invalid page fault in module USER.EXE on PB6.5

[traxsoft1]

Does anyone have a fix for (invalid page fault in module USER.EXE) on
PB6.5. A changes to the code or PB fix.

Thank Jose

trax...@amn.com

[Eric Aling [TeamPS]]

Could you give us the exact address of the error and maybe some information
on what you are doing when this error occurs?

--

- Eric Aling [TeamPS], Cypres Informatisering bv, The Netherlands
Eric's Home & PB Site @ http://utopia.knoware.nl/users/cypr115

traxsoft1 <trax...@email.msn.com> wrote in message
snnmTe4...@forums.powersoft.com...

[Jason Howat]

We've been having the an invalid page fault problem with USER.EXE as well.
This was occuring in PB6.0.00, 6.0.01beta and now 6.5.00. It happens when
running the p-code and machine code executables, and when running the
program through the development environment. The address it occurs at is
always 0011:00000971.

There is one reference to a problem at this address in the Infobase and the
suggested solution was to regenerate the objects. We have regenerated our
objects three times before doing a build and the problem still occurs.

Any ideas?

Jason.

Eric Aling [TeamPS] <_NOSPAM_...@cypres.nl> wrote in article
<eJidNm#z9GA...@forums.powersoft.com>...

> Could you give us the exact address of the error and maybe some
information
> on what you are doing when this error occurs?
>
>

[traxsoft1]

Jason here is the message as it display

PB60 caused an invalid page fault in
module USER.EXE at 0011:0000095f.
Registers:
EAX=00000000 CS=16ff EIP=0000095f EFLGS=00000206
EBX=00015744 SS=489f ESP=0000862e EBP=bff5864c
ECX=00021677 DS=1677 ESI=0002cd7c FS=4897
EDX=00000000 ES=0b87 EDI=801e0002 GS=0000
Bytes at CS:EIP:
26 8b 77 06 eb 03 8b 76 e8 83 7e f6 00 74 09 c4
Stack dump:
0002cd7c 57440006 d94c0b87 cd180002 00000002 00c80000 d94c0002 86940002
38902020 2d680b87 00000b87 0002cd7c 8700489f 00060000 f690801e 9fbc0081

traxsoft1 wrote in message ...

[Eric Aling [TeamPS]]

Hi Jason,

I find it strange that you have this gpf also when running the programm
through development. User.exe is a 16bit exe while PB is 32bit (accessable
throught user32.dll, but still). Do you create your executables as 16bit or
32bit? What OS are you running on, what DBMS are you using. Can you try to
pinpoint when the gpf occurs (that is, when using a treeview control or tab
control)? Have you tried turnng the 'just in time debugging' setting on in
the system options?

HTH
--

- Eric Aling [TeamPS], Cypres Informatisering bv, The Netherlands
Eric's Home & PB Site @ http://utopia.knoware.nl/users/cypr115

Jason Howat <jas...@datacom.co.nz> wrote in message
01bdd095$26da7f00$44d8cbc0@jasonh...

>We've been having the an invalid page fault problem with USER.EXE as well.
>This was occuring in PB6.0.00, 6.0.01beta and now 6.5.00. It happens when
>running the p-code and machine code executables, and when running the
>program through the development environment. The address it occurs at is
>always 0011:00000971.
>
>There is one reference to a problem at this address in the Infobase and the
>suggested solution was to regenerate the objects. We have regenerated our
>objects three times before doing a build and the problem still occurs.
>
>Any ideas?
>
>Jason.
>
>Eric Aling [TeamPS] <_NOSPAM_...@cypres.nl> wrote in article
><eJidNm#z9GA...@forums.powersoft.com>...
>> Could you give us the exact address of the error and maybe some
>information
>> on what you are doing when this error occurs?
>>
>>
>> traxsoft1 <trax...@email.msn.com> wrote in message
>> snnmTe4...@forums.powersoft.com...

[joshua...@dayrunner.com]

I have experienced this, as well. The gpf in user.exe while running my app in
the development environment. Maybe there is some 16-bit code left in PB 6.5? I
am on Win95B connecting to MS SQL Server. Haven't tried compiling yet.

I have a pfc frame window with a treeview (u_tvs) and a datawindow (u_dw),
which shows detail for the item selected in the treeview. The gpf seems to
occur when I would select several items in the treeview in rapid succession.
Depending on the level, the datawindow control changes its dataobject then
does a retrieve.

I began getting this error when I tried not turning on the datawindow caching
service in my application manager (i.e. through "this.of_SetDWCache(TRUE)").
With the caching service NOT enabled, I can cause the gpf every time. When it
IS enabled, I have not yet been able to cause the gpf.

I'm very new to PB, so I will leave it up to everybody else to draw the
appropriate conclusions.

Josh Saint

In article <7wIL#OW19G...@forums.powersoft.com>,

-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum

[traxsoft1]

This GPF happens when running under WIN 95/98 and not NT4.0 running under
compile or pbd and it seem to happen after a retrieve of a datawindow and
just before it displays the data., you are able to see in the background of
the GPF the retrieved data.

any clue

jose

Eric Aling [TeamPS] <_NOSPAM_...@cypres.nl> wrote in message
<7wIL#OW19G...@forums.powersoft.com>...

[Eric Aling [TeamPS]]

Hi Jose,

Do you have graphical objects in your datawindow such as bitmaps, buttons or
graphs? Are you retrieving the data or do you use the sharedata() function
to fill the datawindow? Do you have something coded in the retrieverow or
retrieveend event?

--

- Eric Aling [TeamPS], Cypres Informatisering bv, The Netherlands
Eric's Home & PB Site @ http://utopia.knoware.nl/users/cypr115

traxsoft1 <trax...@email.msn.com> wrote in message
gQKIE5c...@forums.powersoft.com...

[Jason Howat]

Hi Eric,

Our application was originally in 32bit PB4 and has been migrated to PB6.5.
We are producing 32bit executables. We're developing and running the app
on Win 95. The DBMS is Sybase 11. AFAIK, there are no new features of
PB6.5 in use yet, it is simply a migrated version of the PB4 app. I
believe it happens when windows are being opened or closed, so that could
also involve database operations, it is hard to say exactly as there is not
one specific action that we've been able to find that produces the problem.

Surely the just in time debugging option would not be of much use here as
PowerBuilder isn't able to recover from an invalid page fault?

Thanks,
Jason.

Eric Aling [TeamPS] <_NOSPAM_...@cypres.nl> wrote in article

[Eric Aling [TeamPS]]

Hi Jason,

It's very hard to see from here what's going on. What I would suggest (as
last option) is to start the executable with the /PBDEBUG argument. This
produces a DBG file which can show you what the last instruction was before
it crashed. Have you tried that?

HTH
--

- Eric Aling [TeamPS], Cypres Informatisering bv, The Netherlands
Eric's Home & PB Site @ http://utopia.knoware.nl/users/cypr115

Jason Howat <jas...@datacom.co.nz> wrote in message
01bdd63b$39b01ee0$44d8cbc0@jasonh...

[Jason Howat]

Yes, I managed to reproduce the problem on Friday and again today, with the
PBDebug checkbox turned on in the System Options window within the PB
development environment. It produced 37 and 72 meg trace files before the
page fault occurred. In both cases the crash happened after the last line
associated with opening the window had executed (the rowfocuschanged event
of a datawindow on that window). This makes me suspect that the cause is
due to the way that PB is redrawing the screen, or handling the controls on
the screen.

I'm now going to work on cutting the application down so that there is a
bare minimum of code there that still produces the problem.

Thanks,
Jason.

Eric Aling [TeamPS] <_NOSPAM_...@cypres.nl> wrote in article

<GtS1ZZ$19GA...@forums.powersoft.com>...

[Eric Aling [TeamPS]]

Hi Jason,

That's good to hear. Check that event for possible invalid objects, rows
that can be 0, data that can be null etc.

HTH
--

- Eric Aling [TeamPS], Cypres Informatisering bv, The Netherlands
Eric's Home & PB Site @ http://utopia.knoware.nl/users/cypr115

Jason Howat <jas...@datacom.co.nz> wrote in message
01bdda13$1419a260$44d8cbc0@jasonh...