POWER/cpuidle: Generic IBM-POWER cpuidle driver enabled for PSERIES and POWERNV platforms
Subsystems (2)
v5 cover POWER/cpuidle: Generic IBM-POWER cpuidle driver enabled for PSERIES and POWERNV platforms
This patch series consolidates the backend cpuidle driver for pSeries and powernv platforms with minimal code duplication.
Current existing backend driver for pseries has been moved to drivers/cpuidle and has been extended to accommodate powernv idle power mgmt states. As seen in V1 of this patch series, having a separate powernv backend driver results in too much code duplication, which is less elegant and can pose maintenance problems going further.
Using the cpuidle framework to exploit platform low power idle states management can take advantage of advanced heuristics, tunables and features provided by framework. The statistics and tracing infrastructure provided by the cpuidle framework also helps in enabling power management related tools and help tune the system and applications.
Earlier in 3.3 kernel, pSeries idle state management was modified to exploit the cpuidle framework and the end goal of this patch is to have powernv platform also to hook its idle states into cpuidle framework with minimal code duplication between both platforms.
This series aims to maintain compatibility and functionality to existing pseries and powernv idle cpu management code. There are no new functions or idle states added as part of this series. This can be extended by adding more states to this existing framework.
This patch series has been tested on both PSERIES and POWERNV platform and is based on v3.11-rc6 kernel version.
With this patch series, the powernv cpuidle functionalities are on-par with pSeries idle management.
V1 -> http://lkml.org/lkml/2013/7/23/143
V2 -> https://lkml.org/lkml/2013/7/30/872
V3 -> http://comments.gmane.org/gmane.linux.ports.ppc.embedded/63093
V4 -> https://lkml.org/lkml/2013/8/22/25
Changes in V5:
=============
* As per the discussions in the community, this patch series enables cpuidle backend driver only for IBM-POWER platforms. File is re-named from drivers/cpuidle/cpuidle-powerpc.c to drivers/cpuidle/cpuildle-ibm-power.c
New back-end cpuidle driver is called IBM-POWER-Idle.
* General cleanups on the accessors front that was introduced in previous version.
Changes in V4:
=============
* This patch series includes generic backend driver cpuidle cleanups including, replacing the driver and device initialisation routines with cpuidle_register function.
* Enable CPUIDLE framework only for POWER and POWERNV platforms.
Changes in V3:
=============
* This patch series does not include smt-snooze-delay fixes. This will be taken up later on.
* Integrated POWERPC driver in drivers/cpuidle. Enabled for all of POWERPC platform. Currently has PSERIES and POWERNV support. No compile time flags in .c file. This will be one consolidated binary that does a run time detection based on platform and take decisions accordingly.
* Enabled CPUIDLE framwork for all of PPC64.
Changes in V2:
=============
* Merged the backend driver posted out for powernv in V1 with pSeries to create a single powerpc driver but this had compile time flags.
Deepthi Dharwar (5):
pseries/cpuidle: Remove dependency of pseries.h file
pseries: Move plpar_wrapper.h to powerpc common include/asm location. POWER/cpuidle: Generic IBM-POWER backend cpuidle driver. POWER/cpuidle: Enable powernv cpuidle support.
powernv/cpuidle: Enable idle powernv cpu to call into the cpuidle framework.
arch/powerpc/include/asm/plpar_wrappers.h | 325 +++++++++++++++++++++ arch/powerpc/include/asm/processor.h | 2 arch/powerpc/platforms/powernv/setup.c | 14 + arch/powerpc/platforms/pseries/Kconfig | 9 - arch/powerpc/platforms/pseries/Makefile | 1 arch/powerpc/platforms/pseries/cmm.c | 3 arch/powerpc/platforms/pseries/dtl.c | 3 arch/powerpc/platforms/pseries/hotplug-cpu.c | 3 arch/powerpc/platforms/pseries/hvconsole.c | 2 arch/powerpc/platforms/pseries/iommu.c | 3 arch/powerpc/platforms/pseries/kexec.c | 2 arch/powerpc/platforms/pseries/lpar.c | 2 arch/powerpc/platforms/pseries/plpar_wrappers.h | 324 --------------------- arch/powerpc/platforms/pseries/processor_idle.c | 362 ----------------------- arch/powerpc/platforms/pseries/pseries.h | 3 arch/powerpc/platforms/pseries/setup.c | 2 arch/powerpc/platforms/pseries/smp.c | 2 drivers/cpuidle/Kconfig | 7 drivers/cpuidle/Makefile | 2 drivers/cpuidle/cpuidle-ibm-power.c | 335 +++++++++++++++++++++ 20 files changed, 693 insertions(+), 713 deletions(-) create mode 100644 arch/powerpc/include/asm/plpar_wrappers.h delete mode 100644 arch/powerpc/platforms/pseries/plpar_wrappers.h delete mode 100644 arch/powerpc/platforms/pseries/processor_idle.c create mode 100644 drivers/cpuidle/cpuidle-ibm-power.c
-- Deepthi
PATCH v5 1/5 pseries/cpuidle: Remove dependency of pseries.h file
As a part of pseries_idle cleanup to make the backend driver code common to both pseries and powernv.
Remove non-essential smt_snooze_delay declaration in pseries.h header file and pseries.h file inclusion in pseries/processor_idle.c
--- arch/powerpc/platforms/pseries/processor_idle.c | 1 - arch/powerpc/platforms/pseries/pseries.h | 3 --- 2 files changed, 4 deletions(-)
arch/powerpc/platforms/pseries/processor_idle.c +0 -1
@@ -20,7 +20,6 @@#include <asm/runlatch.h> #include "plpar_wrappers.h" -#include "pseries.h" struct cpuidle_driver pseries_idle_driver = { .name ="pseries_idle" ,
arch/powerpc/platforms/pseries/pseries.h +0 -3
@@ -60,9 +60,6 @@ extern struct device_node *dlpar_configure_connector(u32);extern int dlpar_attach_node (struct device_node *);extern int dlpar_detach_node (struct device_node *); -/* Snooze Delay, pseries_idle */ -DECLARE_PER_CPU (long, smt_snooze_delay); -/* PCI root bridge prepare function override for pseries */ struct pci_host_bridge ;int pseries_root_bridge_prepare (struct pci_host_bridge *bridge);
PATCH v5 2/5 pseries: Move plpar_wrapper.h to powerpc common include/asm location.
In reply to Deepthi DharwarAs a part of pseries_idle backend driver cleanup to make the code common to both pseries and powernv platforms, it is necessary to move the backend-driver code to drivers/cpuidle.
As a pre-requisite for that, it is essential to move plpar_wrapper.h to include/asm.
--- arch/powerpc/include/asm/plpar_wrappers.h | 325 +++++++++++++++++++++++ arch/powerpc/platforms/pseries/cmm.c | 3 arch/powerpc/platforms/pseries/dtl.c | 3 arch/powerpc/platforms/pseries/hotplug-cpu.c | 3 arch/powerpc/platforms/pseries/hvconsole.c | 2 arch/powerpc/platforms/pseries/iommu.c | 3 arch/powerpc/platforms/pseries/kexec.c | 2 arch/powerpc/platforms/pseries/lpar.c | 2 arch/powerpc/platforms/pseries/plpar_wrappers.h | 324 ----------------------- arch/powerpc/platforms/pseries/processor_idle.c | 3 arch/powerpc/platforms/pseries/setup.c | 2 arch/powerpc/platforms/pseries/smp.c | 2 12 files changed, 336 insertions(+), 338 deletions(-) create mode 100644 arch/powerpc/include/asm/plpar_wrappers.h delete mode 100644 arch/powerpc/platforms/pseries/plpar_wrappers.h
arch/powerpc/include/asm/plpar_wrappers.h new file +325 -0
@@ -0,0 +1,325 @@ +#ifndef _PSERIES_PLPAR_WRAPPERS_H +#define _PSERIES_PLPAR_WRAPPERS_H + +#include <linux/string.h> +#include <linux/irqflags.h> + +#include <asm/hvcall.h> +#include <asm/paca.h> +#include <asm/page.h> + +/* Get state of physical CPU from query_cpu_stopped */ +int smp_query_cpu_stopped (unsigned int pcpu); +#define QCSS_STOPPED 0 +#define QCSS_STOPPING 1 +#define QCSS_NOT_STOPPED 2 +#define QCSS_HARDWARE_ERROR -1 +#define QCSS_HARDWARE_BUSY -2 + +static inline long poll_pending (void ) +{ +return plpar_hcall_norets (H_POLL_PENDING ); +} + +static inline u8 get_cede_latency_hint (void ) +{ +return get_lppaca ()->cede_latency_hint; +} + +static inline void set_cede_latency_hint (u8 latency_hint) +{ +get_lppaca ()->cede_latency_hint = latency_hint; +} + +static inline long cede_processor (void ) +{ +return plpar_hcall_norets (H_CEDE ); +} + +static inline long extended_cede_processor (unsigned long latency_hint) +{ +long rc; +u8 old_latency_hint =get_cede_latency_hint (); + +set_cede_latency_hint (latency_hint); + + rc =cede_processor (); +#ifdef CONFIG_TRACE_IRQFLAGS +/* Ensure that H_CEDE returns with IRQs on */ +if (WARN_ON (!(mfmsr () &MSR_EE ))) +__hard_irq_enable (); +#endif + +set_cede_latency_hint (old_latency_hint); + +return rc; +} + +static inline long vpa_call (unsigned long flags,unsigned long cpu, +unsigned long vpa) +{ + flags = flags <<H_VPA_FUNC_SHIFT ; + +return plpar_hcall_norets (H_REGISTER_VPA , flags, cpu, vpa); +} + +static inline long unregister_vpa (unsigned long cpu) +{ +return vpa_call (H_VPA_DEREG_VPA , cpu,0 ); +} + +static inline long register_vpa (unsigned long cpu,unsigned long vpa) +{ +return vpa_call (H_VPA_REG_VPA , cpu, vpa); +} + +static inline long unregister_slb_shadow (unsigned long cpu) +{ +return vpa_call (H_VPA_DEREG_SLB , cpu,0 ); +} + +static inline long register_slb_shadow (unsigned long cpu,unsigned long vpa) +{ +return vpa_call (H_VPA_REG_SLB , cpu, vpa); +} + +static inline long unregister_dtl (unsigned long cpu) +{ +return vpa_call (H_VPA_DEREG_DTL , cpu,0 ); +} + +static inline long register_dtl (unsigned long cpu,unsigned long vpa) +{ +return vpa_call (H_VPA_REG_DTL , cpu, vpa); +} + +static inline long plpar_page_set_loaned (unsigned long vpa) +{ +unsigned long cmo_page_sz =cmo_get_page_size (); +long rc =0 ; +int i; + +for (i =0 ; !rc && i <PAGE_SIZE ; i += cmo_page_sz) + rc =plpar_hcall_norets (H_PAGE_INIT ,H_PAGE_SET_LOANED , vpa + i,0 ); + +for (i -= cmo_page_sz; rc && i !=0 ; i -= cmo_page_sz) +plpar_hcall_norets (H_PAGE_INIT ,H_PAGE_SET_ACTIVE , + vpa + i - cmo_page_sz,0 ); + +return rc; +} + +static inline long plpar_page_set_active (unsigned long vpa) +{ +unsigned long cmo_page_sz =cmo_get_page_size (); +long rc =0 ; +int i; + +for (i =0 ; !rc && i <PAGE_SIZE ; i += cmo_page_sz) + rc =plpar_hcall_norets (H_PAGE_INIT ,H_PAGE_SET_ACTIVE , vpa + i,0 ); + +for (i -= cmo_page_sz; rc && i !=0 ; i -= cmo_page_sz) +plpar_hcall_norets (H_PAGE_INIT ,H_PAGE_SET_LOANED , + vpa + i - cmo_page_sz,0 ); + +return rc; +} + +extern void vpa_init (int cpu); + +static inline long plpar_pte_enter (unsigned long flags, +unsigned long hpte_group,unsigned long hpte_v, +unsigned long hpte_r,unsigned long *slot) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall (H_ENTER , retbuf, flags, hpte_group, hpte_v, hpte_r); + + *slot = retbuf[0 ]; + +return rc; +} + +static inline long plpar_pte_remove (unsigned long flags,unsigned long ptex, +unsigned long avpn,unsigned long *old_pteh_ret, +unsigned long *old_ptel_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall (H_REMOVE , retbuf, flags, ptex, avpn); + + *old_pteh_ret = retbuf[0 ]; + *old_ptel_ret = retbuf[1 ]; + +return rc; +} + +/* plpar_pte_remove_raw can be called in real mode. It calls plpar_hcall_raw */ +static inline long plpar_pte_remove_raw (unsigned long flags,unsigned long ptex, +unsigned long avpn,unsigned long *old_pteh_ret, +unsigned long *old_ptel_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall_raw (H_REMOVE , retbuf, flags, ptex, avpn); + + *old_pteh_ret = retbuf[0 ]; + *old_ptel_ret = retbuf[1 ]; + +return rc; +} + +static inline long plpar_pte_read (unsigned long flags,unsigned long ptex, +unsigned long *old_pteh_ret,unsigned long *old_ptel_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall (H_READ , retbuf, flags, ptex); + + *old_pteh_ret = retbuf[0 ]; + *old_ptel_ret = retbuf[1 ]; + +return rc; +} + +/* plpar_pte_read_raw can be called in real mode. It calls plpar_hcall_raw */ +static inline long plpar_pte_read_raw (unsigned long flags,unsigned long ptex, +unsigned long *old_pteh_ret,unsigned long *old_ptel_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall_raw (H_READ , retbuf, flags, ptex); + + *old_pteh_ret = retbuf[0 ]; + *old_ptel_ret = retbuf[1 ]; + +return rc; +} + +/* +* plpar_pte_read_4_raw can be called in real mode. +* ptes must be 8*sizeof(unsigned long) +*/ +static inline long plpar_pte_read_4_raw (unsigned long flags,unsigned long ptex, +unsigned long *ptes) + +{ +long rc; +unsigned long retbuf[PLPAR_HCALL9_BUFSIZE ]; + + rc =plpar_hcall9_raw (H_READ , retbuf, flags |H_READ_4 , ptex); + +memcpy (ptes, retbuf,8 *sizeof (unsigned long )); + +return rc; +} + +static inline long plpar_pte_protect (unsigned long flags,unsigned long ptex, +unsigned long avpn) +{ +return plpar_hcall_norets (H_PROTECT , flags, ptex, avpn); +} + +static inline long plpar_tce_get (unsigned long liobn,unsigned long ioba, +unsigned long *tce_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; + + rc =plpar_hcall (H_GET_TCE , retbuf, liobn, ioba); + + *tce_ret = retbuf[0 ]; + +return rc; +} + +static inline long plpar_tce_put (unsigned long liobn,unsigned long ioba, +unsigned long tceval) +{ +return plpar_hcall_norets (H_PUT_TCE , liobn, ioba, tceval); +} + +static inline long plpar_tce_put_indirect (unsigned long liobn, +unsigned long ioba,unsigned long page,unsigned long count) +{ +return plpar_hcall_norets (H_PUT_TCE_INDIRECT , liobn, ioba, page, count); +} + +static inline long plpar_tce_stuff (unsigned long liobn,unsigned long ioba, +unsigned long tceval,unsigned long count) +{ +return plpar_hcall_norets (H_STUFF_TCE , liobn, ioba, tceval, count); +} + +static inline long plpar_get_term_char (unsigned long termno, +unsigned long *len_ret,char *buf_ret) +{ +long rc; +unsigned long retbuf[PLPAR_HCALL_BUFSIZE ]; +unsigned long *lbuf = (unsigned long *)buf_ret;/* TODO: alignment? */ + + rc =plpar_hcall (H_GET_TERM_CHAR , retbuf, termno); + + *len_ret = retbuf[0 ]; + lbuf[0 ] = retbuf[1 ]; + lbuf[1 ] = retbuf[2 ]; + +return rc; +} + +static inline long plpar_put_term_char (unsigned long termno,unsigned long len, +const char *buffer) +{ +unsigned long *lbuf = (unsigned long *)buffer;/* TODO: alignment? */ +return plpar_hcall_norets (H_PUT_TERM_CHAR , termno, len, lbuf[0 ], + lbuf[1 ]); +} + +/* Set various resource mode parameters */ +static inline long plpar_set_mode (unsigned long mflags,unsigned long resource, +unsigned long value1,unsigned long value2) +{ +return plpar_hcall_norets (H_SET_MODE , mflags, resource, value1, value2); +} + +/* +* Enable relocation on exceptions on this partition +* +* Note: this call has a partition wide scope and can take a while to complete. +* If it returns H_LONG_BUSY_* it should be retried periodically until it +* returns H_SUCCESS. +*/ +static inline long enable_reloc_on_exceptions (void ) +{ +/* mflags = 3: Exceptions at 0xC000000000004000 */ +return plpar_set_mode (3 ,3 ,0 ,0 ); +} + +/* +* Disable relocation on exceptions on this partition +* +* Note: this call has a partition wide scope and can take a while to complete. +* If it returns H_LONG_BUSY_* it should be retried periodically until it +* returns H_SUCCESS. +*/ +static inline long disable_reloc_on_exceptions (void ) +{ +return plpar_set_mode (0 ,3 ,0 ,0 ); +} + +static inline long plapr_set_ciabr (unsigned long ciabr) +{ +return plpar_set_mode (0 ,1 , ciabr,0 ); +} + +static inline long plapr_set_watchpoint0 (unsigned long dawr0,unsigned long dawrx0) +{ +return plpar_set_mode (0 ,2 , dawr0, dawrx0); +} + +#endif /* _PSERIES_PLPAR_WRAPPERS_H */
arch/powerpc/platforms/pseries/cmm.c +1 -2
@@ -40,8 +40,7 @@#include <asm/pgalloc.h> #include <asm/uaccess.h> #include <linux/memory.h> - -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> #define CMM_DRIVER_VERSION "1.0.0"#define CMM_DEFAULT_DELAY 1
arch/powerpc/platforms/pseries/dtl.c +1 -2
@@ -29,8 +29,7 @@#include <asm/firmware.h> #include <asm/lppaca.h> #include <asm/debug.h> - -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> struct dtl {struct dtl_entry *buf;
arch/powerpc/platforms/pseries/hotplug-cpu.c +2 -1
@@ -30,7 +30,8 @@#include <asm/machdep.h> #include <asm/vdso_datapage.h> #include <asm/xics.h> -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> +#include "offline_states.h" /* This version can't take the spinlock, because it never returns */
arch/powerpc/platforms/pseries/hvconsole.c +1 -1
@@ -28,7 +28,7 @@#include <linux/errno.h> #include <asm/hvcall.h> #include <asm/hvconsole.h> -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> /** * hvc_get_chars - retrieve characters from firmware for denoted vterm adatper
arch/powerpc/platforms/pseries/iommu.c +1 -2
@@ -48,8 +48,7 @@#include <asm/ppc-pci.h> #include <asm/udbg.h> #include <asm/mmzone.h> - -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> static void tce_invalidate_pSeries_sw(struct iommu_table *tbl,
arch/powerpc/platforms/pseries/kexec.c +1 -1
@@ -17,9 +17,9 @@#include <asm/mpic.h> #include <asm/xics.h> #include <asm/smp.h> +#include <asm/plpar_wrappers.h> #include "pseries.h" -#include "plpar_wrappers.h" static void pseries_kexec_cpu_down (int crash_shutdown,int secondary) {
arch/powerpc/platforms/pseries/lpar.c +1 -1
@@ -41,8 +41,8 @@#include <asm/smp.h> #include <asm/trace.h> #include <asm/firmware.h> +#include <asm/plpar_wrappers.h> -#include "plpar_wrappers.h" #include "pseries.h" /* Flag bits for H_BULK_REMOVE */
arch/powerpc/platforms/pseries/plpar_wrappers.h deleted +0 -324
324 lines not on this page · raw message
arch/powerpc/platforms/pseries/processor_idle.c +1 -2
@@ -18,8 +18,7 @@#include <asm/machdep.h> #include <asm/firmware.h> #include <asm/runlatch.h> - -#include "plpar_wrappers.h" +#include <asm/plpar_wrappers.h> struct cpuidle_driver pseries_idle_driver = { .name ="pseries_idle" ,
arch/powerpc/platforms/pseries/setup.c +1 -1
@@ -66,8 +66,8 @@#include <asm/firmware.h> #include <asm/eeh.h> #include <asm/reg.h> +#include <asm/plpar_wrappers.h> -#include "plpar_wrappers.h" #include "pseries.h" int CMO_PrPSP =-1 ;
arch/powerpc/platforms/pseries/smp.c +1 -1
@@ -43,8 +43,8 @@#include <asm/cputhreads.h> #include <asm/xics.h> #include <asm/dbell.h> +#include <asm/plpar_wrappers.h> -#include "plpar_wrappers.h" #include "pseries.h" #include "offline_states.h"
PATCH v5 3/5 POWER/cpuidle: Generic IBM-POWER backend cpuidle driver.
In reply to Deepthi DharwarThis patch involves moving the current pseries_idle backend driver code from pseries/processor_idle.c to drivers/cpuidle/cpuidle-ibm-power.c.
It enables the support for pseries platform, such that going forward the same code with minimal efforts can be re-used for a common driver on powernv This removes a lot of code duplicacy, thus making the code elegant.
--- arch/powerpc/include/asm/processor.h | 2 arch/powerpc/platforms/pseries/Kconfig | 9 - arch/powerpc/platforms/pseries/Makefile | 1 arch/powerpc/platforms/pseries/processor_idle.c | 360 ----------------------- drivers/cpuidle/Kconfig | 7 drivers/cpuidle/Makefile | 2 drivers/cpuidle/cpuidle-ibm-power.c | 304 +++++++++++++++++++ 7 files changed, 314 insertions(+), 371 deletions(-) delete mode 100644 arch/powerpc/platforms/pseries/processor_idle.c create mode 100644 drivers/cpuidle/cpuidle-ibm-power.c
arch/powerpc/include/asm/processor.h +1 -1
@@ -430,7 +430,7 @@ enum idle_boot_override {IDLE_NO_OVERRIDE = 0, IDLE_POWERSAVE_OFF};extern int powersave_nap;/* set if nap mode can be used in idle loop */ extern void power7_nap (void ); -#ifdef CONFIG_PSERIES_IDLE +#ifdef CONFIG_CPU_IDLE_IBM_POWER extern void update_smt_snooze_delay (int cpu,int residency);#else static inline void update_smt_snooze_delay (int cpu,int residency) {}
arch/powerpc/platforms/pseries/Kconfig +0 -9
@@ -119,12 +119,3 @@ config DTL which are accessible through a debugfs file. Say N if you are unsure. - -config PSERIES_IDLE - bool "Cpuidle driver for pSeries platforms" - depends on CPU_IDLE - depends on PPC_PSERIES - default y - help - Select this option to enable processor idle state management - through cpuidle subsystem.
arch/powerpc/platforms/pseries/Makefile +0 -1
@@ -21,7 +21,6 @@ obj-$(CONFIG_HCALL_STATS) += hvCall_inst.o obj-$(CONFIG_CMM) += cmm.o obj-$(CONFIG_DTL) += dtl.o obj-$(CONFIG_IO_EVENT_IRQ) += io_event_irq.o -obj-$(CONFIG_PSERIES_IDLE) += processor_idle.o ifeq ($(CONFIG_PPC_PSERIES),y) obj-$(CONFIG_SUSPEND) += suspend.o
arch/powerpc/platforms/pseries/processor_idle.c deleted +0 -360
360 lines not on this page · raw message
drivers/cpuidle/Kconfig +7 -0
@@ -42,6 +42,13 @@ config CPU_IDLE_ZYNQ help Select this to enable cpuidle on Xilinx Zynq processors. +config CPU_IDLE_IBM_POWER + bool "CPU Idle driver for IBM POWER platforms" + depends on PPC_PSERIES || PPC_POWERNV + default y + help + Select this option to enable processor idle state management + on IBM POWER platform. endif config ARCH_NEEDS_CPU_IDLE_COUPLED
drivers/cpuidle/Makefile +2 -0
@@ -8,3 +8,5 @@ obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o obj-$(CONFIG_CPU_IDLE_CALXEDA) += cpuidle-calxeda.o obj-$(CONFIG_ARCH_KIRKWOOD) += cpuidle-kirkwood.o obj-$(CONFIG_CPU_IDLE_ZYNQ) += cpuidle-zynq.o + +obj-$(CONFIG_CPU_IDLE_IBM_POWER) += cpuidle-ibm-power.o
drivers/cpuidle/cpuidle-ibm-power.c new file +304 -0
304 lines not on this page · raw message
RE: [PATCH V5 3/5] POWER/cpuidle: Generic IBM-POWER backend cpuidle driver.
drivers/cpuidle/Kconfig
Quoted 5 lines
drivers/cpuidle/Kconfig +0
-0
Maybe drivers/cpuidle/Kconfig.powerpc is better? Like arm.
Quoted 28 lines
drivers/cpuidle/Kconfig +1
-0+obj-$(CONFIG_CPU_IDLE_IBM_POWER) += cpuidle-ibm-power.odrivers/cpuidle/cpuidle-ibm-power.c new file +20
-0@@ -0,0 +1,304 @@ +/* +* cpuidle-ibm-power - idle state cpuidle driver. +* Adapted from drivers/idle/intel_idle.c and +* drivers/acpi/processor_idle.c +* +*/ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/moduleparam.h> +#include <linux/cpuidle.h> +#include <linux/cpu.h> +#include <linux/notifier.h> + +#include <asm/paca.h> +#include <asm/reg.h> +#include <asm/machdep.h> +#include <asm/firmware.h> +#include <asm/runlatch.h>
drivers/cpuidle/cpuidle-ibm-power.c line 30
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+#include <asm/plpar_wrappers.h> + +struct cpuidle_driver power_idle_driver = { + .name ="IBM-POWER-Idle" , + .owner =THIS_MODULE , +}; + +#define MAX_IDLE_STATE_COUNT 2 + +static int max_idle_state =MAX_IDLE_STATE_COUNT -1 ;
Again, do not use the macro.
Quoted 16 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +16
-0+static struct cpuidle_state *cpuidle_state_table; + +static inline void idle_loop_prolog (unsigned long *in_purr) +{ + *in_purr =mfspr (SPRN_PURR ); +/* +* Indicate to the HV that we are idle. Now would be +* a good time to find other work to dispatch. +*/ +get_lppaca ()->idle =1 ; +} + +static inline void idle_loop_epilog (unsigned long in_purr) +{ +get_lppaca ()->wait_state_cycles +=mfspr (SPRN_PURR ) - in_purr; +get_lppaca ()->idle =0 ;
drivers/cpuidle/cpuidle-ibm-power.c line 56
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+} + +static int snooze_loop (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index) +{ +unsigned long in_purr; + +idle_loop_prolog (&in_purr); +local_irq_enable ();
snooze_loop has already registered in cpuidle framework to handle snooze state. where disable the irq? Why do "enable" here?
Quoted 27 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +27
-0+/* +* States for dedicated partition case. +*/ +static struct cpuidle_state dedicated_states[MAX_IDLE_STATE_COUNT ] = { + {/* Snooze */ + .name ="snooze" , + .desc ="snooze" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =0 , + .target_residency =0 , + .enter = &snooze_loop }, + {/* CEDE */ + .name ="CEDE" , + .desc ="CEDE" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =10 , + .target_residency =100 , + .enter = &dedicated_cede_loop }, +}; + +/* +* States for shared partition case. +*/ +static struct cpuidle_state shared_states[MAX_IDLE_STATE_COUNT ] = { + {/* Shared Cede */ + .name ="Shared Cede" , + .desc ="Shared Cede" ,
drivers/cpuidle/cpuidle-ibm-power.c
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+ .flags = CPUIDLE_FLAG_TIME_VALID, + .exit_latency = 0, + .target_residency = 0, + .enter = &shared_cede_loop }, +}; + +static void __exit power_processor_idle_exit(void) +{ + +unregister_cpu_notifier (&setup_hotplug_notifier);
Remove a blank line.
drivers/cpuidle/cpuidle-ibm-power.c line 301
Quoted 7 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +7
-0+cpuidle_unregister (&power_idle_driver); +return ; +} + +module_init (power_processor_idle_init); +module_exit (power_processor_idle_exit); +
Did you have tested the module? If not tested, please don't use the module.
drivers/cpuidle/cpuidle-ibm-power.c line 304
Quoted 4 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +3
-0+MODULE_AUTHOR ("Deepthi Dharwar <[email protected]>" ); +MODULE_DESCRIPTION ("Cpuidle driver for IBM POWER platforms" ); +MODULE_LICENSE ("GPL" );
���{.n�+�������+%�����ݶ��w��{.n�+����{��G�����{ay�ʇڙ�,j��f���h��������z_�(�階�ݢj"���m������G������?����&���~��iO��z��v�^�m����������?�I�
Re: [PATCH V5 3/5] POWER/cpuidle: Generic IBM-POWER backend cpuidle driver.
Quoted 8 lines · On 08/23/2013 08:46 AM, Wang Dongsheng-B40534 wrote:
drivers/cpuidle/Kconfig
Quoted 5 lines
drivers/cpuidle/Kconfig +0
-0Maybe
drivers/cpuidle/Kconfig.powerpcis better? Like arm.
Yes will do that, going ahead other powerpc archs can use this Kconfig.
Quoted 32 lines
Quoted 32 lines
drivers/cpuidle/Kconfig +1
-0+obj-$(CONFIG_CPU_IDLE_IBM_POWER) += cpuidle-ibm-power.odrivers/cpuidle/cpuidle-ibm-power.c new file +24
-0@@ -0,0 +1,304 @@ +/* +* cpuidle-ibm-power - idle state cpuidle driver. +* Adapted from drivers/idle/intel_idle.c and +* drivers/acpi/processor_idle.c +* +*/ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/moduleparam.h> +#include <linux/cpuidle.h> +#include <linux/cpu.h> +#include <linux/notifier.h> + +#include <asm/paca.h> +#include <asm/reg.h> +#include <asm/machdep.h> +#include <asm/firmware.h> +#include <asm/runlatch.h> +#include <asm/plpar_wrappers.h> + +struct cpuidle_driver power_idle_driver = { + .name ="IBM-POWER-Idle" ,
Quoted 9 lines
drivers/cpuidle/cpuidle-ibm-power.c line 30
Quoted 6 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +6
-0+ .owner = THIS_MODULE, +}; + +#define MAX_IDLE_STATE_COUNT 2 + +static int max_idle_state =MAX_IDLE_STATE_COUNT -1 ;Again, do not use the macro.
Yes, shall use ARRAY_SIZE instead and get away with this macro. That should work
Quoted 21 lines
Quoted 21 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +21
-0+static struct cpuidle_state *cpuidle_state_table; + +static inline void idle_loop_prolog (unsigned long *in_purr) +{ + *in_purr =mfspr (SPRN_PURR ); +/* +* Indicate to the HV that we are idle. Now would be +* a good time to find other work to dispatch. +*/ +get_lppaca ()->idle =1 ; +} + +static inline void idle_loop_epilog (unsigned long in_purr) +{ +get_lppaca ()->wait_state_cycles +=mfspr (SPRN_PURR ) - in_purr; +get_lppaca ()->idle =0 ; +} + +static int snooze_loop (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index)
Quoted 9 lines
drivers/cpuidle/cpuidle-ibm-power.c line 56
Quoted 5 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +5
-0+{ +unsigned long in_purr; + +idle_loop_prolog (&in_purr); +local_irq_enable ();
snooze_loophas already registered in cpuidle framework to handle snooze state. where disable the irq? Why do "enable" here?
On POWER, snooze state is an infinte busy looping state. The CPUs keep looping around lowering their priority until they are interrupted. For this we need to enable irqs in the snooze loop. There is no way a CPU can come out this state if the interrupts are not enabled.
Quoted 40 lines
Quoted 27 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +27
-0+/* +* States for dedicated partition case. +*/ +static struct cpuidle_state dedicated_states[MAX_IDLE_STATE_COUNT ] = { + {/* Snooze */ + .name ="snooze" , + .desc ="snooze" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =0 , + .target_residency =0 , + .enter = &snooze_loop }, + {/* CEDE */ + .name ="CEDE" , + .desc ="CEDE" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =10 , + .target_residency =100 , + .enter = &dedicated_cede_loop }, +}; + +/* +* States for shared partition case. +*/ +static struct cpuidle_state shared_states[MAX_IDLE_STATE_COUNT ] = { + {/* Shared Cede */ + .name ="Shared Cede" , + .desc ="Shared Cede" ,drivers/cpuidle/cpuidle-ibm-power.c
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+ .flags = CPUIDLE_FLAG_TIME_VALID, + .exit_latency = 0, + .target_residency = 0, + .enter = &shared_cede_loop }, +}; + +static void __exit power_processor_idle_exit(void) +{ + +unregister_cpu_notifier (&setup_hotplug_notifier);Remove a blank line.
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c line 301
Quoted 7 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +7
-0+cpuidle_unregister (&power_idle_driver); +return ; +} + +module_init (power_processor_idle_init); +module_exit (power_processor_idle_exit); +Did you have tested the module? If not tested, please don't use the module.
We do want to make it a module. But for now that cannot be done due to some other dependencies. This feature needs to be built in.
Thanks for the review.
Quoted 5 lines
drivers/cpuidle/cpuidle-ibm-power.c line 304
Quoted 4 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +3
-0+MODULE_AUTHOR ("Deepthi Dharwar <[email protected]>" ); +MODULE_DESCRIPTION ("Cpuidle driver for IBM POWER platforms" ); +MODULE_LICENSE ("GPL" );
Regards,
Deepthi
Re: [PATCH V5 3/5] POWER/cpuidle: Generic IBM-POWER backend cpuidle driver.
In reply to Wang Dongsheng-B40534Quoted 40 lines · On 08/23/2013 08:46 AM, Wang Dongsheng-B40534 wrote:
drivers/cpuidle/Kconfig
Quoted 5 lines
drivers/cpuidle/Kconfig +0
-0Maybe
drivers/cpuidle/Kconfig.powerpcis better? Like arm.Quoted 32 lines
drivers/cpuidle/Kconfig +1
-0+obj-$(CONFIG_CPU_IDLE_IBM_POWER) += cpuidle-ibm-power.odrivers/cpuidle/cpuidle-ibm-power.c new file +24
-0@@ -0,0 +1,304 @@ +/* +* cpuidle-ibm-power - idle state cpuidle driver. +* Adapted from drivers/idle/intel_idle.c and +* drivers/acpi/processor_idle.c +* +*/ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/moduleparam.h> +#include <linux/cpuidle.h> +#include <linux/cpu.h> +#include <linux/notifier.h> + +#include <asm/paca.h> +#include <asm/reg.h> +#include <asm/machdep.h> +#include <asm/firmware.h> +#include <asm/runlatch.h> +#include <asm/plpar_wrappers.h> + +struct cpuidle_driver power_idle_driver = { + .name ="IBM-POWER-Idle" ,
Quoted 8 lines
drivers/cpuidle/cpuidle-ibm-power.c line 30
Quoted 6 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +6
-0+ .owner = THIS_MODULE, +}; + +#define MAX_IDLE_STATE_COUNT 2 + +static int max_idle_state =MAX_IDLE_STATE_COUNT -1 ;Again, do not use the macro.
Wanting to re-use CPUIDLE_STATE_MAX macro, defined in linux/cpuidle.h similar to what x86 uses. This is def scalable for various platforms , as demonstrated in drivers/idle/intel_idle.c
Quoted 86 lines
Quoted 17 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +16
-0+static struct cpuidle_state *cpuidle_state_table; + +static inline void idle_loop_prolog (unsigned long *in_purr) +{ + *in_purr =mfspr (SPRN_PURR ); +/* +* Indicate to the HV that we are idle. Now would be +* a good time to find other work to dispatch. +*/ +get_lppaca ()->idle =1 ; +} + +static inline void idle_loop_epilog (unsigned long in_purr) +{ +get_lppaca ()->wait_state_cycles +=mfspr (SPRN_PURR ) - in_purr; +get_lppaca ()->idle =0 ;drivers/cpuidle/cpuidle-ibm-power.c line 56
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+} + +static int snooze_loop (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index) +{ +unsigned long in_purr; + +idle_loop_prolog (&in_purr); +local_irq_enable ();
snooze_loophas already registered in cpuidle framework to handle snooze state. where disable the irq? Why do "enable" here?Quoted 27 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +27
-0+/* +* States for dedicated partition case. +*/ +static struct cpuidle_state dedicated_states[MAX_IDLE_STATE_COUNT ] = { + {/* Snooze */ + .name ="snooze" , + .desc ="snooze" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =0 , + .target_residency =0 , + .enter = &snooze_loop }, + {/* CEDE */ + .name ="CEDE" , + .desc ="CEDE" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =10 , + .target_residency =100 , + .enter = &dedicated_cede_loop }, +}; + +/* +* States for shared partition case. +*/ +static struct cpuidle_state shared_states[MAX_IDLE_STATE_COUNT ] = { + {/* Shared Cede */ + .name ="Shared Cede" , + .desc ="Shared Cede" ,drivers/cpuidle/cpuidle-ibm-power.c
Quoted 10 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +10
-0+ .flags = CPUIDLE_FLAG_TIME_VALID, + .exit_latency = 0, + .target_residency = 0, + .enter = &shared_cede_loop }, +}; + +static void __exit power_processor_idle_exit(void) +{ + +unregister_cpu_notifier (&setup_hotplug_notifier);Remove a blank line.
drivers/cpuidle/cpuidle-ibm-power.c line 301
Quoted 7 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +7
-0+cpuidle_unregister (&power_idle_driver); +return ; +} + +module_init (power_processor_idle_init); +module_exit (power_processor_idle_exit); +Did you have tested the module? If not tested, please don't use the module.
Quoted 4 lines
drivers/cpuidle/cpuidle-ibm-power.c new file +3
-0+MODULE_AUTHOR ("Deepthi Dharwar <[email protected]>" ); +MODULE_DESCRIPTION ("Cpuidle driver for IBM POWER platforms" ); +MODULE_LICENSE ("GPL" );
PATCH v5 4/5 POWER/cpuidle: Enable powernv cpuidle support.
In reply to Deepthi DharwarThe following patch extends the current power backend idle driver to the powernv platform.
--- drivers/cpuidle/cpuidle-ibm-power.c | 37 ++++++++++++++++++++++++++++++++--- 1 file changed, 34 insertions(+), 3 deletions(-)
drivers/cpuidle/cpuidle-ibm-power.c +34 -3
@@ -50,9 +50,11 @@ static int snooze_loop(struct cpuidle_device *dev, struct cpuidle_driver *drv, int index) { -unsigned long in_purr; +unsigned long in_purr =0 ; + +if (firmware_has_feature (FW_FEATURE_SPLPAR )) +idle_loop_prolog (&in_purr); -idle_loop_prolog (&in_purr);local_irq_enable ();set_thread_flag (TIF_POLLING_NRFLAG );
@@ -66,7 +68,8 @@ static int snooze_loop(struct cpuidle_device *dev,clear_thread_flag (TIF_POLLING_NRFLAG );smp_mb (); -idle_loop_epilog (in_purr); +if (firmware_has_feature (FW_FEATURE_SPLPAR )) +idle_loop_epilog (in_purr);return index; }
@@ -129,6 +132,15 @@ static int shared_cede_loop(struct cpuidle_device *dev,return index; } +static int nap_loop (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index) +{ +ppc64_runlatch_off (); +power7_idle (); +return index; +} +/* * States for dedicated partition case. */
@@ -162,6 +174,23 @@ static struct cpuidle_state shared_states[MAX_IDLE_STATE_COUNT] = { .enter = &shared_cede_loop }, }; +static struct cpuidle_state powernv_states[MAX_IDLE_STATE_COUNT ] = { + {/* Snooze */ + .name ="snooze" , + .desc ="snooze" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =0 , + .target_residency =0 , + .enter = &snooze_loop }, + {/* NAP */ + .name ="NAP" , + .desc ="NAP" , + .flags =CPUIDLE_FLAG_TIME_VALID , + .exit_latency =10 , + .target_residency =100 , + .enter = &nap_loop }, +}; +void update_smt_snooze_delay (int cpu,int residency) {struct cpuidle_driver *drv =cpuidle_get_driver ();
@@ -261,6 +290,8 @@ static int power_idle_probe(void) cpuidle_state_table = shared_states; else cpuidle_state_table = dedicated_states; + }else if (firmware_has_feature (FW_FEATURE_OPALv3)) { + cpuidle_state_table = powernv_states; }else return -ENODEV ;
PATCH v5 5/5 powernv/cpuidle: Enable idle powernv cpu to call into the cpuidle framework.
In reply to Deepthi DharwarThis patch enables idle cpu on the powernv platform to hook on to the cpuidle framework, if available, else call on to default idle platform code.
--- arch/powerpc/platforms/powernv/setup.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)
arch/powerpc/platforms/powernv/setup.c +13 -1
@@ -25,6 +25,7 @@#include <linux/of.h> #include <linux/interrupt.h> #include <linux/bug.h> +#include <linux/cpuidle.h> #include <asm/machdep.h> #include <asm/firmware.h>
@@ -175,6 +176,17 @@ static void __init pnv_setup_machdep_rtas(void) }#endif /* CONFIG_PPC_POWERNV_RTAS */ +void powernv_idle (void ) +{ +/* Hook to cpuidle framework if available, else +* call on default platform idle code +*/ +if (cpuidle_idle_call ()) { +HMT_low (); +HMT_very_low (); + } +} +static int __init pnv_probe (void ) {unsigned long root =of_get_flat_dt_root ();
@@ -205,7 +217,7 @@ define_machine(powernv) { .show_cpuinfo = pnv_show_cpuinfo, .progress = pnv_progress, .machine_shutdown = pnv_shutdown, - .power_save = power7_idle, + .power_save = powernv_idle, .calibrate_decr = generic_calibrate_decr, #ifdef CONFIG_KEXEC .kexec_cpu_down = pnv_kexec_cpu_down,