1448238377 Q * Ghislain Quit: Leaving.
1448240360 Q * fstd Remote host closed the connection
1448240377 J * fstd ~fstd@xdsl-87-78-14-190.netcologne.de
1448262223 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:4d03:55d8:7a78:38c2
1448262299 Q * thierryp Remote host closed the connection
1448263722 J * Ghislain ~aqueos@adsl1.aqueos.com
1448265202 Q * aj__ Ping timeout: 480 seconds
1448265765 J * thierryp ~thierry@zeta.inria.fr
1448266839 J * aj__ ~aj@2001:6f8:1337:0:b1c4:4181:5721:47d6
1448273551 J * Gremble ~Gremble@cpc87151-aztw31-2-0-cust755.18-1.cable.virginm.net
1448283560 Q * fstd Remote host closed the connection
1448283572 J * fstd ~fstd@xdsl-81-173-185-216.netcologne.de
1448284539 Q * thierryp Remote host closed the connection
1448284576 J * thierryp ~thierry@zeta.inria.fr
1448285766 N * Bertl_zZ Bertl_oO
1448285789 M * Ghislain hello there
1448291175 M * Ghislain 3.4.110, compile and bot testfs [231]# failed  [213]# failed. [203]# failed. http://pastebin.com/raw.php?i=vgseP8d4
1448291250 M * Ghislain this one  as a dcache.c bug corrected this is why i am particulary interested in testing it while 41. is on the work ^^
1448291587 J * Wermwud ~Wermwud@69-29-150-18.stat.centurytel.net
1448292193 Q * eyck Quit: leaving
1448292249 J * eyck ~eyck@u28n61.nowanet.pl
1448292544 M * Ghislain testme.sh => no error
1448292624 Q * eyck Quit: leaving
1448292665 J * eyck ~eyck@u28n61.nowanet.pl
1448292813 Q * eyck 
1448292864 J * eyck ~eyck@u28n61.nowanet.pl
1448292914 Q * eyck 
1448293071 J * eyck ~eyck@u28n61.nowanet.pl
1448295542 Q * thierryp Remote host closed the connection
1448295717 M * Bertl_oO I wouldn't worry too much about 231 failing with an offset of 4
1448295733 M * Bertl_oO I'm a little more worried that the numbers are negative :)
1448295773 M * Bertl_oO but that might be okay, I do not remember how the test was coded
1448296081 M * Ghislain well this is dark magic wrapped in some vodoo i guess ^^
1448296265 M * Ghislain it mount a FS then compare df and dlimits
1448296467 M * Ghislain it do not seems to write anything the test just check diffs
1448296542 M * Ghislain oh it remove some files rm -f $MNT/cow-300*
1448296658 M * Ghislain cow test create files with dd
1448296692 M * Ghislain it creates 43 files
1448296703 M * undefined 3.10 and 3.14 have failed 203, 213, & 231 forever
1448296776 M * Ghislain it save the limits, creates 43 1k files and compares
1448296832 M * undefined Ghislain: the only difference between what you report and what i encounter (in 3.10 and 3.14) is the difference in the number of nodes
1448296849 M * undefined i'm encountering a difference in inodes of 1
1448296864 M * undefined me: 17 ~ 16
1448296871 M * undefined you: 20 ~ 16
1448296911 M * undefined i'm encountering off-by-1 while you are encountering off-by-4
1448296997 M * undefined maybe difference in preallocation setting(s)
1448297098 M * Ghislain i added /bin/sync before the test
1448297106 M * Ghislain then i have no fails
1448297162 M * Ghislain in do_dlimit_save i added /bin/sync
1448297170 M * Ghislain and in do_dlimit_verify too
1448297218 M * Ghislain 7 times all succedded,  could we call it a win ?
1448297472 M * undefined yeah, currently a sync is only done at the beginning of the test
1448297504 M * Ghislain is my modification making sense or invalidate the test ? i should think it validate as the goal is to see if this is accounted for, not that it is accounted at the second. This means the caching code is not incrementing the dlimits at the same time as the real limit
1448298368 M * Bertl_oO nah, sync is perfectly fine, we'll add that
1448298485 M * undefined i did a sync at the top of do_dlimit_verify and that was sufficient
1448298602 M * undefined does hint that vdlimit is only taking into account allocations made on disk and not to the filesystem but still in-memory/write-cache (journal?)
1448298614 M * undefined at least for ext4
1448298687 M * Bertl_oO whatever the kernel side does to get an inode/block :)
1448298705 M * Bertl_oO i.e. we do not handle preallocations any different than normal allocations
1448299011 M * Ghislain this discrimination toward preallocated inodes !
1448299054 M * Ghislain undefined: oh yes i just tested ext4 selfish that i am
1448299263 M * Ghislain reboot -d -f -i inside the guest works fine also i havent been able to crash it  with our simple manual obvious crash test
1448299283 M * Ghislain so seems the 3.4.110 is fine apart some hunk in patching
1448301634 J * bonbons ~bonbons@ppp-105-93.adsl.restena.lu
1448301780 Q * Gremble Remote host closed the connection
1448307996 Q * aj__ Ping timeout: 480 seconds
1448308203 M * Bertl_oO off to bed now ... have a good one everyone!
1448308205 N * Bertl_oO Bertl_zZ
1448308723 J * thierryp ~thierry@82.226.190.44
1448309521 J * aj__ ~aj@6YRAAAUQ6.tor-irc.dnsbl.oftc.net
1448310353 Q * thierryp Remote host closed the connection
1448312851 J * sannes ~ace@2a02:fe0:c120:9660:6de0:b394:e2dc:673c
1448313071 Q * aj__ Ping timeout: 480 seconds
1448313588 J * aj__ ~aj@tmo-100-240.customers.d1-online.com
1448314747 J * garilber ~garilber@91FAAATUQ.tor-irc.dnsbl.oftc.net
1448314824 M * garilber salut
1448314842 P * garilber 
1448316271 Q * bonbons Quit: Leaving
1448317004 Q * sannes Remote host closed the connection
1448319328 Q * Wermwud Quit: Leaving (Please imagine me slamming the door on my way out)
1448319871 Q * laurens Ping timeout: 480 seconds
1448321177 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:2dd4:2ac3:27ca:f42c
1448321660 Q * thierryp Ping timeout: 480 seconds
1448322770 Q * Ghislain Quit: Leaving.