**** BEGIN LOGGING AT Thu Sep 20 03:00:01 2018 Sep 20 04:50:01 anyone know how to add support for a second SIM slot in modemmanager when it goes to the PCIe slot? Sep 20 06:20:40 It seems that compiling wireguard, maybe other packages too, with the SDK (ralink mt76x8), the assembler files (.S) doesn't get any optimalization. Delay slots are mostly nops. Compiling assebler file by hand with gcc of SDK, delays slots are filled. Sep 20 06:23:05 compile it with: ARCH=mips STAGING_DIR=/mnt/buildbot/openwrt-sdk-ramips-mt76x8_gcc-7.3.0_musl.Linux-x86_64/staging_dir/ CROSS_COMPILE=/mnt/buildbot/openwrt-sdk-ramips-mt76x8_gcc-7.3.0_musl.Linux-x86_64/staging_dir/toolchain-mipsel_24kc_gcc-7.3.0_musl/bin/mipsel-openwrt-linux-musl- Sep 20 06:23:10 KERNELDIR=/mnt/buildbot/openwrt-sdk-ramips-mt76x8_gcc-7.3.0_musl.Linux-x86_64/build_dir/target-mipsel_24kc_musl/linux-ramips_mt76x8/linux-4.14.44 make Sep 20 06:46:09 https://paste.fedoraproject.org/paste/oSwpyIOfQAn3WQS3McwikQ output make V=1 Sep 20 06:55:13 Rene__: did you work out the command line differences? Sep 20 06:55:31 i.e. what makes it work when compiling "by hand" Sep 20 07:03:46 by hand I adding -O2 Sep 20 07:04:51 but the wireguard doesn't seems to add any optimalization to .S files. I see that -Os is added for .c files. Sep 20 07:06:40 Rene__: okay, so its not a bug with SDK vs. buildroot but a Wireguard packaging or build system issue Sep 20 07:07:36 can any one update the mwlwifi driver in master to the latest 10.3.8.0-20180920. Sep 20 07:07:42 https://github.com/kaloz/mwlwifi/commit/4a98c2bec174982676546e599d6f904b65a7179f Sep 20 07:10:43 Is there a router out yet with this new 88W8997 chipset in? Sep 20 07:13:02 jow: ok, I just have to wait when jason is back to help me out. Sep 20 07:13:06 Tapper: https://wikidevi.com/wiki/Special:Ask?title=Special%3AAsk&q=[[Chip1+model::~88W8997*]] (short answer, only wlan cards, apparently no routers yet) Sep 20 07:13:27 OK thanks Sep 20 07:14:42 Tapper: Imre is the openwrt package maintainer...hopefully a polite request to him would sort it ? Sep 20 07:15:15 Ping him on here or on github? Sep 20 07:16:16 He doesn't appear to be here on irc at the moment. His email is in the package Makefile Sep 20 07:17:05 If it were me, I'd try a ping on github. Sep 20 07:18:18 ldir: I doubt that Sep 20 07:18:54 ldir OK thanks mate Sep 20 07:19:10 I will have a go later Sep 20 07:21:05 truth be told so do I - but I've reset my brain to naive optimism mode and a polite request from a '3rd party' might help.... it may also clarify things Sep 20 07:23:40 Tapper: if you're feeling brave you could do it yourself....and then you could test it for us too ;-) Sep 20 11:01:49 nbd: ping Sep 20 11:23:07 jow: top work on luci app merging, super fast :) Sep 20 18:08:46 Hi how do I use git to push a change to my fork of openwrt? Sep 20 18:08:52 I tryed git commit -a -s Sep 20 18:09:09 it says pleas tell me who you are Sep 20 18:09:42 I cant work out the syntax for git config to set my user name Sep 20 18:12:24 Tapper: git push origin Sep 20 18:12:50 so not git commit -a -s Sep 20 18:13:18 yes to commit your work to your local repo Sep 20 18:13:31 but that will not work for me Sep 20 18:13:40 I am not an expert but I do it like this. Sep 20 18:13:50 It says that I have to set a username and set my email address Sep 20 18:13:56 git add Sep 20 18:13:58 git commit -s Sep 20 18:14:26 git config --global user.name "Mona Lisa" Sep 20 18:14:40 git config --global user.email "bla@bla" Sep 20 18:14:45 k thanks Sep 20 18:25:21 Rene__ how do i do a pr for openwrt now mate Sep 20 18:25:57 using github? Sep 20 18:26:16 yeah Sep 20 18:26:37 git push origin Sep 20 18:26:49 so that your branch is on github Sep 20 18:27:03 so Sep 20 18:27:11 go to github page and I think they offer you a pr button Sep 20 18:27:17 git push origin tapper82/openwrt Sep 20 18:27:56 yes Sep 20 18:30:32 My change is not showing up at https://github.com/tapper82/openwrt Sep 20 18:32:00 Tapper: that would push a branch named "tapper82/openwrt" to the origin remote. Sep 20 18:32:07 Tapper: what is the url listed for the "origin" remote when you run `git remote -v`? Sep 20 18:34:01 Hi agb It does not show a url Sep 20 18:34:14 it just prints >>> Sep 20 18:36:23 Rene__ any ideas? Sep 20 18:37:36 git branch Sep 20 18:37:45 with has * for it? Sep 20 18:39:10 Tapper: this is what I see when I run `git remote -v` https://dpaste.de/cbC7 I would at least output for "origin" when you run it. Sep 20 18:40:01 Hi dude is that https://dpaste.de/cbC7 a screen shot? Sep 20 18:40:23 If so I thank you for posting it, but I am blind mate. Sep 20 18:41:09 ah. it was a pastebin, but that site uses javascript silliness that probably breaks accessibility. here's the raw paste: https://dpaste.de/cbC7/raw Sep 20 18:41:55 OK so what the heck is going on with mine then lol Sep 20 18:42:13 don't mind for a min need to reboot brb Sep 20 18:43:11 basically you'll need to make sure you have a remote that points to your fork, whether it is origin or otherwise. then you can run `git push origin branchname` where "origin" should be replaced with the name of the remote that points to your fork, and branchname refers to the branch you committed changes to. Sep 20 18:43:38 I hope the reboot fixes things up. Sep 20 18:49:22 agb I am sorry still don't get it Sep 20 18:49:36 I forcked openwrt Sep 20 18:49:47 I clone tapper82/openwrt Sep 20 18:50:02 I cant push to tapper82/openwrt Sep 20 18:52:04 Tapper: tapper82/openwrt is repo not the brach. Sep 20 18:53:35 to get current branchname: git branch | grep \* | cut -d ' ' -f2 Sep 20 18:55:15 it says master Sep 20 18:56:34 git push origin master Sep 20 18:56:51 but it is better so move your work to branch. Sep 20 18:57:07 error 403 Sep 20 18:57:27 It's trying to push to the openwrt git and i don't have any rits Sep 20 18:57:32 rites Sep 20 18:58:56 so you cloned openwrt git not your github account? Sep 20 18:59:16 I mite of! :-( Sep 20 18:59:34 how do i tell Sep 20 19:00:16 Tapper: `git remote -v` should provide the full repo URL which will indicate which repo you cloned, that's what I was looking for the output of that command. Sep 20 19:01:01 you can also add another remote to your local repo, for example: git remote add tapper git@github.com:tapper82/openwrt.git Sep 20 19:01:02 agb I don't get any output Sep 20 19:01:36 then you could push with `git push tapper branchname` Sep 20 19:02:55 the URI format (git@github.com:user/repo.git) I provided assumes SSH auth. I don't use HTTPS-based auth with git but I know github has documentation on that if you prefer HTTP based auth instead of SSH. Sep 20 19:03:11 Tapper: sorry but I have to go. Sep 20 19:03:40 Rene__ np thanks for halpping mate Sep 20 19:04:06 agb I am getting confused Sep 20 19:04:30 I don't mind how it gets pushed for now Sep 20 19:05:27 Tapper: have you registered an SSH public key with GitHub? Sep 20 19:05:53 agb O shit no do i have to? Sep 20 19:06:53 My god this change is 2 lines man all this for to crappy lines! Sep 20 19:08:29 Tapper: you don't have to register an SSH key, and looking at the github documentation they recommend HTTPS-based (username/password) auth. Sep 20 19:09:11 OK so what do i type to push this change? Sep 20 19:09:47 remote -v shows up what I think it should now Sep 20 19:10:21 ok, could you post that to a pastebin so I can see the current value? Sep 20 19:11:51 first to lines are openwrt and next lines are tapper82/openwrt Sep 20 19:12:07 so lines 3 and 4 are my forck Sep 20 19:12:42 I don't know how to coppy and paste out of the ubuntu window Sep 20 19:13:32 ok. what is the first word on lines 3 and 4? Sep 20 19:14:29 tapper Sep 20 19:15:08 tapper git@github.com/tapper82/openwrt Sep 20 19:15:58 git@github.com:tapper82/openwrt Sep 20 19:16:54 if you don't have an ssh key registered, than let's change that, sorry. run `git remote remote tapper; git remote add tapper https://github.com/tapper82/openwrt.git` Sep 20 19:17:08 this change will let you push with your github username and password instead of an SSH key. Sep 20 19:17:48 once done, it should only require running `git push tapper` or if that gets fussy `git push tapper branchname` where branchname is the branch you've committed this change. Sep 20 19:18:38 adding a key makes everything easier though Sep 20 19:18:43 It says tapper is all read there Sep 20 19:18:49 all ready Sep 20 19:18:55 you've a rypo in the first git remote command - should be git remote remove tapper Sep 20 19:19:58 YES I am winning lol Sep 20 19:20:00 ldir: you're right. that first command should be `git remote remove tapper` Sep 20 19:20:10 thanks people Sep 20 19:24:18 Tapper1: congrats. I see your commit has made it into the repo! Sep 20 19:24:39 now you can make that merge request Sep 20 19:24:54 sorry, pull request, I've been living in gitlab for a while. Sep 20 19:25:28 cool thanks mate I am propper happy my build works and the wifi drive I bumped is working to Sep 20 19:25:43 wifi driver** Sep 20 19:26:46 hooray! Sep 20 19:27:04 So what's the way to PR? Sep 20 19:29:39 Tapper: you can go to https://github.com/tapper82/openwrt/pulls and click the "New pull request" button. Sep 20 19:30:00 dun nice Sep 20 19:30:12 thanks all! Sep 20 19:33:09 ok, so there's a small problem with your commit... the signed-off-by line should have your real name in it... the email address is fine. So... Sep 20 19:33:22 git commit --amend Sep 20 19:33:49 fix the signed-off-by Sep 20 19:35:10 then use the same git push command you used before - this should fail. Tell me the error you get. Sep 20 19:36:16 OK on it now hang on Sep 20 19:36:54 np - you're scaling a vertical leaning curve very well matey ! Sep 20 19:40:29 ldir non fast foreword Sep 20 19:41:16 the tip of your current branch is behind Sep 20 19:41:40 yeah, so what git is basically warning you is that your current commit in your local tree can't be connected to commits in the remote tree. Sep 20 19:42:31 how to fix? Sep 20 19:42:38 it'll suggest that you pull in the remote commits, which is usually sensible, but in this case not. You would end up with two mwlwifi bump commits, one wrong, one correct. This is clearly madness, so... Sep 20 19:44:04 one of the most dangerous git commands of options is -f for force. A force push will tell the remote to ignore the disconnect and just take our version as gospel. It's bloody dangerous! In this case it's the correct thing to do. So just add -f to your git push command. Sep 20 19:44:55 And the really clever bit is that since you've already opened the PR, that will also automagically get updated with your shiny new commit. Sep 20 19:45:58 dun Sep 20 19:46:31 One thing had to go rong it's sods law Sep 20 19:46:41 but nice quick fix Sep 20 19:46:47 ldir is the man! Sep 20 19:47:41 There you go, I can see the PR with the updated commit, got your name in, great stuff. Sep 20 19:48:22 btw it was a joint effort, I saw lots of people helping you. Sep 20 19:49:11 so FYI the -s option on git commit does the signed-off-by line for you based on info in your .gitconfig Sep 20 19:49:45 it might be worth looking at your .gitconfig in the home dir to check it has your full name so you don't fall into this trap again. Sep 20 19:49:58 KK thanks Sep 20 19:50:16 Do you have a dog? Sep 20 19:53:18 No mate Sep 20 19:53:52 I have a wife and 5 kids I would not have time to lookafter a dog proper Sep 20 19:54:27 I do love dogs, I will feel realy sad for it if i didn't have time to work it Sep 20 19:55:03 plus there is people who need one more than me Sep 20 19:55:05 Dammit - I was going to say "if a blind man and his dog can master openwrt patches with github, why can't you!?" :-) Sep 20 19:55:15 hahah Sep 20 19:55:49 has that updated? I cant see the change on my end on https://github.com/openwrt/openwrt/pull/1388 Sep 20 19:56:01 yes I can see the change Sep 20 19:56:09 KK Sep 20 19:57:21 I'm going to try and pull it into my staging tree using a maintainer's script...that I've never used before so it could all go horribly wrong... but let's try. Sep 20 20:06:41 done Sep 20 20:10:15 Tapper: for next time, you want to add a reason why you update this package in the commit message (basically the stuff you've added in your first response to your own PR, just inside the body of the commit message itself) Sep 20 20:16:13 check the netifd or odhcpd version bumps for minimal examples of that Sep 20 20:17:38 indeed, next time we can aim for perfection. This time, considering unix editors, make package/download PKG_HASH=skip & make package/check FIXUP=1, make package/refresh was all new, let alone git commit, let alone git push hassles, were new, let alone....blind (I don't know how he does it) I think a little slack and a lot of cheering for the first patch :-) Sep 20 20:18:18 you guys are just delightful Sep 20 20:19:27 ldir: are you going to the summit? Sep 20 20:19:43 yes :-) Sep 20 20:19:48 oh ok Sep 20 20:19:53 maybe ill come then Sep 20 20:20:23 been trying to weigh up if its worth 300 odd quid and a dumbass multihop return Sep 20 20:20:26 plus hotel Sep 20 20:20:35 I've been meaning to get to something linuxy for years, this year I decided I was going....before I found out it was on our wedding anniversary....and Missus ldir has still let me go! Sep 20 20:20:50 haha Sep 20 20:21:30 last thing i went to was probably the lsst ever breakpoint Sep 20 20:21:35 last Sep 20 20:21:45 You'll all find out how little I actually know! Sep 20 20:21:56 mmm, its tempting Sep 20 20:22:17 gowan gowan gowan gowan gowan Sep 20 20:22:23 the return path is dublin, london then prague though Sep 20 20:22:25 its a joke Sep 20 20:22:53 but i would like to come.... Sep 20 20:23:18 so perhaps i could just visit family in the uk and fly back another day Sep 20 20:23:20 ok deal Sep 20 20:23:27 well it's going to cost me a small fortune Sep 20 20:23:29 pkgadd I will keep it in mind for next time. Thanks for the feed back. Sep 20 20:23:38 yeah i bet Sep 20 20:23:44 lisbon is never cheap Sep 20 20:24:04 but I've been meaning to get to one for ages and I just decided this was the year. Sep 20 20:24:08 ldir Missus ldir sounds like a nice person :-) Sep 20 20:24:29 yes, buy chocolate Sep 20 20:24:43 Missus ldir is a wonderful lady indeed! Sep 20 20:24:54 and flowers if she dont have heayfever Sep 20 20:24:58 lol Sep 20 20:25:39 or stack of hay ;) Sep 20 20:25:49 lol Sep 20 20:25:50 lol Sep 20 20:26:58 Guys any update on marvell's dsa to replace swconfig Sep 20 20:26:59 ? Sep 20 20:29:55 * ldir calls it a night Sep 20 21:11:40 reboot is broken on my oxnas device (pogoplug v3) Sep 20 21:11:52 opkg search /sbin/reboot says busybox Sep 20 21:12:01 any clues what might be broken? Sep 20 21:12:18 reboot just hangs the device Sep 20 21:13:02 root@pogoplug:/# reboot -f Sep 20 21:13:02 [285331.270000] reboot: Restarting system Sep 20 21:13:02 [285331.270000] Reboot failed -- System halted Sep 20 22:25:45 marvell's dsa to replace swconfig. I'm sure I had a patch for that... Sep 20 22:30:27 there is always resistance to push through changes that will soft-brick users' routers over a sysupgrade, for good reasons, unfortunately the wish for a half decent migration script pretty much kills moving from swconfig to dsa in its tracks (and I very much doubt that such a script is possible in the first place) Sep 20 22:31:03 nah i doubt it. i'm pretty sure it's the lack of dual CPU port support Sep 20 22:31:28 I thought there were non-upstream(able) patches for that around Sep 20 22:31:29 there's no consensus on how to achieve that upstream unfortunately Sep 20 22:32:03 current working idea is to alternate the switch ports to cpu ports Sep 20 22:32:25 another one would be to bond the CPU ports so that the traffic gets load balanced Sep 20 22:32:39 but apparently not all switches support that Sep 20 22:32:47 and it does increase latency Sep 20 22:33:41 ah yes and idea 3 is to wire one CPU port to the WAN port and the other to the LAN ports Sep 20 22:33:57 * pkgadd can't really test it on ipq8065/ nbg6817 yet Sep 20 22:34:11 why not? Sep 20 22:35:46 no support yet (if I understood błogıc correctly, there are two modes of operation, the MAC address being encoded in binary at some offset or in ASCII via uboot-env, the later is used in (at least) the nbg6817 and the ea8500), my last attempt bricked my router (at least its ethernet) Sep 20 22:36:11 ah Sep 20 22:36:49 i'm happy to be back to marvell. a proper ethernet driver is very valuable Sep 20 22:37:39 # fw_printenv | grep ethaddr --> ethaddr=60:31:97:XX:XX:X4 Sep 20 22:38:09 all MAC addresses are derived from that one Sep 20 22:38:26 I'm pretty happy with my nbg6817 (but my WAN is only 100/40 MBit/s) Sep 20 22:39:28 I was happy with my r7800, until I realised I had increased ping times while gaming, Sep 20 22:39:43 Wonder if I could have mitigated that with ethtool... Sep 20 22:40:46 I've never been a gamer ;) that last games I played (and which I didn't write myself as exercise to learn programming language, simple stuff like tic-tac-toe) were still DOS based Sep 20 22:42:09 my ping times are around 14-16 ms, good enough for my needs (and if there were bigger latencies I probably wouldn't notice) Sep 20 22:49:45 SIP phones and interactive ssh sessions (among them IRC/ irssi) being the services most depending on latencies Sep 20 23:58:28 Tapper: btw., is it just me or is targetted linkspam increasing over the last couple of days/ weeks (as in incoherent, but remotely/ superficially relevant text combined with links to "support" sites)? Sep 21 01:51:09 yayo Sep 21 01:51:50 with linux kernel version 4.14.71 Sep 21 01:51:58 ok Sep 21 01:51:59 i assume because of this : https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=linux-4.14.y&id=9efcaa7c4afba5628f2650a76f69c798f47eeb18 Sep 21 01:52:19 you will have to delete Sep 21 01:52:41 this https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=target/linux/generic/pending-4.14/206-mips-disable-vdso.patch;h=f3c58dcd3adadbf2457891ac9ec1f12b53381cc7;hb=HEAD Sep 21 01:54:22 nbd, Hauke: ^ Sep 21 02:04:35 hi! is there anyone here who has a Comtrend WAP-5813n? i'm trying to find a working bootloader **** ENDING LOGGING AT Fri Sep 21 03:00:01 2018