=== rsalveti_ is now known as rsalveti [01:29] Talk to me [01:29] What do you want [01:29] Make it quick [01:29] I did not ask you to fuck with my cell phone towers. [01:30] Are you on a top secret mission from a supa dupa spy [01:30] Talk to me [01:30] Make it quick [01:30] liars do not enter Zion [01:31] not by beaming cell phone towers and banging on the floor either [01:31] Talk to me [01:31] Make it quick [01:33] say something useful [01:34] quit bugging [01:41] Hi dayangkun [01:42] Is that linode working out well? === Kaloz_ is now known as Kaloz === broder_ is now known as broder === lag is now known as Guest73764 [03:41] RE: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1313981 I need some help understanding why I loose keyboard and mouse (both USB) when using a previous kernel version === inaddy is now known as tinoco [08:43] hans109h (N,BFTL), ok i see you found which 3.13 kernel was the first one which breaks, and that one includes an interesting change to PNP handling; which considering the debug output my last kernel produced is rather suspicious. i'll try a new test kernel with that reverted, and with some more debug [11:10] Where is the confige file used by dpkg -buildconfig? [11:11] one, i think i've seen that answers probabally 3 times already ... they are in debian.master/config [11:12] it never made it to my screen [11:12] theives and thugs for a neighborhood watch [11:12] they watch out for the feds [11:13] steal mail [11:16] What is an acorn parition? [11:16] maybe from acorn computers? [11:24] indeed a partition table format used on old acorn systems, which might potentially be on an USB drive [11:25] What is the most compatible way to edit the config for use with the dpkg -buildpackage? [11:25] debian/rules editconfigs [11:26] or change the config in the config fragment and run debian/rules updateconfigs [11:26] Does that include editing the localversion label? [11:29] The enforce file says auto adding localversion will cause package failure [11:31] setting localversion breaks the build which relies on the form of the version number [11:43] jibel, hey ... what triggers the dkms test builds against the CKT PPA [11:44] and indeed how can i tell if it has done :) [11:45] apw, Reminds me, not sure you saw the email, I managed to give iscsitarget a basic test in utopic. So whatever you did seems to work [11:46] smb, yeah i did thanks. that is more than i know how to do [11:46] i guess that is something you should teach us [11:46] the basic concepts and bits needed for that to work [11:47] yeah, I suppose I could do it in 2 weeks when we are in the same location [11:47] smb, sounds good indeed [11:47] i assume VMs will do [11:55] Looking at the config in place and comparing with lsmod it looks like the config fragments have things set as y so they are supposed to not be compiled as modules but the lsmod is showing them as loaded modules, there must be something else changing the configs, any idea what that is? [11:55] one, give us an example and i can look [11:56] kvm [11:57] wait that is seperate from config_have_kvm [11:57] Is there not an ncurses package in the repo? [11:57] The menuconfig gives explanations. [11:59] Does kconfig or xconfig also give a summary? [12:00] The package system calls for ncurses if selecting y to modify the config before build === josepht_ is now known as josepht [12:10] apw, a new version of the kernel in the PPA or a new version of a dkms module in proposed. In this case only this module will be tested. [12:12] jibel, in the case a new kernel dropped, it is not easy to tell from jenkins if it has been redone, is tehre something else somewhere to tell me it has [12:12] jibel, i believe it has in this case, as i opened one of the latest builds and checked, but from the dashboardy thing i couldn't tell [12:14] apw, from jenkins it is not easy to tell indeed. On http://d-jenkins.ubuntu-ci:8080/view/DKMS/view/U%20KT-PPA%20-generic/ you'll see that all the tests have run 9h30 ago, and if you open one you'll see the version of the kernel. [12:17] apw, I'll check with the CI team because apparently something changed in jenkins. Previously the version of the kernel and dkms modules was displayed in the list, but it is not there anymore. [12:18] and I cannot find a way to add it [12:21] jibel, great thanks, i am late to this party so i have not seen the live one before now to know it has gone :) [12:22] apw, it should like https://jenkins.qa.ubuntu.com/view/DKMS/view/Trusty%20Release%20Kernel-Team-PPA%20-generic/ [12:28] jibel, and i can't see it on that one either, or i am stupid, possibly the latter [12:32] apw, ah right, the version is just "generic" in the list called 'test statistics grid'. not really useful. [12:46] tjaalton, bug #1317865 looks like it might be yours [12:49] rtg: thanks [12:49] I have other stuff to push too [12:51] ffs, dropbox.. [12:52] oh right, the one smb reported earlier [12:53] smb, sforshee: this one looks interesting: bug #1317811 [12:54] bug 1317811 [12:54] Hm, might be something that sforshee had been lookig at [12:54] ok [12:56] Not completely the same. His was the ride the rocket message and finally ending in a bugon fragments spannign a compound page or so [12:57] perhaps related [12:59] I guess the basic question is why buffers get so big in general. IIRC this is trying to transfer a packet where the skb is more than they can fit into 19 slots (where I had to check again how big a slot is but I think a page) [13:00] that makes sense since the workaround is to restrict MTU to 1500 [13:01] right... hm, has the default really gone up? maybe some negotiation [14:19] apw: there is no updateconfigs in that directory [14:21] that autocomplete , so rules is a script and updateconfigs is an argument for it? [14:23] one yes [14:24] smb, bigger than 19 pages !? [14:27] Is there a package that provides ncurses so that the editconfigs may be run? [14:28] apw, Yes, it counts the number of pages the frags in a skb need and anything above 17 results in that rides the rocket message and a drop of that package [14:29] Or does ubuntu allow for kconfig or other config apps? [14:31] one, the package for ncurses, is libncurses-dev [14:35] It looks like updateconfigs does run kconfig and it is now looping is it safe to ctrl+c it? [14:36] it can make a mess if it is stopped [14:37] it shouldn't look forever [15:11] What is the proper way to clean the kernel source directory I am working on building the kernel a second time removing some unnecisarry things from the config and the dpkg-buildpackage exits with errors and a lot of complaints about not being able to represent changes. [15:11] unrepresentable changes to source [15:15] Is it due to not being able to create a hash diff for nonexistant modules given that there is removal of modules in the new config? [15:15] debian/rules clean [15:16] it will likely fail later unless you fix the ABI as well, adding removed modules to "modules.ignore" in the abi directory [15:17] And what about .. where the package shall go does it append to filename or overwrite considering appending localversion breaks the process? [15:18] the package will be whatever version you have at the top of debian.master/changelog, it is common to add like apw1 to the end of the version to indicate this is your version [15:22] smb, have you configured an openvsiwtch bridge like we did a normal one for xen [15:27] So what reads the symbols to make it happen? [15:30] one, "it" ? [15:35] Whatever the symbol represents. [15:36] Now, how is this build system used to build only one of the supported archs the last time it produced a package labled _all [15:37] one, what symbols are we talking about? i don't see you mention them before and that sentence is context free [15:38] one, _all is the architecture independant packages, common to all architectures [15:38] -b says build the arch independant and arch dependant packages for "this" architecture [15:42] It was said that it builds for the arch it is built on however it says building for i386. [15:42] This is not a 386. [15:43] is it a 32bit machine? [15:43] I did have a 386 in the past not sure where it went. [15:43] 64 [15:43] and does it have 32bit ubuntu installed on it [15:44] Yes it is running 32bit ubuntu. [15:44] then your architecture is "i386" which is a debian architecuture name for the 32bit x86 Os [15:45] The symbols are in the abi directory conscerning the modules. [15:45] No your architechture is not 386 [15:46] ok, believe what you want [15:46] no [15:47] but it is a fact that the "i386" debian architecture is 32bit ubuntu on x86, "amd64" is the debian arch for 64bit ubuntu on x86 [15:47] Do you call Jesus a carpenter like the religious leaders? [15:48] an interesting non-sequitor, i have no reason to call him anything [15:48] If I can do something lesser that is not the sum of my architechure [15:49] In the same way computers are reflections of a mind. [15:49] With a great latency. [15:50] apw: can you remind me where 3.15 is? (the ppa) [15:50] please [15:50] tseliot, the utopic pocket of ppa:canonical-kernel-team/ppa [15:50] apw: both hardware and software [15:50] apw: thanks [15:51] tseliot, np, good luck with the testing :-p [15:51] thanks :) [15:51] well, with test-building [15:51] build testing... [15:51] computer is a physical representation of thoughts a mind had in the past. [15:52] tseliot, don't forget the wl driver. its got some serious problems as well. [15:52] rtg: that's the driver I'm working on [15:52] oh, cool. [15:53] tseliot, did you find the channel somewhere to steal, i found two places you could but who knows if they are sufficient, sigh [15:54] apw: I get it right before sending the variable, so that I'm sure nothing else messed with it. I only need to test the patch [15:54] tseliot, oh nice, be interested to see where you got it from :) [15:54] apw: how is a more advanced arch specified? [15:54] apw: do not run from the truth [15:54] one, that doesn't make sense, the arch is just a piece of text representing the architecture [15:55] it is rather stupidly called "i386" but it is not even compatible with a 386 machine [15:55] tseliot, i thi [15:55] apw: I see the robots interrupt when I begin to reveal [15:56] yupanakernel wants to live forever [15:56] tseliot, i think i have a POS with some of that h/w in it, if you want to throw me a package [15:56] does apw? [15:56] apw: sure, that would help [15:58] such timing! it is to be expected. cross empty or occupied I live forever. [15:59] That anything is here shows that JESUS CHRIST is the FATHER in the flesh. [15:59] our report [16:00] is this the 'we' that apw suspects [16:00] our report [16:00] this is not the right channel to discuss such topics... [16:01] apw: do you accuse me of fault? [16:01] it's #ubuntu-kernel [16:01] share your logic template then [16:02] :) [16:02] bjf, how long does that last ? [16:02] rtg, i don't know [16:23] apw: my patch seems to build. What architecture do you need the packages for? [16:24] tseliot, i think that machine is 32bit [16:26] apw: ok [16:26] tseliot, or throw me the debdiff, and i can build it here [16:36] apw: http://people.canonical.com/~amilone/bcmwl_6.30.223.141+bdcom-0ubuntu3.debdiff [16:36] I'm also creating a 32 bit chroot === Guest73764 is now known as lag [16:52] apw: and here's the package http://people.canonical.com/~amilone/bcmwl-kernel-source_6.30.223.141+bdcom-0ubuntu3_i386.deb [17:01] tseliot, thanks, machine is upgrading, i do not run it often enough it seems [17:01] :) [18:04] apw, Not sure I ever completed to use opevswitch in replacement [22:04] smb, ok thanks