Re: [RFC/PATCH 0/3] Fix ctnetlink regressions

[Pablo Neira Ayuso]

On Mon, Jan 16, 2017 at 09:14:05PM -0800, Kevin Cernekee wrote:
> These patches address a problem I am seeing on Linux 4.4.  They do not
> apply as-is to the master branch.  But I wanted to run them past the list
> first to gather feedback on whether this is a reasonable approach.

1/3 and 3/3 look fine.

Regarding 2/3, I'd suggest a new ctnetlink_update_status() and you use
it from ctnetlink_glue_parse_ct(). This new function allows us to set
status bits leaving what is already set intact, as you proposed.

So we leave ctnetlink_change_status() in place, but not being called
from the userspace helper path.

Thanks for fixing up this mess Kevin.