**** BEGIN LOGGING AT Sun Nov 04 02:59:59 2018 Nov 04 03:35:30 CS undergrad student wants to contribute https://forum.openwrt.org/t/cs-undergrad-student-wants-to-contribute/24509 Nov 04 03:36:59 Tapper: out of curiosity, did you find an answer to this mysterious summit release? Nov 04 03:37:22 Hi pkgadd know mate. Nov 04 09:26:15 Is there any support for linksys EA6350 v3 ? Nov 04 12:00:09 Borromini: thanks for testing. could you please run one more test for me? Nov 04 12:00:18 please copy the mt76 makefile from openwrt master to a 18.06 branch build Nov 04 12:00:24 it should be compatible Nov 04 12:00:34 i want to rule out mac80211 differences between master and 18.06 Nov 04 12:00:35 thanks Nov 04 12:00:36 bbl Nov 04 12:01:17 nbd: will do later and report back to you, thanks Nov 04 16:20:04 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Nov 04 16:33:54 for the lazy among us: find -path '.*/.git/config' -exec sed -i 's:git\.lede-project\.org:git.openwrt.org:g' {} + Nov 04 16:53:51 stintel: how very dare you accuse me of being lazy!!! ;-) Nov 04 17:01:23 hahaha Nov 04 17:01:31 you gave yourself away ;) Nov 04 17:06:16 dammit Nov 04 17:49:01 hello Nov 04 19:01:05 nbd: ping Nov 04 19:14:35 nbd: i just had a similar issue, i was away for the day so i can't say how long the uptime was more or less but it was way better than on 18.06 with the latest mt76. it seems it's particularly prone to happen on association somehow - it often happens when start my laptop e.g. Nov 04 19:20:37 rmilecki: I reported the brcmfmac problem to linux-wireless@ https://www.spinics.net/lists/linux-wireless/msg179703.html Nov 04 19:27:21 rmilecki: would you be OK with me committing https://git.openwrt.org/c48e0f49 ? Nov 04 19:27:28 s/committing/pushing/ Nov 04 19:29:09 I rather not drop the patch that enables roamoff on brcm2708 Nov 04 19:29:41 because I've had my fair share of instability issues with brcmfmac on rpi Nov 04 19:34:44 hmm, my issues happened after that patch was added to openwrt Nov 04 19:35:50 anyway, it doesn't matter, the upstream patch broke something, it should be reverted or fixed. this is not a bug in brcmfmac Nov 04 19:43:36 stintel: huh, I don't have that email. Neither in my linux-wireless folder, nor in spam. I have all other messages from the list though (both Larry's from before and Joe's from after yours) Nov 04 19:48:00 KanjiMonster: well it was definitely received as it's on that archive .. Nov 04 19:48:36 KanjiMonster: do you have access to the logs of your incoming mailserver(s) ? Nov 04 19:48:44 it's gmail Nov 04 19:48:47 hah Nov 04 19:48:50 say no more Nov 04 19:48:57 tons and tons of false positives there Nov 04 19:49:04 up to rejecting mail for no explainable reason Nov 04 19:49:07 :D Nov 04 19:49:47 my mails pass SPF and are validly DKIM-signed Nov 04 19:50:07 gmail is just total unreliable crap Nov 04 19:50:27 Nov 04 19:51:38 I'm used to finding perfectly valid emails in spam, but not at all isn't that common AFAICT Nov 04 19:51:49 microsoft's servers can be unnecessarily stringent as well though (they actually responded to my whitelisting request) Nov 04 19:55:07 mail is a disaster :) Nov 04 19:55:31 KanjiMonster: either way, what's your opinion on pushing my revert Nov 04 19:58:02 I would like to push it, so we have working wifi on brcm2708 again Nov 04 20:04:40 stintel: ok, that's weird.... i also didn't get your e-mail (also GMail here) Nov 04 20:04:46 KanjiMonster: ^^ Nov 04 20:05:00 stintel: revert sounds OK to me, but I'd prefer to ask nbd for an opinion Nov 04 20:05:16 I don't think that revert breaks anything for OpenWrt, but we may still want to ask Nov 04 20:05:26 nbd: there is a regression: https://www.spinics.net/lists/linux-wireless/msg179703.html Nov 04 20:05:34 nbd: in brcmfmac caused by nl80211 commit Nov 04 20:05:47 nbd: do you think we're good with the revert for now? https://git.openwrt.org/?p=openwrt/staging/stintel.git;a=commitdiff;h=c48e0f49e1f34ef4c54183ae151b7eee2b58f914 Nov 04 20:06:05 thanks guys Nov 04 20:06:19 stintel: a really nice report btw, thanks a lot for handling that! Nov 04 20:07:38 rmilecki: iiuc there is only one driver that disables that flag, so the breakage was easily overlooked Nov 04 20:09:28 we should probably also investigate if that roamoff=1 is still needed on brcm2708, but even if it isn't, that doesn't change the fact brcmfmac w/ roamoff=1 is broken Nov 04 20:10:02 stintel: I wonder if it wouldn't be enough to just set the brcmfmac's ops' update_connect_params to NULL when roamoff is enabled Nov 04 20:10:18 stintel: right Nov 04 20:10:31 stintel: we can check if rasberry pi kernrel still carries that patch Nov 04 20:12:10 KanjiMonster: well that's where you guys come in - I have no knowledge about mac80211/nl80211/drivers at all Nov 04 20:12:37 KanjiMonster: I can give that a try though Nov 04 20:14:45 stintel: OTOH, maybe just revert it. 7f9a3e150ec7d3596386449c15aefb59904a1266 didn't introduce the connect param thingy, so it might not be the right solution (just silences the WARN_ON / EINVAL return) Nov 04 20:15:18 KanjiMonster: I think revert until upstream fix is the nicest way, yes Nov 04 20:18:55 for that stability thing, I've got a 2-AP with same SSID network here, which allows roaming, I could probably come up with a test setup to compare roamoff=1 vs roamoff=0 Nov 04 20:19:52 I've been playing with https://vaping.readthedocs.io/en/latest/dev/ recently Nov 04 20:20:03 https://vaping.readthedocs.io/en/latest/ * Nov 04 20:35:03 stintel: your e-mail has arrived :) Nov 04 20:36:13 rmilecki: heh Nov 04 20:36:24 extreme greylisting? Nov 04 20:36:55 right, you can never figure that our with gmail :( Nov 04 21:38:39 ldir: https://www.linux-ipv6.be/OpenWrt/dnsmasq-static-unavailable.txt Nov 04 22:41:23 hello. i'm getting some kind of error when i try to build. base-files is failing. it looks like osme kind of key-build issue, that i'm not used to seeing? link to make log: https://gist.github.com/rektide/6f3e62084523ee3451ec42d6df4b7ba3 Nov 04 23:21:41 "crytographically sign packages" was on. turned off & is building now. Nov 05 01:58:48 Hey, can I submit patches to the packages feed via email or do they only one push requests? Nov 05 01:59:59 because f*** m$ Nov 05 02:00:01 lol Nov 05 02:20:10 dansan: the package feed is maintained via github pull requests Nov 05 02:27:46 stintel: strange! I wonder why - curious about tags: lan, ***known-othernet***, eth0.54 **** ENDING LOGGING AT Mon Nov 05 03:00:00 2018