orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
h0m1 has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
knerten1 has joined #nixos-aarch64
knerten has quit [Ping timeout: 240 seconds]
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
<samueldr>
while that mediatek device is not better than qualcomm devices outright... it's not really worse it looks like
<samueldr>
some things work better... like the ramoops pstore always just works
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
angerman has quit [Read error: Connection reset by peer]
angerman has joined #nixos-aarch64
orivej has joined #nixos-aarch64
pbb has quit [Read error: Connection reset by peer]
pbb has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
orivej_ has joined #nixos-aarch64
orivej has quit [Ping timeout: 260 seconds]
<samueldr>
trying to get pkgsCross.aarch64-multiplatform.buildPackages.gdb to run gdb on aarch64 stuff (a kernel) and it looks like this doesn't have a gdb binary?
<samueldr>
though re-reading I'm not sure it would help me, I already know the function, I would want to know where in it
orivej_ has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-aarch64
rajivr has joined #nixos-aarch64
zupo has joined #nixos-aarch64
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-aarch64
zupo has quit [Client Quit]
orivej has quit [Ping timeout: 265 seconds]
<patagonicus>
thefloweringash: Overall armv7 seems to be ok for rebuilding a system at the nixos-unstable channel, but it's currently failing for me in the test stage of bash-completion as the last actual package to be compiled.
<thefloweringash>
I added a minimal system to my hydra too: don’t know if it includes bash completion, but it succeeds.
<patagonicus>
Specifically the "Unknown pytest.mark.complete - is this a typo?" - but for me those tests are also marked with an "E" in the progress graph, but not on your hydra.
<patagonicus>
I'm currently building git + dependencies so I can bisect it. :)
<thefloweringash>
Interesting. Are we on at the same nixpkgs version?
<samueldr>
ambient impurities like the instruction set?
<thefloweringash>
That would be mildly horrifying, but useful to learn
<samueldr>
also, just checking, nix on nixos on armv7l, right?
<patagonicus>
samueldr: Exynos5422 with 4xCortex A15 and 4xA7
<thefloweringash>
Um, lx2160a, cortex-a72, in kvm with -cpu=host,aarch64=off
<patagonicus>
Yes, armv7l
<samueldr>
yeah, I don't know what can differ instruction-set-wise between A15, A7 and A72 in 32 bit mode
<samueldr>
note, thefloweringash, you are probably running armv8
<samueldr>
which, in addition, I don't know how it differs from armv7
<patagonicus>
The error doesn't seem like something that should be impacted by arch, but, it's computers, you never know. As far as I can tell so far the difference is that my machine treats warnings about an unknown annotation in Python as fatal.
<patagonicus>
Especially since it prints status updates of the tests every 5% of progress, but I noticed that sometimes the lines where one character shorter on my machine.
<thefloweringash>
What does your test output look like, a whole bunch of `E`? Maybe pastebin the log
ryantrinkle has quit [Ping timeout: 240 seconds]
<patagonicus>
Yup, give me a second. git is also in the testing stage, so I can see if that commit is the one breaking it for me soon.
<thefloweringash>
In my experience, git takes as very long time to test :-)
<patagonicus>
Hmm, not sure if I saw the timeout before, that might have been caused by git testing at the same time.
<patagonicus>
But, there's more errors that just the ones in your log, so those are probably "ok" and something else is broken.
<patagonicus>
Oh, no. For some reason trying to nix-shell for nix-bisect is trying to build things like coreutils, gcc and a few others. Really weird since I already have those built at the current nixos channel.
<patagonicus>
Yep, it's the parallel execution that breaks the tests. Which is annoying, because that does speed up the tests quite a bit.
<{^_^}>
#91706 (by Patagonicus, 3 minutes ago, open): pytest_xdist breaks tests for bash-completion on armv7l
<thefloweringash>
Is this a resource contention thing, that is, do you have a failed test run that doesn’t include timeouts?
<thefloweringash>
Does it pass with different parallelism, like only using two cores?
<patagonicus>
So, I wasn't running anything else while doing these builds - if that's not good enough I doubt many other armv7 machines will be able to run it.
<thefloweringash>
Your machine does have an unusually large number of cores for an armv7l machine.
<patagonicus>
I can try doing less parallelism, but it sounds like there's something inherently broken in how the tests are run and chosing the right number of cores might just mask that. The commit introducing the parallelism to the nixpkg says that the tests are flaky and slow and with the change they are at least not slow anymore …
<thefloweringash>
I’m wondering if it’s also file system resources (like a microSD card) that are contributing
<patagonicus>
Maybe. /tmp is on an SSD, though. I can try with it on the SD card.
<patagonicus>
Currently building some other packages, but I'll play around with it later today.
<thefloweringash>
Though yeah, I do agree, I think we should definitely tune the timeout to be much more tolerant. The nix CI machines also tend to be very busy
<patagonicus>
This also seems to be reproducible, I've probably built the broken version at least half a dozen times and it always failed.
<patagonicus>
AFAICT it already fails with 2 cores.
<patagonicus>
Yeah. But I only get three timeouts, so I'm guessing it gets worse the more cores you throw at it.
<patagonicus>
Oh. It passed with 4 cores, but not with 3. Yeah, these are some really unreliable tests.
<patagonicus>
Aaand it's not deterministic. A second run with 2 cores passed. Yeah, I don't trust those thests.
zupo has joined #nixos-aarch64
orivej has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
zupo has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 264 seconds]
orivej_ has joined #nixos-aarch64
orivej has joined #nixos-aarch64
orivej_ has quit [Ping timeout: 264 seconds]
zupo has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
bdju has quit [Quit: Lost terminal]
bdju has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 256 seconds]
orivej_ has joined #nixos-aarch64
LnL has quit [Quit: exit 1]
ehmry has joined #nixos-aarch64
orivej_ has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 258 seconds]
orivej has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
ryantrinkle has joined #nixos-aarch64
orivej has quit [Ping timeout: 256 seconds]
orivej_ has joined #nixos-aarch64
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
orivej_ has quit [Ping timeout: 256 seconds]
zupo has joined #nixos-aarch64
zupo has quit [Client Quit]
WilliButz has joined #nixos-aarch64
WilliButz has quit [Remote host closed the connection]
Guest80019 has joined #nixos-aarch64
WilliButz has joined #nixos-aarch64
Guest80019 is now known as LnL
WilliButz has quit [Remote host closed the connection]
WilliButz has joined #nixos-aarch64
WilliButz has quit [Remote host closed the connection]
WilliButz has joined #nixos-aarch64
orivej has joined #nixos-aarch64
<patagonicus>
"can't open fsimg temp.img: Value too large for defined data type" well, that's an error I didn't expect for building a new sd-image on the odroid itself instead of cross-compiling.
pbb has quit [Ping timeout: 272 seconds]
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
pbb has joined #nixos-aarch64
rajivr has quit [Quit: Connection closed for inactivity]
<{^_^}>
#82718 (by misuzu, 14 weeks ago, open): nixos/lib/make-ext4-fs: use mkfs.ext4 instead of cptofs
<patagonicus>
lopsided98: Nice, thanks!
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
<patagonicus>
Yep, cherry-picked that on top of my fix for bash-completion at the nixos-unstable (or at least where it was yesterday) and it worked. :)
<patagonicus>
lopsided98++
<{^_^}>
lopsided98's karma got increased to 13
<patagonicus>
I doubt I'll have time to before next weekend, but one of the next steps is making my own version of sd-image.nix for the XU4/HC1/HC2 (and probably XU3 as well) that gets rid of the firmware vfat partition and installs the right bootloader.
t184256 has joined #nixos-aarch64
orivej has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
quinn has quit [Ping timeout: 258 seconds]
orivej has quit [Ping timeout: 256 seconds]
orivej has joined #nixos-aarch64
bennofs has joined #nixos-aarch64
bennofs_ has quit [Ping timeout: 264 seconds]
quinn has joined #nixos-aarch64
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 260 seconds]
orivej has joined #nixos-aarch64
quinn_ has joined #nixos-aarch64
quinn has quit [Ping timeout: 260 seconds]
cole-h has joined #nixos-aarch64
<patagonicus>
Wow, now I'm super confused. If I built odroid-xu3-bootloader on the HC2 itself (or rather ubootOdroidXU3) then it correctly tries to load exynos5422-odroidhc1.dtb instead of exynos5422-odroid.dtb. So that seems to be a problem with cross-compiling.
<patagonicus>
But, on the upside: I can built a full sd-image on the HC2, flash it, flash odroid-xu3-bootloader and it works. :)
zupo has joined #nixos-aarch64
orivej has quit [Ping timeout: 264 seconds]
orivej has joined #nixos-aarch64
orivej_ has joined #nixos-aarch64
orivej has quit [Ping timeout: 260 seconds]
cole-h has quit [Quit: Goodbye]
orivej_ has quit [Quit: No Ping reply in 180 seconds.]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 258 seconds]
orivej has joined #nixos-aarch64
orivej_ has joined #nixos-aarch64
orivej has quit [Ping timeout: 256 seconds]
orivej_ has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-aarch64
h0m1 has quit [Quit: WeeChat 2.8]
h0m1 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: 264 seconds]
<samueldr>
I'm quite excited, I just built (in nix) a kernel for a device for which there is AFAICT absolutely no custom ROM, not even a custom kernel build from someone other than the OEM :D
<samueldr>
(though there is source)
<samueldr>
so, first good news: it looks like the source is good
<ashkitten>
!
<ashkitten>
what device?
<samueldr>
asus-x018d
<samueldr>
it was supposed to be my first mediatek port, but a friend donated his old phone to the project (xiaomi-begonia) so now there is no good reason to keep it for a future first MTK port
<samueldr>
the main reason it was chosen is: asus is B+ in bootloader unlocking (phones home, but no waiting time); asus releases sources; sp flash tool + roms to unbrick is known to work
<ashkitten>
ooh
<gchristensen>
nice work!
<samueldr>
right now it's not booting mobile nixos, I had issues with my first try, but it does boot the TWRP build that is using the binary kernel from ASUS
<samueldr>
I at least have a known good build to extrapolate from and figure out what is broken
<ashkitten>
(still no cosmo, btw. they told me last week that they're waiting on dvorak devices to arrive in europe so maybe 1-2 wks)
<samueldr>
yeah :(
<samueldr>
oh, and another important bit, it was a 64 bit device, running 64 bit OS :)
<ashkitten>
i have no expectations at this point until i see a tracking number
<ashkitten>
nice
<samueldr>
so if all goes well, this evening I have the first non-OEM rom going, and it's not even android!
<ashkitten>
maybe in another 2 decades oems will ship devices with linux preloaded
<samueldr>
2 decades is two too many
<ashkitten>
yes, but it's at least realistic
<samueldr>
be careful how it's phrased, or else the monkey paw will curl
<samueldr>
your phone now ships with windows 11, using WSL4
orivej has quit [Ping timeout: 246 seconds]
orivej has joined #nixos-aarch64
<ashkitten>
at that point i abandon having a phone