I'm struggling with exception handling inside my Datasnap REST Service (Delphi XE3 but also tried with Delphi 10 Seattle). I have written half a dozen Windows Service over the years and I always include a TApplicationEvents component so that I can log any Application Exceptions to the Windows Event Log.
However, this behaviour is not happening with a Datasnap Service. The TApplicationEvents.OnException event never gets fired, so I assume something else is eating the exception and handling it before it gets here.
The exception is displayed in the web service method's result, which is fine because it means I can at least display something on the client side, but I'd also like to catch it before then so that I might be able to handle different exceptions server-side.
The only consistent way I have managed so far is to wrap each individual method in a try..except block, and handle the exception in each method, before re-raising the exception. However, with a Web Service of 20 methods and growing, this isn't really going to scale up.
I have also tried implementing the OnError, OnTrace and other events of the some of the Datasnap components (TDSServer, TDSHTTPService, TDSTCPServerTransport, etc.) but these never seem to get fired either.
Has anyone come across anything like this, please?
Tl;Dr: it's not implemented in a usable fashion (in 10.1 Berlin).
I came across the same problem and after reading through a lot of source, I found no practical solution.
So an exemplary (my) StackTrace would look like this:
MyClass::MyServerMethod()
/* skipping some funny unimportant RTTI/TValue handling here */
System::Rtti::TRttiMethod::Invoke
Dsreflect::TDSMethod::Invoke(TObject, TDSMethodValues)
TDSServerConnectionHandler::DbxExecute(const TDBXExecuteMessage)
TDSServerCommand::DerivedExecuteUpdate
TExecuteCallback
TDSService::Execute(const string, const TRequestCommandHandler, TExecuteCallback)
TDSService::ProcessRequest(const string, const TRequestCommandHandler, TExecuteCallback)
TDSRESTService::ProcessREST(const string, const string, const TArray<Byte>, const TRequestCommandHandler)
TDSRESTService::ProcessGETRequest(const string, TStrings, TArray<Byte>, TRequestCommandHandler)
TDSRESTServer::DoDSRESTCommand(TDSHTTPRequest, TDSHTTPResponse, string)
TDSRESTServer::DoCommand(TDSHTTPContext, TDSHTTPRequest, TDSHTTPResponse)
Dshttpwebbroker::TDSRESTWebDispatcher::DispatchRequest(TObject, Web::Httpapp::TWebRequest, Web::Httpapp::TWebResponse)
Note: This depends entirely on your usage of DataSnap. In the above case requests are passed into the DataSnap API through TDSRESTWebDispatcher (comming from TIdCustomHTTPServer).
Every Exception raised in a ServerMethod will end up in TDSService::ProcessRequest.
In this procedure every Exception is caught and ONLY its Message is added to a TRequestCommandHandler->CommandList.
Further down the Message is written as JSON/DBX command to the output.
So we can never handle the Exception Object and access the StackTrace or other information. So this alone is unacceptable and has to change
The good news is, that this procedure is virtual and can be overwritten. The bad news is, that in the example above you would have to extend TDSRESTService with your own ProcessRequest procedure (including your errorhandler), TDSRESTServer with own DoDSRESTCommand (in there the TDSRESTService is created in a monstrously large procedure) and TDSRESTWebDispatcher (depending on your usage).
My personal recommendation is to don't use DataSnap.
Note: At the point of writing this, I haven't found any invoke of the OnError event.
Related
I am using C++ Builder XE7 VCL.
Around August 11 2016 2:00pm UTC, I started receiving multiple complaints from my user base about printing problems. Most of these printing modules have proven stable for many years, and there were no updates to my project within the past 24 hours. I was able to reproduce similar problems on my development/test environment.
Without going into many details of my project, let me present a very simple printing program that is failing:
void __fastcall TForm1::PrintButtonClick(TObject *Sender)
{
// Test Print:
TPrinter *Prntr = Printer();
Prntr->Title = "Test_";
Prntr->BeginDoc();
Prntr->Canvas->Font->Size = 10;
Prntr->Canvas->TextOut(300,1050,"* * * Printing Test * * *");
if (Prntr->Printing) {
Prntr->EndDoc();
}
}
On the first attempt to print, everything works perfectly as expected. If I click the button a second time, TPrinter produces a small PDF, but the PDF file is actually corrupted and appears to have a file handle stuck to it.
If I click the button a third time, I get no printing and the following error message appears:
Printer is not currently printing.
My own test was done using a PDF printer driver, but the complaints I am receiving from users include a variety of local printers, network printers, PDF printers, etc.
In my actual project, I have try/catch exception handling, so the actual results are slightly different, but substantially similar to this result. The results show the hallmarks of instability and/or memory leaks without much in terms of error messages.
I suspect there may have been some Microsoft Windows updates that are tangling with Embarcadero DLLs, but I have not been able to verify this so far.
Is anyone else having similar problems?
The reason using a TPrintDialog or TPrinterSetupDialog "works" to fix the error is because they force the singleton TPrinter object (returned by the Vcl.Printers.Printer() function) to release its current handle to a printer if it has one, thus causing TPrinter.BeginDoc() to create a new handle. TPrinter releases its printer handle when:
it is being destroyed.
its NumCopies, Orientation, or PrinterIndex property is set.
its SetPrinter() method is called (internally by the PrinterIndex property setter and SetToDefaultPrinter() method, and by TPrintDialog and TPrinterSetupDialog).
Without doing that, calling TPrinter.BeginDoc() multiple times will just keep re-using the same printer handle. And apparently something about the recent Microsoft security update has now affected that handle reuse.
So, in short, (without uninstalling the Microsoft update) in between calls to BeginDoc() you need to do something that causes TPrinter to release and recreate its printer handle, and then the problem should go away. At least until Embarcadero can release a patch to TPrinter to address this issue. Maybe they could update TPrinter.EndDoc() or TPrinter.Refresh() to release the current printer handle (they currently do not).
Therefore, the following workaround resolves the printing issue without requiring any changes to the user interface:
void __fastcall TForm1::PrintButtonClick(TObject *Sender)
{
// Test Print:
TPrinter *Prntr = Printer();
Prntr->Title = "Test_";
Prntr->Copies = 1; // Here is the workaround
Prntr->BeginDoc();
if (Prntr->Printing) {
Prntr->Canvas->Font->Size = 10;
Prntr->Canvas->TextOut(300,1050,"* * * Printing Test * * *");
Prntr->EndDoc();
}
}
There are no Embarcadero DLLs involved in printing. TPrinter simply calls Win32 API GDI-based print functions directly.
The "Printer is not currently printing" error occurs when one of the following operations is performed on a TPrinter when its Printing property is false:
TPrinter::NewPage()
TPrinter::EndDoc()
TPrinter::Abort()
a TPrinter::Canvas subproperty is being changed.
the TPrinter::Canvas is being drawn on.
You are performing half of these operations in the test code you have shown, but you did not specify which line of code is actually throwing the error.
The Printing property simply returns the current value of the TPrinter::FPrinting data member, which is set to false only when:
the TPrinter object is initially created (the Printer() function returns a singleton object that is reused for the lifetime of the executable).
the Win32 API StartDoc() function fails inside of TPrinter::BeginDoc() (FPrinting is set to true before StartDoc() is called).
TPrinter::EndDoc() is called when Printing is true.
So, given the test code you have shown, there are two possibilities:
StartDoc() fails and you are not checking for that condition. BeginDoc() will not throw an error (VCL bug?!?), it will simply exit normally but Printing will be false. Add a check for that:
Prntr->BeginDoc();
if (Prntr->Printing) { // <-- here
Prntr->Canvas->Font->Size = 10;
Prntr->Canvas->TextOut(300,1050,"* * * Printing Test * * *");
Prntr->EndDoc();
}
the Printing property is getting set to false prematurely while you are in the process of printing something. The only ways that could happen in the code shown are if:
random memory is being corrupted, and TPrinter happens to be the victim.
multiple threads are manipulating the same TPrinter object at the same time. TPrinter is not thread-safe.
Since you can reproduce the problem in your development system, I suggest you enable Debug DCUs in the project options, then run your app in the debugger, and put a data breakpoint on the TPrinter::FPrinting data member. The breakpoint will be hit when FPrinting changes value, and you will be able to look at the call stack to see exactly which code is making that change.
Based on this information, I am going to go out on a limb and guess that the cause of your error is StartDoc() failing. Unfortunately, StartDoc() is not documented as returning why it fails. You certainly cannot use GetLastError() for that (most GDI errors are not reported by GetLastError()). You might be able to use the Win32 API Escape() or ExtEscape() function to retrieve an error code from the print driver itself (use TPrinter::Canvas::Handle as the HDC to query). But if that does not work, you won't be able to determine the reason of the failure, unless Windows is reporting an error message in its Event Log.
If StartDoc() really is failing, it is because of an Win32 API failure, not a VCL failure. Most likely the printer driver itself is failing internally (especially if a PDF print driver is leaving an open file handle to its PDF file), or Windows is failing to communicate with the driver correctly. Either way, it is outside of the VCL. This is consistent with the fact that the error started happening without you making any changes to your app. A Windows Update likely caused a breaking change in the print driver.
Try removing the following Windows update:
Security Update for Microsoft Windows (KB3177725)
MS16-098: Description of the security update for Windows kernel-mode drivers: August 9, 2016
This appears to resolve the issue for several test cases so far.
It started happening here today too. On my Windows 10 setup this would happen after calling TForm's Print() 3 times. I tried both the Microsoft Print to PDF and the Microsoft XPS Document Writer and both gave the same error.
I did some quick debugging and found that it's the call to StartDoc() that returns a value <= 0.
A temp fix until I can figure out what is really causing this is to re-create the Printer object in Printers by calling
Vcl.Printers.SetPrinter(TPrinter.Create).Free;
after calling anything that is using the Printer object. Might not be advisable to do that, but it solved my issue for now.
Looks like something is not released properly when calling EndDoc()
It seems that this is really a Microsoft problem and they should fix this buggy update. You can find more info at this on the company site.
Using a Printer Setup Dialog is only a workaround, not real solution, for this Microsoft bug. After confirmation, the printer setup dialog always creates a new handle for the printer. The next one or two print jobs will then succeed.
The patch should come from Microsoft, not from Embracadero. Thousands of programs are affected and it would be a massive waste of time and money to implement a workaround in all of them, if there is a bug in MS update.
I started experiencing the same weird behavior at more or less the same time, when running 32 bit Delphi applications built in XE7 on a Windows 10 64 bit system.
After uninstalling the latest security upgrade for Windows 10 (KB3176493), printing from these applications works normally again.
A rather surprising side effect of uninstalling this update seems to be that the file associations - which are the default programs for handling specific file types - are being reverted to Microsoft Windows default values...
The following variation of the code in the question will resolve the problem, but requires a TPrinterSetupDialog component added to the form:
void __fastcall TForm1::PrintButtonClick(TObject *Sender)
{
// Test Print:
TPrinter *Prntr = Printer();
Prntr->Title = "Test_";
PrinterSetupDialog1->Execute();
Prntr->BeginDoc();
if (Prntr->Printing) {
Prntr->Canvas->Font->Size = 10;
Prntr->Canvas->TextOut(300,1050,"* * * Printing Test * * *");
Prntr->EndDoc();
}
}
For program usage, the user will be presented with the printer setup dialog before proceeding to printing.
As to "why", my best guess at this point is that TPrinter used alone does not have all necessary permissions to access all necessary resources from Windows after the Security Update for Microsoft Windows (KB3177725) was implemented on Aug 10, 2016. Somehow a call to TPrinterSetupDialog (or TPrintDialog) before calling BeginDoc() sets up the necessary conditions for TPrinter to perform successfully.
I wrote my routines for using CDO.Message. It's working long time ago.
But now, in some site they installed a certificate and then the priorly working version is making errors with all calls.
I used simple variants to hold the CDO.Message COM object.
When I used Send method(?) it returns a HResult.
But it's interesting, because the HResult is unusable for get the error code, because the Send is seems to be a real method which makes Exception on problems.
So the result code is -1 if I set to this value before.
I tried to get the last error code with GetLastError. But this is 0.
I can catch the Exception, but it contains only the error message what is:
"The transport failed to connect to the server"
The VB codes can get the error code which could provide extra information about the problem (or not).
Do you know about a technic to get the error code value from Delphi XE3?
It would be better if we had some source code at hand. Specifically the Delphi declaration of the COM interface.
My guess is that the method is declared as safecall. What this means is that the compiler understands that the method is actually stdcall returning HRESULT, and re-writes the parameters to match. If the true COM method returns an HRESULT other than S_OK then the compiler writes code to check for that and convert the error into an exception.
The exception that is raised will be EOleSysError and that has the property ErrorCode which contains the HRESULT that you are looking for.
So, you need to:
Add an exception handler to catch EOleSysError.
Read the ErrorCode property of the EOleSysError exception instance that you catch.
This is all a little bit round the houses. If you'd prefer to avoid exception handling, then you can always re-write the COM interface declaration to be a true stdcall method returning an HRESULT.
I am developing this Delhi 2009 app with SQLite database version 3.7.9, accessed by the Zeos components. Started off OK - I can create a DB if none exists, create tables and insert records - the 'normal stuff'. There is a little table that contains serial params for a connection to some old equipments and the serial comms handler needs to be informed of connection parameter changes, but the Zeos components do not support events on SQLite.
I looked at the SQLite API and there is a 'Data change notification callback', so I set about making this happen. The setup prototype is:
void *sqlite3_update_hook(
sqlite3*,
void(*)(void *,int ,char const *,char const *,sqlite3_int64),
void*
);
So, in Delphi, I imported the DLL call:
function sqlite3_update_hook(pDB:pointer;pCallback:pointer;aux:pointer): pointer; stdcall;
..
function sqlite3_update_hook; external 'sqlite3.dll' name 'sqlite3_update_hook';
..and declared an empty callback for testing:
function DBcallback(aux:pointer;CBtype:integer;CBdatabaseName:PChar;CBtableName:PChar;CBrowID:Int64):pointer; stdcall;
begin
end;
..and called the setup:
sqlite3_update_hook(SQLiteHandle,#DBcallback,dmOilmon);
SQLiteHandle is the sqlite3* pointer for the database as retrieved from the driver after connection. dmOilmon is the DataModule instance so that, when working correctly, I can call some method from the callback, (yes, I know callback is in an unknown thread - I can deal with that OK, I'll just be signaling semaphores).
Good news: when I ran the app, the DBcallback was called, aparrently successfully, upon the first insert. Bad news: some little time later, the app blew up with 'Too many exceptions', and usually a CPU window full of '??', (the occasional alternative was system death and a reboot - Vista Ultimate 64). Breaking in the callback, aux, CBtype and CBrowID were all as expected but the debugger tooltip showed the CBdatabaseName / CBtableName PChars to be pointing at a string of Chinese characters..
I tried to trace where the callback is called from - the call chain passes through in the Zeos driver code, for some reason. Breaking there and stepping through, I checked the stack pointer and it's the same before and after the callback.
So, I set up the 'empty' callback, the callback gets called at the expected point but with a coupe of dodgy-looking parameters, the callback returns to whence it came with the SP correct. I seem to have done everything right, but... :((
Has anyone seen this, (or fixed it even:)?
Can you suggest a mechanism whereby an aparrently successful callback can pesuade an app to later generate recursive exceptions?
Does anyone have any suggestions for further debugging?
Rgds,
Martin
You have the problems with correct translation of SQLite header to Delphi and with linking:
Not stdcall, but must be cdecl.
Your function DBcallback must be a procedure. Because void(*)(...) is a pointer to C function returning void.
Not PChar, but must be PAnsiChar. For non-Unicode Delphi's it may be not important, for Unicode Delphi's, it is only a correct translation.
You are using compile-time dynamic linking (external 'sqlite3.dll' ...) to sqlite3.dll. But ZeosLib uses run-time dynamic linking (LoadLibrary & GetProcAddress). That may lead to "DLL hell" issue, when your EXE loads one sqlite3.dll, and ZeosLib plans to load another sqlite3.dll.
Solution found, see my comment below
D5, odbc to mysql database
This code:
with QryCmdPerf do begin
Close;
ParamByName('ACCTID').AsInteger:= AcctId;
ParamByName('FROMDT').AsString:= MySQLDate(FromDt);
ParamByName('TODT').AsString:= MySQLDate(ToDt);
Open;
first;
try
edit;
FieldByName('PnL').AsFloat:= 97979;
ApplyUpdates;
except
close;
end;
end; // with
(specifically the "ApplyUpdates") causes a popup to appear with the text "Update Failed" if the PnL field already has the value 97979, evidently because of this code:
procedure TUpdateSQL.ExecSQL(UpdateKind: TUpdateKind);
begin
with Query[UpdateKind] do
begin
Prepare;
ExecSQL;
if RowsAffected <> 1 then DatabaseError(SUpdateFailed);
end;
end;
in DBTables.pas. Anyway, I want to be able to issue ApplyUpdates, and not have to worry about a popup if it doesn't do any updating. But if "try...except" doesn't work, what will?
TIA
You're confusing the dialog displayed by the debugger with a dialog displayed by your program. Please see this article I wrote a few years ago:
Why do I continue getting error messages even after I have written an exception handler?
It describes several ways to avoid the debugger interfering:
Use "advanced breakpoints" to temporarily disable the debugger around the code that throws exceptions.
Configure the debugger to ignore certain exception types. (Read the debugger's message more carefully to see exactly what exception class you're dealing with.)
Configure the debugger not to interrupt on any exceptions.
Turn off integrated debugger entirely.
The short answer is, you have to set up an eventhandler for OnUpdateError or no amount of "try...except" blocks will block the popup. The long answer is it appears to be a bug with odbc. The repro is here: http://www.codeupload.com/3919 for anyone who wants to take a look at it. You can skip the MySQL stuff, any odbc database will do.
There are two things that can be going wrong here.
Option 1
First, some "very bad code" may be short-circuiting the unwinding of the call stack on exceptions. E.g. ApplyUpdates or one of its child routines may also have a try...except block that calls Application.HandleException directly.
To test this, if you put a breakpoint on QryCmdPerf.Close, do you reach it?
If not, then Application.HandleException (or worse Application.ShowException) has been called directly.
Solving this requires a custom exception handler hooked to the Application.OnException event. You may have to set temporary state to know when this particular exception can be ignored.
Yes it's messy, that why calling Application.HandleException directly, is "very bad code".
Option 2
If you do reach the breakpoint, but the exception is being raised again, then it should be a lot simpler to solve.
The Close method is probably attempting to save any pending changes, so is effectively applying the updates again. Rather than simply closing the data set, call CancelChanges or equivalent.
I'm a member in a team that use Delphi 2007 for a larger application and we suspect heap corruption because sometimes there are strange bugs that have no other explanation.
I believe that the Rangechecking option for the compiler is only for arrays. I want a tool that give an exception or log when there is a write on a memory address that is not allocated by the application.
Regards
EDIT: The error is of type:
Error: Access violation at address 00404E78 in module 'BoatLogisticsAMCAttracsServer.exe'. Read of address FFFFFFDD
EDIT2: Thanks for all suggestions. Unfortunately I think that the solution is deeper than that. We use a patched version of Bold for Delphi as we own the source. Probably there are some errors introduced in the Bold framwork. Yes we have a log with callstacks that are handled by JCL and also trace messages. So a callstack with the exception can lock like this:
20091210 16:02:29 (2356) [EXCEPTION] Raised EBold: Failed to derive ServerSession.mayDropSession: Boolean
OCL expression: not active and not idle and timeout and (ApplicationKernel.allinstances->first.CurrentSession <> self)
Error: Access violation at address 00404E78 in module 'BoatLogisticsAMCAttracsServer.exe'. Read of address FFFFFFDD. At Location BoldSystem.TBoldMember.CalculateDerivedMemberWithExpression (BoldSystem.pas:4016)
Inner Exception Raised EBold: Failed to derive ServerSession.mayDropSession: Boolean
OCL expression: not active and not idle and timeout and (ApplicationKernel.allinstances->first.CurrentSession <> self)
Error: Access violation at address 00404E78 in module 'BoatLogisticsAMCAttracsServer.exe'. Read of address FFFFFFDD. At Location BoldSystem.TBoldMember.CalculateDerivedMemberWithExpression (BoldSystem.pas:4016)
Inner Exception Call Stack:
[00] System.TObject.InheritsFrom (sys\system.pas:9237)
Call Stack:
[00] BoldSystem.TBoldMember.CalculateDerivedMemberWithExpression (BoldSystem.pas:4016)
[01] BoldSystem.TBoldMember.DeriveMember (BoldSystem.pas:3846)
[02] BoldSystem.TBoldMemberDeriver.DoDeriveAndSubscribe (BoldSystem.pas:7491)
[03] BoldDeriver.TBoldAbstractDeriver.DeriveAndSubscribe (BoldDeriver.pas:180)
[04] BoldDeriver.TBoldAbstractDeriver.SetDeriverState (BoldDeriver.pas:262)
[05] BoldDeriver.TBoldAbstractDeriver.Derive (BoldDeriver.pas:117)
[06] BoldDeriver.TBoldAbstractDeriver.EnsureCurrent (BoldDeriver.pas:196)
[07] BoldSystem.TBoldMember.EnsureContentsCurrent (BoldSystem.pas:4245)
[08] BoldSystem.TBoldAttribute.EnsureNotNull (BoldSystem.pas:4813)
[09] BoldAttributes.TBABoolean.GetAsBoolean (BoldAttributes.pas:3069)
[10] BusinessClasses.TLogonSession._GetMayDropSession (code\BusinessClasses.pas:31854)
[11] DMAttracsTimers.TAttracsTimerDataModule.RemoveDanglingLogonSessions (code\DMAttracsTimers.pas:237)
[12] DMAttracsTimers.TAttracsTimerDataModule.UpdateServerTimeOnTimerTrig (code\DMAttracsTimers.pas:482)
[13] DMAttracsTimers.TAttracsTimerDataModule.TimerKernelWork (code\DMAttracsTimers.pas:551)
[14] DMAttracsTimers.TAttracsTimerDataModule.AttracsTimerTimer (code\DMAttracsTimers.pas:600)
[15] ExtCtrls.TTimer.Timer (ExtCtrls.pas:2281)
[16] Classes.StdWndProc (common\Classes.pas:11583)
The inner exception part is the callstack at the moment an exception is reraised.
EDIT3: The theory right now is that the Virtual Memory Table (VMT) is somehow broken. When this happen there is no indication of it. Only when a method is called an exception is raised (ALWAYS on address FFFFFFDD, -35 decimal) but then it is too late. You don't know the real cause for the error. Any hint of how to catch a bug like this is really appreciated!!! We have tried with SafeMM, but the problem is that the memory consumption is too high even when the 3 GB flag is used. So now I try to give a bounty to the SO community :)
EDIT4: One hint is that according the log there is often (or even always) another exception before this. It can be for example optimistic locking in the database. We have tried to raise exceptions by force but in test environment it just works fine.
EDIT5: Story continues... I did a search on the logs for the last 30 days now. The result:
"Read of address FFFFFFDB" 0
"Read of address FFFFFFDC" 24
"Read of address FFFFFFDD" 270
"Read of address FFFFFFDE" 22
"Read of address FFFFFFDF" 7
"Read of address FFFFFFE0" 20
"Read of address FFFFFFE1" 0
So the current theory is that an enum (there is a lots in Bold) overwrite a pointer. I got 5 hits with different address above. It could mean that the enum holds 5 values where the second one is most used. If there is an exception a rollback should occur for the database and Boldobjects should be destroyed. Maybe there is a chance that not everything is destroyed and a enum still can write to an address location. If this is true maybe it is possible to search the code by a regexpr for an enum with 5 values ?
EDIT6: To summarize, no there is no solution to the problem yet. I realize that I may mislead you a bit with the callstack. Yes there are a timer in that but there are other callstacks without a timer. Sorry for that. But there are 2 common factors.
An exception with Read of address FFFFFFxx.
Top of callstack is System.TObject.InheritsFrom (sys\system.pas:9237)
This convince me that VilleK best describe the problem.
I'm also convinced that the problem is somewhere in the Bold framework.
But the BIG question is, how can problems like this be solved ?
It is not enough to have an Assert like VilleK suggest as the damage has already happened and the callstack is gone at that moment. So to describe my view of what may cause the error:
Somewhere a pointer is assigned a bad value 1, but it can be also 0, 2, 3 etc.
An object is assigned to that pointer.
There is method call in the objects baseclass. This cause method TObject.InheritsForm to be called and an exception appear on address FFFFFFDD.
Those 3 events can be together in the code but they may also be used much later. I think this is true for the last method call.
EDIT7: We work closely with the the author of Bold Jan Norden and he recently found a bug in the OCL-evaluator in Bold framework. When this was fixed these kinds of exceptions decreased a lot but they still occasionally come. But it is a big relief that this is almost solved.
You write that you want there to be an exception if
there is a write on a memory address that is not allocated by the application
but that happens anyway, both the hardware and the OS make sure of that.
If you mean you want to check for invalid memory writes in your application's allocated address range, then there is only so much you can do. You should use FastMM4, and use it with its most verbose and paranoid settings in debug mode of your application. This will catch a lot of invalid writes, accesses to already released memory and such, but it can't catch everything. Consider a dangling pointer that points to another writeable memory location (like the middle of a large string or array of float values) - writing to it will succeed, and it will trash other data, but there's no way for the memory manager to catch such access.
I don't have a solution but there are some clues about that particular error message.
System.TObject.InheritsFrom subtracts the constant vmtParent from the Self-pointer (the class) to get pointer to the adress of the parent class.
In Delphi 2007 vmtParent is defined:
vmtParent = -36;
So the error $FFFFFFDD (-35) sounds like the class pointer is 1 in this case.
Here is a test case to reproduce it:
procedure TForm1.FormCreate(Sender: TObject);
var
I : integer;
O : tobject;
begin
I := 1;
O := #I;
O.InheritsFrom(TObject);
end;
I've tried it in Delphi 2010 and get 'Read of address FFFFFFD1' because the vmtParent is different between Delphi versions.
The problem is that this happens deep inside the Bold framework so you may have trouble guarding against it in your application code.
You can try this on your objects that are used in the DMAttracsTimers-code (which I assume is your application code):
Assert(Integer(Obj.ClassType)<>1,'Corrupt vmt');
It sounds like you have memory corruption of object instance data.
The VMT itself isn't getting corrupted, FWIW: the VMT is (normally) stored in the executable and the pages that map to it are read-only. Rather, as VilleK says, it looks like the first field of the instance data in your case got overwritten with a 32-bit integer with value 1. This is easy enough to verify: check the instance data of the object whose method call failed, and verify that the first dword is 00000001.
If it is indeed the VMT pointer in the instance data that is being corrupted, here's how I'd find the code that corrupts it:
Make sure there is an automated way to reproduce the issue that doesn't require user input. The issue may be only reproducible on a single machine without reboots between reproductions owing to how Windows may choose to lay out memory.
Reproduce the issue and note the address of the instance data whose memory is corrupted.
Rerun and check the second reproduction: make sure that the address of the instance data that was corrupted in the second run is the same as the address from the first run.
Now, step into a third run, put a 4-byte data breakpoint on the section of memory indicated by the previous two runs. The point is to break on every modification to this memory. At least one break should be the TObject.InitInstance call which fills in the VMT pointer; there may be others related to instance construction, such as in the memory allocator; and in the worst case, the relevant instance data may have been recycled memory from previous instances. To cut down on the amount of stepping needed, make the data breakpoint log the call stack, but not actually break. By checking the call stacks after the virtual call fails, you should be able to find the bad write.
mghie is right of course. (fastmm4 calls the flag fulldebugmode or something like that).
Note that that works usually with barriers just before and after an heap allocation that are regularly checked (on every heapmgr access?).
This has two consequences:
the place where fastmm detects the error might deviate from the spot where it happens
a total random write (not overflow of existing allocation) might not be detected.
So here are some other things to think about:
enable runtime checking
review all your compiler's warnings.
Try to compile with a different delphi version or FPC. Other compilers/rtls/heapmanagers have different layouts, and that could lead to the error being caught easier.
If that all yields nothing, try to simplify the application till it goes away. Then investigate the most recent commented/ifdefed parts.
The first thing I would do is add MadExcept to your application and get a stack traceback that prints out the exact calling tree, which will give you some idea what is going on here. Instead of a random exception and a binary/hex memory address, you need to see a calling tree, with the values of all parameters and local variables from the stack.
If I suspect memory corruption in a structure that is key to my application, I will often write extra code to make tracking this bug possible.
For example, in memory structures (class or record types) can be arranged to have a Magic1:Word at the beginning and a Magic2:Word at the end of each record in memory. An integrity check function can check the integrity of those structures by looking to see for each record Magic1 and Magic2 have not been changed from what they were set to in the constructor. The Destructor would change Magic1 and Magic2 to other values such as $FFFF.
I also would consider adding trace-logging to my application. Trace logging in delphi applications often starts with me declaring a TraceForm form, with a TMemo on there, and the TraceForm.Trace(msg:String) function starts out as "Memo1.Lines.Add(msg)". As my application matures, the trace logging facilities are the way I watch running applications for overall patterns in their behaviour, and misbehaviour. Then, when a "random" crash or memory corruption with "no explanation" happens, I have a trace log to go back through and see what has lead to this particular case.
Sometimes it is not memory corruption but simple basic errors (I forgot to check if X is assigned, then I go dereference it: X.DoSomething(...) that assumes X is assigned, but it isn't.
I Noticed that a timer is in the stack trace.
I have seen a lot of strange errors where the cause was the timer event is fired after the form i free'ed.
The reason is that a timer event cound be put on the message que, and noge get processed brfor the destruction of other components.
One way around that problem is disabling the timer as the first entry in the destroy of the form. After disabling the time call Application.processMessages, so any timer events is processed before destroying the components.
Another way is checking if the form is destroying in the timerevent. (csDestroying in componentstate).
Can you post the sourcecode of this procedure?
BoldSystem.TBoldMember.CalculateDerivedMemberWithExpression
(BoldSystem.pas:4016)
So we can see what's happening on line 4016.
And also the CPU view of this function?
(just set a breakpoint on line 4016 of this procedure and run. And copy+paste the CPU view contents if you hit the breakpoint). So we can see which CPU instruction is at address 00404E78.
Could there be a problem with re-entrant code?
Try putting some guard code around the TTimer event handler code:
procedure TAttracsTimerDataModule.AttracsTimerTimer(ASender: TObject);
begin
if FInTimer then
begin
// Let us know there is a problem or log it to a file, or something.
// Even throw an exception
OutputDebugString('Timer called re-entrantly!');
Exit; //======>
end;
FInTimer := True;
try
// method contents
finally
FInTimer := False;
end;
end;
N#
I think there is another possibility: the timer is fired to check if there are "Dangling Logon Sessions". Then, a call is done on a TLogonSession object to check if it may be dropped (_GetMayDropSession), right? But what if the object is destroyed already? Maybe due to thread safety issues or just a .Free call and not a FreeAndNil call (so a variable is still <> nil) etc etc. In the mean time, other objects are created so the memory gets reused. If you try to acces the variable some time later, you can/will get random errors...
An example:
procedure TForm11.Button1Click(Sender: TObject);
var
c: TComponent;
i: Integer;
p: pointer;
begin
//create
c := TComponent.Create(nil);
//get size and memory
i := c.InstanceSize;
p := Pointer(c);
//destroy component
c.Free;
//this call will succeed, object is gone, but memory still "valid"
c.InheritsFrom(TObject);
//overwrite memory
FillChar(p, i, 1);
//CRASH!
c.InheritsFrom(TObject);
end;
Access violation at address 004619D9 in module 'Project10.exe'. Read of address 01010101.
Isn't the problem that "_GetMayDropSession" is referencing a freed session variable?
I have seen this kind of errors before, in TMS where objects were freed and referenced in an onchange etc (only in some situations it gave errors, very difficult/impossible to reproduce, is fixed now by TMS :-) ). Also with RemObjects sessions I got something similar (due to bad programming bug by myself).
I would try to add a dummy variable to the session class and check for it's value:
public variable iMagicNumber: integer;
constructor create: iMagicNumber := 1234567;
destructor destroy: iMagicNumber := -1;
"other procedures": assert(iMagicNumber = 1234567)