Progress graphical client stops working

Posted by jmls on 18-Aug-2011 07:06

10.2B04, windows 7 64 bit (32-bit progress)

I have a long-running process that "ticks" over every second, displaying the time. This screen listens for a tcp socket message, and on receipt of this message goes and "does stuff".

After a long period of no activity on the socket, an inbound message causes the client to crash and produce a stack trace.

This stack trace shows the line where the error happened, and it is a different line almost every time. However, the error *always* seems to ocurr on a NEW statement (like a = new foo.demo() ), but the class being instantiated is different in each case

Is the stack trace useful to progress at all ?

Does anyone have a similar problem ?

below is the first part of one of the protraces

=====================================================

PROGRESS stack trace as of Thu Aug 18 12:52:06 2011

=====================================================

Startup parameters:

-pf C:\Progress\OpenEdge\startup.pf,-cpinternal ISO8859-1,-cpstream ISO8859-1,-cpcoll Basic,-cpcase Basic,-d dmy,-numsep 44,-numdec 46,(end .pf),-s 200,-mmax 65534,-T C:\OpenEdge\WRK,-assemblies W:\data\hash42\dotr\dll,-db c:\data\db\temp-db\temp-db.db,-H localhost,-db X:\data\hash42db\hash42.db,-H localhost,-S 49750,-ld hash42,-db C:\data\db\sports\sports.db,-H localhost,-db W:\data\db\Maia.db,-H localhost,-debugalert,-rereadnolock,-inp 16000,-tok 4000,-zn,-p C:\eclipse\3.7\plugins\com.openedge.pdt.debug.core_10.2.1.04\abl\_debuglauncher.p

Exception code: C0000005 ACCESS_VIOLATION

Fault address:  1029B287 01:0029A287 C:\Progress\OpenEdge\bin\prow32.dll

Registers:

EAX:00000000

EBX:1074E7E0

ECX:073A60BC

EDX:0206BC58

ESI:00000000

EDI:073A60C8

CS:EIP:0023:1029B287

SS:ESP:002B:0018E790  EBP:1074E7E0

DS:002B  ES:002B  FS:0053  GS:002B

Flags:00010246

All Replies

Posted by olivier.dunemann on 26-Aug-2011 02:47

Hi Julian,

We've had the same kind of problem.

Could you check the following:

- Did you test the "-filterocxevents" client startup parameter?

- Is there any "PROCESS EVENTS" statement directly or indirectly invoked by the PSTimer's Tick event? If so, try to avoid it;

- Do you use the "MessageBlaster" control (msgblst32.ocx)? If so, try to avoid it. It does not work quite wel under Win7;

- Is your system's "Data Execution Prevention" (DEP) set to "Turn on DEP for essentils Windows programs and services only", rather than "Turn on DEP for all programs..."? (You can find this setting under My Computer/System Properties/Advanced/Performance Options/Data Execution Prevention).

Hope it helps.

Posted by jmls on 26-Aug-2011 03:00

thanks for the tips.

However, we don't use activex controls, just a plain old socket.

however, you've reminded me about -IOeverywhere

I've also managed to get the same error on a progress character client.

Will check out the DEP.

Thanks again

On 26 August 2011 08:47, olivier.dunemann

Posted by Admin on 26-Aug-2011 03:10

however, you've reminded me about -IOeverywhere

-IOEverywhere 1 (Capital E please).

Error message from wrong cased startup parameters can sometimes be very funny

Posted by jmls on 26-Aug-2011 03:20

bloody keybaord ...

Posted by Admin on 26-Aug-2011 03:25

bloody keybaord ...

Was just trying to save your weekend.

This thread is closed