ekleog changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html https://r13y.com | 18.09 release managers: vcunat and samueldr | https://logs.nix.samueldr.com/nixos-dev
<samueldr> Tue Dec 31 19:05:31 EST 2097
<samueldr> guess that laptop without battery needs ntp
<gchristensen> hey at least 2038 wasn't a problem
<gchristensen> everybody should read https://discourse.nixos.org/t/thoughts-from-a-new-nixos-user/2135 and feel really stinkin' proud of themselves
<ekleog> ^
<gchristensen> lol, someone set the nixos-18.09-small jobset to be invisible
pie___ has joined #nixos-dev
drakonis has quit [Quit: WeeChat 2.3]
pie__ has quit [Ping timeout: 245 seconds]
lassulus_ has joined #nixos-dev
lassulus has quit [Ping timeout: 246 seconds]
lassulus_ is now known as lassulus
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 268 seconds]
orivej_ has joined #nixos-dev
FRidh has joined #nixos-dev
phreedom has quit [Remote host closed the connection]
phreedom has joined #nixos-dev
{^_^} has quit [Excess Flood]
{^_^} has joined #nixos-dev
<domenkozar> I'm working on making cachix easier to use on nixos
<domenkozar> would this make life easier?
init_6 has joined #nixos-dev
drakonis has joined #nixos-dev
jtojnar has quit [Read error: Connection reset by peer]
jtojnar has joined #nixos-dev
garbas has joined #nixos-dev
init_6 has quit [Ping timeout: 244 seconds]
orivej_ has quit [Ping timeout: 246 seconds]
FRidh has quit [Quit: Konversation terminated!]
drakonis has quit [Quit: WeeChat 2.3]
tilpner has quit [Quit: WeeChat 2.3]
tilpner has joined #nixos-dev
orivej has joined #nixos-dev
elvishjerricco has quit []
elvishjerricco has joined #nixos-dev
Jackneill has quit [Quit: Leaving]
Jackneill has joined #nixos-dev
<samueldr> since I've been side-tracked and have it on my mind: are there any issues, mostly thinking capacity, in running a limited amount of tests on all mainline kernels, including -rc* on the nixos.org hydra?
<samueldr> maybe having the non-testing kernels fail the channel advance in case of failure (not sure about testing)
yl has joined #nixos-dev
<gchristensen> something for the PM is documentation on channel halts
<gchristensen> FYI: I'm canceling all builds except for release-18.09 to get bumps out, then will restart the other builds
<gchristensen> starting with nixos-unstable-small
<gchristensen> yikes!
<gchristensen> "Sadly, our bugzilla is rarely used" -- Sasha w.r.t. LKML and their bug tracker
<samueldr> :|
<gchristensen> really distressing
<gchristensen> oh well
<gchristensen> aszlig, something to know for the future: don't open bug reports to the linux kernel, send a mail instead
<aszlig> gchristensen: i didn't open a bug report
<gchristensen> yeah
<gchristensen> I know
<gchristensen> just thinking through it all
<gchristensen> at a minimum, we shoul be mailing the ML what we learn, not just posting to a bug report
<aszlig> btw. did anyone proceed on the overlayfs regression in the meantime?
<gchristensen> I doubt it
<gchristensen> not that I have heard
<samueldr> aszlig: didn't notify you, looked at the ML thread, and was thoroughly out of my element to continue with what was asked
<aszlig> okay, well then i'll do that this weekend
garbas has quit [Quit: WeeChat 2.3]
garbas has joined #nixos-dev
<clever> gchristensen: have you looked at ` --option build-repeat 1` for the instructions on https://r13y.com/?
<gchristensen> yeah
<gchristensen> it doesn't really work, iirc, if it substitutes
<clever> yeah, so it would be limited to things your doing locally
<gchristensen> so has to deal with weird edge cases if the path is already in the nix store
<clever> you could also just blanket enable that in nix.conf
<gchristensen> hehe
<gchristensen> yeah, pretty sure that isn't a thing a casual contributor would really want to do
<clever> building '/nix/store/0qmnc5c4i77s0qsb061nrsxa49av6n1f-daedalus-installer-0.1.0.0.drv' (round 1/2)...
<clever> building '/nix/store/0qmnc5c4i77s0qsb061nrsxa49av6n1f-daedalus-installer-0.1.0.0.drv' (round 2/2)...
<clever> output '/nix/store/ksyz564hhwipnb3108xk1dcdfzidjswz-daedalus-installer-0.1.0.0' of '/nix/store/0qmnc5c4i77s0qsb061nrsxa49av6n1f-daedalus-installer-0.1.0.0.drv' differs from previous round
<clever> gchristensen: dang, i dont see the old copy....
<gchristensen> -K :)
<clever> /* If this is the first round of several, then move the output out
<clever> of the way. */
<clever> if (nrRounds > 1 && curRound == 1 && curRound < nrRounds && keepPreviousRound) {
<clever> bool keepPreviousRound = settings.keepFailed || runDiffHook;
<clever> gchristensen: yep, or you can setup an automatic diff hook
<gchristensen> I know
<gchristensen> :P
<gchristensen> but automaticdiff hook doesn't work with --check
<clever> yeah, now it did leave a .check behind, same as --check
<gchristensen> fpletz, globin: I have someone who would like to help push on networkd -- you interested in chatting?
garbas has quit [Quit: WeeChat 2.3]
garbas has joined #nixos-dev
<gchristensen> sometimes when running r13y's checker, I'll set my laptop out on the back porch to help keep this laptop cool.
<cransom> quicker too. combat the the thermal throttling.
<gchristensen> there you go!
init_6 has joined #nixos-dev
init_6 has quit []
<matthewbauer[m]> Could someone on hydra restart aborted builds in nixpkgs-18.09-darwin: https://hydra.nixos.org/eval/1505176
<matthewbauer[m]> They should build fine but they somehow got cancelled causing them to show as failures
<samueldr> matthewbauer[m]: probably this https://logs.nix.samueldr.com/nixos-dev/2019-02-15#1962378;
<matthewbauer[m]> oh ok
<matthewbauer[m]> This issue is a couple of weeks old though - they may have been cancelled for another reason
<samueldr> (to be fair I didn't follow the link, maybe I should have?)
<gchristensen> I'll restart jobs in a bit, matthewbauer[m]
<samueldr> 2d ago, probably unrelated gchristensen?
<gchristensen> there were many aborted darwin jobs before I aborted more
<matthewbauer[m]> Ok thanks!
<gchristensen> ok, 18.09 finished building, time to bump channels and restart like 50k obs :P
drakonis has joined #nixos-dev
orivej has quit [Ping timeout: 258 seconds]
orivej has joined #nixos-dev
averell has quit [Quit: .]
<gchristensen> new report up https://r13y.com/ -- continued syslinux trouble, plus the docs got an old PR merged adding back a section w/out an ID (fixed)
<samueldr> wild, did sylinux break reproducibility after the debian patches landed upstream?
averell has joined #nixos-dev
<gchristensen> not sure
<gchristensen> b1bda29f5ca5063579b0998e1a7f0b4281b5c9d7 included your patches right?
<gchristensen> yea
<gchristensen> it'll be nice having staging merge
<samueldr> yeah, a 2019 syslinux version, first thing I checked
* gchristensen reads about movq to see how that makes sense
<samueldr> 0x15: 00010101 / 0xb: 00001011 so not a bitflip or something similar
<samueldr> or at least, not a direct one
<samueldr> though 10101 101_1 dunno what operation could cause this unexpectedly
<gchristensen> movq moves a quadword from arg2 to arg1
<gchristensen> I wonder if 0x15 and 0x0b are similar somehow
<gchristensen> as registers
* ekleog wonders how we end up with 3 `perl-5.28.1` in the iso closure
<gchristensen> bootstrapping
<gchristensen> its in the build path, not actually in the iso
<ekleog> oh ok, thought that was the runtime closure
<clever> gchristensen: ive heard that the mov family of opcodes, are turing complete
<ekleog> for that autogen error, it looks almost like an overclocked processor's failure mode
<ekleog> oops
<gchristensen> https://twitter.com/alicegoldfuss/status/1096551553182253056 "it's moving different values, not to different registers. the $ means "this literal value" so for one it's moving 0x15 (21 in decimal) and in another it's moving $0xb (11 in decimal) and both are moving to memory location 0x5c8 + whatever is in register %rbp."
<clever> ekleog: yep, thats the one
<sorear> what would help here, symbols
<clever> gchristensen: what is the github repo for our syslinux?
<clever> oh wait, thats autogen
<clever> agen5/autogen.c:inner_main(void * closure, int argc, char ** argv)
<sorear> that's a pretty short function and the offsets in the disassembly are very large
<sorear> there are two ways this can go. one, massive inlining. two, you're looking at a stripped binary and objdump is just blaming the last public symbol
<clever> its just before a series of testb's and a call to getenv
<gchristensen> back in 1.5h
<samueldr> clever: I know you don't need it now, but the repo for syslinux is the upstream one, https://repo.or.cz/syslinux.git
<clever> i think inner_main got inlined into main
<clever> yeah, i would need to build autogen with debug on to get more info
<clever> calls to optionProcess also got inlined into inner main
callahad has joined #nixos-dev