1467937469 J * fstd_ ~fstd@x5ce48ee1.dyn.telefonica.de
1467937936 Q * fstd Ping timeout: 480 seconds
1467937936 N * fstd_ fstd
1467943021 J * derjohn_mobi ~aj@x590c63d4.dyn.telefonica.de
1467943463 Q * aj__ Ping timeout: 480 seconds
1467953526 Q * sannes Quit: Leaving.
1467953570 J * sannes ~ace@2a02:fe0:c131:9070:a99f:c1c5:8f1a:e6a1
1467953599 Q * sannes 
1467953832 J * sannes ~ace@2a02:fe0:c131:9070:203f:59f7:29c7:1cd6
1467961257 J * Ghislain ~aqueos@adsl1.aqueos.com
1467966761 N * Bertl_zZ Bertl
1467966765 M * Bertl morning folks!
1467966863 M * Ghislain helo bertl !
1467967280 Q * derjohn_mobi Ping timeout: 480 seconds
1467967884 Q * tokkee Ping timeout: 480 seconds
1467969053 J * tokkee tokkee@osprey.tokkee.org
1467969947 J * derjohn_mob ~aj@2001:6f8:1337:0:980c:12d0:3c4:faab
1467974279 Q * Aiken Remote host closed the connection
1467974637 N * Bertl Bertl_oO
1467977416 M * Ghislain is there anything i can do to help on 4.1.x thing ?
1467977471 M * Ghislain fir now the kernel runs fine, i have looked a little at the dcache but i did not find any issue in the counting apart from when it reach 0 they have special case that could be the root of the issue
1467977535 M * Ghislain as they seem to return 1 when it reach 0 in certain case htat perhap would need a -- for the counter
1467977548 M * Ghislain but i barely understand the basics here
1467977751 J * Gremble ~Gremble@cpc87179-aztw31-2-0-cust6.18-1.cable.virginm.net
1467978030 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467978683 J * thierryp_ ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467978784 Q * thierryp Read error: Connection reset by peer
1467979594 Q * Gremble Quit: I Leave
1467979667 M * Guy- Ghislain: you could see if you can reproduce my sleep bug :)
1467980248 Q * thierryp_ Quit: ciao folks
1467980334 M * Ghislain how your tar work there not even a /etc/hosts in it
1467980479 M * Ghislain i try vserver sleepbug build   -m skeleton -n sleepbug --context 40003 --initstyle plain
1467980479 M * Ghislain  and put you tgz and see
1467980558 M * Ghislain ok
1467980575 M * Ghislain it runs and the sleep do not trigger any crash or message in dmesg
1467980700 M * Ghislain i do not understand how it runs but it runs and sleep trigger nothing
1467981016 M * Guy- Ghislain: why wouldn't it run without a /etc/hosts? it doesn't even have networking
1467981033 M * Guy- Ghislain: but the magic that triggers the bug is most certainly not in the guest, but in the host kernel
1467981048 M * Ghislain the only dmesg i have is [Fri Jul  8 14:29:12 2016] vxW: [\xffffffbbsystemd\xffffffab,23451:#40003|40003|40003] denied [0x24] access to inode proc:ffff88042cec3d28[#0,4026531986]
1467981066 M * Guy- I had an existing box, with several guests, and no problems other than the samba related unix domain socket bug
1467981074 M * Guy- I upgraded the kernel to get rid of that bug, changed nothing else
1467981080 M * Guy- and now I have this sleep bug
1467981086 M * Guy- so it's a problem with the kernel, not the guests
1467981088 M * Ghislain i run 4.1.27-vs2.3.8.5.2aq on it
1467981095 M * Ghislain yep
1467981097 M * Guy- the way to reproduce it would be to compile a kernel with my config
1467981103 M * Guy- and see if it breaks the same way
1467981106 M * Ghislain use Dtrace..oh no linux do not have that
1467981269 M * Ghislain you do not enable memory accounting with the swap ?
1467981275 M * Guy- no
1467981350 M * Ghislain i would remove this CONFIG_SCHED_AUTOGROUP=y , hum you use deadline is it possible to chnage the scheduer on runtime to see if i can reproduce that ?
1467981383 M * Ghislain you use kvm and xen on those hosts ?
1467981783 M * Ghislain to remove bloat i would remove ext2/3 from the kernel, remove xen KKvm if you not use them, remoeve the CONFIG_SCHED_AUTOGROUP=y because this seesm dangerous to have kernel move things around on a server imho
1467981868 M * Ghislain i do not see big differences in your kernel apart the sceduler and all the xen/kvm stuff
1467981907 M * Guy- Ghislain: no, I don't use xen or kvm, but these options have always been enabled; ext2/ext3 likewise
1467981913 M * Guy- AUTOGROUP is also not new
1467981921 M * Ghislain   yes but allways good to get rid of bloat
1467981969 M * Guy- not always
1467981977 M * Guy- for example, I may conceivably need kvm/xen at short notice
1467982059 M * Guy- Ghislain: this is a diff between the config of the last kernel that didn't have the sleep bug and the first kernel that did
1467982062 M * Guy- Ghislain: http://sprunge.us/IEgO
1467982085 M * Guy- I suspect one of the "kernel hacking" options
1467982121 M * Guy- CONFIG_DEBUG_ATOMIC_SLEEP=y wasn't even turned on before
1467982132 M * Guy- so maybe I always had this problem but it was never reported
1467982215 M * Ghislain what do all +CONFIG_UPROBES=y ? can be also the code that do 'stuff'
1467982428 M * Guy- I was hoping to have better instrumentation to debug future kernel issues
1467982478 M * undefined Guy-: doesn't sound hardware specific, so why not try to test it on some desktop virtualization platform (eg virtualbox, vmware player)
1467982495 M * undefined that's what i did for the last kernel bug i found
1467982526 M * undefined made it a lot easier to trigger and debug than continually rebooting the server i found the problem on
1467982527 M * undefined ;)
1467982541 M * Guy- undefined: yes, that would allow me to find out which kernel config option triggers/masks it
1467982556 M * Guy- unfortunately I have zero free time to do this in
1467982576 M * undefined ah, so you are trying to get Ghislain to do your debugging!  i see now. ;)
1467982601 M * Guy- undefined: he asked if there was any way he could help :)
1467982603 M * Ghislain dam ! this is bad dude ;p
1467982642 M * Guy- (obviously it's not just "my" debugging)
1467983680 M * Ghislain well i tried your guest on my 4.1.27 and this do not trigger it, doyou had the issue before adding the probes into the kernel ?
1467983699 M * Ghislain or did you added the KPROBES in your first 4.1.27 ?
1467984379 M * Guy- yes, I enabled the probes while I built my first 4.1.27
1467984400 M * Guy- I have a hunch that the messages are only logged if CONFIG_DEBUG_ATOMIC_SLEEP=y
1467985380 M * Guy- /o\
1467985408 M * Guy- sorry, wrong window
1467988315 Q * derjohn_mob Ping timeout: 480 seconds
1467988662 J * thierryp ~thierry@82.226.190.44
1467989481 J * Aiken ~Aiken@d63f.h.jbmb.net
1467990398 Q * transacid Remote host closed the connection
1467990417 J * transacid ~transacid@transacid.de
1467993362 Q * thierryp Read error: Connection reset by peer
1467993487 J * thierryp ~thierry@82.226.190.44
1467993594 Q * thierryp Read error: Connection reset by peer
1467993613 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467994106 J * thierryp_ ~thierry@82.226.190.44
1467994170 Q * thierryp Read error: Connection reset by peer
1467994568 J * derjohn_mob ~aj@x590c63d4.dyn.telefonica.de
1467994714 Q * thierryp_ Read error: Connection reset by peer
1467994909 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467995694 J * thierryp_ ~thierry@82.226.190.44
1467995725 Q * thierryp Read error: Connection reset by peer
1467996097 Q * thierryp_ Read error: Connection reset by peer
1467996156 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467997064 Q * thierryp Read error: Connection reset by peer
1467997067 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:dd6:6345:ebee:54c5
1467997248 Q * thierryp 
1468003229 Q * transacid Ping timeout: 480 seconds
1468003748 J * transacid ~transacid@transacid.de
1468004895 Q * Defaultti Quit: Quitting.
1468005005 J * Defaultti defaultti@lakka.kapsi.fi
1468016919 Q * sladen Remote host closed the connection
1468016924 J * sladen ~paul@starsky.19inch.net
1468019710 Q * Ghislain Quit: Leaving.