Re: Strange exception, please help
From: Leroy Casterline (spamfreeme_at_nospam.com)
Date: 11/02/04
- Next message: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Previous message: Steve Jackowski: "Packect Filters/Drivers"
- In reply to: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Next in thread: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Reply: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Messages sorted by: [ date ] [ thread ]
Date: Tue, 02 Nov 2004 13:40:29 -0700
>You should have included the crash dump analysis produced by "!analyze -v".
>Are you sure that the instruction code was the same at the time of the
>crash?
This is happening in an application rather than in a driver. The code is
written in BCB 6, so I'm using Borland's debugger rather than WinDbg.
Yes, the instruction code is identical between crashing and non-crashing
runs. The only difference is the location of the instruction in virtual
memory. Commenting/uncommenting a line of code in my app shifts the
location of the failing instruction by 20 bytes. The failure is
absolutely repeatable.
Baring a hardware problem, do you know of any way that an instruction
could write to an address other than its target address? Perhaps a
problem with virtual memory mapping or some such?
- Next message: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Previous message: Steve Jackowski: "Packect Filters/Drivers"
- In reply to: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Next in thread: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Reply: nospam_at_cristalink.com: "Re: Strange exception, please help"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|