We just implemented a new debugger intercept process at a site and we are noticing that the intercept procedure is getting called twice for each error. I don't think I have noticed this before so I'm curious if this is a known issue. It might very well be something on our end but I didn't want to chase a rabbit down a hole unnecessarily.

Don Bakke

SRP Computer Solutions, Inc.

We just implemented a new debugger intercept process at a site and we are noticing that the intercept procedure is getting called twice for each error. I don't think I have noticed this before so I'm curious if this is a known issue. It might very well be something on our end but I didn't want to chase a rabbit down a hole unnecessarily.

Don Bakke

SRP Computer Solutions, Inc.

Hi, Don. Can you clarify which version of OI you're seeing this with? You have this in the OI 32-bit category…?

Thanks,

- Bryan Shumsky

Revelation Software, Inc.

Hi, Don. Can you clarify which version of OI you're seeing this with? You have this in the OI 32-bit category…?

Thanks,

- Bryan Shumsky

Revelation Software, Inc.

Hi, Bryan. Yes, I should have posted this. It is 9.4.6.

Don Bakke

SRP Computer Solutions, Inc.

Hi, Don. Can you clarify which version of OI you're seeing this with? You have this in the OI 32-bit category…?

Thanks,

- Bryan Shumsky

Revelation Software, Inc.

Hi, Bryan. Yes, I should have posted this. It is 9.4.6.

Don Bakke

SRP Computer Solutions, Inc.

Hi Bryan - Curious if you had thoughts on this since you asked me to clarify the version of OI. I am not asking for anybody to troubleshoot this but was wondering if anybody else has seen this. I can port my code to OI 10 and see if the issue is the same.

Don Bakke

SRP Computer Solutions, Inc.

Hi Don,

I don't recall seeing the intercept activate more than once - and once it's active there's a flag set that should stop it activating again until it's returned (to stop any recursion in case the intercept proc triggers it's own debug condition). The engine should also execute an "abort all" instruction once the intercept has run that should prevent any further processing until the next command is issued.

Is the call stack different between the first and second call? Maybe a profile log would show where the second trigger is coming from?

Regards,

Carl Pates

Hi Don,

I don't recall seeing the intercept activate more than once - and once it's active there's a flag set that should stop it activating again until it's returned (to stop any recursion in case the intercept proc triggers it's own debug condition). The engine should also execute an "abort all" instruction once the intercept has run that should prevent any further processing until the next command is issued.

Is the call stack different between the first and second call? Maybe a profile log would show where the second trigger is coming from?

Regards,

Carl Pates

Hi Carl,

Thanks for the response and confirming what is expected. The call stack is identical, which is what surprises me. I think there is something in my code that is causing interference. I was able to create a very stripped down version of the debugger and the problem has gone away. I'm going to start adding code back to identify the culprit. I'll report my findings.

Don Bakke

SRP Computer Solutions, Inc.

Hi Don,

I don't recall seeing the intercept activate more than once - and once it's active there's a flag set that should stop it activating again until it's returned (to stop any recursion in case the intercept proc triggers it's own debug condition). The engine should also execute an "abort all" instruction once the intercept has run that should prevent any further processing until the next command is issued.

Is the call stack different between the first and second call? Maybe a profile log would show where the second trigger is coming from?

Regards,

Carl Pates

Hi Carl,

Thanks for the response and confirming what is expected. The call stack is identical, which is what surprises me. I think there is something in my code that is causing interference. I was able to create a very stripped down version of the debugger and the problem has gone away. I'm going to start adding code back to identify the culprit. I'll report my findings.

Don Bakke

SRP Computer Solutions, Inc.

Wonder of wonders, it seems to have fixed itself. I stripped down the code and added back my logging, screen grab, and email logic one at a time and now it is working again. I won't look this gift horse in the mouth. :wink:

Don Bakke

SRP Computer Solutions, Inc.

<lol> I hate programming.

Carl Pates

View this thread on the Works forum...