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 | Type 'help' for help | Logs: -> irc.armbian.com
<lanefu> yeah lemme think on how to give you context
<lanefu> also in your check-out.. do `touch .ignore_changes` and thank me later
<wbn> bless
<lanefu> okay
<lanefu> If ther's a armbianEnv.txt
<lanefu> which is most
<lanefu> it will be derived from here.. https://github.com/armbian/build/tree/master/config/bootenv
<lanefu> it's possible that specific board customizations may add additional
<wbn> ok
<wbn> i also totally blanked, im SSHed into my armbian board rn so i c an just check that
<lanefu> actually ther ewill always be some customization
<lanefu> haha
<lanefu> yeah
<lanefu> a big thing is mapping the device tree naming convetion adn managing overlays
<lanefu> wbn: do you need to worry abotu wifi config?
<wbn> thank GOD no
<lanefu> k
<wbn> haha
<wbn> im wondering what armbianEnv.txt has to do with boot time stuff
<wbn> like
<wbn> < wbn> also I want to be able to test if /dev/mmcblk0 is uninitialized, and if it is, overwrite it with the contents of its own block device
<wbn> this is something that i want to do at boot-time
<DigitalMan1983> that's where the boot parameters are
<wbn> where do i put that?
<lanefu> yeah those are boot params
<lanefu> now there is an armbian-first-run
<wbn> hmm
<wbn> i'm not sure if first-run is OK for this case
<lanefu> let me see what hooks are there
<lanefu> yeah wait tell me again what you're wanting to do lol
<wbn> i'll explain a little more about my use case while you investigate (btw, thanks :) )
<wbn> so the basic idea is
<wbn> the rockpi (and orangepi) can boot off a microSD
<wbn> so right now what we're doing is cloning a "known good" OS/config onto like 16 microSD's at a time
<DigitalMan1983> ah OK
<wbn> BUT, the (rock|orange)pi's can also boot off eMMC storage
<wbn> so what im trying to do is
<wbn> given a (r|o)pi and a blank microSD card, create an image to go on the eMMC storage that starts up, copies itself to the microSD, and then rebootsd
<DigitalMan1983> yes, this is the case with my A64, and that's been the cause of some trouble with some of the board configurations I've tried...gets the order of MMC and eMMC screwed up and shits itself on bootup
<wbn> thus we can speed up the duplication process just by purchasing more eMMC chips
<DigitalMan1983> ahh OK
<wbn> which are way more cost effective than duplicators
<DigitalMan1983> I see
<DigitalMan1983> good thinking :)
<wbn> yeah
<wbn> there is one weird catch
<DigitalMan1983> what's that?
<lanefu> so you want teh devices to boot from sdcard, adn intall to emmc? or the other way around?
<wbn> in that the eMMC that the rockpi uses, and the eMMC that the orange pi uses, are completely different in terms of physical connector
<wbn> lanefu: other way around
<lanefu> so after it boots from emmc, and images sdcard, what happens next
<wbn> actually i have to look into the boot order
<wbn> but i think it's configurable
<wbn> so, assuming the boot order is sdcard -> eMMC
<wbn> the first time it boots the sdcard is uninitialized, so it boots to eMMC
<lanefu> it does vary by soc
<wbn> eMMC flashes the sdcard and reboots
<lanefu> msot of the armbian uboto cnfigs will try all devices
<wbn> then it boots into sdcard
<wbn> and we have a working OS
<lanefu> ha okay
<wbn> take the eMMC out and put it in another board with a blank sdcard
<lanefu> you want the eMMC to be there just for the purpose of imaging the sdcard it uses then boot
<wbn> yep
<lanefu> from a wear and tear perscpetive, frequent moving of emmc is probably suboptimal for that connector
<wbn> do agree but they're cheap
<lanefu> will the sdcard in there always be expected to be blank
LanDi has joined #armbian
<DigitalMan1983> ohhhhh it's a snap in module hahaha I get it. My emmc is just the BGA IC soldered to the board
<wbn> lanefu: hmmm good question
<wbn> i haven't thought about the specifics of recovery yet
<wbn> let's say - if it finds a mountable partition on /dev/mmcblk0p1 - just stop and reboot
<wbn> the vast majority of our hardware failures are the microSD card anyway
<wbn> as you can imagine
<lanefu> bookmark this for later https://www.youtube.com/watch?v=f5S8Er1-5Vw
<wbn> thx
<wbn> so it looks like the eMMC is booted first
<wbn> so maybe like
<lanefu> yeah, you may need to do a custom uboot to try sd first
<lanefu> anyway
<wbn> dd the local image to the sdcard, then blink the LED or something
<wbn> to let the operator know to take the card out
<wbn> the eMMC* out
<wbn> oh custom uboot is also a great idea
<wbn> but my blinkylight idea is also pretty good ;)
<lanefu> yeah that would be the simplest approach
<wbn> and doesn't require me to fork uboot
<wbn> haha
<DigitalMan1983> ohhhh good idea! I actually tried modifying the boot parameters and built a u-boot binary the other day, I'll have to give it a shot
<lanefu> yeah the armbian-fu uses a rather intelligent u-boot command, that inherits addtional cofig from armbianEnv.txt so that you dont have to regenerator boot.scr n stuffs
<DigitalMan1983> I believe what I was doing was trying to select the LCD panel and set the resolution or try to specify a device tree, can't remember which
<DigitalMan1983> I poked through the boot.scr and boot.cmd and the armbianEnv.txt and half understood it :)
<lanefu> wbn: also fyi /usr/sbin/nand-sata-install
<DigitalMan1983> It is looking like, while it is going to be a learning curve for me, it's a LOT less convoluted than Rockchip's boot process
<lanefu> DigitalMan1983: yeah all this stuff is a different ballgame.. it takes a while to wrap your head around it
<lanefu> and teh build engine has some architectural quirks, but it's very flexible and can build for 120 SBCs out of the box
<DigitalMan1983> yeah the build scripts are awesome
<DigitalMan1983> it really blew my socks off
<lanefu> people get frushstrated in one of those ways thats like "gentoo is easier to deal with than debian" lol
<lanefu> wbn: PS pleaes try to stay within the built in frameworks for patching and customzing.. unlike a certain "kali" SBC distro recently
<wbn> i will try not to reinvent any wheels o7
<wbn> but you can trust me not to do that, i'm lazy
<wbn> ok, tuning out now, have a great night everyone
<DigitalMan1983> you too wbn, was nice chatting
<lanefu> cya
<DigitalMan1983> +lanefu are you an Armbian developer?
<lanefu> DigitalMan1983: i think i'm allowed to say that.. i'm more of a generalist within the sphere of armbian
ChriChri_ has joined #armbian
<DigitalMan1983> fair enough :)
ChriChri has quit [Ping timeout: 256 seconds]
ChriChri_ is now known as ChriChri
Tenkawa has joined #armbian
_whitelogger has joined #armbian
Tenkawa has quit [Quit: Leaving.]
AmonX has joined #armbian
<AmonX> Another Armbian question regarding the pinebook pro. Does it handle watching youtube 1080p and 1080p60 videos fine? I had issues using my install and was wondering if I am doing something wrong.
on2 has joined #armbian
<nekomancer[m]> <AmonX "Another Armbian question regardi"> no. not enough speed.
<AmonX> Fiddle sticks guess I am stuck with Ayufans Ubuntu at the moment. Thank you. I did try out Armbian :) it is nice. I use to to restore my pbp.
sunshavi has quit [Ping timeout: 256 seconds]
LanDi has quit [Remote host closed the connection]
ced117 has quit [*.net *.split]
archetech_a has quit [*.net *.split]
AmonX has quit [*.net *.split]
DigitalMan1983 has quit [*.net *.split]
Miouyouyou has quit [*.net *.split]
mpmc has quit [*.net *.split]
bw\ has quit [*.net *.split]
Dr{Wh0} has quit [*.net *.split]
silven has quit [*.net *.split]
mpmc has joined #armbian
archetech_a has joined #armbian
ced117 has joined #armbian
AmonX has joined #armbian
Miouyouyou has joined #armbian
DigitalMan1983 has joined #armbian
silven has joined #armbian
bw\ has joined #armbian
Dr{Wh0} has joined #armbian
ced117 has quit [Max SendQ exceeded]
ced117 has joined #armbian
ced117 has joined #armbian
<DigitalMan1983> Successfully booted to command prompt on my 2GB Olinuxino-A64 using 1GB Lime-A64 config!!! Thanks so much to everyone who helped along the way. Built with Kernel 5.7.15, Debian Buster flavor :)
sassinak-work has quit [Ping timeout: 244 seconds]
sassinak-work has joined #armbian
<ArmbianTwitter> @rogerioff88 (Rogério Figueiredo): @orangepixunlong @OPi_Community @Poddingue @cnxsoft @armbian @linuxdevices Thanks to armbian not you u guys you can't even make a working inf of some kind Android is a bunch of bugs and all Linux distro too (2s ago)
_whitelogger has joined #armbian
sunshavi has joined #armbian
_whitelogger has joined #armbian
<ArmbianTwitter> @OPi_Community (Orange-Pi-Community): @rogerioff88 @orangepixunlong @Poddingue @cnxsoft @armbian @linuxdevices We can criticize the Images but not in this context. They used their reach to show something nice off. (15s ago)
<IgorPec> good morning
<Werner> Morning
<AmonX> Good morning
lanefu has quit [Ping timeout: 240 seconds]
xperia64 has quit [Ping timeout: 240 seconds]
lanefu has joined #armbian
xperia64 has joined #armbian
ced117 has quit [Ping timeout: 240 seconds]
ced117 has joined #armbian
ced117 has joined #armbian
ced117 has quit [Changing host]
indy has joined #armbian
thebigfrog has quit [Ping timeout: 256 seconds]
thebigfrog has joined #armbian
Dr{WhO} has joined #armbian
Dr{Wh0} has quit [Ping timeout: 256 seconds]
Dr{Wh0} has joined #armbian
Dr{WhO} has quit [Ping timeout: 240 seconds]
<archetech_a> gday yall DT overlay changed from on n2 recently lane suspects idk the priority of that being looked at today just fyi nic wont recieve on bullseye-dev build
Dr{WhO} has joined #armbian
Dr{Wh0} has quit [Ping timeout: 246 seconds]
Dr{Wh0} has joined #armbian
Dr{WhO} has quit [Ping timeout: 246 seconds]
xwigg has joined #armbian
<xwigg>
dddddd_ is now known as dddddd
poxifide has quit [Quit: ~_~]
poxifide has joined #armbian
poxifide has quit [Quit: ~_~]
xwigg has quit [Ping timeout: 260 seconds]
xwigg has joined #armbian
xwigg has quit [Quit: Leaving]
xwigg has joined #armbian
scope has joined #armbian
scope has quit [Remote host closed the connection]
xwigg has quit [Quit: Leaving]
Miouyouyou has quit [Quit: Idle timeout reached: 172800s]
<on2> good afternoon
<on2> any improvements with the 5.x kernel regarding the audio for odroids?
xwigg has joined #armbian
scoop has joined #armbian
scoop has quit [Client Quit]
xwigg has quit [Quit: Leaving]
scope has joined #armbian
scope has quit [Quit: Laundry time]
scope has joined #armbian
scope_ has joined #armbian
<scope_> znc for the wins
<scope_> brb
scope has quit [Client Quit]
scope_ has quit [Client Quit]
scope has joined #armbian
sassinak-work has quit [Ping timeout: 272 seconds]
sassinak-work has joined #armbian
hipr_c has joined #armbian
Tenkawa has joined #armbian
indy has quit [Ping timeout: 240 seconds]
indy_ has joined #armbian
indy_ is now known as indy
drobo_00 has joined #armbian
indy has quit [Quit: ZNC - http://znc.sourceforge.net]
indy has joined #armbian
azend has quit [Quit: No Ping reply in 180 seconds.]
azend has joined #armbian
indy has quit [Ping timeout: 256 seconds]
indy has joined #armbian
_whitelogger has joined #armbian
indy has quit [Ping timeout: 240 seconds]
indy has joined #armbian
DaRock has quit [Ping timeout: 256 seconds]
<nekomancer[m]> why history on https://github.com/armbian/build so short?
<IgorPec> some is intentional, some by acciedent
<IgorPec> from this branch it should be better
<nekomancer[m]> hm
<nekomancer[m]> git fetch tags/v20.05
<nekomancer[m]> fatal: 'tags/v20.05' does not appear to be a git repository
<Tenkawa> why are you fetching and not cloning?
<Tenkawa> shouldnt it be git clone -b v20.05
Mony_ has quit [Ping timeout: 246 seconds]
<Tenkawa> I used git clone https://github.com/armbian/build.git -b v20.05
<Tenkawa> (i'm read only though.. I dont know if you are editing or not)
<Tenkawa> add another dirname on the end to call it something besides build
<Tenkawa> ie git clone https://github.com/armbian/build.git -b v20.05 mybuild
<nekomancer[m]> because I have repo clone already. I only need to checkout state of tag
<Tenkawa> well thats not a repo
<Tenkawa> its only a tag
<Tenkawa> what if you just run git fetch by itself>
<Tenkawa> does it sync up?
<PPA> Noob question: if I throw a user patch at the armbian build, do I also need to supply a modified .config file?
<IgorPec> no need, but you can
<PPA> If I don't, does that have averse effects? (e.g. kernel features not registering because they're disabled in the default config)
<nekomancer[m]> yes. after `git fetch origin` a did `git checkout v20.05` successfully
<IgorPec> PPA: use KERNEL_CONFIGIRE="yes" when compiling and you will be promped for changes - if exits
<IgorPec> s/KERNEL_CONFIGIRE/KERNEL_CONFIGURE
<ArmbianHelper> IgorPec meant to say: PPA: use KERNEL_CONFIGURE="yes" when compiling and you will be promped for changes - if exits
<PPA> Ah, so the build program will detect if a userpatch changes something about the kernel, and it'll propose to make appropriate changes to the config file on its own then?
<IgorPec> yes, if there is a need for config change
<PPA> Awesome
<PPA> Thanks
<Tenkawa> git checkout tags/v20.05 worked here but remember it does put the source in a detached state
<Tenkawa> so it will need to be remerged
* Tenkawa misses cvs
Mony has joined #armbian
archetech_a has quit [Read error: Connection reset by peer]
archetech has joined #armbian
<nekomancer[m]> LOL. despite of checkout v20.05, compile script update it to current before compilation
<IgorPec> touch .ignore_changes will help
<nekomancer[m]> IgorPec: thank you. Maybe it be good to be in https://docs.armbian.com/Developer-Guide_User-Configurations/ page?
<IgorPec> its a nice opportunity that you try to add :)
<archetech> heh
<nekomancer[m]> ..мне страшно, я не уверен, как правильно формулировать условия
<ArmbianHelper> I'm scared, I'm not sure how to correctly formulate conditions [ru~>eng]
<archetech> login edit logout
<IgorPec> nekomancer[m]: you can't break anything. just send a pull requeest and if its fine, its merged
<DigitalMan1983_> hi everyone, quick question....I did a KERNEL_ONLY build, and now would like to build a full image using that kernel image with the build scripts if I can. is there a way to do that?
<archetech> docs say rootfs_only I think
<archetech> should read
<Werner> Adjust CLEAN_LEVEL to "make"
Tenkawa has left #armbian [#armbian]
<DigitalMan1983_> thanks guys
<lanefu> archetech: good news
<lanefu> I decompile the DTB filees for N2- and N2+
<lanefu> and vimdiff'd them
<lanefu> and there's absolutely No F-ing difference between them as far as Ethernet stuff goes :-|
<archetech> thats bad news
<lanefu> Does anybody else have an N+ with data retrieve performance issues?
<lanefu> IgorPec: 's automated testing does Iperf.. so kinda think he'd notice if it couldnt download
<lanefu> archetech: ...but you said it downloads fine wiht other distros?
<archetech> yes
<archetech> all other distros
<archetech> and lfs
<lanefu> weird
<lanefu> so my suggestions on where to compare. would be to dump our the kernel configs and compare between us adn them.. and then i guess also compare the DTB as well.... https://michlstechblog.info/blog/linux-decompile-show-a-device-tree-binary-file-dtb/
<archetech> whats weird is its lopsided speed
<archetech> that says config to me
<archetech> either NMgr or boot related
<archetech> thks for the diff least thats checked off
<lanefu> network manager feels like a bit of a stretch to be a cause, but yeah it does dbus magic and who knows
<archetech> ill tru sysd-netd vs NM in a while
<archetech> try
<archetech> getting lfs ready for weston
<archetech> we also know that ubu 20.08 works so its maybe deb 11
<lanefu> archetech: Ooooooooo I _didn't_ know it works on ubuntu
<lanefu> forreal?
<lanefu> man that's a whole new level of linuxweird
<archetech> and Tekakat(sp) has deb 11 on his plain N2
<archetech> I didnt hack anything either its plain bulls install plus the pkgs you gave
<archetech> those didnt break it it was broke
<lanefu> hmm
<lanefu> is it possible you're broken on the inside?
<lanefu> just looking at all avenues here
<archetech> we can shelve it too until another build I do or a n2plus comes along
<lanefu> j/k
<lanefu> anyway yeah i'd agree that best to wait for a second opinion on another N2+
<archetech> you could build a bulls-dev and run it on yours too
<lanefu> ahh yeah good call
* DigitalMan1983_ is broken on the inside for sure
<archetech> I am too and contrite
Toast has joined #armbian
fizikz has quit [Ping timeout: 256 seconds]
<wbn> hello again
<wbn> today i may attempt to configure my armbian build
<wbn> if i'm in the mood for it
<wbn> how's it going DigitalMan1983_
Toast has quit [Quit: Toast]
MrFixIt has quit [Remote host closed the connection]
MrFixIt has joined #armbian
<DigitalMan1983_> hi wbn
<DigitalMan1983_> I was able to get a working image built yesterday for my Olinuxino-A64
jl_678 has joined #armbian
<jl_678> Hi all
<jl_678> I have a weird problem on oDroid HC2 running Armbian Debian.
<jl_678> Anyway, I see packet loss on the network interface. Everything seems to work okay, but the large number of dropped packets is troubling. I went to the 5.x and it was much better, but I had a ton of other problems.
<jl_678> Example: In the last 20 minutes, I show 700 packets dropped which seems excessive....
* archetech hmmm debian
<archetech> try an iperf3 test both ways?
<jl_678> Meaning to and from the device from another connected device?
<archetech> yes on the lan
<archetech> and web too
<jl_678> As an added data point, I only see dropped packets on RX and nothing on TX. I have swapped cables.
<jl_678> I cannot recall if I have swapped switch ports and will try that.
<archetech> im having weirdness on my N2 w/ deb bullseye armbian
<archetech> nic
<jl_678> Dropping too?
<archetech> can barely receive
<jl_678> Hmmm, I found 5.x kernel to be better for packet loss, but it introduced other huge network issues.
<archetech> mines 5.7.15
epsilonKNOT has quit [Ping timeout: 240 seconds]
<jl_678> I tried 5.7.5 and 5.7.7. My use case is network intensive and in both cases, under load the network interface just disappeared. Literally the port was just dropped from the system.
<jl_678> A reboot solved it, but this was a deal breaker for me.
<jl_678> Oh, I also found that if you unplug the LAN from the HC2 and plug it back in, the interface would disappear and not come back.
<jl_678> That second problem related to 5.7.5 and not 5.7.7.
<lanefu> jl_678: what SoC is HC2?
<archetech> odroid
<archetech> cpu?
saurabh009 has joined #armbian
<saurabh009> Hi, has anyone tried to enable the touch screen xpt2046 with the armbian ?
<lanefu> saurabh009: probably best to ask that question on forums
<jl_678> Exynos5422
<lanefu> don't get too much into accessories here.. mostly kernels
<jl_678> Oh and I have 5 of them and only two are showing this behavior.
<jl_678> Any particular settings in iPerf3?
<saurabh009> Can't ask things on Armbian IRC ?
<lanefu> jl_678: k where did you get the 5.7 kernels?
<lanefu> saurabh009: you can ask whatever, i'm just telling you where you'll get the bet answers
<lanefu> you're of course welcome here still
<jl_678> armbian-config
<lanefu> jl_678: k you switched to "dev" kernel for 5.7?
<jl_678> It was just there as an option and so I tried it. It was better in that it did not drop the network connection on unplug.
<jl_678> Instead, it seemed to drop the connection on load....
<jl_678> (e.g. network load.)
<lanefu> okay yeah
<lanefu> i've got 3 Exynos boards with different kernels
<lanefu> just iperf3d htem backwards adn fowards
<lanefu> :( http://ix.io/2uaW
<lanefu> that's iperf -R (receive), and then iperf pushing
<lanefu> so yeah outband transfer out of order for sure
<lanefu> BUT>. also on bullseye
<lanefu> damn
hipr_c has quit [Ping timeout: 245 seconds]
<jl_678> So you see it too.
<jl_678> Any idea on what needs to be done to address it? I have no idea if there are specific optimizations or something.
<lanefu> only breadcrumb i can give you here's a optimization for receiee packet steering https://github.com/armbian/build/blob/master/packages/bsp/odroid/90-builtin-net-rps.rules
<lanefu> understanding how that works might help dig a little deeper
<lanefu> otherwise gonna have to go deep
<lanefu> jl_678: what distro / releaes are you using'
<jl_678> Deb Buster
<jl_678> Kernel 4.14.187
<jl_678> Truthfully, that is a little deep for me and I am not comfortable going too nuts.
jl_678 has quit [Remote host closed the connection]
fizikz has joined #armbian
Hokedli has joined #armbian
jl_678 has joined #armbian
<jl_678> Hey sorry, I am back. My laptop went to sleep and dropped the connectiom
<jl_678> I take it that there is not a lot that can be done with this issue unless I want to get my hands dirty.
<DigitalMan1983_> any plans to incorporate kernel 5.8 into build scripts in the near future?
<DigitalMan1983_> (I suppose I could try building my own mainline, but that's work and stuff)
jl_678 has quit [Remote host closed the connection]
fizikz has quit [Ping timeout: 264 seconds]
<lanefu> DigitalMan1983_: 5.8 will the new "dev" kernel for most SoCs in near future
<lanefu> on a SoC by SoC basis
<lanefu> they're wrkng on sunxi right now.. not sur about rockhchip
<lanefu> you shoudl see conversatioon on the forums
MrFixIt has quit [Remote host closed the connection]
<DigitalMan1983_> cool, thanks
fizikz has joined #armbian
MrFixIt has joined #armbian
Hokedli has quit [Quit: Konversation terminated!]
* lanefu needs to try to type slower
* lanefu juggling too many things
fizikz has quit [Quit: Leaving.]
fizikz has joined #armbian
drobo_00 has quit [Quit: drobo_00]
archetech has quit [Quit: Konversation terminated!]
archetech has joined #armbian
mrueg has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
mrueg has joined #armbian