Provide vcpu dispatch statistics
Subsystems (2)
v2 cover Provide vcpu dispatch statistics
Changes since v1:
- Changes have been done primarily to address a few build issues, by moving changes out of mm/numa.c to other places.
- Patches 1-5,8-9 have been rebased with patches 4 and 5 being interchanged, and patch 4 incorporating #ifdef CONFIG_PPC_SPLPAR.
- Patch 6 is new and moves vphn code under pseries.
- Patch 7 has changes to move most of the code from mm/numa.c into platforms/pseries/lpar.c, along with guarding some of these under CONFIG_PPC_SPLPAR.
Changes since RFC:
- Patches 1/8 to 5/8: no changes, except rebase to powerpc/merge
- Patch 6/8: The mutex guarding the vphn hcall has been dropped. It was only meant to serialize hcalls issued when stats are initially enabled. However, in reality, the various per-cpu workers will be scheduled at slightly different times and chances of hcalls for retrieving the same associativity information at the same time is very less. Even in that case, there are no other side effects.
- Patch 6/8: The third column for vcpu dispatches on the same core, but different thread has been dropped and merged with the second column.
- Patch 7/8: new patch to ensure we don't take too much time while enabling/disabling statistics on large systems with heavy workload.
- Patch 8/8: new patch adding a document describing the fields in the procfs file.
--
This series adds a new procfs file /proc/powerpc/vcpudispatch_stats for providing statistics around how the LPAR processors are dispatched by the POWER Hypervisor, in a shared LPAR environment. Patch 7/9 has more details on how the statistics are gathered.
An example output:
$ sudo cat /proc/powerpc/vcpudispatch_stats
cpu0 6839 4126 2683 30 0 6821 18 0
cpu1 2515 1274 1229 12 0 2509 6 0
cpu2 2317 1198 1109 10 0 2312 5 0
cpu3 2259 1165 1088 6 0 2256 3 0
cpu4 2205 1143 1056 6 0 2202 3 0
cpu5 2165 1121 1038 6 0 2162 3 0
cpu6 2183 1127 1050 6 0 2180 3 0
cpu7 2193 1133 1052 8 0 2187 6 0
cpu8 2165 1115 1032 18 0 2156 9 0
cpu9 2301 1252 1033 16 0 2293 8 0
cpu10 2197 1138 1041 18 0 2187 10 0
cpu11 2273 1185 1062 26 0 2260 13 0
cpu12 2186 1125 1043 18 0 2177 9 0
cpu13 2161 1115 1030 16 0 2153 8 0
cpu14 2206 1153 1033 20 0 2196 10 0
cpu15 2163 1115 1032 16 0 2155 8 0In the output above, for vcpu0, there have been 6839 dispatches since statistics were enabled. 4126 of those dispatches were on the same physical cpu as the last time. 2683 were on a different core, but within the same chip, while 30 dispatches were on a different chip compared to its last dispatch.
Also, out of the total of 6839 dispatches, we see that there have been 6821 dispatches on the vcpu's home node, while 18 dispatches were outside its home node, on a neighbouring chip.
- Naveen
Naveen N. Rao (9):
powerpc/pseries: Use macros for referring to the DTL enable mask powerpc/pseries: Do not save the previous DTL mask value powerpc/pseries: Factor out DTL buffer allocation and registration
routines
powerpc/pseries: Introduce helpers to gatekeep DTLB usage powerpc/pseries: Generalize hcall_vphn()
powerpc/pseries: Move mm/book3s64/vphn.c under platforms/pseries/ powerpc/pseries: Provide vcpu dispatch statistics powerpc/pseries: Protect against hogging the cpu while setting up the
stats
powerpc/pseries: Add documentation for vcpudispatch_stats
Documentation/powerpc/vcpudispatch_stats.txt | 68 ++
arch/powerpc/include/asm/lppaca.h | 11 +
arch/powerpc/include/asm/plpar_wrappers.h | 23 +
arch/powerpc/include/asm/topology.h | 6 +
arch/powerpc/mm/book3s64/Makefile | 1 -
arch/powerpc/mm/book3s64/vphn.h | 16 -
arch/powerpc/mm/numa.c | 34 +-
arch/powerpc/platforms/pseries/Makefile | 1 +
arch/powerpc/platforms/pseries/dtl.c | 22 +-
arch/powerpc/platforms/pseries/lpar.c | 647 +++++++++++++++++-
arch/powerpc/platforms/pseries/setup.c | 34 +-
.../{mm/book3s64 => platforms/pseries}/vphn.c | 16 +-
12 files changed, 784 insertions(+), 95 deletions(-)
create mode 100644 Documentation/powerpc/vcpudispatch_stats.txt
delete mode 100644 arch/powerpc/mm/book3s64/vphn.h
rename arch/powerpc/{mm/book3s64 => platforms/pseries}/vphn.c (83%)Signature · 1 line
2.21.0
PATCH v2 1/9 powerpc/pseries: Use macros for referring to the DTL enable mask
Introduce macros to encode the DTL enable mask fields and use those instead of hardcoding numbers.
--- arch/powerpc/include/asm/lppaca.h | 11 +++++++++++ arch/powerpc/platforms/pseries/dtl.c | 8 +------- arch/powerpc/platforms/pseries/lpar.c | 2 +- arch/powerpc/platforms/pseries/setup.c | 2 +- 4 files changed, 14 insertions(+), 9 deletions(-)
arch/powerpc/include/asm/lppaca.h +11 -0
@@ -141,6 +141,17 @@ struct dtl_entry {#define DISPATCH_LOG_BYTES 4096/* bytes per cpu */ #define N_DISPATCH_LOG (DISPATCH_LOG_BYTES / sizeof(struct dtl_entry)) +/* +* Dispatch trace log event enable mask: +* 0x1: voluntary virtual processor waits +* 0x2: time-slice preempts +* 0x4: virtual partition memory page faults +*/ +#define DTL_LOG_CEDE 0x1 +#define DTL_LOG_PREEMPT 0x2 +#define DTL_LOG_FAULT 0x4 +#define DTL_LOG_ALL (DTL_LOG_CEDE | DTL_LOG_PREEMPT | DTL_LOG_FAULT) +extern struct kmem_cache *dtl_cache;/*
arch/powerpc/platforms/pseries/dtl.c +1 -7
@@ -27,13 +27,7 @@ struct dtl { }; static DEFINE_PER_CPU(struct dtl, cpu_dtl); -/* -* Dispatch trace log event mask: -* 0x7: 0x1: voluntary virtual processor waits -* 0x2: time-slice preempts -* 0x4: virtual partition memory page faults -*/ -static u8 dtl_event_mask =0x7 ; +static u8 dtl_event_mask =DTL_LOG_ALL ;/*
arch/powerpc/platforms/pseries/lpar.c +1 -1
@@ -113,7 +113,7 @@ void vpa_init(int cpu)pr_err ("WARNING: DTL registration of cpu %d (hw %d) " "failed with %ld\n" ,smp_processor_id (), hwcpu, ret); -lppaca_of (cpu).dtl_enable_mask =2 ; +lppaca_of (cpu).dtl_enable_mask =DTL_LOG_PREEMPT ; } }
arch/powerpc/platforms/pseries/setup.c +1 -1
@@ -312,7 +312,7 @@ static int alloc_dispatch_logs(void)pr_err ("WARNING: DTL registration of cpu %d (hw %d) failed " "with %d\n" ,smp_processor_id (),hard_smp_processor_id (), ret); -get_paca ()->lppaca_ptr->dtl_enable_mask =2 ; +get_paca ()->lppaca_ptr->dtl_enable_mask =DTL_LOG_PREEMPT ;return 0 ; }
Signature · 1 line
2.21.0
PATCH v2 2/9 powerpc/pseries: Do not save the previous DTL mask value
In reply to Naveen N. RaoWhen CONFIG_VIRT_CPU_ACCOUNTING_NATIVE is enabled, we always initialize DTL enable mask to DTL_LOG_PREEMPT (0x2). There are no other places where the mask is changed. As such, when reading the DTL log buffer through debugfs, there is no need to save and restore the previous mask value.
We don't need to save and restore the earlier mask value if CONFIG_VIRT_CPU_ACCOUNTING_NATIVE is not enabled. So, remove the field from the structure as well.
--- arch/powerpc/platforms/pseries/dtl.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
arch/powerpc/platforms/pseries/dtl.c +1 -3
@@ -42,7 +42,6 @@ struct dtl_ring {struct dtl_entry *write_ptr;struct dtl_entry *buf;struct dtl_entry *buf_end; -u8 saved_dtl_mask; }; static DEFINE_PER_CPU(struct dtl_ring, dtl_rings);
@@ -92,7 +91,6 @@ static int dtl_start(struct dtl *dtl) dtlr->write_ptr = dtl->buf;/* enable event logging */ - dtlr->saved_dtl_mask =lppaca_of (dtl->cpu).dtl_enable_mask;lppaca_of (dtl->cpu).dtl_enable_mask |= dtl_event_mask; dtl_consumer = consume_dtle;
@@ -110,7 +108,7 @@ static void dtl_stop(struct dtl *dtl) dtlr->buf =NULL ;/* restore dtl_enable_mask */ -lppaca_of (dtl->cpu).dtl_enable_mask = dtlr->saved_dtl_mask; +lppaca_of (dtl->cpu).dtl_enable_mask =DTL_LOG_PREEMPT ;if (atomic_dec_and_test (&dtl_count)) dtl_consumer =NULL ;
Signature · 1 line
2.21.0
PATCH v2 3/9 powerpc/pseries: Factor out DTL buffer allocation and registration routines
In reply to Naveen N. RaoIntroduce new helpers for DTL buffer allocation and registration and have the existing code use those.
--- arch/powerpc/include/asm/plpar_wrappers.h | 2 + arch/powerpc/platforms/pseries/lpar.c | 66 ++++++++++++++++------- arch/powerpc/platforms/pseries/setup.c | 34 +----------- 3 files changed, 52 insertions(+), 50 deletions(-)
arch/powerpc/include/asm/plpar_wrappers.h +2 -0
@@ -88,6 +88,8 @@ static inline long register_dtl(unsigned long cpu, unsigned long vpa)return vpa_call (H_VPA_REG_DTL , cpu, vpa); } +extern void register_dtl_buffer (int cpu); +extern void alloc_dtl_buffers (void );extern void vpa_init (int cpu); static inline long plpar_pte_enter(unsigned long flags,
arch/powerpc/platforms/pseries/lpar.c +48 -18
83 lines not on this page · raw message
arch/powerpc/platforms/pseries/setup.c +2 -32
@@ -273,46 +273,16 @@ struct kmem_cache *dtl_cache;*/ static int alloc_dispatch_logs (void ) { -int cpu, ret; -struct paca_struct *pp; -struct dtl_entry *dtl; -if (!firmware_has_feature (FW_FEATURE_SPLPAR ))return 0 ;if (!dtl_cache)return 0 ; -for_each_possible_cpu (cpu) { - pp = paca_ptrs[cpu]; - dtl =kmem_cache_alloc (dtl_cache,GFP_KERNEL ); -if (!dtl) { -pr_warn ("Failed to allocate dispatch trace log for cpu %d\n" , - cpu); -pr_warn ("Stolen time statistics will be unreliable\n" ); -break ; - } - - pp->dtl_ridx =0 ; - pp->dispatch_log = dtl; - pp->dispatch_log_end = dtl +N_DISPATCH_LOG ; - pp->dtl_curr = dtl; - } +alloc_dtl_buffers ();/* Register the DTL for the current (boot) cpu */ - dtl =get_paca ()->dispatch_log; -get_paca ()->dtl_ridx =0 ; -get_paca ()->dtl_curr = dtl; -get_paca ()->lppaca_ptr->dtl_idx =0 ; - -/* hypervisor reads buffer length from this field */ - dtl->enqueue_to_dispatch_time =cpu_to_be32 (DISPATCH_LOG_BYTES ); - ret =register_dtl (hard_smp_processor_id (),__pa (dtl)); -if (ret) -pr_err ("WARNING: DTL registration of cpu %d (hw %d) failed " -"with %d\n" ,smp_processor_id (), -hard_smp_processor_id (), ret); -get_paca ()->lppaca_ptr->dtl_enable_mask =DTL_LOG_PREEMPT ; +register_dtl_buffer (smp_processor_id ());return 0 ; }
Signature · 1 line
2.21.0
PATCH v2 4/9 powerpc/pseries: Introduce helpers to gatekeep DTLB usage
In reply to Naveen N. RaoSince we would be introducing a new user of the DTL buffer in a subsequent patch, add helpers to gatekeep use of the DTL buffer. The current usage of the DTL buffer from debugfs is at a per-cpu level (corresponding to the cpu debugfs file that is opened). Subsequently, we will have users enabling/accessing DTLB for all online cpus. These helpers allow any number of per-cpu users, or a single global user exclusively.
--- arch/powerpc/include/asm/plpar_wrappers.h | 2 ++ arch/powerpc/platforms/pseries/dtl.c | 10 +++++- arch/powerpc/platforms/pseries/lpar.c | 38 +++++++++++++++++++++++ 3 files changed, 49 insertions(+), 1 deletion(-)
arch/powerpc/include/asm/plpar_wrappers.h +2 -0
@@ -88,6 +88,8 @@ static inline long register_dtl(unsigned long cpu, unsigned long vpa)return vpa_call (H_VPA_REG_DTL , cpu, vpa); } +extern bool register_dtl_buffer_access (bool global); +extern void unregister_dtl_buffer_access (bool global);extern void register_dtl_buffer (int cpu);extern void alloc_dtl_buffers (void );extern void vpa_init (int cpu);
arch/powerpc/platforms/pseries/dtl.c +9 -1
@@ -180,11 +180,15 @@ static int dtl_enable(struct dtl *dtl)if (dtl->buf)return -EBUSY ; +if (register_dtl_buffer_access (false)) +return -EBUSY ; + n_entries = dtl_buf_entries; buf =kmem_cache_alloc_node (dtl_cache,GFP_KERNEL ,cpu_to_node (dtl->cpu));if (!buf) {printk (KERN_WARNING "%s: buffer alloc failed for cpu %d\n" , __func__, dtl->cpu); +unregister_dtl_buffer_access (false);return -ENOMEM ; }
@@ -201,8 +205,11 @@ static int dtl_enable(struct dtl *dtl) }spin_unlock (&dtl->lock); -if (rc) +if (rc) { +unregister_dtl_buffer_access (false);kmem_cache_free (dtl_cache, buf); + } +return rc; }
@@ -214,6 +221,7 @@ static void dtl_disable(struct dtl *dtl) dtl->buf =NULL ; dtl->buf_entries =0 ;spin_unlock (&dtl->lock); +unregister_dtl_buffer_access (false); }/* file interface */
arch/powerpc/platforms/pseries/lpar.c +38 -0
@@ -99,6 +99,44 @@ void register_dtl_buffer(int cpu) } } +#ifdef CONFIG_PPC_SPLPAR +static DEFINE_SPINLOCK (dtl_buffer_refctr_lock); +static unsigned int dtl_buffer_global_refctr, dtl_buffer_percpu_refctr; + +bool register_dtl_buffer_access (bool global) +{ +int rc =0 ; + +spin_lock (&dtl_buffer_refctr_lock); + +if ((global && (dtl_buffer_global_refctr || dtl_buffer_percpu_refctr)) + || (!global && dtl_buffer_global_refctr)) { + rc =-1 ; + }else { +if (global) + dtl_buffer_global_refctr++; +else + dtl_buffer_percpu_refctr++; + } + +spin_unlock (&dtl_buffer_refctr_lock); + +return rc; +} + +void unregister_dtl_buffer_access (bool global) +{ +spin_lock (&dtl_buffer_refctr_lock); + +if (global) + dtl_buffer_global_refctr--; +else + dtl_buffer_percpu_refctr--; + +spin_unlock (&dtl_buffer_refctr_lock); +} +#endif /* CONFIG_PPC_SPLPAR */ +void vpa_init (int cpu) {int hwcpu =get_hard_smp_processor_id (cpu);
Signature · 1 line
2.21.0
PATCH v2 5/9 powerpc/pseries: Generalize hcall_vphn()
In reply to Naveen N. RaoH_HOME_NODE_ASSOCIATIVITY hcall can take two different flags and return different associativity information in each case. Generalize the existing hcall_vphn() function to take flags as an argument and to return the result. Update the only existing user to pass the proper arguments.
--- arch/powerpc/mm/book3s64/vphn.h | 8 ++++++++ arch/powerpc/mm/numa.c | 27 +++++++++++++-------------- 2 files changed, 21 insertions(+), 14 deletions(-)
arch/powerpc/mm/book3s64/vphn.h +8 -0
@@ -11,6 +11,14 @@*/ #define VPHN_ASSOC_BUFSIZE (VPHN_REGISTER_COUNT*sizeof(u64)/sizeof(u16) + 1) +/* +* The H_HOME_NODE_ASSOCIATIVITY hcall takes two values for flags: +* 1 for retrieving associativity information for a guest cpu +* 2 for retrieving associativity information for a host/hypervisor cpu +*/ +#define VPHN_FLAG_VCPU 1 +#define VPHN_FLAG_PCPU 2 +extern int vphn_unpack_associativity (const long *packed,__be32 *unpacked);#endif
arch/powerpc/mm/numa.c +13 -14
@@ -1083,6 +1083,17 @@ static void reset_topology_timer(void);static int topology_timer_secs =1 ;static int topology_inited; +static long hcall_vphn (unsigned long cpu,u64 flags,__be32 *associativity) +{ +long rc; +long retbuf[PLPAR_HCALL9_BUFSIZE ] = {0 }; + + rc =plpar_hcall9 (H_HOME_NODE_ASSOCIATIVITY , retbuf, flags, cpu); +vphn_unpack_associativity (retbuf, associativity); + +return rc; +} +/* * Change polling interval for associativity changes. */
@@ -1161,25 +1172,13 @@ static int update_cpu_associativity_changes_mask(void)* Retrieve the new associativity information for a virtual processor's * home node. */ -static long hcall_vphn (unsigned long cpu,__be32 *associativity) -{ -long rc; -long retbuf[PLPAR_HCALL9_BUFSIZE ] = {0 }; -u64 flags =1 ; -int hwcpu =get_hard_smp_processor_id (cpu); - - rc =plpar_hcall9 (H_HOME_NODE_ASSOCIATIVITY , retbuf, flags, hwcpu); -vphn_unpack_associativity (retbuf, associativity); - -return rc; -} -static long vphn_get_associativity (unsigned long cpu,__be32 *associativity) {long rc; - rc =hcall_vphn (cpu, associativity); + rc =hcall_vphn (get_hard_smp_processor_id (cpu), +VPHN_FLAG_VCPU , associativity);switch (rc) {case H_FUNCTION :
Signature · 1 line
2.21.0
PATCH v2 6/9 powerpc/pseries: Move mm/book3s64/vphn.c under platforms/pseries/
In reply to Naveen N. Raohcall_vphn() is specific to pseries and will be used in a subsequent patch. So, move it to a more appropriate place under arch/powerpc/platforms/pseries. Also merge vphn.h into plpar_wrappers.h
---
arch/powerpc/include/asm/plpar_wrappers.h | 19 +++++++++++++++
arch/powerpc/mm/book3s64/Makefile | 1 -
arch/powerpc/mm/book3s64/vphn.h | 24 -------------------
arch/powerpc/mm/numa.c | 13 +---------
arch/powerpc/platforms/pseries/Makefile | 1 +
.../{mm/book3s64 => platforms/pseries}/vphn.c | 16 +++++++++++--
6 files changed, 35 insertions(+), 39 deletions(-)
delete mode 100644 arch/powerpc/mm/book3s64/vphn.h
rename arch/powerpc/{mm/book3s64 => platforms/pseries}/vphn.c (83%)arch/powerpc/include/asm/plpar_wrappers.h +19 -0
@@ -11,6 +11,25 @@#include <asm/paca.h> #include <asm/page.h> +/* The H_HOME_NODE_ASSOCIATIVITY h_call returns 6 64-bit registers. */ +#define VPHN_REGISTER_COUNT 6 + +/* +* 6 64-bit registers unpacked into up to 24 be32 associativity values. To +* form the complete property we have to add the length in the first cell. +*/ +#define VPHN_ASSOC_BUFSIZE (VPHN_REGISTER_COUNT*sizeof(u64)/sizeof(u16) + 1) + +/* +* The H_HOME_NODE_ASSOCIATIVITY hcall takes two values for flags: +* 1 for retrieving associativity information for a guest cpu +* 2 for retrieving associativity information for a host/hypervisor cpu +*/ +#define VPHN_FLAG_VCPU 1 +#define VPHN_FLAG_PCPU 2 + +long hcall_vphn (unsigned long cpu,u64 flags,__be32 *associativity); +static inline long poll_pending (void ) {return plpar_hcall_norets (H_POLL_PENDING );
arch/powerpc/mm/book3s64/Makefile +0 -1
@@ -10,7 +10,6 @@ obj-$(CONFIG_PPC_NATIVE) += hash_native.o obj-$(CONFIG_PPC_RADIX_MMU) += radix_pgtable.o radix_tlb.o obj-$(CONFIG_PPC_4K_PAGES) += hash_4k.o obj-$(CONFIG_PPC_64K_PAGES) += hash_64k.o -obj-$(CONFIG_PPC_SPLPAR) += vphn.o obj-$(CONFIG_HUGETLB_PAGE) += hash_hugetlbpage.o ifdef CONFIG_HUGETLB_PAGE obj-$(CONFIG_PPC_RADIX_MMU) += radix_hugetlbpage.o
arch/powerpc/mm/book3s64/vphn.h deleted +0 -24
@@ -1,24 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ARCH_POWERPC_MM_VPHN_H_ -#define _ARCH_POWERPC_MM_VPHN_H_ - -/* The H_HOME_NODE_ASSOCIATIVITY h_call returns 6 64-bit registers. */ -#define VPHN_REGISTER_COUNT 6 - -/* -* 6 64-bit registers unpacked into up to 24 be32 associativity values. To -* form the complete property we have to add the length in the first cell. -*/ -#define VPHN_ASSOC_BUFSIZE (VPHN_REGISTER_COUNT*sizeof(u64)/sizeof(u16) + 1) - -/* -* The H_HOME_NODE_ASSOCIATIVITY hcall takes two values for flags: -* 1 for retrieving associativity information for a guest cpu -* 2 for retrieving associativity information for a host/hypervisor cpu -*/ -#define VPHN_FLAG_VCPU 1 -#define VPHN_FLAG_PCPU 2 - -extern int vphn_unpack_associativity (const long *packed,__be32 *unpacked); - -#endif
arch/powerpc/mm/numa.c +1 -12
@@ -1064,7 +1064,7 @@ u64 memory_hotplug_max(void)/* Virtual Processor Home Node (VPHN) support */ #ifdef CONFIG_PPC_SPLPAR -#include "book3s64/vphn.h" +#include <asm/plpar_wrappers.h> struct topology_update_data { struct topology_update_data *next;
@@ -1083,17 +1083,6 @@ static void reset_topology_timer(void);static int topology_timer_secs =1 ;static int topology_inited; -static long hcall_vphn (unsigned long cpu,u64 flags,__be32 *associativity) -{ -long rc; -long retbuf[PLPAR_HCALL9_BUFSIZE ] = {0 }; - - rc =plpar_hcall9 (H_HOME_NODE_ASSOCIATIVITY , retbuf, flags, cpu); -vphn_unpack_associativity (retbuf, associativity); - -return rc; -} -/* * Change polling interval for associativity changes. */
arch/powerpc/platforms/pseries/Makefile +1 -0
@@ -25,6 +25,7 @@ obj-$(CONFIG_LPARCFG) += lparcfg.o obj-$(CONFIG_IBMVIO) += vio.o obj-$(CONFIG_IBMEBUS) += ibmebus.o obj-$(CONFIG_PAPR_SCM) += papr_scm.o +obj-$(CONFIG_PPC_SPLPAR) += vphn.o ifdef CONFIG_PPC_PSERIES obj-$(CONFIG_SUSPEND) += suspend.o
arch/powerpc/platforms/pseries/vphn.c renamed from arch/powerpc/mm/book3s64/vphn.c +14 -2
@@ -1,6 +1,7 @@// SPDX-License-Identifier: GPL-2.0 +#include <asm/plpar_wrappers.h> #include <asm/byteorder.h> -#include "vphn.h" +#include <asm/hvcall.h> /* * The associativity domain numbers are returned from the hypervisor as a
@@ -22,7 +23,7 @@* * Convert to the sequence they would appear in the ibm,associativity property. */ -int vphn_unpack_associativity (const long *packed,__be32 *unpacked) +static int vphn_unpack_associativity (const long *packed,__be32 *unpacked) {__be64 be_packed[VPHN_REGISTER_COUNT ];int i, nr_assoc_doms =0 ;
@@ -71,3 +72,14 @@ int vphn_unpack_associativity(const long *packed, __be32 *unpacked)return nr_assoc_doms; } + +long hcall_vphn (unsigned long cpu,u64 flags,__be32 *associativity) +{ +long rc; +long retbuf[PLPAR_HCALL9_BUFSIZE ] = {0 }; + + rc =plpar_hcall9 (H_HOME_NODE_ASSOCIATIVITY , retbuf, flags, cpu); +vphn_unpack_associativity (retbuf, associativity); + +return rc; +}
Signature · 1 line
2.21.0
PATCH v2 7/9 powerpc/pseries: Provide vcpu dispatch statistics
In reply to Naveen N. RaoFor Shared Processor LPARs, the POWER Hypervisor maintains a relatively static mapping of the LPAR processors (vcpus) to physical processor chips (representing the "home" node) and tries to always dispatch vcpus on their associated physical processor chip. However, under certain scenarios, vcpus may be dispatched on a different processor chip (away from its home node). The actual physical processor number on which a certain vcpu is dispatched is available to the guest in the 'processor_id' field of each DTL entry.
The guest can discover the home node of each vcpu through the H_HOME_NODE_ASSOCIATIVITY(flags=1) hcall. The guest can also discover the associativity of physical processors, as represented in the DTL entry, through the H_HOME_NODE_ASSOCIATIVITY(flags=2) hcall.
These can then be compared to determine if the vcpu was dispatched on its home node or not. If the vcpu was not dispatched on the home node, it is possible to determine if the vcpu was dispatched in a different chip, socket or drawer.
Introduce a procfs file /proc/powerpc/vcpudispatch_stats that can be used to obtain these statistics. Writing '1' to this file enables collecting the statistics, while writing '0' disables the statistics. The statistics themselves are available by reading the procfs file. By default, the DTLB log for each vcpu is processed 50 times a second so as not to miss any entries. This processing frequency can be changed through /proc/powerpc/vcpudispatch_stats_freq.
--- arch/powerpc/include/asm/topology.h | 6 + arch/powerpc/mm/numa.c | 16 + arch/powerpc/platforms/pseries/lpar.c | 536 +++++++++++++++++++++++++- 3 files changed, 556 insertions(+), 2 deletions(-)
arch/powerpc/include/asm/topology.h +6 -0
@@ -35,6 +35,7 @@ static inline int pcibus_to_node(struct pci_bus *bus) cpu_all_mask : \ cpumask_of_node(pcibus_to_node(bus))) +extern int cpu_distance (__be32 *cpu1_assoc,__be32 *cpu2_assoc);extern int __node_distance (int ,int );#define node_distance (a, b) __node_distance(a, b)
@@ -84,6 +85,11 @@ static inline int numa_update_cpu_topology(bool cpus_locked)static inline void update_numa_cpu_lookup_table (unsigned int cpu,int node) {} +static inline int cpu_distance (__be32 *cpu1_assoc,__be32 *cpu2_assoc) +{ +return 0 ; +} +#endif /* CONFIG_NUMA */ #if defined(CONFIG_NUMA ) && defined(CONFIG_PPC_SPLPAR )
arch/powerpc/mm/numa.c +16 -0
@@ -163,6 +163,22 @@ static void unmap_cpu_from_node(unsigned long cpu) }#endif /* CONFIG_HOTPLUG_CPU || CONFIG_PPC_SPLPAR */ +int cpu_distance (__be32 *cpu1_assoc,__be32 *cpu2_assoc) +{ +int dist =0 ; + +int i, index; + +for (i =0 ; i < distance_ref_points_depth; i++) { + index =be32_to_cpu (distance_ref_points[i]); +if (cpu1_assoc[index] == cpu2_assoc[index]) +break ; + dist++; + } + +return dist; +} +/* must hold reference to node during call */ static const __be32 *of_get_associativity (struct device_node *dev) {
arch/powerpc/platforms/pseries/lpar.c +534 -2
583 lines not on this page · raw message
Signature · 1 line
2.21.0
PATCH v2 8/9 powerpc/pseries: Protect against hogging the cpu while setting up the stats
In reply to Naveen N. RaoWhen enabling or disabling the vcpu dispatch statistics, we do a lot of work including allocating/deallocating memory across all possible cpus for the DTL buffer. In order to guard against hogging the cpu for too long, track the time we're taking and yield the processor if necessary.
--- arch/powerpc/include/asm/plpar_wrappers.h | 2 +- arch/powerpc/platforms/pseries/lpar.c | 29 ++++++++++++++++------- arch/powerpc/platforms/pseries/setup.c | 2 +- 3 files changed, 22 insertions(+), 11 deletions(-)
arch/powerpc/include/asm/plpar_wrappers.h +1 -1
@@ -110,7 +110,7 @@ static inline long register_dtl(unsigned long cpu, unsigned long vpa)extern bool register_dtl_buffer_access (bool global);extern void unregister_dtl_buffer_access (bool global);extern void register_dtl_buffer (int cpu); -extern void alloc_dtl_buffers (void ); +extern void alloc_dtl_buffers (unsigned long *time_limit);extern void vpa_init (int cpu); static inline long plpar_pte_enter(unsigned long flags,
arch/powerpc/platforms/pseries/lpar.c +20 -9
@@ -62,7 +62,7 @@ static u8 dtl_mask = DTL_LOG_PREEMPT;static u8 dtl_mask;#endif -void alloc_dtl_buffers (void ) +void alloc_dtl_buffers (unsigned long *time_limit) {int cpu;struct paca_struct *pp;
@@ -86,6 +86,11 @@ void alloc_dtl_buffers(void) pp->dispatch_log = dtl; pp->dispatch_log_end = dtl +N_DISPATCH_LOG ; pp->dtl_curr = dtl; + +if (time_limit &&time_after (jiffies, *time_limit)) { +cond_resched (); + *time_limit = jiffies +HZ ; + } } }
@@ -157,7 +162,7 @@ static int vcpudispatch_stats_freq = 50;static __be32 *vcpu_associativity, *pcpu_associativity; -static void free_dtl_buffers (void ) +static void free_dtl_buffers (unsigned long *time_limit) {#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE int cpu;
@@ -172,6 +177,11 @@ static void free_dtl_buffers(void) pp->dispatch_log =0 ; pp->dispatch_log_end =0 ; pp->dtl_curr =0 ; + +if (time_limit &&time_after (jiffies, *time_limit)) { +cond_resched (); + *time_limit = jiffies +HZ ; + } }#endif }
@@ -464,7 +474,7 @@ static void reset_global_dtl_mask(void)lppaca_of (cpu).dtl_enable_mask = dtl_mask; } -static int dtl_worker_enable (void ) +static int dtl_worker_enable (unsigned long *time_limit) {int rc =0 , state;
@@ -483,13 +493,13 @@ static int dtl_worker_enable(void)set_global_dtl_mask (DTL_LOG_ALL );/* Setup dtl buffers and register those */ -alloc_dtl_buffers (); +alloc_dtl_buffers (time_limit); state =cpuhp_setup_state (CPUHP_AP_ONLINE_DYN ,"powerpc/dtl:online" , dtl_worker_online, dtl_worker_offline);if (state <0 ) {pr_err ("vcpudispatch_stats: unable to setup workqueue for DTL processing\n" ); -free_dtl_buffers (); +free_dtl_buffers (time_limit);reset_global_dtl_mask ();unregister_dtl_buffer_access (1 ); rc = -EINVAL ;
@@ -503,14 +513,14 @@ static int dtl_worker_enable(void)return rc; } -static void dtl_worker_disable (void ) +static void dtl_worker_disable (unsigned long *time_limit) {mutex_lock (&dtl_worker_mutex); dtl_worker_refctr--;if (!dtl_worker_refctr) {cpuhp_remove_state (dtl_worker_state); dtl_worker_state =0 ; -free_dtl_buffers (); +free_dtl_buffers (time_limit);reset_global_dtl_mask ();unregister_dtl_buffer_access (1 ); }
@@ -520,6 +530,7 @@ static void dtl_worker_disable(void)static ssize_t vcpudispatch_stats_write (struct file *file,const char __user *p,size_t count,loff_t *ppos) { +unsigned long time_limit = jiffies +HZ ;struct vcpu_dispatch_data *disp;int rc, cmd, cpu;char buf[16 ];
@@ -552,13 +563,13 @@ static ssize_t vcpudispatch_stats_write(struct file *file, const char __user *p, disp->last_disp_cpu =-1 ; } - rc =dtl_worker_enable (); + rc =dtl_worker_enable (&time_limit);if (rc) {destroy_cpu_associativity ();return rc; } }else { -dtl_worker_disable (); +dtl_worker_disable (&time_limit);destroy_cpu_associativity (); }
arch/powerpc/platforms/pseries/setup.c +1 -1
@@ -279,7 +279,7 @@ static int alloc_dispatch_logs(void)if (!dtl_cache)return 0 ; -alloc_dtl_buffers (); +alloc_dtl_buffers (0 );/* Register the DTL for the current (boot) cpu */ register_dtl_buffer (smp_processor_id ());
Signature · 1 line
2.21.0
PATCH v2 9/9 powerpc/pseries: Add documentation for vcpudispatch_stats
In reply to Naveen N. RaoAdd a document describing the fields provided by/proc/powerpc/vcpudispatch_stats.
--- Documentation/powerpc/vcpudispatch_stats.txt | 68 ++++++++++++++++++++ 1 file changed, 68 insertions(+) create mode 100644 Documentation/powerpc/vcpudispatch_stats.txt
Documentation/powerpc/vcpudispatch_stats.txt new file +68 -0
@@ -0,0 +1,68 @@ +VCPU Dispatch Statistics: +========================= + +For Shared Processor LPARs, the POWER Hypervisor maintains a relatively +static mapping of the LPAR processors (vcpus) to physical processor +chips (representing the "home" node) and tries to always dispatch vcpus +on their associated physical processor chip. However, under certain +scenarios, vcpus may be dispatched on a different processor chip (away +from its home node). + +/proc/powerpc/vcpudispatch_stats can be used to obtain statistics +related to the vcpu dispatch behavior. Writing '1' to this file enables +collecting the statistics, while writing '0' disables the statistics. +By default, the DTLB log for each vcpu is processed 50 times a second so +as not to miss any entries. This processing frequency can be changed +through /proc/powerpc/vcpudispatch_stats_freq. + +The statistics themselves are available by reading the procfs file +/proc/powerpc/vcpudispatch_stats. Each line in the output corresponds to +a vcpu as represented by the first field, followed by 8 numbers. + +The first number corresponds to: +1. total vcpu dispatches since the beginning of statistics collection + +The next 4 numbers represent vcpu dispatch dispersions: +2. number of times this vcpu was dispatched on the same processor as last + time +3. number of times this vcpu was dispatched on a different processor core + as last time, but within the same chip +4. number of times this vcpu was dispatched on a different chip +5. number of times this vcpu was dispatches on a different socket/drawer +(next numa boundary) + +The final 3 numbers represent statistics in relation to the home node of +the vcpu: +6. number of times this vcpu was dispatched in its home node (chip) +7. number of times this vcpu was dispatched in a different node +8. number of times this vcpu was dispatched in a node further away (numa +distance) + +An example output: + $ sudo cat /proc/powerpc/vcpudispatch_stats + cpu0 6839 4126 2683 30 0 6821 18 0 + cpu1 2515 1274 1229 12 0 2509 6 0 + cpu2 2317 1198 1109 10 0 2312 5 0 + cpu3 2259 1165 1088 6 0 2256 3 0 + cpu4 2205 1143 1056 6 0 2202 3 0 + cpu5 2165 1121 1038 6 0 2162 3 0 + cpu6 2183 1127 1050 6 0 2180 3 0 + cpu7 2193 1133 1052 8 0 2187 6 0 + cpu8 2165 1115 1032 18 0 2156 9 0 + cpu9 2301 1252 1033 16 0 2293 8 0 + cpu10 2197 1138 1041 18 0 2187 10 0 + cpu11 2273 1185 1062 26 0 2260 13 0 + cpu12 2186 1125 1043 18 0 2177 9 0 + cpu13 2161 1115 1030 16 0 2153 8 0 + cpu14 2206 1153 1033 20 0 2196 10 0 + cpu15 2163 1115 1032 16 0 2155 8 0 + +In the output above, for vcpu0, there have been 6839 dispatches since +statistics were enabled. 4126 of those dispatches were on the same +physical cpu as the last time. 2683 were on a different core, but within +the same chip, while 30 dispatches were on a different chip compared to +its last dispatch. + +Also, out of the total of 6839 dispatches, we see that there have been +6821 dispatches on the vcpu's home node, while 18 dispatches were +outside its home node, on a neighbouring chip.
Signature · 1 line
2.21.0