Perf-percobaan 4.docx
Transcript of Perf-percobaan 4.docx
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 1/11
pi@raspberrypi:~/evaluasi/naive $ perf stat -e cpu-cycles,instructions,cache-references,cache-misses,branch-instructions,cpu-clock,task-clock,page-
faults,context-switches,cpu-migrations,minor-faults,major-faults,alignment-faults,emulation-faults ./loop_blockingCouldnt open result !le
"erformance counter stats for ./loop_blocking:
#$%&'() cycles * &.)(+ '&(+) instructions * &.# insns per cycle#'+$(' cache-references * %$.)+ 0/sec%%+ cache-misses * #.(1& 2 of all cache refs''$%# branches * $.&') 0/sec
+.''1$$1 cpu-clock
+.''#('( task-clock * &.'(' C"3s utilied1+ page-faults * &.&$# 0/sec# context-switches * &.)1' 4/sec& C"3-migrations * &.&&& 4/sec1+ minor-faults * &.&$# 0/sec& major-faults * &.&&& 4/sec& alignment-faults * &.&&& 4/sec& emulation-faults * &.&&& 4/sec
&.&&)#')&% seconds time elapsedpi@raspberrypi:~/evaluasi/naive $ perf record -e cpu-cycles,instructions,cache-references,cache-misses,branch-instructions,cpu-clock,task-clock,page-faults,context-switches,cpu-migrations,minor-faults,major-faults,alignment-faults,emulation-faults ./loop_blocking
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 2/11
Couldnt open result !le5 perf record: 6oken up $ times to write data 75 perf record: Captured and wrote &.&## 08 perf.data 9('( samples; 7
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 3/11
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 4/11
pi@raspberrypi:~/evaluasi/naive $ perf report --stdio --sort comm,dso* <<<<<<<<* captured on: =un >pr $& &(:):)% #&$'* hostname : raspberrypi* os release : 1.$.$-?%@* perf ?ersion : +.'.&* arch : arm?%l* nrcpus online : 1* nrcpus a?ail : 1* cpudesc : 9null;* total memory : (1&'& k8* cmdline : /usr/bin/perf record -e cpu-cycles,instructions,cache-references,cac* e?ent : name < cycles, type < &, con!g < &x&, con!g$ < &x&, con!g# < &x&, e* e?ent : name < instructions, type < &, con!g < &x$, con!g$ < &x&, con!g# <
* e?ent : name < cache-references, type < &, con!g < &x#, con!g$ < &x&,con!g* e?ent : name < cache-misses, type < &, con!g < &x+, con!g$ < &x&, con!g#<* e?ent : name < branches, type < &, con!g < &x1, con!g$ < &x&, con!g# <&x&,* e?ent : name < cpu-clock, type < $, con!g < &x&, con!g$ < &x&, con!g# <&x&* e?ent : name < task-clock, type < $, con!g < &x$, con!g$ < &x&, con!g# <&x* e?ent : name < page-faults, type < $, con!g < &x#, con!g$ < &x&, con!g# <&
* e?ent : name < context-switches, type < $, con!g < &x+, con!g$ < &x&,con!g
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 5/11
* e?ent : name < C"3-migrations, type < $, con!g < &x1, con!g$ < &x&,con!g#* e?ent : name < minor-faults, type < $, con!g < &x), con!g$ < &x&, con!g# <* e?ent : name < major-faults, type < $, con!g < &x', con!g$ < &x&, con!g# <* e?ent : name < alignment-faults, type < $, con!g < &x%, con!g$ < &x&, con!g
* e?ent : name < emulation-faults, type < $, con!g < &x, con!g$ < &x&, con!g* A>BA_C"3_DE"EFEG info a?ailable, use -H to display* <<<<<<<<** =amples: '& of e?ent cycles* A?ent count 9approx.;: 1++11%** E?erhead Command =hared Ebject* ........ ............. .................* '.$'2 loop_blocking 5kernel.kallsyms7 $&.1$2 loop_blocking ld-#.$(.so
+.1+2 loop_blocking libc-#.$(.so
* =amples: )1 of e?ent instructions* A?ent count 9approx.;: $1%1(** E?erhead Command =hared Ebject* ........ ............. .................* #.+&2 loop_blocking 5kernel.kallsyms7 $&.')2 loop_blocking ld-#.$(.so
%.&)2 loop_blocking libc-#.$(.so
* =amples: )% of e?ent cache-references* A?ent count 9approx.;: '%#')** E?erhead Command =hared Ebject* ........ ............. .................* +.'2 loop_blocking 5kernel.kallsyms7 (.'+2 loop_blocking ld-#.$(.so
'.)$2 loop_blocking libc-#.$(.so
* =amples: 1( of e?ent cache-misses* A?ent count 9approx.;: #$(1#** E?erhead Command =hared Ebject* ........ ............. .................* ($.%$2 loop_blocking 5kernel.kallsyms7 1.+2 loop_blocking libc-#.$(.so
+.($2 loop_blocking ld-#.$(.so
* =amples: )) of e?ent branches
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 6/11
* A?ent count 9approx.;: $%1$$** E?erhead Command =hared Ebject* ........ ............. .................*
%%.&$2 loop_blocking 5kernel.kallsyms7 $'.%+2 loop_blocking ld-#.$(.so'.#'2 loop_blocking libc-#.$(.so
* =amples: +1 of e?ent cpu-clock* A?ent count 9approx.;: )&&&&&** E?erhead Command =hared Ebject* ........ ............. .................* #.+)2 loop_blocking 5kernel.kallsyms7
$1.%$2 loop_blocking ld-#.$(.so#.(12 loop_blocking libc-#.$(.so
* =amples: +1 of e?ent task-clock* A?ent count 9approx.;: )&&&&&** E?erhead Command =hared Ebject* ........ ............. .................* #.+)2 loop_blocking 5kernel.kallsyms7 $1.%$2 loop_blocking ld-#.$(.so
#.(12 loop_blocking libc-#.$(.so
* =amples: $1 of e?ent page-faults* A?ent count 9approx.;: )$** E?erhead Command =hared Ebject* ........ ............. .................* '&.%2 loop_blocking ld-#.$(.so
++.++2 loop_blocking libc-#.$(.so
+.(#2 loop_blocking 5kernel.kallsyms7 $.('2 loop_blocking perf
* =amples: 1 of e?ent context-switches* A?ent count 9approx.;: 1** E?erhead Command =hared Ebject* ........ ............. .............* $&&.&&2 loop_blocking 5unknown7
* =amples: & of e?ent C"3-migrations
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 7/11
* A?ent count 9approx.;: &** E?erhead Command =hared Ebject* ........ ....... .............*
* =amples: $1 of e?ent minor-faults* A?ent count 9approx.;: )$** E?erhead Command =hared Ebject* ........ ............. .................* '&.%2 loop_blocking ld-#.$(.so
++.++2 loop_blocking libc-#.$(.so+.(#2 loop_blocking 5kernel.kallsyms7
$.('2 loop_blocking perf
* =amples: & of e?ent major-faults* A?ent count 9approx.;: &** E?erhead Command =hared Ebject* ........ ....... .............*
* =amples: & of e?ent alignment-faults* A?ent count 9approx.;: &** E?erhead Command =hared Ebject* ........ ....... .............*
* =amples: & of e?ent emulation-faults* A?ent count 9approx.;: &** E?erhead Command =hared Ebject* ........ ....... .............
*
pi@raspberrypi:~/evaluasi/naive $ perf report --stdio--dsos<loop_blocking,libc-#.$(.so* <<<<<<<<* captured on: =un >pr $& &(:):)% #&$'* hostname : raspberrypi* os release : 1.$.$-?%@* perf ?ersion : +.'.&* arch : arm?%l* nrcpus online : 1
* nrcpus a?ail : 1* cpudesc : 9null;
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 8/11
* total memory : (1&'& k8* cmdline : /usr/bin/perf record -e cpu-cycles,instructions,cache-references,cac* e?ent : name < cycles, type < &, con!g < &x&, con!g$ < &x&, con!g# < &x&, e* e?ent : name < instructions, type < &, con!g < &x$, con!g$ < &x&, con!g# <* e?ent : name < cache-references, type < &, con!g < &x#, con!g$ < &x&,
con!g* e?ent : name < cache-misses, type < &, con!g < &x+, con!g$ < &x&, con!g#<* e?ent : name < branches, type < &, con!g < &x1, con!g$ < &x&, con!g# <&x&,* e?ent : name < cpu-clock, type < $, con!g < &x&, con!g$ < &x&, con!g# <&x&* e?ent : name < task-clock, type < $, con!g < &x$, con!g$ < &x&, con!g# <&x* e?ent : name < page-faults, type < $, con!g < &x#, con!g$ < &x&, con!g# <&* e?ent : name < context-switches, type < $, con!g < &x+, con!g$ < &x&,
con!g* e?ent : name < C"3-migrations, type < $, con!g < &x1, con!g$ < &x&,con!g#* e?ent : name < minor-faults, type < $, con!g < &x), con!g$ < &x&, con!g# <* e?ent : name < major-faults, type < $, con!g < &x', con!g$ < &x&, con!g# <* e?ent : name < alignment-faults, type < $, con!g < &x%, con!g$ < &x&, con!g* e?ent : name < emulation-faults, type < $, con!g < &x, con!g$ < &x&, con!g* A>BA_C"3_DE"EFEG info a?ailable, use -H to display* <<<<<<<<** =amples: $ of e?ent cycles* A?ent count 9approx.;: $1%%%** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: # of e?ent instructions* A?ent count 9approx.;: $&+((** E?erhead Command =hared Ebject =ymbol
* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: # of e?ent cache-references* A?ent count 9approx.;: 11%)(** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 9/11
* =amples: $ of e?ent cache-misses* A?ent count 9approx.;: ('&** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............
* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: # of e?ent branches* A?ent count 9approx.;: $&($&** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: $ of e?ent cpu-clock* A?ent count 9approx.;: #)&&&&** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: $ of e?ent task-clock* A?ent count 9approx.;: #)&&&&** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ............* $&&.&&2 loop_blocking libc-#.$(.so 5.7 _dl_addr
* =amples: + of e?ent page-faults* A?ent count 9approx.;: $%** E?erhead Command =hared Ebject =ymbol
* ........ ............. ............. ....................* 1%.&'2 loop_blocking libc-#.$(.so 5.7 __H___libc_open 1%.&'2 loop_blocking libc-#.$(.so 5.7 __init_misc
).2 loop_blocking libc-#.$(.so 5.7 strchr
* =amples: 1 of e?ent context-switches* A?ent count 9approx.;: 1** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ......
* $&&.&&2 loop_blocking 5unknown7 5.7 &&&&&&&&
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 10/11
* =amples: & of e?ent C"3-migrations* A?ent count 9approx.;: &*
* E?erhead Command =hared Ebject =ymbol* ........ ....... ............. ......*
* =amples: + of e?ent minor-faults* A?ent count 9approx.;: $%** E?erhead Command =hared Ebject =ymbol* ........ ............. ............. ....................* 1%.&'2 loop_blocking libc-#.$(.so 5.7 __H___libc_open
1%.&'2 loop_blocking libc-#.$(.so 5.7 __init_misc).2 loop_blocking libc-#.$(.so 5.7 strchr
* =amples: & of e?ent major-faults* A?ent count 9approx.;: &** E?erhead Command =hared Ebject =ymbol* ........ ....... ............. ......*
* =amples: & of e?ent alignment-faults* A?ent count 9approx.;: &** E?erhead Command =hared Ebject =ymbol* ........ ....... ............. ......*
* =amples: & of e?ent emulation-faults* A?ent count 9approx.;: &*
* E?erhead Command =hared Ebject =ymbol* ........ ....... ............. ......*
** 9Ior a higher le?el o?er?iew, try: perf report --sort comm,dso;*9AJB;9AJB;
pi@raspberrypi:~/evaluasi/naive $ perf record -e cpu-clock --freK<&&&./loop_blocking
8/16/2019 Perf-percobaan 4.docx
http://slidepdf.com/reader/full/perf-percobaan-4docx 11/11
Couldnt open result !le5 perf record: 6oken up $ times to write data 75 perf record: Captured and wrote &.&&' 08 perf.data 9#'( samples; 7
pi@raspberrypi:~/evaluasi/naive $ perf e?list -I
cpu-clock: sample_freK<&&&
pi@raspberrypi:~/evaluasi/naive $ perf report --stdio --show-nr-sample--dsos<loop_blockingArror:
Dhe perf.data !le has no samplesL* dso: loop_blocking* <<<<<<<<* captured on: =un >pr $& $&:&':$$ #&$'* hostname : raspberrypi* os release : 1.$.$-?%@* perf ?ersion : +.'.&
* arch : arm?%l* nrcpus online : 1* nrcpus a?ail : 1* cpudesc : 9null;* total memory : (1&'& k8* cmdline : /usr/bin/perf record -e cpu-clock --freK<&&& ./loop_blocking* e?ent : name < cpu-clock, type < $, con!g < &x&, con!g$ < &x&, con!g# <&x&* A>BA_C"3_DE"EFEG info a?ailable, use -H to display* <<<<<<<<*