Watchdog-timer system reset not working on atmega328p

Hello again,

While using, your awesome, SimulIDE 1.0.0 we might have encountered a bug regarding the
Simulation of the Watchdog-timer on the atmega328p. The problem is that the MCU’s system reset is not being triggered when it should be.

To demonstrate the issue, we have created a minimal not working example attached below.

It would be very nice if you could take a look at it.

As always thank you for your wonderful project and have a good day.

Hi Fabian.
Thanks for reporting and for the example.
I will have a look.

There is an error in how the watchdog is enabled/disabled:
Currently some options determined by config word (fuses) are configured in Properties->Config.

There you have "Enable Watchdog", this is equivalent to WDTON fuse.
This fuse can force to enable the Watchdog, but not to disable it.
Currently if you don't enable Watchdog in Properties dialog then it will not work.

So until I fix this issue you need to enable Watchdog in Properties->Config.

Solved at trunk Rev. 1338, 1.0.1 Rev. 1339, 1.0.0 Rev. 1133.

Thank you very much that is helpful!

SimulIDE_1.0.0-RC3 is  available for download:
https://www.simulide.com/p/downloads.html

This issue should be solved in this version.

Hello Arcachofo,

there still seems to be unexpected behavior on RC2 (with enabling Watchdog) and on RC3
when the fuze time is set to 2s or lower like in the demo I have sent you. In this case, the system reset is triggered permanently without the fuze time elapsing.

It only works as expected for me when the fuze time is set to 4s or more.

there still seems to be unexpected behavior on RC2 (with enabling Watchdog) and on RC3
when the fuze time is set to 2s or lower like in the demo I have sent you. In this case, the system reset is triggered permanently without the fuze time elapsing.

It only works as expected for me when the fuze time is set to 4s or more.

Solved at trunk Rev 1341.

Ok, I have to apologize something went wrong when I pulled Revision 1134 leading me to believe it is still bugged. It functions perfectly fine now.

We will continue working with rc3.
Thank you very much.

I forgot to mention that last problem about wrong prescaler was also solved in branch 1.0.0, at Rev 1136.

But I'm not sure if you are using executable from download page or building from Launchpad repository.

I am building from Launchpad.