|
i assumed that much. but TRACE, as i know it, starts off with a strng pointer, and then a variable number of parameters. it doesn't take two parameters and then a format string and then the varargs.
Cleek | Image Toolkits | Thumbnail maker
|
|
|
|
|
Try this instead:
CString strLogMsg;
strLogMsg.Format(_T("Query Data Failed - Error: %lx"), e.Error());
TRACE(strLogMsg);
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
I can't use the MFC class
|
|
|
|
|
Are you sure that logmsg is large enough?
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
yes i tried by increasing the size. Before this i was thinking that we will never get crash in application by using try catch block however this is strange observation and it's happenning with release build only.
|
|
|
|
|
itkid wrote: this is strange observation and it's happenning with release build only.
It's probably happening in both but because of the extra stuff that is added to a Debug-build, it goes unnoticed.
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
A buffer with a length of 200 chars is more than large enough for a format specification of Error: %lx.
The crash is also happening outside of the try/catch block - it is happening when you return from the function. IME, a crash on return from a function is due to a fulblungered return address. Generally, this is due to stepping on your stack somehow. you can poke around by increasing the stack used both in the function and around any buffers populated by the function.
For example, add TCHAR __caCrashBuf1[ 1024 ] at the top of the function, and something like TCHAR __caCrashBufzyx[ 1024 ] around any buffers used by the function. If you do that and the crash stops or changes, you can narrow down the location that is stepping on the stack.
I would also be investigating that TGTTRACE(...) function. FWIW, it also looks like you can get rid of the logmsg buffer by building the error string in the TGTTRACE(...) function (no need to build strings twice):
TGTTRACE( llItemMedium, lLogError, L"Query Data Failed - Error: %lx", e.Error() ); Peace!
-=- James
If you think it costs a lot to do it right, just wait until you find out how much it costs to do it wrong!
Tip for new SUV drivers: Professional Driver on Closed Course does not mean your Dumb Ass on a Public Road!
DeleteFXPFiles & CheckFavorites (Please rate this post!)
|
|
|
|
|
Two questions:
1) What is the definition of the TGTTRACE macro? Is it defined in the release build?
2) Is this a Unicode build? Seems odd to mix the _T(..) and L... constructs.
|
|
|
|
|
TGTTrace is use to log the error into the file. It's unicode release
-- modified at 11:47 Wednesday 9th November, 2005
|
|
|
|
|
Hello,
Arn't trace macro's removed in release builds?
Anyway, maybe your stack gets corrupted by a non NULL terminated string. Use _sntprintf() instead. You can specify the max number of characters. This way you won't corrupt the stack.
Hope this helps.
Behind every great black man...
... is the police. - Conspiracy brother
Blog[^]
|
|
|
|
|
Hello,
ok my problem is that i don't know how to parse a config file. The file is read line by line into a char[255]. The config file looks like this:
some_proberty = some_value
test2 = test3
url = www.codeproject.com
So there is always the name an then = and then the value. In Visual C++ I would use CStrings which have the Trim, Left, Right Functions, etc. But it seems there are no such functions for the char. So how can I read my config file ?
With best regards,
Benedikt
|
|
|
|
|
|
Thank you !!!!
|
|
|
|
|
|
RedDragon2k wrote: In Visual C++ I would use CStrings which have the Trim, Left, Right Functions, etc. But it seems there are no such functions for the char.
Sure there is. If they did not exist for char, then the CString class would not have such methods (since it is based on a char type). Why not just copy the code from the three CString methods into your project? It'll work fine.
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
RedDragon2k wrote: ok my problem is that i don't know how to parse a config file. The file is read line by line into a char[255]. The config file looks like this:
Two Option :-
#1. Use GetPrivateProfileString() api to extract data from file
#2. use CStdioFile::ReadString() method (which read the File Line by line) and then you have to parse the data manaully
"Opinions are neither right nor wrong. I cannot change your opinion. I can, however, change what influences your opinion." - David Crow
cheers,
Alok Gupta
VC Forum Q&A :- I/ IV
|
|
|
|
|
is there a way to change values of domain users HKEY_CURRENT_USERS key on a particualr machine while that user is not logged on.
a programmer traped in a thugs body
|
|
|
|
|
How will the HKCU hive exist if there is no current user? Perhaps you'll want to update the HKU hive instead.
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
If you have a local user its current user comes from one the HKU key. However we don't have roaming profiles in our environment but a domain user can log onto a machine make changes to his HKCU values and it is retained however if I'm logged on as an admin I can't find a key under HKU tat corresponds to that user.
a programmer traped in a thugs body
|
|
|
|
|
My guess is that the registry hive for the user on the domain is located on a server somewhere, in which case you need a remote registry connection and look for user profile on the domain controller or other system containing the information.
If a 'domain' user logs onto a local system and he has not been added to that local system explicitly, his prfile is probably temporary while he is logged in and thus not found within the registry of the current (local) system.
|
|
|
|
|
Take a look at the LoadUserProfile function.
The hProfile member of PROFILEINFO parameter is a registry key handle opened to the root of the user's hive
|
|
|
|
|
Can some one tell me how to add an entry to a string table. Is it as simple as inserting a String.I was looking at inserting IDS_ABC.
Thanks,
M
-- modified at 13:04 Tuesday 8th November, 2005
|
|
|
|
|
Well, you can edit the .rc and .h files manually, or you can do it via the IDE. Click the ResourceView tab on Workspace window. Right-click String Table and select Insert String Table from the context menu. Now in the right pane, either press the Insert key or right-click and select New String. Does that help?
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
|
Thanks, That was what I was doing. Just wanted to be sure if it was done differently. Could you please explain the use of String table.
Thanks
M
|
|
|
|
|
TUMB wrote: Could you please explain the use of String table.
String resources are simply null-terminated Unicode strings that can be loaded when needed from the file. Using string literals instead would make your code harder to translate into other languages, plus it makes the memory footprint larger. String resources are not loaded into memory unless they are asked for.
"Take only what you need and leave the land as you found it." - Native American Proverb
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
