IRC logs for #baserock for Wednesday, 2015-01-21

*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]01:53
*** zoli_ [] has joined #baserock03:03
*** zoli_ [] has quit [Changing host]03:03
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock03:03
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]04:29
*** mike [] has joined #baserock07:34
mike is now known as Guest596107:34
*** zoli_ [] has joined #baserock07:46
*** zoli_ [] has quit [Changing host]07:46
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock07:46
*** grahamfinney_ [] has joined #baserock08:01
*** mariaderidder [] has joined #baserock08:14
*** wdutch [] has joined #baserock08:23
*** bashrc [] has joined #baserock09:03
*** ssam2 [~ssam2@] has joined #baserock09:37
Mode #baserock +v ssam2 by ChanServ09:37
*** jonathanmaw [~jonathanm@] has joined #baserock09:44
*** sambishop [~sambishop@] has joined #baserock09:45
*** CTtpollard [] has joined #baserock10:03
bashrcon login I'm getting: Error: OpenID failure: time_bad_sig: Return_to signature is not valid. 10:09
jmacsHmm, seems OK here. What id provider did you log in with?10:10
*** franred [] has joined #baserock10:10
*** Krin [] has joined #baserock10:12
* Kinnison can log in, using stackoverflow's provider10:13
mwilliams_ctI seem to recall someone saying they'd seen that error intermittently, will check irc logs10:14
*** lachlanmackenzie [~lachlan@] has joined #baserock10:16
bashrcI created an account on, so I don't have to use OpenID10:21
KinnisonI have no idea how you would have10:21
Zara_bashrc: I get that error every so often; I don't know why it happens on some days but not others. One day it's fine, the next I have to use git to edit. So I'm also curious about this.10:37
Zara_mwilliams_ct: that was probably me10:37
bashrcI think it depends upon the emotional state of the wiki10:38
Zara_Maybe it only works for me on odd-numbered days.10:39
bashrcor maybe it's not the wiki, it could be Google10:40
* bashrc finds comments like "This is basically a hacked version of..."10:42
bashrcand "it is not a reasonable fix"10:42
bashrcin morph10:45
jmalkgiven that git send-email isn't available in baserock, what's the easiest way to mail a patch to the list?10:54
Zara_jmalk: I use git send-email from outside of baserock10:55
SotKjmalk: I push the branch somewhere, clone it on my debian machine and use git send-email there10:55
jmalkSotK, Zara_ thanks10:55
petefoth jmalk: or mount drive in the baserock VM that contains the source in your host using sshfs. I think there are destructions on how to do that on wiki.baserock.org10:58
tiagogomes_reading Sam's email, no `moph petrify` will not come back?10:59
persiatiagogomes_: We all hope not10:59
tiagogomes_why not? Does one still will have to manually edit definitions to replace branches names with SHA1s?11:00
ssam2tiagogomes_: see the suggestion of a `morph set-ref` command, in the same mail11:08
tiagogomes_ah ok, when I read it in the first time, the suggestion of the optional --tag flag made me thinking that it would only allow you to change a ref to a SHA1, and not to other types of ref11:15
bashrchow old is the systemd version in baserock?11:49
richard_mawhow new do you need it to be?11:51
* richard_maw is checking which version is available11:51
bashrcas new as possible11:51
richard_mawwe have v217, we're only one released version behind11:52
*** a1exhughe5 [] has joined #baserock11:53
richard_mawwhat in particular do you need systemd to be able to do?11:53
bashrcI was just wondering if it had a certain bug11:54
bashrcwhich was probably fixed in 21411:55
bashrcshould there be any .network files within /usr/lib/systemd/network ?12:10
ssam2i think so, yeah12:10
bashrcmaybe that's the problem12:10
ssam2the systemd chunk install one, if I recall correctly12:11
ssam2ah, no it doesn't12:11
ssam2but my Baserock chroot has  /etc/systemd/network/, which must come from somewhere L)12:12
ssam2it does indeed come from the systemd chunk according to the metadata in /baserock12:13
richard_mawnetwork config in /usr/lib is a bit weird, since it needs to be valid network config for every system12:14
ssam2oh, I missed that bashrc was talking about /usr/lib/systemd/network not /etc/systemd/network12:14
ssam2ignore me12:14
richard_mawwhich means its uses are pretty much limited to networking within the machine12:15
richard_mawso containers and VMs etc.12:15
richard_mawand AIUI v217 doesn't do any of that, but later releases do12:15
richard_mawhence why the /usr/lib network config path is empty12:15
*** a1exhughe5 [] has quit [Ping timeout: 256 seconds]12:22
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]12:27
*** zoli_ [] has joined #baserock12:30
*** zoli_ [] has quit [Changing host]12:30
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock12:30
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]12:32
*** wdutch [] has quit [Quit: Quit]13:59
*** wdutch [] has joined #baserock13:59
*** wdutch [] has quit [Client Quit]14:00
*** wdutch [] has joined #baserock14:00
straycatheh, hg help config contains info for use with plan914:08
*** zoli_ [] has joined #baserock14:09
*** zoli_ [] has quit [Changing host]14:09
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock14:09
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]14:26
*** zoli_ [] has joined #baserock14:43
*** zoli_ [] has quit [Changing host]14:43
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock14:43
*** a1exhughe5 [] has joined #baserock15:02
jmalkjust sent v2 of my cross-bootstrap branch to the list - AFAIK it's functionally the same but followed rjek's recommended naming convention15:02
bashrcshould baserock have a resolv.conf ?15:12
KinnisonIt should automatically make one when it DHCPs15:19
pedroalvareznowadays resolv.conf is a symlink to somewhere created by systemd15:26
paulsherwooddid anyone discuss or think about ssam2's idea to adopt ostree?15:48
paulsherwoodwhen i looked into ostree it seemed quite a major thing. would it be even applicable for embedded, for edxample?15:49
robtaylori see no reason not15:49
robtaylorits really just the equivelent of our chunks and chunk server, really15:50
robtaylor'git for file system contents'15:50
*** Guest5961 [] has quit [Quit: Leaving]15:55
paulsherwoodi think it does rather more than that :)15:59
robtaylortrue, but the rest is about deploying from that store, and can be done different ways16:13
paulsherwoodit seems to be a potential basis for atomic updates iiuc16:14
robtaylorthere's a hardlink tree deployment16:15
paulsherwoodso the question is whether that would be something to consider for baserock target systems? 16:16
* pedroalvarez kicks a build of the toolchain-upgrade branch on ppc64 and armv7lhf16:16
paulsherwoodpedroalvarez: :-)16:17
pedroalvarezI'll have some feedback in some hours :)16:17
tiagogomes_thanks pedroalvarez16:22
pedroalvareznp ;)16:22
ssam2well done tiagogomes_ !!16:24
tiagogomes_This is a very nice reading about cross-compilation:
tiagogomes_very clear16:27
ssam2tiagogomes_: did you test cross-bootstrap with your updated toolchain by the way?16:29
ssam2it seems to have many users these days16:29
tiagogomes_ssam2 not yet16:29
KinnisonTiago's work was done in part to facilitate cross-bootstrapping to aarch6416:29
KinnisonWe hope to begin testing that v.soon16:30
Kinnisonwe're rushing to try and get something up on our HP moonshot ready for the baserock meetup in a couple of weeks16:30
Kinnisonv. exciting times16:30
ssam2would it be helpful to have this patch series merged quickly, given that if cross-bootstrap breaks in master you'll be fixing it soon enough anyway?16:31
ssam2or should we wait?16:31
KinnisonLet's wait until we have a little more testing done on it16:32
Kinnisonbut if any of you lot can try it out we'd appreciate it16:33
Kinnisonit has taken a lot of work for Tiago16:33
ssam2I suppose it won't really bitrot unless we merge other toolchain stuff in the meantime16:33
Kinnisonthere's the risk the series might bitrot if Paul's morph-arch stuff was merged16:34
ssam2yes, I suggest we leave that til after16:34
Kinnisonbut frankly I'm tempted to ask that we hold off on that and merge it as part of a series to add aarch64 support16:34
KinnisonMay as well deal with it in one series16:35
ssam2well, ok. I'm happy either way16:35
pedroalvarezI will try to test tiago's branch asap (also cross-bootstraping) so we can merge it soon so it doesn't need rebasing16:36
Kinnisonpedroalvarez: thanks16:37
pedroalvareztiagogomes_: first error:  stage1-gcc in armv7lhf:
pedroalvarezI can get the full log if needed16:39
tiagogomes_damn it16:40
pedroalvarezppc64 has built it though16:40
paulsherwoodthat's a w00t :)16:45
pedroalvareztiagogomes_: ppc64 - stage2-make16:50
Kinnisonfailed at, or is currently doing?16:51
pedroalvarezoops, forgot to attach the link16:52
pedroalvarezfailed at16:52
Kinnisontiagogomes_: Hmm ^^^ looks like that missing sysroot-passed-to-linker issue16:54
Kinnisontiagogomes_: yet -Wl,--sysroot is on the cmdline16:54
*** mariaderidder [] has quit [Quit: Ex-Chat]16:58
*** a1exhughe5 [] has quit [Quit: Leaving]16:58
tiagogomes_Kinnison, I don't know from where it comes from, there is no -Wl,--sysroot in definitions17:03
tiagogomes_I am assuming that if there is a stage2-morph in definitions, the one in the repo is ignored17:08
tiagogomes_Because either stage2-morph in the repo is being favored, or not the right system branch is being checked out17:09
*** jonathanmaw [~jonathanm@] has quit [Quit: Leaving]17:09
*** wdutch [] has quit [Quit: Quit]17:12
ssam2tiagogomes_: if there is a 'morph' field for that chunk in the stratum it will look in definitions for its morph file17:13
ssam2without the 'morph' field I'm not sure17:13
tiagogomes_it does17:14
tiagogomes_anyway, I'll not go anywhere just by looking at the logs. Is it possible for me to have access to the ppc64 machine?17:15
pedroalvareztiagogomes_: I can give you access to one ppc64 box that I have around. Give me a sec and I'll /msg you17:16
tiagogomes_ok, I'll start with the ppc64 failure as it is almost end of the day and it seems easier :)17:17
tiagogomes_ppc64 sorted!17:20
tiagogomes_It was my branch, but not the right ref17:21
tiagogomes_an old ref17:21
tiagogomes_2015-01-21 17:21:13 Updating cached git repository upstream:binutils-redhat for ref b1d3b01332ae49a60ff5d6bf53d3a5b1805769c817:22
tiagogomes_ERROR: Git directory /srv/distbuild/gits/git___ct_mcr_1_ducie_codethink_co_uk_delta_binutils_redhat has no commit at ref b1d3b01332ae49a60ff5d6bf53d3a5b1805769c8^{commit}.17:22
tiagogomes_I was trying a `morph build`, should I do a distbuild?17:22
*** grahamfinney_ [] has quit [Ping timeout: 245 seconds]17:24
*** Krin [] has quit [Remote host closed the connection]17:26
pedroalvareztiagogomes_: hm.. oh yeah, the machine I gave you access is using a different trove, and maybe is taking some time to propagate 17:27
tiagogomes_pedroalvarez, could the same being happening to the armv7lhf machine?17:32
pedroalvareztiagogomes_: I don't think that's the problem. I've changed the configuration of the ppc64 machine. It should build now.17:33
pedroalvareztiagogomes_: so is f2532705c6e110bf0012cac0933abcbf1c5ed848 the right sha1?17:35
tiagogomes_pedroalvarez, ok, I'll wait news from you on the ppc64 side. Can I have access to the armv7lhf machine?17:35
tiagogomes_pedroalvarez, yep17:35
pedroalvarezok, started building again17:35
*** ssam2 [~ssam2@] has quit [Remote host closed the connection]17:37
robtaylorcame across this randomly today
paulsherwoodwoah! 17:48
robtaylorRob Landley rings a bell for me, but I can't place him17:48
paulsherwoodwho is this masked superhero?17:48
tiagogomes_pedroalvarez, in which file should I look for the distbuild logs? /srv/distbuild/morph.log?17:51
Zara_Landley's website says he wrote a lot of Busybox, so possibly that's where he's familiar from?17:52
richard_mawrobtaylor, paulsherwood: Rob Landley was the primary busybox maintainer17:52
pedroalvareztiagogomes_: is not distbuilding. Is just doing a local build in a distbuild node. You will get the logs in /srv/distbuild/morph.log, yeah17:52
robtaylorrichard_maw: ahh17:54
robtaylorthat;s the one. probably interacted with him back when I was fixing bugs in busybox..17:54
jmacsI can't see any reason for the checks for "eth0:" and "eth1:" in get_host_interface in virtualbox-ssh.write18:00
jmacsMaking a patch to remove them unless someone can enlighten me18:00
*** sambishop [~sambishop@] has quit [Remote host closed the connection]18:02
*** bashrc [] has quit [Quit: Lost terminal]18:03
robtaylorIt's clearly a day of cool things. also found this http://ellcc.org18:03
pedroalvareztiagogomes_: good news. In ppc64 is now building gawk18:20
pedroalvarezlast one, stage 318:21
pedroalvareznow gcc. If it builds, I guess everything else will be ok18:21
tiagogomes_pedroalvarez great! 18:23
*** mdunford [] has quit [Ping timeout: 256 seconds]18:31
*** franred [] has quit [Quit: Leaving]18:57
*** lachlanmackenzie [~lachlan@] has quit [Ping timeout: 240 seconds]19:20
*** rdale [] has joined #baserock19:28
*** rdale_ [] has quit [Ping timeout: 272 seconds]19:31
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]20:02
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock20:02
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]20:04
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock20:04
*** rdale_ [] has joined #baserock20:36
*** rdale [] has quit [Ping timeout: 252 seconds]20:37
*** rdale [] has joined #baserock20:43
*** rdale_ [] has quit [Read error: Connection reset by peer]20:43
*** rdale [] has quit [Ping timeout: 246 seconds]20:49
*** rdale [] has joined #baserock20:53
*** rdale_ [] has joined #baserock20:55
*** rdale__ [] has joined #baserock20:57
*** rdale [] has quit [Ping timeout: 272 seconds]20:59
*** rdale [] has joined #baserock21:00
*** rdale_ [] has quit [Ping timeout: 276 seconds]21:00
*** rdale__ [] has quit [Ping timeout: 240 seconds]21:01
*** brlogger [] has joined #baserock21:11
*** brlogger [] has quit [Client Quit]21:11
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]21:39
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock21:47
persiaI keep seeing people say "nowadays, foo is <some relation to systemd>". Please do consider that there are also systems that do not run systemd, especially in the resource-constrained space.23:01
persiaAboriginal linux looks really interesting, but I'd prefer to see bionic fixed, rather than switching to musl23:01
paulsherwoodpersia: is the foo systemd comment related to ostree, or something else?23:02
persiaThe trigger was "nowadays resolv.conf is a symlink to somewhere created by systemd" which came up just before the OSTree dsicussion, but I've seen the same class of assertion for a number of other things.23:06
persiaAnd since we have reference systems in definitions master that do not use systemd, I think it is dangerous to assume that all-things-use-systemd-all-the-time.23:06
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]23:31

Generated by 2.15.3 by Marius Gedminas - find it at!