That's not the case. What happened is that the Samsung UART patches to support the M1 got merged into tty-next first, and it is preferable to have this series b

That's not the case. What happened is that the Samsung UART patches to support the M1 got merged into tty-next first, and it is preferable to have this series based on those - otherwise, the devicetree would fail validation due to the missing bindings, and there would be no serial to debug or do anything. It helps keeping things ordered for bisection. So we decided to merge a specific point from tty-next into this branch first, before the rest of the patches. Those patches are already in the pipeline for upstream inclusion. So it pulled in everything in tty-next that was before the Samsung changes, and it shows up in the diffstat, but those aren't really part of this series, it's just caused by the merge from tty-next.