00:23 <\x> hurricos: just in time for no NAT november

01:10 <russell--> where did lan6 come from in /etc/config/network? i'm seeing it on my mr24

01:23 <\x> I wonder which drivers supports this https://w1.fi/cgit/hostap/commit/?id=3c79173c328e005da03c8115e10ff5e58f28055d

01:39 <russell--> git grep lan6 doesn't turn up anything relevant, afaict

01:44 <slh> \x: scroll to the very bottom of the patch, QCA_WLAN_VENDOR_FEATURE_HT_VHT_TWT_RESPONDER is kind of telling ;)

01:44 <\x> slh: i wonder if they gonna backport it to their mac80211 drivers

01:45 <slh> good question, I fear I can guess that answer

01:45 <\x> slh: i tried enabling it on ipq60xx, like set 2.4ghz as N and i dont see twt reponder on beacon so yeah, it aint there

01:45 <slh> well, that kind of development at this probably targets qcn6124

01:46 <slh> so wifi7/ 11be

01:57 <\x> slh: maybe its only available on their nl80211 drivers (vendor ones)

02:02 <slh> probably

02:03 <\x> or maybe its another firmware lotto

02:03 <\x> like on ipq40xx's ftm feature

02:05 <\x> like it only works on 3.2.x firmware (tested on 3.2.1.1), but not on 3.6 which is uploaded onto linux-firmware

02:06 <slh> it's probably quite some time, before we get to see ipq957x support, to really know

05:34 <owrt-images-builds> Build [#405](https://buildbot.openwrt.org/images/#/builders/153/builds/405) of `master_bmips/bcm6328` failed.

05:34 <owrt-images-builds> Build [#359](https://buildbot.openwrt.org/images/#/builders/237/builds/359) of `master_bcm27xx/bcm2712` failed.

06:53 <owrt-images-builds> Build [#406](https://buildbot.openwrt.org/images/#/builders/167/builds/406) of `master_bmips/bcm6362` completed successfully.

08:17 <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.)

08:35 <f00b4r0> nbd: hi, did you get a chance to find out what's wrong with the 23.05 mt76 backport?

09:04 <nbd> sorry, didn't have time for that yet

09:11 <russell--> aha, i understand (kind of) why i'm seeing a lan6 ... package/base-files/files/bin/config_generate is matching the lan interface with proto dhcp on mr24

09:14 <dwfreed> russell--: sounds like the board.d/02_network for it is wrong

09:16 <dwfreed> or have you customized it so it comes up as a dumb AP with dhcp?

09:18 <russell--> it's a single ethernet interface, i think the idea is that for initial configuration the vanilla ethernet interface has to be accessible (not firewall'd, so LAN) ... i think once upon a time it provided a dhcp server. i'm overriding the vanilla configuration, so any changes tend to break my custom fixups

09:20 <russell--> target/linux/apm821xx/base-files/etc/board.d/02_network has ucidef_set_interface_lan "eth0" "dhcp"

09:21 <russell--> this commit has the justification: abb4083fa068

09:22 <dwfreed> that's not a justification

09:28 <russell--> i didn't offer it and don't defend it, i just reported what was presented

09:59 <f00b4r0> nbd: ok no problem. I wonder if it would also help with MT6000. We're seeing repeated crashes in 23.05.5. Instrumentation to be added shortly to get crash logs.

10:02 <owrt-images-builds> Build [#316](https://buildbot.openwrt.org/images/#/builders/238/builds/316) of `master_d1/generic` failed.

10:25 <xback> nbd: Hi, since the backports bump to 6.11 i'm seeing these splats on boot on ipq40xx combined with ath10-ct. any idea why? :-) it seems all functional for the rest. https://pastebin.com/raw/MdTvvNGw

10:27 <\x> i also have ipq4019, best combo for perf and stability is ath10k + 3.2.1.1 firmware from quic repo

10:28 <\x> https://git.codelinaro.org/clo/ath-firmware/ath10k-firmware/-/tree/main/QCA4019/hw1.0/3.2.1.1

10:29 <xback> true, but I need WDS, which does not work on native ath10k :-)

10:29 <\x> ah yeah

10:29 <\x> no AP/VLAN

10:29 <\x> 11s works though no idea on apup, havent tried

10:30 <xback> looks like upstream is really obsessed with default AP/STA since ath10k and later. all other modes get unsupported (like IBSS, etc)

10:30 <\x> even on current ath11k theres no ap/vlan

11:41 <schmars[m]> i mean, IBSS has been deprecated for like a decade

11:41 <schmars[m]> but lack of ap/vlan sucks - come to mediatek land, the grass is green here :-)

11:42 <f00b4r0> when it works ;P

11:42 <schmars[m]> hehehe

11:43 <schmars[m]> you have pretty concrete reliability requirements

11:43 <f00b4r0> i think "not crashing" is a pretty basic one, no? ;P

11:45 <robimarko_> Now you are asking much :)

11:45 <schmars[m]> i know i know - i'm just happy that it recovers itself, even if there's downtime of ~15s. for APs somewhere in the public that's fine. and with recently with mediatek we get so much for so little $$

11:46 <f00b4r0> schmars[m]: I'm having other troubles with mt7986 though now ;(

11:46 <f00b4r0> robimarko: heh ;)

11:47 <schmars[m]> yeah godspeed man :(

11:47 <f00b4r0> have to say it's fairly frustrating to have such a poor experience with what is supposed to be the "best in class" platform

11:48 <schmars[m]> yeah i feel you - for me its even more frustrating when the drivers come from the unaccountable powers that be, and i can only hope for fixes/improvements in the next release eventually

11:48 <schmars[m]> n b d is a godsent in comparison :-)

11:49 <f00b4r0> ath9k forever :)

11:49 <schmars[m]> legends never die

11:49 <f00b4r0> rock solid, amazing range, now that was something

11:50 <schmars[m]> the mesh story though... "deaf node" problem, not everything was great

11:58 <\x> f00b4r0: is it time to hoard all those AR9380's

12:07 <nbd> xback: wiphy->retry_long is initialized to 4 by default in cfg80211. mac80211 tells the driver to set it as long retry count via drv_config. driver returns an error code because of firmware capabilities (see warning)

12:14 <\x> https://www.youtube.com/watch?v=F4vUY33acW0 mt7981, 2.5GbE rj45 + SFP + 1GbE rj45

12:16 <\x> thats some weird io setup but ok

14:39 <xback> nbd: Thanks for your insights :-)

14:41 <schmars[m]> anyone aware of an mt7530-mdio issue, where the link suddenly goes down, and then recovers after anything between 2 seconds and multiple minutes?

14:48 <Slimey> https://www.youtube.com/watch?v=m3Ny3j5nH0U

14:49 <schmars[m]> nbd: the god damn 7th :) SYS_RESET_COUNT = 7 and this time recovery failed without any obviously useful messages - logread+sysrq: https://gist.github.com/pktpls/6e4a483a34337f20cf0888bc6934dd59

14:50 <schmars[m]> device is the same you still have access to - 2001:bf7:830:c001:bad5:26ff:feff:9fbc

14:50 <schmars[m]> we'll unfortunately have to tear that site down next week

15:03 <f00b4r0> Slimey: cool, albeit probably very expensive toy ;)

15:19 <nbd> schmars[m]: sorry, seems that i accidentally crashed it while looking for information

15:20 <schmars[m]> :-)

15:21 <schmars[m]> if there's any more info i could grab automatically in the next few days, let me know. i've been meaning to set up remote logging for a while

15:23 <nbd> i don't have any ideas about that at the moment

18:27 <owrt-images-builds> Build [#410](https://buildbot.openwrt.org/images/#/builders/206/builds/410) of `master_bcm47xx/legacy` failed.

21:36 <owrt-images-builds> Build [#406](https://buildbot.openwrt.org/images/#/builders/153/builds/406) of `master_bmips/bcm6328` completed successfully.

22:27 <Mangix> aparcar: ping

23:30 <KGB-0> https://tests.reproducible-builds.org/openwrt/openwrt_omap.html has been updated. (11.1% images and 100.0% packages reproducible in our current test framework.)