Merge tag 'core-rcu-2020-06-01' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull RCU updates from Ingo Molnar:
 "The RCU updates for this cycle were:

   - RCU-tasks update, including addition of RCU Tasks Trace for BPF use
     and TASKS_RUDE_RCU

   - kfree_rcu() updates.

   - Remove scheduler locking restriction

   - RCU CPU stall warning updates.

   - Torture-test updates.

   - Miscellaneous fixes and other updates"

* tag 'core-rcu-2020-06-01' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (103 commits)
  rcu: Allow for smp_call_function() running callbacks from idle
  rcu: Provide rcu_irq_exit_check_preempt()
  rcu: Abstract out rcu_irq_enter_check_tick() from rcu_nmi_enter()
  rcu: Provide __rcu_is_watching()
  rcu: Provide rcu_irq_exit_preempt()
  rcu: Make RCU IRQ enter/exit functions rely on in_nmi()
  rcu/tree: Mark the idle relevant functions noinstr
  x86: Replace ist_enter() with nmi_enter()
  x86/mce: Send #MC singal from task work
  x86/entry: Get rid of ist_begin/end_non_atomic()
  sched,rcu,tracing: Avoid tracing before in_nmi() is correct
  sh/ftrace: Move arch_ftrace_nmi_{enter,exit} into nmi exception
  lockdep: Always inline lockdep_{off,on}()
  hardirq/nmi: Allow nested nmi_enter()
  arm64: Prepare arch_nmi_enter() for recursion
  printk: Disallow instrumenting print_nmi_enter()
  printk: Prepare for nested printk_nmi_enter()
  rcutorture: Convert ULONG_CMP_LT() to time_before()
  torture: Add a --kasan argument
  torture: Save a few lines by using config_override_param initially
  ...

1 files changed, 59 insertions, 19 deletions

diff --git a/arch/arm64/include/asm/hardirq.h b/arch/arm64/include/asm/hardirq.h
index 87ad961f3c97..985493af704b 100644
--- a/arch/arm64/include/asm/hardirq.h
+++ b/arch/arm64/include/asm/hardirq.h
@@ -32,30 +32,70 @@ u64 smp_irq_stat_cpu(unsigned int cpu);
 
 struct nmi_ctx {
 	u64 hcr;
+	unsigned int cnt;
 };
 
 DECLARE_PER_CPU(struct nmi_ctx, nmi_contexts);
 
-#define arch_nmi_enter()							\
-	do {									\
-		if (is_kernel_in_hyp_mode()) {					\
-			struct nmi_ctx *nmi_ctx = this_cpu_ptr(&nmi_contexts);	\
-			nmi_ctx->hcr = read_sysreg(hcr_el2);			\
-			if (!(nmi_ctx->hcr & HCR_TGE)) {			\
-				write_sysreg(nmi_ctx->hcr | HCR_TGE, hcr_el2);	\
-				isb();						\
-			}							\
-		}								\
-	} while (0)
+#define arch_nmi_enter()						\
+do {									\
+	struct nmi_ctx *___ctx;						\
+	u64 ___hcr;							\
+									\
+	if (!is_kernel_in_hyp_mode())					\
+		break;							\
+									\
+	___ctx = this_cpu_ptr(&nmi_contexts);				\
+	if (___ctx->cnt) {						\
+		___ctx->cnt++;						\
+		break;							\
+	}								\
+									\
+	___hcr = read_sysreg(hcr_el2);					\
+	if (!(___hcr & HCR_TGE)) {					\
+		write_sysreg(___hcr | HCR_TGE, hcr_el2);		\
+		isb();							\
+	}								\
+	/*								\
+	 * Make sure the sysreg write is performed before ___ctx->cnt	\
+	 * is set to 1. NMIs that see cnt == 1 will rely on us.		\
+	 */								\
+	barrier();							\
+	___ctx->cnt = 1;                                                \
+	/*								\
+	 * Make sure ___ctx->cnt is set before we save ___hcr. We	\
+	 * don't want ___ctx->hcr to be overwritten.			\
+	 */								\
+	barrier();							\
+	___ctx->hcr = ___hcr;						\
+} while (0)
 
-#define arch_nmi_exit()								\
-	do {									\
-		if (is_kernel_in_hyp_mode()) {					\
-			struct nmi_ctx *nmi_ctx = this_cpu_ptr(&nmi_contexts);	\
-			if (!(nmi_ctx->hcr & HCR_TGE))				\
-				write_sysreg(nmi_ctx->hcr, hcr_el2);		\
-		}								\
-	} while (0)
+#define arch_nmi_exit()							\
+do {									\
+	struct nmi_ctx *___ctx;						\
+	u64 ___hcr;							\
+									\
+	if (!is_kernel_in_hyp_mode())					\
+		break;							\
+									\
+	___ctx = this_cpu_ptr(&nmi_contexts);				\
+	___hcr = ___ctx->hcr;						\
+	/*								\
+	 * Make sure we read ___ctx->hcr before we release		\
+	 * ___ctx->cnt as it makes ___ctx->hcr updatable again.		\
+	 */								\
+	barrier();							\
+	___ctx->cnt--;							\
+	/*								\
+	 * Make sure ___ctx->cnt release is visible before we		\
+	 * restore the sysreg. Otherwise a new NMI occurring		\
+	 * right after write_sysreg() can be fooled and think		\
+	 * we secured things for it.					\
+	 */								\
+	barrier();							\
+	if (!___ctx->cnt && !(___hcr & HCR_TGE))			\
+		write_sysreg(___hcr, hcr_el2);				\
+} while (0)
 
 static inline void ack_bad_irq(unsigned int irq)
 {