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

Groups > linux.kernel > #1346299 > unrolled thread

[PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI handler

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 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI handler Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com> - 2016-03-01 03:20 +0100
    Re: [PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI  handler Guenter Roeck <linux@roeck-us.net> - 2016-03-01 04:00 +0100
      RE: [PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI  handler 河合英宏 / KAWAI，HIDEHIRO   <hidehiro.kawai.ez@hitachi.com> - 2016-03-01 05:00 +0100

#1346299 — [PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI handler

commit 58c5661f2144 ("panic, x86: Allow CPUs to save registers even
if looping in NMI context") introduced nmi_panic() which prevents
concurrent/recursive execution of panic().  It also saves registers
for the crash dump on x86.

hpwdt driver can call panic() from NMI handler, so replace it with
nmi_panic().

Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
Cc: Thomas Mingarelli <thomas.mingarelli@hpe.com>
Cc: Wim Van Sebroeck <wim@iguana.be>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: linux-watchdog@vger.kernel.org
---
 drivers/watchdog/hpwdt.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/watchdog/hpwdt.c b/drivers/watchdog/hpwdt.c
index 92443c3..fd0486f 100644
--- a/drivers/watchdog/hpwdt.c
+++ b/drivers/watchdog/hpwdt.c
@@ -496,11 +496,12 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
 
 	if (!is_icru && !is_uefi) {
 		if (cmn_regs.u1.ral == 0) {
-			panic("An NMI occurred, "
+			nmi_panic(regs, "An NMI occurred, "
 				"but unable to determine source.\n");
+			goto handled;
 		}
 	}
-	panic("An NMI occurred. Depending on your system the reason "
+	nmi_panic(regs, "An NMI occurred. Depending on your system the reason "
 		"for the NMI is logged in any one of the following "
 		"resources:\n"
 		"1. Integrated Management Log (IML)\n"
@@ -508,6 +509,13 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
 		"3. OA Forward Progress Log\n"
 		"4. iLO Event Log");
 
+handled:
+	/*
+	 * Returning from nmi_panic() means this CPU was processing panic()
+	 * before NMI.  Return NMI_HANDLED and go back to panic routines.
+	 */
+	return NMI_HANDLED;
+
 out:
 	return NMI_DONE;
 }

[toc] | [next] | [standalone]

#1346319 — Re: [PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI handler

On 02/29/2016 05:50 PM, Hidehiro Kawai wrote:
> commit 58c5661f2144 ("panic, x86: Allow CPUs to save registers even
> if looping in NMI context") introduced nmi_panic() which prevents
> concurrent/recursive execution of panic().  It also saves registers
> for the crash dump on x86.
>
> hpwdt driver can call panic() from NMI handler, so replace it with
> nmi_panic().
>
> Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
> Cc: Thomas Mingarelli <thomas.mingarelli@hpe.com>
> Cc: Wim Van Sebroeck <wim@iguana.be>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: linux-watchdog@vger.kernel.org
> ---
>   drivers/watchdog/hpwdt.c |   12 ++++++++++--
>   1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/watchdog/hpwdt.c b/drivers/watchdog/hpwdt.c
> index 92443c3..fd0486f 100644
> --- a/drivers/watchdog/hpwdt.c
> +++ b/drivers/watchdog/hpwdt.c
> @@ -496,11 +496,12 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
>
>   	if (!is_icru && !is_uefi) {
>   		if (cmn_regs.u1.ral == 0) {
> -			panic("An NMI occurred, "
> +			nmi_panic(regs, "An NMI occurred, "
>   				"but unable to determine source.\n");

This message should really be in a single line.
Sure, strictly speaking that should be done in a separate patch,
but since you touch the line you might as well fix it.

> +			goto handled;

The goto is unnecessary. Just return NMI_HANDLED.

>   		}
>   	}
> -	panic("An NMI occurred. Depending on your system the reason "
> +	nmi_panic(regs, "An NMI occurred. Depending on your system the reason "
>   		"for the NMI is logged in any one of the following "
>   		"resources:\n"
>   		"1. Integrated Management Log (IML)\n"
> @@ -508,6 +509,13 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
>   		"3. OA Forward Progress Log\n"
>   		"4. iLO Event Log");
>
> +handled:
> +	/*
> +	 * Returning from nmi_panic() means this CPU was processing panic()
> +	 * before NMI.  Return NMI_HANDLED and go back to panic routines.
> +	 */

I don't think this comment adds much if any value.

> +	return NMI_HANDLED;
> +
>   out:

Any goto to this label does no longer serve a useful purpose (if it ever served one),
and should be replaced with "return NMI_DONE;"

>   	return NMI_DONE;
>   }
>
>
>

[toc] | [prev] | [next] | [standalone]

#1346350 — RE: [PATCH 3/3] hpwdt: Use nmi_panic() when kernel panics in NMI handler

Hi Guenter,

Thanks for the review.

> On 02/29/2016 05:50 PM, Hidehiro Kawai wrote:
> > commit 58c5661f2144 ("panic, x86: Allow CPUs to save registers even
> > if looping in NMI context") introduced nmi_panic() which prevents
> > concurrent/recursive execution of panic().  It also saves registers
> > for the crash dump on x86.
> >
> > hpwdt driver can call panic() from NMI handler, so replace it with
> > nmi_panic().
> >
> > Signed-off-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
> > Cc: Thomas Mingarelli <thomas.mingarelli@hpe.com>
> > Cc: Wim Van Sebroeck <wim@iguana.be>
> > Cc: Guenter Roeck <linux@roeck-us.net>
> > Cc: linux-watchdog@vger.kernel.org
> > ---
> >   drivers/watchdog/hpwdt.c |   12 ++++++++++--
> >   1 file changed, 10 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/watchdog/hpwdt.c b/drivers/watchdog/hpwdt.c
> > index 92443c3..fd0486f 100644
> > --- a/drivers/watchdog/hpwdt.c
> > +++ b/drivers/watchdog/hpwdt.c
> > @@ -496,11 +496,12 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
> >
> >   	if (!is_icru && !is_uefi) {
> >   		if (cmn_regs.u1.ral == 0) {
> > -			panic("An NMI occurred, "
> > +			nmi_panic(regs, "An NMI occurred, "
> >   				"but unable to determine source.\n");
> 
> This message should really be in a single line.
> Sure, strictly speaking that should be done in a separate patch,
> but since you touch the line you might as well fix it.

OK, I'll combine them.

> > +			goto handled;
> 
> The goto is unnecessary. Just return NMI_HANDLED.

Sure, I'll do that.

> 
> >   		}
> >   	}
> > -	panic("An NMI occurred. Depending on your system the reason "
> > +	nmi_panic(regs, "An NMI occurred. Depending on your system the reason "
> >   		"for the NMI is logged in any one of the following "
> >   		"resources:\n"
> >   		"1. Integrated Management Log (IML)\n"
> > @@ -508,6 +509,13 @@ static int hpwdt_pretimeout(unsigned int ulReason, struct pt_regs *regs)
> >   		"3. OA Forward Progress Log\n"
> >   		"4. iLO Event Log");
> >
> > +handled:
> > +	/*
> > +	 * Returning from nmi_panic() means this CPU was processing panic()
> > +	 * before NMI.  Return NMI_HANDLED and go back to panic routines.
> > +	 */
> 
> I don't think this comment adds much if any value.

So, I remove this comment.  NMI_HANDLED would be self-explanatory.

> > +	return NMI_HANDLED;
> > +
> >   out:
> 
> Any goto to this label does no longer serve a useful purpose (if it ever served one),
> and should be replaced with "return NMI_DONE;"

I'll fix it.  It will become a bit simpler.

Thanks,

--
Hidehiro Kawai
Hitachi, Ltd. Research & Development Group

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web