IRC logs for #baserock for Wednesday, 2015-12-23

*** gtristan_ has joined #baserock02:23
*** gtristan has quit IRC02:47
*** gtristan_ has quit IRC02:48
*** gtristan_ has joined #baserock03:25
*** gtristan has joined #baserock03:25
*** gtristan_ has quit IRC03:27
*** gtristan_ has joined #baserock03:29
*** gtristan has quit IRC03:29
*** gtristan_ is now known as gtristan06:28
gtristanjjardon, btw, were you able to login to the system again ? and... did your GNOME system have internet ?08:24
gtristanfwiw, I have had a better experience with VirtualBox than with kvm08:24
gtristanI usually run with VirtualBox with 2 network adaptors and audio enabled, first one is NAT for internet, and second one is a host only adaptor which lets you ssh into it08:27
*** gtristan has quit IRC08:40
*** toscalix has joined #baserock09:05
*** paulw has joined #baserock09:25
*** gtristan has joined #baserock09:27
*** ssam2 has joined #baserock10:12
*** ChanServ sets mode: +v ssam210:12
* gtristan runs raspbian inside qemu10:18
gtristanafter downloading like 3 different kernel + image combinations that didnt work10:18
gtristanbut yay a command line :)10:18
*** tiagogomes_ has joined #baserock10:20
tiagogomes_Gerrit is down10:50
tiagogomes_Fact, it was up 30s ago10:51
tiagogomes_mm, is up again10:51
ssam2sorry Tiago that was me taking a backup10:51
tiagogomes_ah ok10:52
ssam2mason seems to be failing due to some new exciting bug:
ssam2one build succeeds, the other 5 fail, all with this error:11:09
ssam2ERROR: /srv/distbuild/tmp/staging: File exists11:09
ssam2so presumably something (Morph?) is creating a temporay directory with a non-unique name11:10
ssam2or maybe Morph just needs to check whether that dir exists before trying to create it11:10
tiagogomes_odd, If it is the 'staging' directory, the directory where the temporary directories will be created with unique names for each staging area, morph checks if that directory exist11:15
tiagogomes_At least of local builds11:15
ssam2also, seems minimal-system-x86_64 build is broken11:24
ssam2i guess depmod is not in mimimal-syste,11:24
ssam2that tmp/staging thing might be a race, seems it doesn't happen often11:26
ssam2I guess broke minimal-system11:30
ssam2maybe it's possible to make depmod available at buildtime11:30
tiagogomes_It is on foundation...11:33
ssam2yeah... which is not in minimal-system11:33
ssam2perhaps it should be moved, but another fix would be to wrap the depmod call in an `if which depmod`11:33
ssam2i'll try taht11:33
tiagogomes_the second one ?11:34
tiagogomes_yep easier than add kmod to minimal-system11:34
ssam2changing a system-integration command seems to cause a rebuilt of that chunk, which is silly11:37
ssam2I guess it's by design with the current way system-integration is implemented11:37
tiagogomes_oh, right, system integration commands are included in the cache key computation11:42
persiaRight, be cause the script is embedded in the artifact.11:42
tiagogomes_Yes, but the script could still be embedded in the artifact and not be considered for the cache key computation11:49
tiagogomes_I mean, the script could still be embedded in the artifact, but the system integration commands not be considered for the cache key computation11:52
ssam2that wouldn't work because two artifacts with different system-integration commands might have the same cachekey11:54
ssam2they could be stored separately though, we already support multi-file artifacts in a kind of rubbish way11:54
ssam2although that still wouldn't solve the problem :-) I guess to make this fast, build tools need to be able to look for an artifact with all cache id items the same /except/ for one thing that they want to change..;.. which is not simple11:55
tiagogomes_ssam2 if we got rid of the system artifact, it would be fine having two artifacts with different system-integration commands might have the same cachekey11:57
ssam2unless you tried to send both of them to the same cache11:58
jjardongtristan: yeah, I could login but sometimes doesn't work, as you said. And yeah I have internet (using qemu here)12:12
jjardonssam2: ah, sorry about that; other solution is to build those drivers inside the kernel instead as a module12:16
ssam2seems annoying to prevent everyone from using modules, though12:22
ssam2using `if` seems to have worked12:22
ssam2 is the patch12:23
*** qiu has joined #baserock15:25
qiuAnyone who can give a help about baserock of GENIVI15:26
persiaqiu: Lots of folk, although at this time of year, "lots" is smaller than usual.15:27
persiaJust ask questions, and if anyone knows, they will answer.15:27
qiuI am GENIVI IVI amateur in china, Now I meet a problem when building Baserock hardware, could you please help to solve it?15:29
qiuFirst, I have bought the develop board NVIDIA Jetson TK1, and going to set up a Jetson TK1 as a Baserock ARMv7 development machine base on the guide:15:29
qiuAnd when I executed the command :15:29
qiubash baserock-current-build-system-armv7lhf-jetson.img jetson-tk115:29
qiuFail log print: IOError: [Errno 2] No such file or directory: '/root/nvidia-tools/_out_tools/baserock-flashing-tools/tmp/jetson-workdir/jetson-tk1/u-boot-spl-entry'15:29
qiuSo, could you please help to find out why it happens and how to solve it?15:29
qiuroot@ubuntu:~/nvidia-tools/_out_tools/baserock-flashing-tools# bash baserock-15.34-build-system-armv7lhf-jetson.img jetson-tk115:29
qiuPlease put your Jetson into recovery mode now15:29
qiuHit enter when this is done, and the board is attached to the flashing host15:29
qiuflash u-boot15:29
qiuNew BCT file PM375_Hynix_2GB_H5TC4G63AFR_RDA_924MHz.bct has been successfully generated!15:29
qiuImage file jetson-tk1-emmc.img has been successfully generated!15:29
persiaThat looks promising.  What happens when you boot the Jetson?15:31
qiuWhen I bought the Jetson board , it can boot normally. And what I do is flash Genivi baserock image to board15:34
qiuI do this base on the step:
persiaAnd you've just completed step 2?15:37
qiuno, stop when excute "bash baserock-current-build-system-armv7lhf-jetson.img jetson-tk1"15:38
qiuError happen -> IOError: [Errno 2] No such file or directory: '/root/nvidia-tools/_out_tools/baserock-flashing-tools/tmp/jetson-workdir/jetson-tk1/u-boot-spl-entry'15:39
persiaI thought that was the last command under "2) Flashing the Jetson": am I seeing a different page?15:39
qiuSorry for that I misunderstood you. I fail in the last command of "2)".15:43
qiuAnd the command above it can be excuted normally.15:44
persiaSo, when you ran the last command, the output was what you pasted above?15:44
persiaTo me, that looks like it worked.15:44
qiuyes, that is15:45
jmacsqiu: Are you logged in as root, or another user?15:45
persiaWhat happens when you disconnect the Jetson from the flashing host, and boot it?15:45
qiuI logged in root all the time15:45
jmacsHmm, OK15:46
qiuMaybe it is nothing to do with the jetson, because when I disconnect the board, has the same problem15:48
persiaMissing  '/root/nvidia-tools/_out_tools/baserock-flashing-tools/tmp/jetson-workdir/jetson-tk1/u-boot-spl-entry' ?15:51
qiuThe folder "tmp" is created automatically when excute the last command. And I do not konw why  '/root/nvidia-tools/_out_tools/baserock-flashing-tools/tmp/jetson-workdir/jetson-tk1/u-boot-spl-entry' is missing15:56
ratmice______hmm, that commit is rather recent, I wonder if the wiki instructions aren't lagging behind it15:57
ratmice______that said i'm not familiar with it so not any real help in solving it :(15:59
persiaIndeed: that does look like the culprit.  Any ideas as to what to put there to make it work?15:59
persiaOh well.16:00
jmacsI don't think this is the problem qiu is seeing, but that curl command doesn't work.16:00
jmacs gives a 301 moved permanently and curl doesn't follow it.16:01
persiaHrm?  That commit adds "u-boot-spl-entry" as a new file.16:01
qiuSo, what I need to do now to setup the jetson?16:02
jmacsqiu, do you have a file called baserock-current-build-system-armv7lhf-jetson.img.gz ?16:02
jmacsOK, it's not the curl problem then.16:03
qiuThat file is which I download to flash to jetson16:04
ratmice______so previously in the instructions it'd gone build-tools, but that creation of the file happens in 'build' script16:07
jmacscurl command is fixed on the wiki, anyway.16:09
ratmice______so, not sure if build-tools should run build, or if that needs to be done manually as 'cd tegra-uboot-flasher-scripts' './build-tools' './build', or what16:12
qiudears, I think this issue is easy to reproduce, just operate step by step with "", it will reproduce 100%.  By the way , I use VM ubuntu 32-bit 14.0416:16
qiuIf I download older verdion image to flash, can it avoid this problem?16:29
benbrown_qiu: Could you run the command with '-x' passed to bash and put it a pastebin?16:30
benbrown_Don't know much about flashing myself, but it'd be useful to know where the script is exiting.16:30
qiuroot@ubuntu:~/nvidia-tools/baserock-flashing-tools# bash -x baserock-current-build-system-armv7lhf-jetson.img jetson-tk116:34
qiu+ set -e16:34
qiu+ mkdir -p tmp/16:34
qiu+ baserock_image=baserock-current-build-system-armv7lhf-jetson.img16:34
qiu+ device_type=jetson-tk116:34
qiu+ trap cleanup EXIT16:34
qiu+ '[' baserock-current-build-system-armv7lhf-jetson.img = '' ']'16:34
*** qiu has quit IRC16:34
*** qiu has joined #baserock16:34
benbrown_qiu: Paste the log in
benbrown_To avoid the above :)16:35
jmacsI get the same error, fwiw16:35
jmacsUnfortunately the person who wrote those instructions is ill today16:36
*** gtristan has quit IRC16:37
qiuSorry for my action, I will paste log to other established website16:38
*** gtristan has joined #baserock16:39
qiuIt is at deep night in china, and I will go back to follow up this issue torrow or later on. Ohh, forget to say, Merry christmas!! (●'◡'●)16:51
benbrown_qiu: ratmice______ seems to be on the right track, cd into ~/nvidia-tools/tegra-uboot-flasher-scripts and run `./build build`16:51
benbrown_qiu: Very well, Merry Christmas :)16:53
benbrown_That doesn't succeed anyway, I'll leave this for someone who knows what they're doing.16:55
qiuEnm, OK ! Good night dear~16:57
*** qiu has quit IRC16:57
*** ssam2 has quit IRC17:07
*** gtristan has quit IRC17:09
*** toscalix has quit IRC17:35
*** tiagogomes_ has quit IRC17:58
pedroalvarezWell, if that's the problem we could workaround it by cloning an specific version of those scripts..22:50

Generated by 2.14.0 by Marius Gedminas - find it at!