Re: [PATCH 02/11] rcu: fix bug when rcu_exp_handler() in nested interrupt

From: Lai Jiangshan
Date: Thu Oct 31 2019 - 11:14:42 EST

On 2019/10/31 10:31 äå, Paul E. McKenney wrote:
On Thu, Oct 31, 2019 at 06:47:31AM -0700, Paul E. McKenney wrote:
On Thu, Oct 31, 2019 at 10:07:57AM +0000, Lai Jiangshan wrote:
These is a possible bug (although which I can't triger yet)
since 2015 8203d6d0ee78
(rcu: Use single-stage IPI algorithm for RCU expedited grace period)

->rcu_read_lock_nesting = -RCU_NEST_BIAS;
interrupt(); // before or after rcu_read_unlock_special()
fetch some rcu protected pointers
// exp GP starts in other cpu.
some works
NESTED interrupt for rcu_exp_handler();

Also, which platforms support nested interrupts? Last I knew, this was

report exp qs! BUG!

Why would a quiescent state for the expedited grace period be reported
here? This CPU is still in an RCU read-side critical section, isn't it?

And I now see what you were getting at here. Yes, the current code
assumes that interrupt-disabled regions, like hardware interrupt
handlers, cannot be interrupted. But if interrupt-disabled regions such
as hardware interrupt handlers can be interrupted (as opposed to being
NMIed), wouldn't that break a whole lot of stuff all over the place in
the kernel? So that sounds like an arch bug to me.

I don't know when I started always assuming hardware interrupt
handler can be nested by (other) interrupt. I can't find any
documents say Linux don't allow nested interrupt handler.
Google search suggests the opposite.

grep -rIni nested Documentation/memory-barriers.txt Documentation/x86/
It still have some words about nested interrupt handler.

The whole patchset doesn't depend on this patch, and actually
it is reverted later in the patchset. Dropping this patch
can be an option for next round.


Thanx, Paul

// exp GP completes and pointers are freed in other cpu
some works with the pointers. BUG
->rcu_read_lock_nesting = 0;

Although rcu_sched_clock_irq() can be in nested interrupt,
there is no such similar bug since special.b.need_qs
can only be set when ->rcu_read_lock_nesting > 0

Signed-off-by: Lai Jiangshan <laijs@xxxxxxxxxxxxxxxxx>
kernel/rcu/tree_exp.h | 5 +++--
kernel/rcu/tree_plugin.h | 9 ++++++---
2 files changed, 9 insertions(+), 5 deletions(-)

diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h
index 6dec21909b30..c0d06bce35ea 100644
--- a/kernel/rcu/tree_exp.h
+++ b/kernel/rcu/tree_exp.h
@@ -664,8 +664,9 @@ static void rcu_exp_handler(void *unused)
* Otherwise, force a context switch after the CPU enables everything.
rdp->exp_deferred_qs = true;
- if (!(preempt_count() & (PREEMPT_MASK | SOFTIRQ_MASK)) ||
- WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs())) {
+ if (rcu_preempt_need_deferred_qs(t) &&
+ (!(preempt_count() & (PREEMPT_MASK | SOFTIRQ_MASK)) ||
+ WARN_ON_ONCE(rcu_dynticks_curr_cpu_in_eqs()))) {
} else {
diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index d4c482490589..59ef10da1e39 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -549,9 +549,12 @@ rcu_preempt_deferred_qs_irqrestore(struct task_struct *t, unsigned long flags)
static bool rcu_preempt_need_deferred_qs(struct task_struct *t)
- return (__this_cpu_read(rcu_data.exp_deferred_qs) ||
- READ_ONCE(t->rcu_read_unlock_special.s)) &&
- t->rcu_read_lock_nesting <= 0;
+ return (__this_cpu_read(rcu_data.exp_deferred_qs) &&
+ (!t->rcu_read_lock_nesting ||
+ t->rcu_read_lock_nesting == -RCU_NEST_BIAS))
+ ||
+ (READ_ONCE(t->rcu_read_unlock_special.s) &&
+ t->rcu_read_lock_nesting <= 0);