Decrease console output from CSD-lock timeouts for v6.6
PATCH 1/1 Decrease console output from CSD-lock timeouts for v6.6
Hello!
This series contains a couple of patches that reduce the console output produced by CSD-lock timeouts.
1. Reduce logging due to dump_stack of CSD waiters, courtesy of
Imran Khan.
2. Reduce NMI traffic from CSD waiters to CSD destination, courtesy
of Imran Khan.
Thanx, Paul
------------------------------------------------------------------------
b/kernel/smp.c | 3 ++- kernel/smp.c | 10 +++++++++- 2 files changed, 11 insertions(+), 2 deletions(-)
PATCH 1/2 smp: Reduce logging due to dump_stack of CSD waiters
From: Imran Khan <[email protected]>
If a waiter is waiting for CSD lock, its call stack will not change between first and subsequent hang detection for the same CSD lock. Therefore, do dump_stack only for first-time detection for a given waiter.
This avoids excessive logging on systems with hundreds of CPUs where repetitive dump_stack from hundreds of CPUs would otherwise flood the console.
--- kernel/smp.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
kernel/smp.c +2 -1
@@ -259,7 +259,8 @@ static bool csd_lock_wait_toolong(struct __call_single_data *csd, u64 ts0, u64 *arch_send_call_function_single_ipi (cpu); } } -dump_stack (); +if (firsttime) +dump_stack (); *ts1 = ts2;return false;
Signature · 1 line
2.40.1
PATCH 2/2 smp: Reduce NMI traffic from CSD waiters to CSD destination
In reply to Paul E. McKenneyFrom: Imran Khan <[email protected]>
On systems with hundreds of CPUs, if most of the CPUs detect a CSD hang, then all of these waiting CPUs send an NMI to the destination CPU in order to dump its backtrace.
Given enough NMIs, the destination CPU will spent much of its time producing backtraces, thus further delaying that CPU's response to the original CSD IPI. In the worst case, by the time destination CPU is done producing all of these backtrace NMIs, the CSD wait timeout will have elapsed so that the waiters resend their backtrace NMIs again, further delaying forward progress.
Therefore, to avoid these delays, issue the backtrace NMI only from the first waiter. The destination CPU's other waiters can make use of backtrace obtained from the first waiter's NMI.
--- kernel/smp.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
kernel/smp.c +9 -1
@@ -46,6 +46,8 @@ static DEFINE_PER_CPU_ALIGNED(struct call_function_data, cfd_data); static DEFINE_PER_CPU_SHARED_ALIGNED(struct llist_head, call_single_queue); +static DEFINE_PER_CPU(atomic_t, trigger_backtrace) = ATOMIC_INIT(1); +static void __flush_smp_call_function_queue (bool warn_cpu_offline);int smpcfd_prepare_cpu (unsigned int cpu)
@@ -253,7 +255,8 @@ static bool csd_lock_wait_toolong(struct __call_single_data *csd, u64 ts0, u64 * *bug_id, !cpu_cur_csd ? "unresponsive" : "handling this request"); }if (cpu >=0 ) { -dump_cpu_task (cpu); +if (atomic_cmpxchg_acquire (&per_cpu (trigger_backtrace, cpu),1 ,0 )) +dump_cpu_task (cpu);if (!cpu_cur_csd) {pr_alert ("csd: Re-sending CSD lock (#%d) IPI from CPU#%02d to CPU#%02d\n" , *bug_id,raw_smp_processor_id (), cpu);arch_send_call_function_single_ipi (cpu);
@@ -434,9 +437,14 @@ static void __flush_smp_call_function_queue(bool warn_cpu_offline)struct llist_node *entry, *prev;struct llist_head *head;static bool warned; +atomic_t *tbt;lockdep_assert_irqs_disabled (); +/* Allow waiters to send backtrace NMI from here onwards */ + tbt =this_cpu_ptr (&trigger_backtrace); +atomic_set_release (tbt,1 ); + head =this_cpu_ptr (&call_single_queue); entry =llist_del_all (head); entry =llist_reverse_order (entry);
Signature · 1 line
2.40.1