diff mbox series

[04/23] context_tracking: Add a note about noinstr VS unsafe context tracking functions

Message ID 20220620231029.3844583-4-paulmck@kernel.org (mailing list archive)
State Mainlined
Commit f67671baadf6dbae8eca46cb95e85b762e462b2b
Headers show
Series RCU/context-tracking updates for v5.20 | expand

Commit Message

Paul E. McKenney June 20, 2022, 11:10 p.m. UTC
From: Frederic Weisbecker <frederic@kernel.org>

Some context tracking functions enter or exit into/from RCU idle mode
while using trace-able and lockdep-aware IRQs (un-)masking. As a result
those functions can't get tagged as noinstr. This is unlikely to be
fixed since these are obsolete APIs. Drop a note about this matter.

[ paulmck: Apply Peter Zijlstra feedback. ]

Reported-by: Peter Zijlstra <peterz@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Uladzislau Rezki <uladzislau.rezki@sony.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Nicolas Saenz Julienne <nsaenz@kernel.org>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Xiongfeng Wang <wangxiongfeng2@huawei.com>
Cc: Yu Liao <liaoyu15@huawei.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Paul Gortmaker<paul.gortmaker@windriver.com>
Cc: Alex Belits <abelits@marvell.com>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
---
 kernel/context_tracking.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)
diff mbox series

Patch

diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c
index 36a98c48aedc7..3082332f64765 100644
--- a/kernel/context_tracking.c
+++ b/kernel/context_tracking.c
@@ -103,6 +103,15 @@  void noinstr __context_tracking_enter(enum ctx_state state)
 }
 EXPORT_SYMBOL_GPL(__context_tracking_enter);
 
+/*
+ * OBSOLETE:
+ * This function should be noinstr but the below local_irq_restore() is
+ * unsafe because it involves illegal RCU uses through tracing and lockdep.
+ * This is unlikely to be fixed as this function is obsolete. The preferred
+ * way is to call __context_tracking_enter() through user_enter_irqoff()
+ * or context_tracking_guest_enter(). It should be the arch entry code
+ * responsibility to call into context tracking with IRQs disabled.
+ */
 void context_tracking_enter(enum ctx_state state)
 {
 	unsigned long flags;
@@ -125,6 +134,14 @@  void context_tracking_enter(enum ctx_state state)
 NOKPROBE_SYMBOL(context_tracking_enter);
 EXPORT_SYMBOL_GPL(context_tracking_enter);
 
+/*
+ * OBSOLETE:
+ * This function should be noinstr but it unsafely calls local_irq_restore(),
+ * involving illegal RCU uses through tracing and lockdep.
+ * This is unlikely to be fixed as this function is obsolete. The preferred
+ * way is to call user_enter_irqoff(). It should be the arch entry code
+ * responsibility to call into context tracking with IRQs disabled.
+ */
 void context_tracking_user_enter(void)
 {
 	user_enter();
@@ -168,6 +185,15 @@  void noinstr __context_tracking_exit(enum ctx_state state)
 }
 EXPORT_SYMBOL_GPL(__context_tracking_exit);
 
+/*
+ * OBSOLETE:
+ * This function should be noinstr but the below local_irq_save() is
+ * unsafe because it involves illegal RCU uses through tracing and lockdep.
+ * This is unlikely to be fixed as this function is obsolete. The preferred
+ * way is to call __context_tracking_exit() through user_exit_irqoff()
+ * or context_tracking_guest_exit(). It should be the arch entry code
+ * responsibility to call into context tracking with IRQs disabled.
+ */
 void context_tracking_exit(enum ctx_state state)
 {
 	unsigned long flags;
@@ -182,6 +208,14 @@  void context_tracking_exit(enum ctx_state state)
 NOKPROBE_SYMBOL(context_tracking_exit);
 EXPORT_SYMBOL_GPL(context_tracking_exit);
 
+/*
+ * OBSOLETE:
+ * This function should be noinstr but it unsafely calls local_irq_save(),
+ * involving illegal RCU uses through tracing and lockdep. This is unlikely
+ * to be fixed as this function is obsolete. The preferred way is to call
+ * user_exit_irqoff(). It should be the arch entry code responsibility to
+ * call into context tracking with IRQs disabled.
+ */
 void context_tracking_user_exit(void)
 {
 	user_exit();