00:20 <Mangix> stintel: replaced with what?

10:35 <ynezz> while comparing config diffs of https://github.com/openwrt/openwrt/pull/12917 at https://gist.github.com/ynezz/4372b7900bb426ee23aa5d47c36cf45a I've noticed, that `+# CONFIG_TARGET_ROOTFS_JFFS2_NAND is not set` appeared due to https://github.com/openwrt/openwrt/pull/12917/commits/1bf6021705de1604f17634ff8d8b4fb8e2451590

10:36 <ynezz> makes me wonder if we should remove support for JFFS2_NAND completely as there seems to be no users and folks should be using UBI for NAND flashes

10:39 <stintel> KanjiMonster: I see you're already reviewing PR from the Apple / CSA guy, cool!

10:59 <ynezz> ugh that rabbit hole is so deep, now I'm wondering if anyone is actually using mediatek/mt7629 devices, seems like the WiFi is crippled from the start? :P I cant see wpad-basic nor mt76 kernel drivers being included

11:17 <stintel> hehe

12:00 <stintel> closing in on the hostapd commit that causes #13156

12:01 <stintel> I suspect it might be https://w1.fi/cgit/hostap/commit/?id=e978072baaca886943c391b9a99066771ddc01a8

12:06 <stintel> after this I might have a go at sending BSS Transition Request with BSS Termination bit set to all stations when doing sysupgrade for example

12:06 <stintel> to allow clients to gracefully migrate

12:12 <stintel> also this dashboard I made while I was doing Wi-Fi stuff at Meta helps in triggering the bug: https://codeberg.org/stintel/usteer-web-ucode

12:13 <stintel> beats manually constructing the ubus call to send the BSS Transition Request

13:01 <stintel> e978072baaca886943c391b9a99066771ddc01a8 is the first bad commit

13:13 <stintel> ynezz: do you want me to CC you in the mail I'm sending to hostap ML ?

13:15 <ynezz> stintel: yep, thanks, nice catch

13:16 <stintel> doesn't revert cleanly unfortunately

13:17 <stintel> ugh, moderator approval

13:17 <stintel> oh well I CC'd the person who introduced the bug, let's hope they come up with a fix soon

13:18 <stintel> because this bug is seriously annoying

13:18 <stintel> happens all the time

13:21 <stintel> I suspect prune_associations will need to be updated

13:31 <stintel> dangole: xdp-utils fails to build for me because bfpeb doesn't support -stack-protector-strong, even if I disable stack protector completely in menuconfig, for both userspace and kernel... any ideas ?

13:32 <stintel> something must be leaking that but I fail to figure out what

14:00 <hauke> ynezz: I think we do not support the wifi included in mt7629

14:00 <dangole> stintel: hm, haven't tried xdp-utils in a while, but will get there soon when it comes to support XDP with MT7988. Can you give me a hint on how to reproduce this? @PolynomialDivision may have done something with xdp-utils more recently than me, maybe ping as well...

14:00 <dangole> hauke, ynezz: correct. we don't support mt7629 wifi at all.

14:01 <stintel> dangole: I just enabled xdp-utils (as I'm working on suricata 7 which support af_xdp which requires libxdp), I am using host clang 16, might be related to that too

14:01 <stintel> last change was remove autorelease, then your bump to 1.2.9

14:04 <ynezz> hauke, dangole: ah, thanks!

17:40 <Slimey> heh

17:46 <rmilecki> Ansuel: NVMEM access to EEPROM works great

17:46 <rmilecki> thanks for your work on that

17:48 <Slimey> i heard you like cpus https://cdn.discordapp.com/attachments/561746595067068418/1134542642974294056/0728231245.jpg

19:26 <fda> jow: the overview page is broken: https://ibb.co/kBCVzmp for avm7320 - but NOT for rpi2, e8450, avm1200

19:26 <fda> "TypeError" - "stats is undefined"

19:27 <fda> maybe this is because avm7320 has 2 lan ports, but only 1 works with openwrt. and this 1 then only supports 1gbit and no 10/100

19:27 <fda> on the other 3 devices i noticed the new (and ugly) "Port status"

19:27 <stintel> nbd: so basically you're suggesting I do merge https://git.openwrt.org/?p=openwrt/staging/stintel.git;a=commitdiff;h=7a02662469e72fde013fd7fd8e9b56f63e3d4665 ?

19:28 <stintel> if you confirm, I'll update the commit message in the patch to explain a few things, including what you said about mt7915

19:39 <fda> jow: workaround: "rm /www/luci-static/resources/view/status/include/29_ports.js" and it overview works again: https://ibb.co/VN3k71x

19:39 <fda> please highlight if i should test something

19:42 <fda> btw, why does openwrt cripple the js files??? file system with compression dont need much space for space/tab/newline but edit such file is a pain

19:42 <stintel> dangole: looks like I found a solution for the xdp-tools stack protector issue

19:45 <stintel> dangole: https://git.openwrt.org/3daf15caf827d8fc029f0d6c1bd6588c4dd6ab35

20:12 <stintel> suddenly those 4GB RAM in the M300 come in handy with suricata using 1.3G :)

21:10 <Slimey> lol

21:11 <Slimey> DDR3-ECC ?

21:13 <Slimey> i kid you not i must have pulled a TB of ram from decommissioned servers

21:14 <stintel> :)

22:04 <stintel> Slimey: ECC yes, dunno if it's DDR3