**** BEGIN LOGGING AT Fri Apr 12 02:59:57 2019
Apr 12 05:21:40 <blogic> ynezz: +ath79_perfcount_irq = irq_create_mapping(domain, ATH79_MISC_PERF_IRQ);
Apr 12 05:21:59 <blogic> does the domain remap not handle the controller offset ?>
Apr 12 05:22:18 <blogic> so you should pass it 5 and not ATH79_MISC_IRQ(5); ?
Apr 12 05:42:34 <jow> cotequeiroz: simply always enable engine support. If one is forced to use openssl, it is a lost cause anyway
Apr 12 06:00:09 <blogic> jow: nicely put
Apr 12 06:21:00 <blogic> [   56.374276] wlan0: RX AssocResp from 00:03:7f:12:ae:e3 (capab=0x1 status=0 aid=1)
Apr 12 06:21:03 <blogic> [   56.380733] wlan0: HE AP is missing HE capability/operation
Apr 12 06:21:05 <blogic> yet another step closer
Apr 12 06:21:07 <blogic> [   56.422802] ath11k c000000.wifi: Removed peer: 00:03:7f:12:ae:e3 for VDEV: 0
Apr 12 07:41:00 <ldir> I have a very bizarre issue https://pastebin.com/Lq4NCidC see the last 2 'df' outputs - wth? !
Apr 12 07:41:58 <ldir> it's a freshly booted device, initially mounts are correct, then at some point over the 10 seconds of 'df' usage, it changes.
Apr 12 07:46:40 <ynezz> I remember seening something like that on imx6 with /boot partition
Apr 12 07:47:18 <ynezz> it was mounted twice, but I've simply thought, that it might be some issue with init scripts
Apr 12 07:49:41 <ldir> The mounts in question are on 2 different USB devices.  - not using extroot.  and the really odd bit is that initially it's correct.  then it goes wrong. and umount/mount get's it wrong from the command line too.
Apr 12 07:53:42 <ynezz> blogic: well, I wasn't simply sure why 13 was previously selected as a good candidate for the perfcount virq, to me the proper fix would be to simply allocate any free virq available at the time of perfcount init
Apr 12 07:54:09 <blogic> ynezz: well right now you are allocating irq13 of the misc controller
Apr 12 07:54:12 <blogic> but you want 5
Apr 12 07:54:20 <ynezz> really ?
Apr 12 07:54:24 <blogic> yes
Apr 12 07:54:31 <blogic> perfirq is line 5 on the misc controller
Apr 12 07:54:33 <ynezz> that macro expands to 8+5
Apr 12 07:54:42 <blogic> which is wrong
Apr 12 07:54:51 <blogic> as the domain xlate function already does the 8+
Apr 12 07:54:53 <ynezz> ah ok, makes sense
Apr 12 07:55:02 <blogic> well it does not, its obselete thx to the xlate
Apr 12 07:55:27 <blogic> previously there was a direct 1-1 relation between linux irq numbers and hw irqs
Apr 12 07:55:35 <blogic> this is gone soince we introduces domains
Apr 12 07:55:51 <blogic> there is alegcay domain whihc still does the predicatble 1-1 foo but we dont use that
Apr 12 07:56:07 <blogic> so the code looks good but you prolly wanna drop the macro wrapper
Apr 12 07:56:43 <xback> jow: regarding release bumps.  Nothing to add from me.  I don't see anything critical in current kernel queues.
Apr 12 07:57:18 <ynezz> blogic: ok, thanks, I'll investigate it
Apr 12 07:57:41 <blogic> ynezz: look at how ralink does it the fix is already upstream there
Apr 12 07:57:55 <blogic> ynezz: do you have the atheros datasheets ?
Apr 12 07:58:39 * blogic pms ynezz the github repo link :)
Apr 12 07:58:59 <ynezz> indeed, I don't have other sources :)
Apr 12 07:59:08 <blogic> fine
Apr 12 07:59:20 <ynezz> actually I've got the inspiration for this fix in lantiq
Apr 12 07:59:36 <blogic> lantiq uses the legacy 1-1 mappe domain i think
Apr 12 07:59:41 <blogic> ralink is a better reference
Apr 12 07:59:46 <ynezz> ok, noted
Apr 12 08:13:04 <mangix> wow, people are really interested in mt7625
Apr 12 08:47:57 <blogic> 7625 or 7615 :
Apr 12 08:48:03 <blogic> ?
Apr 12 09:03:51 <mangix> p15
Apr 12 09:03:57 <mangix> typo
Apr 12 09:07:27 <_lore_> mangix: if you mean 7615 802.11ac chipset, we are currently working on it
Apr 12 09:39:29 <Monkeh> _lore_: How's that going?
Apr 12 09:40:30 <_lore_> a v1 has been already posted upstream
Apr 12 10:52:52 <blogic> w00t
Apr 12 10:53:15 <blogic> got 2 aath11k units running ath11k talinkg 11ax 4x4 pumping traffic
Apr 12 10:53:31 <blogic> need to cleanup the patches and send them upstream
Apr 12 10:56:44 <stintel> blogic: are you working for the crack addicts now ? :P
Apr 12 11:01:32 <blogic> no
Apr 12 11:02:10 <blogic> https://twitter.com/BitsCrispy/status/1116657525015949312
Apr 12 11:02:12 <blogic> :-D
Apr 12 11:03:10 <stintel> cool stuff
Apr 12 11:03:21 <stintel> what devices are you using ?
Apr 12 11:03:29 <blogic> eval kits
Apr 12 11:04:12 <blogic> [    0.000000] Boot CPU: AArch64 Processor [410fd034]
Apr 12 11:07:00 <stintel> https://www.kickstarter.com/projects/librecomputer/la-frite-open-source-fries is gonna be shipped soonish. when I get it I'm gonna do some cleanup on the mesongx target that's been sitting in my staging tree for too long, and hope to push it with support for that board and odroid c2
Apr 12 11:09:20 <blogic> looks nice
Apr 12 11:09:29 <blogic> never really used any of the SBC stuff myself
Apr 12 11:09:33 <blogic> but still nice
Apr 12 11:10:43 <stintel> the odroid c2 runs 4.19 without patches, which I think is amazing
Apr 12 11:10:49 <stintel> especially compared to the rpi mess
Apr 12 11:10:50 <blogic> yarp
Apr 12 11:11:01 <blogic> but thats b0rkedcom
Apr 12 11:15:05 <stintel> now I just need to find something like that, that can do 4k60 over HDMI and has 11ac wifi strong enough to actually stream 4k video
Apr 12 11:15:25 <stintel> or find an apartment that has UTP everywhere
Apr 12 11:16:34 <stintel> ah
Apr 12 11:16:35 <stintel> https://www.khadas.com/vim
Apr 12 11:21:26 <dansan> stintel: wow, that's darned cute!
Apr 12 11:21:40 <stintel> too bad the wifi seems completely unsupported in mainline linux
Apr 12 11:22:04 <dansan> aww man
Apr 12 11:22:12 <stintel> I guess that's why I didn't already buy it
Apr 12 11:22:12 <dansan> what type of wifi hardware is it?
Apr 12 11:22:56 <dansan> I've dug into the MediaTek wifi drivers last December because I got paid to fix it, but dang that is complex!
Apr 12 11:23:39 <stintel> "AP6398S"
Apr 12 11:24:31 <dansan> oh, it's part of the µC?
Apr 12 11:25:41 <stintel> hmm, could be AMPAK which could be broadcom based
Apr 12 11:25:52 <dansan> oh, I see ... 7749 4664 5000
Apr 12 11:26:03 <dansan> oops, wrong clip board Ampak AP6398S
Apr 12 11:26:09 <dansan> what you just said, lol
Apr 12 11:32:39 <stintel> https://github.com/khadas/fenix/commit/267cc0ad9fe7c2962d638095101b04270a7a37a3#diff-2d6daa7e47666ce33a44189dfcecd0ea
Apr 12 11:32:58 <stintel> bcm4359
Apr 12 11:33:24 <stintel> it might actually be supported
Apr 12 11:41:48 <dansan> stintel: oh, very cool!
Apr 12 11:41:58 <stintel> I guess I'll have to order one
Apr 12 11:42:17 <stintel> but even on amazon.de it will be shipped from China and I hate dealing with customs here
Apr 12 11:43:41 <stintel> oh. there's a distributer on Romania, that's a neighboring country
Apr 12 11:47:17 <dansan> where are you from?  I'm in the states
Apr 12 11:49:31 <stintel> I'm in Bulgaria
Apr 12 11:51:04 <dansan> It's amazing how many inexpensive yet powerful devices are out now
Apr 12 11:56:45 <stintel> lol I just missed that there is also a distributor in Bulgaria
Apr 12 11:58:45 <stintel> but they all seem to have the original Vim2
Apr 12 12:12:52 <dansan> They have a board for audio too called the tone board.  I'm trying to understand it's exact function :)
Apr 12 12:14:12 <dansan> I guess it just has high quality DACs.
Apr 12 13:28:01 <ldir> I am seeing some really strange behaviour - https://pastebin.com/YKn7zfP8  - this is as my AP boots.
Apr 12 13:30:18 <ldir> so everything mounted fine, then some cross mounting
Apr 12 13:39:38 <russell--> ldir: wtf is dsb1?
Apr 12 13:40:57 <russell--> oh, just a mount point
Apr 12 13:41:20 <abenz> I have an ipq40xx target, 4 arm cores. I figured irqbalance would be good to run
Apr 12 13:41:46 <abenz> but I cant see any difference, I tried stressing the system with max transfers over ath10k etc
Apr 12 13:42:00 <abenz> but I dont see a tangible difference, does it make a difference for some of you?
Apr 12 13:42:11 <abenz> perhaps my workloads are not the best way to test
Apr 12 14:04:05 <abenz> ^ answering above, it seems it does help in certain situations. https://forum.openwrt.org/t/sqm-on-netgear-r7800-choppy-audio-video-teksavvy/3323/22?u=abenz
Apr 12 15:14:17 <ldir> goes in search of a good commit to prime a git bisect session
Apr 12 15:20:13 <stintel> sooo ... Device Tree. where do I even start
Apr 12 15:20:42 <stintel> I've just opened this Ubnt LBE-5AC-Gen2
Apr 12 15:21:02 <stintel> serial header is working and can boot ath79 image via initramfs
Apr 12 15:50:36 <stintel> ok so I guess the WA.v8.4.2# prompt hints that it's a WA board
Apr 12 16:19:18 * gch981213 wonders if there is any mt7621 router using the new u-boot-2018.09 for mt7621 now
Apr 12 16:46:27 <stintel> lol that was easier than expected. the rebranded qca988x is already supported in upstream ath10k and ath10k-ct
Apr 12 16:46:49 <stintel> and I have working ethernet, 2.4 and 5GHz wifi
Apr 12 16:48:20 <ldir> stintel: ship it! ;-)
Apr 12 17:06:39 <stintel> ldir: well I'm gonna figure out the LEDs and reset button and then I'll post a commit for review
Apr 12 17:09:17 <stintel> would be nice if I can get that sorted by the end of the day coz I am away with gf for the weekend
Apr 12 17:17:07 <stintel> heh even the button works already
Apr 12 17:17:49 <blogic> whooping
Apr 12 17:18:09 <blogic> i wrecked my left elbow just now doing weights
Apr 12 17:18:24 <blogic> no, the right actually ...
Apr 12 17:18:34 <blogic> hurts like fooo
Apr 12 17:18:46 <blogic> not that anyone cares
Apr 12 17:19:04 <blogic> and i just add twt responder support to ath11k
Apr 12 17:19:13 <blogic> that is more exciting news i guess :-D
Apr 12 17:19:30 <blogic> coming up, add/del BA support
Apr 12 17:23:47 <Tapper> blogic I did mine in about 2 years ago in the gym to. The pane was bad. Best thing I can tell you is get some ice on it even if it dont hert that bad now
Apr 12 17:24:22 <blogic> Tapper: thx mate, got a cooling cushion on it right now actually
Apr 12 17:24:29 <Tapper> It was my left elbow  I did. I was doing pullups
Apr 12 17:24:45 <blogic> i just did this doing dips
Apr 12 17:24:59 <Tapper> I waas to fat for my own elbows hahah
Apr 12 17:25:10 <blogic> calling me fat ? :-)
Apr 12 17:25:20 <Tapper> No dude me hahah
Apr 12 17:25:21 <blogic> i am down to 92kg at 186
Apr 12 17:25:33 <blogic> anyhow ...
Apr 12 17:25:55 <Tapper> I was same wate!
Apr 12 17:26:47 <Tapper> I had my dedlift up to 225 KG and bench to 150 KG then messed up my elbow and have not bin to the gym for 2 years.
Apr 12 17:27:27 <Tapper> I keep saying I am going back because elbow  is ok now but never get around to signing  back up.
Apr 12 17:29:52 <Tapper> If I go back going to stay off the big waits I am getting a bit old now.
Apr 12 17:30:32 <Tapper> at 36 it's time to just stay fit and stop trying to get hench to make the girls look at me lol
Apr 12 17:35:25 <stintel> https://git.openwrt.org/?p=openwrt/staging/stintel.git;a=shortlog;h=refs/heads/lbe-5ac-gen2
Apr 12 17:35:44 <stintel> so simple, lol
Apr 12 17:36:49 <stintel> ondertussen draait er al OpenWrt op ;)
Apr 12 17:37:03 <stintel> wrong window lol
Apr 12 17:37:47 <Monkeh> stintel: Yay, common platform
Apr 12 17:38:05 * Monkeh still has to get his old UAP Pro going on ath79
Apr 12 17:39:40 <stintel> Monkeh: I should do the same for my DAP-2695
Apr 12 17:39:55 <Monkeh> Does yours have a headache inducing bootloader oddity too? :P
Apr 12 17:40:39 <stintel> there was something but I don't remember exactly
Apr 12 17:40:56 <stintel> I think you couldn't flash through webinterface but it has a console port and via tftp it works just fine
Apr 12 17:41:34 <stintel> none of the GPIOs seem to have impact on the LEDs
Apr 12 17:52:45 <nmrh> gch981213: ty
Apr 12 18:16:28 <mamarley> Hauke: So I did what you recommended yesterday and tested hostapd from upstream git master.  I haven't had a chance to check with Android Q yet, but with Android 9 on the OG Pixel XL, it still refuses to connect when WPA3 and 802.11r are enabled at the same time without any log messages on either end.
Apr 12 18:52:21 <starlays> heloo, I have a question regarding transmission package, does anyone know why it is not using the config from /etc/config/transmission? It is building his own .json config in his own directory that has to be owned by transmission. The wiki from here is indicating something else: https://openwrt.org/docs/guide-user/services/downloading_and_filesharing/transmission
Apr 12 19:56:47 <cotequeiroz> jow: late ack for your openssl suggestion.  I'll do it on Monday.  Time to go fishing now.
Apr 12 20:48:21 <brucethompson> I have q question about the OpenWrt host build environment (staging_dir/hostpkg/bin)
Apr 12 20:48:49 <brucethompson> Is there to rebuild host packages here without rebuilding the world?
Apr 12 20:49:08 <blogic> yes
Apr 12 20:49:21 <brucethompson> OK
Apr 12 20:49:33 <blogic> make package/procd/{clean,compile} V=99
Apr 12 20:49:37 <blogic> or
Apr 12 20:49:40 <blogic> make package/procd/{clean,compile} V=s
Apr 12 20:49:52 <blogic> i think 99 is old school and the new kids in town use s
Apr 12 20:49:56 <brucethompson> So the target is procd then?
Apr 12 20:50:01 <blogic> yarp
Apr 12 20:50:14 <brucethompson> Let me try. Thx
Apr 12 20:50:17 <blogic> sure
Apr 12 20:50:36 <blogic> there is also
Apr 12 20:50:43 <blogic> make package/procd/{clean,compile} V=s QUILT=1
Apr 12 20:50:58 <blogic> that stage a packeg + patches using quilt
Apr 12 20:51:11 <blogic> you can then enter the build dir and use quilt to edit the series
Apr 12 20:51:30 <blogic> once you did a quilt refresh you can then go back to $(TOP_DIR) and do a
Apr 12 20:51:37 <blogic> make package/procd/update
Apr 12 20:51:43 <brucethompson> I have not done anything with Quilt
Apr 12 20:51:57 <blogic> that'll pull the new patch series back to packages/systm/procd/patches/
Apr 12 20:52:13 <blogic> okie ;)
Apr 12 20:52:56 <ynezz> blogic: https://github.com/ynezz/linux/commit/66f26173c4e97599c126840f068ec28703bf1164 (v2 of perfcount fix for ath79)
Apr 12 20:53:10 <brucethompson> I just did make/package/procd/clean and it did not remove anything in (staging_dir/hostpkg/bin)
Apr 12 20:53:32 <blogic> there ynezz drop line 135 and 127
Apr 12 20:53:37 <blogic> there ynezz drop line 125 and 127
Apr 12 20:53:57 <blogic> brucethompson: ok, host packages
Apr 12 20:54:06 <blogic> which host pkg you trying to clean ?
Apr 12 20:54:11 <brucethompson> pip
Apr 12 20:54:24 <brucethompson> I think its screwed up
Apr 12 20:55:29 <blogic> make package/opkg/host/clean
Apr 12 20:55:35 <blogic> that cleans the host opkg
Apr 12 20:55:43 <blogic> make package/opkg/host/{clean,compile}
Apr 12 20:55:52 <blogic> assuming pip is actually called pip try
Apr 12 20:55:57 <blogic> make package/opkg/pip/{clean,compile}
Apr 12 20:56:18 <brucethompson> OK. I'll try this.
Apr 12 21:01:58 <brucethompson> I did make package/opkg/host/{clean,compile}. Does not seem to do anything to the (staging_dir/hostpkg/bin) directory
Apr 12 21:05:45 <brucethompson> Maybe I need make dirclean? Then I'll just have to wait an hour or so.
Apr 12 21:09:28 <ynezz> blogic: thanks, sent
Apr 12 21:14:24 <brucethompson> Its definitely pip2.7 that's messed up
Apr 12 21:20:34 <brucethompson> I was wrong. Its the build of python that's messed up.
Apr 12 21:54:17 <brucethompson> Looks like my last questions ended in a dead end.
Apr 12 21:54:41 <brucethompson> is there any documentation on how host packages are built?
Apr 12 22:33:56 <blogic> brucethompson: what is this documentation dark magic thou is speaking of ?!
Apr 12 22:39:07 <brucethompson> Well.... Now I'm just doing a total rebuild. I'll see if that works.
Apr 12 23:32:07 <brucethompson> blogic: You still there??
**** ENDING LOGGING AT Sat Apr 13 02:59:57 2019