On Tue, Sep 12, 2023 at 01:46:29PM +0200, Jinpu Wang wrote: > Hi Greg and Stable folks. > > We've noticed regression in raid1 due to following commits: > 79dabfd00a2b ("md/raid1: hold the barrier until handle_read_error() finishes") > caeed0b9f1ce ("md/raid1: free the r1bio before waiting for blocked rdev") I'll drop them from all queues, but can you test 6.6-rc1 to be sure that all is ok there? thanks, greg k-h