autkin has left #openwrt-devel [#openwrt-devel]
Mangix has quit [Read error: Connection reset by peer]
Tapper has quit [Ping timeout: 480 seconds]
SpectreDev_01 has quit [Quit: Connection closed for inactivity]
tSYS has quit [Quit: *squeak*]
tSYS has joined #openwrt-devel
dansan has quit [Ping timeout: 480 seconds]
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
DonkeyHotei has quit [Ping timeout: 480 seconds]
Mangix has joined #openwrt-devel
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
DonkeyHotei has joined #openwrt-devel
rua has quit [Quit: Leaving.]
rua has joined #openwrt-devel
<KGB-2> https://tests.reproducible-builds.org/openwrt/openwrt_x86.html has been updated. (100.0% images and 100.0% packages reproducible in our current test framework.)
Borromini has joined #openwrt-devel
Borromini has quit [Quit: Lost terminal]
SpectreDev_01 has joined #openwrt-devel
<SpectreDev_01> PaulFertser: yeah well this is interesting it seems to reboot at 4am, it reboot yesterday also at 4am
<PaulFertser> SpectreDev_01: are you recording the serial log?
<PaulFertser> Also, are you checking for new leaks automatically?
<PaulFertser> Is it possible it rebooted due to other reasons such as power glitch on mains?
<SpectreDev_01> Well I cant since it's rebooted at 4am while I was asleep
<SpectreDev_01> Ive not had anything else reboot, so I doubt it's a power glitch
<PaulFertser> You can leave some automation running to store the log and to run the check periodically.
<SpectreDev_01> Hmm ok
<PaulFertser> It's odd how it can depend on time though
<PaulFertser> I mean absolute time
<SpectreDev_01> Yeah when I check the uptime it's basically 5hrs and 20mins and I wake up at 9am so it's basically around 4am, but I was streaming last night at 9pm no reboot it ran fine for 9hrs I go to sleep and boom it's rebooted itself while not even in use
<SpectreDev_01> But I'll get a build up with usb and have it save the logs to the usb
<PaulFertser> Hm, why not just leave another computer running and saving from serial?
<PaulFertser> That would be more reliable
<PaulFertser> It might be an oops that makes usb inoperable.
<SpectreDev_01> Ah true, I'll get my pi attached and have it log it then, since that's on 24/7 anyways
<mrkiko> SpectreDev_01: what target/device?
<PaulFertser> Yes, and you can have it cat the kmemleak file to "| logger" periodically
<SpectreDev_01> mrkiko: Linksys MX4200 IPQ8174 basically IPQ807x
danitool has joined #openwrt-devel
<mrkiko> SpectreDev_01: thanks!
<PaulFertser> SpectreDev_01: do you have time to hack that script to not emit bss_color specification so that hostapd wouldn't enable that feature and retry with HE40 ?
<PaulFertser> (to not emit the unrecognised config item when it's disabled)
<SpectreDev_01> PaulFertser: got plenty of time, first just gotta flash kmemleak build and then we can do it
<SpectreDev_01> Just give me about 5-10 mins
<PaulFertser> I'm not in a hurry
autkin has joined #openwrt-devel
mrkiko has quit [Quit: leaving]
<SpectreDev_01> PaulFertser: all ready to go
<PaulFertser> SpectreDev_01: then I'd try disabling color in config again and revert other settings to those when you could easily reproduce the issue.
<SpectreDev_01> PaulFertser: that config don't work, it keeps saying invalid, I also tried config he_bss_color_disabled '1' which hostapd just ignores
<SpectreDev_01> Hmm weird in the wireless overview section it doesn't say the modem name just says Qcom Atheros IPQ8074
<SpectreDev_01> Unless that's just intended to be done like that
<PaulFertser> SpectreDev_01: with invalid config I suggest you comment out the line in the script that adds the invalid option.
<PaulFertser> SpectreDev_01: so that uci would not be adding any bss_color options to hostapd.conf at all.
<SpectreDev_01> Ehm where's the script at
<PaulFertser> 12:59 < PaulFertser> So probably the way to try is to make the script not emit he_bss_color option and not emit he_bss_color_disabled either.
<PaulFertser> 12:59 < PaulFertser> SpectreDev_01: /lib/netifd/wireless/mac80211.sh
<SpectreDev_01> Ah I Found he_bss_color_enabled 1 in the script
<PaulFertser> When I was looking at it I saw it adding he_bss_color_disabled to config trying to disable it.
<PaulFertser> But you say hostapd doesn't understand he_bss_color_disabled so the idea is to not emit it.
<SpectreDev_01> Yeah I'll have to comment it out completely
Tapper has joined #openwrt-devel
<SpectreDev_01> PaulFertser: yep that did the trick fully commented out and it doesn't show he_bss_color=128 in config when it boots up
<PaulFertser> SpectreDev_01: so let's see how it goes with HE40...
<PaulFertser> SpectreDev_01: btw, why are you using 2.4 GHz band?
<SpectreDev_01> Well I have 1 camera and a smart laundry that connect to 2.4ghz only unfortunately otherwise I'd just disable 2.4ghz
<PaulFertser> Do they support 802.11ax?
<SpectreDev_01> Ehm I don't think the camera does
<SpectreDev_01> Not sure on the washing machine
<PaulFertser> I'd say unlikely.
<PaulFertser> It's just that you mentioned HE40 as if it was actually used by anything.
rua has quit [Quit: Leaving.]
DonkeyHotei has quit [Ping timeout: 480 seconds]
<SpectreDev_01> Oh right yeah 1 of my super old laptop connects to 2.4ghz, but that's gonna go soon
<SpectreDev_01> PaulFertser: yeah robimarko responded over forum and said that's a 1 time kernel memory leak
GNUmoon has joined #openwrt-devel
DonkeyHotei has joined #openwrt-devel
autkin has left #openwrt-devel [#openwrt-devel]
fakuivan has quit [Remote host closed the connection]
fakuivan has joined #openwrt-devel
parazyd has quit [Ping timeout: 480 seconds]
parazyd has joined #openwrt-devel
rua has joined #openwrt-devel
ienke has joined #openwrt-devel
<ienke> 2/16MiB is easily doable, dropping support for 4/16MiB is absolute bullshit
<ienke> v6.x.x is going to suck the life out of more boards, but this is just the decline
<ienke> >12:01 <neggles> if all else fails, openwrt chroot :P
<ienke> pivot_root/switch_root must be hard for US people, inshallah
<ienke> Ansuel: you should hang yourself, you nigger cattle, you are why openwrt is failing
ienke has quit [autokilled: This host violated network policy. Mail support@oftc.net if you think this is in error. (2023-12-17 12:56:29)]
<SpectreDev_01> PaulFertser: caught the OOM log katb.in/avakebijulu
<PaulFertser> SpectreDev_01: and is something unreclaimable in /proc/meminfo increasing?
<SpectreDev_01> I'll have a look but this didn't cause a reboot for some reason
<SpectreDev_01> I had To manually reboot
<SpectreDev_01> PaulFertser: 76352kb unreclaimable
<SpectreDev_01> Doesn't seem to increase
<PaulFertser> SpectreDev_01: hm, looks like the DMA string from our log increased over that time
<PaulFertser> [ 7419.803782] DMA: 530*4kB (UME) 1011*8kB (UME) 489*16kB (UME) 407*32kB (UME) 48*64kB (UME) 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 34128kB
<PaulFertser> [ 7422.982400] DMA: 774*4kB (UME) 1028*8kB (UME) 482*16kB (UME) 394*32kB (UME) 46*64kB (UE) 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 34584kB
<SpectreDev_01> Hmm
<PaulFertser> SpectreDev_01: and that with kmemleak enabled and not reporting anything, right?
<SpectreDev_01> Nope I've enabled hss color no kmemleak
rua has quit [Quit: Leaving.]
rua has joined #openwrt-devel
<SpectreDev_01> PaulFertser: yep it did it again, OOM
goliath has joined #openwrt-devel
<SpectreDev_01> It rebooted again
<PaulFertser> SpectreDev_01: my idea was to somehow pinpoint where the leak happens.
<PaulFertser> SpectreDev_01: I have an idea: after startup call periodically (say, every minute) echo m > /proc/sysrq-trigger
<SpectreDev_01> Hmm ok
<SpectreDev_01> Katb.in/ifeyulizexe
<SpectreDev_01> PaulFertser: here we are
<hauke> &bu5
<SpectreDev_01> ?
<PaulFertser> SpectreDev_01: yes, this might be useful, just make it run every minute and collect full log till oom.
<PaulFertser> SpectreDev_01: also, check kmemleak periodically.
<PaulFertser> And show it all to robimarko :)
<enyc> Hrrrrrrrrrrrrrrrm
<enyc> wrong chanen sorry
torv has quit [Remote host closed the connection]
torv has joined #openwrt-devel
<SpectreDev_01> PaulFertser: katb.in/itiojozi yeah I think it fills up because the packets end up becoming pending and eventually just causes OOM the system hangs
danitool has quit [Remote host closed the connection]
<PaulFertser> SpectreDev_01: the last link doesn't work.
<SpectreDev_01> Oh crap
<SpectreDev_01> Well it was about packets pending
<SpectreDev_01> Been running well for 35mins so far
hanetzer1 has joined #openwrt-devel
hanetzer has quit [Ping timeout: 480 seconds]
danitool has joined #openwrt-devel
minimal has joined #openwrt-devel
xes has joined #openwrt-devel
<SpectreDev_01> PaulFertser: yeah I removed to patches related to packet flushing and get some kernel warning now katb.in/vopakojukij
<SpectreDev_01> As for the above link it's here katb.in/itiojoxozi
<SpectreDev_01> As for the above link it's here katb.in/itikojoxozi
<SpectreDev_01> Yeah that didn't last long, it was 25% free ram until I hit 1hr mark
<SpectreDev_01> Oh wait those patches didn't exist hmm
Danct12 has quit [Quit: ZNC 1.8.2 - https://znc.in]
Danct12 has joined #openwrt-devel
<swalker> updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html
Shados has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
<Ansuel> well guys it's happening... first PHY network driver written in rust
<tidalf> cool
Shados has joined #openwrt-devel
tidalf is now known as Guest10757
Guest10757 has quit [Ping timeout: 480 seconds]
<Mangix> Ansuel: what hardware uses it?
<Mangix> 135 lines of code...sounds barebone
DonkeyHotei has quit [Read error: Connection reset by peer]
<minimal> must be a very old network card if its rusty ;-)
<Mangix> hrm industrial.
<Mangix> this driver looks so barebone.
<schmars[m]> released 2000, but there's a ton of stuff that doesn't need the latest tech
tidalf has joined #openwrt-devel
fakuivan has quit [Ping timeout: 480 seconds]
<Mangix> LMAO "IMHO - Rust is for programmers that failed at learning to write proper C production code ... since I doubt that anyone can write proper C++ production code."
SpectreDev_01 has quit [Quit: Connection closed for inactivity]
<Mangix> it's interesting though. now two drivers have to be maintained for no good reason.
GNUmoon has quit [Remote host closed the connection]
f23xhxxx has joined #openwrt-devel
<Mangix> oh this is funny. original driver is 131 lines of code
<Mangix> *rust
tidalf is now known as Guest10767
Guest10767 has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
SpectreDev_01 has joined #openwrt-devel
<SpectreDev_01> That's pretty cool rust PHY
tidalf has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
Tapper has quit [Quit: Tapper]
tidalf is now known as Guest10776
Guest10776 has quit [Ping timeout: 480 seconds]
DonkeyHotei has joined #openwrt-devel
<SpectreDev_01> PaulFertser: on 23.05 after an hour I now get decent free memory around 80mb
DonkeyHotei has quit [Read error: Connection reset by peer]
DonkeyHotei has joined #openwrt-devel
tidalf has joined #openwrt-devel
tidalf is now known as Guest10778
Guest10778 has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
tidalf has quit [Ping timeout: 480 seconds]
tidalf has joined #openwrt-devel
goliath has quit [Quit: SIGSEGV]
tidalf is now known as Guest10786
Obi-Wan has quit [Remote host closed the connection]
Guest10786 has quit [Ping timeout: 480 seconds]
Obi-Wan has joined #openwrt-devel