Hi, i'm back at work with my patch for automatic CD loading, but the target moves faster than i can follow. Meanwhile the sr part of my change plays on a fresh construction site of linux-scsi: https://lore.kernel.org/linux-scsi/730eced4-c804-a78f-3d52-2a448dbd1b84@inte... "rework check_disk_change()" 2020-09-02 14:11 Christoph Hellwig Currently our plans overlap by [PATCH 17/19] "sr: use bdev_check_media_change" although they do not (yet) really conflict https://lore.kernel.org/linux-scsi/730eced4-c804-a78f-3d52-2a448dbd1b84@inte... It seems unwise to propose changes while Christoph Hellwig is actively working on the same code. (And i don't even know where to get the git branch with such a recent change when it is committed.) What is the socially acceptable way to coordinate plans with him ? - Shall i patiently wait until he is done with cdrom and sr ? - Shall i announce my interest in fixing automatic tray loading and other things in a brief mail to the list ? - Shall i post my patches (*) as they are and just wait for instructions ? (*) Studying Christoph's work in linux-scsi brought me to the prediction that they will want separate patches for cdrom and sr. Also i reworked my implementation along his design pattern of factoring out and exporting functions and then calling them where beneficial. (I.e. my -EDEADLK hack is dead. Long live cdrom_handle_open_tray(). Larger cdrom patch, but also nicer looking code.) Have a nice day :) Thomas