Re: FILESYS exception with hive registry in 4.2

hi,

thanks for responding. I have the following queries:

1> That was an example (in the previous post) with the system running in
debug mode. The following is the call stack when the system is in release
mode:

Call Stack: iesample.exe: 0x87D5C4F2 12:39:05 06/11/2005 India Standard
Time
FILESYS!HivesFromHKEY(HKEY__ * 0x80048bc0, long * 0xffffdac0) line 131
FILESYS!prgRegFlushKey(HKEY__ * 0x01f590e0 __imp_RegOpenKeyExW) line
1277 + 16 bytes
FILESYS!prgRegCloseKey(HKEY__ * 0xffffffff) line 2130 + 12 bytes
COREDLL!xxx_RegCloseKey(HKEY__ * 0x03174e90 IID_IInternetBindInfo) line
21 + 68 bytes
URLMON!GetUserAgentString(void) line 2838
URLMON!CINet::GetUserAgentString(void) line 3293
URLMON!CINet::INetAsyncOpen(void) line 1236 + 8 bytes
URLMON!CINet::INetAsyncStart(void) line 1132 + 16 bytes
URLMON!CINet::Start(unsigned short const *,struct IInternetProtocolSink
*,struct IInternetBindInfo *,unsigned long,unsigned long) line 659 + 8
bytes
URLMON!COInetProt::Start(const unsigned short * 0xfffffffe,
IInternetProtocolSink * 0x00000004, IInternetBindInfo * 0x03174cd0
IID_IInternetPriority, unsigned long 0x00000000, unsigned long 0x00000000,
IInternetPriority * 0x000468bc) line 106
URLMON!CTransaction::Start(const unsigned short * 0x00045f50,
IInternetProtocolSink * 0x00045e18, IInternetBindInfo * 0x00000004, unsigned
long 0x02000330, unsigned long 0x00000000, unsigned long 0x00000003) line
2028
URLMON!CBinding::StartBinding(const unsigned short * 0x00000000,
IBindCtx * 0x00000000, const _GUID & {...}, int 0x00000001, unsigned short *
* 0x0e0ae468, void * * 0x0e0ae500, int 0x00000001) line 1022 + 40 bytes
URLMON!CUrlMon::StartBinding(int 0x00000000, IBindCtx * 0x00000000,
IMoniker * 0x00000000, const _GUID & {...}, void * * 0x00000000, CBinding *
0x00740068) line 292
End Call Stack: iesample.exe: 0x87D5C4F2 12:39:05 06/11/2005 India Standard
Time

In this case, xxx_RegCloseKey seems to be the one who's starting off
things with 0xff_ff_ff_ff or in other words URLMon is not passing that...so,
why this difference between the release build and the debug build?

I get the same output on the KITL though :

Exception 004 Thread=87c85800 Proc=a7fb626a
AKY=00000043 PC=00020ca4 RA=00022f10 BVA=ffffffff
Process 'filesys.exe'

I am a little bit surprised as to how come BVA happens to be 0xff_ff_ff_ff
when the first argument passed into HivesFromHKEY happens to be 0x80048bc0.
I think the call stacks not probably showing the correct ones ?!? At the
point of exception a0 shows 0xff_ff_ff_ff in contrast to 0x80048bc0.

2> what i am actually worried about is how come i don't see any
discussions / reports on this (even if URLMON is the culprit)....It is so
very consistent here that i think if at all it is present everywhere, it
would have been talked about. Is there any chance by which you guys can try
this out (i am using a MIPS based platform, hive through flash/Hard Disk) ?
Even the exception through CEPlayer(another example application mentioned in
the first post) is through URLMon!, when i try to give File->Open-><some IP
addr:8080> and i don't get it when i use CEPlayer to open a media on the
system itself. The two call stacks that i have provided so far is by
opening IE.

Can you help me with this? thanks for the response again.

--
cheers,

Joseph Garibaldi.
B'lore, India.

"Andrew Rogers [MSFT]" <arogers@xxxxxxxxxxxxxxxxxxxx> wrote in message
news:%23elD51hbFHA.2520@xxxxxxxxxxxxxxxxxxxxxxx
> It appears that urlmon is passing -1 to RegCloseKey. Inside the hive
> registry code in filesys.exe, bad parameters (other than NULL) are caught
> using exception handling. So, this exception is handled inside filesys.exe
> and RegCloseKey returns failure. Technically, ulrmon should not be
> passing -1 to RegCloseKey, but the failure is completely benign.


.

Relevant Pages

  • Re: Application.ThreadException Doesnt Work Right. Why?
    ... I understand that, but what this is, is if any exception that is unintended ... we trap it along with all of the stack information etc. and send it ... >> it only functions when in debug mode. ... of any SQL Exceptions ...
    (microsoft.public.dotnet.framework.windowsforms)
  • Re: Try Catch Else Finally
    ... line numbers come up only when compiled in debug mode. ... You CAN get the line that caused the exception if you number your lines ... REM This program doesn't do much ... 50: Catch ex As DivideByZeroException ...
    (microsoft.public.dotnet.languages.vb)
  • Re: Object reference not set to an instance of an object.
    ... You will get more exception details if you are running the ... application in debug mode, than if you ran it in release mode. ... > Object reference not set to an instance of an object. ...
    (microsoft.public.dotnet.languages.vb)
  • Re: Having to "print" before method invocation?
    ... Hey Fredrik, thanks for responding. ... foo that I've been using without much fuss for a few months now. ... if I add print statements before trying to invoke methods on ... the interpreter won't raise the exception immediately (since it expected you to ...
    (comp.lang.python)
  • IDE debugging question
    ... Sub Main ... a divide by zero exception when attempting to assign to c. ... I am now in debug mode in the IDE and can I have the ...
    (microsoft.public.dotnet.languages.vb)