| *** claneys has quit IRC | 00:14 | |
| *** waltminer has quit IRC | 00:16 | |
| fury | sigh. have to update my build machine to ubuntu 16.10 just to have a working version of e2fsck :P | 00:30 |
|---|---|---|
| fury | debating whether to just throw caution to the wind and go 17.04 | 00:31 |
| fury | all so i can back up this fast 64 gig SD card (it has a raspbian install that has Qt creator on it, i was testing how fast QML was over there...) | 00:34 |
| fury | and use it for the minnowboard instead | 00:34 |
| fury | cause in order to get a compact image, i need to resize the fs to the minimum possible size, apparently in order to do that i need to first mark it clean with e2fsck | 00:38 |
| fury | it should be easier to do this :P | 00:38 |
| *** mdurnev has joined #automotive | 00:39 | |
| fury | think this old sd card is dying anyway O_o | 00:48 |
| *** waltminer has joined #automotive | 00:56 | |
| *** onoffon has joined #automotive | 02:25 | |
| *** waltminer has quit IRC | 02:26 | |
| fury | well, that was a terrible idea | 02:29 |
| fury | now the sd card keeps getting unmounted whenever i try to read a bunch of files from it or try to resize the partition | 02:30 |
| mranostay | fury: time to go to Walmart and get another dozen cards :) | 02:34 |
| fury | or order on amazon :P there's faster cards out there now... | 02:34 |
| fury | ok | 02:48 |
| fury | now i'm pretty sure i'm doing something wrong | 02:48 |
| fury | whenever i create a .hddimg through bitbake and then mkefi-agl.sh it, it does not boot up | 02:48 |
| fury | whenever i mkefi-agl.sh one of the released .hddimgs on download.automotivelinux.org, it works | 02:49 |
| fury | compared to the working boot setup, i get no ramdisk, vfs, devtmpfs, and systemd never starts | 02:51 |
| *** onoffon has quit IRC | 02:52 | |
| fury | pretty sure i pulled the same stuff as in dab rc2 | 02:56 |
| fury | cause that just happened today | 02:56 |
| fury | something is causing the initrd that i generate to not work | 03:34 |
| fury | copied over the initrd from the dab rc2 released image over to my created sd card | 03:34 |
| fury | and it works | 03:34 |
| *** mdurnev has quit IRC | 03:54 | |
| *** AlisonChaiken has joined #automotive | 04:31 | |
| *** onoffon has joined #automotive | 04:52 | |
| *** dabukalam_ has joined #automotive | 05:08 | |
| *** dabukalam has quit IRC | 05:22 | |
| *** leon-anavi-cloud has quit IRC | 05:22 | |
| *** leon-anavi-cloud has joined #automotive | 05:22 | |
| *** albanc has quit IRC | 05:30 | |
| *** mdurnev has joined #automotive | 05:30 | |
| *** albanc has joined #automotive | 05:32 | |
| *** mfritzsch has joined #automotive | 05:40 | |
| *** clopez has quit IRC | 05:44 | |
| *** clopez has joined #automotive | 05:47 | |
| *** _corrupt has quit IRC | 06:03 | |
| *** msisov has joined #automotive | 06:17 | |
| *** dabukalam_ has quit IRC | 06:21 | |
| *** dabukalam has joined #automotive | 06:22 | |
| *** Vamsi has joined #automotive | 06:40 | |
| *** jobol has joined #automotive | 06:54 | |
| *** onoffon has quit IRC | 06:55 | |
| *** onoffon has joined #automotive | 06:58 | |
| jobol | hi | 07:07 |
| jobol | is smurray at its keyboard now? | 07:07 |
| *** rajm has joined #automotive | 07:10 | |
| *** claneys has joined #automotive | 07:12 | |
| *** ToxicGumbo-work1 has joined #automotive | 07:12 | |
| mranostay | jobol: unlikely 3a in his timezone | 07:12 |
| * mranostay sleeps now as well | 07:13 | |
| jobol | mranostay, thx | 07:13 |
| *** ToxicGumbo-work has quit IRC | 07:14 | |
| *** kooltux__ has joined #automotive | 07:21 | |
| *** bruce has joined #automotive | 07:29 | |
| *** bruce is now known as Guest19422 | 07:30 | |
| *** kooltux__ has quit IRC | 07:31 | |
| *** onoffon has quit IRC | 07:41 | |
| *** [AD]Turbo has joined #automotive | 07:45 | |
| [AD]Turbo | hi there | 07:45 |
| *** kbroulik has joined #automotive | 07:55 | |
| *** rdale has joined #automotive | 07:57 | |
| *** toscalix has joined #automotive | 08:04 | |
| *** dominig has joined #automotive | 08:08 | |
| *** rajm has quit IRC | 08:36 | |
| *** rajm has joined #automotive | 08:42 | |
| *** fredw has joined #automotive | 08:45 | |
| *** fredw has quit IRC | 09:00 | |
| *** gunnarx has quit IRC | 09:02 | |
| *** fredw has joined #automotive | 09:03 | |
| *** praneeth has quit IRC | 09:44 | |
| *** praneeth has joined #automotive | 09:44 | |
| *** User3209 has joined #automotive | 09:47 | |
| *** fredw has quit IRC | 10:27 | |
| *** User3209 has quit IRC | 10:49 | |
| *** zmc has quit IRC | 11:03 | |
| fury | g’mornin | 11:06 |
| *** fredw has joined #automotive | 11:44 | |
| *** gunnarx has joined #automotive | 12:02 | |
| *** gunnarx has joined #automotive | 12:02 | |
| smurray | jobol: still there? | 12:07 |
| jobol | yes | 12:07 |
| *** gunnarx has quit IRC | 12:07 | |
| jobol | smurray, I was reading what was said this night and found you wrote "slight issue atm is the systemd reloads required to have the framework see them has been removed" | 12:13 |
| smurray | jobol: right | 12:13 |
| jobol | smurray, do you see it, a blank home screen, at first boot? | 12:14 |
| jobol | I don't | 12:14 |
| jobol | anyway I filled https://jira.automotivelinux.org/browse/SPEC-725 | 12:14 |
| smurray | jobol: I'll test again, but I was seeing behavior that I thought would be caused by that | 12:15 |
| smurray | jobol: I've ruled out the systemd reload being a root cause of the tty issue, so if you need to put it back, it should be fine | 12:16 |
| *** Guest19422 has quit IRC | 12:20 | |
| smurray | jobol: I'll retest today with an up to date clean build w/o any local changes | 12:24 |
| *** Vamsi has quit IRC | 12:26 | |
| jobol | smurray, great, thx for retesting, i'll probably no go back because doing reload at the end is enough | 12:27 |
| smurray | jobol: yes, that would be good. Do you have a change for that in the works? | 12:27 |
| *** caiortp has joined #automotive | 12:27 | |
| jobol | smurray, but if afm-user-daemon starts before end of post-install / first boot, the issue will exist | 12:28 |
| jobol | smurray, Ronan works on SPEC-725, he was near achievment in the morning (our morning) so soon | 12:28 |
| smurray | jobol: that should be manageable with a dependency, I think. Just need to add the right thing to Before in the run-agl-postinsts | 12:29 |
| *** waltminer has joined #automotive | 12:34 | |
| *** zmc has joined #automotive | 12:54 | |
| *** onoffon has joined #automotive | 12:59 | |
| *** msisov has quit IRC | 13:10 | |
| *** mdurnev has quit IRC | 13:12 | |
| *** hitendra has joined #automotive | 13:14 | |
| hitendra | hi, I'm trying to get wifi dongle working on m3ulcb, the settings app comes up, but it does not list the wifi connections, I see this error message in journal | 13:15 |
| hitendra | Jul 03 22:39:27 m3ulcb afb-daemon[4523]: while calling GetProperties method [../../git/binding-wifi/wifi-connman.c:117,wifi_state] Jul 03 22:39:27 m3ulcb afb-daemon[4523]: ERROR: [API wifi-manager] Error GDBus.Error:org.freedesktop.DBus.Error.UnknownObject: Method "GetProperties" with signature "" on interface "net.connman.Technology" doesn't exist | 13:15 |
| hitendra | https://pastebin.com/AqpHQW76 | 13:16 |
| fury | i get that on mine when i don't have the dongle connected... does it show up in ifconfig? | 13:16 |
| hitendra | nope | 13:20 |
| fury | hmm | 13:20 |
| smurray | that would suggest the driver isn't getting picked up | 13:21 |
| fury | check /proc/config.gz i think, make sure CONFIG_ATH9K_HTC is set to m or y | 13:21 |
| hitendra | it recognizes Bluetooth dongle | 13:21 |
| fury | if it is not, then it didn't build the kernel with that driver | 13:21 |
| hitendra | CONFIG_ATH9K_HTC=m | 13:27 |
| hitendra | CONFIG_ATH9K_HW=m | 13:27 |
| *** bruce has joined #automotive | 13:27 | |
| fury | hmm O_o | 13:27 |
| hitendra | CONFIG_ATH9K_COMMON=m | 13:27 |
| fury | journalctl -f | 13:27 |
| hitendra | but, # CONFIG_ATH9K is not set | 13:27 |
| fury | then unplug the dongle and plug it back in | 13:27 |
| *** bruce is now known as Guest84707 | 13:27 | |
| fury | oh | 13:28 |
| fury | https://wireless.wiki.kernel.org/en/users/Drivers/ath9k_htc | 13:28 |
| fury | this one says you need CONFIG_ATH_COMMON too maybe | 13:28 |
| fury | i don't know where i put my dongle >.> | 13:29 |
| fury | hmm | 13:32 |
| fury | aac files play on porter and pi but not minnowboard | 13:32 |
| hitendra | CONFIG_ATH_COMMON is also set, CONFIG_ATH_COMMON=m | 13:33 |
| CTtpollard | fury: just aac? | 13:33 |
| *** onoffon has quit IRC | 13:34 | |
| fury | well, i already had gstreamer1.0-plugins-ugly-mpg123 in my image install append so it plays mp3s | 13:34 |
| fury | i don't know which plugin does AAC... my first thought was faad, but it doesn't find it in any of the feeds | 13:35 |
| JEEB | libavcodec has AAC decoding | 13:35 |
| JEEB | (which either comes from FFmpeg or Libav, so whatever gst plugin uses that :)) | 13:35 |
| JEEB | (FAAD had a reason to exist about eight-nine years ago, but most of the bugs in lavc regarding something that faad could decode have been since closed AFAIK) | 13:36 |
| JEEB | otherwise you're using some other multimedia thing that isn't OSS on the thing where it works (such as HW decoding or so) | 13:37 |
| *** gunnarx has joined #automotive | 13:37 | |
| *** gunnarx has joined #automotive | 13:37 | |
| fury | makes sense | 13:38 |
| *** onoffon has joined #automotive | 13:42 | |
| *** mfritzsch has quit IRC | 13:53 | |
| *** AlisonChaiken has quit IRC | 14:11 | |
| fury | dominig: yep, the missing netboot was the problem. after 2 years of fiddling around with AGL and never needing netboot, i kinda just skimmed over the aglsetup.sh line entirely thinking "it's just the same as porter, qemu, raspberry pi, wandboard, etc..." | 14:25 |
| fury | thanks, and sorry for taking up so much of your time trying to figure it out :D | 14:25 |
| *** waltminer has quit IRC | 14:36 | |
| *** waltminer has joined #automotive | 14:46 | |
| *** hitendra has quit IRC | 15:07 | |
| *** onoffon has quit IRC | 15:15 | |
| *** hitendra has joined #automotive | 15:25 | |
| *** jobol has quit IRC | 15:28 | |
| *** [AD]Turbo has quit IRC | 15:29 | |
| *** onoffon has joined #automotive | 15:37 | |
| *** mdurnev has joined #automotive | 15:44 | |
| *** fredw has quit IRC | 15:50 | |
| fury | JEEB: adding gstreamer1.0-libav to my image did the trick, thanks for pointing me in the right direction :D (i think the other images include it by default in their BSP layer) | 15:51 |
| *** onoffon has quit IRC | 15:51 | |
| *** rajm has quit IRC | 15:52 | |
| mdp | fury, are you saying that the images produced without meta-netboot's initrd enabled are non-functional? | 15:58 |
| *** onoffon has joined #automotive | 16:02 | |
| JEEB | fury: coolio | 16:05 |
| *** mdurnev has quit IRC | 16:05 | |
| *** onoffon has quit IRC | 16:10 | |
| fury | mdp: yes–aglsetup.sh without agl-netboot produces a dud initrd for Minnowboard. When I replaced it with the initrd from the .hddimg that was released for dab rc2 it worked fine | 16:10 |
| fury | and when I did a -f including agl-netboot it worked fine as well | 16:11 |
| mranostay | fury: qemu build? | 16:11 |
| fury | intel-corei7-64 | 16:11 |
| fury | for Minnowboard | 16:11 |
| fury | the docs for that board say to include agl-netboot but I didn’t see it, so I spent a good day or so scratching my head | 16:12 |
| smurray | that's odd, it was not required for CC, kind of seems like a regression, tbh | 16:13 |
| *** onoffon has joined #automotive | 16:13 | |
| fury | I don’t mind either way, I just wish the script was a little more idiot proof for my sake :D but now that I got the Minnowboard quad core, I’m probably not going to switch boards for a while | 16:15 |
| fury | if a machine type requires certain features I’d think those features would be baked into the include in agl-distro or wherever it is that defines all the types it supports | 16:16 |
| fury | but yeah, also seemed odd to me that netboot is required to build the right initrd | 16:16 |
| fury | I suppose I should look into what it’d take to use netboot anyway, I’m kinda tired of flashing SD cards | 16:17 |
| fury | is it a writable filesystem once it boots up, can I change it on the fly (like to add libhomescreen which seems to be missing from master at the moment?) | 16:18 |
| fury | half the apps fail to start due to no libhomescreen so I had to add that to my image install append and copy the .so to /usr/lib | 16:19 |
| fury | that was pretty cool that I could just add a new library without rebooting | 16:19 |
| fury | I’m still reeling over the fact that (aside from whole image reflashes) changing something and seeing the result only takes seconds, not minutes, whether it’s changing a QML template, adding a new library, or recompiling an app | 16:20 |
| fury | 5-6 years in microcontroller programming is hard to untrain :P | 16:21 |
| smurray | fury: it's based on nbd, so yeah, rootfs is writeable | 16:21 |
| mdp | that's a clear regression. but not entirely surprising because 1) AGL CI requires netboot 2) a good number of core AGL developers build mostly with netboot because it's part of their deployment workflow | 16:21 |
| fury | cool, that’s even easier than flashing SD cards | 16:22 |
| mdp | fury, +1. once you switch you can help to break non-netboot as well ;) | 16:22 |
| fury | :P | 16:22 |
| *** kbroulik has quit IRC | 16:25 | |
| fury | how do I get multitouch to work? I only seem to be getting a single touch in my app | 16:27 |
| fury | in all the apps I guess | 16:30 |
| smurray | I suspect you'll need to hit the mailing list for that one | 16:36 |
| *** onoffon has quit IRC | 16:41 | |
| *** onoffon has joined #automotive | 16:42 | |
| *** toscalix has quit IRC | 16:58 | |
| waltminer | fury - maybe you could write a little post about the benefits of netboot for those old school guys who like SD cards | 16:58 |
| *** AlisonChaiken has joined #automotive | 16:59 | |
| waltminer | people who have been using it for a while forget the amazing things it enables compared to the “old way” | 16:59 |
| fury | I’ll see if I can whip something up, soon as I figure out how to do it :D | 17:00 |
| *** clopez has quit IRC | 17:00 | |
| *** clopez has joined #automotive | 17:04 | |
| fury | is there a place on the docs site or the wiki where it describes how to set up a netboot or is it more complicated than that? | 17:10 |
| mdp | waltminer: true, but the old school people have been using netboot forever..it's just that it was previously nfsroot rather than nbd..the latter does has advantages..best to focus on those. | 17:12 |
| mdp | waltminer: the regression is a problem for people moving on from development to actual deployment | 17:12 |
| waltminer | see! | 17:12 |
| waltminer | you proved my point | 17:13 |
| mdp | just trying to be your Yes Man | 17:14 |
| mdp | I do work PT in MarCom ;) | 17:14 |
| fury | eeh, some of these guides I’m googling are talking about installing an FTP server and DHCP server? what if the network already DHCPs and I don’t want to anger the IT department (too much)? is it possible to just set up the PXE part of it on my build server, or do they have to go hand in hand? | 17:17 |
| mranostay | no jobol when you need him... pesky timezones :) | 17:19 |
| smurray | fury: DHCP and TFTP server. The nbd bit you can run on demand as your user | 17:20 |
| fury | it’s all windows servers :( | 17:21 |
| smurray | ouch | 17:21 |
| fury | the only Linux boxes in the building are the ones I’ve set up, as far as I know | 17:21 |
| smurray | I use a private network for my netboot testing, using a USB ethernet adapter off of a box | 17:22 |
| mdp | +1 for private network | 17:24 |
| fury | hmm. I think there was the possibility of setting up a separate network for such things, but that was a while back and I didn’t think it’d be necessary. Now that I know I need a whole DHCP/TFTP/PXE setup for this netboot (and coincidentally also for a Container Linux kubernetes cluster) it might be time to ask if it’s still possible | 17:24 |
| mdp | with all the buggy bsps out there, you generally don't want target boards on your public network | 17:24 |
| fury | hmm | 17:24 |
| fury | true, but also need to be able to hook up a PC or a phone to it (either to SSH in or connect to a server app running on AGL) | 17:25 |
| fury | on the same network as all those phones and PCs, easy peasy...private network tho, not every one of them has a second Ethernet port, etc. | 17:26 |
| fury | But on a private network not as worried to nmap and try to root all the SSHs because the minnowboard for whatever reason likes to get a new IP every time :P | 17:28 |
| smurray | you can fix that with your own network and DHCP server ;) | 17:28 |
| fury | oh yeah | 17:29 |
| fury | I have a feeling he’ll be displeased that I didn’t think of this 6-12 months ago when the subject of a private network came up :D | 17:31 |
| fury | cause the wiring has to run through like half the building | 17:31 |
| fury | in order to hit all the boxes it needs to hit | 17:31 |
| mranostay | just plug into the corporate network.. i'm sure IT won't mind :) | 17:32 |
| fury | I’m already on the corporate network :P but it already has DHCP, windows server 2000something | 17:32 |
| smurray | VPN an option? | 17:32 |
| fury | hmm. not sure. there’s a VPN I log into to get to the network from outside? | 17:33 |
| fury | but didn’t know you could make a VPN on the inside | 17:33 |
| fury | much to learn, I still have | 17:33 |
| fury | I believe I created one when I fiddled around with a bunch of virtual machines to try Container Linux | 17:34 |
| smurray | sorry, s/VPN/VLAN/ | 17:34 |
| fury | cause that created a PXE that made the other two nodes boot | 17:34 |
| fury | oh | 17:34 |
| fury | yeah that’s what it made | 17:34 |
| mranostay | fury: setup your own DHCP server.. IT loves that as well :) | 17:36 |
| *** akilaw has joined #automotive | 17:36 | |
| fury | :D | 17:36 |
| akilaw | Hi | 17:37 |
| fury | would it even be possible to netboot via a wifi dongle? | 17:38 |
| mranostay | fury: reminds me the time i misconfigured a DHCP server in a lab... had the grumpy IT person at my desk in 5 minutes :) | 17:38 |
| smurray | mranostay: heh, I worked on a product with a DHCP server in it, IT ended up putting my dept on its own subnet ;) | 17:38 |
| fury | and hi akilaw | 17:38 |
| mranostay | same IT person loved the reason we engineers needed root to nfsroot off a server :) (this is pre docker containers) | 17:39 |
| mranostay | smurray: guessing netboot solves that too? :) | 17:39 |
| smurray | the nbd stuff uses a port > 10k, so no worries | 17:40 |
| *** claneys has quit IRC | 17:45 | |
| mdp | smurray: my colleague wiped out the entire Moto comp group public network with his buggy ethernet driver on a vme board. | 17:56 |
| mdp | smurray: "back in the day" | 17:56 |
| mdp | smurray: half a day for IT to restore networking for 1000+ employees there ;) | 17:57 |
| smurray | mdp: heh | 17:58 |
| mranostay | mdp: how he manage that one? | 17:59 |
| mranostay | ARP flood or something? | 17:59 |
| mdp | mranostay: it was spraying malformed packets as he was debugging his driver..whatever switches they had were triggered into some deadlocked state by it..as I understand something to do with their logging/recovery mechanism which was buggy. | 18:00 |
| mdp | buggy bugs is the industry norm ;) | 18:00 |
| mdp | mranostay: our isolated network lab (remote accessible) was on floor 1 of the building but our cubicles were on floor 2..so he was violating policy, being lazy hacking at his desk plugged into the public network. hehe | 18:02 |
| mranostay | mdp: get a nice talking to? | 18:02 |
| mdp | mranostay: those things tend to give you religion about keeping boards off important public networks ;) | 18:02 |
| mdp | mranostay: yeah..but he was a rock star..untouchable..learned way too much from him :P | 18:03 |
| mranostay | mdp: reminds me of former employer who's network got flood by the QA guys "testing" video streaming :) | 18:03 |
| mdp | mranostay: it's not like anybody can afford to get rid of the top 5% | 18:04 |
| mranostay | seems the multicast rules were off :) | 18:04 |
| mdp | lol | 18:04 |
| mranostay | also there was TIPC floods off carrier boards we couldn't ever find :) | 18:04 |
| mdp | "It's software all the way down" | 18:05 |
| *** waltminer has quit IRC | 18:12 | |
| khem | now a days you would qemu for debugging | 18:32 |
| mdp | except when qemu doesn't work | 18:37 |
| mdp | and then you debug qemu | 18:37 |
| *** Guest84707 has quit IRC | 18:38 | |
| *** waltminer has joined #automotive | 18:54 | |
| fury | is it supposed to auto mount a flash drive when i plug it in and start scanning it for music files? | 19:01 |
| fury | cause it totally isn't | 19:01 |
| mranostay | fury: which platform? | 19:02 |
| fury | i can see it get attached but nothing seems to happen after that (no mounting) | 19:02 |
| fury | minnowboard | 19:02 |
| fury | corei7-64 | 19:02 |
| mranostay | it is working on m3ulcb last i checked | 19:02 |
| mranostay | systemctl status udisks | 19:02 |
| mranostay | what does that show? | 19:02 |
| fury | https://www.irccloud.com/pastebin/C2nT1glS/ | 19:03 |
| fury | failed to start | 19:03 |
| fury | rmdir: failed to remove '/media/*': No such file or directory | 19:04 |
| mranostay | heh lovely.. well i built a m3ulcb image so i'll test that | 19:05 |
| mranostay | architecture shouldnt' matter (tm) :) | 19:05 |
| mdp | mranostay: too bad qemu usb isn't better, we could have just stayed there. | 19:06 |
| fury | why not rm -rf /media/* | 19:06 |
| fury | that doesn't result in an error when it's already empty | 19:07 |
| fury | oh i guess it's the &> /dev/null after the || true | 19:12 |
| fury | if i remove that bit and systemctl start udisks it works fine | 19:12 |
| fury | until a minute and a half later when Jul 01 11:03:32 intel-corei7-64 systemd[1]: udisks.service: Start-post operation timed out. Stopping. | 19:13 |
| fury | i guess it doesn't like that automount.sh runs continuously? | 19:19 |
| mranostay | hmm maybe | 19:22 |
| fury | and since it killed in the background i guess it didn't auto clear lightmediascanner's database or sometihng cause i sqlited in to see and it still has files from the other two flash drives i plugged in | 19:23 |
| mranostay | fury: yeah the database stays | 19:23 |
| mranostay | that is how lms works | 19:23 |
| fury | oh ok | 19:23 |
| fury | ooh, Jul 01 11:14:43 intel-corei7-64 afb-daemon[27073]: (mediaplayer:27076): GStreamer-CRITICAL **: gstsystemclock: write control failed in wakeup_async: 24:Too many open files | 19:24 |
| fury | i guess that's a bad thing? | 19:24 |
| fury | only 13,621 files open | 19:24 |
| mranostay | heh | 19:24 |
| mranostay | fury: 13k files of what? | 19:25 |
| mranostay | but yeah i probably need to do some systemd service magic | 19:26 |
| fury | i have every single app open | 19:27 |
| fury | each of them has several libraries open | 19:27 |
| fury | but mediaplayer especially, it has 2k files open | 19:27 |
| fury | across many different threads i guess | 19:27 |
| fury | i guess it's opening every mp3 file at once and grabbing id3 tags and whatnot | 19:29 |
| mranostay | yeah it is.. | 19:29 |
| mranostay | and putting it on the playlist.. | 19:29 |
| fury | that'd prolly do it. i plugged in a 16 gig flash drive full of songs | 19:29 |
| mranostay | yeah it is known design issue.. | 19:30 |
| fury | 723 mp3s | 19:30 |
| fury | my bad :D | 19:30 |
| fury | navi had a bunch of threads going too | 19:30 |
| fury | root@intel-corei7-64:~# lsof | wc -l | 19:35 |
| fury | 415244 | 19:35 |
| fury | notbad.jpg | 19:35 |
| mranostay | heh | 19:46 |
| hitendra | @fury, I just cannot get AGL on m3ulcb recognize the wifi dongle, I have run out of ideas, any last thoughts before I post it on ML? | 19:48 |
| mranostay | er does it have firmware you need to put in /lib/firmware | 19:49 |
| hitendra | root@m3ulcb:/# connmanctl enable wifi | 19:51 |
| hitendra | Error wifi: Method "SetProperty" with signature "sv" on interface "net.connman.Technology" doesn't exist | 19:51 |
| hitendra | @mranostay, this is what i see under /lib/firmware https://pastebin.com/RZ4ZfBre | 19:54 |
| mranostay | i assume you are using a wifi device with a realtek chipset? | 19:56 |
| dl9pf | ronan_: ping | 19:58 |
| dl9pf | fury: maybe this one helps with some pointers on the tftp and dns , just skip the lava stuff ... https://docs.google.com/document/d/1K7RgFsl2Z-JT6_fvxHBDA66sVciK5MGmhNPIW8zXUo0/edit?usp=sharing | 20:01 |
| dl9pf | fury: note - the lava package is old, redoing it atm | 20:03 |
| smurray | dl9pf: just wondering, did you ever cook up a workaround for the power button on the m3 board for the CI setup? | 20:03 |
| mranostay | the floor is lava! :) | 20:03 |
| dl9pf | yes, renesas provided a solution with an atmega attached to the fan for power and one gpio soldered to the power button :/ ... | 20:04 |
| fury | sweet | 20:04 |
| dl9pf | so still poor man's solution might be bridging the 0 ohm resitor | 20:04 |
| dl9pf | pads that were left | 20:05 |
| dl9pf | unpopulated | 20:05 |
| dl9pf | their solution is fully 'firmware upgrade proof' ... | 20:05 |
| *** onoffon has quit IRC | 20:05 | |
| fury | hitendra: follow the journalctl and see what it shows when you disconnect and reconnect the dongle - i'm curious to know if it's even seeing the thing pop up on the bus | 20:06 |
| dl9pf | smurray: do you want the # of the resistor ? don't remember it fttomh . | 20:06 |
| smurray | dl9pf: nah, that's okay | 20:07 |
| fury | dl9pf: that's really cool, so you basically can just set up a raspberry pi or any linux box with dnsmasq and tftpd operating on the secondary eth | 20:09 |
| dl9pf | yes | 20:09 |
| dl9pf | plus power control over relay or whatever | 20:09 |
| fury | now i just need to hunt down a usb to ethernet or some box with two ethernet ports | 20:09 |
| dl9pf | and you can go for a pina collada with your laptop and run a test | 20:09 |
| fury | that sounds fantastic :D | 20:09 |
| dl9pf | fury: find some cheap usb dongles with asix usb2 (100mbit) ... these are very reliable and even work with most u-boot versions (when connected to usb on the board). | 20:10 |
| fury | how would i make it do UI tests tho, i have a feeling that's where i'm going to be spending most of my time testing things...robot with a stylus on it? :P | 20:10 |
| dl9pf | for UI tests look at e.g. openQA ... from SUSE (suse obviously and meanwhile also fedora are using it) | 20:11 |
| dl9pf | it spawns kvm vm's | 20:11 |
| fury | ooh | 20:11 |
| dl9pf | but a lot of work to setup the "needles" aka snapshots of the expected UI on the monitor | 20:12 |
| dl9pf | e.g. look at this: https://openqa.opensuse.org/tests/440384 | 20:13 |
| hitendra | @mranostay, i'm using CES2017 recommended dongle which is TL-WN722N, built on AR9271 chipset and requires ath9k_htc drivers | 20:14 |
| fury | ... https://openqa.opensuse.org/tests/440384/file/video.ogv that is the coolest thing i think i've ever seen in my life | 20:15 |
| hitendra | @fury, journalctl only shows USB disconnect and connect events, https://pastebin.com/s77tapLk | 20:17 |
| smurray | hitendra: I've got the h/w here, give me a bit and I'll try building an image to reproduce | 20:18 |
| fury | yep...i guess driver isn't detecting it for some reason | 20:18 |
| fury | smurray: you'll have to add the kernel module for it, it doesn't seem to be in the default m3ulcb image | 20:18 |
| smurray | yeah, I just did that in my tree | 20:19 |
| smurray | missing the driver for the rtl-sdr dongle as well from the looks of it | 20:19 |
| fury | speaking of kernel modules... for the intel-corei7-64, the kernel is up to 4.8 but the bbappend that adds the usb device modules, it's a 4.4.bbappend...should that be bumped up to 4.8.bbappend or what? | 20:19 |
| smurray | likely, or move to 4.% | 20:20 |
| fury | sounds good | 20:20 |
| hitendra | i copied over *.cfg from meta-agl/meta-agl-bsp/meta-renesas/recipes-kernel/linux/linux to meta-agl/meta-agl-bsp/meta-rcar-gen3 | 20:21 |
| fury | i've been doing some impromptu agl demos for people in the building and i've been able to completely lock it up a couple times... i think both times it was navi related | 20:22 |
| fury | like something that it's doing is causing the whole thing to lock up, SSH and all | 20:22 |
| *** onoffon has joined #automotive | 20:23 | |
| smurray | hitendra: note that just copying the .cfg file isn't enough, you need to tweak the bbappend to apply it | 20:25 |
| smurray | fury: the navi app is a bit of a black box, unfortunately | 20:26 |
| fury | hmm... there ought to be a non-black box out there somewhere :P | 20:26 |
| fury | or maybe the effort it takes to make one of those things is enough that everyone who's done it well so far wants to get paid for it | 20:27 |
| fury | guess it could be either way | 20:27 |
| fury | open street map lists a load of open source offerings that use their data | 20:28 |
| smurray | yes | 20:28 |
| fury | http://doc.qt.io/qt-5/location-plugin-osm.html | 20:29 |
| smurray | one thing to note is there's no location service binding or the like yet | 20:30 |
| fury | yup | 20:30 |
| hitendra | @smurray, sorry i meant say both, *.cfg and corrponding bbapends | 20:31 |
| hitendra | Are these the correct firmware files for TL-WN722N? https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/ath9k_htc | 20:37 |
| hitendra | if so, i see both those files under /lib/firmware | 20:37 |
| dl9pf | fury: submit a patch for the 4.% | 20:38 |
| dl9pf | fury: navi is pure egl app ... not sure its really bug-free | 20:38 |
| fury | sounds good | 20:38 |
| *** dominig has quit IRC | 20:39 | |
| JEEB | if it's C you can run clang's static analyzer on it | 20:40 |
| JEEB | among probably non-OSS ones which corporations most likely have access to | 20:40 |
| smurray | hitendra: copying the .cfg and adding it to the bbappend seems to have worked as expected here, stuff shows up in the wifi settings as I expect | 20:44 |
| hitendra | hmm, did you also copy over .patch and hibernation/ stuff? | 20:48 |
| hitendra | also, what do you see in "zcat proc/config.gz | grep ATH" ? | 20:49 |
| dl9pf | JEEB: fury: https://github.com/gpsnavi/gpsnavi thats the app | 20:49 |
| smurray | no, I copied over the ath9k_htc.cfg and added it to the SRC_URI in meta-agl-bsp/meta-rcar-gen3/recipes-kernel/linux/linux-renesas_%.bbappend | 20:50 |
| dl9pf | well ... https://github.com/AGLExport/gpsnavi | 20:50 |
| hitendra | just that one file? | 20:50 |
| hitendra | not even cfg_mac_80211.cfg ? | 20:50 |
| smurray | that one is not required | 20:51 |
| hitendra | i started with these 2 files first, but it did not work, so copied over all .cfg and even that did not work :) | 20:51 |
| smurray | https://www.irccloud.com/pastebin/2fCdV5a6/ | 20:51 |
| smurray | there's the grep of /proc/config.gz | 20:52 |
| smurray | I'm not sure what's wrong on your side | 20:52 |
| smurray | I'll upload the change to gerrit, as it makes sense to go in for demos with the rcar3 | 20:52 |
| fury | oh man, i'm a little rusty, how do i submit a change for meta-agl again? | 20:54 |
| fury | repo upload says no branches ready for upload | 20:54 |
| fury | git review asks me for gerrit credentials | 20:54 |
| smurray | git review is what I usually use | 20:55 |
| fury | i thought it was supposed to use my SSH key? | 20:55 |
| *** onoffon has quit IRC | 20:55 | |
| fury | oh. it probably would if i had added it in the first place ._. | 20:56 |
| hitendra | @smurray, i will rebuild the image with just ath9k_htc.cfg, see if that does it, thanks for your help | 20:56 |
| fury | could not connect to gerrit O_o | 20:57 |
| dl9pf | ? | 21:01 |
| dl9pf | git review -s | 21:01 |
| dl9pf | git review | 21:01 |
| fury | ok, got it | 21:02 |
| fury | had to git review -s and type in my username first, then it picked up my ssh key | 21:02 |
| fury | https://gerrit.automotivelinux.org/gerrit/#/c/10071/ | 21:02 |
| fury | is there a quick way to add that to dab as well, or just do the same change in my local copy of dab? | 21:02 |
| smurray | you can cherry-pick in the gerrit web interface | 21:03 |
| fury | cool | 21:03 |
| fury | https://gerrit.automotivelinux.org/gerrit/#/c/10073/ | 21:05 |
| fury | i assigned it to myself in https://jira.automotivelinux.org/browse/SPEC-732 - should i assign someone else or just leave it? | 21:05 |
| fury | whoops. i broke 4.4% | 21:07 |
| fury | unless it includes 4.% before 4.4% | 21:09 |
| fury | deleted FILESEXTRAPATHS_prepend := "${THISDIR}/files:" from 4.4% and put it in 4.% - does it still find the files in 4.4%? | 21:09 |
| mdp | fury, hehe, we keep warning about that "many files" mediaplayer limitation here. That would go away when the apps are rewritten to use a qt independent media indexer and engine. | 21:11 |
| fury | :D | 21:11 |
| mdp | "patches welcome" | 21:11 |
| smurray | fury: it's safer to put the FILESEXTRAPATHS_prepend in each bbappend | 21:13 |
| fury | makes sense | 21:13 |
| fury | i'll fix that | 21:13 |
| fury | i think | 21:13 |
| fury | now what happens with my cherry picked change, do i have to cherry pick it again or go into the other change and hit rebase or something? | 21:15 |
| fury | i am such a git n00b | 21:15 |
| smurray | you did it with the gerrit UI? | 21:16 |
| fury | yeah | 21:16 |
| smurray | hmm, I've not been in that situation before | 21:17 |
| smurray | might be a question for dl9pf | 21:17 |
| fury | guess i shoulda waited, i do that a lot, commit broken stuff and only realize it's broken after i commit it, then i make another commit to clean it up | 21:17 |
| fury | if i had a good side by side diff tool on linux i'd prolly be able to catch that before it gets onto gerrit :D | 21:18 |
| smurray | sdiff? | 21:18 |
| fury | i use beyond compare on windows | 21:18 |
| smurray | I've seen something, forget the name of it | 21:18 |
| smurray | I work in a terminal almost all the time | 21:19 |
| *** onoffon has joined #automotive | 21:19 | |
| fury | i'm a gui pleb... drag and drop, DRAG AND DROP | 21:19 |
| smurray | fury: if you do happen to have a dab tree created, you could cherry-pick the change, fix it there, then re-upload it, but that's more involved | 21:20 |
| fury | https://www.youtube.com/watch?v=1Nyn0k1Btns | 21:20 |
| fury | ah | 21:20 |
| *** caiortp has quit IRC | 21:23 | |
| fury | ok | 21:26 |
| fury | yeah, that's what i did | 21:26 |
| fury | wasn't sure whether to hit the cherry pick button again or just do this, this seemed safer? | 21:27 |
| fury | repo download meta-agl 10073, fix the file, git commit --amend, git review, ezpz | 21:27 |
| smurray | yeah | 21:28 |
| fury | lucky i had the dab tree already sitting there from when i was trying to figure out what i did wrong making the image for minnowboard | 21:28 |
| smurray | it's possible gerrit is smart enough to update if you cherry-pick between branches a second time, but maybe not | 21:28 |
| fury | hot damn, beyond compare gave me a free upgrade to pro edition multiplatform | 21:31 |
| fury | the way that company does their trial period and software licensing gives me hope for humanity | 21:31 |
| fury | cause the trial period is "30 days" but if you only use it once a month, it only knocks off one day on days that you actually use it | 21:32 |
| fury | and then when you finally buy the license, they keep giving you upgrades to the new version for free | 21:32 |
| dl9pf | fury: looking, but I'm not halfway to bed ... | 21:32 |
| dl9pf | so might propose the fix in the morning | 21:32 |
| fury | dl9pf: no worries - i amended the commit in the other branch, all seems well (except i'm burning up a couple of CI cycles with the original patches now...) | 21:33 |
| dl9pf | no worries ... | 21:33 |
| fury | is there a way for me to tell it to stop the cycle for the original patch, or just don't worry bout it? | 21:33 |
| fury | also, i love this amend stuff. i've always wanted to be able to use it in-house, but don't have gerrit or gitlab or any code review type tool that works like that...i installed gitlab on ubuntu but i really want it to integrate with a container tool so i can deploy sites/apps through it, and setting up kubernetes feels like a chinese finger trap O_o | 21:36 |
| dl9pf | just leave it running ... usually it stops the build of the 1st commit when there is a 2nd ... if not ... leave it running | 21:37 |
| fury | ok cool | 21:37 |
| dl9pf | ttyl | 21:37 |
| fury | have a good one :D | 21:37 |
| *** waltminer has quit IRC | 22:32 | |
| fury | hmm | 23:22 |
| fury | minnowboard just locked up | 23:22 |
| fury | wasn't doing anything in particular, just bluetooth streaming | 23:22 |
| fury | right out of the blue | 23:22 |
| fury | no response to SSH or touch | 23:22 |
| *** khem has quit IRC | 23:26 | |
| *** rdale has quit IRC | 23:29 | |
| *** khem has joined #automotive | 23:31 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!