[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20210109020536.127953-8-frederic@kernel.org>
Date: Sat, 9 Jan 2021 03:05:35 +0100
From: Frederic Weisbecker <frederic@...nel.org>
To: Peter Zijlstra <peterz@...radead.org>,
"Paul E . McKenney" <paulmck@...nel.org>
Cc: LKML <linux-kernel@...r.kernel.org>,
Frederic Weisbecker <frederic@...nel.org>,
"Rafael J . Wysocki" <rafael.j.wysocki@...el.com>,
Ingo Molnar <mingo@...nel.org>,
Thomas Gleixner <tglx@...utronix.de>, stable@...r.kernel.org
Subject: [RFC PATCH 7/8] entry: Report local wake up on resched blind zone while resuming to user
The last rescheduling opportunity while resuming to user is in
exit_to_user_mode_loop(). This means that any wake up performed on
the local runqueue after this point is going to have its rescheduling
silently ignored.
Perform sanity checks to report these situations.
Signed-off-by: Frederic Weisbecker <frederic@...nel.org>
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Thomas Gleixner <tglx@...utronix.de>
Cc: Ingo Molnar<mingo@...nel.org>
Cc: Paul E. McKenney <paulmck@...nel.org>
Cc: Rafael J. Wysocki <rafael.j.wysocki@...el.com>
---
kernel/entry/common.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/kernel/entry/common.c b/kernel/entry/common.c
index 8f3292b5f9b7..1dfb97762336 100644
--- a/kernel/entry/common.c
+++ b/kernel/entry/common.c
@@ -5,6 +5,7 @@
#include <linux/highmem.h>
#include <linux/livepatch.h>
#include <linux/audit.h>
+#include <linux/sched.h>
#include "common.h"
@@ -23,6 +24,8 @@ static __always_inline void __enter_from_user_mode(struct pt_regs *regs)
instrumentation_begin();
trace_hardirqs_off_finish();
instrumentation_end();
+
+ sched_resched_local_allow();
}
void noinstr enter_from_user_mode(struct pt_regs *regs)
@@ -206,6 +209,7 @@ static void exit_to_user_mode_prepare(struct pt_regs *regs)
if (unlikely(ti_work & EXIT_TO_USER_MODE_WORK))
ti_work = exit_to_user_mode_loop(regs, ti_work);
+ sched_resched_local_forbid();
arch_exit_to_user_mode_prepare(regs, ti_work);
/* Ensure that the address limit is intact and no locks are held */
--
2.25.1
Powered by blists - more mailing lists