On Mon, 4 Jun 2018 14:45:57 +0900 Hoeun Ryu <hoeun....@lge.com.com> wrote:
> From: Hoeun Ryu <hoeun....@lge.com> > > Many console device drivers hold the uart_port->lock spinlock with irq > enabled > (using spin_lock()) while the device drivers are writing characters to their > devices, > but the device drivers just try to hold the spin lock (using spin_trylock()) > if > "oops_in_progress" is equal or greater than 1 to avoid deadlocks. > > There is a case ocurring a deadlock related to the lock and > oops_in_progress. A CPU > could be stopped by smp_send_stop() while it was holding the port lock > because irq was > enabled. Once a CPU stops, it doesn't respond interrupts anymore and the lock > stays > locked forever. > > console_flush_on_panic() is called during panic() and it eventually holds > the uart > lock but the lock is held by another stopped CPU and it is a deadlock. By > moving > bust_spinlocks(0) after console_flush_on_panic(), let the console device > drivers > think the Oops is still in progress to call spin_trylock() instead of > spin_lock() to > avoid the deadlock. hm. Sergey, is this at all related to the UART printk deadlock change which you're presently discussing in http://lkml.kernel.org/r/20180615093919.559-1-sergey.senozhat...@gmail.com? > --- a/kernel/panic.c > +++ b/kernel/panic.c > @@ -233,8 +233,6 @@ void panic(const char *fmt, ...) > if (_crash_kexec_post_notifiers) > __crash_kexec(NULL); > > - bust_spinlocks(0); > - > /* > * We may have ended up stopping the CPU holding the lock (in > * smp_send_stop()) while still having some valuable data in the console > @@ -246,6 +244,8 @@ void panic(const char *fmt, ...) > debug_locks_off(); > console_flush_on_panic(); > > + bust_spinlocks(0); > + > if (!panic_blink) > panic_blink = no_blink; >
