fche changed the topic of #systemtap to: http://sourceware.org/systemtap; email systemtap@sourceware.org if answers here not timely, conversations may be logged
khaled has quit [Quit: Konversation terminated!]
orivej has quit [Ping timeout: 258 seconds]
orivej has joined #systemtap
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #systemtap
hpt has joined #systemtap
orivej has quit [Ping timeout: 256 seconds]
orivej_ has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
wcohen has quit [Remote host closed the connection]
orivej_ has quit [Ping timeout: 246 seconds]
orivej has joined #systemtap
derek0883 has joined #systemtap
wcohen has joined #systemtap
orivej has quit [Ping timeout: 272 seconds]
orivej_ has joined #systemtap
orivej_ has quit [Ping timeout: 264 seconds]
orivej has joined #systemtap
orivej has quit [Ping timeout: 272 seconds]
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek088_ has joined #systemtap
derek0883 has quit [Ping timeout: 246 seconds]
derek088_ has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 260 seconds]
khaled has joined #systemtap
ema has quit [Remote host closed the connection]
ema has joined #systemtap
orivej has joined #systemtap
<linus2> fche: why do the ppfunc() is null in mark probe?
<fche> linus2, try pp()
<linus2> fche: pp is too long
<linus2> May pp() increace performance problem and caused the result is not precise?
tonyj has quit [Ping timeout: 246 seconds]
<fche> try probefunc()
<fche> (they each have man pages that explain the difference: man function::probefunc
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #systemtap
tonyj has joined #systemtap
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #systemtap
orivej_ has joined #systemtap
orivej has quit [Ping timeout: 272 seconds]
hpt has quit [Ping timeout: 256 seconds]
_whitelogger has joined #systemtap
<linus2> fche: str1 = "abc" str2 = "efg" how to connect the two string(str3 = "abcefg")?
hpt has joined #systemtap
orivej_ has quit [Ping timeout: 256 seconds]
hpt has quit [Ping timeout: 240 seconds]
<fche> str1 . str2
<linus2> fche: how to strobe openmp code?
<linus2> I found that the openmp code can not be probe
<linus2> statement("*@xx.c:6") can't be found
<fche> is this with CFLAGS=-g ?
<linus2> yes
<linus2> do you know openmp parallel code?
<fche> not really
<linus2> wait a minite, I send you a openmp example code
<linus2> fche: I spent too many time learning stp to probe openmp code, but now I found that it doesn't work. crying...
<fche> oh dear :)
<fche> have you tried running it under gdb ? which compiler version are you using?
<linus2> there is example code. You can copy it and compile it using -fopenmp. Then stp script print the line number of "printf" in C code.
<linus2> fche: g++ (GCC) 7.3.0
<linus2> I am running the binary compiled by g++ not running in gdb
<linus2> fche: it seems the code surrounded by the openmp parallel will translate into other form first.
<linus2> fche: what's going on?
orivej has joined #systemtap
tromey has joined #systemtap
amerey has joined #systemtap
orivej has quit [Ping timeout: 256 seconds]
amerey has quit [Remote host closed the connection]
amerey has joined #systemtap
sapatel has joined #systemtap
derek0883 has joined #systemtap
_whitelogger has joined #systemtap
wcohen has quit [Quit: Leaving]
orivej has joined #systemtap
sapatel has quit [Read error: error:1408F10B:SSL routines:ssl3_get_record:wrong version number]
sapatel has joined #systemtap
DTEIT has joined #systemtap
sapatel has quit [Read error: error:1408F10B:SSL routines:ssl3_get_record:wrong version number]
sapatel has joined #systemtap
nrb has joined #systemtap
<nrb> hi guys, i have a hard time installing systemtap
<nrb> it keeps saying kbuild exited with status 2
<nrb> i am using a custom kernel, i know what i have to apply in .config, i know that i have to do make headers_install and modules_install, also i copied vmlinux over to /lib/modules/ver/
<fche> nrb tried stap -r /path/to/build/tree ?
<nrb> which tree?
<nrb> i have downloaded it from apt
<fche> The kernel build tree
<fche> i.e., whereever you'd say 'make headers_install'
<nrb> just tried, tells that a script has to be specified
<fche> stap does? yes
<nrb> want me to try script + this what you said?
<fche> stap -r /path/to... FOO.stp etc.
<fche> yes
<nrb> just tried, doesn't work
<fche> ok, what are you seeing exactly
<nrb> Pass 1: parsed user script and 465 library scripts using 114972virt/48908res/5972shr/43284data.
<nrb> WARNING: kbuild exited with status: 2
<fche> (fpaste or whatever for completeness if easy)
<nrb> WARNING: kbuild exited with status: 2
<fche> is this a pure normal native build
<nrb> no, custom kernel
<fche> is the kernel version in range of support of systemtap?
<nrb> i use 5.7.7
<fche> custom - but I mean native vs. cross ?
<nrb> native
<fche> OK, try stap -r /etc.... -v -v ..... <<< i.e., add some -v's
<fche> the more the merrier
<nrb> hmm i don't understand
<nrb> i do stap -r pathtolinux -v mystp
<fche> run stap however you ran it, adding a few -v -v options
<fche> for more verbosity
<nrb> it's running through addresses of the kernel now
<nrb> very fast
<fche> it is chatty with more -v's
<nrb> address 't'/'r' somefunction
<nrb> it doesn't stop
<fche> it stops eventually :)
<nrb> cc1: fatal error: /tmp/staphv6ihO/stapconf_8c3ba6acfe358e981da1dd557ae4ee1b_613.h: No such fily
<nrb> compilation terminated.
<fche> ok, so I'd look back a few steps to see what it was doing
<fche> fpaste those
<nrb> fpaste is what?
<nrb> filepaste?
<nrb> this was running a lot of lines, immense output
<fche> fpaste is a fedora command line tool to upload blogs of text to an online service
<fche> yeah the last few hundredish should tell the tale
<nrb> i don't run fedora
<nrb> i can rerun this, redirect the output to a file and then send it
<nrb> but this will be huge
<fche> 2>&1 | tail -200
<nrb> k doing it
<nrb> where do you want it uploaded
<fche> paste.centos.org is an easy one
<fche> if your distro has a favorite different one, whatever you like
<fche> ok, did I ask stap -V version yet?
<nrb> Systemtap translator/driver (version 3.1/0.168, Debian version 3.1-2)
<nrb> Copyright (C) 2005-2017 Red Hat, Inc. and others
<nrb> tested kernel versions: 2.6.18 ... 4.10-rc8
<nrb> This is free software; see the source for copying conditions.
<nrb> enabled features: AVAHI LIBSQLITE3 NLS NSS
<fche> ok, that's three years old and doesn't claim to support kernel 5.7
<nrb> i think i will be good off compiling it myself from github
<fche> consider updating to a newer version?
<nrb> sure, but this is the newest one from my repo
<fche> understood; might want to open a bug against debian to bump that up to a newer release or somesuch
<fche> in general, using an old systemtap against a newer kernel is iffy
<nrb> where can the newest version be found?
<nrb> github?
<nrb> aight gonna head for a break while this compiles
<nrb> thanks for now tho
<fche> righto
<fche> https://sourceware.org/git/?p=systemtap.git;a=blob_plain;f=README;hb=HEAD << has stap build instructions incl. for debian hosts
<fche> (to help with prereqs etc.)
<nrb> i have built it, tried to run, still kbuild exited with status 2
<nrb> i think it's gcc now
<nrb> it's 6.3.0 version
<fche> I seem to recall one stap change a few months ago that could manifest itself as mystery kbuild failures like that one
<fche> but your version should be all good, unless it's invoking the /usr/bin/ system copy accidentally
<nrb> i keep having gcc errors
<nrb> maybe i should try update gcc
<fche> probably not
<nrb> so what would you propose at this point?
<nrb> it's a mess
<fche> ok how about stap --vp 0004 -r /path/to/build/tree -e 'probe oneshot { log("hi") }' 2>&1 | $paste_the_whole_thing_to_pastebin
<fche> eww yeah that looks like some compiler problem
<fche> did you build the kernel with the exact same compiler that stap's using (/usr/bin/gcc probably) ?
<nrb> no, not the same
<nrb> on host machine i have a 9.3.0
<fche> ok, you will need the sameish compiler
<fche> you can run systemtap on the host machine too, even if you're running the stap job somewhere else
<nrb> yeah but i need 5.7.7
<nrb> and i don't have it on my host machine + i will crash a lot as i am debugging
<fche> (you said this was a native build, I guess I should've been more specific, this sounds like a cross-machine cross)
<nrb> i will get the gcc version somewhere and recompile it
<nrb> i thought of intel -> arm
<fche> nrb, no, I mean use stap --remote $TARGET to build the stap on the host machine, and run it on the $TARGET host (via ssh; that just needs a part of the stap runtime binaries installed)
<nrb> but why remote tho
<nrb> this will only make it more difficult
<nrb> why not just run it on the machine
<fche> you can, but then you must deal with a non-native kernel build tree (hoping that it will match / be usable on the target side as is)
<fche> stap --remote is a good way to do a clean cross (machine or arch) run
<nrb> is there a documentation for this
<fche> man stap
<nrb> ok, i will shortly take a look into it
<nrb> i guess you have been debugging the kernel already, can you tell me if it is possible to step line by line in kgdb
<fche> not sure; stepi / step with any luck.
<nrb> i have it all set up, connected to the qemu machine with kernel, i can see the source code and break, but if i do n/si it will drop me out to due to some interrupts and timers
<fche> yeah, not familiar enough to say
<fche> I'm sure it's a solved problem tho
<nrb> i am trying this remote thingy now
<nrb> stap -v test.stp -remote root@localhost, the machine is on 10021
<fche> --remote methinks
<nrb> well yeah whatever
<fche> so you'll need stapsh, staprun, stapio on the target vm
<nrb> and where can i get this from?
<nrb> ok everything besides stapio shows up
<fche> they're part of a systemtap build, so if you have built it on the target it should be ok
<fche> stapio is in /usr/libexec/systemtap
<fche> should be fine
<nrb> eh this is exactly what i expected
<nrb> i will spend at least 1hr on research on how to make this work
<nrb> systemtap is so nice but i always fail at making it work
<fche> your scenario is unusual - and didn't make it clear that you had a host/target vm setup until just a few mins ago
<nrb> how can i give it a particular port
<fche> since this just passes the --remote string to ssh, you may need a .ssh/config clause to identify host/port/etc.
orivej has quit [Ping timeout: 272 seconds]
nrb has quit [Quit: Lost terminal]
<fche> nrb, by the way there is also a libvirt level of stap --remote connection; see the man page and [man stapvirt]
<fche> so something that doesn't require ssh
gavinguo__ has joined #systemtap
tromey has quit [Quit: ERC (IRC client for Emacs 28.0.50)]
ggherdov has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
sapatel has quit [Read error: error:1408F10B:SSL routines:ssl3_get_record:wrong version number]
sapatel has joined #systemtap
derek0883 has joined #systemtap
amerey has quit [Quit: Leaving]
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 272 seconds]
derek0883 has joined #systemtap
sapatel has quit [Remote host closed the connection]
<linus2> fche: stap can't probe openmp code, right?