Messages in this thread | | | Date | Tue, 19 Jun 2018 14:18:46 -0700 | From | "Paul E. McKenney" <> | Subject | Re: [RFC PATCH 2/2] rcu,debug_core: allow the kernel debugger to reset the rcu stall timer |
| |
On Mon, Aug 09, 2010 at 12:12:12AM -0500, Jason Wessel wrote: > When returning from the kernel debugger allow a reset of the rcu > jiffies_stall value to prevent the rcu stall detector from sending NMI > events which stack dumps on all the cpus in the system.
Not sure where the 2010 date came from, but it almost fooled me into deleting your emails unread. ;-)
> Signed-off-by: Jason Wessel <jason.wessel@windriver.com> > CC: Dipankar Sarma <dipankar@in.ibm.com> > CC: Paul E. McKenney <paulmck@linux.vnet.ibm.com> > CC: Ingo Molnar <mingo@elte.hu> > --- > include/linux/rcupdate.h | 8 ++++++++ > kernel/debug/debug_core.c | 2 ++ > kernel/rcutree.c | 9 +++++++++ > 3 files changed, 19 insertions(+), 0 deletions(-) > > diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h > index 9fbc54a..abd3ab6 100644 > --- a/include/linux/rcupdate.h > +++ b/include/linux/rcupdate.h > @@ -599,4 +599,12 @@ static inline void debug_rcu_head_unqueue(struct rcu_head *head) > #define rcu_dereference_index_check(p, c) \ > __rcu_dereference_index_check((p), (c)) > > +#ifdef CONFIG_RCU_CPU_STALL_DETECTOR > +extern void rcu_cpu_stall_reset(void); > +#else /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ > +static inline void rcu_cpu_stall_reset(void) > +{ > +} > +#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ > + > #endif /* __LINUX_RCUPDATE_H */ > diff --git a/kernel/debug/debug_core.c b/kernel/debug/debug_core.c > index e4d6819..1600e90 100644 > --- a/kernel/debug/debug_core.c > +++ b/kernel/debug/debug_core.c > @@ -47,6 +47,7 @@ > #include <linux/pid.h> > #include <linux/smp.h> > #include <linux/mm.h> > +#include <linux/rcupdate.h> > > #include <asm/cacheflush.h> > #include <asm/byteorder.h> > @@ -474,6 +475,7 @@ static void dbg_touch_watchdogs(void) > { > touch_softlockup_watchdog_sync(); > clocksource_touch_watchdog(); > + rcu_cpu_stall_reset(); > } > > static int kgdb_cpu_enter(struct kgdb_state *ks, struct pt_regs *regs) > diff --git a/kernel/rcutree.c b/kernel/rcutree.c > index d5bc439..209b755 100644 > --- a/kernel/rcutree.c > +++ b/kernel/rcutree.c > @@ -532,6 +532,9 @@ static void check_cpu_stall(struct rcu_state *rsp, struct rcu_data *rdp) > > if (rcu_cpu_stall_panicking) > return; > + /* Gracefully handle a watch dog reset when jiffies_stall == 0 */ > + if (!rsp->jiffies_stall) > + return;
Why not just use the existing rcu_cpu_stall_reset()? It sets the next stall a long way into the future, like 2 billion jiffies on 32-bit systems.
> delta = jiffies - rsp->jiffies_stall; > rnp = rdp->mynode; > if ((rnp->qsmask & rdp->grpmask) && delta >= 0) { > @@ -561,6 +564,12 @@ static void __init check_cpu_stall_init(void) > atomic_notifier_chain_register(&panic_notifier_list, &rcu_panic_block); > } > > +void rcu_cpu_stall_reset(void) > +{ > + rcu_sched_state.jiffies_stall = 0; > + rcu_bh_state.jiffies_stall = 0;
This should get you a compiler warning given the existing rcu_cpu_stall_reset(). It also fails to do anything about rcu_preempt_state on PREEMPT=y kernels.
What happens if you just remove the rcutree.c changes from your series and test with the result?
Thanx, Paul
> +} > + > #else /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */ > > static void record_gp_stall_check_time(struct rcu_state *rsp) > -- > 1.6.3.3 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ >
| |