While running new RSC tests, I hit a flaky test: it took over one second for NDIS to re-attach the XDPLWF to the stack after an MTU change. This is outside of our control (the logs did not show any delays in any XDP or test code) so simply treat both the detach and the re-attach as being a partial restart from a timeout point-of-view.
While running new RSC tests, I hit a flaky test: it took over one second for NDIS to re-attach the XDPLWF to the stack after an MTU change. This is outside of our control (the logs did not show any delays in any XDP or test code) so simply treat both the detach and the re-attach as being a partial restart from a timeout point-of-view.