Portfast edge on trunk L2 loops behaviour?

Hi, has anyone tried to purposely create a L2 loop in a lab environment?

I have tried to do this a couple of ways now but haven't been able to. The only thing I haven't tried is bpdufilter on the specific interfaces.

But alas, I know that any port that is set to portfast/edge if it receives a BPDU on it, it will revert to spanning tree mode normal thus preventing L2 loops that would be caused by having it in the forwarding state straight away.

Especially if that port was connected to another switch along with another port (2x single different trunk connections to an opposite switch) i thought i could create one by setting both single trunk interfaces to portfast/edge trunk then i thought broadcasts would loop around each of the ports and back to the original sending switch because they're both set to forwarding. But STP worked anyway, I'm guessing this because the interfaces reverted back to STP mode normal because it received a BPDU on the interfaces, thus normal spanning resumed.

I guess its a testament to how far we have come in terms of having built in security in our switches (good ones anyway) that prevents this type of thing from ever happening.

The only option I see left is to put BPDUfilter on both interfaces and watch everything go down I assume?

I would've thought initially that setting port type edge and to a trunk would've caused it, but alas as I said that didn't happen.