Werner changed the topic of #armbian to: armbian - Linux for ARM development boards | www.armbian.com | Github: github.com/armbian | Commits: #armbian-commits | Forums Feed: #armbian-rss | This channel is logged -> irc.armbian.com
ChriChri_ has joined #armbian
ChriChri has quit [Ping timeout: 246 seconds]
ChriChri_ is now known as ChriChri
dddddd has quit [Ping timeout: 256 seconds]
sunshavi has quit [Remote host closed the connection]
sunshavi has joined #armbian
<Tony_mac32> IgorPec no, I haven't unfortunately
_whitelogger has joined #armbian
pikudoz has joined #armbian
<nekomancer[m]> images distribution. Maybe.
<lanefu> oh yeah that could be interesting to experiement with.. i kind of assume performance isnt great
R0b0t1 has quit [Ping timeout: 272 seconds]
R0b0t1 has joined #armbian
R0b0t1` has quit [Quit: WeeChat 2.9-dev]
R0b0t1` has joined #armbian
pikudoz has quit [Ping timeout: 256 seconds]
<nekomancer[m]> possible.
<Tony_mac32> sigh I love doing everything over and over again
<nekomancer[m]> like cron?
<Tony_mac32> lol
macc24 has quit [Ping timeout: 258 seconds]
<Tony_mac32> in 128 patches somehow the device trees for the boards were untouched, concerning the audio
<nekomancer[m]> to keep work as patches — so complicated
<nekomancer[m]> very fragile
<lanefu> Tony_mac32: so... if the device tree is borked.. _would_ the kernel modules still load/build?
<Tony_mac32> it does, there's nothing wrong with the kernel module as far as I can tell, the kernel just doesn't know to load it because it's not described as part of the board's hardware
<TomTheDragon> yeah audio is... missing from a lot of boards in mainline, i've experienced
<Tony_mac32> I did exactly what I'd doing right now a couple months ago on 5.4, lol
<TomTheDragon> rockpi4 seems to have recently got a patch, but not sure if armbian's actually being built with it.
<nekomancer[m]> hdmi audio countedas part of boadr hardware?
<Tony_mac32> the I2S audio block, yes
<TomTheDragon> pretty much everything is counted as board hardware... IIRC even on chip stuff
<Tony_mac32> *especially on chip stuff*
<nekomancer[m]> but if I not plug hdmi, or plug monitor without sound — there no hdmi sound device
<nekomancer[m]> on pc it pnp. but on arm?
<TomTheDragon> I never realized DTB was so... convoluted. It seems like from kernel version to kernel version, nothing's compatible.
<Tony_mac32> hdmi has i2c
<Tony_mac32> for handshaking/hotplug
<Tony_mac32> so the kernel can know when it is/isn't attached
<nekomancer[m]> do i2c allows discovering?
<nekomancer[m]> or just plug something already known from dtb?
<Tony_mac32> I don't know exactly what causes the handshake to start, could be a grounded pin triggering an interrupt, I've not gone that deep
<Tony_mac32> but the DDC is i2c
<Tony_mac32> and is where you get the EDID
<Tony_mac32> TomTheDragon the problem with this audio stuff on Meson64 is there have been 3 WIP versions, each one different
archetech has quit [Quit: Konversation terminated!]
<lanefu> device tree groundhog day
<lanefu> Tony_mac32: anythig i can do to be helpful tha isnt hellaciously over my head?
<Tony_mac32> sadly no, this is a bad case of multiple unrelated patches stacked on top and missing some critical pieces
<Tony_mac32> my patches match Jerome's but don't work, so something is jacked up
_whitelogger has joined #armbian
<lanefu> k, cuz im also running out of jokes, so its getting difficult
<Tony_mac32> wut
<lanefu> its true
<lanefu> I guess I could review all the broken commitments i've made to Igor and see what i can follow-up on lol
<Tony_mac32> Careful what you dig up, you'll wind up like me :P
<lanefu> yeah
<lanefu> we were like "COME BACK TO IRC AND PAY YOUR DUES"
<Tony_mac32> I added gxbb to it for 5.4
<Tony_mac32> but since those handles don't exist in our 5.6, it must be missing entirely
<lanefu> woohoo
<Tony_mac32> there are 4 disabled patches that might have been supposed to do that, but this one is from Jerome's 5.6 integration branch, so I'll git a go
_whitelogger has joined #armbian
xec has joined #armbian
xecutertool has quit [Ping timeout: 244 seconds]
redentor has quit [Quit: Leaving]
drobo_01 has joined #armbian
drobo_00 has quit [Ping timeout: 246 seconds]
drobo_01 is now known as drobo_00
ed_peguillan has quit [Remote host closed the connection]
<Werner> lanefu, You're welcome
<Werner> Good morning
<Tony_mac32> good morning
<Tony_mac32> IgorPec, you up too?
<Tony_mac32> I want to wipe out the CMA size patches for the MEson64 vdec, and see if we can add it as an overlay option, I think it is wrecking the 512 MB La Frite
<IgorPec> good morning
<IgorPec> yes
<Tony_mac32> So after some runaround and some surfing patchwork, I see the device tree part of the whole audio mess is only a few weeks old, so I'll try to pull it in
<Tony_mac32> aiu driver has been in for a while, but no boards were using it
<IgorPec> great, so we are moving somewhere, but we still have reboot problem, right?
<Tony_mac32> Nope, I purged the Rockchip vdec patches that wree in there somehow and lanefu confirmed Le Potato reboots
<Tony_mac32> so does mine
<IgorPec> i manage to do fix repository bugs (hopefuilly), currently testing
<IgorPec> aha, ok. we need more testings then. my test gear is now hooked to nightly and i will see fast if something is wrong
<IgorPec> all those devices are rebooting properly
<IgorPec> others i have to check
<Tony_mac32> Tinkerboard 30.204 Uptime: 1:23
<Tony_mac32> that's good to see, given the insane shit I had to put in there...
<Tony_mac32> :P
<IgorPec> tinker is doing quite fine, yes. btw. perfromance numbers are bogus ... i only test upgrade + reboot
<Tony_mac32> reminds me I need to revisit u-boot for that one, see if the bugs are gone
<Tony_mac32> N2 I got u-boot booting up, but the N2 kernel was crashing immediately, I'm guessing it doesn't like something in the standard Meson64 boot script
<IgorPec> true. i already looked once and there is some rfc needed for this usb loading
<Tony_mac32> mainline uboot*
<IgorPec> yeah, but there are some patches to get that working ... not sure if that is in mainline
<Tony_mac32> hm ok
<IgorPec> my test odroid n2 seems fine
<IgorPec> i'll run stress test and see if all boards survived :)
<Tony_mac32> My N2 has been running rosetta for weeks, I only brought it down to try the mainline ubootr
<Tony_mac32> :)
<IgorPec> ahaa, mainline
antenagora has joined #armbian
<Tony_mac32> hmmmmm, I'm going to have to look at this more tomorrow.
macc24 has joined #armbian
turker has joined #armbian
ed_peguillan has joined #armbian
macc24 has quit [Quit: WeeChat 2.8]
macc24 has joined #armbian
dddddd has joined #armbian
lids has joined #armbian
Elpaulo has joined #armbian
DaRock has joined #armbian
<HerculeP> moin
<HerculeP> lanefu: still no reboot, no sound (c2)
<HerculeP> btw, filename didn't change!?
macc24 has quit [Quit: WeeChat 2.8]
xec has quit [Remote host closed the connection]
macc24 has joined #armbian
xecuter has joined #armbian
Net147 has quit [Quit: Quit]
Net147 has joined #armbian
Elpaulo has quit [Quit: Elpaulo]
Elpaulo has joined #armbian
Strykar has quit [Quit: /quit]
Strykar has joined #armbian
gnom has quit [Ping timeout: 258 seconds]
torv has quit [Remote host closed the connection]
torv has joined #armbian
<plntyk> IgorPec: i think i found some odroid-c1 issues that might keep the images from booting - Linux Mint *might* be to blame
<plntyk> installing "Linux Mint 19.3" in a VM (libvirt/qemu based) and running a build inside it I ended up with a not bootable image
<IgorPec> plntyk: yes, i also suspected that
<plntyk> 1st (different) error: linux mint errors on writing : print_req_error: I/O error, dev loop0, sector 8192 flags 80700 in dmesg
<plntyk> this prevents even the boot.ini loading so stops before the other issue
<IgorPec> yes, this way it was for me ... could be in fact qemu issue.
<plntyk> mounting an image created on mint on host os (arch) and doing : lsattr <sddisk-mount>/boot results in errors on symlinks
<IgorPec> i also have clean bionic VM here, i will do a build now
<IgorPec> yes, file system is corrupted somehow
<plntyk> aka the uinitrd, zimage files that are symlinks in /boot
<IgorPec> but strange is that this is not happening for other images
<IgorPec> mint is my desktop, where i do most of my testings
<IgorPec> can you check if nightly images are working for you
<IgorPec> i will do the same
Tenkawa has joined #armbian
<IgorPec> corrupted ranmdisk, but other things are loaded
<lanefu> HerculeP: hmm... yeah i replaced the file.... (i think :P)
<IgorPec> lanefu: what have you done to our friend Jenkins ? :)
<HerculeP> lanefu: can you just make a minimal img for testing reboot? last download took 30 minutes :(
<lanefu> HerculeP: sure
<lanefu> IgorPec: ha let me see
<lanefu> IgorPec: looks okay to me
<lanefu> looks like last nights test failed on wireless
<IgorPec> yes, thats been fixed
<lanefu> ha okay.. is there somethign else wrong iwth Jenkins?
<IgorPec> i guess its a network problem. i was not able to access it
<IgorPec> now it works again
<lanefu> :P was it a timeout or did you get an error message?
<IgorPec> like host unreachable
<lanefu> lame
<IgorPec> it could be on my side as well, dunno
<IgorPec> my ISP is crap ;)
<plntyk> IgorPec : same error with uInitrd invalid extent , deleting the uinitrd symlink and creating a copy makes the image boot -> the copy does not have an invalid extent block, the uinitrd-5.4.40-meson does have one
<plntyk> (when stopping the autoboot and issuing extload in bootloader)
archetech has joined #armbian
<IgorPec> plntyk: true
<IgorPec> plntyk: what a bizzare issue
<IgorPec> but this image was build on a Ubuntu Bionic
<plntyk> well "stable" ext4 filesystem is funny when looking at https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/log/fs/ext4
antenagora has quit [Remote host closed the connection]
Net147 has quit [Ping timeout: 258 seconds]
sassinak-work has quit [Ping timeout: 246 seconds]
sassinak-work has joined #armbian
Net147 has joined #armbian
drobo_00 has quit [Quit: drobo_00]
<plntyk> lates change to 8811cu may result in build breakage (at least on my end)
DaRock has quit [Ping timeout: 260 seconds]
<HerculeP> lanefu: still no reboot, still that watchdog error msg :(
<lanefu> interesting
<lanefu> lemme check ebay for a c2
<HerculeP> LOL
<lanefu> or mayble i should ask internet for donation
* HerculeP is still glad to help testing
<lanefu> ha. yeah you've been super helpful
<HerculeP> minimal.img just took 5 min to download & burn this time :D
<lanefu> word.. yeah maybe i'll start gzippign since that's not too slow to decompress
<buZz> doesnt have to be fast, SD cards are slow to begin with :P
<lanefu> hmmm okay yeah weird tater didnt reboot either now
<lanefu> IgorPec: what's the fix for that network driver error that makes the build die.. I'm having trouble iwth it again against #1956
<IgorPec> plntyk: 8811cu was fixed in the morning
<IgorPec> git pull
<Tenkawa> 8811cu is a nightmare
* Tenkawa uses one on his x86/64
<lanefu> thanks.... so weird.. my checkout should have that..
<IgorPec> all those are bad ;)
<lanefu> i need to learn more about how the github special repo endpoints like pull work
<lanefu> this should be the PR against master
<IgorPec> lanefu: i never used this trick before to set to commit id
<lanefu> lane@billroyall:~/GIT/build$ cat userpatches/getPR.sh
<lanefu> #!/bin/bash
<lanefu> REMOTE_REPO=upstream
<lanefu> PR=${1}
<lanefu> git fetch -v ${REMOTE_REPO} +refs/pull/${PR}*:refs/remotes/${REMOTE_REPO}/pull/${PR}*
<lanefu> git checkout ${REMOTE_REPO}/pull/${PR}/merge
<ArmbianTwitter> @armbian (armbian): RT @lanefu: Anybody have an Odroid C2 they wanna send me to use for bench testing? (9s ago)
<IgorPec> lanefu: do you have that change in the source?
<IgorPec> i choose to try commit id instead of making a patch / PR
<lanefu> IgorPec: ha sorry i was probably confusing you.. i wasn't getting that change. i was debugging my checkout script ofr testing PRs
redentor has joined #armbian
<lanefu> i've got it now
<lanefu> i think i needed a -f in the checkout
<IgorPec> aha, ok
<lanefu> thanks. sorry for the confusion
<IgorPec> well, driver code is broken upstream ... setting to commit id is a proper way anyway
<lanefu> git is hard :P
<lanefu> yep def the right fix
<IgorPec> the more you learn it, the better you come along ;)
<IgorPec> ""High Gitfu gets you best girl in town" ;)
<lanefu> rofl
redentor has quit [Remote host closed the connection]
<lanefu> okay reboot worked this time on potato now that i built correctly.. make new c2image for herc
<IgorPec> 5.6.y?
<lanefu> yep
<lanefu> from 1956
<IgorPec> good. so "only" sound is left ... or more?
<plntyk> IgorPec: debugfs shows a very strange output of the non-working odroid-c1 : https://gist.github.com/plntyk/adeddd835024c72a52782e08c3d94781
<plntyk> debugfs + dump_extents option show really strange output
<IgorPec> plntyk: we can do it with a workaround
<lanefu> I think sound.. Tony_mac32 was curious about USB.. so i suppose there's something evaluation we can do there
<IgorPec> 1. boot fat partition 2. change scripts to force copy and not link
<lanefu> grrr u-boot is failing for C2 on 1956 now
<lanefu> hahah okay that is my doing :P
<IgorPec> lanefu: my C2 in the test lab is 2020.04 / 5.4.40 and it seems alright. i can put some stres to see if i can crash huim
<lanefu> pro-tip: no _ in ${SUBREVISION}
<IgorPec> yeah, it can make shit ... yep
<IgorPec> c2 got a task: 10 x (sbcbench + reboot) ... lets see if it survives that ;)
* HerculeP just tested nightly trunk133/5.4 - no reboot, no sound
<HerculeP> (c2) ofc on the GUI, select audio output showed only "analog output" as option, no HDMI
macc24 has quit [Ping timeout: 260 seconds]
macc24 has joined #armbian
<archetech> got one of those for rock64 v2 yet ?
macc24_ has joined #armbian
* HerculeP needs a coffee 1st
macc24 has quit [Ping timeout: 265 seconds]
<lanefu> archetech: this is for testing Amlogic stuff out of a PR
<archetech> I meant the 5.6.13 image
* Tenkawa needs much coffee
* Tenkawa is on a multi month coffee withdrawal :(
<lanefu> Tenkawa: that's a terrible plan
<nekomancer[m]> <Tony_mac32 "My N2 has been running rosetta f"> rosetta? are you talking about boinc project? did boinc projects now available for arm sbc?
<lanefu> nekomancer[m]: yes
<Tenkawa> lanefu: no choice... lockdown
<nekomancer[m]> :wow:
macc24_ is now known as macc24
drobo_00 has joined #armbian
<HerculeP> lanefu: no reboot :( , "watchdog did not stop"
<lanefu> damn
<lanefu> thanks HerculeP
<HerculeP> np
<IgorPec> lanefu: so it hangs on reboot, right?
<IgorPec> it doesn't complete the reboot
<lanefu> IgorPec: correct according to HerculeP , C2 does not complete reboot
<lanefu> le potato is successful, just not c2
<lanefu> archetech: i don't think we're planning on moving Rock64 to 5.6 on the 'current" branch.. but it's 5.6 on dev https://armbian.lane-fu.com/Armbian_20.05.0-trunk-archetect_Rock64_buster_dev_5.6.13_minimal.img.gz
<IgorPec> on the other hand C2 reboot works pitch perfect with 5.4.y
<IgorPec> if we find the reson, we can switch this in last moment
<archetech> thanks lanefu
<archetech> ls
<lanefu> cache compile.sh config lib LICENSE output packages patch README.md userpatches VERSION
<TomTheDragon> nekomancer[m]: I'm.. honestly kinda surprised boinc is a closed platform
<TomTheDragon> or, maybe it's not
<TomTheDragon> appears that it is actually open source
<lanefu> the project workloads and their arch capality is what varies
turker has quit [Remote host closed the connection]
<HerculeP> IgorPec: 5.6 img from lanefu command "reboot" shuts down /w error "watchdog did not stop", shuts off HDMI but stays off instead of starting again
<HerculeP> all my testing from SD (cant find my emmc :( )
<IgorPec> ok, perhaps we have a "classic" amlogic SD card problem
<IgorPec> i have done my tests with eMMC and last with I have similar trobles with C4. worked well on eMMC, very bad on SD
<IgorPec> we talked with chewitt the other day and there might be some fixes ...
redentor has joined #armbian
<lanefu> Ooh nice find
<IgorPec> plntyk: i propose to use this FAT boot solution and proceed here https://forum.armbian.com/topic/13888-odroid-c1-s805/?tab=comments#comment-101591
<HerculeP> yay, found my emmc,
<HerculeP> reboot works!
<HerculeP> lanefu: Armbian_20.05.0-trunk_Odroidc2_buster_current_5.6.12_desktop.img
<lanefu> oh sweet okay hten
<plntyk> nakadamia16+16
<plntyk> yay
<HerculeP> but still no sound :(
jakogut has quit [Remote host closed the connection]
<IgorPec> sound is not critical issue, we can add that later
redentor has quit [Remote host closed the connection]
<lanefu> IgorPec: trying to thing of hte best way to pull chewitt's patches and test
<IgorPec> just grab that patch
<IgorPec> it applies cleanly
macc24 has quit [Ping timeout: 256 seconds]
macc24_ has joined #armbian
macc24_ is now known as macc24
<lanefu> ha okay couldnt figure out hwo to do come up with that sweet url :P building
<HerculeP> lanefu: no reboot
<IgorPec> dman
<IgorPec> is reboot always hangs or sometimes?
<HerculeP> reboot doesnt "hang" its equvalent to "shutdown now" (doesn't restart)
<HerculeP> always
<IgorPec> ok
<IgorPec> did you power cycle after kernel change?
<HerculeP> yup
<HerculeP> complete re-image & re-power
<IgorPec> aaha, true. well, damn
<IgorPec> we need to sleep over and dig in more
<HerculeP> almost midnight indeed
<IgorPec> for me too :)
<HerculeP> at least we found out its SD related
<IgorPec> yes, and its some progress
macc24 has quit [Quit: WeeChat 2.8]
DaRock has joined #armbian
Tenkawa has quit [Ping timeout: 265 seconds]
<Tony_mac32> @igorpec THAT is the patch I was talking about, couldn't find!
Tenkawa has joined #armbian
toketin has quit [Quit: ZNC 1.7.5 - https://znc.in]
toketin has joined #armbian
jakogut has joined #armbian
archetech has quit [Quit: Konversation terminated!]