Op 06-03-13 01:30, Mandeep Singh Baines schreef:
> This check is turning up a lot of code paths which need to be
> fixed so while those paths are fixed, let's make this check
> optional so that folks can still use lockdep.
I think the config option should be inverted, and make it more clear that
you're not just not reporting some real bugs by disabling a check that should be on by default.
> CC: Tejun Heo <tj@xxxxxxxxxx>
> CC: Jeff Layton <jlayton@xxxxxxxxxx>
> CC: "Myklebust, Trond" <Trond.Myklebust@xxxxxxxxxx>
> CC: Oleg Nesterov <oleg@xxxxxxxxxx>
> CC: Ming Lei <ming.lei@xxxxxxxxxxxxx>
> CC: "Rafael J. Wysocki" <rjw@xxxxxxx>
> CC: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> CC: Ingo Molnar <mingo@xxxxxxxxxx>
> ---
> include/linux/freezer.h | 2 ++
> lib/Kconfig.debug | 12 ++++++++++++
> 2 files changed, 14 insertions(+)
>
> diff --git a/include/linux/freezer.h b/include/linux/freezer.h
> index 043a5cf..03bdc54 100644
> --- a/include/linux/freezer.h
> +++ b/include/linux/freezer.h
> @@ -49,8 +49,10 @@ extern void thaw_kernel_threads(void);
>
> static inline bool try_to_freeze(void)
> {
> +#ifdef CONFIG_DEBUG_LOCK_HELD_FREEZING
> if (!(current->flags & PF_NOFREEZE))
> debug_check_no_locks_held();
> +#endif
> might_sleep();
> if (likely(!freezing(current)))
> return false;
> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> index 28be08c..bddda5f 100644
> --- a/lib/Kconfig.debug
> +++ b/lib/Kconfig.debug
> @@ -547,6 +547,18 @@ config DEBUG_MUTEXES
> This feature allows mutex semantics violations to be detected and
> reported.
>
> +config DEBUG_LOCK_HELD_FREEZING
> + bool "Lock debugging: detect when locks are held during freeze"
> + depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
> + select DEBUG_SPINLOCK
> + select DEBUG_MUTEXES
> + select LOCKDEP
> + help
> + This feature will check whether any lock is incorrectly held
> + while freezing. If a task freezes with a lock held it will
> + block any other task that is waiting on that lock from freezing.
> + In the case of cgroup_freezer, this can cause a deadlock.
> +
> config DEBUG_LOCK_ALLOC
> bool "Lock debugging: detect incorrect freeing of live locks"
> depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
