Trapping an exception in the kernel

[Andrew Chalk]

My app. sees an exception from the kernel at an obscure address, 0x7c59ba9d.
When running under the VC++ 6.0 debugger, I can trap this each time it
occurs. If I want to trap it in my program and just tell the program to just
continue (i.e. not pass the exception up the chain) how do I do this?

Any idea how I can find out what the cause of the exception is? In the call
stack my program doesn't appear. MSCORWKS (which I think is .NET, the
framework used by the app. that calls my DLL) and KERNEL32 do.

Many thanks for any help that you can give.

[Ivan Brugiolo [MSFT]]

The address 0x7c59ba9d does not appear to be a kernel address.
It's a user mode address.

You have not told us which kind of exception this is:
one ACCESS_VIOLATION ?
a CLR Exception ?

on average, the way to "continue" the exceptio is something along these
lines.
The ExceptionFilter may be incorrect for your case,
since you have not provided enough information,
and we even don't knonw if the exception is continuable.

DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
CONTEXT * pContext = ExceptionPointers->ContextRecord;
if (0x7c59ba9d == (ULONG_PTR)Context->eip){
return EXCEPTION_CONTINUE_EXECUTION;
} else {
return EXCEPTION_CONTINUE_SEARCH.
}
}

Function(){
__try {
// your code here
} __except (ExceptionFilter(GetExceptionInformation())) {
// nothing
}
}

--
This posting is provided "AS IS" with no warranties, and confers no rights.
Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:eA**************@TK2MSFTNGP10.phx.gbl...

My app. sees an exception from the kernel at an obscure address, 0x7c59ba9d. When running under the VC++ 6.0 debugger, I can trap this each time it
occurs. If I want to trap it in my program and just tell the program to just continue (i.e. not pass the exception up the chain) how do I do this?

Any idea how I can find out what the cause of the exception is? In the call stack my program doesn't appear. MSCORWKS (which I think is .NET, the
framework used by the app. that calls my DLL) and KERNEL32 do.

Many thanks for any help that you can give.

[Andrew Chalk]

Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call
stack:

KERNEL32! 7c59ba9d()
MSCORWKS! 792206a8()
MSCORWKS! 7922063e()
MSCORWKS! 792205f1()
MSCORWKS! 792205d1()
MSCORWKS! 792a85fc()
MSCORWKS! 792a8ef9()
MSCORWKS! 79248678()
03f21d1e()
056ad957()
055d8422()
0474a865()
0474a5a7()
MSCORWKS! 791ece8a()
MSCORWKS! 791eb8a0()
MSCORWKS! 791f3941()
MSCORWKS! 791f38ff()
MSCORWKS! 792cc5f8()
MSCORWKS! 792cc6be()
KERNEL32! 7c57b382()

From this, it appears to be a CLR exception. Correct?

Thanks for the rest of your answer. What should I bracket with the __try()
block? My code is inside an ActiveX control that is loaded by a VB.NET
application.

Regards,

Andrew
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:Oo**************@TK2MSFTNGP11.phx.gbl...

The address 0x7c59ba9d does not appear to be a kernel address.
It's a user mode address.

You have not told us which kind of exception this is:
one ACCESS_VIOLATION ?
a CLR Exception ?

on average, the way to "continue" the exceptio is something along these
lines.
The ExceptionFilter may be incorrect for your case,
since you have not provided enough information,
and we even don't knonw if the exception is continuable.

DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
CONTEXT * pContext = ExceptionPointers->ContextRecord;
if (0x7c59ba9d == (ULONG_PTR)Context->eip){
return EXCEPTION_CONTINUE_EXECUTION;
} else {
return EXCEPTION_CONTINUE_SEARCH.
}
}

Function(){
__try {
// your code here
} __except (ExceptionFilter(GetExceptionInformation())) {
// nothing
}
}

--
This posting is provided "AS IS" with no warranties, and confers no rights. Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:eA**************@TK2MSFTNGP10.phx.gbl...

My app. sees an exception from the kernel at an obscure address,

0x7c59ba9d.

When running under the VC++ 6.0 debugger, I can trap this each time it
occurs. If I want to trap it in my program and just tell the program to

just

continue (i.e. not pass the exception up the chain) how do I do this?

Any idea how I can find out what the cause of the exception is? In the

call

stack my program doesn't appear. MSCORWKS (which I think is .NET, the
framework used by the app. that calls my DLL) and KERNEL32 do.

Many thanks for any help that you can give.

[Ivan Brugiolo [MSFT]]

If you could use a system debugger like cdb/ntsd/windbg from
http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
and report the output of the '~*kb' command, that would help to see more
things.

The stack below is equvalent (as an educate guess) to a call to
kernel32!RaiseException.
The CLR raises C++ exception, CLR exceptin and some native exception as
well,
plus some intentional Access Violation in order to implement the
System.NullReferenceException.

As a general suggestion, you should add this constructs in your VB.NET code,
since the approach suggested below does not look feasible in you case.

Try
' your code here
Catch E as Exception
' your cleanup code here
End Catch
--
This posting is provided "AS IS" with no warranties, and confers no rights.
Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:uz**************@TK2MSFTNGP12.phx.gbl...

Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call
stack:

KERNEL32! 7c59ba9d()
MSCORWKS! 792206a8()
MSCORWKS! 7922063e()
MSCORWKS! 792205f1()
MSCORWKS! 792205d1()
MSCORWKS! 792a85fc()
MSCORWKS! 792a8ef9()
MSCORWKS! 79248678()
03f21d1e()
056ad957()
055d8422()
0474a865()
0474a5a7()
MSCORWKS! 791ece8a()
MSCORWKS! 791eb8a0()
MSCORWKS! 791f3941()
MSCORWKS! 791f38ff()
MSCORWKS! 792cc5f8()
MSCORWKS! 792cc6be()
KERNEL32! 7c57b382()

From this, it appears to be a CLR exception. Correct?

Thanks for the rest of your answer. What should I bracket with the __try()
block? My code is inside an ActiveX control that is loaded by a VB.NET
application.

Regards,

Andrew
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:Oo**************@TK2MSFTNGP11.phx.gbl...

The address 0x7c59ba9d does not appear to be a kernel address.
It's a user mode address.

You have not told us which kind of exception this is:
one ACCESS_VIOLATION ?
a CLR Exception ?

on average, the way to "continue" the exceptio is something along these
lines.
The ExceptionFilter may be incorrect for your case,
since you have not provided enough information,
and we even don't knonw if the exception is continuable.

DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
CONTEXT * pContext = ExceptionPointers->ContextRecord;
if (0x7c59ba9d == (ULONG_PTR)Context->eip){
return EXCEPTION_CONTINUE_EXECUTION;
} else {
return EXCEPTION_CONTINUE_SEARCH.
}
}

Function(){
__try {
// your code here
} __except (ExceptionFilter(GetExceptionInformation())) {
// nothing
}
}

--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:eA**************@TK2MSFTNGP10.phx.gbl...

My app. sees an exception from the kernel at an obscure address,

0x7c59ba9d.

When running under the VC++ 6.0 debugger, I can trap this each time it
occurs. If I want to trap it in my program and just tell the program

to just

continue (i.e. not pass the exception up the chain) how do I do this?

Any idea how I can find out what the cause of the exception is? In the

call

stack my program doesn't appear. MSCORWKS (which I think is .NET, the
framework used by the app. that calls my DLL) and KERNEL32 do.

Many thanks for any help that you can give.

[Andrew Chalk]

Thanks. What additional information would windbg give me. I already get an
assembler listing and a call stack in VC++?

Regards
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:%2****************@TK2MSFTNGP09.phx.gbl...

If you could use a system debugger like cdb/ntsd/windbg from
http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
and report the output of the '~*kb' command, that would help to see more
things.

The stack below is equvalent (as an educate guess) to a call to
kernel32!RaiseException.
The CLR raises C++ exception, CLR exceptin and some native exception as
well,
plus some intentional Access Violation in order to implement the
System.NullReferenceException.

As a general suggestion, you should add this constructs in your VB.NET code, since the approach suggested below does not look feasible in you case.

Try
' your code here
Catch E as Exception
' your cleanup code here
End Catch
--
This posting is provided "AS IS" with no warranties, and confers no rights. Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:uz**************@TK2MSFTNGP12.phx.gbl...

Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call
stack:

KERNEL32! 7c59ba9d()
MSCORWKS! 792206a8()
MSCORWKS! 7922063e()
MSCORWKS! 792205f1()
MSCORWKS! 792205d1()
MSCORWKS! 792a85fc()
MSCORWKS! 792a8ef9()
MSCORWKS! 79248678()
03f21d1e()
056ad957()
055d8422()
0474a865()
0474a5a7()
MSCORWKS! 791ece8a()
MSCORWKS! 791eb8a0()
MSCORWKS! 791f3941()
MSCORWKS! 791f38ff()
MSCORWKS! 792cc5f8()
MSCORWKS! 792cc6be()
KERNEL32! 7c57b382()

From this, it appears to be a CLR exception. Correct?

Thanks for the rest of your answer. What should I bracket with the __try()

block? My code is inside an ActiveX control that is loaded by a VB.NET
application.

Regards,

Andrew
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:Oo**************@TK2MSFTNGP11.phx.gbl...

The address 0x7c59ba9d does not appear to be a kernel address.
It's a user mode address.

You have not told us which kind of exception this is:
one ACCESS_VIOLATION ?
a CLR Exception ?

on average, the way to "continue" the exceptio is something along these lines.
The ExceptionFilter may be incorrect for your case,
since you have not provided enough information,
and we even don't knonw if the exception is continuable.

DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
CONTEXT * pContext = ExceptionPointers->ContextRecord;
if (0x7c59ba9d == (ULONG_PTR)Context->eip){
return EXCEPTION_CONTINUE_EXECUTION;
} else {
return EXCEPTION_CONTINUE_SEARCH.
}
}

Function(){
__try {
// your code here
} __except (ExceptionFilter(GetExceptionInformation())) {
// nothing
}
}

--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:eA**************@TK2MSFTNGP10.phx.gbl...
> My app. sees an exception from the kernel at an obscure address,
0x7c59ba9d.
> When running under the VC++ 6.0 debugger, I can trap this each time it > occurs. If I want to trap it in my program and just tell the program

to just
> continue (i.e. not pass the exception up the chain) how do I do this? >
> Any idea how I can find out what the cause of the exception is? In the call
> stack my program doesn't appear. MSCORWKS (which I think is .NET, the > framework used by the app. that calls my DLL) and KERNEL32 do.
>
> Many thanks for any help that you can give.
>
>

[Ivan Brugiolo [MSFT]]

something like the param passed to the RaiseException function,
so that I can see which exception it is and from which callstack with GOOD
SYMBOLS ?
The stack below is only good for a guess.

--
This posting is provided "AS IS" with no warranties, and confers no rights.
Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:O0**************@TK2MSFTNGP10.phx.gbl...

Thanks. What additional information would windbg give me. I already get an
assembler listing and a call stack in VC++?

Regards
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:%2****************@TK2MSFTNGP09.phx.gbl...

If you could use a system debugger like cdb/ntsd/windbg from
http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
and report the output of the '~*kb' command, that would help to see more
things.

The stack below is equvalent (as an educate guess) to a call to
kernel32!RaiseException.
The CLR raises C++ exception, CLR exceptin and some native exception as
well,
plus some intentional Access Violation in order to implement the
System.NullReferenceException.

As a general suggestion, you should add this constructs in your VB.NET code,

since the approach suggested below does not look feasible in you case.

Try
' your code here
Catch E as Exception
' your cleanup code here
End Catch
--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:uz**************@TK2MSFTNGP12.phx.gbl...

Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call stack:

KERNEL32! 7c59ba9d()
MSCORWKS! 792206a8()
MSCORWKS! 7922063e()
MSCORWKS! 792205f1()
MSCORWKS! 792205d1()
MSCORWKS! 792a85fc()
MSCORWKS! 792a8ef9()
MSCORWKS! 79248678()
03f21d1e()
056ad957()
055d8422()
0474a865()
0474a5a7()
MSCORWKS! 791ece8a()
MSCORWKS! 791eb8a0()
MSCORWKS! 791f3941()
MSCORWKS! 791f38ff()
MSCORWKS! 792cc5f8()
MSCORWKS! 792cc6be()
KERNEL32! 7c57b382()

From this, it appears to be a CLR exception. Correct?

Thanks for the rest of your answer. What should I bracket with the __try() block? My code is inside an ActiveX control that is loaded by a VB.NET
application.

Regards,

Andrew
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message news:Oo**************@TK2MSFTNGP11.phx.gbl...
> The address 0x7c59ba9d does not appear to be a kernel address.
> It's a user mode address.
>
> You have not told us which kind of exception this is:
> one ACCESS_VIOLATION ?
> a CLR Exception ?
>
> on average, the way to "continue" the exceptio is something along these > lines.
> The ExceptionFilter may be incorrect for your case,
> since you have not provided enough information,
> and we even don't knonw if the exception is continuable.
>
> DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
> CONTEXT * pContext = ExceptionPointers->ContextRecord;
> if (0x7c59ba9d == (ULONG_PTR)Context->eip){
> return EXCEPTION_CONTINUE_EXECUTION;
> } else {
> return EXCEPTION_CONTINUE_SEARCH.
> }
> }
>
> Function(){
> __try {
> // your code here
> } __except (ExceptionFilter(GetExceptionInformation())) {
> // nothing
> }
> }
>
> --
> This posting is provided "AS IS" with no warranties, and confers no
rights.
> Use of any included script samples are subject to the terms

specified at > http://www.microsoft.com/info/cpyright.htm
>
>
> "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
> news:eA**************@TK2MSFTNGP10.phx.gbl...
> > My app. sees an exception from the kernel at an obscure address,
> 0x7c59ba9d.
> > When running under the VC++ 6.0 debugger, I can trap this each

time

it > > occurs. If I want to trap it in my program and just tell the

program

to

> just
> > continue (i.e. not pass the exception up the chain) how do I do

this? > >
> > Any idea how I can find out what the cause of the exception is? In the > call
> > stack my program doesn't appear. MSCORWKS (which I think is .NET, the > > framework used by the app. that calls my DLL) and KERNEL32 do.
> >
> > Many thanks for any help that you can give.
> >
> >
>
>

[Andrew Chalk]

So windbg shows symbols?
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:O%****************@TK2MSFTNGP11.phx.gbl...

something like the param passed to the RaiseException function,
so that I can see which exception it is and from which callstack with GOOD
SYMBOLS ?
The stack below is only good for a guess.

--
This posting is provided "AS IS" with no warranties, and confers no rights. Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:O0**************@TK2MSFTNGP10.phx.gbl...

Thanks. What additional information would windbg give me. I already get an

assembler listing and a call stack in VC++?

Regards
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:%2****************@TK2MSFTNGP09.phx.gbl...

If you could use a system debugger like cdb/ntsd/windbg from
http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
and report the output of the '~*kb' command, that would help to see more things.

The stack below is equvalent (as an educate guess) to a call to
kernel32!RaiseException.
The CLR raises C++ exception, CLR exceptin and some native exception as well,
plus some intentional Access Violation in order to implement the
System.NullReferenceException.

As a general suggestion, you should add this constructs in your VB.NET

code,

since the approach suggested below does not look feasible in you case.

Try
' your code here
Catch E as Exception
' your cleanup code here
End Catch
--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:uz**************@TK2MSFTNGP12.phx.gbl...
> Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call > stack:
>
> KERNEL32! 7c59ba9d()
> MSCORWKS! 792206a8()
> MSCORWKS! 7922063e()
> MSCORWKS! 792205f1()
> MSCORWKS! 792205d1()
> MSCORWKS! 792a85fc()
> MSCORWKS! 792a8ef9()
> MSCORWKS! 79248678()
> 03f21d1e()
> 056ad957()
> 055d8422()
> 0474a865()
> 0474a5a7()
> MSCORWKS! 791ece8a()
> MSCORWKS! 791eb8a0()
> MSCORWKS! 791f3941()
> MSCORWKS! 791f38ff()
> MSCORWKS! 792cc5f8()
> MSCORWKS! 792cc6be()
> KERNEL32! 7c57b382()
>
> From this, it appears to be a CLR exception. Correct?
>
> Thanks for the rest of your answer. What should I bracket with the

__try()

> block? My code is inside an ActiveX control that is loaded by a VB.NET > application.
>
> Regards,
>
> Andrew
>
>
> "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message > news:Oo**************@TK2MSFTNGP11.phx.gbl...
> > The address 0x7c59ba9d does not appear to be a kernel address.
> > It's a user mode address.
> >
> > You have not told us which kind of exception this is:
> > one ACCESS_VIOLATION ?
> > a CLR Exception ?
> >
> > on average, the way to "continue" the exceptio is something along

these

> > lines.
> > The ExceptionFilter may be incorrect for your case,
> > since you have not provided enough information,
> > and we even don't knonw if the exception is continuable.
> >
> > DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
> > CONTEXT * pContext = ExceptionPointers->ContextRecord;
> > if (0x7c59ba9d == (ULONG_PTR)Context->eip){
> > return EXCEPTION_CONTINUE_EXECUTION;
> > } else {
> > return EXCEPTION_CONTINUE_SEARCH.
> > }
> > }
> >
> > Function(){
> > __try {
> > // your code here
> > } __except (ExceptionFilter(GetExceptionInformation())) {
> > // nothing
> > }
> > }
> >
> > --
> > This posting is provided "AS IS" with no warranties, and confers no > rights.
> > Use of any included script samples are subject to the terms specified

at

> > http://www.microsoft.com/info/cpyright.htm
> >
> >
> > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
> > news:eA**************@TK2MSFTNGP10.phx.gbl...
> > > My app. sees an exception from the kernel at an obscure address,
> > 0x7c59ba9d.
> > > When running under the VC++ 6.0 debugger, I can trap this each

time

it

> > > occurs. If I want to trap it in my program and just tell the

program to
> > just
> > > continue (i.e. not pass the exception up the chain) how do I do

this?

> > >
> > > Any idea how I can find out what the cause of the exception is?

In the

> > call
> > > stack my program doesn't appear. MSCORWKS (which I think is

..NET, the

> > > framework used by the app. that calls my DLL) and KERNEL32 do.
> > >
> > > Many thanks for any help that you can give.
> > >
> > >
> >
> >
>
>

[Ivan Brugiolo [MSFT]]

If you point it to the public symbol server, it will give a pretty decent
stack trace.
Follow the instruction on the page I pointed you out, and report the output
of `~*kb'.

--
This posting is provided "AS IS" with no warranties, and confers no rights.
Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:Oj**************@tk2msftngp13.phx.gbl...

So windbg shows symbols?
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:O%****************@TK2MSFTNGP11.phx.gbl...

something like the param passed to the RaiseException function,
so that I can see which exception it is and from which callstack with GOOD

SYMBOLS ?
The stack below is only good for a guess.

--
This posting is provided "AS IS" with no warranties, and confers no rights.

Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:O0**************@TK2MSFTNGP10.phx.gbl...

Thanks. What additional information would windbg give me. I already get an assembler listing and a call stack in VC++?

Regards
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in

message news:%2****************@TK2MSFTNGP09.phx.gbl...
> If you could use a system debugger like cdb/ntsd/windbg from
> http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
> and report the output of the '~*kb' command, that would help to see

more > things.
>
> The stack below is equvalent (as an educate guess) to a call to
> kernel32!RaiseException.
> The CLR raises C++ exception, CLR exceptin and some native exception as > well,
> plus some intentional Access Violation in order to implement the
> System.NullReferenceException.
>
> As a general suggestion, you should add this constructs in your VB.NET code,
> since the approach suggested below does not look feasible in you case. >
> Try
> ' your code here
> Catch E as Exception
> ' your cleanup code here
> End Catch
>
>
> --
> This posting is provided "AS IS" with no warranties, and confers no
rights.
> Use of any included script samples are subject to the terms specified at

> http://www.microsoft.com/info/cpyright.htm
>
>
> "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
> news:uz**************@TK2MSFTNGP12.phx.gbl...
> > Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call

> > stack:
> >
> > KERNEL32! 7c59ba9d()
> > MSCORWKS! 792206a8()
> > MSCORWKS! 7922063e()
> > MSCORWKS! 792205f1()
> > MSCORWKS! 792205d1()
> > MSCORWKS! 792a85fc()
> > MSCORWKS! 792a8ef9()
> > MSCORWKS! 79248678()
> > 03f21d1e()
> > 056ad957()
> > 055d8422()
> > 0474a865()
> > 0474a5a7()
> > MSCORWKS! 791ece8a()
> > MSCORWKS! 791eb8a0()
> > MSCORWKS! 791f3941()
> > MSCORWKS! 791f38ff()
> > MSCORWKS! 792cc5f8()
> > MSCORWKS! 792cc6be()
> > KERNEL32! 7c57b382()
> >
> > From this, it appears to be a CLR exception. Correct?
> >
> > Thanks for the rest of your answer. What should I bracket with the
__try()
> > block? My code is inside an ActiveX control that is loaded by a VB.NET > > application.
> >
> > Regards,
> >
> > Andrew
> >
> >
> > "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in

message

> > news:Oo**************@TK2MSFTNGP11.phx.gbl...
> > > The address 0x7c59ba9d does not appear to be a kernel address.
> > > It's a user mode address.
> > >
> > > You have not told us which kind of exception this is:
> > > one ACCESS_VIOLATION ?
> > > a CLR Exception ?
> > >
> > > on average, the way to "continue" the exceptio is something

along these
> > > lines.
> > > The ExceptionFilter may be incorrect for your case,
> > > since you have not provided enough information,
> > > and we even don't knonw if the exception is continuable.
> > >
> > > DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
> > > CONTEXT * pContext = ExceptionPointers->ContextRecord;
> > > if (0x7c59ba9d == (ULONG_PTR)Context->eip){
> > > return EXCEPTION_CONTINUE_EXECUTION;
> > > } else {
> > > return EXCEPTION_CONTINUE_SEARCH.
> > > }
> > > }
> > >
> > > Function(){
> > > __try {
> > > // your code here
> > > } __except (ExceptionFilter(GetExceptionInformation())) {
> > > // nothing
> > > }
> > > }
> > >
> > > --
> > > This posting is provided "AS IS" with no warranties, and confers

no > > rights.
> > > Use of any included script samples are subject to the terms

specified

at
> > > http://www.microsoft.com/info/cpyright.htm
> > >
> > >
> > > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message > > > news:eA**************@TK2MSFTNGP10.phx.gbl...
> > > > My app. sees an exception from the kernel at an obscure address, > > > 0x7c59ba9d.
> > > > When running under the VC++ 6.0 debugger, I can trap this each

time

it
> > > > occurs. If I want to trap it in my program and just tell the

program

> to
> > > just
> > > > continue (i.e. not pass the exception up the chain) how do I do this?
> > > >
> > > > Any idea how I can find out what the cause of the exception

is? In the
> > > call
> > > > stack my program doesn't appear. MSCORWKS (which I think is .NET, the
> > > > framework used by the app. that calls my DLL) and KERNEL32 do.
> > > >
> > > > Many thanks for any help that you can give.
> > > >
> > > >
> > >
> > >
> >
> >
>
>

[Andrew Chalk]

Thanks. I've d/l the debugger and am testing now.

Regards

"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:%2****************@tk2msftngp13.phx.gbl...

If you point it to the public symbol server, it will give a pretty decent
stack trace.
Follow the instruction on the page I pointed you out, and report the output of `~*kb'.

--
This posting is provided "AS IS" with no warranties, and confers no rights. Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:Oj**************@tk2msftngp13.phx.gbl...

So windbg shows symbols?
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:O%****************@TK2MSFTNGP11.phx.gbl...

something like the param passed to the RaiseException function,
so that I can see which exception it is and from which callstack with GOOD SYMBOLS ?
The stack below is only good for a guess.

--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:O0**************@TK2MSFTNGP10.phx.gbl...
> Thanks. What additional information would windbg give me. I already get

an

> assembler listing and a call stack in VC++?
>
> Regards
> "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message > news:%2****************@TK2MSFTNGP09.phx.gbl...
> > If you could use a system debugger like cdb/ntsd/windbg from
> > http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
> > and report the output of the '~*kb' command, that would help to see more

> > things.
> >
> > The stack below is equvalent (as an educate guess) to a call to
> > kernel32!RaiseException.
> > The CLR raises C++ exception, CLR exceptin and some native

exception

as

> > well,
> > plus some intentional Access Violation in order to implement the
> > System.NullReferenceException.
> >
> > As a general suggestion, you should add this constructs in your VB.NET > code,
> > since the approach suggested below does not look feasible in you case. > >
> > Try
> > ' your code here
> > Catch E as Exception
> > ' your cleanup code here
> > End Catch
> >
> >
> > --
> > This posting is provided "AS IS" with no warranties, and confers

no > rights.
> > Use of any included script samples are subject to the terms

specified

at

> > http://www.microsoft.com/info/cpyright.htm
> >
> >
> > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
> > news:uz**************@TK2MSFTNGP12.phx.gbl...
> > > Sorry. My sloppy terminology -- In a call to KERNEL32. Here is the call
> > > stack:
> > >
> > > KERNEL32! 7c59ba9d()
> > > MSCORWKS! 792206a8()
> > > MSCORWKS! 7922063e()
> > > MSCORWKS! 792205f1()
> > > MSCORWKS! 792205d1()
> > > MSCORWKS! 792a85fc()
> > > MSCORWKS! 792a8ef9()
> > > MSCORWKS! 79248678()
> > > 03f21d1e()
> > > 056ad957()
> > > 055d8422()
> > > 0474a865()
> > > 0474a5a7()
> > > MSCORWKS! 791ece8a()
> > > MSCORWKS! 791eb8a0()
> > > MSCORWKS! 791f3941()
> > > MSCORWKS! 791f38ff()
> > > MSCORWKS! 792cc5f8()
> > > MSCORWKS! 792cc6be()
> > > KERNEL32! 7c57b382()
> > >
> > > From this, it appears to be a CLR exception. Correct?
> > >
> > > Thanks for the rest of your answer. What should I bracket with the > __try()
> > > block? My code is inside an ActiveX control that is loaded by a

VB.NET

> > > application.
> > >
> > > Regards,
> > >
> > > Andrew
> > >
> > >
> > > "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in
message
> > > news:Oo**************@TK2MSFTNGP11.phx.gbl...
> > > > The address 0x7c59ba9d does not appear to be a kernel address.
> > > > It's a user mode address.
> > > >
> > > > You have not told us which kind of exception this is:
> > > > one ACCESS_VIOLATION ?
> > > > a CLR Exception ?
> > > >
> > > > on average, the way to "continue" the exceptio is something

along > these
> > > > lines.
> > > > The ExceptionFilter may be incorrect for your case,
> > > > since you have not provided enough information,
> > > > and we even don't knonw if the exception is continuable.
> > > >
> > > > DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){
> > > > CONTEXT * pContext = ExceptionPointers->ContextRecord;
> > > > if (0x7c59ba9d == (ULONG_PTR)Context->eip){
> > > > return EXCEPTION_CONTINUE_EXECUTION;
> > > > } else {
> > > > return EXCEPTION_CONTINUE_SEARCH.
> > > > }
> > > > }
> > > >
> > > > Function(){
> > > > __try {
> > > > // your code here
> > > > } __except (ExceptionFilter(GetExceptionInformation())) {
> > > > // nothing
> > > > }
> > > > }
> > > >
> > > > --
> > > > This posting is provided "AS IS" with no warranties, and confers no

> > > rights.
> > > > Use of any included script samples are subject to the terms
specified
> at
> > > > http://www.microsoft.com/info/cpyright.htm
> > > >
> > > >
> > > > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message > > > > news:eA**************@TK2MSFTNGP10.phx.gbl...
> > > > > My app. sees an exception from the kernel at an obscure address, > > > > 0x7c59ba9d.
> > > > > When running under the VC++ 6.0 debugger, I can trap this

each time
> it
> > > > > occurs. If I want to trap it in my program and just tell the
program
> > to
> > > > just
> > > > > continue (i.e. not pass the exception up the chain) how do I

do > this?
> > > > >
> > > > > Any idea how I can find out what the cause of the exception

is?

In

> the
> > > > call
> > > > > stack my program doesn't appear. MSCORWKS (which I think is

.NET,

> the
> > > > > framework used by the app. that calls my DLL) and KERNEL32 do. > > > > >
> > > > > Many thanks for any help that you can give.
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>

[Andrew Chalk]

I ran my executable from the WinDbg File->Open Executable menu and when the
first chance exception that I am looking for occurred, execution did not
halt. Any idea what I am doing wrong?

Regards,

Andrew
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:eY**************@tk2msftngp13.phx.gbl...

Thanks. I've d/l the debugger and am testing now.

Regards

"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message
news:%2****************@tk2msftngp13.phx.gbl...

If you point it to the public symbol server, it will give a pretty decent

stack trace.
Follow the instruction on the page I pointed you out, and report the output

of `~*kb'.

--
This posting is provided "AS IS" with no warranties, and confers no

rights.

Use of any included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm
"Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
news:Oj**************@tk2msftngp13.phx.gbl...

So windbg shows symbols?
"Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in message news:O%****************@TK2MSFTNGP11.phx.gbl...
> something like the param passed to the RaiseException function,
> so that I can see which exception it is and from which callstack with

GOOD

> SYMBOLS ?
> The stack below is only good for a guess.
>
> --
> This posting is provided "AS IS" with no warranties, and confers no
rights.
> Use of any included script samples are subject to the terms

specified at > http://www.microsoft.com/info/cpyright.htm
>
>
> "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in message
> news:O0**************@TK2MSFTNGP10.phx.gbl...
> > Thanks. What additional information would windbg give me. I

already get

an
> > assembler listing and a call stack in VC++?
> >
> > Regards
> > "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in

message

> > news:%2****************@TK2MSFTNGP09.phx.gbl...
> > > If you could use a system debugger like cdb/ntsd/windbg from
> > > http://www.microsoft.com/whdc/ddk/de...t.mspx?gssnb=1
> > > and report the output of the '~*kb' command, that would help to see more
> > > things.
> > >
> > > The stack below is equvalent (as an educate guess) to a call to
> > > kernel32!RaiseException.
> > > The CLR raises C++ exception, CLR exceptin and some native exception as
> > > well,
> > > plus some intentional Access Violation in order to implement the
> > > System.NullReferenceException.
> > >
> > > As a general suggestion, you should add this constructs in your

VB.NET

> > code,
> > > since the approach suggested below does not look feasible in you

case.

> > >
> > > Try
> > > ' your code here
> > > Catch E as Exception
> > > ' your cleanup code here
> > > End Catch
> > >
> > >
> > > --
> > > This posting is provided "AS IS" with no warranties, and confers no > > rights.
> > > Use of any included script samples are subject to the terms

specified

at
> > > http://www.microsoft.com/info/cpyright.htm
> > >
> > >
> > > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in

message > > > news:uz**************@TK2MSFTNGP12.phx.gbl...
> > > > Sorry. My sloppy terminology -- In a call to KERNEL32. Here is

the > call
> > > > stack:
> > > >
> > > > KERNEL32! 7c59ba9d()
> > > > MSCORWKS! 792206a8()
> > > > MSCORWKS! 7922063e()
> > > > MSCORWKS! 792205f1()
> > > > MSCORWKS! 792205d1()
> > > > MSCORWKS! 792a85fc()
> > > > MSCORWKS! 792a8ef9()
> > > > MSCORWKS! 79248678()
> > > > 03f21d1e()
> > > > 056ad957()
> > > > 055d8422()
> > > > 0474a865()
> > > > 0474a5a7()
> > > > MSCORWKS! 791ece8a()
> > > > MSCORWKS! 791eb8a0()
> > > > MSCORWKS! 791f3941()
> > > > MSCORWKS! 791f38ff()
> > > > MSCORWKS! 792cc5f8()
> > > > MSCORWKS! 792cc6be()
> > > > KERNEL32! 7c57b382()
> > > >
> > > > From this, it appears to be a CLR exception. Correct?
> > > >
> > > > Thanks for the rest of your answer. What should I bracket with the > > __try()
> > > > block? My code is inside an ActiveX control that is loaded by a VB.NET
> > > > application.
> > > >
> > > > Regards,
> > > >
> > > > Andrew
> > > >
> > > >
> > > > "Ivan Brugiolo [MSFT]" <iv******@online.microsoft.com> wrote in > message
> > > > news:Oo**************@TK2MSFTNGP11.phx.gbl...
> > > > > The address 0x7c59ba9d does not appear to be a kernel address. > > > > > It's a user mode address.
> > > > >
> > > > > You have not told us which kind of exception this is:
> > > > > one ACCESS_VIOLATION ?
> > > > > a CLR Exception ?
> > > > >
> > > > > on average, the way to "continue" the exceptio is something

along

> > these
> > > > > lines.
> > > > > The ExceptionFilter may be incorrect for your case,
> > > > > since you have not provided enough information,
> > > > > and we even don't knonw if the exception is continuable.
> > > > >
> > > > > DWORD ExceptionFilter(LPEXCEPTION_POINTERS ExceptionPointers){ > > > > > CONTEXT * pContext = ExceptionPointers->ContextRecord;
> > > > > if (0x7c59ba9d == (ULONG_PTR)Context->eip){
> > > > > return EXCEPTION_CONTINUE_EXECUTION;
> > > > > } else {
> > > > > return EXCEPTION_CONTINUE_SEARCH.
> > > > > }
> > > > > }
> > > > >
> > > > > Function(){
> > > > > __try {
> > > > > // your code here
> > > > > } __except (ExceptionFilter(GetExceptionInformation())) { > > > > > // nothing
> > > > > }
> > > > > }
> > > > >
> > > > > --
> > > > > This posting is provided "AS IS" with no warranties, and confers no
> > > > rights.
> > > > > Use of any included script samples are subject to the terms
> specified
> > at
> > > > > http://www.microsoft.com/info/cpyright.htm
> > > > >
> > > > >
> > > > > "Andrew Chalk" <ac****@XXXmagnacartasoftware.com> wrote in

message

> > > > > news:eA**************@TK2MSFTNGP10.phx.gbl...
> > > > > > My app. sees an exception from the kernel at an obscure

address,

> > > > > 0x7c59ba9d.
> > > > > > When running under the VC++ 6.0 debugger, I can trap this each > time
> > it
> > > > > > occurs. If I want to trap it in my program and just tell the > program
> > > to
> > > > > just
> > > > > > continue (i.e. not pass the exception up the chain) how do I do

> > this?
> > > > > >
> > > > > > Any idea how I can find out what the cause of the

exception

is?

In
> > the
> > > > > call
> > > > > > stack my program doesn't appear. MSCORWKS (which I think

is .NET,
> > the
> > > > > > framework used by the app. that calls my DLL) and KERNEL32

do. > > > > > >
> > > > > > Many thanks for any help that you can give.
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >
> > > >
> > >
> > >
> >
> >
>
>

[Pavel Lebedinsky]

"Andrew Chalk" wrote:

I ran my executable from the WinDbg File->Open Executable menu and when the first chance exception that I am looking for occurred, execution did not
halt. Any idea what I am doing wrong?

To enable 1st chance break on specific exceptions, do 'sxe <exception
code>'.

If you are interested in .NET exceptions, you can also do 'sxe clr'.

[Andrew Chalk]

Thanks. I'll try this.

"Pavel Lebedinsky" <m_pll ./. hotmail ./. com> wrote in message
news:%2******************@tk2msftngp13.phx.gbl...

"Andrew Chalk" wrote:

I ran my executable from the WinDbg File->Open Executable menu and when

the

first chance exception that I am looking for occurred, execution did not
halt. Any idea what I am doing wrong?

To enable 1st chance break on specific exceptions, do 'sxe <exception
code>'.

If you are interested in .NET exceptions, you can also do 'sxe clr'.

[Andrew Chalk]

Many thanks to yourself and Ivan. I eventually trapped the error.

Regards
"Pavel Lebedinsky" <m_pll ./. hotmail ./. com> wrote in message
news:%2******************@tk2msftngp13.phx.gbl...

"Andrew Chalk" wrote:

I ran my executable from the WinDbg File->Open Executable menu and when

the

first chance exception that I am looking for occurred, execution did not
halt. Any idea what I am doing wrong?

To enable 1st chance break on specific exceptions, do 'sxe <exception
code>'.

If you are interested in .NET exceptions, you can also do 'sxe clr'.