**** BEGIN LOGGING AT Sat Sep 29 03:00:00 2018 Sep 29 05:19:43 * russell-- just built for an mt76x8 target, and no wifi Sep 29 05:28:25 blew away tmp, now better Sep 29 07:21:14 interestingly, the glinet gl-mt300n-v2 (mt7628) is working in station mode with reboot-8175-gfcfb9e4ded, but the same version is not working on the linkit 7688 Sep 29 07:48:59 jow: Are there no packages for IPQ401X or IPQ806X in openwrt ?? I dont find them in "http://archive.openwrt.org/chaos_calmer/" .. Am I looking correct ?? Sep 29 08:00:50 obaid: You're looking at an older release, which is not maintained. The current is 18.06.1 and you can download it at https://downloads.openwrt.org/releases/18.06.1/ Sep 29 08:02:36 luaraneda: Thankyou .. Sep 29 10:21:44 greearb: I've tried the latest firmware (ath10k-9984-10-4b/ath10k-fw-beta), but the mfp is still not set: Sep 29 10:21:57 greearb: ath10k_pci 0000:01:00.0: firmware ver 10.4b-ct-9984-fH-011-f9581cd api 5 features peer-flow-ctrl,txstatus-noack,wmi-10.x-CT,ratemask-CT,regdump-CT,txrate-CT,flush-all-CT,pingpong-CT,ch-regs-CT,nop-CT,htt-mgt-CT,set-special-CT,cust-stats-CT crc32 c019071e Sep 29 14:29:21 greearb_: it would be nice if you could add this patch: https://git.kernel.org/linus/91493e8e10f0f495b04a5c32096d56ea1f254c93 Sep 29 14:29:49 this should not cause any problems with older mac80211 versions Sep 29 15:29:50 luaraneda, please open a bug on my ath10k-ct github page. I'll build a series of binaries for you to bisect in hopes that will show where the problem lies. Sep 29 15:36:38 Hauke, ok, I applied that to my 4.16 tree...will get it into ath10k-ct driver soonish Sep 29 15:38:33 huaracheguarache, I must have some bug in my packaging tool, I'll work on that Sep 29 15:38:40 greearb: thanks Sep 29 15:39:46 it built, so I went ahead and pushed it to ath10k-ct... Sep 29 16:06:57 I'm having trouble with my ifdef logic, well logic in general if I'm honest. Could someone take a look at https://pastebin.com/2XRFFJ7U and suggest how I should re-write it. Sep 29 16:22:57 i think it would be more readable to write ((LINUX_VERSION_CODE > KERNEL_VERSION(3, 18, 0)) && (LINUX_VERSION_CODE <= KERNEL_VERSION(3, 18, 124)) ... Sep 29 16:29:33 greearb: ok, I'll keep an eye on the download directory Sep 29 17:28:08 nbd: true (or is that false? :-) Sep 29 17:50:58 stintel: I am looking at the brcm2708 patches needed for kernel 4.14 Sep 29 17:51:14 therea re about 450 patches in https://github.com/raspberrypi/linux Sep 29 17:51:18 *there are Sep 29 18:19:54 Hauke, you can find them here: https://github.com/Ysurac/openmptcprouter/tree/master/root/target/linux/brcm2708/patches-4.14 at least it's what I use even if it's not really clean ;) Sep 29 18:23:56 Ycarus: thanks Sep 29 19:02:08 How's OpenWrt's support for NXP/Freescale P1020 based hardware? Sep 29 20:00:52 voxadam: looks like it is supported: https://downloads.openwrt.org/releases/18.06.1/targets/mpc85xx/p1020/ Sep 29 20:26:59 ldir: ping Sep 29 20:48:45 huaracheguarache: pong Sep 29 20:49:22 hi. i tried out sqm cake with 6rd and experienced slow speeds on ipv6 speedstests for some reason. Sep 29 20:49:49 should cake be set to run on the ipv4 wan interface? Sep 29 20:50:32 that's where I'd expect it to be, yes. Sep 29 20:50:53 any other traffic? cpu usage? Sep 29 20:51:47 ipv4 seemed to be working alright, but I didn't have time to look more closely at cpu usage Sep 29 20:52:02 I can probably take a look later tonight or tomorrow Sep 29 20:52:33 do you run any ipv6 tunneling on your setup? Sep 29 20:53:16 I'm on holiday for a week so it won't be any form of priority - try putting a message in the forum, sebastian moeller is good with this sort of thing. Sep 29 20:53:33 I'm native ipv6 (at home, that is) Sep 29 20:53:44 wish I had native =( Sep 29 20:54:15 my brother has native ipv6 and I've set up openwrt with cake on his router. everything works a-ok Sep 29 20:55:07 I'll make a post in the forums. have a nice holiday ) Sep 29 20:55:44 hello, im having some problems with my router on fedora. The problem is, everything I try to download the down speed starts at the maximum and slowly starts to drop to around 100kb/s when it stays constant there, sometimes it spikes back up and drops again. This happens using wired connection with the router but if i use a wired connection to the modem the problem does not happen. Anyone would have a guess what might be causing this? I've been Sep 29 20:55:46 I used to have a 6in4 tunnel with hurricane electric - cake seemed to get on well, but this was a couple of years ago. Sep 29 20:55:46 trying many things in the past 3 or 4 days with the help of some guy who works in the area but so far we couldn't figure out what is happenning Sep 29 20:56:40 * ldir calls it a night anyway. Sep 29 20:57:14 ldir: thanks, gn Sep 29 20:57:21 im on OpenWrt 18.06.1 r7258-5eb055306f on a TP-Link WR740N v4.23 Sep 29 20:57:44 the problem doesn't seem to affect windows10, only on fedora 28 Sep 29 20:59:25 so the speed is more or less stable with windows 10? Sep 29 20:59:44 huaracheguarache: native IPv6 is of course faster and more stable than tunneling (6in4/ he.net in my case), but dynamic prefixes are a big nuisance) Sep 29 20:59:49 on windows 10 is always stable, at least it seems so Sep 29 21:00:35 opposed to linux i cant run something like wget "some url" and test various links, so the main tool would be speedtests Sep 29 21:01:06 so im not 100% sure but the majority of things in win10 i get at least 80%+ of the down speed Sep 29 21:01:13 fjunqueira: you can download iperf3 for windows Sep 29 21:01:33 huaracheguarache: ok, when i get to windows ill try that out Sep 29 21:01:44 I've used he.net for a couple of years, before getting native dual-stack IPv6 last year - with dynamic prefixes... (still looking for a good solution to combine both (he.net with a static prefix) and native IPv6 for outgoing connections (speed)) Sep 29 21:01:47 currently on fedora now, trying to figure this thing out Sep 29 21:02:28 its being quite painful to update the system since most of the updates even from official repos most of the time dont go past 300kb/s Sep 29 21:03:21 huaracheguarache: anyway, do you have any clue of what might cause this behaviour? Sep 29 21:04:23 fjunqueira: not really =/ Sep 29 21:05:07 huaracheguarache: ok, thanks Sep 29 21:05:28 pkgadd: what are prefixes and what makes them a pain? (i'm ignorant) Sep 29 21:05:41 huaracheguarache: i guess ill boot into windows now and try iperf3 since ive kinda got tired of trying to fix it for today Sep 29 21:06:15 fjunqueira: you can also try post on the forums to see if anyone there can help you out Sep 29 21:07:58 huaracheguarache: with IPv4 and NAT, changing IPv4 addresses just cause a short service interruption. with IPv6 and dynamic prefixes (meaning you get a different /56 prefix on each PPPoE connection), the situation is slightly different - as all your LAN devices get a global IPv6 address, so each device needs its own amount of time to notice the new prefix (and to stop using the old one) Sep 29 21:08:40 oh, that sounds bad Sep 29 21:08:41 I mean, you can retract an old prefix by advertising it with a zero lifetime... Sep 29 21:08:49 addressing and local DNS are another issue, just as well as configuring allowed IP ranges for incoming connections Sep 29 21:10:26 even with ULA prefixes local traffic might use the global IPv6 addresses Sep 29 21:10:34 but that prefix isn't stable... Sep 29 21:11:46 this squid config is /very/ useful, acl localnet src 2003::/23, but unavoidable... (relying on the router to do its job) Sep 29 21:12:06 it works, but dynamic prefixes are a pain Sep 29 21:20:19 pkgadd: maybe I'll have to deal with the same thing when my isp comes around to implementing dual stack Sep 29 21:20:48 I'll cross that bridge when (potentially) I get there Sep 29 21:24:07 I'd have switched to a business contract to avoid dynamic prefixes and get a static one, but those are only available on the pricier contracts and not the entry level one. and at some point it goes beyond what I'm willing to spend on it each month Sep 29 22:02:11 are the device definitions supposed to be in alphabetic order in, say, target/linux/ramips/image/mt76x8.mk ? Sep 29 22:46:41 russell--: yes Sep 29 23:31:13 pkgadd: this one is out of order... Sep 29 23:31:15 commit e1d4aa36b0e08d7e4b77b8b1a7ddae44188e8eb1 Sep 29 23:31:15 Author: INAGAKI Hiroshi Sep 29 23:32:29 possible was confused by the TPLINK entry at the top Sep 29 23:56:45 russell--: bugs like that happen all the time and aren't always apparent while reviewing (or simply missed), patches to correct the ordering are usually accepted Sep 29 23:57:16 it gets worse when target images are renamed (e.g. for DTS conversions) Sep 30 01:30:03 can someone please clarify how the compatible = {} field is supposed to look like in the .dts/i files so that sysupgrade works correctly? I'm getting "not supported by this image". I have added my target name in that list in quotes after a comma, is that the right way? Sep 30 01:38:14 nvm i forgot to add the board to lib/upgrade/platform.sh **** ENDING LOGGING AT Sun Sep 30 03:00:00 2018