Re: Buffer Overrun in VC++ using VFP9

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance



There is at least one known bug in VFP that will cause this. (The one I
bumped into is fixed in SP1 for VFP9, which should be released shortly.)

The bug involves very specific data conditions and buffering. In my case,
the support team gave me a workaround (toggle buffering off and then back on
inside the loop where the error was occurring).

If you can reliably reproduce the problem, I recommend you call the official
support line and open a paid support incident. If your issue is caused by a
bug in the product, you won't be charged for the incident and you'll get
either a workaround or a patch. Either way, the only remedy with buffer
overruns is the official support route, and it has the added benefit of
getting fixes for these things into the stream for all of us.

If you cannot reliably reproduce the problem (i.e. on demand), there isn't
much of a workaround. Sorry.

Dan



John wrote:
> Every so often I get a a dialogue displaying a message stating:
>
> Buffer Overrun Detected!
> Program: c:\ASM GeMS\Sapphire.exe
> A buffer overrun has been detected which has corrupted the program's
> internal state. The program cannot safely continue execution and must
> now be terminated.
>
> The application, Sapphire.exe was originally written in VFP8 and was
> upgraded. It will run quite happily for several weeks and then, when
> it is started up sometimes throws the above error. Does anyone have
> any suggestions on how to go about debugging this? I notice that SP 1
> contains changes that prevent this error with large Scan loops but
> there is no such construct in this application.
>
> The environment is Windows 2000 Server SP4 with latest hotfix. The
> machine has 512 MB of RAM and a 2.4 GHz processor.
>
> This never occurred when the app was in VFP8. Nothing has changed
> other than a recompile. Leaving the application in VFP 8 is not a
> viable option due to company support policies.
>
> TIA
>
> John


.



Relevant Pages

  • Re: Heap corruption problem
    ... No. Valgrind is supported only on Linux -- fully on Pentiums and partly ... > I found the bug. ... > This buffer then had a fixed size, even I could get greater strings to ... Does it support Windows and the Open Watcom ...
    (comp.programming)
  • Re: 2.6.23-rc8-mm2: BUG near reiserfs_xattr_set
    ... I've got this BUG a few seconds after I logged in into Gnome desktop: ... unable to handle kernel NULL pointer dereference at virtual address 00000000 ... # Firmware Drivers ... # SCSI support type ...
    (Linux-Kernel)
  • SoftBUG on massive copying
    ... The system is running a vanilla 2.6.16.25 kernel, filesystems ... BUG: ... # ACPI Support ...
    (Linux-Kernel)
  • [BUG] 2.6.16-rc5-mm2/3 : oops on rmmod snd-ens1371
    ... I've got the following BUG when I rmmod snd-ens1371. ... unable to handle kernel paging request at virtual address 6b6b6b6b ... # ACPI Support ... # CPU Frequency scaling ...
    (Linux-Kernel)
  • [BUG 2.6.16-rt18] BUG at kernel/rtmutex.c:639!
    ... On an IBM Intellistation A-Pro I am seeing the bug regularly as seen below. ... to add 'irqpoll' to the kernel command line. ... # Firmware Drivers ... # ACPI Support ...
    (Linux-Kernel)