A positive value of rcupdate.rcu_task_stall_timeout is an interval
in seconds rather than jiffies.
Signed-off-by: Zhenzhong Duan <[email protected]>
---
Documentation/RCU/stallwarn.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Documentation/RCU/stallwarn.txt b/Documentation/RCU/stallwarn.txt
index 1ab70c3..13e88fc 100644
--- a/Documentation/RCU/stallwarn.txt
+++ b/Documentation/RCU/stallwarn.txt
@@ -153,7 +153,7 @@ rcupdate.rcu_task_stall_timeout
This boot/sysfs parameter controls the RCU-tasks stall warning
interval. A value of zero or less suppresses RCU-tasks stall
warnings. A positive value sets the stall-warning interval
- in jiffies. An RCU-tasks stall warning starts with the line:
+ in seconds. An RCU-tasks stall warning starts with the line:
INFO: rcu_tasks detected stalls on tasks:
--
1.8.3.1
As stated in "Documentation/lockup-watchdogs.txt:line 22", the default
behaivor after 'hardlockup' is to stay locked up rather than panic.
Signed-off-by: Zhenzhong Duan <[email protected]>
---
Documentation/admin-guide/kernel-parameters.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 2b8ee90..fcc9579 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -2769,7 +2769,7 @@
0 - turn hardlockup detector in nmi_watchdog off
1 - turn hardlockup detector in nmi_watchdog on
When panic is specified, panic when an NMI watchdog
- timeout occurs (or 'nopanic' to override the opposite
+ timeout occurs (or 'nopanic' which is the opposite
default). To disable both hard and soft lockup detectors,
please see 'nowatchdog'.
This is useful when you use a panic=... timeout and
--
1.8.3.1
On Sun, Apr 14, 2019 at 11:11:03AM +0800, Zhenzhong Duan wrote:
> A positive value of rcupdate.rcu_task_stall_timeout is an interval
> in seconds rather than jiffies.
>
> Signed-off-by: Zhenzhong Duan <[email protected]>
Good catch, applied, thank you!
Your other patch I must leave to others.
Thanx, Paul
> ---
> Documentation/RCU/stallwarn.txt | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/Documentation/RCU/stallwarn.txt b/Documentation/RCU/stallwarn.txt
> index 1ab70c3..13e88fc 100644
> --- a/Documentation/RCU/stallwarn.txt
> +++ b/Documentation/RCU/stallwarn.txt
> @@ -153,7 +153,7 @@ rcupdate.rcu_task_stall_timeout
> This boot/sysfs parameter controls the RCU-tasks stall warning
> interval. A value of zero or less suppresses RCU-tasks stall
> warnings. A positive value sets the stall-warning interval
> - in jiffies. An RCU-tasks stall warning starts with the line:
> + in seconds. An RCU-tasks stall warning starts with the line:
>
> INFO: rcu_tasks detected stalls on tasks:
>
> --
> 1.8.3.1
>
On 2019/4/14 11:11, Zhenzhong Duan wrote:
> As stated in "Documentation/lockup-watchdogs.txt:line 22", the default
> behaivor after 'hardlockup' is to stay locked up rather than panic.
>
> Signed-off-by: Zhenzhong Duan <[email protected]>
> ---
> Documentation/admin-guide/kernel-parameters.txt | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index 2b8ee90..fcc9579 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -2769,7 +2769,7 @@
> 0 - turn hardlockup detector in nmi_watchdog off
> 1 - turn hardlockup detector in nmi_watchdog on
> When panic is specified, panic when an NMI watchdog
> - timeout occurs (or 'nopanic' to override the opposite
> + timeout occurs (or 'nopanic' which is the opposite
> default). To disable both hard and soft lockup detectors,
> please see 'nowatchdog'.
> This is useful when you use a panic=... timeout and
Hi Maintainers,
Any comment?
Thanks
Zhenzhong
As nobody else commented, I will ;)
Hi Zhenzhong!
On Sun, 14 Apr 2019 11:11:04 +0800
Zhenzhong Duan <[email protected]> wrote:
> As stated in "Documentation/lockup-watchdogs.txt:line 22", the default
> behaivor after 'hardlockup' is to stay locked up rather than panic.
That actually says:
A 'hardlockup' is defined as a bug that causes the CPU to loop in
kernel mode for more than 10 seconds (see "Implementation" below for
details), without letting other interrupts have a chance to run.
Similarly to the softlockup case, the current stack trace is displayed
upon detection and the system will stay locked up unless the default
behavior is changed, which can be done through a sysctl,
'hardlockup_panic', a compile time knob, "BOOTPARAM_HARDLOCKUP_PANIC",
and a kernel parameter, "nmi_watchdog"
If your config has:
CONFIG_BOOTPARAM_HARDLOCKUP_PANIC=y
The kernel will panic on hard lockup by default unless you add nopanic.
If your config has:
# CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is not set
Then the default will be not to panic unless you add "panic" to the
kernel command line.
>
> Signed-off-by: Zhenzhong Duan <[email protected]>
> ---
> Documentation/admin-guide/kernel-parameters.txt | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index 2b8ee90..fcc9579 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -2769,7 +2769,7 @@
> 0 - turn hardlockup detector in nmi_watchdog off
> 1 - turn hardlockup detector in nmi_watchdog on
> When panic is specified, panic when an NMI watchdog
> - timeout occurs (or 'nopanic' to override the opposite
> + timeout occurs (or 'nopanic' which is the opposite
> default). To disable both hard and soft lockup detectors,
Honestly, I think the original text states what it does better than
your update. Because the nopanic is added to override the "opposite
default" which is if the config was set to do so.
That said, this all still can be explained better. What about:
nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
Format: [panic,][nopanic,][num]
Valid num: 0 or 1
0 - turn hardlockup detector in nmi_watchdog off
1 - turn hardlockup detector in nmi_watchdog on
When panic is specified, panic when an NMI watchdog
timeout occurs (or 'nopanic' to not panic on an NMI
watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
To disable both hard and soft lockup detectors,
please see 'nowatchdog'.
This is useful when you use a panic=... timeout and
need the box quickly up again.
-- Steve
> please see 'nowatchdog'.
> This is useful when you use a panic=... timeout and
Hi Steven,
On 2019/5/11 2:47, Steven Rostedt wrote:
> As nobody else commented, I will ;)
>
> Hi Zhenzhong!
>
> On Sun, 14 Apr 2019 11:11:04 +0800
> Zhenzhong Duan <[email protected]> wrote:
>
>> As stated in "Documentation/lockup-watchdogs.txt:line 22", the default
>> behaivor after 'hardlockup' is to stay locked up rather than panic.
> That actually says:
>
> A 'hardlockup' is defined as a bug that causes the CPU to loop in
> kernel mode for more than 10 seconds (see "Implementation" below for
> details), without letting other interrupts have a chance to run.
> Similarly to the softlockup case, the current stack trace is displayed
> upon detection and the system will stay locked up unless the default
> behavior is changed, which can be done through a sysctl,
> 'hardlockup_panic', a compile time knob, "BOOTPARAM_HARDLOCKUP_PANIC",
> and a kernel parameter, "nmi_watchdog"
>
> If your config has:
>
> CONFIG_BOOTPARAM_HARDLOCKUP_PANIC=y
>
> The kernel will panic on hard lockup by default unless you add nopanic.
>
> If your config has:
>
> # CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is not set
>
> Then the default will be not to panic unless you add "panic" to the
> kernel command line.
>
>> Signed-off-by: Zhenzhong Duan <[email protected]>
>> ---
>> Documentation/admin-guide/kernel-parameters.txt | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
>> index 2b8ee90..fcc9579 100644
>> --- a/Documentation/admin-guide/kernel-parameters.txt
>> +++ b/Documentation/admin-guide/kernel-parameters.txt
>> @@ -2769,7 +2769,7 @@
>> 0 - turn hardlockup detector in nmi_watchdog off
>> 1 - turn hardlockup detector in nmi_watchdog on
>> When panic is specified, panic when an NMI watchdog
>> - timeout occurs (or 'nopanic' to override the opposite
>> + timeout occurs (or 'nopanic' which is the opposite
>> default). To disable both hard and soft lockup detectors,
> Honestly, I think the original text states what it does better than
> your update. Because the nopanic is added to override the "opposite
> default" which is if the config was set to do so.
>
> That said, this all still can be explained better. What about:
>
> nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
> Format: [panic,][nopanic,][num]
> Valid num: 0 or 1
> 0 - turn hardlockup detector in nmi_watchdog off
> 1 - turn hardlockup detector in nmi_watchdog on
> When panic is specified, panic when an NMI watchdog
> timeout occurs (or 'nopanic' to not panic on an NMI
> watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
> To disable both hard and soft lockup detectors,
> please see 'nowatchdog'.
> This is useful when you use a panic=... timeout and
> need the box quickly up again.
Thanks Seeve for your comments, so the default behavior of hardlockup
depends on the setting of CONFIG_BOOTPARAM_HARDLOCKUP_PANIC.
It did confused me previously. You suggested fix is better, I'll fix it
later.
Zhenzhong
On Fri, May 10, 2019 at 02:47:49PM -0400, Steven Rostedt wrote:
>
> As nobody else commented, I will ;)
>
> Hi Zhenzhong!
>
> On Sun, 14 Apr 2019 11:11:04 +0800
> Zhenzhong Duan <[email protected]> wrote:
>
> > As stated in "Documentation/lockup-watchdogs.txt:line 22", the default
> > behaivor after 'hardlockup' is to stay locked up rather than panic.
>
> That actually says:
>
> A 'hardlockup' is defined as a bug that causes the CPU to loop in
> kernel mode for more than 10 seconds (see "Implementation" below for
> details), without letting other interrupts have a chance to run.
> Similarly to the softlockup case, the current stack trace is displayed
> upon detection and the system will stay locked up unless the default
> behavior is changed, which can be done through a sysctl,
> 'hardlockup_panic', a compile time knob, "BOOTPARAM_HARDLOCKUP_PANIC",
> and a kernel parameter, "nmi_watchdog"
>
> If your config has:
>
> CONFIG_BOOTPARAM_HARDLOCKUP_PANIC=y
>
> The kernel will panic on hard lockup by default unless you add nopanic.
>
> If your config has:
>
> # CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is not set
>
> Then the default will be not to panic unless you add "panic" to the
> kernel command line.
>
> >
> > Signed-off-by: Zhenzhong Duan <[email protected]>
> > ---
> > Documentation/admin-guide/kernel-parameters.txt | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> > index 2b8ee90..fcc9579 100644
> > --- a/Documentation/admin-guide/kernel-parameters.txt
> > +++ b/Documentation/admin-guide/kernel-parameters.txt
> > @@ -2769,7 +2769,7 @@
> > 0 - turn hardlockup detector in nmi_watchdog off
> > 1 - turn hardlockup detector in nmi_watchdog on
> > When panic is specified, panic when an NMI watchdog
> > - timeout occurs (or 'nopanic' to override the opposite
> > + timeout occurs (or 'nopanic' which is the opposite
> > default). To disable both hard and soft lockup detectors,
>
> Honestly, I think the original text states what it does better than
> your update. Because the nopanic is added to override the "opposite
> default" which is if the config was set to do so.
>
> That said, this all still can be explained better. What about:
>
> nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels
> Format: [panic,][nopanic,][num]
> Valid num: 0 or 1
> 0 - turn hardlockup detector in nmi_watchdog off
> 1 - turn hardlockup detector in nmi_watchdog on
> When panic is specified, panic when an NMI watchdog
> timeout occurs (or 'nopanic' to not panic on an NMI
> watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set)
> To disable both hard and soft lockup detectors,
> please see 'nowatchdog'.
> This is useful when you use a panic=... timeout and
> need the box quickly up again.
Steven, your update makes it more clear to me. If patch is being written with
such update, here is my Reviewed-by:
Reviewed-by: Joel Fernandes (Google) <[email protected]>
thanks!
- Joel
>
> -- Steve
>
>
> > please see 'nowatchdog'.
> > This is useful when you use a panic=... timeout and
>