Re: bug 1592

tor 2006-09-07 klockan 23:01 +0800 skrev Adrian Chadd:
> > What my stuff does is to forcibly call the abort handlers - which will unwind them top down. It kills the ftpstate, then it kills the forward state through the abort handler. This seems to DTRT in the error case.
>
> .. it does this by calling comm_close(ctrl.fd). Like in 2.6.
>
> How are we meant to do it in -3?

I think for 3.0 we should be similar to 2.6. Protocol registers the
error in fwdState and then closes the connection or otherwise
deassociates the fd from the fwdState.

For 3.1 we may rethink stuff. Find the current (2.6) delay of actual
error propagation a bit odd and hard to follow unless you have already
been down that path once.

Regards
Henrik