#archlinux32 | Logs for 2018-05-25

Back
[01:48:39] -!- winircuser-742 has joined #archlinux32
[02:01:02] winircuser-742 is now known as charwhee
[02:13:42] -!- charwhee has quit [Quit: Powered by WinIRC]
[07:20:42] -!- NoobAlice has quit [Quit: Leaving.]
[07:48:10] -!- abaumann has joined #archlinux32
[07:48:10] <buildmaster> Hi abaumann!
[07:54:36] <abaumann> tyzoid: ah, a mighty new build-slave: tyzoid-srv0-bs0 :-)
[08:05:09] -!- deep42thought has joined #archlinux32
[08:05:09] <buildmaster> Hi deep42thought!
[08:05:25] <deep42thought> Good morning, everyone!
[08:14:31] <abaumann> morning. :-)
[08:19:37] -!- eworm has joined #archlinux32
[08:28:45] <abaumann> mmh. I think, my build slave eurobuild3 is somehow borked, I cannot get new assignment: 'come back (shortly) later - I cannot lock build list.
[08:28:48] <abaumann> get-assignment told me:
[08:28:51] <abaumann> come back (shortly) later - I was running already
[08:28:53] <abaumann> ' over and over.
[08:29:15] <deep42thought> this indicates a problem on the build-master
[08:29:21] <deep42thought> "it cannot lock build-list"
[08:29:24] <abaumann> all other slaves work.
[08:29:59] <abaumann> a stale lock file somewhere?
[08:30:07] <deep42thought> they're locked with flock
[08:30:14] <deep42thought> so the file itself does not lock anything
[08:30:24] <abaumann> build_list_lock_file, build-list.lock
[08:32:08] <abaumann> touch ../work/build-list.lock
[08:32:15] <abaumann> the file was not there, so it couldn't be locked.
[08:32:46] <deep42thought> build-list.lock should be on the master
[08:32:47] <deep42thought> not the slave
[08:32:53] <abaumann> ah. true.
[08:32:54] <deep42thought> you do not need to lock anything :-)
[08:32:56] <deep42thought> well
[08:33:03] <deep42thought> it _does_ create a lock file
[08:33:10] <abaumann> now it works.
[08:33:16] <abaumann> * abaumann scratches his head..
[08:33:27] <deep42thought> so the ping-to-master can abort instantly when the build finishes
[08:33:32] <deep42thought> well, temporary error
[08:33:35] <abaumann> ah
[08:33:42] <deep42thought> should I print a process table, when locking fails?
[08:33:57] <deep42thought> filtered, of course
[08:34:10] <abaumann> mmh. maube an idea.
[08:34:12] <abaumann> *maybe
[08:34:28] <abaumann> it was the broken firefox build which left all kind of artifacts on my buildslave.
[08:34:37] <abaumann> I cleaned it now completly.. this may also have helped
[08:35:00] <deep42thought> there was a sanity check running, when you reported the error
[08:35:09] <deep42thought> maybe, this took longer than expected
[08:35:35] <abaumann> yeah. now everything works perfectly again.
[08:35:45] <abaumann> oh. file-5.33.
[08:35:57] <deep42thought> I'll print a list of processes holding a lock on the lock file
[08:36:05] <deep42thought> this seems like a perfect solution :-)
[08:36:08] <abaumann> makes debugging easier anyway.
[08:36:35] <abaumann> ah. file is just a rebuild.
[08:45:09] <abaumann> namcap: downloading "http://k42.ch/mirror/archlinux/community/os/x86_64/jruby-9.2.0.0-1-any.pkg.tar.xz" ... ok.
[08:45:14] <abaumann> This doesn't look right.
[08:45:21] <deep42thought> huh?
[08:45:27] <deep42thought> you're sure, you build "file"?
[08:45:45] <abaumann> So, I'm building a 32-bit package on the slave and then I'm running a namcap on the 64-bit host package.
[08:45:50] <deep42thought> yes
[08:45:52] <abaumann> before
[08:45:53] <deep42thought> that's right
[08:46:00] <abaumann> ah?
[08:46:09] <abaumann> just looks odd. :-)
[08:46:16] <deep42thought> it only sends the diff between namcap on 32 bit and 64 bit
[08:46:23] <abaumann> ah.
[08:46:27] <abaumann> that's clever. :-)
[08:46:42] <deep42thought> the effect is smaller than you might expect
[08:47:06] <abaumann> that's sad. ;-)
[08:47:37] <deep42thought> I *think*, mainly, because the makepkg-namcap is run, when the makedepends are still installed
[08:48:07] <deep42thought> or at least, the dependencies are all "development" dependencies (testing/staging)
[08:48:53] <deep42thought> there were a lot differences, where on x86_64 it complained "you have unsatisfied dependency x" and on i686 it said "dependency x is not needed" (same "x")
[08:51:23] -!- oaken-source has joined #archlinux32
[08:55:34] -!- davor has quit [Ping timeout: 260 seconds]
[08:58:10] -!- oaken-source has quit [Ping timeout: 245 seconds]
[09:00:59] -!- davor has joined #archlinux32
[09:01:24] -!- abaumann has quit [Quit: leaving]
[09:02:08] -!- abaumann has joined #archlinux32
[09:02:09] <buildmaster> Hi abaumann!
[09:28:53] -!- oaken-source has joined #archlinux32
[09:33:54] -!- abaumann has quit [Quit: leaving]
[10:00:08] -!- abaumann has joined #archlinux32
[10:00:08] <buildmaster> Hi abaumann!
[10:14:09] -!- bill-auger has joined #archlinux32
[10:20:23] <deep42thought> abaumann: I found the issue with the sanity-check
[10:20:52] <deep42thought> I changed the meaning of "-w" from "print to webserver" to "wait for lock"
[10:20:57] <deep42thought> ... in the crontab
[10:21:26] <deep42thought> so the locking sanity-checks would accumulate
[10:23:24] <abaumann> ah. good.
[11:00:02] * buildmaster goes insane.
[11:05:01] * buildmaster resumes sanity.
[11:09:45] <abaumann> deep42thought: I'm testing the backport of the binutils patch for glibc compilation, but it takes a while till the whole toolchain is tested again..
[11:11:07] <deep42thought> ok, thanks, take your time
[11:16:52] <deep42thought> the location for the configuration-overrides changed: overrides should now be in conf/common.conf and conf/slave.conf (rather than conf/local.conf)
[11:17:06] <deep42thought> The old location will still be sourced for a while, so nothing should break right now ;-)
[11:18:10] <abaumann> ok. I'll adapt..
[11:22:20] -!- AndrevS has joined #archlinux32
[11:23:13] -!- eduardoeae has quit [Ping timeout: 248 seconds]
[12:19:00] -!- abaumann has quit [Remote host closed the connection]
[12:31:02] -!- oaken-source has quit [Ping timeout: 268 seconds]
[13:33:16] -!- oaken-source has joined #archlinux32
[13:38:22] -!- eduardoeae has joined #archlinux32
[13:54:06] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.stdin.2018-05-25T13:54:05.R02PU0"?.
[14:06:47] <eworm> Can anybody tell me how and where archlinux32's source files are applied?
[14:19:30] * buildmaster resumes sanity.
[14:20:05] -!- abaumann has joined #archlinux32
[14:20:06] <buildmaster> Hi abaumann!
[14:21:26] <abaumann> eworm: quite simple, take the original package, for instance with 'asp export xxx'
[14:21:43] <abaumann> then check out the packages github repo git://github.com
[14:22:04] <abaumann> cat extra/xxx (or core or community/xx)/PKGBUILD >> PKGBUILD
[14:22:18] <abaumann> add i686 to the 'arch' field in PKGBUILD
[14:22:40] <abaumann> in some packages remove lib32- prefixes before packages (for instance in depends, makedepends)
[14:23:22] <abaumann> and yes, we should have a guide somewhere in the wiki for that.
[14:24:54] <eworm> Ah, the code is just appended. Thanks!
[14:30:33] <abaumann> yes, ususally a bunch of bash array manipulations, seds and evals.
[14:35:13] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.stdin.2018-05-25T14:35:13.Z9bkgW"?.
[14:38:19] <abaumann> hi buildmaster
[14:38:28] <abaumann> buildermaster: hello
[14:38:34] <abaumann> buildmaster: hi
[15:13:11] <deep42thought> buildmaster: what's up?
[15:13:12] <buildmaster> up? I'm up for 2 days, 8 hours, 41 minutes, load average: 0.58, 0.42, 0.46 ... and I'm insane :-D
[15:13:16] <tyzoid> deep42thought: Where are the build logs kept on the slaves?
[15:13:22] <deep42thought> nowhere
[15:13:31] <tyzoid> I see.
[15:13:41] <tyzoid> I'm trying to see if I've successfully built a single package yet
[15:14:05] <deep42thought> they're created in work/tmp/tmp.build-packages.*/*.build-log
[15:14:25] <deep42thought> in case you want to have a look when the build is still running
[15:14:37] <tyzoid> `tyzoid-srv0-bs0 tyzoid firefox 2018-05-25 13:12:59 2018-05-24 20:17:48 22949 building`
[15:14:41] <deep42thought> I think, the ssh log has a filter, which should accomplish what you need
[15:14:41] <tyzoid> from the slave status
[15:14:49] <tyzoid> so looks like it is building, just very slowly
[15:15:13] <deep42thought> or it got stuck
[15:15:19] <tyzoid> that too
[15:15:36] <tyzoid> Is there a verbose option I can pass to the script?
[15:15:43] <deep42thought> no :-/
[15:15:47] <tyzoid> I think it'd be really nice to have that available via journalctl
[15:15:56] <deep42thought> "tail --follow builder/work/tmp.*/*.build-log"
[15:16:43] <tyzoid> I mean, the machine does have 100% memory usage
[15:16:48] <tyzoid> so it's kinda stalled on stuf
[15:16:57] <tyzoid> stuff*, but seems to be progressing slowly
[15:16:59] <deep42thought> and is quite slow to respond :-/
[15:17:14] <tyzoid> yeah, I'm going to abort, and give it more ram
[15:18:39] <deep42thought> wait a sec
[15:18:53] <deep42thought> ah, too late
[15:19:38] -!- isacdaavid has joined #archlinux32
[15:20:29] <tyzoid> yeah, sorry
[15:20:52] <deep42thought> np
[15:22:10] -!- isacdaavid has quit [Read error: Connection reset by peer]
[15:22:26] <tyzoid> total used free shared buff/cache available
[15:22:26] <tyzoid> Mem: 8192 21 8166 252 4 8170
[15:22:27] <tyzoid> Swap: 8191 0 8191
[15:22:32] <tyzoid> so much room for activities now
[15:23:09] <deep42thought> btw: https://github.com
[15:23:11] <phrik> Title:builder/build-packages at master · archlinux32/builder · GitHub (at github.com)
[15:24:33] * buildmaster resumes sanity.
[15:25:20] <tyzoid> tee ftw
[15:25:48] <deep42thought> swallows the exit code
[15:25:56] <tyzoid> ah, right
[15:27:45] <deep42thought> ping-to-master could read the conten asynchronous, but that script is a) new and b) meant for something else
[15:30:07] <tyzoid> hey deep42thought: can you join #archlinux-ports for a sec? I want to test something
[15:31:06] <tyzoid> Thanks. I set up autovoice, so people working on ports get bumped to +v
[15:31:25] <deep42thought> yeah, I read it
[15:32:15] <tyzoid> deep42thought: Looks like it's taking a second crack at firefox
[15:32:18] <tyzoid> let's see how this goes
[15:32:25] <deep42thought> yeah
[15:32:36] <deep42thought> the buildmaster assumes, you're still building it and just forgot
[15:32:46] <deep42thought> so you'll get it deterministically each time now
[15:32:56] <deep42thought> until you a) report back or b) hit the timeout
[15:33:04] <tyzoid> what's the timeout?
[15:33:19] <deep42thought> 1h + some jitter for the cronjob
[15:33:24] <deep42thought> I think
[15:33:27] <tyzoid> ah, so timeout on no messages
[15:33:35] <tyzoid> not total build timeout
[15:33:35] <deep42thought> timeout on no ping, right
[15:33:46] <deep42thought> exactly
[15:34:03] <deep42thought> that's, what ping-to-master is for
[15:47:13] <tyzoid> man, that's a huge repo
[15:47:29] <tyzoid> deep42thought: You have your buildmaster configured to stress test every new build slave?
[15:47:33] <tyzoid> Or just mine?
[15:47:56] <deep42thought> all of them
[15:48:01] <tyzoid> lol
[15:48:11] <deep42thought> there is a hidden option to skip the stress test, but I'm not saying which
[15:52:17] -!- isacdaavid has joined #archlinux32
[15:56:07] <deep42thought> tyzoid: if you prefer to build a specific package, there is "-p" option for build-packages - but it won't work, when the buildmaster thinks, you're building something else
[15:56:27] <tyzoid> Yup, doesn't really matter in my case
[15:56:45] <deep42thought> we could think of a "no-huge-memory-slave" option, though
[15:56:46] <tyzoid> I can wait, just want to see something succeed to make sure everything works as expected
[15:57:12] <tyzoid> well, I bumped the container to 8gb memory, so... that solves that issue
[15:57:24] <tyzoid> but yeah, for other vms/containers would be nice
[15:57:38] <tyzoid> that way, the big ones can build the big stuff
[15:58:19] <deep42thought> problem is, we'd need to identify all big-memory packages
[15:58:35] <tyzoid> report memory usage to buildmaster?
[15:58:40] <deep42thought> :-/
[15:59:26] <deep42thought> we'd need to store this permanently somehow
[15:59:49] <deep42thought> archlinuxarm sets some flags in their PKGBUILDs iirc
[16:03:11] <abaumann> yes, plugbuild has such options
[16:03:22] <abaumann> and the slave reports back some cpu and memory info IIRC
[16:03:39] <tyzoid> putting reports in mysql would work
[16:04:14] <tyzoid> like status lines "load avg, mem usage, etc"
[16:04:32] <deep42thought> this would be attached to what identifier?
[16:04:34] <deep42thought> the pkgbase?
[16:04:50] <deep42thought> ah, no
[16:04:54] <deep42thought> to the build_assignment
[16:04:56] <tyzoid> joint key of (build-slave-name, package)
[16:05:07] <tyzoid> with a timestamp
[16:05:12] <tyzoid> in the key as well
[16:05:30] <deep42thought> I'd rather make that a column in build_assignments
[16:05:42] <deep42thought> each build assignment is built only once successfully
[16:06:46] <tyzoid> sure
[16:06:59] <deep42thought> I'll add a todo :-)
[16:07:02] <tyzoid> but then we'd probably want to report mem stats after the build
[16:07:09] <tyzoid> like peak memory / avg memory
[16:07:21] <deep42thought> yes
[16:07:33] <deep42thought> this could be done similar to the report of other stuff
[16:07:41] <deep42thought> e.g. required libraries, namcap log, ...
[16:08:11] <tyzoid> lol, yeah. Now using 100% cpu xD
[16:08:13] <deep42thought> btw: there is now a per-build-slave-filtered log linked from https://packages.archlinux32.org
[16:08:14] <phrik> Title:list of build slaves (at packages.archlinux32.org)
[16:08:20] <tyzoid> before it was all stalled on hdd
[16:08:45] <tyzoid> nice
[16:08:58] <abaumann> deep42thought: my chromium build is running since ages. You just mentioned this 1 hour 'guilloutine'..
[16:09:20] <tyzoid> abaumann: the 1 hour timeout is if the master hasn't had a ping from the slave
[16:09:23] <tyzoid> not total build time
[16:09:27] <deep42thought> abaumann: as long as you ping the master, it assumes, there is progress
[16:09:31] <abaumann> ah.
[16:09:35] <abaumann> ok then.
[16:09:41] <deep42thought> 1h is too few for some packages
[16:09:51] <deep42thought> abaumann: is there still some progress?
[16:09:58] <deep42thought> yes, there is
[16:10:00] <abaumann> oh yes.
[16:10:11] <abaumann> [15756/32880] CXX obj/media/formats/formats/webm_crypto_helpers.o
[16:10:13] <deep42thought> I just saw the number of logged lines increasing :-)
[16:10:24] <abaumann> that's actually a good feature. :-)
[16:10:36] <deep42thought> number of trials would also be a nice info
[16:10:54] <deep42thought> but I did some other work in the last time (mainly preparing to enable a multi-arches build master)
[16:11:27] <tyzoid> looks like firefox isn't actually using more than ~3.5-4gb ram
[16:11:43] <deep42thought> but 8 to 10GB hdd
[16:11:58] <deep42thought> (or was it thunderbird?)
[16:12:02] <deep42thought> I keep forgetting
[16:12:08] <tyzoid> idk, I put 64gb for the disk
[16:12:14] <deep42thought> that's plenty
[16:12:16] <tyzoid> I'd rather just not worry about it
[16:12:20] <deep42thought> :-)
[16:12:34] <tyzoid> I've got close to 1tb free space
[16:25:57] -!- Alina-malina has quit [Ping timeout: 240 seconds]
[16:39:23] -!- deep42thought has quit [Quit: Leaving.]
[16:41:27] <tyzoid> huh, looks like the rust compilers are memory hungry
[16:41:29] <tyzoid> 5.5gb used
[16:52:28] -!- abaumann has quit [Remote host closed the connection]
[16:59:51] -!- abaumann has joined #archlinux32
[16:59:52] <buildmaster> Hi abaumann!
[17:00:00] -!- eworm has quit [Quit: Leaving.]
[17:00:53] <tyzoid> wb
[17:01:22] -!- abaumann has quit [Remote host closed the connection]
[17:08:16] <buildmaster> girls, my database is dirty again ...
[17:17:10] -!- isacdaavid has quit [Read error: Connection reset by peer]
[17:22:44] -!- eduardoeae_ has joined #archlinux32
[17:42:37] <tyzoid> hmm
[17:42:37] <tyzoid> 89:43.63 ../../build/unix/gold/ld: fatal error: libxul.so: mmap: failed to allocate 1704008984 bytes for output file: Cannot allocate memory
[17:49:35] -!- AndrevS has quit [Quit: umount /dev/irc]
[17:58:56] -!- deep42thought has joined #archlinux32
[17:58:57] <buildmaster> Hi deep42thought!
[18:04:27] <tyzoid> wb
[18:04:35] <tyzoid> so, it seems 8gigs isn't enough for firefox
[18:04:45] <deep42thought> :-/
[18:05:03] <tyzoid> It failed after attempting to mmap 1.5gib
[18:05:53] * buildmaster resumes sanity.
[18:12:47] <eschwartz> I've built on an HDD with 2gb ram quite often
[18:13:18] <eschwartz> Took forever but it did work
[18:13:59] <deep42thought> eschwartz: on how many cores?
[18:15:59] <eschwartz> Core2 duo
[18:16:23] <deep42thought> that's 2 physical cores with 2 pipes each?
[18:16:29] <eschwartz> I did it on a single-core Celeron even, but that laptop was really old
[18:16:41] <deep42thought> I think, less cores is better here
[18:16:47] <deep42thought> since less stuff will run in parallel
[18:16:54] <deep42thought> and tendentially use less ram
[18:17:06] <eschwartz> I think it was one physical core with two virtual cores
[18:17:29] <eschwartz> MAKEFLAGS=-j1 ?
[18:17:30] <deep42thought> tyzoid: you may want to try reducing the number of cpu cores :-D
[18:18:45] <eschwartz> I haven't used the Celeron in forever
[18:19:08] <eschwartz> The core2 I used as late as November but the screen cracked
[18:21:16] <tyzoid> deep42thought: is there a way we can tell the current job that? Or would you just like me to remove two cores from the container?
[18:21:38] <deep42thought> you can hot-edit the PKGBUILD ... that's what I usually do
[18:21:46] <deep42thought> but the clean way would be to remove the cores :-)
[18:23:33] <tyzoid> so... reboot?
[18:23:36] <tyzoid> or did you want to do something?
[18:23:45] <deep42thought> reboot :-)
[18:23:57] <eschwartz> Add MAKEFLAGS to arch32-packages with the note "#dont clobber our build machine"
[18:24:20] <deep42thought> we could default to -j1
[18:24:23] <deep42thought> ;-P
[18:24:35] <tyzoid> It's really only the packages that have a large amount of memory used per worker process
[18:24:47] <eschwartz> Would you want to do that, if it means regular packages are slow?
[18:24:48] <tyzoid> otherwise, I'm fine letting the build system use four of my 8 threads.
[18:25:04] <eschwartz> Firefox is just special
[18:25:16] <eschwartz> And deserves special flags
[18:27:01] <deep42thought> you're right, firefox might deserve its own flags for that
[18:28:17] <buildmaster> girls, my database is dirty again ...
[18:28:32] <deep42thought> never change a running system ...
[18:32:19] -!- Alina-malina has joined #archlinux32
[18:32:58] * buildmaster resumes sanity.
[18:34:17] <buildmaster> girls, my database is dirty again ...
[18:39:14] * buildmaster resumes sanity.
[18:50:00] -!- Velines has joined #archlinux32
[18:50:17] <buildmaster> girls, my database is dirty again ...
[18:50:38] <Velines> hi everyone
[18:50:44] <deep42thought> Hi Velines
[18:52:16] <buildmaster> girls, my database is dirty again ...
[18:52:22] <deep42thought> I know, I know
[18:53:25] <Velines> I'm planning to upgrade an old arch installation from November 2016 to archlinux32. It's just a very minimal server installation with a small number of packages, but is there any obvious show blocker I can expect?
[18:53:51] <deep42thought> https://packages.archlinux32.org
[18:53:52] <phrik> Title:Blacklisted packages (at packages.archlinux32.org)
[18:54:08] <deep42thought> anything on that list + dependent packages
[18:54:50] <Velines> thanks :)
[18:55:06] -!- Alina-malina has quit [Ping timeout: 252 seconds]
[18:55:14] <Velines> looks fine though, I'm pretty sure none of that is installed
[18:55:34] <deep42thought> yeah, mostly desktop stuff :-)
[18:57:54] <Velines> yep. my box is just a boring samba+dlna machine ;)
[18:58:32] <tyzoid> Velines: You may want to update through the archive, as you may run into issues
[18:58:40] <Velines> ok, I'll try my luck when the backup is finished :)
[18:59:04] <deep42thought> tyzoid: what issues? there are no issues :-D
[18:59:21] <tyzoid> Velines: Upstream currently has i686 packages in their archive, but if they remove it by the time you do the upgrade, we have a copy of their 2017 archive: https://archive.archlinux32.org
[18:59:22] <phrik> Title:Index of / (at archive.archlinux32.org)
[19:00:16] <Velines> that sounds like a good idea. Upgrade to 2017 archive first, then to current arch32?
[19:00:39] <deep42thought> I'm pretty convinced, 2017 is not too far in the past
[19:01:11] <tyzoid> Velines: Yup. You can try going direct, but you may run into some issues.
[19:01:19] <tyzoid> deep42thought: November 2016 is what they said
[19:01:28] <tyzoid> 1.5 years is a big jump for arch
[19:01:45] * deep42thought cannot read
[19:01:57] <deep42thought> and I cannot memorize numbers
[19:02:07] <Velines> :)
[19:02:07] * tyzoid helps keep deep42thought sane
[19:02:23] <deep42thought> 3.14159265358979323846264338327950
[19:03:50] <Velines> I'll try via the archive, but I expect nothing really bad will happen. At most I'd have to do a clean re-install, but I can save all my config files beforehand
[19:06:27] <deep42thought> reinstall is usually not necessary
[19:06:42] <eschwartz> Velines: consider all news posts since the last time you updated
[19:07:01] <eschwartz> There are multiple "manual intervention is needed" things
[19:07:24] <tyzoid> Also the xorgproto issue
[19:07:33] <tyzoid> which isn't in the news, for some reason
[19:07:35] <deep42thought> ah right, I forgot to mention: you should watch archlinux' news even when you run archlinux32
[19:08:50] <Velines> okay, thanks for the heads up. I run arch on my work machine, so I'm usually quite up to date on manual interventions
[19:11:42] <eschwartz> Js52 is needed by polkit, but you'll be skipping the bad version anyway :p
[19:12:17] <tyzoid> xorgproto is the undocumented roadblock that you'll only hit by doing a longer upgrade cycle.
[19:12:49] <eschwartz> Assuming you've got any xorg things installed on a server
[19:13:13] <tyzoid> I mean, I've done that before
[19:13:41] <tyzoid> having a remote vnc workstation with tons of memory is kinda nice
[19:13:53] <tyzoid> esp. if it's got a fast connection and tons of storage to boot
[19:14:04] <tyzoid> though if it's i686, probably not
[19:14:59] <eschwartz> Velines yep. my box is just a boring samba+dlna machine ;)
[19:15:18] <eschwartz> Quasseldroid copy paste sucks...
[19:16:33] <Velines> its only a dual core with 2gb RAM, so no X anyway ;)
[19:17:53] <eschwartz> My main laptop from December was too, and it had X...
[19:25:08] <deep42thought> oh, I just noticed, my server still seeds the 2017.09.03 isos :-/
[19:47:41] -!- NoobAlice has joined #archlinux32
[19:48:41] -!- abaumann has joined #archlinux32
[19:48:41] <buildmaster> Hi abaumann!
[19:48:50] <abaumann> tyzoid: firefox, exactly the error I got.. *sigh*
[19:49:18] <tyzoid> deep42thought: Yeah, don't remove those yet
[19:49:25] <tyzoid> they're not included in the archive
[19:49:32] <deep42thought> I removed the torrent
[19:49:35] <tyzoid> ah
[19:49:38] <tyzoid> cool
[19:49:42] <deep42thought> isos are still there
[19:49:49] <deep42thought> where should I push them?
[19:58:14] <tyzoid> uh, I'll just make a manual copy on the archive
[19:58:16] <tyzoid> only'll take a sec
[20:00:03] <tyzoid> deep42thought: Alright, they're copying
[20:00:03] <tyzoid> https://archive.archlinux32.org
[20:00:05] <phrik> Title:Index of /isos (at archive.archlinux32.org)
[20:00:17] <tyzoid> iso replacement is slow enough that I can manually maintain the archive for it
[20:00:22] <tyzoid> plus I do the isos anyway xD
[20:00:26] <deep42thought> :-)
[20:00:36] <deep42thought> we should create an iso a day
[20:00:39] <tyzoid> so it'll be done in a few mins, I'll let you know when it's safe to delete
[20:00:42] <tyzoid> lol xD
[20:01:17] <tyzoid> I'd do that if I can get my build scripts rewritten for running in a container
[20:01:27] <tyzoid> right now requires vagrant, which can't run
[20:08:06] <tyzoid> deep42thought: Everything copied, you can remove what you think is appropriate.
[20:08:16] <deep42thought> the hard links :-)
[20:08:34] <tyzoid> I'd probably say we should keep back to february, since IIRC that was the last one that ran on qemu
[20:08:53] <deep42thought> yeah
[20:11:23] <abaumann> jippie: my slave crashed with rebooting on chromium
[20:12:36] <deep42thought> :-/
[20:18:56] -!- Alina-malina has joined #archlinux32
[20:32:40] -!- abaumann has quit [Quit: leaving]
[20:40:28] -!- Velines has quit [Quit: Page closed]
[22:22:30] <NoobAlice> Hi there! I have a few probably-dumb questions and only a couple of months experience with Arch-Official on my daily-use laptop. I was given a tiny netbook, an HP 2133 Mini-Note, that has been sitting in a garage for two to ten years but boots fine and seems to be in great condition.
[22:22:43] <NoobAlice> 1) It has a 32-bit Windows 7 OS, so am I correct in assuming that it should be able to run Arch32?
[22:22:53] <NoobAlice> 2) Is installation of Arch32 as simple as downloading an Arch32 ISO and following the Arch-Official wiki installation guide? Or are there changes in the process for Arch32?
[22:22:57] <NoobAlice> 3) Just checking: If I download the dual-bootable ISO, will that be appropriate as both installation media for Arch32 and recovery media for Arch32 and Arch-Official?
[22:23:11] <deep42thought> 1) yes
[22:23:33] <deep42thought> 2) yes (the mirror list will be different, but that's already reflected in the iso)
[22:24:01] <deep42thought> 3) exactly
[22:24:12] <NoobAlice> deep42thought: Oh beautiful! Thank you so much!
[22:24:54] <NoobAlice> I wanted to double-check before starting anything, after I spent a couple hours trying to figure out which version of Arch-ARM I needed... That was a moronic mistake.
[22:25:40] <deep42thought> speaking of archlinuxarm: I still haven't got my raspi3 booting it :-(
[22:27:04] <NoobAlice> deep42thought: That stinks. The ARM stuff looks messy...
[22:28:10] <deep42thought> hmm, probably still some mistake on my end
[22:30:47] <NoobAlice> I'm not sure how they keep track of things, with every device (board? processor?) having a different set of installation instructions and tiny wiki.
[22:32:46] <deep42thought> there are different instructions for different devices?
[22:32:59] <NoobAlice> Yep!
[22:33:14] <NoobAlice> Raspi 1: https://archlinuxarm.org
[22:33:15] <phrik> Title:Raspberry Pi | Arch Linux ARM (at archlinuxarm.org)
[22:33:25] <deep42thought> ah you're talking about alarm :-)
[22:33:42] <NoobAlice> ...yes? You aren't?
[22:33:55] <NoobAlice> Is there another Arch ARM???
[22:34:10] <deep42thought> no, I thought, you were referring to the x86_64 installation guide
[22:34:55] <NoobAlice> Oh! Nope. The standard Arch Wiki installation makes sense to me.
[22:36:18] * buildmaster goes insane.
[22:36:33] <NoobAlice> I was just referring to you not getting raspi3 going.
[22:36:54] -!- oaken-source has quit [Ping timeout: 260 seconds]
[22:37:09] * NoobAlice gives buildmaster a sedative
[22:38:41] <buildmaster> thanks, NoobAlice
[22:38:41] <buildmaster> np
[22:38:54] <deep42thought> ah, I didn't foresee the "np" ...
[22:42:41] * buildmaster resumes sanity.
[22:52:57] <tyzoid> lol
[22:58:56] <tyzoid> deep42thought: same error on firefox again
[22:58:57] <tyzoid> 181:32.77 ../../build/unix/gold/ld: fatal error: libxul.so: mmap: failed to allocate 1704008984 bytes for output file: Cannotallocate memory
[22:59:28] <deep42thought> hmm, I don't know
[22:59:34] <deep42thought> :'-(
[23:00:21] <tyzoid> how would I change the pkgbuild to use -j2 instead of what it's got?
[23:00:47] <tyzoid> or if it's easier, you've got access to that box if you want to do that quickly
[23:01:03] <deep42thought> nah, I'll only break stuff ;-)
[23:01:08] <deep42thought> I'm a little overdue
[23:06:50] <tyzoid> ok. I'll stop the build-slave service for now. I'l start it up later and try to get a different package
[23:07:07] <deep42thought> i can manually remove the assignment if you like
[23:07:25] <tyzoid> Sure, if that works for you. I was gonna wait for the 1hr timeout
[23:07:51] <deep42thought> ok, done
[23:13:22] <tyzoid> get-assignment told me:
[23:13:22] <tyzoid> come back (shortly) later - I was running already
[23:13:29] <tyzoid> deep42thought: ^ ?
[23:13:45] <deep42thought> exit code = 1
[23:13:47] <tyzoid> been doing this for about 5 mins. This normal?
[23:13:56] <deep42thought> what's the line above that?
[23:14:11] <tyzoid> come back (shortly) later - I cannot lock build list.
[23:14:36] <deep42thought> it should also print a process list right above that
[23:15:08] <deep42thought> rechenknecht tries to upload qemu
[23:15:11] <tyzoid> that's all I'm getting back
[23:19:08] <deep42thought> ah, stdout gets swallowed
[23:21:07] <deep42thought> now you should see more
[23:21:23] <tyzoid> stop and pull, I assume?
[23:21:30] <deep42thought> no
[23:21:34] <deep42thought> it was a change on the buildmaster
[23:21:38] <deep42thought> it outputs now to stderr
[23:21:46] <tyzoid> ah
[23:21:46] <tyzoid> COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
[23:21:47] <tyzoid> get-assignment 12000 slave 9w REG 254,1 0 103714 /home/slave/builder/work/build-list.lock
[23:21:47] <tyzoid> return-assignme 22205 slave 9w REG 254,1 0 103714 /home/slave/builder/work/build-list.lock
[23:21:48] <tyzoid> tar 22336 slave 9w REG 254,1 0 103714 /home/slave/builder/work/build-list.lock
[23:21:48] <tyzoid> FYI: I am 12000.
[23:21:49] <tyzoid> come back (shortly) later - I cannot lock build list.
[23:21:49] <tyzoid> get-assignment told me:
[23:21:50] <tyzoid> come back (shortly) later - I was running already
[23:21:55] <deep42thought> there you go
[23:22:21] <tyzoid> I see, so it's just waiting on those to finish
[23:22:52] <deep42thought> yes
[23:23:00] <deep42thought> because this will influence the build-list
[23:23:04] <deep42thought> and so, it locks it
[23:24:49] <deep42thought> maybe, something is broken
[23:25:04] <deep42thought> but I'll need to get some sleep before looking into this
[23:25:29] <deep42thought> ah, no, now something else is being uploaded
[23:26:14] <deep42thought> but qemu is not in [staging], so something about the upload went wrong :-/
[23:26:17] <tyzoid> I see, so this is normal (mostly)
[23:26:21] <tyzoid> or that
[23:26:23] <tyzoid> lol
[23:26:54] <deep42thought> but I really need to go to bed now
[23:26:57] <deep42thought> cu tomorrow
[23:27:06] -!- deep42thought has quit [Quit: Leaving.]
[23:27:26] <tyzoid> lol, worked right as you left