linux-arch.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
To: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linux-arch@vger.kernel.org,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	linux-kernel@vger.kernel.org, linux-acpi@vger.kernel.org,
	Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>,
	Arun Bharadwaj <arun@linux.vnet.ibm.com>,
	Ingo Molnar <mingo@elte.hu>,
	linuxppc-dev@lists.ozlabs.org
Subject: Re: [v10 PATCH 8/9]: pSeries: implement pSeries processor idle module
Date: Fri, 4 Dec 2009 13:45:39 +0530	[thread overview]
Message-ID: <20091204081539.GA5883@linux.vnet.ibm.com> (raw)
In-Reply-To: <1259894858.2076.1251.camel@pasglop>

* Benjamin Herrenschmidt <benh@kernel.crashing.org> [2009-12-04 13:47:38]:

> On Wed, 2009-12-02 at 15:32 +0530, Arun R Bharadwaj wrote:
> > * Arun R Bharadwaj <arun@linux.vnet.ibm.com> [2009-12-02 15:24:27]:
> > 
> > This patch creates arch/powerpc/platforms/pseries/processor_idle.c,
> > which implements the cpuidle infrastructure for pseries.
> > It implements a pseries_cpuidle_loop() which would be the main idle loop
> > called from cpu_idle(). It makes decision of entering either
> > dedicated_snooze_loop or dedicated_cede_loop for dedicated lpar and
> > shared_cede_loop for shared lpar processor based on the
> > decision taken by the cpuidle governor.
> 
> So unless I'm mistaken, you removed our powerpc "generic" idle loop that
> calls into ppc_md.power_save(), and replaced it by some pseries specific
> idle loops... Now what about all the other powerpc platforms ? native
> 970 (aka G5) ? 6xx ? Cell ? Or are you still calling ppc_md.power_save
> somewhere that I missed ?
> 
> Cheers,
> Ben.

Hi Ben,

I forgot to attach the patch which enables cpuidle for the rest of the
POWER platforms. Attaching it below.

So for these platforms, ppc_md.power_save will be called from from the
cpuidle_idle_call idle loop itself. Also, this cpuidle_idle_call is
not a pseries specific idle loop. It is a common loop for Intel and
PPC which use cpuidle infrastructure.

arun



This patch enables cpuidle for the rest of the POWER platforms like
44x, Cell, Pasemi etc.

Signed-off-by: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
---
 arch/powerpc/include/asm/system.h       |    2 ++
 arch/powerpc/kernel/idle.c              |   28 ++++++++++++++++++++++++++++
 arch/powerpc/kernel/setup_32.c          |    8 ++++++--
 arch/powerpc/platforms/44x/idle.c       |    2 ++
 arch/powerpc/platforms/cell/pervasive.c |    2 ++
 arch/powerpc/platforms/pasemi/idle.c    |    2 ++
 arch/powerpc/platforms/ps3/setup.c      |    2 ++
 7 files changed, 44 insertions(+), 2 deletions(-)

Index: linux.trees.git/arch/powerpc/include/asm/system.h
===================================================================
--- linux.trees.git.orig/arch/powerpc/include/asm/system.h
+++ linux.trees.git/arch/powerpc/include/asm/system.h
@@ -551,8 +551,10 @@ void cpu_idle_wait(void);
 
 #ifdef CONFIG_CPU_IDLE
 extern void update_smt_snooze_delay(int snooze);
+extern void setup_cpuidle_ppc(void);
 #else
 static inline void update_smt_snooze_delay(int snooze) {}
+static inline void setup_cpuidle_ppc(void) {}
 #endif
 
 #endif /* __KERNEL__ */
Index: linux.trees.git/arch/powerpc/kernel/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/kernel/idle.c
+++ linux.trees.git/arch/powerpc/kernel/idle.c
@@ -129,6 +129,34 @@ void default_idle(void)
 	HMT_very_low();
 }
 
+#ifdef CONFIG_CPU_IDLE
+DEFINE_PER_CPU(struct cpuidle_device, ppc_idle_devices);
+struct cpuidle_driver cpuidle_ppc_driver = {
+	.name =         "cpuidle_ppc",
+};
+
+static void ppc_idle_loop(struct cpuidle_device *dev, struct cpuidle_state *st)
+{
+	ppc_md.power_save();
+}
+
+void setup_cpuidle_ppc(void)
+{
+	struct cpuidle_device *dev;
+	int cpu;
+
+	cpuidle_register_driver(&cpuidle_ppc_driver);
+
+	for_each_online_cpu(cpu) {
+		dev = &per_cpu(ppc_idle_devices, cpu);
+		dev->cpu = cpu;
+		dev->states[0].enter = ppc_idle_loop;
+		dev->state_count = 1;
+		cpuidle_register_device(dev);
+	}
+}
+#endif
+
 int powersave_nap;
 
 #ifdef CONFIG_SYSCTL
Index: linux.trees.git/arch/powerpc/kernel/setup_32.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/kernel/setup_32.c
+++ linux.trees.git/arch/powerpc/kernel/setup_32.c
@@ -133,14 +133,18 @@ notrace void __init machine_init(unsigne
 
 #ifdef CONFIG_6xx
 	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
+	    cpu_has_feature(CPU_FTR_CAN_NAP)) {
 		ppc_md.power_save = ppc6xx_idle;
+		setup_cpuidle_ppc();
+	}
 #endif
 
 #ifdef CONFIG_E500
 	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
+	    cpu_has_feature(CPU_FTR_CAN_NAP)) {
 		ppc_md.power_save = e500_idle;
+		setup_cpuidle_ppc();
+	}
 #endif
 	if (ppc_md.progress)
 		ppc_md.progress("id mach(): done", 0x200);
Index: linux.trees.git/arch/powerpc/platforms/44x/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/44x/idle.c
+++ linux.trees.git/arch/powerpc/platforms/44x/idle.c
@@ -24,6 +24,7 @@
 #include <linux/of.h>
 #include <linux/kernel.h>
 #include <asm/machdep.h>
+#include <asm/system.h>
 
 static int mode_spin;
 
@@ -46,6 +47,7 @@ int __init ppc44x_idle_init(void)
 		/* If we are not setting spin mode 
                    then we set to wait mode */
 		ppc_md.power_save = &ppc44x_idle;
+		setup_cpuidle_ppc();
 	}
 
 	return 0;
Index: linux.trees.git/arch/powerpc/platforms/cell/pervasive.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/cell/pervasive.c
+++ linux.trees.git/arch/powerpc/platforms/cell/pervasive.c
@@ -35,6 +35,7 @@
 #include <asm/pgtable.h>
 #include <asm/reg.h>
 #include <asm/cell-regs.h>
+#include <asm/system.h>
 
 #include "pervasive.h"
 
@@ -128,5 +129,6 @@ void __init cbe_pervasive_init(void)
 	}
 
 	ppc_md.power_save = cbe_power_save;
+	setup_cpuidle_ppc();
 	ppc_md.system_reset_exception = cbe_system_reset_exception;
 }
Index: linux.trees.git/arch/powerpc/platforms/pasemi/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/pasemi/idle.c
+++ linux.trees.git/arch/powerpc/platforms/pasemi/idle.c
@@ -27,6 +27,7 @@
 #include <asm/machdep.h>
 #include <asm/reg.h>
 #include <asm/smp.h>
+#include <asm/system.h>
 
 #include "pasemi.h"
 
@@ -81,6 +82,7 @@ static int __init pasemi_idle_init(void)
 
 	ppc_md.system_reset_exception = pasemi_system_reset_exception;
 	ppc_md.power_save = modes[current_mode].entry;
+	setup_cpuidle_ppc();
 	printk(KERN_INFO "Using PA6T idle loop (%s)\n", modes[current_mode].name);
 
 	return 0;
Index: linux.trees.git/arch/powerpc/platforms/ps3/setup.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/ps3/setup.c
+++ linux.trees.git/arch/powerpc/platforms/ps3/setup.c
@@ -33,6 +33,7 @@
 #include <asm/prom.h>
 #include <asm/lv1call.h>
 #include <asm/ps3gpu.h>
+#include <asm/system.h>
 
 #include "platform.h"
 
@@ -214,6 +215,7 @@ static void __init ps3_setup_arch(void)
 	prealloc_ps3flash_bounce_buffer();
 
 	ppc_md.power_save = ps3_power_save;
+	setup_cpuidle_ppc();
 	ps3_os_area_init();
 
 	DBG(" <- %s:%d\n", __func__, __LINE__);

WARNING: multiple messages have this Message-ID (diff)
From: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
To: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Ingo Molnar <mingo@elte.hu>,
	Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>,
	Dipankar Sarma <dipankar@in.ibm.com>,
	Balbir Singh <balbir@in.ibm.com>,
	Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>,
	linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
	linux-arch@vger.kernel.org, linux-acpi@vger.kernel.org,
	Arun Bharadwaj <arun@linux.vnet.ibm.com>
Subject: Re: [v10 PATCH 8/9]: pSeries: implement pSeries processor idle module
Date: Fri, 4 Dec 2009 13:45:39 +0530	[thread overview]
Message-ID: <20091204081539.GA5883@linux.vnet.ibm.com> (raw)
Message-ID: <20091204081539.Coh_pY_2y_ZrHAYIyYn_LITzTk_p_QQ3jukr0Elho-E@z> (raw)
In-Reply-To: <1259894858.2076.1251.camel@pasglop>

* Benjamin Herrenschmidt <benh@kernel.crashing.org> [2009-12-04 13:47:38]:

> On Wed, 2009-12-02 at 15:32 +0530, Arun R Bharadwaj wrote:
> > * Arun R Bharadwaj <arun@linux.vnet.ibm.com> [2009-12-02 15:24:27]:
> > 
> > This patch creates arch/powerpc/platforms/pseries/processor_idle.c,
> > which implements the cpuidle infrastructure for pseries.
> > It implements a pseries_cpuidle_loop() which would be the main idle loop
> > called from cpu_idle(). It makes decision of entering either
> > dedicated_snooze_loop or dedicated_cede_loop for dedicated lpar and
> > shared_cede_loop for shared lpar processor based on the
> > decision taken by the cpuidle governor.
> 
> So unless I'm mistaken, you removed our powerpc "generic" idle loop that
> calls into ppc_md.power_save(), and replaced it by some pseries specific
> idle loops... Now what about all the other powerpc platforms ? native
> 970 (aka G5) ? 6xx ? Cell ? Or are you still calling ppc_md.power_save
> somewhere that I missed ?
> 
> Cheers,
> Ben.

Hi Ben,

I forgot to attach the patch which enables cpuidle for the rest of the
POWER platforms. Attaching it below.

So for these platforms, ppc_md.power_save will be called from from the
cpuidle_idle_call idle loop itself. Also, this cpuidle_idle_call is
not a pseries specific idle loop. It is a common loop for Intel and
PPC which use cpuidle infrastructure.

arun



This patch enables cpuidle for the rest of the POWER platforms like
44x, Cell, Pasemi etc.

Signed-off-by: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
---
 arch/powerpc/include/asm/system.h       |    2 ++
 arch/powerpc/kernel/idle.c              |   28 ++++++++++++++++++++++++++++
 arch/powerpc/kernel/setup_32.c          |    8 ++++++--
 arch/powerpc/platforms/44x/idle.c       |    2 ++
 arch/powerpc/platforms/cell/pervasive.c |    2 ++
 arch/powerpc/platforms/pasemi/idle.c    |    2 ++
 arch/powerpc/platforms/ps3/setup.c      |    2 ++
 7 files changed, 44 insertions(+), 2 deletions(-)

Index: linux.trees.git/arch/powerpc/include/asm/system.h
===================================================================
--- linux.trees.git.orig/arch/powerpc/include/asm/system.h
+++ linux.trees.git/arch/powerpc/include/asm/system.h
@@ -551,8 +551,10 @@ void cpu_idle_wait(void);
 
 #ifdef CONFIG_CPU_IDLE
 extern void update_smt_snooze_delay(int snooze);
+extern void setup_cpuidle_ppc(void);
 #else
 static inline void update_smt_snooze_delay(int snooze) {}
+static inline void setup_cpuidle_ppc(void) {}
 #endif
 
 #endif /* __KERNEL__ */
Index: linux.trees.git/arch/powerpc/kernel/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/kernel/idle.c
+++ linux.trees.git/arch/powerpc/kernel/idle.c
@@ -129,6 +129,34 @@ void default_idle(void)
 	HMT_very_low();
 }
 
+#ifdef CONFIG_CPU_IDLE
+DEFINE_PER_CPU(struct cpuidle_device, ppc_idle_devices);
+struct cpuidle_driver cpuidle_ppc_driver = {
+	.name =         "cpuidle_ppc",
+};
+
+static void ppc_idle_loop(struct cpuidle_device *dev, struct cpuidle_state *st)
+{
+	ppc_md.power_save();
+}
+
+void setup_cpuidle_ppc(void)
+{
+	struct cpuidle_device *dev;
+	int cpu;
+
+	cpuidle_register_driver(&cpuidle_ppc_driver);
+
+	for_each_online_cpu(cpu) {
+		dev = &per_cpu(ppc_idle_devices, cpu);
+		dev->cpu = cpu;
+		dev->states[0].enter = ppc_idle_loop;
+		dev->state_count = 1;
+		cpuidle_register_device(dev);
+	}
+}
+#endif
+
 int powersave_nap;
 
 #ifdef CONFIG_SYSCTL
Index: linux.trees.git/arch/powerpc/kernel/setup_32.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/kernel/setup_32.c
+++ linux.trees.git/arch/powerpc/kernel/setup_32.c
@@ -133,14 +133,18 @@ notrace void __init machine_init(unsigne
 
 #ifdef CONFIG_6xx
 	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
+	    cpu_has_feature(CPU_FTR_CAN_NAP)) {
 		ppc_md.power_save = ppc6xx_idle;
+		setup_cpuidle_ppc();
+	}
 #endif
 
 #ifdef CONFIG_E500
 	if (cpu_has_feature(CPU_FTR_CAN_DOZE) ||
-	    cpu_has_feature(CPU_FTR_CAN_NAP))
+	    cpu_has_feature(CPU_FTR_CAN_NAP)) {
 		ppc_md.power_save = e500_idle;
+		setup_cpuidle_ppc();
+	}
 #endif
 	if (ppc_md.progress)
 		ppc_md.progress("id mach(): done", 0x200);
Index: linux.trees.git/arch/powerpc/platforms/44x/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/44x/idle.c
+++ linux.trees.git/arch/powerpc/platforms/44x/idle.c
@@ -24,6 +24,7 @@
 #include <linux/of.h>
 #include <linux/kernel.h>
 #include <asm/machdep.h>
+#include <asm/system.h>
 
 static int mode_spin;
 
@@ -46,6 +47,7 @@ int __init ppc44x_idle_init(void)
 		/* If we are not setting spin mode 
                    then we set to wait mode */
 		ppc_md.power_save = &ppc44x_idle;
+		setup_cpuidle_ppc();
 	}
 
 	return 0;
Index: linux.trees.git/arch/powerpc/platforms/cell/pervasive.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/cell/pervasive.c
+++ linux.trees.git/arch/powerpc/platforms/cell/pervasive.c
@@ -35,6 +35,7 @@
 #include <asm/pgtable.h>
 #include <asm/reg.h>
 #include <asm/cell-regs.h>
+#include <asm/system.h>
 
 #include "pervasive.h"
 
@@ -128,5 +129,6 @@ void __init cbe_pervasive_init(void)
 	}
 
 	ppc_md.power_save = cbe_power_save;
+	setup_cpuidle_ppc();
 	ppc_md.system_reset_exception = cbe_system_reset_exception;
 }
Index: linux.trees.git/arch/powerpc/platforms/pasemi/idle.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/pasemi/idle.c
+++ linux.trees.git/arch/powerpc/platforms/pasemi/idle.c
@@ -27,6 +27,7 @@
 #include <asm/machdep.h>
 #include <asm/reg.h>
 #include <asm/smp.h>
+#include <asm/system.h>
 
 #include "pasemi.h"
 
@@ -81,6 +82,7 @@ static int __init pasemi_idle_init(void)
 
 	ppc_md.system_reset_exception = pasemi_system_reset_exception;
 	ppc_md.power_save = modes[current_mode].entry;
+	setup_cpuidle_ppc();
 	printk(KERN_INFO "Using PA6T idle loop (%s)\n", modes[current_mode].name);
 
 	return 0;
Index: linux.trees.git/arch/powerpc/platforms/ps3/setup.c
===================================================================
--- linux.trees.git.orig/arch/powerpc/platforms/ps3/setup.c
+++ linux.trees.git/arch/powerpc/platforms/ps3/setup.c
@@ -33,6 +33,7 @@
 #include <asm/prom.h>
 #include <asm/lv1call.h>
 #include <asm/ps3gpu.h>
+#include <asm/system.h>
 
 #include "platform.h"
 
@@ -214,6 +215,7 @@ static void __init ps3_setup_arch(void)
 	prealloc_ps3flash_bounce_buffer();
 
 	ppc_md.power_save = ps3_power_save;
+	setup_cpuidle_ppc();
 	ps3_os_area_init();
 
 	DBG(" <- %s:%d\n", __func__, __LINE__);

  parent reply	other threads:[~2009-12-04  8:15 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-12-02  9:54 [v10 PATCH 0/9] cpuidle: cleanup cpuidle/ introduce cpuidle to POWER Arun R Bharadwaj
2009-12-02  9:54 ` Arun R Bharadwaj
2009-12-02  9:55 ` [v10 PATCH 1/9]: cpuidle: Design documentation patch Arun R Bharadwaj
2009-12-02  9:55   ` Arun R Bharadwaj
2009-12-02  9:57 ` [v10 PATCH 2/9]: cpuidle: cleanup drivers/cpuidle/cpuidle.c Arun R Bharadwaj
2009-12-02  9:57   ` Arun R Bharadwaj
2009-12-04 22:20   ` Torsten Duwe
2009-12-06  5:19     ` Arun R Bharadwaj
2009-12-06  5:19       ` Arun R Bharadwaj
2009-12-02  9:58 ` [v10 PATCH 3/9]: cpuidle: implement a list based approach to register a set of idle routines Arun R Bharadwaj
2009-12-02  9:58   ` Arun R Bharadwaj
2009-12-02  9:59 ` [v10 PATCH 4/9]: x86: refactor x86 idle power management code, remove all instances of pm_idle Arun R Bharadwaj
2009-12-02  9:59   ` Arun R Bharadwaj
2009-12-02 10:00 ` [v10 PATCH 5/9]: POWER: enable cpuidle for POWER Arun R Bharadwaj
2009-12-02 10:00   ` Arun R Bharadwaj
2009-12-02 10:01 ` [v10 PATCH 6/9]: pSeries/cpuidle: refactor pseries idle loops Arun R Bharadwaj
2009-12-02 10:01   ` Arun R Bharadwaj
2009-12-04  2:45   ` Benjamin Herrenschmidt
2009-12-04  2:45     ` Benjamin Herrenschmidt
2009-12-02 10:01 ` [v10 PATCH 7/9]: POWER: add a default_idle idle loop for POWER Arun R Bharadwaj
2009-12-02 10:01   ` Arun R Bharadwaj
2009-12-02 10:02 ` [v10 PATCH 8/9]: pSeries: implement pSeries processor idle module Arun R Bharadwaj
2009-12-02 10:02   ` Arun R Bharadwaj
2009-12-04  2:47   ` Benjamin Herrenschmidt
2009-12-04  2:47     ` Benjamin Herrenschmidt
2009-12-04  8:15     ` Arun R Bharadwaj [this message]
2009-12-04  8:15       ` Arun R Bharadwaj
2009-12-04 10:00       ` Benjamin Herrenschmidt
2009-12-04 10:00         ` Benjamin Herrenschmidt
2009-12-15 11:49         ` Arun R Bharadwaj
2009-12-15 11:49           ` Arun R Bharadwaj
2009-12-02 10:03 ` [v10 PATCH 9/9]: POWER: Enable default_idle when power_save=off Arun R Bharadwaj
2009-12-02 10:03   ` Arun R Bharadwaj
2009-12-02 17:41   ` Daniel Walker
2009-12-03  6:33   ` Arun R Bharadwaj

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20091204081539.GA5883@linux.vnet.ibm.com \
    --to=arun@linux.vnet.ibm.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=benh@kernel.crashing.org \
    --cc=linux-acpi@vger.kernel.org \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=mingo@elte.hu \
    --cc=venkatesh.pallipadi@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).