On Wed, Oct 18, 2023 at 9:57 AM Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> wrote: > That's odd given that as was pointed out, this test does not even touch > the code paths that this patch changed. I think I mixed up the "vmsplice" and the "splice" tests, and my conclusion was wrong, sorry. This performance regression is about the "splice" test which indeed uses /dev/zero. I'll have a closer look and get back to you.
