linux-kernel - [tip:irq/core] irq_work: Do not raise an IPI when queueing work on the local CPU

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <tip-471ba0e686cb13752bc1ff3216c54b69a2d250ea@git.kernel.org>
Date:   Thu, 18 Apr 2019 05:10:53 -0700
From:   tip-bot for Nicholas Piggin <tipbot@...or.com>
To:     linux-tip-commits@...r.kernel.org
Cc:     clg@...d.org, rostedt@...dmis.org, sjitindarsingh@...il.com,
        frederic@...nel.org, npiggin@...il.com, hpa@...or.com,
        paulus@...ba.org, bigeasy@...utronix.de, tglx@...utronix.de,
        mingo@...nel.org, linux-kernel@...r.kernel.org,
        peterz@...radead.org, torvalds@...ux-foundation.org
Subject: [tip:irq/core] irq_work: Do not raise an IPI when queueing work on
 the local CPU

Commit-ID:  471ba0e686cb13752bc1ff3216c54b69a2d250ea
Gitweb:     https://git.kernel.org/tip/471ba0e686cb13752bc1ff3216c54b69a2d250ea
Author:     Nicholas Piggin <npiggin@...il.com>
AuthorDate: Tue, 9 Apr 2019 19:34:03 +1000
Committer:  Ingo Molnar <mingo@...nel.org>
CommitDate: Thu, 18 Apr 2019 14:07:52 +0200

irq_work: Do not raise an IPI when queueing work on the local CPU

The QEMU PowerPC/PSeries machine model was not expecting a self-IPI,
and it may be a bit surprising thing to do, so have irq_work_queue_on
do local queueing when target is the current CPU.

Suggested-by: Steven Rostedt <rostedt@...dmis.org>
Reported-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Tested-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Signed-off-by: Nicholas Piggin <npiggin@...il.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@...radead.org>
Reviewed-by: Frederic Weisbecker <frederic@...nel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@...radead.org>
Cc: =?UTF-8?q?C=C3=A9dric=20Le=20Goater?= <clg@...d.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: Paul Mackerras <paulus@...ba.org>
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Suraj Jitindar Singh <sjitindarsingh@...il.com>
Cc: Thomas Gleixner <tglx@...utronix.de>
Link: https://lkml.kernel.org/r/20190409093403.20994-1-npiggin@gmail.com
[ Simplified the preprocessor comments.
  Fixed unbalanced curly brackets pointed out by Thomas. ]
Signed-off-by: Ingo Molnar <mingo@...nel.org>
---
 kernel/irq_work.c | 75 +++++++++++++++++++++++++++++++------------------------
 1 file changed, 42 insertions(+), 33 deletions(-)

diff --git a/kernel/irq_work.c b/kernel/irq_work.c
index 6b7cdf17ccf8..73288914ed5e 100644
--- a/kernel/irq_work.c
+++ b/kernel/irq_work.c
@@ -56,61 +56,70 @@ void __weak arch_irq_work_raise(void)
 	 */
 }
 
-/*
- * Enqueue the irq_work @work on @cpu unless it's already pending
- * somewhere.
- *
- * Can be re-enqueued while the callback is still in progress.
- */
-bool irq_work_queue_on(struct irq_work *work, int cpu)
+/* Enqueue on current CPU, work must already be claimed and preempt disabled */
+static void __irq_work_queue_local(struct irq_work *work)
 {
-	/* All work should have been flushed before going offline */
-	WARN_ON_ONCE(cpu_is_offline(cpu));
-
-#ifdef CONFIG_SMP
-
-	/* Arch remote IPI send/receive backend aren't NMI safe */
-	WARN_ON_ONCE(in_nmi());
+	/* If the work is "lazy", handle it from next tick if any */
+	if (work->flags & IRQ_WORK_LAZY) {
+		if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) &&
+		    tick_nohz_tick_stopped())
+			arch_irq_work_raise();
+	} else {
+		if (llist_add(&work->llnode, this_cpu_ptr(&raised_list)))
+			arch_irq_work_raise();
+	}
+}
 
+/* Enqueue the irq work @work on the current CPU */
+bool irq_work_queue(struct irq_work *work)
+{
 	/* Only queue if not already pending */
 	if (!irq_work_claim(work))
 		return false;
 
-	if (llist_add(&work->llnode, &per_cpu(raised_list, cpu)))
-		arch_send_call_function_single_ipi(cpu);
-
-#else /* #ifdef CONFIG_SMP */
-	irq_work_queue(work);
-#endif /* #else #ifdef CONFIG_SMP */
+	/* Queue the entry and raise the IPI if needed. */
+	preempt_disable();
+	__irq_work_queue_local(work);
+	preempt_enable();
 
 	return true;
 }
+EXPORT_SYMBOL_GPL(irq_work_queue);
 
-/* Enqueue the irq work @work on the current CPU */
-bool irq_work_queue(struct irq_work *work)
+/*
+ * Enqueue the irq_work @work on @cpu unless it's already pending
+ * somewhere.
+ *
+ * Can be re-enqueued while the callback is still in progress.
+ */
+bool irq_work_queue_on(struct irq_work *work, int cpu)
 {
+#ifndef CONFIG_SMP
+	return irq_work_queue(work);
+
+#else /* CONFIG_SMP: */
+	/* All work should have been flushed before going offline */
+	WARN_ON_ONCE(cpu_is_offline(cpu));
+
 	/* Only queue if not already pending */
 	if (!irq_work_claim(work))
 		return false;
 
-	/* Queue the entry and raise the IPI if needed. */
 	preempt_disable();
-
-	/* If the work is "lazy", handle it from next tick if any */
-	if (work->flags & IRQ_WORK_LAZY) {
-		if (llist_add(&work->llnode, this_cpu_ptr(&lazy_list)) &&
-		    tick_nohz_tick_stopped())
-			arch_irq_work_raise();
+	if (cpu != smp_processor_id()) {
+		/* Arch remote IPI send/receive backend aren't NMI safe */
+		WARN_ON_ONCE(in_nmi());
+		if (llist_add(&work->llnode, &per_cpu(raised_list, cpu)))
+			arch_send_call_function_single_ipi(cpu);
 	} else {
-		if (llist_add(&work->llnode, this_cpu_ptr(&raised_list)))
-			arch_irq_work_raise();
+		__irq_work_queue_local(work);
 	}
-
 	preempt_enable();
 
 	return true;
+#endif /* CONFIG_SMP */
 }
-EXPORT_SYMBOL_GPL(irq_work_queue);
+
 
 bool irq_work_needs_cpu(void)
 {