perf: Make some functions generic
cover perf: Make some functions generic
This patch just converts some private functions into global ones that can be used by other tools like the c2c tool I am trying to merge.
Don Zickus (4):
perf: Allow ability to map cpus to nodes easily
perf, kmem: Utilize the new generic cpunode_map
perf, callchain: Add generic report parse callchain callback function perf, report: Use new generic report parse callchain callback
tools/perf/builtin-kmem.c | 78 +------------------------ tools/perf/builtin-report.c | 77 +----------------------- tools/perf/util/callchain.c | 83 ++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + tools/perf/util/cpumap.c | 139 ++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/cpumap.h | 35 +++++++++++ 6 files changed, 262 insertions(+), 151 deletions(-)
Signature · 1 line
1.7.11.7
PATCH 1/4 perf: Allow ability to map cpus to nodes easily
This patch figures out the max number of cpus and nodes that are on the system and creates a map of cpu to node. This allows us to provide a cpu and quickly get the node associated with it.
It was mostly copied from builtin-kmem.c and tweaked slightly to use less memory (use possible cpus instead of max). It also calculates the max number of nodes.
V3: simplify function names
--- tools/perf/util/cpumap.c | 139 +++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/cpumap.h | 35 ++++++++++++ 2 files changed, 174 insertions(+)
tools/perf/util/cpumap.c +139 -0
@@ -317,3 +317,142 @@ int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) {return cpu_map__build_map (cpus, corep, cpu_map__get_core); } + +/* setup simple routines to easily access node numbers given a cpu number */ +static int __set_max_num (FILE *fp,int *max) +{ +int num; +char buf[256 ]; + + num =fread (&buf,1 ,sizeof (buf), fp); +if (!num) +return -1 ; + + buf[num] ='\0' ; + +/* start on the right, to find highest node num */ +while (--num) { +if ((buf[num] ==',' ) || (buf[num] =='-' )) { + num++; +break ; + } + } +if (sscanf (&buf[num],"%d" , max) <1 ) +return -1 ; + +/* convert from 0-based to 1-based */ + (*max)++; + +return 0 ; +} + +/* Determine highest possible cpu in the system for sparse allocation */ +static void set_max_cpu_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */ + max_cpu_num =4096 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/cpu/possible" ,"r" ); +if (!fp) + goto out; + + ret =__set_max_num (fp, &max_cpu_num); + +fclose (fp); + +out: +if (ret) +pr_err ("Failed to read max cpus, using default of %d\n" , + max_cpu_num); +return ; +} + +/* Determine highest possible node in the system for sparse allocation */ +static void set_max_node_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */ + max_node_num =8 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/node/possible" ,"r" ); +if (!fp) + goto out; + + ret =__set_max_num (fp, &max_node_num); + +fclose (fp); + +out: +if (ret) +pr_err ("Failed to read max nodes, using default of %d\n" , + max_node_num); +return ; +} + +static int init_cpunode_map (void ) +{ +int i; + +set_max_cpu_num (); +set_max_node_num (); + + cpunode_map =calloc (max_cpu_num,sizeof (int )); +if (!cpunode_map) { +pr_err ("%s: calloc failed\n" , __func__); + goto out; + } + +for (i =0 ; i < max_cpu_num; i++) + cpunode_map[i] =-1 ; + +return 0 ; +out: +return -1 ; +} + +#define PATH_SYS_NODE "/sys/devices/system/node" + +int cpu__setup_cpunode_map (void ) +{ +struct dirent *dent1, *dent2; +DIR *dir1, *dir2; +unsigned int cpu, mem; +char buf[PATH_MAX ]; + +/* initialize globals */ +if (init_cpunode_map ()) +return -1 ; + + dir1 =opendir (PATH_SYS_NODE ); +if (!dir1) +return 0 ; + +/* walk tree and setup map */ +while ((dent1 =readdir (dir1)) !=NULL ) { +if (dent1->d_type !=DT_DIR || +sscanf (dent1->d_name,"node%u" , &mem) <1 ) +continue ; + +snprintf (buf,PATH_MAX ,"%s/%s" ,PATH_SYS_NODE , dent1->d_name); + dir2 =opendir (buf); +if (!dir2) +continue ; +while ((dent2 =readdir (dir2)) !=NULL ) { +if (dent2->d_type !=DT_LNK || +sscanf (dent2->d_name,"cpu%u" , &cpu) <1 ) +continue ; + cpunode_map[cpu] = mem; + } +closedir (dir2); + } +closedir (dir1); +return 0 ; +} +
tools/perf/util/cpumap.h +35 -0
@@ -4,6 +4,9 @@#include <stdio.h> #include <stdbool.h> +#include "perf.h" +#include "util/debug.h" +struct cpu_map {int nr;int map[];
@@ -46,4 +49,36 @@ static inline bool cpu_map__empty(const struct cpu_map *map) return map ? map->map[0] == -1 : true; } +int max_cpu_num; +int max_node_num; +int *cpunode_map; + +int cpu__setup_cpunode_map (void ); + +static inline int cpu__max_node (void ) +{ +if (unlikely (!max_node_num)) +pr_debug ("cpu_map not initiailzed\n" ); + +return max_node_num; +} + +static inline int cpu__max_cpu (void ) +{ +if (unlikely (!max_cpu_num)) +pr_debug ("cpu_map not initiailzed\n" ); + +return max_cpu_num; +} + +static inline int cpu__get_node (int cpu) +{ +if (unlikely (cpunode_map ==NULL )) { +pr_debug ("cpu_map not initialized\n" ); +return -1 ; + } + +return cpunode_map[cpu]; +} +#endif /* __PERF_CPUMAP_H */
Signature · 1 line
1.7.11.7
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 8 lines · Don Zickus
This patch figures out the max number of cpus and nodes that are on the system and creates a map of cpu to node. This allows us to provide a cpu and quickly get the node associated with it.
It was mostly copied from
builtin-kmem.cand tweaked slightly to use less memory (use possible cpus instead of max). It also calculates the max number of nodes.V3: simplify function names
some nits below, but I think Arnaldo should
see/check the namespace ;-)
thanks,
jirka
Quoted 72 lines
Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/cpumap.c | 139 +++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/cpumap.h | 35 ++++++++++++ 2 files changed, 174 insertions(+)tools/perf/util/cpumap.c +57
-0@@ -317,3 +317,142 @@ int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) {return cpu_map__build_map (cpus, corep, cpu_map__get_core); } + +/* setup simple routines to easily access node numbers given a cpu number */ +static int __set_max_num (FILE *fp,int *max) +{ +int num; +char buf[256 ]; + + num =fread (&buf,1 ,sizeof (buf), fp); +if (!num) +return -1 ; + + buf[num] ='\0' ; + +/* start on the right, to find highest node num */ +while (--num) { +if ((buf[num] ==',' ) || (buf[num] =='-' )) { + num++; +break ; + } + } +if (sscanf (&buf[num],"%d" , max) <1 ) +return -1 ; + +/* convert from 0-based to 1-based */ + (*max)++; + +return 0 ; +} + +/* Determine highest possible cpu in the system for sparse allocation */ +static void set_max_cpu_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */ + max_cpu_num =4096 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/cpu/possible" ,"r" ); +if (!fp) + goto out; + + ret =__set_max_num (fp, &max_cpu_num); + +fclose (fp); + +out: +if (ret) +pr_err ("Failed to read max cpus, using default of %d\n" , + max_cpu_num); +return ; +} + +/* Determine highest possible node in the system for sparse allocation */ +static void set_max_node_num (void ) +{
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 386
Quoted 10 lines
tools/perf/util/cpumap.c +10
-0+FILE *fp; +int ret =-1 ; + +/* set up default */ + max_node_num =8 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/node/possible" ,"r" ); +if (!fp) + goto out;
any reason why not put fopen/fclose into __set_max_num ?
also looks like 'get_max_num' would fit better for __set_max_num ;-)
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 391
Quoted 5 lines
tools/perf/util/cpumap.c +5
-0+ + ret =__set_max_num (fp, &max_node_num); + +fclose (fp); +
^^^ whitespace
Quoted 57 lines
tools/perf/util/cpumap.c +57
-0+out: +if (ret) +pr_err ("Failed to read max nodes, using default of %d\n" , + max_node_num); +return ; +} + +static int init_cpunode_map (void ) +{ +int i; + +set_max_cpu_num (); +set_max_node_num (); + + cpunode_map =calloc (max_cpu_num,sizeof (int )); +if (!cpunode_map) { +pr_err ("%s: calloc failed\n" , __func__); + goto out; + } + +for (i =0 ; i < max_cpu_num; i++) + cpunode_map[i] =-1 ; + +return 0 ; +out: +return -1 ; +} + +#define PATH_SYS_NODE "/sys/devices/system/node" + +int cpu__setup_cpunode_map (void ) +{ +struct dirent *dent1, *dent2; +DIR *dir1, *dir2; +unsigned int cpu, mem; +char buf[PATH_MAX ]; + +/* initialize globals */ +if (init_cpunode_map ()) +return -1 ; + + dir1 =opendir (PATH_SYS_NODE ); +if (!dir1) +return 0 ; + +/* walk tree and setup map */ +while ((dent1 =readdir (dir1)) !=NULL ) { +if (dent1->d_type !=DT_DIR || +sscanf (dent1->d_name,"node%u" , &mem) <1 ) +continue ; + +snprintf (buf,PATH_MAX ,"%s/%s" ,PATH_SYS_NODE , dent1->d_name); + dir2 =opendir (buf); +if (!dir2) +continue ; +while ((dent2 =readdir (dir2)) !=NULL ) { + if (dent2->d_type != DT_LNK ||
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 458
Quoted 10 lines
tools/perf/util/cpumap.c +10
-0+ sscanf(dent2->d_name, "cpu%u", &cpu) < 1) +continue ; + cpunode_map[cpu] = mem; + } +closedir (dir2); + } +closedir (dir1); +return 0 ; +} +
^^^ white space
Quoted 18 lines
tools/perf/util/cpumap.h +3
-0@@ -4,6 +4,9 @@#include <stdio.h> #include <stdbool.h> +#include "perf.h" +#include "util/debug.h" +struct cpu_map {int nr;int map[];@@ -46,4 +49,36 @@ static inline bool cpu_map__empty(const struct cpu_map *map)return map ? map->map[0 ] ==-1 : true; }
tools/perf/util/cpumap.h static inline bool cpu_map__empty(const struct cpu_map *map) · line 61
Quoted 10 lines
tools/perf/util/cpumap.h +10
-0+int max_cpu_num; +int max_node_num; +int *cpunode_map; + +int cpu__setup_cpunode_map (void ); + +static inline int cpu__max_node (void ) +{ +if (unlikely (!max_node_num)) +pr_debug ("cpu_map not initiailzed\n" );
typo: initialized
tools/perf/util/cpumap.h static inline bool cpu_map__empty(const struct cpu_map *map) · line 69
Quoted 8 lines
tools/perf/util/cpumap.h +8
-0+ +return max_node_num; +} + +static inline int cpu__max_cpu (void ) +{ +if (unlikely (!max_cpu_num)) +pr_debug ("cpu_map not initiailzed\n" );
ditto
tools/perf/util/cpumap.h static inline bool cpu_map__empty(const struct cpu_map *map) · line 77
Quoted 8 lines
tools/perf/util/cpumap.h +8
-0+ +return max_cpu_num; +} + +static inline int cpu__get_node (int cpu) +{ +if (unlikely (cpunode_map ==NULL )) { +pr_debug ("cpu_map not initialized\n" );
ditto
Quoted 10 lines
tools/perf/util/cpumap.h +6
-0+ return -1; + } + +return cpunode_map[cpu]; +} +#endif /* __PERF_CPUMAP_H */ Signature · 1 line
1.7.11.7
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 7 lines · Jiri Olsa
Quoted 7 lines
tools/perf/util/cpumap.c +7
-0+/* Determine highest possible node in the system for sparse allocation */ +static void set_max_node_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */
Quoted 8 lines
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 386
Quoted 6 lines
tools/perf/util/cpumap.c +6
-0+ max_node_num =8 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/node/possible" ,"r" ); +if (!fp) + goto out;any reason why not put fopen/fclose into
__set_max_num?
I either had to pass in a path string pointer or a file pointer, I just chose a file pointer. Thought it fit with the rest of perf passing FPs around. :-D It doesn't matter to me either way.
Quoted 2 lines
also looks like '
get_max_num' would fit better for__set_max_num;-)
Yeah, I split it out based on your suggestion and didn't realize the details kinda changed.
Cheers,
Don
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 17 lines · Don Zickus
Quoted 7 lines · Jiri Olsa
Quoted 7 lines
tools/perf/util/cpumap.c +7
-0+/* Determine highest possible node in the system for sparse allocation */ +static void set_max_node_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */ Quoted 8 lines
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 386
Quoted 6 lines
tools/perf/util/cpumap.c +6
-0+ max_node_num =8 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/node/possible" ,"r" ); +if (!fp) + goto out;any reason why not put fopen/fclose into
__set_max_num?
Quoted 3 lines
I either had to pass in a path string pointer or a file pointer, I just chose a file pointer. Thought it fit with the rest of perf passing FPs around. :-D It doesn't matter to me either way.
well, if there's no need to call it extra then the function can do it
jirka
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
In reply to Don ZickusHi Don,
(Adding Boris to CC as he might be interested)
Quoted 53 lines · Don Zickus
This patch figures out the max number of cpus and nodes that are on the system and creates a map of cpu to node. This allows us to provide a cpu and quickly get the node associated with it.
It was mostly copied from
builtin-kmem.cand tweaked slightly to use less memory (use possible cpus instead of max). It also calculates the max number of nodes.V3: simplify function names
Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/cpumap.c | 139 +++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/cpumap.h | 35 ++++++++++++ 2 files changed, 174 insertions(+)tools/perf/util/cpumap.c +30
-0@@ -317,3 +317,142 @@ int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) {return cpu_map__build_map (cpus, corep, cpu_map__get_core); } + +/* setup simple routines to easily access node numbers given a cpu number */ +static int __set_max_num (FILE *fp,int *max) +{ +int num; +char buf[256 ]; + + num =fread (&buf,1 ,sizeof (buf), fp); +if (!num) +return -1 ; + + buf[num] ='\0' ; + +/* start on the right, to find highest node num */ +while (--num) { +if ((buf[num] ==',' ) || (buf[num] =='-' )) { + num++; +break ; + } + } +if (sscanf (&buf[num],"%d" , max) <1 ) +return -1 ; + +/* convert from 0-based to 1-based */ + (*max)++; + +return 0 ; +} + +/* Determine highest possible cpu in the system for sparse allocation */
tools/perf/util/cpumap.c int cpu_map__build_core_map(struct cpu_map *cpus, struct cpu_map **corep) · line 359
Quoted 10 lines
tools/perf/util/cpumap.c +10
-0+static void set_max_cpu_num (void ) +{ +FILE *fp; +int ret =-1 ; + +/* set up default */ + max_cpu_num =4096 ; + +/* get the highest possible cpu number for a sparse allocation */ + fp =fopen ("/sys/devices/system/cpu/possible" ,"r" );
More generally, this sysfs access needs to check actual mountpoint using sysfs__mountpoint() IMHO.
Also this API can be generalized like reading int value from a sysfs file as the filename itself represents the content in most cases.
So how about changing this way? It might reside on somewhere in tools/lib/api/fs/.
max_cpu_num = sysfs__read_int("devices/system/cpu/possible");
max_node_num = sysfs__read_int("devices/system/node/possible");Hmm.. looking at the code, perf already has filename__read_{int,str} API in util/util.c. Maybe you can just use it instead.
Thanks,
Namhyung
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 1 line · Namhyung Kim
Hi Don,
SNIP
Quoted 9 lines
More generally, this sysfs access needs to check actual mountpoint using
sysfs__mountpoint()IMHO.Also this API can be generalized like reading int value from a sysfs file as the filename itself represents the content in most cases.
So how about changing this way? It might reside on somewhere in
tools/lib/api/fs/.
Quoted 7 lines
max_cpu_num = sysfs__read_int("devices/system/cpu/possible"); max_node_num = sysfs__read_int("devices/system/node/possible");Hmm.. looking at the code, perf already has filename__read_{int,str} API in
util/util.c. Maybe you can just use it instead.
I think those just read the file, while Don needs to parse the map to get max cpu number
jirka
Namhyung Kim
maintainerRe: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 11 lines · Jiri Olsa
Quoted 10 lines · Namhyung Kim
More generally, this sysfs access needs to check actual mountpoint using
sysfs__mountpoint()IMHO.Also this API can be generalized like reading int value from a sysfs file as the filename itself represents the content in most cases.
So how about changing this way? It might reside on somewhere in
tools/lib/api/fs/.max_cpu_num = sysfs__read_int("devices/system/cpu/possible");
Quoted 8 lines
Quoted 5 lines
max_node_num = sysfs__read_int("devices/system/node/possible");Hmm.. looking at the code, perf already has filename__read_{int,str} API in
util/util.c. Maybe you can just use it instead.I think those just read the file, while Don needs to parse the map to get max cpu number
Ah, missed that... sorry for noise.
Thanks,
Namhyung
Re: [PATCH 1/4] perf: Allow ability to map cpus to nodes easily
Quoted 15 lines · Namhyung Kim
Quoted 14 lines · Jiri Olsa
Quoted 13 lines · Namhyung Kim
More generally, this sysfs access needs to check actual mountpoint using
sysfs__mountpoint()IMHO.Also this API can be generalized like reading int value from a sysfs file as the filename itself represents the content in most cases.
So how about changing this way? It might reside on somewhere in
tools/lib/api/fs/.max_cpu_num = sysfs__read_int("devices/system/cpu/possible"); max_node_num = sysfs__read_int("devices/system/node/possible");
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
Hmm.. looking at the code, perf already has filename__read_{int,str} API in
util/util.c. Maybe you can just use it instead.I think those just read the file, while Don needs to parse the map to get max cpu number
Ah, missed that... sorry for noise.
Well, I already implemented using filename__read_str. It is just added overhead, but the end result is the same, a string needs to be read from the file. I just had to remember to free the memory at the end of the function.
Cheers,
Don
PATCH 2/4 perf, kmem: Utilize the new generic cpunode_map
In reply to Don ZickusUse the previous patch implementation of cpunode_map for builtin-kmem.c Should not be any functional difference.
--- tools/perf/builtin-kmem.c | 78 ++--------------------------------------------- 1 file changed, 3 insertions(+), 75 deletions(-)
tools/perf/builtin-kmem.c +3 -75
@@ -14,6 +14,7 @@#include "util/parse-options.h" #include "util/trace-event.h" #include "util/data.h" +#include "util/cpumap.h" #include "util/debug.h"
@@ -31,9 +32,6 @@ static int caller_lines = -1;static bool raw_ip; -static int *cpunode_map; -static int max_cpu_num; -struct alloc_stat {u64 call_site;u64 ptr;
@@ -55,76 +53,6 @@ static struct rb_root root_caller_sorted;static unsigned long total_requested, total_allocated;static unsigned long nr_allocs, nr_cross_allocs; -#define PATH_SYS_NODE "/sys/devices/system/node" - -static int init_cpunode_map (void ) -{ -FILE *fp; -int i, err =-1 ; - - fp =fopen ("/sys/devices/system/cpu/kernel_max" ,"r" ); -if (!fp) { - max_cpu_num =4096 ; -return 0 ; - } - -if (fscanf (fp,"%d" , &max_cpu_num) <1 ) { -pr_err ("Failed to read 'kernel_max' from sysfs" ); - goto out_close; - } - - max_cpu_num++; - - cpunode_map =calloc (max_cpu_num,sizeof (int )); -if (!cpunode_map) { -pr_err ("%s: calloc failed\n" , __func__); - goto out_close; - } - -for (i =0 ; i < max_cpu_num; i++) - cpunode_map[i] =-1 ; - - err =0 ; -out_close: -fclose (fp); -return err; -} - -static int setup_cpunode_map (void ) -{ -struct dirent *dent1, *dent2; -DIR *dir1, *dir2; -unsigned int cpu, mem; -char buf[PATH_MAX ]; - -if (init_cpunode_map ()) -return -1 ; - - dir1 =opendir (PATH_SYS_NODE ); -if (!dir1) -return 0 ; - -while ((dent1 =readdir (dir1)) !=NULL ) { -if (dent1->d_type !=DT_DIR || -sscanf (dent1->d_name,"node%u" , &mem) <1 ) -continue ; - -snprintf (buf,PATH_MAX ,"%s/%s" ,PATH_SYS_NODE , dent1->d_name); - dir2 =opendir (buf); -if (!dir2) -continue ; -while ((dent2 =readdir (dir2)) !=NULL ) { -if (dent2->d_type !=DT_LNK || -sscanf (dent2->d_name,"cpu%u" , &cpu) <1 ) -continue ; - cpunode_map[cpu] = mem; - } -closedir (dir2); - } -closedir (dir1); -return 0 ; -} -static int insert_alloc_stat (unsigned long call_site,unsigned long ptr,int bytes_req,int bytes_alloc,int cpu) {
@@ -235,7 +163,7 @@ static int perf_evsel__process_alloc_node_event(struct perf_evsel *evsel,int ret =perf_evsel__process_alloc_event (evsel, sample);if (!ret) { -int node1 = cpunode_map[sample->cpu], +int node1 =cpu__get_node (sample->cpu), node2 =perf_evsel__intval (evsel, sample,"node" );if (node1 != node2)
@@ -770,7 +698,7 @@ int cmd_kmem(int argc, const char **argv, const char *prefix __maybe_unused)if (!strncmp (argv[0 ],"rec" ,3 )) {return __cmd_record (argc, argv); }else if (!strcmp (argv[0 ],"stat" )) { -if (setup_cpunode_map ()) +if (cpu__setup_cpunode_map ())return -1 ;if (list_empty (&caller_sort))
Signature · 1 line
1.7.11.7
Re: [PATCH 2/4] perf, kmem: Utilize the new generic cpunode_map
Quoted 33 lines · Don Zickus
Use the previous patch implementation of
cpunode_mapforbuiltin-kmem.cShould not be any functional difference.Cc Li Zefan <[email protected]>Signed-off-by Don Zickus <[email protected]>--- tools/perf/builtin-kmem.c | 78 ++--------------------------------------------- 1 file changed, 3 insertions(+), 75 deletions(-)tools/perf/builtin-kmem.c +1
-3@@ -14,6 +14,7 @@#include "util/parse-options.h" #include "util/trace-event.h" #include "util/data.h" +#include "util/cpumap.h" #include "util/debug.h" @@ -31,9 +32,6 @@ static int caller_lines = -1;static bool raw_ip; -static int *cpunode_map; -static int max_cpu_num; -struct alloc_stat {u64 call_site;u64 ptr;@@ -55,76 +53,6 @@ static struct rb_root root_caller_sorted;static unsigned long total_requested, total_allocated;
tools/perf/builtin-kmem.c static struct rb_root root_caller_sorted; · line 65
Quoted 10 lines
tools/perf/builtin-kmem.c +0
-8static unsigned long nr_allocs, nr_cross_allocs; -#define PATH_SYS_NODE "/sys/devices/system/node" - -static int init_cpunode_map (void ) -{ -FILE *fp; -int i, err =-1 ; - - fp =fopen ("/sys/devices/system/cpu/kernel_max" ,"r" );
so the factored code from previous patches now reads the max_cpu_num value from:
/sys/devices/system/cpu/possible
is this intentional?
I think we want to have separate patches for code changes and for changing the file with some comment.
jirka
Re: [PATCH 2/4] perf, kmem: Utilize the new generic cpunode_map
Quoted 49 lines · Jiri Olsa
Quoted 33 lines · Don Zickus
Use the previous patch implementation of
cpunode_mapforbuiltin-kmem.cShould not be any functional difference.Cc Li Zefan <[email protected]>Signed-off-by Don Zickus <[email protected]>--- tools/perf/builtin-kmem.c | 78 ++--------------------------------------------- 1 file changed, 3 insertions(+), 75 deletions(-)tools/perf/builtin-kmem.c +1
-3@@ -14,6 +14,7 @@#include "util/parse-options.h" #include "util/trace-event.h" #include "util/data.h" +#include "util/cpumap.h" #include "util/debug.h" @@ -31,9 +32,6 @@ static int caller_lines = -1;static bool raw_ip; -static int *cpunode_map; -static int max_cpu_num; -struct alloc_stat {u64 call_site;u64 ptr;@@ -55,76 +53,6 @@ static struct rb_root root_caller_sorted;static unsigned long total_requested, total_allocated;tools/perf/builtin-kmem.c static struct rb_root root_caller_sorted; · line 65
Quoted 10 lines
tools/perf/builtin-kmem.c +0
-8static unsigned long nr_allocs, nr_cross_allocs; -#define PATH_SYS_NODE "/sys/devices/system/node" - -static int init_cpunode_map (void ) -{ -FILE *fp; -int i, err =-1 ; - - fp =fopen ("/sys/devices/system/cpu/kernel_max" ,"r" );so the factored code from previous patches now reads the
max_cpu_numvalue from:
/sys/devices/system/cpu/possible
Quoted 1 line
is this intentional?
Yeah, I was trying to save bits. No need to allocate for 5100 cpus when only 128 are used.
Quoted 3 lines
I think we want to have separate patches for code changes and for changing the file with some comment.
So you want me to split the previous patch into two. One for code movement, the other for the path change?
Cheers,
Don
Quoted 2 lines
jirka
Re: [PATCH 2/4] perf, kmem: Utilize the new generic cpunode_map
On Mon, Mar 31, 2014 at 10:42:30PM -0400, Don Zickus wrote:
SNIP
Quoted 19 lines
Quoted 13 lines
tools/perf/builtin-kmem.c static struct rb_root root_caller_sorted; · line 65
Quoted 8 lines
tools/perf/builtin-kmem.c +0
-8-#define PATH_SYS_NODE "/sys/devices/system/node" - -static int init_cpunode_map (void ) -{ -FILE *fp; -int i, err =-1 ; - - fp =fopen ("/sys/devices/system/cpu/kernel_max" ,"r" );so the factored code from previous patches now reads the
max_cpu_numvalue from:
/sys/devices/system/cpu/possibleQuoted 1 line
is this intentional?
Yeah, I was trying to save bits. No need to allocate for 5100 cpus when only 128 are used.
Quoted 1 line
Quoted 5 lines
Quoted 2 lines
I think we want to have separate patches for code changes and for changing the file with some comment.
So you want me to split the previous patch into two. One for code movement, the other for the path change?
I think thats the proper way.. single logic change for patch
thanks,
jirka
Re: [PATCH 2/4] perf, kmem: Utilize the new generic cpunode_map
Quoted 23 lines · Jiri Olsa
On Mon, Mar 31, 2014 at 10:42:30PM -0400, Don Zickus wrote:
SNIP
Quoted 19 lines
Quoted 13 lines
tools/perf/builtin-kmem.c static struct rb_root root_caller_sorted; · line 65
Quoted 8 lines
tools/perf/builtin-kmem.c +0
-8-#define PATH_SYS_NODE "/sys/devices/system/node" - -static int init_cpunode_map (void ) -{ -FILE *fp; -int i, err =-1 ; - - fp =fopen ("/sys/devices/system/cpu/kernel_max" ,"r" );so the factored code from previous patches now reads the
max_cpu_numvalue from:
/sys/devices/system/cpu/possibleQuoted 1 line
is this intentional?
Yeah, I was trying to save bits. No need to allocate for 5100 cpus when only 128 are used.
Quoted 1 line
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
I think we want to have separate patches for code changes and for changing the file with some comment.
So you want me to split the previous patch into two. One for code movement, the other for the path change?
I think thats the proper way.. single logic change for patch
Hmm, the whole change is getting completely messy with the breakdown of common functions, renames, use of internal functions, etc... I'll just break out the small change (kernel_max -> possible) from that mess.
Cheers,
Don
PATCH 3/4 perf, callchain: Add generic report parse callchain callback function
In reply to Don ZickusThis takes the parse_callchain_opt function and copies it into the callchain.c file. Now the c2c tool can use it too without duplicating.
--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)
tools/perf/util/callchain.c +83 -0
@@ -25,6 +25,89 @@ __threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg) +{ +//struct report *rep = (struct report *)opt->value; +char *tok, *tok2; +char *endptr; + + symbol_conf.use_callchain = true; + +if (!arg) +return 0 ; + + tok =strtok ((char *)arg,"," ); +if (!tok) +return -1 ; + +/* get the output mode */ +if (!strncmp (tok,"graph" ,strlen (arg))) + callchain_param.mode =CHAIN_GRAPH_ABS ; + +else if (!strncmp (tok,"flat" ,strlen (arg))) + callchain_param.mode =CHAIN_FLAT ; + +else if (!strncmp (tok,"fractal" ,strlen (arg))) + callchain_param.mode =CHAIN_GRAPH_REL ; + +else if (!strncmp (tok,"none" ,strlen (arg))) { + callchain_param.mode =CHAIN_NONE ; + symbol_conf.use_callchain = false; + +return 0 ; + } + +else +return -1 ; + +/* get the min percentage */ + tok =strtok (NULL ,"," ); +if (!tok) + goto setup; + + callchain_param.min_percent =strtod (tok, &endptr); +if (tok == endptr) +return -1 ; + +/* get the print limit */ + tok2 =strtok (NULL ,"," ); +if (!tok2) + goto setup; + +if (tok2[0 ] !='c' ) { + callchain_param.print_limit =strtoul (tok2, &endptr,0 ); + tok2 =strtok (NULL ,"," ); +if (!tok2) + goto setup; + } + +/* get the call chain order */ +if (!strncmp (tok2,"caller" ,strlen ("caller" ))) + callchain_param.order =ORDER_CALLER ; +else if (!strncmp (tok2,"callee" ,strlen ("callee" ))) + callchain_param.order =ORDER_CALLEE ; +else +return -1 ; + +/* Get the sort key */ + tok2 =strtok (NULL ,"," ); +if (!tok2) + goto setup; +if (!strncmp (tok2,"function" ,strlen ("function" ))) + callchain_param.key =CCKEY_FUNCTION ; +else if (!strncmp (tok2,"address" ,strlen ("address" ))) + callchain_param.key =CCKEY_ADDRESS ; +else +return -1 ; +setup: +if (callchain_register_param (&callchain_param) <0 ) { +pr_err ("Can't register callchain params\n" ); +return -1 ; + } +return 0 ; +} +static void rb_insert_callchain (struct rb_root *root,struct callchain_node *chain,enum chain_mode mode)
tools/perf/util/callchain.h +1 -0
@@ -157,4 +157,5 @@ int sample__resolve_callchain(struct perf_sample *sample, struct symbol **parentint hist_entry__append_callchain (struct hist_entry *he,struct perf_sample *sample);extern const char record_callchain_help[]; +int report_parse_callchain_opt (const char *arg);#endif /* __PERF_CALLCHAIN_H */
Signature · 1 line
1.7.11.7
Re: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
Quoted 14 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0
tools/perf/util/callchain.c line 31
Quoted 7 lines
tools/perf/util/callchain.c +4
-0@@ -25,6 +25,89 @@ __threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg) +{ +//struct report *rep = (struct report *)opt->value;
please erase ^^^
and I think it's better to merge this with the
actual removal in following patch:
perf, report: Use new generic report parse callchain callback
jirka
Namhyung Kim
maintainerRe: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
In reply to Don ZickusQuoted 14 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0
tools/perf/util/callchain.c line 29
Quoted 5 lines
tools/perf/util/callchain.c +2
-0@@ -25,6 +25,89 @@ __threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg)
As it eliminated any dependency to the report code, the name can omit 'report' as well.
Thanks,
Namhyung
Re: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
Quoted 16 lines · Namhyung Kim
Quoted 15 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0@@ -25,6 +25,89 @@
Quoted 7 lines
tools/perf/util/callchain.c line 29
Quoted 4 lines
tools/perf/util/callchain.c +2
-0__threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg)As it eliminated any dependency to the report code, the name can omit 'report' as well.
Hmm, shrinking this down to 'parse_callchain_opt' seems to conflict with builtin-top.c's defintion which is a wrapper around record_parse_callchain_opt.
I can modify builtin-top.c's defintion to top_parse_callchain_opt, but do we want to make the function parse_callchain_opt name more obvious that it is for 'report' style parsing (which means it looks for the embedded callchain flag)? Vs. the record_parse_callchain_opt function, which is for setting up the callchain flag?
Cheers,
Don
Re: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
Quoted 29 lines · Don Zickus
Quoted 16 lines · Namhyung Kim
Quoted 15 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0@@ -25,6 +25,89 @@Quoted 7 lines
tools/perf/util/callchain.c line 29
Quoted 4 lines
tools/perf/util/callchain.c +2
-0__threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg)As it eliminated any dependency to the report code, the name can omit 'report' as well.
Hmm, shrinking this down to '
parse_callchain_opt' seems to conflict with builtin-top.c's defintion which is a wrapper aroundrecord_parse_callchain_opt.
Quoted 5 lines
I can modify builtin-top.c's defintion to
top_parse_callchain_opt, but do we want to make the functionparse_callchain_optname more obvious that it is for 'report' style parsing (which means it looks for the embedded callchain flag)? Vs. therecord_parse_callchain_optfunction, which is for setting up the callchain flag?
that sounds good to me.. (record|report)_parse_callchain_opt
jirka
Namhyung Kim
maintainerRe: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
Hi Jiri and Don,
Quoted 30 lines · Jiri Olsa
Quoted 29 lines · Don Zickus
Quoted 16 lines · Namhyung Kim
Quoted 15 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0@@ -25,6 +25,89 @@Quoted 7 lines
tools/perf/util/callchain.c line 29
Quoted 4 lines
tools/perf/util/callchain.c +2
-0__threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg)As it eliminated any dependency to the report code, the name can omit 'report' as well.
Hmm, shrinking this down to '
parse_callchain_opt' seems to conflict with builtin-top.c's defintion which is a wrapper aroundrecord_parse_callchain_opt.
Quoted 7 lines
Quoted 5 lines
I can modify builtin-top.c's defintion to
top_parse_callchain_opt, but do we want to make the functionparse_callchain_optname more obvious that it is for 'report' style parsing (which means it looks for the embedded callchain flag)? Vs. therecord_parse_callchain_optfunction, which is for setting up the callchain flag?that sounds good to me.. (record|report)_parse_callchain_opt
Agreed. Maybe parse_callchain_(record|report)_opt?
Thanks,
Namhyung
Re: [PATCH 3/4] perf, callchain: Add generic report parse callchain callback function
Quoted 39 lines · Namhyung Kim
Hi Jiri and Don,
Quoted 36 lines · Jiri Olsa
Quoted 34 lines · Don Zickus
Quoted 16 lines · Namhyung Kim
Quoted 15 lines · Don Zickus
This takes the
parse_callchain_optfunction and copies it into thecallchain.cfile. Now the c2c tool can use it too without duplicating.Signed-off-by Don Zickus <[email protected]>--- tools/perf/util/callchain.c | 83 +++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/callchain.h | 1 + 2 files changed, 84 insertions(+)tools/perf/util/callchain.c +0
-0@@ -25,6 +25,89 @@Quoted 7 lines
tools/perf/util/callchain.c line 29
Quoted 4 lines
tools/perf/util/callchain.c +2
-0__threadstruct callchain_cursor callchain_cursor; +int +report_parse_callchain_opt (const char *arg)As it eliminated any dependency to the report code, the name can omit 'report' as well.
Hmm, shrinking this down to '
parse_callchain_opt' seems to conflict with builtin-top.c's defintion which is a wrapper aroundrecord_parse_callchain_opt.I can modify builtin-top.c's defintion to
top_parse_callchain_opt, but do we want to make the functionparse_callchain_optname more obvious that it is for 'report' style parsing (which means it looks for the embedded callchain flag)? Vs. therecord_parse_callchain_optfunction, which is for setting up the callchain flag?
Quoted 3 lines
Quoted 1 line
that sounds good to me.. (record|report)_parse_callchain_opt
Agreed. Maybe parse_callchain_(record|report)_opt?
ook, jirka
PATCH 4/4 perf, report: Use new generic report parse callchain callback
In reply to Don ZickusUse the new routine.
--- tools/perf/builtin-report.c | 77 +-------------------------------------------- 1 file changed, 1 insertion(+), 76 deletions(-)
tools/perf/builtin-report.c +1 -76
@@ -576,8 +576,6 @@ static int parse_callchain_opt(const struct option *opt, const char *arg, int unset) { struct report *rep = (struct report *)opt->value; -char *tok, *tok2; -char *endptr;/* * --no-call-graph
@@ -587,80 +585,7 @@ parse_callchain_opt(const struct option *opt, const char *arg, int unset)return 0 ; } - symbol_conf.use_callchain = true; - -if (!arg) -return 0 ; - - tok =strtok ((char *)arg,"," ); -if (!tok) -return -1 ; - -/* get the output mode */ -if (!strncmp (tok,"graph" ,strlen (arg))) - callchain_param.mode =CHAIN_GRAPH_ABS ; - -else if (!strncmp (tok,"flat" ,strlen (arg))) - callchain_param.mode =CHAIN_FLAT ; - -else if (!strncmp (tok,"fractal" ,strlen (arg))) - callchain_param.mode =CHAIN_GRAPH_REL ; - -else if (!strncmp (tok,"none" ,strlen (arg))) { - callchain_param.mode =CHAIN_NONE ; - symbol_conf.use_callchain = false; - -return 0 ; - } - -else -return -1 ; - -/* get the min percentage */ - tok =strtok (NULL ,"," ); -if (!tok) - goto setup; - - callchain_param.min_percent =strtod (tok, &endptr); -if (tok == endptr) -return -1 ; - -/* get the print limit */ - tok2 =strtok (NULL ,"," ); -if (!tok2) - goto setup; - -if (tok2[0 ] !='c' ) { - callchain_param.print_limit =strtoul (tok2, &endptr,0 ); - tok2 =strtok (NULL ,"," ); -if (!tok2) - goto setup; - } - -/* get the call chain order */ -if (!strncmp (tok2,"caller" ,strlen ("caller" ))) - callchain_param.order =ORDER_CALLER ; -else if (!strncmp (tok2,"callee" ,strlen ("callee" ))) - callchain_param.order =ORDER_CALLEE ; -else -return -1 ; - -/* Get the sort key */ - tok2 =strtok (NULL ,"," ); -if (!tok2) - goto setup; -if (!strncmp (tok2,"function" ,strlen ("function" ))) - callchain_param.key =CCKEY_FUNCTION ; -else if (!strncmp (tok2,"address" ,strlen ("address" ))) - callchain_param.key =CCKEY_ADDRESS ; -else -return -1 ; -setup: -if (callchain_register_param (&callchain_param) <0 ) { -pr_err ("Can't register callchain params\n" ); -return -1 ; - } -return 0 ; +return report_parse_callchain_opt (arg); }int
Signature · 1 line
1.7.11.7