IRC logs for #automotive for Tuesday, 2017-07-25

« Monday, 2017-07-24 Index Wednesday, 2017-07-26 »
*** jlrmagnus has quit IRC00:15
*** mdurnev has joined #automotive00:34
*** waltminer has quit IRC00:49
*** martinkelly1 has quit IRC01:04
*** mdurnev has quit IRC02:05
furyis it bad that i'm filtering out my craigslist car searches, skipping over any car I can't easily put a double din radio in?02:17
furyi'm not asking for much... just that A/C works, it has all four wheels, a good engine and transmission...and a place for me to cram in a $1,500 touchscreen radio (when I make it...)02:18
furyi guess i am asking too much for that to be like $1,000-$1,500 :P02:20
mdpzmc, well, got busy at my destination..so here it is: https://www.dropbox.com/s/560i6k6kh62864h/agl-demo-platform-qemux86-64-20170718132302.vmdk?dl=002:50
mdpfury, get a used 2017 corvette..join the GM developer program and write html5 apps for it.02:51
furysweet02:52
furysign me up02:52
furythis looks good, but must be too good to be true - https://louisville.craigslist.org/cto/d/hyundia-sonata/6221322863.html02:54
zmcmdp: Thanks! Downloading now. I'll try it out in the morning03:04
*** mdurnev has joined #automotive03:58
*** willdye has quit IRC05:04
*** jobol has joined #automotive05:56
*** claneys has joined #automotive07:03
*** rajm has joined #automotive07:08
*** leon has joined #automotive07:11
*** leon is now known as leon-anavi07:11
leon-anavimorning07:11
*** bruce has joined #automotive07:31
*** bruce is now known as Guest3027507:32
*** mfritzsch has joined #automotive07:36
*** [AD]Turbo has joined #automotive07:39
[AD]Turbohi there07:40
*** fredw has joined #automotive07:45
*** rdale has joined #automotive07:57
*** toscalix has joined #automotive07:58
*** kbroulik has joined #automotive08:12
*** jacobo has joined #automotive08:14
*** Guest30275 has quit IRC08:35
*** bruce has joined #automotive08:35
*** bruce is now known as Guest4674708:36
*** claneys has quit IRC08:44
*** claneys has joined #automotive08:56
*** rdale has quit IRC08:58
*** rdale has joined #automotive08:59
*** idonob has joined #automotive10:28
*** kooltux has joined #automotive10:46
*** araujo has quit IRC10:51
*** zeenix has joined #automotive10:54
*** fredw has quit IRC11:01
*** araujo has joined #automotive11:03
*** araujo has joined #automotive11:03
Tartarusdl9pf: Was that vmdk failure on master or something else?11:27
Tartarusdl9pf: Was that vmdk failure on master or something else?11:27
dl9pfyep11:27
Tartarusmaster? OK, I'll take a look at it all again shortly, I suspect I know what's going on, sigh11:28
*** waltminer has joined #automotive11:45
*** kooltux has quit IRC11:56
Tartarusdl9pf: ugh, sorry, I'll upload a new version shortly, forgot to grab one other class we'll need to import12:01
TartarusThis ends up being cleaner upstream at least, just need to get some underlying bugfixes in first12:01
dl9pfTartarus: yes, good work upstream.12:10
*** Vamsi has joined #automotive12:33
*** zeenix has quit IRC12:43
* Tartarus kicks repo12:46
*** mdurnev has quit IRC13:23
*** cajun-rat has joined #automotive13:37
*** chbae has joined #automotive13:46
furyi must've broken my host system or something14:00
furycause completely fresh build dir, sstate cache, downloads dir, and i'm still getting segfaults when trying to compile with qt14:00
*** jlrmagnus has joined #automotive14:31
*** cajun-rat has quit IRC14:33
waltminermdp I reopened SPEC-517 because the boot steps are not rendering correctly on the docs page14:37
mdpyeah, I've mentioned in there why that is14:40
mdpwhat happens is that docs.automotivelinux.org is using a different converter (one that is far more pedantic than pandoc's default markdown to html converter) and so even if you preview your markdown for correctness, docs renders it differently.14:41
mdpand I sent a PR yesterday to correct that14:42
mdpwe probably need somebody that did the docs. converter to update docs-agl's recommended preview commands to match14:42
mdpI know even github-style markdown and that renderer shows my .md file fine14:43
waltminerI saw that.14:44
mdphttps://github.com/automotive-grade-linux/docs-agl/pull/7114:44
waltminerEverywhere else I looked on our site people seemed to just manually create a list in the md14:44
mdpI recall seeing this in some .md parser ..most are more tolerant of white space before code blocks and the unordered list whitespace requirement14:45
mdpwaltminer: it's just an indent issue on the unordered list14:45
waltminerOk. I see the pull request. I will merge it in14:46
*** chbae has quit IRC14:52
*** jlrmagnus has quit IRC15:08
zmcmdp: I just booted that image that you posted. Works great! Thanks alot!15:08
mdpzmc, great, np!15:10
*** Vamsi has quit IRC15:24
*** jlrmagnus has joined #automotive15:31
*** jlrmagnus has quit IRC15:37
*** martinkelly has joined #automotive15:39
*** [AD]Turbo has quit IRC15:46
*** rajm has quit IRC15:48
*** mfritzsch has quit IRC15:48
*** jlrmagnus has joined #automotive15:49
waltminermdp - looks better.16:05
waltminerplease re resolve16:05
*** iSaul has quit IRC16:17
*** kbroulik has quit IRC16:22
*** Guest46747 has quit IRC16:23
*** Net147 has quit IRC16:28
*** Net147 has joined #automotive16:30
*** waltminer has quit IRC16:37
*** claneys has quit IRC16:42
*** iSaul has joined #automotive16:44
*** toscalix has quit IRC16:47
* mranostay hopes fury didn't start an arduino 101 AGL port 17:05
*** CTtpollard has quit IRC17:10
mranostayjobol: so about the race condition with rfkill on bt17:27
*** leon-anavi has quit IRC17:30
*** jacobo has quit IRC17:44
jobolhi17:45
jobolsystemd-rfkill?17:45
jobolmranostay, when is arriving the race condition?17:49
mranostayyeah thay doesn't work17:50
mranostay*that17:50
mranostaybecause systemd-rfkill doesn't understand multiple bluetooth interfaces17:50
jobolmultiple interfaces... I'm not used with bluetooth17:52
jobolSo the scenario is when a new interface is plug or when there is more than one interface17:53
jobolhello how is it hangin? (politess...)17:53
mranostaymore than one.. then again this only tested on rpi317:53
mranostayi can double check today17:54
jobolI can put more than one bluetooth dongle on my board17:54
mranostayyeah but does system-rfkill work? :)17:54
jobolI don't know17:55
jobolI just discovered systemd-rfkill few our ago while doing something else17:55
mranostayit is probably the correct way (tm)17:56
mranostayor doing a ioctl call to /dev/rfkill17:56
mranostaybecause that system() call is just horrible in the BT binding :)17:57
jobolI wonder the power attribute of bluez adaptors is a kind of rfkill but I really dont know17:58
jobolon my board I have systemd-rfkill running17:58
jobolIts effect is to restablish the state thet rfkill had at power down17:59
mranostayjobol: i think it was because the usb dongle settles after the system-rfkill runs18:00
jobol       #include <stdlib.h>18:00
jobol       int system(const char *command);18:00
mranostaysystemd and the kernel is all one big race condition18:00
jobolis it the system() you wrote about?18:00
jobolbig ! :)18:01
mranostayjobol: well it is a horrible syscall to use18:04
mranostaysecurity for one18:04
jobolI installed a service that also made rfkill unblock 0, removing it restarting the board and the 2 hci are unblocked18:04
jobolI block the 2 hci and restart and none of them is blocked18:07
mranostayheh18:10
mranostayheh maybe my USB hub is slow.. i'll test that is afternoon18:10
joboland files in /var/lib/systemd/rfkill are containing zero!18:10
jobolJul 25 18:06:48 m3ulcb bluetoothd[3397]: Failed to set mode: Blocked through rfkill (0x12)18:12
jobolJul 25 18:06:48 m3ulcb connmand[3483]: Bluetooth device /org/bluez/hci1 not enabled Blocked through rfkill18:12
jobolUnderstood: I use the version with system()18:21
*** ZenEntropy has joined #automotive18:27
mdpmranostay: can't you just listen to /dev/rfkill in the binding for change events to before initializing to avoid the race?18:36
mranostayah that is a idea18:36
mranostayhad to put a wait condition in for the autoconnect anyway18:37
mdpsorry, that's sorta what I was implying the other day and didn't elaborate.18:37
mdpthat's the problem in general with systemd units is they complete and stuff isn't _done_...so you really need to go directly to the kernel abi and verify your interface is powered up before allowing the binding to init.18:38
*** akilaw has joined #automotive18:49
akilawasaktogta918:50
*** waltminer has joined #automotive18:50
RzRakilaw, you can create an httpd server in gdp using nodejs it's in18:54
jobolmranostay, it make 50mn that I look around rfkill and I think that using /dev/rfkill directly is a valuable solution18:56
mranostayit is the correct way as well18:56
mranostayi'll look at this afternoon..18:57
jobolThanks for using picocom, syt18:59
*** jobol has quit IRC18:59
akilawHere what I have done inside my pc https://at.projects.genivi.org/wiki/display/PROJ/GENIVI+GSoC+2017+-+Testing+Zebra+UI+mockups19:11
ZenEntropyHi mdp, I tried building a VBox machine following instructions on your doc, downloaded the AGL image, but can't even pass the "Booting the kernel" message. Any hack I would have to apply before booting? or within Virtual Box config?19:20
akilawHow can I create a http server in gdp?19:21
Tartarusmartinkelly: ping19:21
mdpZenEntropy: might have broken since the 18th I haven't tried a recent build19:25
mdpZenEntropy: I'll update and take a look19:26
ZenEntropyOK thanks, let me know.19:26
martinkellyTartarus: pong19:26
Tartarusmartinkelly: hey, I'm Tom Rini.  Lets sort SPEC-784 out a bit :)19:27
martinkellysure :)19:27
TartarusCan you ls tmp/deploy/images/qemux86-6419:27
Tartarusthe ext4 should be there19:27
martinkellyit's not19:28
martinkellymartin@columbia:~/agl/build/qemux86-64$ ls tmp/deploy/images/qemux86-64/agl-demo-platform-qemux86-64-20170711192817.qemuboot.conf19:28
martinkellyagl-demo-platform-qemux86-64-20170711205856.qemuboot.conf19:28
martinkellyagl-demo-platform-qemux86-64-20170711230623.qemuboot.conf19:28
martinkellyagl-demo-platform-qemux86-64-20170725183733.qemuboot.conf19:28
martinkellyagl-demo-platform-qemux86-64-20170725183733.rootfs.ext4.xz19:28
martinkellyagl-demo-platform-qemux86-64-20170725183733.rootfs.manifest19:28
martinkellyagl-demo-platform-qemux86-64-20170725183733.rootfs.tar.xz19:28
martinkellyagl-demo-platform-qemux86-64.ext4.xz19:28
martinkellyagl-demo-platform-qemux86-64.manifest19:28
martinkellyagl-demo-platform-qemux86-64.qemuboot.conf19:28
martinkellyagl-demo-platform-qemux86-64.tar.xz19:28
martinkellybzImage19:28
TartarusHmm, interesting, OK19:28
Tartarusmaster or other branch?19:28
martinkellymaster19:28
martinkellyso you're saying the intended behavior is to generate both?19:28
RzRmartinkelly, dont you use qmeu ?19:29
RzRqemu19:29
Tartarusmartinkelly: Well, what's confusing is that I don't think I turned off other types, but, hmm, sec19:29
martinkellyinteresting19:30
TartarusI think I see it19:31
Tartaruswaiting on bitbake19:32
TartarusWell hmm19:33
Tartarusmartinkelly: For the moment, pop open meta-agl/meta-agl-bsp/conf/include/agl_qemux86-64.inc and add ext4 to the list after vmdk.xz19:34
TartarusI can reproduce the problem, so I'll poke more19:34
*** fredw has joined #automotive20:07
martinkellykk20:07
Tartarusmartinkelly: OK, I don't know when it stopped working, but, we need to be explicit in needing ext4, so we'll make it be explicit20:08
*** akila_w has joined #automotive20:09
dl9pfmartinkelly:  just    xz -d -k  the ext4.xz  in tmp/deploy/images ?20:11
*** akilaw has quit IRC20:12
dl9pfmartinkelly: Tartarus: runqemu only likes uncompressed imagetypes20:16
Tartarusdl9pf: pong20:17
*** praneeth has quit IRC20:17
*** praneeth has joined #automotive20:18
martinkellydl9pf: I know, I think we should just build both types if there are no objections20:23
dl9pfI'll have to special-case CI ... too much space if we go with uncompressed images .20:23
martinkellyah, got it20:36
martinkellyfor my own company's CI, we upload the images to S3 and let jenkins manage the lifetime20:36
martinkellyi've also been building in /dev/shm with fairly beefy instances to save on EBS volumes.20:36
martinkellypotentially some tricks like that could help as well20:36
*** ZenEntropy has quit IRC20:44
*** fredw has quit IRC20:50
*** fredw has joined #automotive20:59
dl9pfmartinkelly: testdrive this: https://gerrit.automotivelinux.org/gerrit/1038721:11
*** dabukalam has quit IRC21:11
*** dabukalam has joined #automotive21:13
*** fredw has quit IRC21:53
*** AlisonChaiken has quit IRC22:03
*** ronan__ has quit IRC22:14
*** AlisonChaiken has joined #automotive22:19
*** ronan__ has joined #automotive22:37
*** AlisonChaiken has quit IRC23:03
*** AlisonChaiken has joined #automotive23:04
*** martinkelly has quit IRC23:14
*** mdurnev has joined #automotive23:50
*** deep-book-gk_ has joined #automotive23:58
« Monday, 2017-07-24 Index Wednesday, 2017-07-26 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!