On 6/29/21 11:23 AM, Michael Johnson wrote:
I tracked it down to a problem with capabilities in the systemd service. I have
sent a patch.
Yep, nice catch. I don't run systemd, otherwise I would have caught this sooner.
Along with the fix, I'm thinking it would be good to have extra
ell/dhcp.c to make these kinds of errors easier to find.
Right. I think we were a bit sloppy here.
Would you be open to a patch which adds `l_error()` output if the
fails? I know that `ell/dhcp.c` doesn't currently
Yes, definitely. We should at least log this failure via CLIENT_DEBUG.
have any error output but instead bubbles the errors up. Also I
don't know how
fatal this kind of error should really be so maybe
logging isn't enough?
We could. I think we issue l_dhcp_client_stop in a few failure cases already.
Perhaps we can also add an extra event to signal that a fatal error occurred.
Would make it easier to know something went wrong even if IWD_DHCP_DEBUG isn't set.