07:21 <stintel> Mangix: mdnsresponder - 2200.100.94.0.2-r1

15:14 <mrkiko> rmilecki: do you happen to have an EA9500 around?

15:15 <Slimey> yankee hotel foxtrot

15:48 <nbd_> f00b4r0: did it crash earlier or later than expected, or about the same?

16:24 <Mangix> stintel: so it works. that's good.

17:07 <Mangix> Ansuel: ping

17:07 <Ansuel> ?

17:07 <Mangix> Ansuel: https://github.com/openwrt/openwrt/blob/main/target/linux/mpc85xx/p1010/config-default <-- this should be using the new QCA phy, no?

17:08 <Mangix> it uses a QCA8327 switch

17:14 <Mangix> OK, that's two non working devices on this target. lovely.

17:14 <Ansuel> Yes It should

17:19 <Mangix> Ansuel: should I add extra commits to that PR fixing this mess or a new PR?

17:20 <Ansuel> Maybe a new pr is better

18:19 <f00b4r0> nbd: it's highly random so I can't say for sure. It crashed within 48h AFAICT, which is on the faster side of things, but not the fastest (we have records of crash in a few hours only)

18:19 <f00b4r0> do you need anything from the device or can I reboot it?

18:28 <nbd> you can reboot it. can you please enable wm firmware debug for the next round?

18:28 <nbd> i will let you know when i've updated the branch for testing

18:36 <f00b4r0> can you remind me how to do that please, I can't remember

18:42 <Mangix> Ansuel: posted PR. I'd look into git grep CONFIG_AT803X_PHY=y. A lot of these seem to need conversion.

18:44 <Mangix> From a quick look, target/linux/ipq806x/config-6.6 , target/linux/qualcommax/config-6.6 , target/linux/ramips/mt7621/config-6.6

18:58 <robimarko> Mangix: Did you manually update it?

18:58 <robimarko> Cause, QCOM_NET_PHYLIB should be selected as well

19:07 <Mangix> robimarko: manual, yes

19:08 <Mangix> I noticed make kernel_oldconfig doesn't really work with these config-default files

19:10 <Mangix> funny, description for QCA83XX_PHY is wrong

20:12 <owrt-images-builds> Build [#335](https://buildbot.openwrt.org/images/#/builders/205/builds/335) of `master_mpc85xx/p1020` completed successfully.

20:13 <rmilecki> mrkiko: i never had EA9500

20:39 <dermoth> Hi... I'm curious if anyone knows what i'm missing exactly, I used to get a /48 from my ISP, set very specific 4-nibble prefixes for each of my two LANs and let dhcp/autoconf assign /64 to hosts. Some hosts there get a static suffix that is either used for internal communication on the ULA or used for forwarding rules and dynamic DNS...

20:42 <stintel> meh, my switch started randomly rebooting: https://gist.github.com/stintel/80e4417e504c74dd561e9d2e5cb2946b

20:42 <stintel> seems to be realtek-otto-watchdog 18003150.watchdog: phase 1 timeout

20:43 <stintel> but also a bunch of rcu: INFO: rcu_sched detected stalls on CPUs/tasks:

20:43 <dermoth> My service was recently transfert as part of ISP buyout, I now get a /56.... So I have only 2 nibbles to play with on the lans, but no mater what I tri the LAN interfaces use 0 and 0x100 respectively, so on one I could just change the prefix mask / leave the last two nibbles blank, for the other the interface it gets ULA only, and no external address, since the 3rd nibble is past of the /56

20:43 <dermoth> *part of the /56

20:46 <dermoth> For the nibbles, I'm talking of the "IPv6 assignment hint" - which I used to get specific and distinct prefixes on the lans... so it was either:

20:46 <dermoth> ULA (/48) + 4 nibbles + /64

20:46 <dermoth> Public (/48) + 4 nibbles + /64

20:47 <dermoth> Now I would want to have:

20:47 <dermoth> Public (/56) + 2 nibbles + /64

20:47 <dermoth> ULA (/48) + 00 + 2 nibbles + /64

20:49 <stintel> sounds like you need option ip6hint

20:49 <dermoth> I'm not sure if I should change the ULA prefix, or use the ipv6 prefix filter... (and just realized now I missed the part where I specific the requested prefix length, which clearly is being ignored by my isp now)

20:50 <dermoth> stintel: it's exactly what I used, and it worked when I was receiving a /48 - now with a /56 even if I trim it town it's being ignored

20:50 <stintel> it works for me with a /60

20:51 <stintel> https://gist.github.com/stintel/9f161596fcfe75a63b649ec5448ce815

20:51 <dermoth> So you have 1-hex hint that end up part of your ip address correct?

20:51 <stintel> yes

20:52 <dermoth> Did you set your ULA to /60 as well?

20:52 <stintel> I do not use ULA

20:52 <dermoth> ah....

20:52 <dermoth> I guess i'll try that

20:52 <stintel> could be a bug where something doesn't like ULA vs public of different lengths, so yeah could be worth trying

20:54 <dermoth> oh you know what, I think I accidentally set the interface delegation to /56, I guess I shouldn't play with the settings when I'm too tired!

20:55 <dermoth> obviously the first interface would get the whole space and nothing left for the 2nd

20:56 <stintel> hah, that could also explain things ;)

23:23 <KGB-2> 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.)

23:41 <schmars[m]> is this new? https://www.tp-link.com/us/support/gpl-code/

23:42 <schmars[m]> shit ton of code dumps. grabbed a random omada realtek switch and it has kernel, uboot, buildroot, and some kind of sdk

23:42 <schmars[m]> * realtek switch tarball (260 MB) and it