On 7/12/2022 2:02 PM, David Marchand wrote:
On Tue, Jul 12, 2022 at 7:58 PM Greg Rose <gvrose8...@gmail.com> wrote:
It is time to remove support for the ovs kernel driver and push
towards use of the upstream linux openvswitch kernel driver
in it's place [1].
1. https://mail.openvswitch.org/pipermail/ovs-dev/2022-april/393292.html
Changes in V2:
* Removed the Linux datapath and fixed up various Makefiles to adjust
* Moved openvswitch.h from datapath/linux/compat/include to include/linux
and fixed up the Makefiles to adjust
* Removed further instances in utilities and auxiliary files where the
--with-linux configuration command was still being used.
* Changed order of patches to (hopefully) fix up some robot auto-test
errors
Changes in V3:
* Squashed the two patches removing rhel specs due to interdependence
between the two that caused build errors between application of each
of the patches
* Moved the changes to acinclude.m4 towards the end of the series
Changes in V4:
* Changed up the order of the patches again to fix robot errors
* Added back most of the tests in .github/workflows/build-and-test.yml
* Removed xenserver and related documentation, test scripts and makefile
dependencies
There are probably more cleanups to do for xenserver, but the focus
should be on OVS kmod removal and this xenserver stuff can be finished
later, so I am fine with the current patch.
Still to do:
* Figure out why a certain 4 tests always fail with 'make check' in the
github workflows but do not fail on my own system.
Allez !
I think we are close to something working.
The v5 will be the one :-).
Yep, getting down to the last details here. I found that the last
patch to remove xenserver is what is causing a 'make check' failure
in ci. I'll see if I can track that down as well.
Thanks for all the reviews David, you've been a great help.
- Greg
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
