richedit(?1) + windows 98 + em_replacesel = codepage problem

TheMSsForum.com: The Microsoft Software Forum

The MSS Forum ‹ VC
- Archive
  - Biz
  - MCSE
  - CRM
  - Drivers
  - Framework
  - ADO
  - ASP
  - Compact
  - Forms
  - Dotnet
  - C#
  - VB
  - FontpageGen
  - Excel
  - WorkSheet
  - Exchange
  - Setup
  - Fox
  - Fontpage
  - ASP
  - IIS
  - Entourage
  - Money
  - Messanger
  - PocketPC
  - Powerpoint
  - Project
  - Publisher
  - Excel
  - VB
  - Security
  - Portal
  - Services
  - SQLServerDev
  - SVCS
  - SQLServer
  - VB
  - VC
  - MFC
  - ExcelGen

- Previous
  - 1
    - deadlock on SMP machines - perhaps caused by wldap32.dll? Hello, Apologies in advance, this is a very long post, but I'm trying to be as complete as possible. I'm also not sure if this is the correct group, but it seemed appropriate. We have a process that runs as part of our product that has been working fine on Windows. However, we recently had an issue onsite where these processes (there are usually multiple instances) would hang and become completely unresponsive on SMP (Compaq 8-way) machines running Windows NT and 2000. I've managed to recreate the issue in our test lab on a dual Xeon (hyperthreaded) box running Windows 2003. There is definitely some sort of deadlock situation here, but I'm at a loss how to proceed after banging my head against this for a week now. There is some evidence that suggests the DllInit function of wldap32 is somehow a player in this - the details are further down. First, let me provide a little context: This service (it's not really an application) is written in VC++ 6.0, and is spawned via a daemon that is also part of our application. In a typical deployment there is one of these service processes for each processor on the local machine. The service instantiates a Java Virtual Machine via JNI, and most of our core code executes within the VM. One of the functions of this service is to spawn processes and feed them input and capture their output. This is acheived by the VM loading a VC++ 6.0 bridge library, which in turn loads another VC++ 6.0 library (we'll call this the "work library") which does the actual work. When all of the assigned work is through, the work library is unloaded until the next time. The work library is fairly simple, and the core behavior involves invoking CreateProcess (usually cmd.exe, which in my test setup just echoes some msg), and creating threads to write to the input handle, and read from the output and error handles. This has been working on all flavors of Windows (uniprocessor) for quite some time. Now, to the details. Here is the basic sequence leading up to the hang: 1) Process loads the work library 2) Process performs some number of brief "work units" (CreateProcess() + accompanying threads, and then cleanup) happily and healthily 3) Process CPU utilization drops to 0, and there is no more activity, ever. This lockup appears to happen faster when more instances are running. Attaching cdb to these hung processes consistently reveals two scenarios: Scenario A - the "Is wldap32.dll doing something bad?" scenario ---------- The work library thread which called CreateProcess has successfully waited on the process to exit. It waits on an Event that will be set by the thread which reads the output from the process, but that Event is never set. Mainly because those helper threads are stuck. The debugger always has to forcibly break into the process b/c the loader lock is held. !locks reveals two locks in contention: CritSec ntdll!LdrpLoaderLock+0 at 77FC2340 LockCount 8 RecursionCount 1 OwningThread b80 EntryCount 2a ContentionCount 2a *** Locked CritSec WLDAP32!PerThreadListLock+0 at 76F34010 LockCount 1 RecursionCount 1 OwningThread d64 EntryCount 1 ContentionCount 1 *** Locked ---- Stack trace for thread b80: 0f5dfba8 77f43741 77f5d64e 000005a0 00000000 SharedUserData!SystemCallStub+0x4 0f5dfbac 77f5d64e 000005a0 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0f5dfbe8 77f42044 000005a0 76f1129f 76f34010 ntdll!RtlpWaitForCriticalSection+0x126 0f5dfbf0 76f1129f 76f34010 00000000 0f5dfc1c ntdll!RtlEnterCriticalSection+0x46 0f5dfc00 76f11257 00000b80 77f47dc9 76f10000 WLDAP32!AddPerThreadEntry+0x2f 0f5dfc08 77f47dc9 76f10000 00000002 00000000 WLDAP32!LdapDllInit+0x3e 0f5dfc28 77f5d067 76f11180 76f10000 00000002 ntdll!LdrpCallInitRoutine+0x14 0f5dfcbc 77f5cef6 0f5dfd30 0f5dfd30 00000000 ntdll!LdrpInitializeThread+0xd4 0f5dfd1c 77f45241 0f5dfd30 77f40000 00000000 ntdll!LdrpInitialize+0x171 00000000 00000000 00000000 00000000 00000000 ntdll!KiUserApcDispatcher+0x7 To me, this begs the question: why is this library referred to in the first place? We certainly don't use LDAP and we aren't using ActiveDirectory, either. 'lm' reveals that it is an unloaded module. My only guess is that the library init routine is called in order to see if LDAP is enabled in case it's needed to check process/thread creation security tokens. The interesting thing about WLDAP32!PerThreadListLock+0 is that the owning thread (d64) is nowhere to be found! So, no stack trace. There are several other threads waiting on locks. None of them are in my input/output/error read/write user code, they are all in LdrpInitialize or kernel32!ExitThread. I suspect the exiting threads are mine. 5 threads have the same trace (with differing addresses, but waiting on the same critical section - 77fc2340): 0f6dfbec 77f43741 77f5d64e 00000330 00000000 SharedUserData!SystemCallStub+0x4 0f6dfbf0 77f5d64e 00000330 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0f6dfc2c 77f42044 00000330 77f5cfcc 77fc2340 ntdll!RtlpWaitForCriticalSection+0x126 0f6dfc34 77f5cfcc 77fc2340 00000000 7ffdf000 ntdll!RtlEnterCriticalSection+0x46 0f6dfcbc 77f5cef6 0f6dfd30 0f6dfd30 00000000 ntdll!LdrpInitializeThread+0x2f 0f6dfd1c 77f45241 0f6dfd30 77f40000 00000000 ntdll!LdrpInitialize+0x171 00000000 00000000 00000000 00000000 00000000 ntdll!KiUserApcDispatcher+0x7 And here are the exiting threads: c18: ---- 0e0dfe70 77f43741 77f5d64e 00000330 00000000 SharedUserData!SystemCallStub+0x4 0e0dfe74 77f5d64e 00000330 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0e0dfeb0 77f42044 00000330 77f7c658 77fc2340 ntdll!RtlpWaitForCriticalSection+0x126 0e0dfeb8 77f7c658 77fc2340 00000000 7ffa4000 ntdll!RtlEnterCriticalSection+0x46 0e0dff44 77e57eb3 00000000 0c0bd418 00a87b88 ntdll!LdrShutdownThread+0x30 0e0dff7c 77bc917d 00000000 77bc91f3 00000000 kernel32!ExitThread+0x41 0e0dff84 77bc91f3 00000000 00000000 00000000 msvcrt!_endthreadex+0x25 0e0dffb8 77e4a990 00a87b88 00000000 00000000 msvcrt!_endthreadex+0x9b 0e0dffec 00000000 77bc917e 00a87b88 00000000 kernel32!BaseThreadStart+0x34 9cc: ---- 0e1dfe70 77f43741 77f5d64e 00000330 00000000 SharedUserData!SystemCallStub+0x4 0e1dfe74 77f5d64e 00000330 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0e1dfeb0 77f42044 00000330 77f7c658 77fc2340 ntdll!RtlpWaitForCriticalSection+0x126 0e1dfeb8 77f7c658 77fc2340 00000000 7ffa3000 ntdll!RtlEnterCriticalSection+0x46 0e1dff44 77e57eb3 00000000 0b2e6c48 0b38b590 ntdll!LdrShutdownThread+0x30 0e1dff7c 77bc917d 00000000 77bc91f3 00000000 kernel32!ExitThread+0x41 0e1dff84 77bc91f3 00000000 00000000 00000000 msvcrt!_endthreadex+0x25 0e1dffb8 77e4a990 0b38b590 00000000 00000000 msvcrt!_endthreadex+0x9b 0e1dffec 00000000 77bc917e 0b38b590 00000000 kernel32!BaseThreadStart+0x34 I can provide further information if it is needed. ------------------------------------------------------------------------------- Scenario B - the "what happened to the malloc lock?" scenario ---------- As in scenario A, the work library thread which called CreateProcess has successfully waited on the process to exit. It waits on an Event that will be set by the thread which reads the output from the process, but that Event is never set. These helper threads are also stuck, but this time they are actually in user code. The debugger does not have to forcibly break into this process. !locks only reveals one lock in contention: CritSec MSVCRTD!__app_type+34 at 10264780 LockCount 1 RecursionCount 0 OwningThread 0 EntryCount 4 ContentionCount 4 *** Locked The owning thread is 0. Uh oh. My two helper threads also happen to be waiting on this lock: ea8: ---- 0f2eed90 77f43741 77f5d64e 00000680 00000000 SharedUserData!SystemCallStub+0x4 0f2eed94 77f5d64e 00000680 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0f2eedd0 77f42044 00000680 1020b5a3 10264780 ntdll!RtlpWaitForCriticalSection+0x126 0f2eedd8 1020b5a3 10264780 0f2eeec4 0f2eee18 ntdll!RtlEnterCriticalSection+0x46 0f2eede8 1021352a 00000009 0f2efed8 0f2eee74 MSVCRTD!_lock+0x93 0f2eee18 102134ce 0de09de0 00000001 0f2eee34 MSVCRTD!_free_dbg+0x2a 0f2eee28 104bd0ed 0de09de0 0f2eee44 104ad0f0 MSVCRTD!free+0xe 0f2eee34 104ad0f0 0de09de0 0f2eeec4 0f2eee5c MSVCP60D!operator delete+0xd 0f2eee44 1048be9e 0de09de0 00000021 ffffffff MSVCP60D!std::allocator<char>::deallocate+0x10 0f2eee5c 10489b71 00000001 0f2eeec4 0f2efee4 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::_Tidy+0x6e 0f2eee6c 0c0e3c99 0f2eff74 0f2efeec 0de09ce0 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::~basic_string<char,std::char_traits<char>,std::allocator<char> >+0x11 0f2efee4 0c0d9ddb 00000000 00000000 0de09ce0 worklib!NullCopyRunner::run+0x119 0f2eff84 1020bf53 0de09c98 00000000 00000000 worklib!JobThread::execute+0x4b 0f2effb8 77e4a990 0de09ce0 00000000 00000000 MSVCRTD!_beginthreadex+0x133 0f2effec 00000000 1020bee0 0de09ce0 00000000 kernel32!BaseThreadStart+0x34 7d0: ---- 0f1ee64c 77f43741 77f5d64e 00000680 00000000 SharedUserData!SystemCallStub+0x4 0f1ee650 77f5d64e 00000680 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc 0f1ee68c 77f42044 00000680 1020b5a3 10264780 ntdll!RtlpWaitForCriticalSection+0x126 0f1ee694 1020b5a3 10264780 00000021 0f1ee6d8 ntdll!RtlEnterCriticalSection+0x46 0f1ee6a4 102129da 00000009 00000000 0f1ee7cc MSVCRTD!_lock+0x93 0f1ee6d8 102129a6 00000021 00000001 00000001 MSVCRTD!_nh_malloc_dbg+0x2a 0f1ee6f4 1020e2cf 00000021 00000001 0f1eff44 MSVCRTD!_nh_malloc+0x16 0f1ee708 104b7a69 00000021 0f1ee728 104ad0d2 MSVCRTD!operator new+0xf 0f1ee714 104ad0d2 00000021 00000000 0f1eff44 MSVCP60D!std::_Allocate+0x19 0f1ee728 1048ba18 00000021 00000000 00000000 MSVCP60D!std::allocator<char>::allocate+0x12 0f1ee770 1048bccc 00000017 00000000 0f1eff44 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::_Copy+0x58 0f1ee784 1048a005 00000017 00000001 0f1eff44 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::_Grow+0x10c 0f1ee798 1048a060 0c136674 00000017 0f1eff44 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::assign+0x15 0f1ee7ac 10489ad7 0c136674 0f1eff44 0f1eff84 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::assign+0x20 0f1ee7bc 0c0d9e1f 0c136674 0f1eff40 1020d797 MSVCP60D!std::basic_string<char,std::char_traits<char>,std::allocator<char> >::basic_string<char,std::char_traits<char>,std::allocator<char> >+0x27 0f1eff84 1020bf53 0de09ba8 00000000 00000000 worklib!JobThread::execute+0x8f 0f1effb8 77e4a990 0de09bf0 00000000 00000000 MSVCRTD!_beginthreadex+0x133 0f1effec 00000000 1020bee0 0de09bf0 00000000 kernel32!BaseThreadStart+0x34 So it appears that the malloc/heap/whatever-you-call-it lock is somehow munged. None of the other threads reveal anything interesting. ------------------------------------------------------------------------------- Summary: I considered the possibility of heap contamination, but using gflags I enabled full page heap and every other form of heap checking, but there appear to be no problems there. I am suspicious that there is something else at work because these scenarios happen exactly the same every time. As I mentioned before, this code has been working perfectly on Windows uniprocessor systems for more than a year now. The only variable is the multiprocessor environment. I am seriously thinking that this is system related. I've noticed that there have been some deadlock related fixes for wldap32.dll, including a recent hotfix that postdates the release of Windows 2003. I'm wondering if this would fix our problem? I eagerly await any new ideas on this issue! I am pretty much out of them at this point. If I can provide further information on this, please ask. Thanks for reading! Regards, Marc Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232683
  - 2
    - Can rebasing affect correctness of information in the map file generated before? Hi everybody For some strange reasons offsets in dlls loaded into memory do not match those, mentioned in map file. We suspect that this happens due to automatic rebasing of all the binary modules involved in workspace at the end of complete build of all the projects. Can this possibly be the case and how to avoid it? It looks like order of some sections is changed. Quite often offsets exceed the last offset mentioned in the map file. We generate map file with options /map /MAPINFO:LINES /MAPINFO:EXPORTS during building of a particular project. Rebasing is made after all the projects were built and involves all the modules that will be loaded: rebase.exe -b 0x60060000 [module ...] Kind regards, Antonio Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232682
  - 3
    - how to compile ".hh" file? (how can i assoiate a tool) Hi, I am having ".... No compile tool is associated with the file extention" message when I try to compile a .hh file. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232672
  - 4
    - convert unsigned char array from managed to unmanaged Do you know how to convert unsigned char array from managed to unmanaged e.g. unsigned char charArray __gc[] convert it to unsigned char charArray_[] and vice versa Thanks Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232670
  - 5
    - deallocating vector storage The following is a thread from c.l.c++...since the behavior is apparantly platform-dependent, I'm looking for the best solution for MSVC++ 6.0 and also for 7.1 if it is different. Regards, Ryan Tino wrote: > In TC++PL (third edition), Section 16.3.8, Stroustrup gives the > following as a way to give memory from a std::vector back to the > system: > > vector<int> tmp = v; // v.size() == 0 > v.swap( tmp ); // v now has capacity of tmp (default), extra capacity > returned to system > > We need to have a general method that returns memory back to the > system for vectors which are no longer needed, but have not yet gone > out of scope. Are there alternative suggestions for accomplishing > this? There is no guarantee that the storage will be returned. Swapping with an empty vector, clearing the vector, resizing it to 0, all are good attempts to accomplish what you need, but the Standard makes no promises WRT the memory allocation behaviour. It's all platform-dependent and implementation-dependent, I believe. V Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232665
  - 6
    - Which interfaces should be referenced in coclass definition of idl file? Hi everybody, Is there any rule or principle based on which one can decide which interfaces from those supported by the object to include into coclass definition and which not? What influence does it have depending on where component will be used: C++, Visual Basic, VBScript (Internet Explorer), Delphi etc. ? For example in the following case when I try to insert control into Visual Basic, it just hangs. [ uuid(BBBBBBBB-BBBB-BBBB-BBBB-BBBBBBBBBBBB), // actual guid replaced helpstring("MyClient class") ] coclass MyClient { [default] interface IMyDualInterfaces; [default, source] dispinterface _IMySourceEvents; interface IMyCustomInterfaces0; interface IDispatch; interface IMyCustomInterfaces1; interface IMyCustomInterfaces2; interface IViewObjectEx; interface IViewObject2; interface IViewObject; interface IOleInPlaceObjectWindowless; interface IOleInPlaceObject; interface IOleWindow; interface IOleInPlaceActiveObject; interface IOleControl; interface IOleObject; interface IPersistStreamInit; interface IPersistPropertyBag; interface IPersist; interface ISupportErrorInfo; interface IConnectionPointContainer; interface ISpecifyPropertyPages; interface IQuickActivate; interface IPersistStorage; interface IDataObject; interface IProvideClassInfo; interface IProvideClassInfo2; interface IMarshal; interface IObjectSafety; interface IConnectionPointContainer; }; Any reference to an article that might help will be greatly appreciated. Thanks in advance, Antonio Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232664
  - 7
    - Re: windows taskbar and start menu Concerning the context menu, I use a tool called HPS HwndSpy (http://www.highplains.net/hwndspy.html) and it displays the HWND of the context menu. I'm not sure this will really help you - the HWND is recreated each time you right-click on the taskbar. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232663
  - 8
    - DLL with imports that should be used from different mainprogr Dear Gurus, I am on MSDEV 6.0 working from the command line. My Code is C Code. I am trying to generate a DLL that imports symbols from a main program. The DLL should be used by different main programs that all export the same symbols. The usual approach, if the DLL is only used by one main program would be to simply link the DLL against the export library of the main program. However if I do exactly that, then the other main programs will crash with the DLL when the DLL calls functions from the main program. (the reason being that the dll has somehow incorporated the name of the main program of which the symbols were imported at DLL link time). So how can I properly create a DLL that imports symbols from the main program without telling the DLL anything else then the names of the symbols that will be imported? And how can I tell the DLL the imported symbolnames without using the export lib of a main program? thank you very much for any insights Wolfgang Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232657
  - 9
    - array of handles arithmetic C, win32, VC7 Suppose I have an array of handles, like this: HANDLE* thread_handle; int max_threads; ... max_threads=256; ... thread_handle=calloc(max_threads, sizeof(*thread_handle)); if (thread_handle==NULL) return 1; There are more handles than a function like WaitForMultipleObjects can use. http://msdn.microsoft.com/library/en-us/dllproc/base/waitformultipleobjects.asp The handles may point to threads that have already done _endthreadex(0), there are never more than 64 threads in existence at the same moment. Threads are usually short lived (they wait for a UDP network answer). The maximum amount of thread handles is 64, but there may be up to 256 used. Is it possible to point to the 65th, 129th, 193th handle and run that function again? You can do this with strings because they are contiguous array of bytes, but does it work the same way with handles? May be I'm doing something stupid here, help is appreciated. Bert Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232653
  - 10
    - How to check from destructor whether it was called due to exception or normal exit from the scope? Hi, everybody Is there a reliable way of detecting whether destructor of a local object was called due to stack unwind (exception) or normal exit from the scope? This check has to be made inside destructor of this local object. If there is no portable C++ way of doing this, hacks for Windows and/or MSVC6 generated code (with checking some registers etc.) will do. Thanks in advance, Antonio Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232646
  - 11
    - How to create a timer event in VC++ which does the functionalities of VB's 'Timer control' - This should be compatible to Unix OS also. Hai, I am having a winsock applicaiton, in which I should check for messages in regular time period. This can be implemented in VB by using the timer control, what is the equivalent of 'timer control' in VC++? Can any one send me a sample application for creating a timer event in VC++ which does the functionalities of VB's timer control, or is there any other solution? The solution should be compatible with Unix OS also, as the application is going to be implemented in both Windows and Unix OS. Thanks. Uma Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232642
  - 12
    - rpc server not available What would be the most likely place that causes the return value "RPC server not available" when trying to connect to a remote machine? o I have checked DCOM settings Auth=None,Impersonate - both PC's have the same settings. Server is launched by the administrator account with correct password. Server is already running as I try to connect. o I am using the correct GUID o Both machines are in the same workgroup although I am using the IP address of the remote PC to connect in the COSERVERINFO struct (other fields in COSERVERINFO are 0) o CoCreateInstanceEx is used with CLSCTX_INPROC_SERVER|CLSCTX_LOCAL_SERVER|CLSCTX_REMOTE_SERVER tia/ak -- You are only young once, but you can stay immature indefinitely. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232632
  - 13
    - Get remote env. variable Anyone know of an API to get an enviornment variable on a remote machine? Thanks Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232626
  - 14
    - GetVolumeNameForVolumeMountPoint I am trying to use GetVolumeNameForVolumeMountPoint, but it keeps giving me an undeclared identifier error with VC 6.0. According to the MSDN I am supposed to include Windows.h and use the Kernel32.lib library, which I am doing without success. Any ideas? Thanks, Robin By the way here is the link to the msdn: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/fileio/base/getdrivetype.asp -- Using Opera's revolutionary e-mail client: http://www.opera.com/m2/ Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232617
  - 15
    - How To find out New Incoming Mail in Outlook? Hi Guys, I catch the New Mail Event in Outlook. When New mail is coming, my code gives an alert msg. But I can't find Which is the current new mail. I have an idea to get Unread mails. But It gives Today unread mail, Previous day Unread mails and New Mail. From these How can i get the New mail? Pls help me. -- Computer Ganesh Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232601
  - 16
    - Suspicious error VC 7.1 won't compile this ///////////////////////////////////////////////////////////////////// #include <map> namespace N { struct K {}; } struct T {}; // namespace N { bool operator<(const K&, const K&){ return true; } } bool operator<(const N::K&, const N::K&){ return true; } int main () { std::map<N::K, T>().insert(std::pair<N::K, T>(N::K(), T())); } ///////////////////////////////////////////////////////////////////// Uncommenting the namespaced operator will help. Sergei Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232596
  - 17
    - First Enum Default value When I open an enum structure and I don't name it with a value such as enum abc { APPLE, ORANGE, etc } What is the default value of "APPLE"? Is it 0 or 1? Thanks Jack Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232590
  - 18
    - CreateProcess I have a MFC SDI program that calls CreateProcess from a Dialog. The following is part of the code: case ID_COMPILE_CODE: { CString strFile("nmake /X 'k:\\Test\\Debug\\Output.txt' /f test.mak CFG='Test - Win32 Debug'"); CString strAPI("API=Win32"); CString strPath("e:\\microsoft visual studio\\vc98\\bin;"); strPath += "e:\\microsoft visual studio\\common\\msdev98\\bin;"; strPath += "k:\\test"; CString strLib = "e:\\microsoft visual studio\\vc98\\lib"; CString strInclude = "e:\\microsoft visual studio\\vc98\\include"; strPath = "PATH=" + strPath; strLib = "LIB=" + strLib; strInclude = "INCLUDE=" + strInclude; _putenv((LPCSTR)strPath); _putenv((LPCSTR)strLib); _putenv((LPCSTR)strInclude); _putenv((LPCSTR)strAPI); STARTUPINFO si; ZeroMemory(&si, sizeof(STARTUPINFO)); si.cb = sizeof(STARTUPINFO); PROCESS_INFORMATION pf; ZeroMemory(&pf, sizeof(PROCESS_INFORMATION)); if (CreateProcess(NULL, strFile.GetBuffer(strFile.GetLength()), NULL, NULL, TRUE, HIGH_PRIORITY_CLASS, NULL, NULL, &si, &pf)) { if (WaitForSingleObject(pf.hProcess, INFINITE) != WAIT_FAILED) SetErrorMsg(CString("No Errors")); else SetErrorMsg(CString("WaitForSingleObject(...)")); CloseHandle(pf.hThread); CloseHandle(pf.hProcess); } else SetErrorMsg(CString("CreateProcess(...)")); } return TRUE; . . . void CDlg::SetErrorMsg(CString strMes) { LPVOID lpMsgBuf; FormatMessage(FORMAT_MESSAGE_ALLOCATE_BUFFER | FORMAT_MESSAGE_FROM_SYSTEM | FORMAT_MESSAGE_IGNORE_INSERTS, NULL, GetLastError(), MAKELANGID(LANG_NEUTRAL, SUBLANG_DEFAULT), (LPTSTR)&lpMsgBuf, 0, NULL); CString strErrMsg("Error in processing " + strMes + "\n" + (LPTSTR)lpMsgBuf); AfxMessageBox(strErrMsg); LocalFree(lpMsgBuf); } I get the message that the process was successful, but there is no .obj nor .exe files in the directory. Any Ideals? Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232584
  - 19
    - about <advanced windows 3rd> chapter 5 The writer says that : In Windows NT,the highest 64KB memory of the lower 2GB address space of a process can protect the upper 2GB system space from destruction. But what I puzzled was: how can the 64KB memory do? It seems that I can use some memory lower of the 64KB ,and use the WriteProcessMemory function to get into the upper 2GB. thanks (I'm a beginner) Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232582
  - 20
    - How to flush the ostringstream buffer? Hi, I have a quetion about to refresh the ostringstream buffer: like this. ostringstream buffer; buffer << 245; // then the buffer.str() = "245"; ..... //next I want using this buffer again but I don't need the content of buffer before, so buffer.flush() buffer << 456; //here I hope I can get "456" from buffer.str() but the buffer still is 245456, seem the buffer.flush not take function? why? who can help me? Thanks adanveced, Bill Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232581
  - 21
    - Explicit specialization of class template member functions Is this supported in VS C++ 2005? For example, template <class T> class foo { void goo(void) { } }; template <> void foo::goo<int>(void) { } Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232577
  - 22
    - Remote file exists ? hi I am using IBindStatusCallBack to get files on a server. Whats the easiest way to find out if a URL points to an existing file ? thanks Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232568
  - 23
    - Bit-related question My question mentions MFC, but it is actually just a matter of bit arithmetic. If you are working on an MFC application having a frame, seeing the following statement is not uncommon: cs.style &= ~FWS_ADDTITLE; // remove extra stuff from title My question is, why is the value simply not subtracted like: cs.style -= FWS_ADDTOTITLE; The resulting style is the same for both. Is there a style (i.e., value) that could exist such that subtracting would produce a different result than ANDing the one's complement? I've tried with several starting values but both ways produce the same value. Thanks, DC Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232558
  - 24
    - clock() Is it a known issue that in VC++ RTL the function 'clock()' reports the wall clock instead of CPU time used? I ran this test program on Windows and Unix: ----------------------------------------- #include <iostream> #include <ctime> using namespace std; int main() { time_t t0 = time(0); clock_t c0 = clock(); for (int i = 0; i < 10; ++i) { cout << "push > " << flush; char a; cin >> a; cout << endl; } cout << "Difference in wall clock is " << difftime(time(0), t0) << endl; cout << "Difference in CPU ticks is " << double(clock() - c0) / CLOCKS_PER_SEC << endl; return 0; } ------------------------------------------- and got very similar values on Windows and quite different values on Unix. It seems that on Windows it just takes the time() at the beginning and then 'clock' simply takes the difference between current time and the stored value and multiplies it with CLOCKS_PER_SEC... I know of GetProcessTimes() function, but couldn't 'clock()' be implemented to use it? Thanks. Victor Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232556
  - 25
    - CDaoDatabase + Access97 Limits Does anyone if there is a limit on the number of fields or the number of tables in a CDaoDatabase database? My code suddenly won't run after I have added a few more fields in one table and I wonder if I have exceeded something? If Yes, then does anyone know any workarounds? GT Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232551
- Next
  - 1
    - Gettin a compile error error C2229: struct '__unnamed' has an illegal zero-sized array i am getting above error for this " unsigned char buffer[4096]; " at this line static int iax_send(struct iax_session *pvt, struct ast_frame *f, unsigned int ts, int seqno, int now, int transfer, int final) { /* Queue a packet for delivery on a given private structure. Use "ts" for timestamp, or calculate if ts is 0. Send immediately without retransmission or delayed, with retransmission */ struct ast_iax2_full_hdr *fh; struct ast_iax2_mini_hdr *mh; struct { struct iax_frame fr2;/* Buffer -- must preceed fr2 */ unsigned char buffer[4096]; //--------------- this line has the error } buf; struct iax_frame *fr; ............. Code continues thanks Rj Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232544
  - 2
    - External Dependencies Hi guys How i can insert files in External dependencies folder ??? Thank Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232542
  - 3
    - How to convert Format Outlook mail Received Time? Hai Guys, I have an addin for Outlook. I get the received time for incoming mails. But time is the type of DATE datatype(i.e. in double) How can I convert that into Our Normal Time and Date Format? For Examble I got the DATE as 38097.542242106 I want as 7.30 PM July 7 2002. How can I format that? Pls reply me. Thanks. -- Computer Ganesh Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232533
  - 4
    - Help in VC I need to integrate the InfoPath documents in VC++ ActiveX control and I have gone through this article (311765) in the site, http://support.microsoft.com/default.aspx?scid=kb;en-us;311765 Using above article utility (DSOFramer), I'm able to host office documents except InfoPath. I tried to modify the DSO framer code base, even though, I was getting error saying that ActiveX Server doesn't support this kind of document like that. Could you please help me, why this problem is coming for InfoPath. What are necessary changes required for the above article for hosting InfoPath documents in VB or HTML. Please give the solution/suggestions. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232531
  - 5
    - BUG Report: Failed When a Native Event is Unhooked with with Multi-Heritance Hi, i need to use the native event in a ATL/WTL C++ application. I found out there was some error in the __unhook generated code when your receiver has multi-heritance __unhook fail in these case to find the right hook to delete, and doesnt. It leaks memory and the hook isnt removed so if you __raise, you keep getting the event. To work around this problem, use the following method: native_event.cpp: #include <stdio.h> class A { int a; }; class B { int b; }; [event_source(native)] class CSource { public: __event void MyEvent(int nValue); }; [event_receiver(native)] class CReceiver : public A, public B //multi-heritance { public: void MyHandler1(int nValue) { printf("MyHandler1 was called with value %d.\n", nValue); } void MyHandler2(int nValue) { printf("MyHandler2 was called with value %d.\n", nValue); } }; int main() { CSource source; CReceiver receiver; __unhook(&CSource::MyEvent, &source, &CReceiver::MyHandler2, &receiver); __hook(&CSource::MyEvent, &source, &CReceiver::MyHandler1, &receiver); __raise source.MyEvent(1); __unhook(&CSource::MyEvent, &source, &CReceiver::MyHandler1, &receiver);//failed __hook(&CSource::MyEvent, &source, &CReceiver::MyHandler2, &receiver); __raise source.MyEvent(2); __unhook(&CSource::MyEvent, &source, &CReceiver::MyHandler2, &receiver);//failed } Compile: cl /Fx /Fas native_event.cpp Result: MyHandler1 was called with value 1. MyHandler2 was called with value 2. MyHandler1 was called with value 2. MyHandler2 can't be unhooked。 Applies to: Microsoft Visual C++ .NET (2002) Microsoft Visual C++ .NET (2003) Visual C++ 2005 Express Edition Beta solution: To solve this problem we can change native_event.mrg.cpp from int __isEqual(void* p, void* pfn) { return ((T*) p == this->p) && (__eventingGetAddr<T>::__getMAddr(pfn) == (void ( T::*) ()) pmfn); } to: int __isEqual(void* p, void* pfn) { return ((T*) p == this->p) && pfn == __eventingGetAddr<T>::__getVAddr((void (T::*) ()) pmfn)); } But, I don't know what is the best soluction. Because of this BUG, Native event can't be used in ATL/WTL. Regards. Baryon Lee Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232524
  - 6
    - ERROR_IO_PENDING Hi All, When I am trying to delete a file with DeleteFile API ( file is created with fopen() call) sometimes returns ERROR_IO_PENDING as it's GetLastError(). Any ideas on why I am getting it. Thanks in advance, Sateesh. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232523
  - 7
    - signed and unsigned Is data types of "short" and "int" for sure to be signed in Vc++ .net? How to set unsigned char/unsigned int/unsigned short in a project? Thanks! Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232522
  - 8
    - _wtof not in stdlib.h or wchar.h Hi, I am trying to convert one wchar_t data type to float, the function in MSDN is _wtof, but it is not defined in stdlib.h or in wchar.h. I am using service pack 4 of visual studio 6. Any help on this ?, as of how can I convert it to float ? Regards Alpha Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232519
  - 9
    - Template class : problem with static member initialization Microsoft C++ 13.00.9466 has problems with code below. GNU g++ 3.3.1 compiles this code with no errors. What is wrong? ------ foo.cpp : BEGIN ------ #include <string> using namespace std; template<class T> class Foo { private: static string func_(); public: static const string data_; }; template<class T> string Foo<T>::func_() { return string(); } template<class T> const string Foo<T>::data_(Foo<T>::func_()); int main () { Foo<int>::data_; return 0; } ------ foo.cpp : END -------- ------ Compilation : BEGIN ------ $ cl /EHsc yoo.cpp Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 13.00.9466 for 80x86 Copyright (C) Microsoft Corporation 1984-2001. All rights reserved. foo.cpp foo.cpp(20) : error C2061: syntax error : identifier 'func_' foo.cpp(20) : error C2059: syntax error : ')' foo.cpp(20) : error C2143: syntax error : missing ')' before '<end Parse>' foo.cpp(20) : error C2063: 'Foo<T>::data_' : not a function foo.cpp(20) : error C2373: 'Foo<T>::data_' : redefinition; different type modifiers foo.cpp(10) : see declaration of 'Foo<T>::data_' foo.cpp(25) : error C2039: 'data_' : is not a member of 'Foo<T>' with [ T=int ] foo.cpp(25) : error C2065: 'data_' : undeclared identifier ------ Compilation : END -------- -- Alex Vinokur http://mathforum.org/library/view/10978.html http://sourceforge.net/users/alexvn Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232518
  - 10
    - Linking a menu item to a dialog box Hi, Using visual C++, how can I link a menu item in project A to a dialog box in project B, without using the class wizard? Thank you, Rosie Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232516
  - 11
    - IOCompletionPort Hi, I am trying to understand the how IOCompletionPort works. Any help from the gurus out there is highly appreciated. Thanks. Regards, Soumitra Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232507
  - 12
    - machine dependent identifier I am working on a program using Visual C++ and MFC. We want to be able to set it up so that it will only work on one computer. Is there a computer or processor specific identifier that can be accessed? If so, how can I access it? Dave Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232503
  - 13
    - WinInet-FtpGetFile errors 12003 and 12030 (Note, this happens under both Windows XP (VC++6.0) or WinCE 4.2 (eVC++4.0 SP3), which is why I'm posting this question under both newsgroups). I'm using FtpGetFile() repeatedly to copy a folder structure that contains 222 files (total 27Mb), of which one file is 17Mb and the rest are smaller. The 17Mb file is one of the first (#8) to copy, and it does OK, but after 14-80 files are copied (around 18-19 Mb) I get error code 12003 from the first file that fails, then 12030 from subsequent files. Codes in the 12000 range are not included in my message list in either environment, so I don't know what this error means. Questions: 1. What do error codes 12003 and 12030 mean? (I can't find them in MSDN or CE Help). What is actually failing here? 2. Is there any way to get the WinInet error messages (I've seen other Ftp routines report 12xxx error codes) included in my XP and CE environments? I thought maybe this was a limitation of our FTP site, but our IT group says they can copy Gbs of data with no problem. I tried skipping over the 17Mb file, and the other 221 files (10Mb) copy just fine. Then, even more mysterious, I tried copying the 17Mb file last (after the other 221), and that works fine as well! Any ideas? Can I depend on the workaround of copying the biggest file last always working? Thanks, Michael Schwab Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232482
  - 14
    - Class and static member Hi, I have a class. In this class i have a static method member and some (not static) data members are used by this static method. However, when static method uses a not static data, VC++ compiler generates an error.... so i decided to recast my data member to static but only when they are used inside this static method... but it doesn't work... even if i use static_cast<bool> mydata; i would like to not define as static my data member into my class because some other methods (which are not static) use these data... thanks for help, Maileen Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232481
  - 15
    - check class inherits from 2 interfaces Hi Guys, I would be grateful if you could give me a hint. I am writing a plugin based program. i have 2 interfaces (abstract classes) IPlugin and IPluginMenu, all plugins inherit from IPlugin but plugins that need menu need to inherit from IPluginMenu and implement OnMenu which is pure virtual. Now in my SDK I have a function called SetPluginMenu and there as a parameter I want to pass a instance of class that inherits from IPluginMenu, but the problem is that I also need to make sure that caller also inherits from IPlugin, so that someone inheriting only from IPluginMenu won't be able to call SetMenu as he has to inherit from both classes. Thanks in advance Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232474
  - 16
    - Get Application path in unmanaged C? Can someone point me to the function or technique to get the path from whence the current application was launched? TIA, jack Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232465
  - 17
    - OpenFileDialog troubles... Ummm... Kindah new to VS... I learnt C++ on Codewarrior 4 palm OS... so basically all this MFC stuff and the windows API in general is still setting in... Anyways, been looking for a nice easy way to setup an open file dialog without converting my c++ app to C# or java... I found some info on "OpenFileDialog" from MSDN which seems to be part of some crazy language independant class, but I can't make heads or tails of it cause there isn't any info on which file(s) I need to #include. Also I just don't like the interface to the object... well I hate that it's an object... 0.o So i'm just wondering if there is an easy way to show the standard "Open File" dialog in c++ without adding in the static linked MFC stuff or crazy object stuff (I'm one of them anti-MFC if at all possible people), and if at all possible something that looks more like a functions that require parameters... In a few lines of code... BTW... tried MSDN (both my local and the online versions), google, and looked through all the sample source I could find that came with my VS .NET 2003, still couldn't find anything that helped :/ #include <window.h> //... //declarations and stuff //... WinMain(...) { // stuff... strcpy (strFilePath, /* This is the thingy i'm looking for */ SimpleOpenFileDialog (strDefaultPath, strDefaultFileType, dwDisplayTypeBitsOrOtherOptions) ); // stuff... } //... // more stuff //... Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232464
  - 18
    - anybody knows details about this bug? http://support.microsoft.com/?kbid=843148 Dynamic initialization does not work as expected in Visual C++ .NET 2003 -- The set of solutions is never empty. Two solutions together form a new problem. -- Mycroft Holmes Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232440
  - 19
    - One Instance What is the best way to check if the executable is already running to prevent multiple instances of one program. If you understand what I mean. Thanks in advance, Robin -- Using Opera's revolutionary e-mail client: http://www.opera.com/m2/ Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232423
  - 20
    - vector<Object*>::assign() hi i posted a simmilar question, but i lost the thread somewhere. How would i make this work : vector<SomeClass> vo; vector<SomeClass*> vpo; vpo.assign( vo.begin(), vo.end() ); I get a type conversion error at compile. Any ideas ? Thanks. Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232422
  - 21
    - Pure virtual function question Consider the following class Base { virtual int GetValue() const { return(0); } }; class Object1 : public Base { virtual int GetValue() const { return(1); } }; class ISpecialObject : public Object1 { virtual int GetValue() const = 0; }; class MySpecialObject : public ISpecialObject { virtual int GetValue() const { return(2); } }; Are there going to be any problems with forcing ISpecialObject::GetValue() to be pure virtual, even though its base class already has an implementation? Thanks Murrgon Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232420
  - 22
    - IViewObject::Draw hi Can somebody please provide a sample of something on how to "set" the 4th argument in the function : IViewObject::Draw( , , , DVTARGETDEVICE * ptd, ,, ); Basicaly i need to fill this item with enough info to allow the Com to create a 32bit buffer internaly. I looked at MSDN frominfo on DVTARGETDEVICE , but its hazy at best ... Any ideas ? thanks Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232417
  - 23
    - Compare Function in CListCtrl (ListView) Hi, I have some trouble with my Compare function to sort my Items in my ListView. i tried to apply the MS example but without success. in my *.h file : in my class / in public : static int CALLBACK CompareFunction(LPARAM lParam1, LPARAM lParam2, LPARAM lParamSort) in my *.cpp file : //in my OnColumnClick method pListCtrl->SortItems(CompareFunction, (LPARAM) pListCtrl); int CALLBACK CompareFunction(LPARAM lParam1, LPARAM lParam2, LPARAM lParamSort) { // lParamSort contains a pointer to the list view control. CListCtrl* pListCtrl = (CListCtrl*) lParamSort; // sort on 1st column for this test CString strItem1 = pListCtrl->GetItemText(lParam1, 0); CString strItem2 = pListCtrl->GetItemText(lParam2, 0); return strcmp(strItem2, strItem1); } but it doesn't work..nothing is sorted :( every time just the first item of the first column is processed :( could you help me please ? thx, maileen Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232416
  - 24
    - Process' User Name Hi, Is there an API to get the User a process is running under? I can get the process ids using the ToolHelp APIs but there is no mention of the user Vin Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232413
  - 25
    - Disconnecting Mapped Drives When the user Ends Session I want to be able to disconnect all Mapped drives. I am using the WM_QUERYENDSESSION message to run a shellex to do the following net use * /d /y But the command never runs and the mapped drives stay connected. I thought that I could stall the shutdown/restart run the command then continue with the Shutdown/Restart. However I am unable to determine what the original request was (Shut down or restart). Everything works except what I just mentioned. Is there a way to disconnect mapped drives at End Session that I am over looking? Thanks in advance, Robin -- Using Opera's revolutionary e-mail client: http://www.opera.com/m2/ Tag: richedit(?1) + windows 98 + em_replacesel = codepage problem Tag: 232408

I have richedit (control) with some russian text in it.
Then I make a selection - SendMessage(Handle, EM_EXSETSEL, 0, ...));
And then I replace a selection - SendMessage(Handle, EM_REPLACESEL, 0,
buffer) with some russian text.
The result is OK on windows2000 and NT, but on windows 98(russian edition)
inserted text is in wrong codepage (appeared as unreadable abracadabra).
so help me please.

--
--- With best wishes, Kirill.
* ...chasing dreams...

Top