PATCH] s390: fix machine check handling

PATCH] s390: fix machine check handling

[Sebastian Ott]

Hi,

would it be possible to have this one for 3.18 (if not, please add
Cc: stable@vger.kernel.org # v3.18)

Regards,
Sebastian

>8-----

>From ca198a6ff5766f19645104b7f8d621774524c4b4 Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Fri, 28 Nov 2014 15:40:57 +0100
Subject: [PATCH] s390: fix machine check handling

Commit eb7e7d76 "s390: Replace __get_cpu_var uses" broke machine check
handling.

We copy machine check information from per-cpu to a stack variable for
local processing. Next we should zap the per-cpu variable, not the
stack variable.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
---
 arch/s390/kernel/nmi.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index dd1c24c..3f51cf4 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -54,12 +54,8 @@ void s390_handle_mcck(void)
 	 */
 	local_irq_save(flags);
 	local_mcck_disable();
-	/*
-	 * Ummm... Does this make sense at all? Copying the percpu struct
-	 * and then zapping it one statement later?
-	 */
-	memcpy(&mcck, this_cpu_ptr(&cpu_mcck), sizeof(mcck));
-	memset(&mcck, 0, sizeof(struct mcck_struct));
+	mcck = *this_cpu_ptr(&cpu_mcck);
+	memset(this_cpu_ptr(&cpu_mcck), 0, sizeof(mcck));
 	clear_cpu_flag(CIF_MCCK_PENDING);
 	local_mcck_enable();
 	local_irq_restore(flags);
-- 
1.8.4.2

Re: PATCH] s390: fix machine check handling

[Christoph Lameter]

On Fri, 28 Nov 2014, Sebastian Ott wrote:

> would it be possible to have this one for 3.18 (if not, please add
> Cc: stable@vger.kernel.org # v3.18)

Here is my ack if it helps:

Acked-by: Christoph Lameter <cl@linux.com>

[PATCH percpu/for-3.18-fixes] s390: fix machine check regression caused by the conversion from __get_cpu_var() to this_cpu_ptr()

[Tejun Heo]

>From 1bb43105791dae7e9d4d1b1047729b12c5d5c3be Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Fri, 28 Nov 2014 16:03:14 +0100

Commit eb7e7d766326 ("s390: Replace __get_cpu_var uses") broke machine
check handling.

We copy machine check information from per-cpu to a stack variable for
local processing. Next we should zap the per-cpu variable, not the
stack variable.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Christoph Lameter <cl@linux.com>
Fixes: eb7e7d766326 ("s390: Replace __get_cpu_var uses")
Signed-off-by: Tejun Heo <tj@kernel.org>
---
Sorry about the delay.  Was on vacation.

Patch applied to percpu/for-3.18-fixes w/ minor updates to the patch
subject and description.  Will push out to Linus soon.

Thanks.

 arch/s390/kernel/nmi.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index dd1c24c..3f51cf4 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -54,12 +54,8 @@ void s390_handle_mcck(void)
 	 */
 	local_irq_save(flags);
 	local_mcck_disable();
-	/*
-	 * Ummm... Does this make sense at all? Copying the percpu struct
-	 * and then zapping it one statement later?
-	 */
-	memcpy(&mcck, this_cpu_ptr(&cpu_mcck), sizeof(mcck));
-	memset(&mcck, 0, sizeof(struct mcck_struct));
+	mcck = *this_cpu_ptr(&cpu_mcck);
+	memset(this_cpu_ptr(&cpu_mcck), 0, sizeof(mcck));
 	clear_cpu_flag(CIF_MCCK_PENDING);
 	local_mcck_enable();
 	local_irq_restore(flags);
-- 
2.1.0

Re: [PATCH percpu/for-3.18-fixes] s390: fix machine check regression caused by the conversion from __get_cpu_var() to this_cpu_ptr()

[Sebastian Ott]

On Tue, 2 Dec 2014, Tejun Heo wrote:
> Sorry about the delay.  Was on vacation.
> 
> Patch applied to percpu/for-3.18-fixes w/ minor updates to the patch
> subject and description.  Will push out to Linus soon.
> 

I've seen that Martin took the patch and it's already in Linus' tree -
commit 2cb4a18262fd0108cb8abd875710c59d0aa66f1d

Regards,
Sebastian

> Thanks.
> 
>  arch/s390/kernel/nmi.c | 8 ++------
>  1 file changed, 2 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
> index dd1c24c..3f51cf4 100644
> --- a/arch/s390/kernel/nmi.c
> +++ b/arch/s390/kernel/nmi.c
> @@ -54,12 +54,8 @@ void s390_handle_mcck(void)
>  	 */
>  	local_irq_save(flags);
>  	local_mcck_disable();
> -	/*
> -	 * Ummm... Does this make sense at all? Copying the percpu struct
> -	 * and then zapping it one statement later?
> -	 */
> -	memcpy(&mcck, this_cpu_ptr(&cpu_mcck), sizeof(mcck));
> -	memset(&mcck, 0, sizeof(struct mcck_struct));
> +	mcck = *this_cpu_ptr(&cpu_mcck);
> +	memset(this_cpu_ptr(&cpu_mcck), 0, sizeof(mcck));
>  	clear_cpu_flag(CIF_MCCK_PENDING);
>  	local_mcck_enable();
>  	local_irq_restore(flags);
> -- 
> 2.1.0
> 
> 

Re: [PATCH percpu/for-3.18-fixes] s390: fix machine check regression caused by the conversion from __get_cpu_var() to this_cpu_ptr()

[Tejun Heo]

On Tue, Dec 02, 2014 at 06:32:28PM +0100, Sebastian Ott wrote:
> On Tue, 2 Dec 2014, Tejun Heo wrote:
> > Sorry about the delay.  Was on vacation.
> > 
> > Patch applied to percpu/for-3.18-fixes w/ minor updates to the patch
> > subject and description.  Will push out to Linus soon.
> > 
> 
> I've seen that Martin took the patch and it's already in Linus' tree -
> commit 2cb4a18262fd0108cb8abd875710c59d0aa66f1d

Ah, okay.  Dropping from percpu/for-3.18-fixes then.

Thanks.

-- 
tejun