**** BEGIN LOGGING AT Mon Aug 26 03:03:02 2019
Aug 26 05:05:26 <rmilecki> morning
Aug 26 05:06:40 <rmilecki> jow: would you take a look at https://patchwork.ozlabs.org/patch/1151976/ [PATCH 1/2] base-files: use JSON for storing firmware validation info
Aug 26 05:06:52 <rmilecki> blogic: maybe you want to check that as well ^^
Aug 26 05:52:16 <bookworm> well, the "ugly" tag was unnecessary, no matter whether it's true or not. "needs work" provides the same information and isn't insulting
Aug 26 05:52:31 <rubberduck> good morning
Aug 26 05:52:48 <rubberduck> i managed to get my wanted vlans on my fritzbox 4040 (ipq4018 based)
Aug 26 05:53:13 <rubberduck> it turned out that if i define vlan i MUST define 1 and 2 as the first vlans
Aug 26 05:54:14 <rubberduck> and i managed to patch the board config so that luci shows the wan port in the switch config.
Aug 26 06:23:06 <mangix> ausjke: i take great offense to that. what failures are you talking about?
Aug 26 06:25:27 <blogic> rmilecki: looking good
Aug 26 06:25:33 <blogic> i just did a quick skim though
Aug 26 08:20:19 <rmilecki> can my ubus method handler call another ubus method?
Aug 26 08:20:44 <rmilecki> i'm using ubus_invoke() but after specified timeout it returns UBUS_STATUS_TIMEOUT
Aug 26 10:06:39 <jow> rmilecki: yes, but you need to use the *_async invoke family
Aug 26 10:06:49 <jow> rmilecki: otherwise your both blocking handlers will deadlock
Aug 26 10:06:51 <rmilecki> ah
Aug 26 10:06:56 <rmilecki> thanks
Aug 26 10:07:22 <jow> you can find some example implementations in rpcd
Aug 26 10:07:48 <jow> it uses async handlers since plugins it invokes could invoke ubus methods as well
Aug 26 10:08:34 <rmilecki> it's obvious now to me once you pointed that
Aug 26 10:16:45 <rmilecki> jow: did you see my request for looking at the base-files patch?
Aug 26 10:19:39 <jow> rmilecki: yes, but am in a full day meeting so likely not before tomorrow
Aug 26 10:19:59 <rmilecki> np
Aug 26 10:20:26 <rmilecki> if you just saw it, it's fine :)
Aug 26 10:21:12 <rubberduck> can you please tell me who is responsible for the ipq40xx tree?
Aug 26 10:39:29 <gch981213> rubberduck: There is a maintainer defined in target/linux/ipq40xx/Makefile (blogic). If you just want to submit a patch, other core developers could merge it as well.
Aug 26 10:39:48 <rubberduck> i'm not good in creating patches
Aug 26 10:39:59 <rubberduck> so i wanted to contact the maintainer.
Aug 26 10:44:55 <blogic> hi
Aug 26 10:45:08 <blogic> what's up with ipq40xx ?
Aug 26 10:45:34 <rubberduck> i have a fritzbox 4040 and the WAN Port is missing in the switch configuration in luci
Aug 26 10:45:42 <rubberduck> (port5)
Aug 26 10:45:46 <blogic> and you have a patch ?
Aug 26 10:45:50 <rubberduck> no
Aug 26 10:46:39 <rubberduck> i have managed to change the /etc/board.json file so that luci does correct now
Aug 26 10:47:09 <blogic> rubberduck: can you build a test image ?
Aug 26 10:47:12 <rubberduck> i found https://github.com/chunkeey/LEDE-IPQ40XX/commit/a04cf208fe317074502f7ea81dafa828c89b74bb
Aug 26 10:47:15 <blogic> or do you use snapshot images ?
Aug 26 10:47:26 <rubberduck> 18.06.4
Aug 26 10:47:42 <blogic> https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=target/linux/ipq40xx/base-files/etc/board.d/02_network;h=e5ba7260f3bcd4168def852d77646d2afc5ac947;hb=refs/heads/master#l31
Aug 26 10:47:48 <blogic> there is no wan port define here
Aug 26 10:48:25 <rubberduck> port 5 is wan
Aug 26 10:48:28 <blogic> ok
Aug 26 10:48:48 <blogic> i fragged my 404 a couple of months back
Aug 26 10:48:56 <rubberduck> fragged?
Aug 26 10:49:01 <blogic> killed
Aug 26 10:49:25 <rubberduck> completely bricked?
Aug 26 10:49:27 <rubberduck> -> bad
Aug 26 10:49:52 <blogic> yup, typo while flashing, jacking the eva
Aug 26 10:49:59 <rubberduck> jtag?
Aug 26 10:50:04 <blogic> too lazy
Aug 26 10:50:37 <blogic> so it comes up as eth0.2 now i guess
Aug 26 10:50:40 <blogic> and eth0.1 is lan
Aug 26 10:50:49 <rubberduck> eth0.2 = eth1
Aug 26 10:51:02 <rubberduck> this is hardwired in the nic driver
Aug 26 10:51:20 <rubberduck> needs to be changed _OR_ documented i think
Aug 26 10:51:21 <blogic> i thought we merged the untanlge patch
Aug 26 10:51:50 <rubberduck> which patch?
Aug 26 10:52:07 <blogic> the one that stops it adding vlans inside the eth driver
Aug 26 10:52:49 <rubberduck> i found out that with 18.06.4 you MUST define VLAN2 (after VLAN1) or the system will completely disconnect from network
Aug 26 10:53:04 <rubberduck> blogic: 18.06.4 or 19.07?
Aug 26 10:56:09 <blogic> its in neither
Aug 26 10:56:31 <rubberduck> i mentioned a patch earlier here
Aug 26 10:56:39 <rubberduck> the first part patching 01_leds
Aug 26 10:56:58 <blogic> yep, that is the one i mean
Aug 26 10:57:11 <rubberduck> why wasn't that integrated? directly mirroring the link state of port 5 (WAN) to the WAN LED should be better for most people
Aug 26 10:57:27 <rubberduck> https://github.com/chunkeey/LEDE-IPQ40XX/commit/a04cf208fe317074502f7ea81dafa828c89b74bb
Aug 26 10:57:37 <blogic> no idea, its not my staging tree
Aug 26 10:57:42 <blogic> try mailing chunkey
Aug 26 10:59:18 <rubberduck> i will do so
Aug 26 10:59:43 <blogic> thx !
Aug 26 11:00:37 <rubberduck> r u german?
Aug 26 11:49:16 <Tapper> GSoC 2019 - OpenWrt Firmware Wizard - Final Update https://t.co/RahUSTE5E2
Aug 26 11:49:36 <Tapper> This could be grate for OpenWrt!
Aug 26 11:51:17 <rubberduck> Tapper: it would be nice if there would be a way to use Freifunk parallel to "normal" openwrt on the same router
Aug 26 11:52:06 <Tapper> The sysupgrade will be a part of luci
Aug 26 11:52:17 <rubberduck> have several openwrt Wifi AccessPoints with Multi-SSID.
Aug 26 11:52:26 <Tapper> To do in place upgrades saving packages
Aug 26 11:52:48 <rubberduck> and my fritzbox 4040 has enough ram and CPU Power to do freifunk routing, mesh _AND_ AccessPoint
Aug 26 11:55:02 <woshty> If I do a sysupgrade, installed packages get lost but everything set in uci stays? What else does get lost/stays? Is there an overview somewhere?
Aug 26 12:01:42 <MY_R> woshty, sysupgrade -l / --list-backup
Aug 26 12:11:15 <Tapper> When the package is dun it will save configs and all installed packages as far is I understande it.
Aug 26 13:31:30 <rmilecki> i noticed that "block" tries to scare me with "hotplug-call call failed" errors
Aug 26 13:31:42 <rmilecki> it's coming from fstools's block.c
Aug 26 13:32:01 <rmilecki> # hotplug-call foo; echo $?
Aug 26 13:32:03 <rmilecki> 1
Aug 26 13:32:15 <rmilecki> it seems that calling hotplug-call for non-existing directory result in an error
Aug 26 13:32:18 <rmilecki> is that expected?
Aug 26 13:32:20 <rmilecki> blogic?
Aug 26 13:33:15 <woshty> thx
Aug 26 13:57:04 <jow> rmilecki: I'd say it makes sense that hotplug-call fails when it cannot actually call something
Aug 26 13:57:14 <rmilecki> ok
Aug 26 13:57:31 <jow> rmilecki: if the caller wants to catch the error it should probably stat /etc/hotplug.d/xxx before
Aug 26 13:57:46 <rmilecki> jow: ok
Aug 26 14:07:29 <blogic> i agree
Aug 26 14:07:30 <blogic> :-)
Aug 26 14:08:51 <jow> or a very simple fix would be adjusting the .ipk package to ship an /etc/hotplug.d/block/README or a postinstall script doing mkdir -p ${IPKG_INSTROOT}/etc/hotplug.d/block
Aug 26 18:07:41 <mrkiko> blocktrron: ping
Aug 26 18:08:13 <mrkiko> Hauke: hello! Are you active?
Aug 26 18:08:42 <mrkiko> woshty: maybe sysupgrade -l tells you somethings
Aug 26 18:09:11 <mrkiko> woshty: uci should stay, plus things lsted in sysupgrade -l + you can modify sysupgrade.conf; and maybe it's -L and not -l, sorry
Aug 26 18:09:27 <mrkiko> MY_R: sorry, didn't notice your message
Aug 26 18:31:37 <mrkiko> anyone has any ideas on where I can gather some more informations about lantiq xrx300 platform?
Aug 26 18:31:48 <mrkiko> I found something for xrx500 in k3c github repo, but
Aug 26 19:05:17 <ausjke> why is sunxi devices all have sd-image even though some has 16MB flash
Aug 26 19:08:30 <Hauke> mrkiko: hi
Aug 26 19:08:39 <Hauke> xrx300 is similar to v49
Aug 26 19:08:41 <Hauke> vr9
Aug 26 19:08:51 <Hauke> grx500 is different
Aug 26 19:10:32 <hellsenberg> lantiq is pure undocumentium imho
Aug 26 19:14:13 <Hauke> mrkiko: which device are you looking at?
Aug 26 19:14:47 <Hauke> the grx300 and grx330 are using a MIPS 34KEc, they have no DSL but Wifi integarted
Aug 26 19:14:57 <Hauke> the switch is similar to the VR9
Aug 26 19:15:40 <Hauke> GRX500 (grx350 / GRX550) is a MIPS InterAptive dual core, there are many changes in the datapath
Aug 26 19:16:03 <hellsenberg> reminds me of my ARV7510PW22, which should be supported but I can't get to work
Aug 26 19:16:28 <Hauke> recent source code for the GRX500 can be found here: https://gitlab.com/gplmirror/telekom-speedport-smart-3
Aug 26 19:16:48 <Hauke> this is based on kernel 4.9
Aug 26 19:17:20 <hellsenberg> nice
Aug 26 19:19:23 <mrkiko> Hauke: oh sorry, I wasn't here
Aug 26 19:19:34 <mrkiko> Hauke: thanks! I am looking at FRITZ!BOX 3272
Aug 26 19:19:38 <mrkiko> it has DSL
Aug 26 19:19:45 <mrkiko> and the CPU looks to be 34ck
Aug 26 19:20:07 <mrkiko> the kernel, reports it as XRX300 SOC, rev 1.2
Aug 26 19:20:29 <Hauke> there is also the ARX300 or something like this
Aug 26 19:20:34 <Hauke> which supports ADSL only
Aug 26 19:20:59 <blocktrron> mrkiko: a bit, whats up?
Aug 26 19:21:00 <Hauke> it is probably pretty similar to the VR9 / VRX200
Aug 26 19:21:09 <mrkiko> Hauke: this has both Wi-Fi and ADSL; I don't exactly know where to get more info about this device, and I can't look at labels due to the fact I am blind
Aug 26 19:22:02 <mrkiko> blocktrron: hello! I was asking you for some help on the device I am trying to port; Martin suggested me in the ML it's similar to the HomeHub 4A
Aug 26 19:22:08 <mrkiko> for which I found no sources
Aug 26 19:23:11 <mrkiko> Hauke: yeah, martin told me in the ML. I followed its suggestions - and so I tried to build a minimal DTS with memory and CPU only. But the kernel panicked at sysctrl.c, especially at line
Aug 26 19:23:20 <mrkiko> if (!np_pmu || !np_cgu || !np_ebu)
Aug 26 19:24:23 <blocktrron> hmm, im not really deep into lantiq - is there a specific question?
Aug 26 19:24:35 <mrkiko> So I copied most of VR9 - icu0 and so on. Unfortunately I am not able to understand most parts of it, so I have no cognition of what I am doing. you can do ports like this when similarities are enough but ... After copying things, the kernel panicked at plat_soc_init
Aug 26 19:24:44 <Hauke> mrkiko: pmu, cgu and ebu are mandatory and they are in the SoC
Aug 26 19:25:54 <mrkiko> blocktrron: thank for the help, I would like to ask you if you had access to any sources that derived from the apparently ongoing effort to add support for the BT HomeHub 4A, which should have a similar chipset (XRX300 or ARX300, I am getting confused), so I could have a look at them; I also took a look at UGW
Aug 26 19:26:29 <mrkiko> I found UGW here - https://github.com/paldier/K3C
Aug 26 19:26:34 <mrkiko> don't know if it's complete or even the right thing
Aug 26 19:27:31 <mrkiko> Hauke: yeah - I am thinking I am specifying some wrong IO addresses or something like that. Unfortunately I have no source code to look at or datasheets
Aug 26 19:27:56 <mrkiko> so any help would be greatly apreciated to bootstrap
Aug 26 19:28:40 <mrkiko> I also took a look in sysfs from the original firmware ... but I guess I didn't look in the right place
Aug 26 19:30:00 <blocktrron> i cant help much in that regard, as i've never dealt with that specific soc.
Aug 26 19:32:58 <mrkiko> blocktrron: thank you in any case for the kindness... If you have any suggestions, I am here.
Aug 26 19:33:35 <mrkiko> BTW - my EVA bootloader env is at http://ix.io/1Tyw ; full bootlog at http://ix.io/1Tyx
Aug 26 19:41:34 <mrkiko> Hauke: may the panic log be useful'
Aug 26 19:43:00 <Hauke> mrkiko: when it crashes in the line you gave, you removed something essential from the device tree
Aug 26 19:43:07 <Hauke> like the cgu, pmu or ebu entry
Aug 26 19:43:34 <mrkiko> i was thinking to look at the GPL code for TP-Link TD-W9980
Aug 26 19:46:45 <mrkiko> Hauke: yeah, infact. I understood it from the code so I added them from vr9 ... but now I think I am specifying something very wrong. my crash now is:
Aug 26 19:47:41 <mrkiko> Hauke: http://ix.io/1Tyz
Aug 26 20:01:29 <mrkiko> Hauke: BTW, on the TP-Link firmware I found an ar10_ref_board.c; do you think I'm heading in the right direction?
Aug 26 20:01:58 <Hauke> I do not know this probably uses a very old code base
Aug 26 20:43:24 <mrkiko> mhm...
Aug 26 20:43:46 <mrkiko> so - my platform is completely different from GRX500 or do you think it's worth checking that out as well?
Aug 26 20:44:49 <mrkiko> Hauke: I guessyou saw the mails I sent, you where in CC, if all went well with the mail system. Martin told me you're working on Lantiq Ethernet, and I also saw your kernel commits... congrats.
Aug 26 21:31:24 <mrkiko> Hauke: wow, anyway the lantiq design seems pretty ... how to say, elaborated!
Aug 26 22:21:33 <mrkiko> have a good night all guys!
**** ENDING LOGGING AT Tue Aug 27 02:59:57 2019