[RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

Subsystems: the rest

STALE2678d

Revision v2 of 4 in this series.

5 messages, 3 authors, 2019-04-09 · open the first message on its own page

[RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

From: Ricardo Neri <hidden>
Date: 2019-02-27 16:06:33

CPU architectures that have an NMI watchdog use arch_touch_nmi_watchdog()
to briefly ignore the hardlockup detector. If the architecture does not
have an NMI watchdog, one can be constructed using a source of non-
maskable interrupts. In this case, arch_touch_nmi_watchdog() is common
to any underlying hardware resource used to drive the detector and needs
to be available to other kernel subsystems if hardware different from perf
drives the detector.

There exists perf-based and HPET-based implementations. Make it available
to the latter.

For clarity, wrap this function in a separate preprocessor conditional
from functions which are truly specific to the perf-based implementation.

Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ashok Raj <redacted>
Cc: Andi Kleen <redacted>
Cc: Tony Luck <tony.luck@intel.com>
Cc: "Rafael J. Wysocki" <redacted>
Cc: Don Zickus <redacted>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Babu Moger <redacted>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <redacted>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Philippe Ombredanne <redacted>
Cc: Colin Ian King <redacted>
Cc: Byungchul Park <redacted>
Cc: "Paul E. McKenney" <redacted>
Cc: "Luis R. Rodriguez" <mcgrof@kernel.org>
Cc: Waiman Long <longman@redhat.com>
Cc: Josh Poimboeuf <redacted>
Cc: Randy Dunlap <redacted>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Christoffer Dall <redacted>
Cc: Marc Zyngier <redacted>
Cc: Kai-Heng Feng <redacted>
Cc: Konrad Rzeszutek Wilk <redacted>
Cc: David Rientjes <rientjes@google.com>
Cc: "Ravi V. Shankar" <redacted>
Cc: x86@kernel.org
Cc: sparclinux@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Ricardo Neri <redacted>
---
 include/linux/nmi.h | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/include/linux/nmi.h b/include/linux/nmi.h
index 5a8b19749769..bf5ebcfdd590 100644
--- a/include/linux/nmi.h
+++ b/include/linux/nmi.h

@@ -94,8 +94,16 @@ static inline void hardlockup_detector_disable(void) {}
 # define NMI_WATCHDOG_SYSCTL_PERM	0444
 #endif
 
-#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF) || \
+    defined(CONFIG_X86_HARDLOCKUP_DETECTOR_HPET)
 extern void arch_touch_nmi_watchdog(void);
+#else
+# if !defined(CONFIG_HAVE_NMI_WATCHDOG)
+static inline void arch_touch_nmi_watchdog(void) {}
+# endif
+#endif
+
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
 extern void hardlockup_detector_perf_stop(void);
 extern void hardlockup_detector_perf_restart(void);
 extern void hardlockup_detector_perf_disable(void);

-- 
2.17.1

Re: [RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

From: Paul E. McKenney <hidden>
Date: 2019-02-27 16:18:16

On Wed, Feb 27, 2019 at 08:05:13AM -0800, Ricardo Neri wrote:

quoted hunk

CPU architectures that have an NMI watchdog use arch_touch_nmi_watchdog()
to briefly ignore the hardlockup detector. If the architecture does not
have an NMI watchdog, one can be constructed using a source of non-
maskable interrupts. In this case, arch_touch_nmi_watchdog() is common
to any underlying hardware resource used to drive the detector and needs
to be available to other kernel subsystems if hardware different from perf
drives the detector.

There exists perf-based and HPET-based implementations. Make it available
to the latter.

For clarity, wrap this function in a separate preprocessor conditional
from functions which are truly specific to the perf-based implementation.

Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ashok Raj <redacted>
Cc: Andi Kleen <redacted>
Cc: Tony Luck <tony.luck@intel.com>
Cc: "Rafael J. Wysocki" <redacted>
Cc: Don Zickus <redacted>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Babu Moger <redacted>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <redacted>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Philippe Ombredanne <redacted>
Cc: Colin Ian King <redacted>
Cc: Byungchul Park <redacted>
Cc: "Paul E. McKenney" <redacted>
Cc: "Luis R. Rodriguez" <mcgrof@kernel.org>
Cc: Waiman Long <longman@redhat.com>
Cc: Josh Poimboeuf <redacted>
Cc: Randy Dunlap <redacted>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Christoffer Dall <redacted>
Cc: Marc Zyngier <redacted>
Cc: Kai-Heng Feng <redacted>
Cc: Konrad Rzeszutek Wilk <redacted>
Cc: David Rientjes <rientjes@google.com>
Cc: "Ravi V. Shankar" <redacted>
Cc: x86@kernel.org
Cc: sparclinux@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Ricardo Neri <redacted>
---
 include/linux/nmi.h | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/include/linux/nmi.h b/include/linux/nmi.h
index 5a8b19749769..bf5ebcfdd590 100644
--- a/include/linux/nmi.h
+++ b/include/linux/nmi.h

@@ -94,8 +94,16 @@ static inline void hardlockup_detector_disable(void) {}
 # define NMI_WATCHDOG_SYSCTL_PERM	0444
 #endif

-#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF) || \
+    defined(CONFIG_X86_HARDLOCKUP_DETECTOR_HPET)

Why not instead make CONFIG_X86_HARDLOCKUP_DETECTOR_HPET select
CONFIG_HARDLOCKUP_DETECTOR_PERF?  Keep the arch-specific details
in the arch-specific files and all that.

							Thanx, Paul

 extern void arch_touch_nmi_watchdog(void);
+#else
+# if !defined(CONFIG_HAVE_NMI_WATCHDOG)
+static inline void arch_touch_nmi_watchdog(void) {}
+# endif
+#endif
+
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
 extern void hardlockup_detector_perf_stop(void);
 extern void hardlockup_detector_perf_restart(void);
 extern void hardlockup_detector_perf_disable(void);
-- 
2.17.1

Re: [RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

From: Ricardo Neri <hidden>
Date: 2019-03-01 01:17:21

On Wed, Feb 27, 2019 at 08:17:58AM -0800, Paul E. McKenney wrote:

On Wed, Feb 27, 2019 at 08:05:13AM -0800, Ricardo Neri wrote:

quoted

CPU architectures that have an NMI watchdog use arch_touch_nmi_watchdog()
to briefly ignore the hardlockup detector. If the architecture does not
have an NMI watchdog, one can be constructed using a source of non-
maskable interrupts. In this case, arch_touch_nmi_watchdog() is common
to any underlying hardware resource used to drive the detector and needs
to be available to other kernel subsystems if hardware different from perf
drives the detector.

There exists perf-based and HPET-based implementations. Make it available
to the latter.

For clarity, wrap this function in a separate preprocessor conditional
from functions which are truly specific to the perf-based implementation.

Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ashok Raj <redacted>
Cc: Andi Kleen <redacted>
Cc: Tony Luck <tony.luck@intel.com>
Cc: "Rafael J. Wysocki" <redacted>
Cc: Don Zickus <redacted>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Babu Moger <redacted>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <redacted>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Philippe Ombredanne <redacted>
Cc: Colin Ian King <redacted>
Cc: Byungchul Park <redacted>
Cc: "Paul E. McKenney" <redacted>
Cc: "Luis R. Rodriguez" <mcgrof@kernel.org>
Cc: Waiman Long <longman@redhat.com>
Cc: Josh Poimboeuf <redacted>
Cc: Randy Dunlap <redacted>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Christoffer Dall <redacted>
Cc: Marc Zyngier <redacted>
Cc: Kai-Heng Feng <redacted>
Cc: Konrad Rzeszutek Wilk <redacted>
Cc: David Rientjes <rientjes@google.com>
Cc: "Ravi V. Shankar" <redacted>
Cc: x86@kernel.org
Cc: sparclinux@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Signed-off-by: Ricardo Neri <redacted>
---
 include/linux/nmi.h | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/include/linux/nmi.h b/include/linux/nmi.h
index 5a8b19749769..bf5ebcfdd590 100644
--- a/include/linux/nmi.h
+++ b/include/linux/nmi.h

@@ -94,8 +94,16 @@ static inline void hardlockup_detector_disable(void) {}
 # define NMI_WATCHDOG_SYSCTL_PERM	0444
 #endif

-#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF) || \
+    defined(CONFIG_X86_HARDLOCKUP_DETECTOR_HPET)

Why not instead make CONFIG_X86_HARDLOCKUP_DETECTOR_HPET select
CONFIG_HARDLOCKUP_DETECTOR_PERF?  Keep the arch-specific details
in the arch-specific files and all that.

Thanks for your feedback, Paul! The HPET implementation does not use
perf. Thus, in my opinion is not correct for the HPET HLD to select
the perf implementation. Patch 8 of this series splits the perf-specific
code and the generic hardlockup detector code. Does this make sense?

Thanks and BR,
Ricardo

Re: [RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

From: Thomas Gleixner <hidden>
Date: 2019-03-26 21:21:33

On Thu, 28 Feb 2019, Ricardo Neri wrote:

quoted

-#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF) || \
+    defined(CONFIG_X86_HARDLOCKUP_DETECTOR_HPET)

Why not instead make CONFIG_X86_HARDLOCKUP_DETECTOR_HPET select
CONFIG_HARDLOCKUP_DETECTOR_PERF?  Keep the arch-specific details
in the arch-specific files and all that.

Thanks for your feedback, Paul! The HPET implementation does not use
perf. Thus, in my opinion is not correct for the HPET HLD to select
the perf implementation. Patch 8 of this series splits the perf-specific
code and the generic hardlockup detector code. Does this make sense?

That's what intermediate config symbols are for.

config HARDLOCKUP_DETECTOR_CORE
       bool

And make both PERF and HPET select it.

Thanks,

	tglx

Re: [RFC PATCH v2 09/14] watchdog/hardlockup: Make arch_touch_nmi_watchdog() to hpet-based implementation

From: Ricardo Neri <hidden>
Date: 2019-04-09 02:06:43

On Tue, Mar 26, 2019 at 10:20:41PM +0100, Thomas Gleixner wrote:

On Thu, 28 Feb 2019, Ricardo Neri wrote:

quoted

-#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF) || \
+    defined(CONFIG_X86_HARDLOCKUP_DETECTOR_HPET)

Why not instead make CONFIG_X86_HARDLOCKUP_DETECTOR_HPET select
CONFIG_HARDLOCKUP_DETECTOR_PERF?  Keep the arch-specific details
in the arch-specific files and all that.

Thanks for your feedback, Paul! The HPET implementation does not use
perf. Thus, in my opinion is not correct for the HPET HLD to select
the perf implementation. Patch 8 of this series splits the perf-specific
code and the generic hardlockup detector code. Does this make sense?

That's what intermediate config symbols are for.

config HARDLOCKUP_DETECTOR_CORE
       bool

And make both PERF and HPET select it.

I'll implement it in this manner.

Thanks and BR,
Ricardo

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help