Skip to contentSkip to conversation

FAILED: patch "[PATCH] ARM: imx6q: cpuidle: fix bug that CPU might not wake up at" failed to apply to 3.18-stable tree

PATCH [email protected] 2 participants 3 messages Mar 29, 2019 – Apr 1, 2019

Subsystems (4)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 3.18-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 91740fc8242b4f260cfa4d4536d8551804777fae Mon Sep 17 00:00:00 2001
From: Kohji Okuno <[email protected]>
Date: Tue, 26 Feb 2019 11:34:13 +0900
Subject: [PATCH] ARM: imx6q: cpuidle: fix bug that CPU might not wake up at
expected time

In the current cpuidle implementation for i.MX6q, the CPU that sets 'WAIT_UNCLOCKED' and the CPU that returns to 'WAIT_CLOCKED' are always the same. While the CPU that sets 'WAIT_UNCLOCKED' is in IDLE state of "WAIT", if the other CPU wakes up and enters IDLE state of "WFI" istead of "WAIT", this CPU can not wake up at expired time.
Because, in the case of "WFI", the CPU must be waked up by the local
timer interrupt. But, while 'WAIT_UNCLOCKED' is set, the local timer is stopped, when all CPUs execute "wfi" instruction. As a result, the local timer interrupt is not fired.
In this situation, this CPU will wake up by IRQ different from local
timer. (e.g. broacast timer)

So, this fix changes CPU to return to 'WAIT_CLOCKED'.

Fixes e5f9dec8ff5f ("ARM: imx6q: support WAIT mode using cpuidle")
arch/arm/mach-imx/cpuidle-imx6q.c +10 -17
index bfeb25aaf9a2..326e870d7123 100644
@@ -16,30 +16,23 @@
 #include "cpuidle.h"
 #include "hardware.h"
 
-static atomic_t master = ATOMIC_INIT(0);
-static DEFINE_SPINLOCK(master_lock);
+static int num_idle_cpus = 0;
+static DEFINE_SPINLOCK(cpuidle_lock);
 
 static int imx6q_enter_wait(struct cpuidle_device *dev,
 			    struct cpuidle_driver *drv, int index)
 {
-	if (atomic_inc_return(&master) == num_online_cpus()) {
-		/*
-		 * With this lock, we prevent other cpu to exit and enter
-		 * this function again and become the master.
-		 */
-		if (!spin_trylock(&master_lock))
-			goto idle;
+	spin_lock(&cpuidle_lock);
+	if (++num_idle_cpus == num_online_cpus())
 		imx6_set_lpm(WAIT_UNCLOCKED);
-		cpu_do_idle();
-		imx6_set_lpm(WAIT_CLOCKED);
-		spin_unlock(&master_lock);
-		goto done;
-	}
+	spin_unlock(&cpuidle_lock);
 
-idle:
 	cpu_do_idle();
-done:
-	atomic_dec(&master);
+
+	spin_lock(&cpuidle_lock);
+	if (num_idle_cpus-- == num_online_cpus())
+		imx6_set_lpm(WAIT_CLOCKED);
+	spin_unlock(&cpuidle_lock);
 
 	return index;
 }
#1 lore raw

Hi Greg and Shawn,

Quoted 4 lines

The patch below does not apply to the 3.18-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

Could you please confirm?

Regards,
Kohji Okuno

-------------------------------------------------------------------------

From 6000156db44bbdae0bd720c46e68a2622639d9d8 Mon Sep 17 00:00:00 2001
From: Kohji Okuno <[email protected]>
Date: Mon, 1 Apr 2019 14:22:02 +0900
Subject: [PATCH] ARM: imx6q: cpuidle: fix bug that CPU might not wake up at
expected time

commit 91740fc8242b4f260cfa4d4536d8551804777fae upstream.

In the current cpuidle implementation for i.MX6q, the CPU that sets 'WAIT_UNCLOCKED' and the CPU that returns to 'WAIT_CLOCKED' are always the same. While the CPU that sets 'WAIT_UNCLOCKED' is in IDLE state of "WAIT", if the other CPU wakes up and enters IDLE state of "WFI" istead of "WAIT", this CPU can not wake up at expired time.
Because, in the case of "WFI", the CPU must be waked up by the local
timer interrupt. But, while 'WAIT_UNCLOCKED' is set, the local timer is stopped, when all CPUs execute "wfi" instruction. As a result, the local timer interrupt is not fired.
In this situation, this CPU will wake up by IRQ different from local
timer. (e.g. broacast timer)

So, this fix changes CPU to return to 'WAIT_CLOCKED'.

Fixes e5f9dec8ff5f ("ARM: imx6q: support WAIT mode using cpuidle")

[okuno.kohji: backport to 3.18]

---
 arch/arm/mach-imx/cpuidle-imx6q.c | 27 ++++++++++-----------------
 1 file changed, 10 insertions(+), 17 deletions(-)
arch/arm/mach-imx/cpuidle-imx6q.c +10 -17
index aa935787b743..9d219b626949 100644
@@ -15,30 +15,23 @@
 #include "cpuidle.h"
 #include "hardware.h"
 
-static atomic_t master = ATOMIC_INIT(0);
-static DEFINE_SPINLOCK(master_lock);
+static int num_idle_cpus = 0;
+static DEFINE_SPINLOCK(cpuidle_lock);
 
 static int imx6q_enter_wait(struct cpuidle_device *dev,
 			    struct cpuidle_driver *drv, int index)
 {
-	if (atomic_inc_return(&master) == num_online_cpus()) {
-		/*
-		 * With this lock, we prevent other cpu to exit and enter
-		 * this function again and become the master.
-		 */
-		if (!spin_trylock(&master_lock))
-			goto idle;
+	spin_lock(&cpuidle_lock);
+	if (++num_idle_cpus == num_online_cpus())
 		imx6q_set_lpm(WAIT_UNCLOCKED);
-		cpu_do_idle();
-		imx6q_set_lpm(WAIT_CLOCKED);
-		spin_unlock(&master_lock);
-		goto done;
-	}
+	spin_unlock(&cpuidle_lock);
 
-idle:
 	cpu_do_idle();
-done:
-	atomic_dec(&master);
+
+	spin_lock(&cpuidle_lock);
+	if (num_idle_cpus-- == num_online_cpus())
+		imx6q_set_lpm(WAIT_CLOCKED);
+	spin_unlock(&cpuidle_lock);
 
 	return index;
 }
Signature · 1 line
2.17.1

From 6000156db44bbdae0bd720c46e68a2622639d9d8 Mon Sep 17 00:00:00 2001
From: Kohji Okuno <[email protected]>
Date: Mon, 1 Apr 2019 14:22:02 +0900
Subject: [PATCH] ARM: imx6q: cpuidle: fix bug that CPU might not wake up at
expected time

commit 91740fc8242b4f260cfa4d4536d8551804777fae upstream.

In the current cpuidle implementation for i.MX6q, the CPU that sets 'WAIT_UNCLOCKED' and the CPU that returns to 'WAIT_CLOCKED' are always the same. While the CPU that sets 'WAIT_UNCLOCKED' is in IDLE state of "WAIT", if the other CPU wakes up and enters IDLE state of "WFI" istead of "WAIT", this CPU can not wake up at expired time.
Because, in the case of "WFI", the CPU must be waked up by the local
timer interrupt. But, while 'WAIT_UNCLOCKED' is set, the local timer is stopped, when all CPUs execute "wfi" instruction. As a result, the local timer interrupt is not fired.
In this situation, this CPU will wake up by IRQ different from local
timer. (e.g. broacast timer)

So, this fix changes CPU to return to 'WAIT_CLOCKED'.

Fixes e5f9dec8ff5f ("ARM: imx6q: support WAIT mode using cpuidle")

[okuno.kohji: backport to 3.18]

---
 arch/arm/mach-imx/cpuidle-imx6q.c | 27 ++++++++++-----------------
 1 file changed, 10 insertions(+), 17 deletions(-)
arch/arm/mach-imx/cpuidle-imx6q.c +10 -17
index aa935787b743..9d219b626949 100644
@@ -15,30 +15,23 @@
 #include "cpuidle.h"
 #include "hardware.h"
 
-static atomic_t master = ATOMIC_INIT(0);
-static DEFINE_SPINLOCK(master_lock);
+static int num_idle_cpus = 0;
+static DEFINE_SPINLOCK(cpuidle_lock);
 
 static int imx6q_enter_wait(struct cpuidle_device *dev,
 			    struct cpuidle_driver *drv, int index)
 {
-	if (atomic_inc_return(&master) == num_online_cpus()) {
-		/*
-		 * With this lock, we prevent other cpu to exit and enter
-		 * this function again and become the master.
-		 */
-		if (!spin_trylock(&master_lock))
-			goto idle;
+	spin_lock(&cpuidle_lock);
+	if (++num_idle_cpus == num_online_cpus())
 		imx6q_set_lpm(WAIT_UNCLOCKED);
-		cpu_do_idle();
-		imx6q_set_lpm(WAIT_CLOCKED);
-		spin_unlock(&master_lock);
-		goto done;
-	}
+	spin_unlock(&cpuidle_lock);
 
-idle:
 	cpu_do_idle();
-done:
-	atomic_dec(&master);
+
+	spin_lock(&cpuidle_lock);
+	if (num_idle_cpus-- == num_online_cpus())
+		imx6q_set_lpm(WAIT_CLOCKED);
+	spin_unlock(&cpuidle_lock);
 
 	return index;
 }
Signature · 1 line
2.17.1
author maintainer
#2 lore raw
Quoted 8 lines · Kohji Okuno

Hi Greg and Shawn,

Quoted 4 lines

The patch below does not apply to the 3.18-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

Could you please confirm?

Looks good to me, thanks, now queued up.

greg k-h