On 1/15/25 1:26 PM, Jann Horn wrote: > The locking in the buffer cloning code is somewhat complex because it goes > back and forth between locking the source ring and the destination ring. > > Make it easier to reason about by locking both rings at the same time. > To avoid ABBA deadlocks, lock the rings in ascending kernel address order, > just like in lock_two_nondirectories(). I'm going to let this one sit for review for a day or two, and since it depends changes slated for 6.13, I'll queue it up in a post-merge branch for 6.14 if all goes well. Thanks! -- Jens Axboe
