fche changed the topic of #systemtap to: http://sourceware.org/systemtap; email systemtap@sourceware.org if answers here not timely, conversations may be logged
orivej has quit [Ping timeout: 248 seconds]
orivej has joined #systemtap
hpt has joined #systemtap
orivej has quit [Ping timeout: 240 seconds]
scox has quit [Ping timeout: 240 seconds]
scox has joined #systemtap
nkambo has joined #systemtap
slowfranklin has joined #systemtap
lightydo has quit [Remote host closed the connection]
lightydo has joined #systemtap
hpt has quit [Quit: Lost terminal]
scox has quit [Ping timeout: 248 seconds]
gromero has joined #systemtap
orivej has joined #systemtap
gila has joined #systemtap
scox has joined #systemtap
gila has quit [Read error: Connection reset by peer]
drsmith_away is now known as drsmith
mbenitez has joined #systemtap
mbenitez has joined #systemtap
mbenitez has quit [Changing host]
mjw has joined #systemtap
orivej has quit [Ping timeout: 256 seconds]
brolley has joined #systemtap
<lightydo> Hi, can someone provide an example of counting all sw page faults per process?
<lightydo> I am trying to see if I can replicate the work done by capsule8 (meltdown detection tool - https://github.com/capsule8/capsule8/blob/master/examples/meltdown/main.go) via systemtap.
<fche> lightydo, hi
<fche> https://sourceware.org/systemtap/examples/#memory/pfaults.stp <- this one's a very basic log
<fche> https://sourceware.org/systemtap/examples/#memory/numa_faults.stp <-- this one's more process-aggregated
<lightydo> Hi fche, I thought the approach would be to use something like : probe perf.sw.page_faults.counter("a") { }
<fche> that sort of thing would probably work too, so basically don't trap them all just try to collect counts
<lightydo> yes, the question is where to aggregate the counter... page_faults_cnt[pid()] <<< @perf("a")
<fche> those @perf("...") type count values are only accessible in userspace probes (due to limitations in older kernels; current kernel should let us remove that restrictions, but that's some work)
<fche> so basically probe process("...").function("main").return { } kind of thing
<fche> -or-
<fche> use systemwide (normal, .counter-less) perf probes, and then count counter hits against the then-current process
<fche> basically profiling
<fche> so then
<fche> # stap /usr/share/systemtap/examples/general/eventcount.stp 'perf.sw.page_faults'
<lightydo> thanks, i'll look into it.
slowfranklin has quit [Quit: slowfranklin]
mbenitez has quit [Quit: To lunch / office]
orivej has joined #systemtap
slowfranklin has joined #systemtap
mbenitez has joined #systemtap
mbenitez has joined #systemtap
mbenitez has quit [Changing host]
slowfranklin has quit [Quit: slowfranklin]
gromero has quit [Quit: Leaving]
mjw has quit [Quit: Leaving]
drsmith is now known as drsmith_away
drsmith_away is now known as drsmith
mbenitez has quit [Quit: Leaving]
scox has quit [Ping timeout: 256 seconds]
higgins has quit [Quit: Leaving]
higgins has joined #systemtap
brolley has left #systemtap [#systemtap]
drsmith is now known as drsmith_away