We have had several discussions with the other printk developers and together decided that we would like more time to investigate the implicit expectations [1] of each and every console driver because now we are concerned that there may be more hidden implicit expectations. We much appreciate the chance to have this in 5.19 release candidates because the wider user base helped us to get feedback from user cases and architectures that we were not able to cover. And we see that the patchset needs some more love. We are sorry for the inconvenience and thanks a lot for the feedback and help with debugging and testing. This patchset reverts only the final patches adding the kthreads and per-console. It keeps the refactoring of the console_unlock() code that makes sense on its own. [1] https://lore.kernel.org/all/YrBdjVwBOVgLfHyb@alley/ PS: I am going to prepare a proper pull request. I just wanted to send this in advance just in case anyone wants to comment on it. Petr Mladek (6): Revert "printk: Wait for the global console lock when the system is going down" Revert "printk: Block console kthreads when direct printing will be required" Revert "printk: remove @console_locked" Revert "printk: extend console_lock for per-console locking" Revert "printk: add kthread console printers" Revert "printk: add functions to prefer direct printing" drivers/tty/sysrq.c | 2 - include/linux/console.h | 17 -- include/linux/printk.h | 16 - kernel/hung_task.c | 11 +- kernel/panic.c | 6 - kernel/printk/internal.h | 2 - kernel/printk/printk.c | 593 ++++-------------------------------- kernel/printk/printk_safe.c | 32 -- kernel/rcu/tree_stall.h | 2 - kernel/reboot.c | 16 +- kernel/watchdog.c | 4 - kernel/watchdog_hld.c | 4 - 12 files changed, 65 insertions(+), 640 deletions(-) -- 2.35.3
