On Tue, Feb 27, 2024 at 10:16:39AM -0500, Mike Snitzer wrote:

That's the mainline issue a bunch of MD (and dm-raid) oriented
engineers are working hard to fix, they've been discussing on
linux-raid (with many iterations of proposed patches).

It regressed due to 6.8 MD changes (maybe earlier).

Do you know if there is a way to skip specific tests to get a useful
baseline value (and to complete the run?)