<clever> MESS:00:00:09.498347:0: uart: Baud rate change done...
<clever> MESS:00:00:09.501779:0: uart: Baud rate
<clever> samueldr: heh, it cut itself off!
<samueldr> reset from early print console to "actual" console?
<clever> samueldr: that was the start.elf file, cutting itself off
<clever> it changed the uart config, in the middle of printing debug to the uart
chiefgoat has joined #nixos-aarch64
<clever> samueldr: let me try with the stock kernel too
chiefgoat has quit [Client Quit]
<clever> [ 1.927871] VFS: Cannot open root device "mmcblk0p2" or unknown-block(0,0): error -6
chiefgoat has joined #nixos-aarch64
<clever> its ignoring the initrd...
<clever> because i didnt copy it!
<clever> samueldr: that boots
<samueldr> so it's not busybox ash being screwey then, I guess?
<clever> i guess
<clever> but i dont see how my kernel could have broken ash so badly
<clever> nor how the firmware could mess with such a specific task
chiefgoat has quit [Quit: ZNC 1.7.5 - https://znc.in]
<samueldr> my question stands: could it be the kernel code that somehow "resets" the UART in a bad way?
<samueldr> asssuming everything before is from the earlyprintk console
chiefgoat has joined #nixos-aarch64
<clever> samueldr: nope, because remote echo still works, as does a lot of other debug
<clever> samueldr: let me pastebin more...
<samueldr> remote echo and other debug inside the kernel or outside?
<samueldr> and eh, I'm just spitballing here, you're clearly the one with the better knowledge :)
<clever> i know more of the hardware, but am out of ideas on what else to check
<clever> samueldr: lines 5-117 come from the gpu side of things
<clever> samueldr: 118-179 is the arm chainloader (a custom blob on the arm side, does u-boot like things)
<clever> 180-486 is obviously linux
<clever> 487-489 is the gpu again, it has a timer setup to fire ~100 seconds after boot has begun
<clever> [ 130.545880] random: crng init done
<clever> and linux then chimes in with this
<clever> samueldr: so the uart hardware is still working, and linux is still capable of printing to it...
<samueldr> right
<samueldr> what if you prefixed a custom "set -- foo bar" at the top of the script?
<samueldr> oh!
<clever> 3 echo before
<clever> 5 echo after
<clever> 4 set -- a b c
<samueldr> well, maybe not the /tmp mkdir, but at least the log redirection
<samueldr> this happens RIGHT after a read -r line
<samueldr> oh no
<samueldr> don't mind that, there are other read -r line working fine
<clever> before
<clever> after
<clever> <<< NixOS Stage 1 >>>
<clever> samueldr: thats it, and i had turned off `set -x`, let me turn it back on...
<samueldr> okay, so not set itself
orivej has quit [Ping timeout: 258 seconds]
<clever> samueldr: `set -x` isnt catching everything, i cant see the case block doing things
<clever> + IFS='='
<clever> + echo root /dev/mmcblk0p2
<clever> + set -- root /dev/mmcblk0p2
<samueldr> at that point I would be stumped, and start asking on IRC :)
<clever> already hit that step :P
orivej has joined #nixos-aarch64
<clever> 1 #! @shell@ -v
<clever> samueldr: this adds even more debug
<clever> samueldr: though nothing of use
<clever> samueldr: i removed the root= from the cmdline, and now i'm seeing some systemd action...
<clever> + systemd-udevd --daemon
<clever> Starting version 243
<clever> but it hangs there
<clever> 1811 if (arg_daemonize) {
<clever> >>1814 log_info("Starting version " GIT_VERSION);
worldofpeace has quit [Ping timeout: 264 seconds]
ryantrinkle has quit [Ping timeout: 265 seconds]
<samueldr> hmmm
<samueldr> IIRC it failed to me hanging on that when there was no proper console or a "wrong" console on the cli on one device
<samueldr> clever: though not sure here that would apply
<clever> + strace systemd-udevd --daemon
<clever> strace: error while loading shared libraries: libgcc_s.so.1: cannot open shared object file: No such file or directory
<clever> samueldr: its failing to properly embed strace into the initrd
<clever> and pkgsStatic also cant find gcc_s
<clever> samueldr: ah
<clever> 22 cp ${pkgs.stdenv.cc.cc}/armv7l-unknown-linux-gnueabihf/lib/libgcc_s.so $out/lib/ -v
<clever> i tried that, but nada
<samueldr> weird, in my init it works fine
<clever> cp: missing destination file operand after '/nix/store/81i544mmyijmv33hidpjz1khfq8gmnxl-extra-utils-armv7l-unknown-linux-gnueabihf/lib'
h0m1 has quit [Ping timeout: 252 seconds]
<clever> samueldr: its not in glibc, on my end
<clever> samueldr: confirmed with an echo + ls
<samueldr> oh
<samueldr> I never tried it since the switch to gcc9
<samueldr> I would assume this changed something
h0m1 has joined #nixos-aarch64
{^_^} has quit [Remote host closed the connection]
{^_^} has joined #nixos-aarch64
<clever> [ 13.445993] potentially unexpected fatal signal 11.
<clever> [ 13.450889] CPU: 0 PID: 92 Comm: udevadm Not tainted 4.19.89-v7+ #143
<{^_^}> https://github.com/NixOS/nixpkgs/pull/143 (by falsifian, 7 years ago, merged): upgrade firmware-linux-nonfree to 0.36
<clever> samueldr: thats not good...
t184256 has left #nixos-aarch64 ["Disconnected: Replaced by new connection"]
t184256 has joined #nixos-aarch64
zupo has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
wavirc22 has quit [Ping timeout: 248 seconds]
wavirc22 has joined #nixos-aarch64
wavirc22 has quit [Read error: Connection reset by peer]
wavirc22 has joined #nixos-aarch64
t184256 has left #nixos-aarch64 ["Disconnected: Replaced by new connection"]
t184256 has joined #nixos-aarch64
t184256 has left #nixos-aarch64 ["Disconnected: Replaced by new connection"]
t184256 has joined #nixos-aarch64
orivej has quit [Ping timeout: 265 seconds]
zupo has joined #nixos-aarch64
t184256 has left #nixos-aarch64 ["Error from remote client"]
t184256 has joined #nixos-aarch64
angerman has quit []
angerman has joined #nixos-aarch64
t184256 has left #nixos-aarch64 [#nixos-aarch64]
NekomimiScience has quit []
NekomimiScience has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ryantrinkle has joined #nixos-aarch64
misuzu has quit [Quit: leaving]
misuzu has joined #nixos-aarch64
pkral has quit []
pkral has joined #nixos-aarch64
t184256 has joined #nixos-aarch64
orivej has joined #nixos-aarch64
rajivr___ has quit [Quit: Connection closed for inactivity]
zupo has joined #nixos-aarch64
<sphalerite> gchristensen: was nix-env intentionally broken on the aarch64 community box?
<gchristensen> I don't think so?
<sphalerite> for some reason I have no dir in /nix/var/nix/profiles/per-user
<gchristensen> what hapens if you use nix-env ?
<sphalerite> err actually I think the problem is that I have no .nix-profile in my home
<gchristensen> this all seems normal
<gchristensen> like, pretty standard I think, for a brand new user?
<clever> i think the bashrc deals with creating it on login
<gchristensen> maybe it is a bug with when the daemon started making the profiles/per-user dirs?
<sphalerite> what happens: it tries to build a new generation of /nix/var/nix/profiles/default
<sphalerite> which is what nix-env does if there's no ~/.nix-profile symlink
<sphalerite> iirc
<gchristensen> [grahamc@aarch64:~]$ nix-env -iA nixos.hello
<gchristensen> error: getting information about '/home/grahamc/.nix-defexpr': No such file or directory
<gchristensen> hm
<sphalerite> hm, that could be related
<samueldr> sphalerite: I did hit the same issue the other day
<clever> nix-defexpr is to do with channels
<sphalerite> `ln -s /nix/var/nix/profiles/per-user/lheckemann/profile .nix-profile` makes nix-env work
<clever> lrwxrwxrwx 1 clever users 44 Oct 11 2015 .nix-defexpr/channels_root -> /nix/var/nix/profiles/per-user/root/channels
<gchristensen> it semes nix isn't initailizing on login
<samueldr> *something* is off, but no idea what
<sphalerite> (not for nix-defexpr stuff presumably, but at least it can do _something_)
<clever> this should exist
<sphalerite> it makes `nix-env -f https://sphalerite.org/dotfiles.tar.gz -iA basic` work :p
<sphalerite> but yeah I'm guessing the nonexistence of .nix-profile and .nix-defexpr have the same root cause
<clever> yeah
<gchristensen> I feel like I have two critical questions. (1) does this happen on any brand new install of NixOS unstable (2) is this a difference in how this box is built
<gchristensen> I imagen (1) is probably no because surlely (?) we have a test (?) which ould cover this (?)
<flokli> I assume nixos/tests/boot.nix should test that (we nix-env --dry-run -iA nixos.procps)
<samueldr> I was thinking the same
<sphalerite> I guess that tests .nix-defexpr but not .nix-rpofile
craige has quit [Ping timeout: 268 seconds]
zupo has quit [Ping timeout: 268 seconds]
zupo has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
zupo has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
zupo has joined #nixos-aarch64
zupo has quit [Ping timeout: 240 seconds]
zupo has joined #nixos-aarch64
<samueldr> if you're ever tight in a serial console which wraps weirdly, open a new tab or tmux window or whatever outside the serial connection
<samueldr> $ echo "stty rows $(tput lines) cols $(tput cols)"
<samueldr> that'll give you the proper stty invocation to fix the wrapping
<gchristensen> throw that right up in to the wiki
<gchristensen> this is very pretty https://wiki.kobol.io/helios64/intro/
<samueldr> RK3399 +1 for a platform that is pretty well maintained and known
<samueldr> so I figure that little box holds one of those RK3399 and HDD caddies, and not multiple of those RK3399 boards
<samueldr> nixos support should be a cinch
<samueldr> more pretty pictures https://kobol.io/
<sphalerite> yeah, someone mentioned that in here the other day. I'd really like one, but I already have a hackier equivalent, so I'm not sure I can justify it
<sphalerite> I'm already struggling to decide whether I want a pinebook pro or not
<samueldr> that price is really good
<samueldr> it's also standard form factor, and can even be powered from an ATX power supply
<samueldr> so once they ship non full-bundle, you could mount it in a standard case
<sphalerite> well, I'm powering my nanopi from a standard ATX power supply too
<samueldr> >> USB Type-C : DAS Mode
<samueldr> curious what's doing it
<samueldr> if it's u-boot or the kernel
<gchristensen> wow that is a great price
<samueldr> look carefully at all the features
<samueldr> I think they dropped an order of magnitude accidentally :)
<samueldr> not really, but similar featured things are not cheap!
<samueldr> it wouldn't have surprised me to see the board alone at that price point, the box at that price too
<gchristensen> same...
<flokli> awww <3
<gchristensen> hm?
<flokli> helios64 really sounds like something I'd want to have
<samueldr> same here, though have no real need to
<gchristensen> same here, except I sort of(?) have reason to (ish?)
<samueldr> I sort of could move the drives from my older computer tower into that
<samueldr> that computer is currently doing what that box would
<samueldr> but it's not like it's needed
<flokli> well, the current box holding my harddrives probably has a worse cpu than that thing :-)
<flokli> Intel(R) Celeron(R) CPU G1610T @ 2.30GHz
<flokli> :-)
<gchristensen> :o
<samueldr> >> Intel(R) Xeon(R) CPU E5440 @ 2.83GHz
<samueldr> that's running in the wrong socket
<gchristensen> hah
<samueldr> LGA771 CPUs can generally run in LGA775 boards with a pin mod, you could buy them for dirt cheap on aliexpress pre-modded
<samueldr> I think I paid 20$ for what, at the time, was a really good upgrade
<gchristensen> nice...
<samueldr> especially considering the Pentium chip it replaced didn't have VT
<samueldr> sphalerite (if you're around) your unbind/bind dance for gru-bob, how long does it usually take for you?
zarel has quit [Ping timeout: 252 seconds]
zarel has joined #nixos-aarch64
<samueldr> turns out my code was bad :)
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ryantrinkle has quit [Ping timeout: 240 seconds]