Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1198419 > unrolled thread

[patch 1/7] x86/irq: Protect smp_cleanup_move

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [patch 1/7] x86/irq: Protect smp_cleanup_move Thomas Gleixner <tglx@linutronix.de> - 2015-08-02 22:40 +0200

#1198419 — [patch 1/7] x86/irq: Protect smp_cleanup_move

smp_cleanup_move fiddles without protection in the interrupt
descriptors and the vector array. A concurrent irq setup/teardown or
affinity setting can pull the rug under that operation.

Add proper locking.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 arch/x86/kernel/apic/vector.c |   18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)

Index: tip/arch/x86/kernel/apic/vector.c
===================================================================
--- tip.orig/arch/x86/kernel/apic/vector.c
+++ tip/arch/x86/kernel/apic/vector.c
@@ -539,6 +539,9 @@ asmlinkage __visible void smp_irq_move_c
 
 	entering_ack_irq();
 
+	/* Prevent vectors vanishing under us */
+	raw_spin_lock(&vector_lock);
+
 	me = smp_processor_id();
 	for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
 		int irq;
@@ -546,6 +549,7 @@ asmlinkage __visible void smp_irq_move_c
 		struct irq_desc *desc;
 		struct apic_chip_data *data;
 
+	retry:
 		irq = __this_cpu_read(vector_irq[vector]);
 
 		if (irq <= VECTOR_UNDEFINED)
@@ -555,12 +559,16 @@ asmlinkage __visible void smp_irq_move_c
 		if (!desc)
 			continue;
 
+		if (!raw_spin_trylock(&desc->lock)) {
+			raw_spin_unlock(&vector_lock);
+			cpu_relax();
+			raw_spin_lock(&vector_lock);
+			goto retry;
+		}
+
 		data = apic_chip_data(&desc->irq_data);
 		if (!data)
-			continue;
-
-		raw_spin_lock(&desc->lock);
-
+			goto unlock;
 		/*
 		 * Check if the irq migration is in progress. If so, we
 		 * haven't received the cleanup request yet for this irq.
@@ -589,6 +597,8 @@ unlock:
 		raw_spin_unlock(&desc->lock);
 	}
 
+	raw_spin_unlock(&vector_lock);
+
 	exiting_irq();
 }
 


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [standalone]

Back to top | Article view | linux.kernel

csiph-web