**** BEGIN LOGGING AT Mon Sep 30 03:00:08 2019
Sep 30 06:10:49 <gpeskens> pkgadd, sorry for no response, was off to bed, thanks! It indeed looks like 2 independent SOCs with a glue driver. Figuring out how it boots should be 'relatively' easy, and your approach sounds like the right one (get OpenWRT to boot, get the Lantiq HW in order, and then figure out how to boot WiFi), I'll try and see what comes out of it!
Sep 30 06:13:40 <pkgadd> good luck, it won't be easy, but nevertheless a pretty interesting device
Sep 30 06:15:26 <pkgadd> I think the very interesting 3390 is in the same boat as well
Sep 30 06:21:00 <pkgadd> err, s/3390/3490/
Sep 30 06:43:15 <gpeskens> 5490 as well (the 2 I have ;) )
Sep 30 06:43:21 <gpeskens> anyway, off to work!
Sep 30 06:46:15 <dynamite_bud> hey blogic man if you're awake help me out
Sep 30 06:46:47 <dynamite_bud> or Anyone else if you could guide me on how to start with OpenWrt/LEDE development.
Sep 30 08:11:24 <ynezz> aparcar[m]: what doesn't work in "however it doesn't work" ?
Sep 30 08:19:23 <aparcar[m]> ynezz: the kmod-gpio still contains the whole build path even tho it should be replaced
Sep 30 08:22:09 <aparcar[m]> ynezz: It only happens to gpio and wireguard kmods
Sep 30 08:22:49 <aparcar[m]> Oh and if you have the time please have a look at this: https://lists.infradead.org/pipermail/openwrt-devel/2019-September/019166.html
Sep 30 08:40:40 <ldir> aparcar[m]: a potential clue here is that they are out-of-tree kmods
Sep 30 08:42:29 <aparcar[m]> ldir: uhm so not mainline if I guess correct? But wouldn't they still be affected by the remap command of gcc8?
Sep 30 08:43:12 <ldir> yes, not mainline - not provided with the kernel if you like.
Sep 30 08:44:34 <ldir> In terms of build process, all I know is it is different.  eg. when building on macos external modules fail to build if the kernel build system can find 'libelf'
Sep 30 08:45:09 <ldir> the kernel's objtool is not cross-compile friendly
Sep 30 08:47:25 <aparcar[m]> ldir: this is currently beyond by compileing skill
Sep 30 08:47:42 <ldir> likewise!
Sep 30 08:47:46 <aparcar[m]> I guess I'll hope ynezz is hooked by that problem
Sep 30 08:48:41 <ynezz> well, you need to build with V=sc and see if that remap params are passed down to that particular module/file
Sep 30 08:49:38 <aparcar[m]> how can I build the modules? It's target/compile right?
Sep 30 08:49:50 <aparcar[m]> what is the difference between V=s und V=sc?
Sep 30 08:50:15 <ynezz> more verbose output
Sep 30 08:50:19 <ldir> AFAIK sc shows the commands used
Sep 30 08:51:48 <aparcar[m]> ty
Sep 30 08:51:51 <ynezz> can't you just `make package/kernel/gpio-button-hotplug/{clean,compile} V=sc` ?
Sep 30 08:51:53 * ldir has to go
Sep 30 08:54:53 <ynezz> it seems to me, that the issue is with the KERNEL_MAKE_FLAGS
Sep 30 08:58:54 <aparcar[m]> ynezz: I'll do some testing now
Sep 30 09:19:20 <aparcar[m]> ynezz: the command looks like that http://ix.io/1JL5 and still the build path is contained as a string http://ix.io/1JKS
Sep 30 09:25:55 <jow> anyone here has experience with 802.11s + SAE ?
Sep 30 09:28:34 <ynezz> aparcar[m]: that's not actual command
Sep 30 09:29:26 <aparcar[m]> <jow "anyone here has experience with "> jow: yes. Me and dangowrt
Sep 30 09:29:48 <aparcar[m]> ynezz: sorry what do you mean by that?
Sep 30 09:30:21 <ynezz> aparcar[m]: make package/kernel/gpio-button-hotplug/{clean,compile} V=sc 2>&1 | tee build.log; pastebinit build.log
Sep 30 09:30:27 <jow> aparcar[m]: which ciphers are used by 802.11s in SAE mode?
Sep 30 09:30:35 <jow> just CCMP?
Sep 30 09:31:33 <aparcar[m]> ynezz: http://ix.io/1JLA
Sep 30 09:33:16 <aparcar[m]> jow: ieee80211s_encryption='psk2+aes'
Sep 30 09:34:01 <aparcar[m]> I kinda remember daniel patched wpad to allow 80211s with sae
Sep 30 09:34:22 <ynezz> aparcar[m]: it's that line with gpio-button-hotplug.c
Sep 30 09:35:19 <ynezz> you've something strange going on -fmacro-prefix-map=./=
Sep 30 09:36:05 <aparcar[m]> yes I think thats legacy. I'll remove that for testing
Sep 30 09:38:35 <ynezz> anyway as you can see, you're lacking the proper -ffile-prefix-map in that command line
Sep 30 09:40:09 <Pepe> I'm curious about progress about OpenWrt 19.07. In the while, there's going to be the end of September and still no insights about RC. :(
Sep 30 09:42:45 <aparcar[m]> ynezz: where is that added? grep -r can't find it...
Sep 30 09:43:27 <aparcar[m]> only in package/libs/openssl/patches/120-strip-cflags-from-binary.patch but that doesn't seem related
Sep 30 09:47:30 <ynezz> you're adding -ffile-prefix-map in your rules.mk, don't you?
Sep 30 09:49:09 <aparcar[m]> ynezz: yes
Sep 30 09:50:13 <aparcar[m]> but apperently somewhere else the fmacro is added
Sep 30 09:50:56 <ynezz> not here, so it must be something local
Sep 30 09:52:10 <ynezz> anyway, my whole point is, that you need to get -ffile-prefix-map into the kernel module build commandline as kernel doesn't utilize TARGET_CFLAGS
Sep 30 09:53:22 <jow> Pepe: currently requires luci backports and fixes to libustream
Sep 30 09:53:31 <aparcar[m]> I see, okay gotta sleep now but try to figure it out asap
Sep 30 09:53:31 <jow> Pepe: heopefully an rc can be made within the next two weeks
Sep 30 13:34:31 <Quazil> jow: 802.11s is not appropriate for automotive applications
Sep 30 13:34:47 <Quazil> You need .... Batman.
Sep 30 14:26:51 <rmilecki> hello
Sep 30 14:27:01 <rmilecki> jow: did you have a moment to look at my luci patches?
Sep 30 15:51:31 <kristrev> blogic: Hi! I made a second attempt at getting 4.19 to work on my board, but no luck. The boot gets stuck on RCU stalls. I have put the bootlog here if you have time to take a look: https://pastebin.com/ULaGHpgK
Sep 30 15:52:43 <kristrev> blogic: In order to get a "working" image for the board, I hade to adjust the patch adding the U7623 DTS slightly. I also noticed that there is no config-4.19 for mt7623, so as a starting poing I just copied the 4.14 config
Sep 30 16:48:03 <gch981213> This guy makes me actually want to insult him now: https://github.com/openwrt/openwrt/pull/2023#issuecomment-536629254
Sep 30 16:50:44 <kristrev> haha
Sep 30 16:51:04 <kristrev> selecting random offset ... bold
Sep 30 17:21:14 <karlp> like they say, it will work most of the time...
Sep 30 17:36:50 <Hauke> gpeskens: what chips are used in the 5490?
Sep 30 17:37:34 <Hauke> I am mostly interested how the fiber part works
Sep 30 18:47:21 <gpeskens> Hauke, its a lantiq SOC + an atheros SOC connected via the lantiq switch
Sep 30 18:47:45 <gpeskens> also a spartan for, I guess, DSP
Sep 30 18:48:20 <gpeskens> lemme dive into the tree what it says about the fiber port
Sep 30 18:53:34 <gpeskens> https://pastebin.com/YxkJfSu3 <-- that's the header file that describes the switchconfig as far as I can see, unfortunately not much in the DTS files that AVM releases in the source besides GPIO
Sep 30 18:54:17 <gpeskens> optic is a coretek opto ct-1212tet-hf4c-d avme
Sep 30 18:58:52 <Hauke> gpeskens: thanks
Sep 30 19:02:25 <gpeskens> yw
**** ENDING LOGGING AT Tue Oct 01 02:59:57 2019