Access to \\AnotherMachine\SharedFolder\MyFile.dat was denied."
If MyService is as valid NT account, error message
"An unknown error occurred while accessing \\AnotherMachine\SharedFolder\MyFile.dat"
Thanks for your help!
Chris Wang
|
|
|
|
|
Your second case sounds very suspicious. Have you tried stepping through the code in your service to see what happens to the file open?
What file access methods do you use? Please post the code that opens the file along with the code that produces these messages.
My guess is that the first case can be taken at face value, when the service runs as the local system account, it may not have permissions on the network resource. (One solution is to just give it permission).
The second case does not look like a permissions problem at all, but rather some other bug in the code.
HOpe this helps,
Bill
|
|
|
|
|
Hi all,
With VC7(.NET), following code give me a surprise!!!
-------snippet-------
int main(void) {
char x = 0xf0;
return 0;
}
with VC7, x = 0, with VC6 sp5, x = -15(0xf0).
when i change assignment to followings:
1) char x = static_cast<char>(0xf0);
2) char x = (char)0xf0;
3) memcpy(&x, "\xf0", 1);
4) unsigned char x = (unsigned char)0xf0;
nothing changed, x = 0.
what's the problem? me, ANSI/ISO standard, or VC7?
|
|
|
|
|
(First of all, I guess there's a typo in your post and you meant -16 instead of -15.)
This is most surprising! According to the standard, the behavior you're describing is compliant: you're trying to convert a value outside (signed) char range (namely 0xf0=240), and the standard marks this as undefined behavior. 1) 2) are equivalent, about 3) I'm not sure what's the expected behavior, and as for 4) the expected value of x is 240, not 0 (Could you check 4) again?)
What strikes me is that, despite the undefined behavior, the simplest solution, in terms of resulting code, would be that x was set to -16 (no overflow checking, direct bit pattern reinterpretation). Moreover, it is highly unlikely that MS decides to change the way its compiler behaves in such a delicate subject, thus breaking backward compatibility to gain esentially nothing. Could it be that this is a debug mode feature, and in release mode x is -16 like in the old days? Could you check that?
Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo
|
|
|
|
|
I think 3 would set x to the value of 'x'. You will copy one character from the string "\xf0" to the variable x. I don't think x is a standard escape character, so the \ will produce a warning and disappear at compile time leaving "xf0". The code copies 1 character so that would be the 'x'.
Thanks for the help,
Bill
|
|
|
|
|
Something else is going on. It works as expected (-16) on my VC7.
Tim Smith
I know what you're thinking punk, you're thinking did he spell check this document? Well, to tell you the truth I kinda forgot myself in all this excitement. But being this here's CodeProject, the most powerful forums in the world and would blow your head clean off, you've got to ask yourself one question, Do I feel lucky? Well do ya punk?
|
|
|
|
|
My main program dialog is supposed to be non-maximizable,
it has a border style of thin and no maximize button.
The problem is if I run the program using a shortcut and set the
window style to maximize, it runs it maximized and it
looks like cr^P.
I've tried to then send messages to the dialog to
simulate the user clicking the restore menu item, but
this just maximizes it again.
Is there any way to intercept this maximize command and/or
dynamically resize the form back 2 its normal size?
|
|
|
|
|
Handle the WM_WINDOWPOSCHANGING message. Windows sends this message when it is about to resize a window, you can override the proposed settings to the size of your dialog.
|
|
|
|
|
If you handle the WM_GETMINMAXINFO message you can specify the allowable dimensions of your window.
Cheers
Steen.
"To claim that computer games influence children is rediculous. If Pacman had influenced children born in the 80'ies we would see a lot of youngsters running around in dark rooms eating pills while listening to monotonous music"
|
|
|
|
|
How in MFC would I intercept messages prior to them being sent or mapped/directed to the message handlers...?
I need to perform special operations on WM_PAINT, WM_LBUTTONUP etc before they're mapped to OnPaint() OnLButtonDown(nFlags, CPoint) etc...?
Can this be accomplished using OnCmdMsg()...?
Anyone know any good explanations of OnCmdMsg..?
How would I use hooks in MFC, this would allow me to hijack a message and perform some calculations, change some WPARAM's before letting it continue. Basically I want the ability to do what was possible with plain jane SDK C code.
OnCommand() WndProc() don't lend themselves nicely to what i'm trying to get done.
"An expert is someone who has made all the mistakes in his or her field" - Niels Bohr
|
|
|
|
|
HockeyDude wrote:
I need to perform special operations on WM_PAINT, WM_LBUTTONUP etc before they're mapped to OnPaint()
WM_ERASEBKGND 
HockeyDude wrote:
How would I use hooks in MFC
I don't know if it helps you but there is a article about
keyboards hook.
Mazy
"So,so you think you can tell,
Heaven from Hell,
Blue skies from pain,...
How I wish,how I wish you were here."
Wish You Were Here-Pink Floyd-1975
|
|
|
|
|
|
Of course....Nish...your a genius...man if that'll let me do what I need done...wow...what a relief...only...will give me the handle to window the control is for, like SDK?
This is where one of the problems comes in..?
I wonder if theres a function for getting the MSG structure of the current message...i'm pretty sure that has the HWND...
in anycase...PreTranslateMessage sounds like a solid plan.
Cheers and thanx u kindly!
"An expert is someone who has made all the mistakes in his or her field" - Niels Bohr
|
|
|
|
|
|
Thanx Mike, but what I really need is the message AND the window handle that sent the message...so if the MSG structure's hwnd field is not NULL then i'm in business.
Thanx!
"An expert is someone who has made all the mistakes in his or her field" - Niels Bohr
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
