[PATCH 1/2 -tip] irq: remove unneeded desc->chip->ack check

[PATCH 1/2 -tip] irq: remove unneeded desc->chip->ack check

[Wang Chen]

desc->chip->ack is initialized to ack_bad().
It will not be NULL.

Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
---
 kernel/irq/chip.c   |    3 +--
 kernel/irq/handle.c |   14 +++++---------
 2 files changed, 6 insertions(+), 11 deletions(-)

diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index f63c706..b929803 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -530,8 +530,7 @@ handle_percpu_irq(unsigned int irq, struct irq_desc *desc)
 
 	kstat_incr_irqs_this_cpu(irq, desc);
 
-	if (desc->chip->ack)
-		desc->chip->ack(irq);
+	desc->chip->ack(irq);
 
 	action_ret = handle_IRQ_event(irq, desc->action);
 	if (!noirqdebug)
diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c
index 6492400..4408672 100644
--- a/kernel/irq/handle.c
+++ b/kernel/irq/handle.c
@@ -337,11 +337,9 @@ unsigned int __do_IRQ(unsigned int irq)
 		/*
 		 * No locking required for CPU-local interrupts:
 		 */
-		if (desc->chip->ack) {
-			desc->chip->ack(irq);
-			/* get new one */
-			desc = irq_remap_to_desc(irq, desc);
-		}
+		desc->chip->ack(irq);
+		/* get new one */
+		desc = irq_remap_to_desc(irq, desc);
 		if (likely(!(desc->status & IRQ_DISABLED))) {
 			action_ret = handle_IRQ_event(irq, desc->action);
 			if (!noirqdebug)
@@ -352,10 +350,8 @@ unsigned int __do_IRQ(unsigned int irq)
 	}
 
 	spin_lock(&desc->lock);
-	if (desc->chip->ack) {
-		desc->chip->ack(irq);
-		desc = irq_remap_to_desc(irq, desc);
-	}
+	desc->chip->ack(irq);
+	desc = irq_remap_to_desc(irq, desc);
 	/*
 	 * REPLAY is when Linux resends an IRQ that was dropped earlier
 	 * WAITING is used by probe to mark irqs that are being tested
-- 
1.5.3.4

Re: [PATCH 1/2 -tip] irq: remove unneeded desc->chip->ack check

[Ingo Molnar]

* Wang Chen <wangchen@cn.fujitsu.com> wrote:

> desc->chip->ack is initialized to ack_bad().
> It will not be NULL.

hm, that is only true of no_irq_chip() - is it true of all irq_chip 
definitions on all architectures?

	Ingo

[PATCH -tip] irq: check chip->ack before calling (WAS: irq: remove unneeded desc->chip->ack check)

[Wang Chen]

Ingo Molnar said the following on 2008-12-26 21:27:
> * Wang Chen <wangchen@cn.fujitsu.com> wrote:
> 
>> desc->chip->ack is initialized to ack_bad().
>> It will not be NULL.
> 
> hm, that is only true of no_irq_chip() - is it true of all irq_chip 
> definitions on all architectures?
> 

No, some arch's irq_chip doesn't have ack routine.
So I was wrong.
But this enlighten another thought that generic irq layer doesn't know
whether irq_chip has ack routine on some architectures.
Upon that, before calling chip->ack, should check it's not NULL.

Impact: fix theoretic NULL dereference

Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
---
diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index f63c706..9a7fbb8 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -290,7 +290,8 @@ static inline void mask_ack_irq(struct irq_desc *desc, int irq)
 		desc->chip->mask_ack(irq);
 	else {
 		desc->chip->mask(irq);
-		desc->chip->ack(irq);
+		if (desc->chip->ack)
+			desc->chip->ack(irq);
 	}
 }
 
@@ -475,7 +476,8 @@ handle_edge_irq(unsigned int irq, struct irq_desc *desc)
 	kstat_incr_irqs_this_cpu(irq, desc);
 
 	/* Start handling the irq */
-	desc->chip->ack(irq);
+	if (desc->chip->ack)
+		desc->chip->ack(irq);
 	desc = irq_remap_to_desc(irq, desc);
 
 	/* Mark the IRQ currently in progress.*/

Re: [PATCH -tip] irq: check chip->ack before calling (WAS: irq: remove unneeded desc->chip->ack check)

[Ingo Molnar]

* Wang Chen <wangchen@cn.fujitsu.com> wrote:

> Ingo Molnar said the following on 2008-12-26 21:27:
> > * Wang Chen <wangchen@cn.fujitsu.com> wrote:
> > 
> >> desc->chip->ack is initialized to ack_bad().
> >> It will not be NULL.
> > 
> > hm, that is only true of no_irq_chip() - is it true of all irq_chip 
> > definitions on all architectures?
> > 
> 
> No, some arch's irq_chip doesn't have ack routine.
> So I was wrong.
> But this enlighten another thought that generic irq layer doesn't know
> whether irq_chip has ack routine on some architectures.
> Upon that, before calling chip->ack, should check it's not NULL.
> 
> Impact: fix theoretic NULL dereference
> 
> Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>

applied to tip/irq/genirq, thanks!

(find below how the final commit message ended up looking like.)

	Ingo

------------------>
>From efdc64f0c792ea744bcc9203f35b908e66d42f41 Mon Sep 17 00:00:00 2001
From: Wang Chen <wangchen@cn.fujitsu.com>
Date: Mon, 29 Dec 2008 13:35:11 +0800
Subject: [PATCH] genirq: check chip->ack before calling

Impact: fix theoretical NULL dereference

The generic irq layer doesn't know whether irq_chip has ack routine on some
architectures or not. Upon that, before calling chip->ack, we should check
that it's not NULL.

Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 kernel/irq/chip.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index 6eb3c79..0ad02d7 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -290,7 +290,8 @@ static inline void mask_ack_irq(struct irq_desc *desc, int irq)
 		desc->chip->mask_ack(irq);
 	else {
 		desc->chip->mask(irq);
-		desc->chip->ack(irq);
+		if (desc->chip->ack)
+			desc->chip->ack(irq);
 	}
 }
 
@@ -475,7 +476,8 @@ handle_edge_irq(unsigned int irq, struct irq_desc *desc)
 	kstat_incr_irqs_this_cpu(irq, desc);
 
 	/* Start handling the irq */
-	desc->chip->ack(irq);
+	if (desc->chip->ack)
+		desc->chip->ack(irq);
 	desc = irq_remap_to_desc(irq, desc);
 
 	/* Mark the IRQ currently in progress.*/