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 [Remote host closed the connection]
khaled has joined #systemtap
khaled_ has joined #systemtap
khaled has quit [Ping timeout: 265 seconds]
tanislav000 has quit [Ping timeout: 272 seconds]
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 [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 quit [Quit: Konversation terminated!]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
<kerneltoast> fche, around?
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
<fche> barely
<kerneltoast> so the same as always
<fche> yes
<kerneltoast> push time?
hpt has joined #systemtap
<fche> hmmmmmm
<fche> is the effect of this new lock to totally serialize context usage across all cpus?
<kerneltoast> yes, to replace rcu as the synchronization mechanism
<fche> the point of the per-cpu context objects is to allow concurrent probe execution across cpus
<fche> hmmm
<kerneltoast> so nothing can use contexts after they're freed
<kerneltoast> concurrent probe execution is still allowed
<fche> how about calling it _stp_contexts_busy_ctr or something?
<fche> it should be the sum of all the c->busy values, istm
<kerneltoast> sure
<kerneltoast> good otherwise?
<fche> looks good
<kerneltoast> SHIPPIT
<irker576> systemtap: sultan systemtap.git:master * release-4.4-14-g0cc239e6f / runtime/linux/runtime_context.h: runtime_context: replace _stp_context_lock with an atomic variable
<kerneltoast> SHIPPT
<fche> ok
<fche> with any luck by the 'morn we'll have all green on this little panel I am looking at
<kerneltoast> gotta make the panels green
<kerneltoast> reminds me of my days at canonical
<kerneltoast> my manager would tell me that the little test panels weren't green
<kerneltoast> and i had to go make them green
<fche> a small matter of css
<kerneltoast> also, it looks like not even raw_* lock variants get the tracepoint out
<kerneltoast> so much for being raw
<kerneltoast> gordon ramsay would be ashamed
* fche is searching for a family-friendly ramsey meme
<fche> all the good ones contain swearing, sorry
<kerneltoast> how about this: https://imgur.com/yGkdLGE
<fche> nice
<kerneltoast> not quite the right meme but it is in fact a family friendly ramsay meme
<fche> ok I see some of the bots engaging your newest code-child already; will report the findings tomorrow
<fche> and with that, I bid you adieu, looking for some nice warm tea, burned?
<kerneltoast> fche, running the testsuite with my print patch adjusted for the lock tracepoint insanity: https://gist.github.com/kerneltoast/3536057c98a7900da6801b9d1e46eaaa
<kerneltoast> yes i would like a noice cuppa burned tea
<kerneltoast> adieu, mon capitaine
orivej has joined #systemtap
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 [Remote host closed the connection]
derek0883 has joined #systemtap
orivej has quit [Ping timeout: 256 seconds]
derek0883 has quit [Remote host closed the connection]
khaled has joined #systemtap
irker576 has quit [Quit: transmission timeout]
tanislav000 has joined #systemtap
mjw has joined #systemtap
hpt has quit [Ping timeout: 264 seconds]
orivej has joined #systemtap
lickingball has quit [Ping timeout: 268 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 272 seconds]
khaled has quit [Quit: Konversation terminated!]
khaled has joined #systemtap
<fche> kerneltoast, the rawhide buildbot has respectfully rejected even the last patch, back to the uprobes_onthefly.exp hang
tromey has joined #systemtap
amerey has joined #systemtap
orivej has quit [Ping timeout: 240 seconds]
<kerneltoast> fche, ouch, maybe you're hitting the mutex trylock now
<kerneltoast> can't get vmcores off the buildbots?
orivej has joined #systemtap
irker962 has joined #systemtap
<irker962> systemtap: alizhang systemtap.git:concious_language * release-4.4-13-gb923b6872 / runtime/Doxyfile runtime/linux/kprobes.c scripts/update-htmldocs systemtap.spec tapsets.cxx: replace more master to main/primary
amerey has quit [Ping timeout: 256 seconds]
amerey has joined #systemtap
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
lickingball has joined #systemtap
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 [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
<kerneltoast> fche yoyo
<fche> dude
<kerneltoast> i was idly pondering stap, as one does
<kerneltoast> and realized that the reason why lock tracepoints haven't nuked us till now is because of the runtime context reentrancy protection
<kerneltoast> any locks used without the context being held can explode
<fche> not that simple; we use locks all the time without context being held
derek0883 has joined #systemtap
<kerneltoast> we do?
<kerneltoast> maybe it's
<kerneltoast> locks used inside probes
<kerneltoast> which can be used without the context being held
<kerneltoast> outside the probes
<kerneltoast> is what explodes
<kerneltoast> fche, gah i forgot to sprinkle in preempt_disable after adding that atomic patch yesterday
<kerneltoast> g a h
derek0883 has quit [Ping timeout: 260 seconds]
lickingball has quit [K-Lined]
<irker962> systemtap: sultan systemtap.git:master * release-4.4-15-gfbab0ea35 / runtime/linux/runtime_context.h: runtime_context: disable preempt while holding the context
<kerneltoast> fche, just pushed what i hope is the last nail in the coffin for the un-RCU patch
<fche> ok, buildbots are on the job; rawhide lagging behind (it's retrying that crash from last night)
<kerneltoast> fche, how would you feel about this: https://gist.github.com/kerneltoast/68db137e55f3fa1add03540b5673818d
<fche> no strong opinion; rhel6 is 2.6.32 and the current code is working there
<kerneltoast> so opposed to upstreaming?
<fche> no strong opinion; maybe the version threshold is wrong?
<kerneltoast> rhel6 has backportz
<kerneltoast> but some of the older rhel6 kernels don't
<kerneltoast> and autoconfing this seems xtreme
<fche> is there a #ifdef thing to detect nonexistence of kfree_rcu ?
<kerneltoast> oh actually
<kerneltoast> kfree_rcu is a macro
<kerneltoast> hah
<fche> SHIP IT
<kerneltoast> WEW
<fche> after an #ifdef kfree_rcu vs. the kernel-version #if
<kerneltoast> of course
<kerneltoast> only the finest cuisine for my esteemed stap colleague
derek0883 has joined #systemtap
<irker962> systemtap: sultan systemtap.git:master * release-4.4-16-gaedc044d5 / runtime/task_finder_vma.c: task_finder_vma: add kfree_rcu() compat for old kernels
mjw has quit [Quit: Leaving]
derek0883 has quit [Ping timeout: 265 seconds]
tanislav000 has quit [Read error: Connection reset by peer]
orivej has quit [Ping timeout: 264 seconds]
gregwork has joined #systemtap
tromey has quit [Quit: ERC (IRC client for Emacs 27.1)]
derek0883 has joined #systemtap
modem has quit [Read error: Connection reset by peer]
modem has joined #systemtap
derek0883 has quit [Ping timeout: 256 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Remote host closed the connection]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 246 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 260 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 260 seconds]
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: 256 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 272 seconds]
derek0883 has joined #systemtap
derek0883 has quit [Ping timeout: 256 seconds]
irker962 has quit [Quit: transmission timeout]
<kerneltoast> fche, how are the buildbots doing