**** BEGIN LOGGING AT Mon Oct 21 02:59:58 2019 Oct 21 04:20:57 blocktrron: PR updated. Oct 21 05:49:17 patchwork hiccup http://lists.infradead.org/pipermail/openwrt-devel/2019-October/019391.html ? Oct 21 05:50:09 hm, nope, it's already marked as Superseded Oct 21 06:05:49 DonkeyHotei: thanks, commented already **** BEGIN LOGGING AT Mon Oct 21 06:43:31 2019 Oct 21 06:49:21 blocktrron: replied. Oct 21 07:38:59 blogic: Hi. I got a notification that you had accepted my patch adding support for u4019, but I cant find it in any of the git-repos. Has the change not been pushed, or maybe accept was not the correct verdict? :) Oct 21 10:10:42 kristrev: they get accepted into developer's staging trees for testing/cleanup/whatever and assuming then get pushed into master. Oct 21 10:11:03 assuming ok then get pushed... Oct 21 10:21:05 ldir: Thanks for letting me know :) I was expecting that to be the process, but I don Oct 21 10:21:41 dont see it in Johns tree. I am probably just a bit impatient, but someone asked me about the device and I then noticed accept but no commit anywhere :) Oct 21 10:25:09 john doesn't tend to push too often to his staging tree which leads to the perterbing situation you describe "been accepted... but you can't find where) Oct 21 10:25:28 it does (usually) appear in the end. Oct 21 10:28:30 kristrev: done Oct 21 10:28:49 kristrev: sorry forgot to do that last night Oct 21 10:31:20 blogic: Thanks and no worries, thanks for taking a look and accepting the patch Oct 21 11:28:09 mangix: I wouldn't say it. It's enabled for OpenWrt 19.07 and OpenWrt master. So, it shouldn't be an issue to backport it. Users requested it and it makes sense for some users. Oct 21 11:30:43 blocktrron: https://patchwork.ozlabs.org/patch/1176288/ this looks goood too me and it works on mt 8970 Oct 21 11:30:51 is there anything holding us from merging it ? Oct 21 11:32:07 Run Testing was on my ToDo Oct 21 11:32:23 works here Oct 21 11:32:25 But if you did this already - feel free to take care about it Oct 21 11:32:29 mind if i pick it up ? Oct 21 11:32:33 cool Oct 21 11:33:08 In other news - has anyone tinkered around with Arubas new AP11 already? Oct 21 11:33:44 im still using oldstable on 8970 because the newer versions just freeze after a while Oct 21 11:44:20 rmilecki: could you verify this one -> https://patchwork.ozlabs.org/patch/1171272/ Oct 21 11:44:35 blogic: i'll check that Oct 21 11:44:56 rmilecki: it looks correct to me Oct 21 11:45:06 just would like an extra set of eyes to have a look Oct 21 11:45:14 if you agree, let me know and i can merge it Oct 21 11:45:20 its already in my procd-staging Oct 21 12:09:20 kristrev_: this one fails to apply, could you rebase/send the series please --> https://patchwork.ozlabs.org/patch/1166494/ Oct 21 12:51:33 blogic: Yes, will do, real life and work got in the way so I havent had a chance to work with the device after Adrians changes Oct 21 12:51:46 Will have access to the device again in the middle of next week, so will rebase then Oct 21 12:51:49 sure, i know the problem ;) Oct 21 13:00:37 build #26 of brcm2708/bcm2708 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/openwrt-19.07/images/builders/brcm2708%2Fbcm2708/builds/26 Oct 21 13:22:42 ooh, bots are back, hail to the jow! Oct 21 13:45:17 build #124 of ar7/ac49x is complete: Failure [failed cleanbase kmodconfig] Build details are at http://buildbot.openwrt.org/master/images/builders/ar7%2Fac49x/builds/124 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , Kristian Evensen , Josef Schlehofer , Oct 21 13:45:17 Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz , Tim Harvey Oct 21 13:45:17 , Michal Cieslakiewicz , Andre Heider Oct 21 14:01:00 blogic: ping Oct 21 14:08:28 build #122 of malta/be is complete: Failure [failed cleanbase kmodconfig] Build details are at http://buildbot.openwrt.org/master/images/builders/malta%2Fbe/builds/122 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , Kristian Evensen , Josef Schlehofer , Oct 21 14:08:28 Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz , Tim Harvey Oct 21 14:08:28 , Michal Cieslakiewicz , Andre Heider Oct 21 14:14:11 apparently one of the patches i pushed broke stuff Oct 21 14:14:21 Monkeh: i'll nudge you later Oct 21 14:14:42 Hokay Oct 21 14:16:22 Horshore Oct 21 14:23:07 build #121 of ath79/generic is complete: Failure [failed cleanbase kmodconfig] Build details are at http://buildbot.openwrt.org/master/images/builders/ath79%2Fgeneric/builds/121 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , Kristian Evensen , Josef Schlehofer Oct 21 14:23:07 , Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz Oct 21 14:23:07 , Tim Harvey , Michal Cieslakiewicz , Andre Heider Oct 21 14:24:34 also ramips build failed with latest master: mipsel-openwrt-linux-musl-cpp: error: ../dts/_unielec_u7621-06-256m-16m.dts: No such file or directory Oct 21 14:24:39 blogic: Makefile:218: *** missing 'endif'. Stop. Oct 21 14:24:54 blogic: https://git.openwrt.org/?p=openwrt/openwrt.git;a=commitdiff;h=4a45e69d190f72ed94878487b271ed7651dd9efa Oct 21 14:25:04 aparcar[m]: *please* test your patches Oct 21 14:26:03 blogic: endif is missing in the ifdef CONFIG_INSTALL_LOCAL_KEY Oct 21 14:26:45 I've reverted it Oct 21 14:26:48 aparcar[m]: ^^ Oct 21 14:27:14 thanks, you've beat me :) Oct 21 14:27:23 wifi on the train is too slow... Oct 21 14:33:30 reverting this commit 4408723d42e0 ramips: remove RAM size from device name for UniElec devices, fixes ramips build issue Oct 21 14:45:47 build #120 of samsung/s5pv210 is complete: Failure [failed cleanbase kmodconfig] Build details are at http://buildbot.openwrt.org/master/images/builders/samsung%2Fs5pv210/builds/120 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , Kristian Evensen , Josef Schlehofer Oct 21 14:45:47 , Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz Oct 21 14:45:47 , Tim Harvey , Michal Cieslakiewicz , Andre Heider Oct 21 15:00:59 build #117 of mpc85xx/p2020 is complete: Failure [failed cleanbase kmodconfig] Build details are at http://buildbot.openwrt.org/master/images/builders/mpc85xx%2Fp2020/builds/117 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , Kristian Evensen , Josef Schlehofer Oct 21 15:00:59 , Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz Oct 21 15:00:59 , Tim Harvey , Michal Cieslakiewicz , Andre Heider Oct 21 16:11:47 build #115 of ramips/mt7621 is complete: Failure [failed images] Build details are at http://buildbot.openwrt.org/master/images/builders/ramips%2Fmt7621/builds/115 blamelist: Jeff Kletsky , Tomasz Maciej Nowak , John Crispin , Kristian Evensen , Josef Schlehofer Oct 21 16:11:48 , Yousong Zhou , Paul Spooren , Adrian Schmutzler , Masafumi UTSUGI , Hauke Mehrtens , Rosen Penev , Daniel Gimpelevich , Eneas U de Queiroz Oct 21 16:11:48 , Tim Harvey , Michal Cieslakiewicz , Andre Heider Oct 21 20:22:10 good, it actually works as advertised https://gitlab.com/openwrtorg/project/urngd/pipelines/90373006 Oct 21 20:24:04 5 jobs for master in 2 minutes and 45 seconds (queued for 3 seconds), benefits of building when most of the EU is in the bars :p Oct 21 20:34:11 build #27 of rb532/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/openwrt-19.07/images/builders/rb532%2Fgeneric/builds/27 Oct 21 20:45:46 ynezz: i have a patch local for procd to start urgnd as an internal early job Oct 21 20:45:58 it kills those warning lines you get on some device Oct 21 20:53:14 blogic: sorry I really have to setup a local buildbot before proposing such patches... Oct 21 20:53:14 I'll send a v2 later Oct 21 20:53:31 aparcar[m]: no worries this stuff happens Oct 21 20:54:09 you'd be suprised, i once broke net-next with my flow-dissector fixes for DSA for a whole rcX Oct 21 20:54:12 :-) Oct 21 20:54:32 aparcar[m]: did you manage to log into the server Oct 21 21:39:52 blogic: yes server login works. but only 8 cores tsts Oct 21 21:41:06 8 cores Oct 21 21:41:25 my laptop has 4, that machine is a miracle :-) Oct 21 21:48:45 build #26 of brcm2708/bcm2709 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/openwrt-19.07/images/builders/brcm2708%2Fbcm2709/builds/26 Oct 21 21:51:30 which devices uses a rtw88 wifi card? Oct 21 22:01:53 Hauke: it's a m.2 card. Haven't seen one yet, however x86 is most likely Oct 21 22:02:39 Bought one because i was annoyed with how much of a wreck the QCA6174 is. Oct 21 22:03:26 The choices are rather limited if you want a m.2 card doing AP mode. Oct 21 22:03:56 qca6174 is a wifi card > Oct 21 22:04:02 ? Oct 21 22:04:48 Yes? Oct 21 22:05:01 Or am I mixing things up here? Oct 21 22:05:06 ath10k ? Oct 21 22:05:34 yup Oct 21 22:06:00 no idea, i saw qac60xx a couple of time and was not sure what is is Oct 21 22:06:02 kk Oct 21 22:06:07 seems to be different from the other chips (doesn't support multi-vif, etc. pp.) Oct 21 22:06:25 dafuhq Oct 21 22:06:37 no nulti vap ? Oct 21 22:07:31 nope Oct 21 22:08:01 It's targeted towards laptops (supports Bluetooth), so i think anything advanced was out of scope for them Oct 21 22:08:17 ah ok Oct 21 22:08:31 Dell still ships them with their current models, however it only seems to be sold by Rivet Networks? Oct 21 22:09:05 If you have an idea what to pick if it has to be m.2 - that would be great. iwlwifi doesn't support 5GHz AP mode sadly :/ Oct 21 22:10:18 iwlwifi only supports 2.4 ap on ax200, i got that confirmed by johill Oct 21 22:10:31 no idea about ac Oct 21 22:10:57 but on ax they only have support in fw on 2.4 Oct 21 22:11:06 no 5 ap Oct 21 22:11:21 same story for AC Oct 21 22:11:35 must be 5g dfs cac Oct 21 22:12:05 i;ll ask johill to confirm Oct 21 22:12:39 thanks. maybe i should simply look out for a janky m.2 --> pcie adapter from ali Oct 21 22:14:02 there will be ath11k hastings silicon soon I am told Oct 21 22:14:38 do they plan providing pcie radios compatible with ath11k? Oct 21 22:14:39 i actually have 2 pre-productions samples in front of me, arrived 3 days ago Oct 21 22:14:45 yup Oct 21 22:14:55 in m2 Oct 21 22:15:02 <3 Oct 21 22:15:29 the future is now Oct 21 22:15:47 yup Oct 21 22:32:32 qca6174 is a wifi card > --- It's a piece of crap Oct 21 22:33:01 firmware crashing during normal operations, can barely do monitor mode (as in just capture) Oct 21 22:33:55 kinda feels like they contracted broadcom to design it Oct 21 22:51:10 Mister_X: :-) Oct 21 22:51:41 i dare say ath11k will be better Oct 21 22:52:41 we are hitting a stable 2gbit with he160 over 24h iperf/chariot Oct 21 22:59:50 nice to hear :) Oct 21 23:49:07 any devs around? got a question Oct 21 23:50:53 Just ask... Oct 21 23:55:36 i've got a wavlink device/wifi extender... WL-WN560N2-BL Oct 21 23:55:56 it's driven by a mediatek MT7628KN Oct 21 23:59:27 trying to figure out if & how to run openwrt (or any linux version) on it Oct 21 23:59:37 and how to go about setting this up Oct 21 23:59:57 https://openwrt.org/tag/mt7628?do=showtag&tag=MT7628 Oct 22 00:00:18 there's that, but this particular device isn't on there... wondering if I can just run another version Oct 22 00:03:33 it's got a flash chip that I'm trying to ID Oct 22 00:03:36 the naswer to that would be in 99% of all cases, a firm "no". you won't be able to avoid porting your device to OpenWrt properly Oct 22 00:04:07 https://imgur.com/a/19xh97z Oct 22 00:04:44 sdk for MT7628 seems to be based on openwrt kernel Oct 22 00:05:04 i don't mind porting Oct 22 00:05:09 just need assistance Oct 22 00:05:15 figuring out where to start Oct 22 00:05:50 first step I was thinking was to tie into 3.3V UART pinout on board using usb adapter and seeing if I can pull up a console prompt Oct 22 00:06:24 assuming its there & I can even log in (no root password?) Oct 22 00:07:35 my guess is that it would be similar to other devices based on the same chip, but simpler, since it is only has 1 eth port out of the embedded 5port switch Oct 22 00:10:29 build #142 of layerscape/armv8_64b is complete: Failure [failed images] Build details are at http://buildbot.openwrt.org/master/images/builders/layerscape%2Farmv8_64b/builds/142 blamelist: Rosen Penev , Tomasz Maciej Nowak , John Crispin , Kristian Evensen , Yousong Zhou Oct 22 00:10:29 , Paul Spooren , Adrian Schmutzler , Daniel Gimpelevich , Petr ?tetiar , Tim Harvey , Michal Cieslakiewicz , Andre Heider Oct 22 00:12:12 ok, thinking/guessing it might be a winbond flash chip clone Oct 22 00:12:31 W25Q16 clone? Oct 22 00:13:30 so, 2MB spi flash Oct 22 00:15:10 what's the lower limit for external flash for running openwrt? Oct 22 00:15:30 I'd have no prob replacing/upgrading the flash chip Oct 22 00:16:12 >>4 MB Oct 22 00:17:20 pkgadd: ok cool... know of a decent flash chip in the same form factor? Oct 22 00:17:59 SOIC Oct 22 00:19:36 they're all ~interchangeable Oct 22 00:19:55 you can get up to ~16meg cheaply if you're prepared to be resoldering, but, youre way out in the woods there Oct 22 00:20:38 and at this point I would strongly reconsider if it's worth the effort Oct 22 00:21:05 * karlp would concur Oct 22 00:21:35 lol, the openwrt dev channel discouraging diy/dev efforts? Oct 22 00:21:46 soldering/reflowing a chip is cake Oct 22 00:21:59 damn ebay prices have gone up Oct 22 00:22:07 that's the only thing that sucks Oct 22 00:22:21 i just need a basic kernel running on this so i can do a AP Oct 22 00:22:40 the stock firmware sucks, and only supports doing a NAT'ed AP w/ dhcp server Oct 22 00:23:42 gonna have to dig around the junk box & see if I can scavenge an spi flash chip from something else Oct 22 00:23:43 Not discouraging... but heard same Q over and over and "always" failing.. I'd say Oct 22 00:24:06 well diff is I'm willing to do work to port, and not just "do this for me" Oct 22 00:24:29 i just need feedback & help getting going in right direction Oct 22 00:25:52 that doesn't change the fact that your device ships with 2 MB flash, which makes it unsuitable for running OpenWrt from the get go. yes, you can replace the flash chip, but that makes it your very own custom hardware no one else can buy. and adding support for a one-of-a-kind device to OpenWrt doesn't make sense Oct 22 00:26:10 yes, this is custom Oct 22 00:26:19 not trying to publish an image for this device Oct 22 00:27:09 looking to compile & get it to work on a certain device, and it can just serve an FYI "openwrt was ported to such&such device" for the historical archive Oct 22 00:27:20 you do realize that you can buy already supported devices with 16 MB flash and 128 MB RAM for 16-20 USD delivered? Oct 22 00:27:44 yeah, prolly, but im a diy'er, and have this device in front of me Oct 22 00:27:45 and used ones for 1-5 EUR plus shipping Oct 22 00:28:31 flash chips used to be like $2 (before trump drama), now theyre $$$...so, again, thats really the only part that sucks...but im pretty sure i can scavenge one off something Oct 22 00:31:02 still plenty cheap if you really want themÞ https://lcsc.com/product-detail/FLASH_Macronix-MX25L12835FM2I-10G_C71945.html Oct 22 00:31:28 https://www.ebay.com/itm/1Piece-New-W25Q64BVSIG-25Q64BVSIG-64M-BIT-Serial-Flash-Memory-SOIC-8-SPI-IC-Chip/231122518979 Oct 22 00:33:19 before looking any further, how much RAM does you device have? 2 MB flash 'usually' go hand in hand with 8 MB RAM... Oct 22 00:37:30 64mb Oct 22 00:37:35 its a soc Oct 22 00:37:48 mediatek7628KN Oct 22 00:37:59 https://openwrt.org/tag/mt7628?do=showtag&tag=MT7628 Oct 22 00:38:20 that would be a strange combination, but well, stranger things have happened Oct 22 00:38:44 strange combination how? Oct 22 00:38:57 it's a simple wifi extender Oct 22 00:39:43 it is very strange on behalf of a vendor to put 64 MB RAM into a device, but only 2 MB flash Oct 22 00:40:02 https://imgur.com/a/zkpJCCh Oct 22 00:40:18 the chip itself has the ram built-in Oct 22 00:40:28 you usually only see 2 MB flash with 8, maybe 16-, MB RAM - and usually running a non-linux OS Oct 22 00:40:32 2MB flash == cheap device, they didn't want to spend I guess Oct 22 00:41:39 honestly, I'm not 100% positive that it's a 2MB flash chip, but, based on the markings, it would seem so Oct 22 00:41:57 I'd have to remove it & use my soic adapter to talk to it & find out Oct 22 00:45:02 https://imgur.com/a/IdG0OvJ Oct 22 00:45:06 full board visible there Oct 22 01:41:16 jow: ping Oct 22 01:41:33 I just hit an infinite loop on nlbwmon, version 2019-06-06-4574e6e8-1 Oct 22 01:42:01 Here is the strace: https://pastebin.com/ikjAxgB6 Oct 22 01:42:48 I'll let the process running in case you need more data, as it's not always reproducible Oct 22 01:44:32 Near the bottom of strace's output there is a different message Oct 22 01:57:13 build #27 of brcm63xx/smp is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/openwrt-19.07/images/builders/brcm63xx%2Fsmp/builds/27 Oct 22 02:32:01 got an odd build failure. I did make -j1 V=s and it looks like this is the failure? Oct 22 02:32:04 Package ip-full is missing dependencies for the following libraries: Oct 22 02:32:07 libcap.so.2 Oct 22 02:32:30 current master **** ENDING LOGGING AT Tue Oct 22 03:00:24 2019