**** BEGIN LOGGING AT Thu Mar 21 02:59:57 2019 Mar 21 05:30:29 q Mar 21 05:31:06 Sorry. Selected the wrong window :( Mar 21 05:50:22 gch981213: in case breed settings go wrong, any recourse? Mar 21 08:34:30 mangix: Which kind of mistake? Overclocking too much? Mar 21 08:36:49 gch981213: yeah Mar 21 08:37:19 i'm trying to flash the stock bootloader back but i can't seem to get my reader to pick it up Mar 21 08:37:24 *spi programmer Mar 21 08:39:59 mangix: If your reset button or WPS button happened to be the gpio used by breed you can hold the button and power it on. Otherwise you can only use a spi flash progranmmer. Mar 21 09:30:15 Have someone done a iperf3 test recently with MT7621 device> Mar 21 09:31:53 I notice on my Ubiquity ER-X-SFP Download is around ~475mbit and upload 935mbit Mar 21 09:33:30 I wondering if other users has simulair results. Mar 21 09:40:19 mangix: what's the problem? Mar 21 09:41:20 * russell-- has been keeping his spi programmer warm recently Mar 21 09:48:27 Rene__: thats a bad result with an mt7621 device. Mar 21 09:48:34 I have a newifi d2, Gigabit up and down no problems. Mar 21 09:48:59 try multithreaded iperf Mar 21 09:49:17 it might be your pc/laptop/whatever bottlenecking it Mar 21 09:49:36 also install and run irqbalance. Mar 21 10:02:40 I have 2 pc Mar 21 10:03:58 Intel i3 and pentium g4650. Mar 21 10:06:07 I see it also with v1.9 and new v2.0.1 firmware of ubuquity. Mar 21 10:30:03 Funny that I see the same issue with pc to pc Mar 21 10:30:27 try multithreaded iperf Mar 21 10:30:28 ^^ Mar 21 10:30:39 iperf3 -h Mar 21 10:33:53 -P 2 same max bw of 475mbit Mar 21 10:34:42 thats weird. anyway, atleast you know that its not the ubnt that is the problem. Mar 21 10:37:06 I see it on the ubnt with 2 different PC Mar 21 10:37:17 I see it on the ubnt with 2 different PC's and also PC to PC Mar 21 10:38:07 dev PC has 2 different network cards. Intel is connected to LAN and Realtek is directly connected to UBNT. Mar 21 10:38:31 UBNT is also connected via SFP module to LAN, LAN to my server Mar 21 10:42:25 This issue is for this evening to fix. Mar 21 12:01:35 Rene__: is the sfp interface working on that device? /me recalls reading that it wasn't supported yet some time ago Mar 21 12:30:45 feckert ping Mar 21 13:06:47 karlp: Re: PR#1702, if I understand your comment properly you suggest, that we should provide (and support) factory/sysupgrade firmware images for all possible board+cape/hat combination? Mar 21 13:11:26 'even just a second dts file if it's a comonly sold readily available baseboard people use.' -> this would probably mean `sun8i-h2-plus-orangepi-zero-with-sata-usb-hat-ext4-sdcard.img.gz` in this case right? Mar 21 13:11:56 _we_ (openwrt) shoudl provide based on the machines available upstream, Mar 21 13:12:32 IMO, upstream should have dts files for _products_ that are readily available, yes, not necessarily arbitrary mixmatchs of hats/capes/extensions that users modded together. Mar 21 13:12:56 ok, so we agree, thanks :) Mar 21 13:13:07 a path for users to have a hope of managing overlays might be nice, though, of ocurse. Mar 21 13:13:44 yeah, there is so many interesting projects, but just 24 hours in the day :) Mar 21 13:37:27 russell--: Yes, but with 5.0 kernel, phylink API and at803x patches Mar 21 13:39:00 It should be easy to support it in current openwrt, mt7530 and at803x needs a patch to get it running Mar 21 13:56:32 * gch981213 is trying to send commit "c7ca224299 ramips: fix cpu clock of mt7621 and add dt clk devices" upstream and got confused by this checkpatch.pl complaining: https://paste.ubuntu.com/p/z7pZhtmP3g/ Mar 21 13:57:30 Why should I add an empty line before the prediv_tbl array? Mar 21 13:59:03 "because" ? Mar 21 13:59:09 uninit vs init vars? Mar 21 14:03:23 gch981213: I think, that the problem is also definition of multiple variables on one line Mar 21 14:03:59 this declaration `u32 xtal_clk, cpu_clk, bus_clk;` is not allowed by kernel coding style if I remember it correctly Mar 21 14:05:22 * gch981213 is reading the long coding style documentation... Mar 21 14:23:43 gch981213: maybe it wants the pre-defined stuff first, then your un-init stuff ? Mar 21 14:35:35 ldir: Yeah. That actually fixed the problem. Mar 21 14:35:57 but /me has sent the patch out :( Mar 21 14:37:06 quick, send a v2 before you get burned alive in the lake of linus hellfire :-) Mar 21 14:39:13 my last patch to iproute2-next was accepted without comment...and snuck into iproute2-5.0.0 as well! I was stunned. Mind you it only fixed a couple of error messages but I felt sure there would be something wrong :-) Mar 21 14:45:13 * gch981213 feels so bad about spamming lists because he sent a broken v2 :( Mar 21 14:49:49 gch981213: think of the spam you're saving by stopping the reviewer/s from sending 'fix this up--->>>' emails :-) Mar 21 14:55:02 Weird. Gmail folded the entire patch body, which made me thinking that I forgot to commit the change before sending it. Mar 21 16:36:11 ynezz: ack, that's a good reason, I'll send v2 today. Mar 21 17:00:59 hey guys. out of curiosity: is there a release manager like there was for 18.06 and 17.01? :-) Mar 21 17:06:37 Is anyone here aware of any device that supports ch. 183-196 (apparently JP only, with "registration needed") Mar 21 17:07:43 (not looking to buy, but to determine some logic around channel selection, as they are 4915-4980 GHz, which doesn't match [5][0-9]{3}) Mar 21 17:08:32 gch981213: i think static const is prefered to const static Mar 21 17:10:01 russell--: i'm using a ch341a spi programmer. i have two. none of them can read the winbond flash Mar 21 18:22:48 anyone here? Mar 21 18:23:28 no Mar 21 18:28:33 jeffsf, maybe ubiquiti stuff Mar 21 18:29:10 Thanks -- It's only another dozen characters in a script, so I'm putting it in Mar 21 18:29:53 (49|5[0-9])[0-9][0-9] vs. 5[0-9]{3} Mar 21 19:08:46 mangix: using flashrom? is the particular chip supported? /me had to use a git version of flashrom for a particular chip that hadn't been supported until recently. Mar 21 20:12:22 russell--: both flashrom and ch341prog Mar 21 22:55:01 tmn505: thanks, BTW what about that wifi phy path migration/upgrade script, there seems to be another potential user for this https://github.com/openwrt/openwrt/pull/1638#pullrequestreview-217136953 Mar 21 22:58:18 KanjiMonster: one more pls https://github.com/openwrt/openwrt/pull/1771 ? I promise it's the last one this month :) **** ENDING LOGGING AT Fri Mar 22 02:59:57 2019