| *** bharmon has quit IRC | 00:04 | |
| *** bbranch has quit IRC | 00:36 | |
| *** pespin has joined #automotive | 07:14 | |
| *** gunnarx has joined #automotive | 07:22 | |
| gunnarx | Reading the logs from yesterday... | 07:23 |
|---|---|---|
| gunnarx | CTtpollard: I think I've had that audiomanagement error, but otoh I've had probably every error available. So I can't remember the details now. | 07:25 |
| gunnarx | Yes, I have some vague recollection of having this in combination with CommonAPI changes. | 07:26 |
| gunnarx | Try comparing to johnnytk/meta-ivi which my project uses and see if you did the same changes. What I think we did was to cherry pick the common API fix from a later meta-ivi version (8.0 or so) | 07:29 |
| *** jacobo has joined #automotive | 07:30 | |
| gunnarx | weston: #qt channel can help you better with QtCreator integration, Often many helpful and intelligent people there. But I suppose the consensus was that Qt C++ is C++ so there is much information on static analysis tools. | 07:31 |
| *** joone has quit IRC | 07:41 | |
| *** kratz00_work has quit IRC | 07:41 | |
| *** ctgriffiths has quit IRC | 07:41 | |
| *** joone has joined #automotive | 07:42 | |
| *** kratz00_work has joined #automotive | 07:42 | |
| *** ctgriffiths has joined #automotive | 07:42 | |
| CTtpollard | thanks gunnarx I'll take a look | 07:43 |
| *** joone1 has joined #automotive | 07:45 | |
| *** kratz00_1ork has joined #automotive | 07:46 | |
| *** joone has quit IRC | 07:46 | |
| *** kratz00_work has quit IRC | 07:46 | |
| *** ctgriffiths has quit IRC | 07:46 | |
| *** jonathanmaw has joined #automotive | 07:47 | |
| *** ctgriffiths has joined #automotive | 07:48 | |
| paulsherwood | does anyone have a strong view about whether release tags should have a description or not? | 07:57 |
| *** CTtpollard has quit IRC | 07:58 | |
| rjek | paulsherwood: On one hand, a tag should be self-describing. On the other, more meta data the better | 08:02 |
| gunnarx | paulsherwood: Depends on the project ? Is it a general or specific question? | 08:03 |
| *** CTtpollard has joined #automotive | 08:04 | |
| gunnarx | Maybe a tangent, but a couple of times we have touched on (Pavel strongly advocating) that individual components should not refer to, for example, GENIVI platform releases, only the opposite direction. | 08:07 |
| gunnarx | So if projects start putting info in the meta data "this is compatible with GENIVI platform version X", is that good or bad? | 08:07 |
| gunnarx | It violate's Pavel's Law(tm) :) Which I can see the merit of also tbh. | 08:08 |
| gunnarx | So what I mean is, metadata is good but it depends on what info you put there most likely. | 08:08 |
| weston | gunnarx: I got the info. thanks | 08:09 |
| gunnarx | s/violate's/violates/ | 08:09 |
| rjek | gunnarx: I suspect it depends on the complexity of a component; it may have many dependancies for which it is simpler to say "compatible with all the things in release X" | 08:13 |
| CTtpollard | meta_ivi contains audiomanager_6.1.bb | 08:36 |
| CTtpollard | on commit 764abc4, as the fork you used it gunnarx | 08:36 |
| gunnarx | OK. I'm not really sure on what is behind that. I think we just "made it work"... | 08:48 |
| gunnarx | I know the AM stuff is a big mess - there is a giant patch updating the generated-src to a newer CommonAPI version... | 08:49 |
| gunnarx | But the first step I guess is "make it work", then we need to tackle that. Most likely by updating the baseline to use the latest AM (which uses the latest CommonAPI 3.0) and then updating GDP to the latest baseline. | 08:50 |
| CTtpollard | yup | 09:05 |
| CTtpollard | it's building for my by manually fixing the common api uri | 09:05 |
| CTtpollard | so there's something else that's coming with that commit which is breaking things | 09:05 |
| *** wschaller_ has joined #automotive | 09:05 | |
| *** wschaller_ has quit IRC | 09:12 | |
| *** gmacario has joined #automotive | 09:16 | |
| *** mdunford has joined #automotive | 09:21 | |
| *** wschaller_ has joined #automotive | 09:21 | |
| *** brlogger has joined #automotive | 09:24 | |
| *** pedroalvarez has joined #automotive | 09:25 | |
| paulsherwood | gunnarx: i'm about to tag an agl repo | 10:07 |
| paulsherwood | (unless i can get jonathanmaw to volunteer instead) | 10:08 |
| *** fredcadete has quit IRC | 10:13 | |
| *** jeremiah has joined #automotive | 10:13 | |
| *** pespin has quit IRC | 10:13 | |
| jeremiah | ohai! | 10:13 |
| CTtpollard | hi jeremiah | 10:15 |
| rjek | :) | 10:19 |
| jeremiah | 0/~ | 10:22 |
| jeremiah | I've got the mad IRC Sk!llZ | 10:22 |
| CTtpollard | jeremiah: are you feeling ok? | 10:23 |
| * jeremiah ETOOMUCHCAFFEINE | 10:23 | |
| gunnarx | oh save us... | 10:25 |
| rjek | ETOOLITTLECAFFEINE | 10:26 |
| * rjek keyboardfaces | 10:26 | |
| *** gunnarx has quit IRC | 10:41 | |
| *** gunnarx has joined #automotive | 10:54 | |
| *** mdunford has quit IRC | 10:56 | |
| *** mdunford has joined #automotive | 11:10 | |
| *** wschaller_ has quit IRC | 11:23 | |
| paulsherwood | jeremiah: i think i did git tag it? | 11:23 |
| paulsherwood | git tag 0.2015.33 -m "Initial 'release' known to provide a working image on Porter board" | 11:23 |
| paulsherwood | is that what you meant? | 11:23 |
| CTtpollard | ok, so it looks like the fork gunnar has for meta-ivi has a recipe for audiomanager_6.1, but using the same upstream commit from yocto uses 6.2. So hence it can't append to 6.1 during build time | 11:26 |
| * pedroalvarez has moved Baserock to audiomanager 7.0 | 11:28 | |
| CTtpollard | so I can't update the readme for GDP, as the patch wil fix the common api bug but will then break audiomanager | 11:30 |
| CTtpollard | gunnarx ^ | 11:32 |
| *** wschaller_ has joined #automotive | 11:37 | |
| jeremiah | paulsherwood: yep, that is what I meant. | 11:38 |
| jeremiah | pedroalvarez: Nice. :) | 11:38 |
| jeremiah | I'm going to be working in both wikis and issue trackers now, unless somebody warns me otherwise. | 11:39 |
| jeremiah | It is suboptimal, but we have to begin to use the tools at some point and I'm going to be a dog and eat the dogfood | 11:39 |
| jeremiah | By "both" I mean http://wiki.genivi.projects.org *and* https://genivi-oss.atlassian.net/wiki/display/ds/Projects+Wiki | 11:40 |
| jeremiah | And then Bugzilla and JIRA | 11:40 |
| jeremiah | I note that my assumption is that the URL for the JIRA and confluence will change but not the content. | 11:40 |
| jeremiah | I'm going to follow up with Joel regarding that assumption. | 11:41 |
| *** emaj has joined #automotive | 12:18 | |
| gunnarx | jeremiah: Well that's the assumption that has been communicated on PMO and TT. https://irclogs.baserock.org/automotive/%23automotive.2015-08-10.log.html#t2015-08-10T13:05:54 | 12:33 |
| gunnarx | I say communicated, I am not standing behind any guarantee of this, to be clear. | 12:33 |
| jeremiah | :) | 12:34 |
| jeremiah | I'm working with the CommonAPI guys to get the CAPI 3 bug tracking set up for SOMEIP. I'm going to create entries in both JIRA and Bugzilla but I plan in the future to only put new requests in JIRA. | 12:36 |
| jeremiah | I'll mention this in the various internal meetings to make sure everyone thinks its okay. | 12:36 |
| * CTtpollard thinks GDP needs to be patched to audio manager 6.2 | 12:39 | |
| jeremiah | http://gstreamer.freedesktop.org/conference/2015/index.html << Gstreamer conference | 12:42 |
| *** gunnarx has quit IRC | 12:43 | |
| * rjek wonders if Pexip will be there demonstrating their ludicrous pipelines | 12:45 | |
| jeremiah | heh | 12:45 |
| *** gunnarx has joined #automotive | 13:03 | |
| *** bbranch has joined #automotive | 13:15 | |
| *** gunnarx has quit IRC | 13:49 | |
| *** gunnarx has joined #automotive | 13:50 | |
| *** mythi_ has quit IRC | 14:06 | |
| jeremiah | Anyone who has a Porter board and has tried to get GDP running on it, a question for you: Are there *two* zip files for the GFX drivers? | 14:16 |
| jeremiah | I only see one. | 14:16 |
| jeremiah | Would be nice to have an MD5 sum as well just to know I have the complete zip file. | 14:16 |
| *** bbranch has quit IRC | 14:18 | |
| paulsherwood | CTtpollard: ^^ | 14:18 |
| CTtpollard | jeremiah: hi | 14:20 |
| CTtpollard | direct from Genivi there is r-car_series_evaluation_software_package_for_linux-20150515 & r-car_series_evaluation_software_package_of_linux_drivers-20150515 | 14:21 |
| CTtpollard | md5's in the same order | 14:23 |
| CTtpollard | bcca2517afdd2c3bd8bf4c2cb0a64d06 | 14:23 |
| CTtpollard | bcca2517afdd2c3bd8bf4c2cb0a64d06 | 14:24 |
| CTtpollard | *e1e83ce1882c48faedcc02adc9853e17 | 14:24 |
| *** gunnarx has quit IRC | 14:28 | |
| *** gunnarx has joined #automotive | 14:29 | |
| *** emaj has quit IRC | 14:29 | |
| jeremiah | CTtpollard: Thanks! | 14:34 |
| jeremiah | Where does one pull those from? | 14:34 |
| wschaller_ | jeremiah, they're both on http://www.renesas.com/secret/r_car_download/rcar_demoboard.jsp - the second is hidden under "To download related Linux drivers, please click here." | 14:35 |
| jeremiah | aha!! | 14:35 |
| jeremiah | wschaller_: Thank you, that helps. I clicked on the big orange button and just got graphics. :-) | 14:36 |
| wschaller_ | np, it confused me too :) | 14:36 |
| * wschaller_ clarifies the wiki instructions | 14:36 | |
| jeremiah | wschaller_++ | 14:37 |
| jeremiah | May I request md5 sums for sanity? | 14:38 |
| jeremiah | e1e83ce1882c48faedcc02adc9853e17 r-car_series_evaluation_software_package_of_linux_drivers-20150515.zip | 14:38 |
| CTtpollard | :) | 14:38 |
| wschaller_ | good idea | 14:38 |
| jeremiah | bcca2517afdd2c3bd8bf4c2cb0a64d06 r-car_series_evaluation_software_package_for_linux-20150515.zip | 14:38 |
| jeremiah | Those are mine. | 14:38 |
| CTtpollard | they are correct | 14:38 |
| jeremiah | w00t! | 14:39 |
| * jeremiah does a dance of immodest proportions | 14:39 | |
| CTtpollard | I'm not totally set putting them on the wiki, if renesas releases new ones | 14:39 |
| CTtpollard | the checksums | 14:39 |
| jeremiah | Ah. | 14:41 |
| CTtpollard | the scripts for instance only parse the name of the tar up till the date tag prefix | 14:41 |
| CTtpollard | and then wildcard the rest | 14:41 |
| jeremiah | Yes, I read them, they're of limited utility shall we say | 14:41 |
| CTtpollard | *suffix I should say | 14:41 |
| jeremiah | I'd prefer just instructions of which file to move where. | 14:41 |
| CTtpollard | jeremiah: those instructions are on the wiki now | 14:42 |
| jeremiah | Rockin' | 14:42 |
| *** gunnarx has quit IRC | 14:45 | |
| *** gunnarx has joined #automotive | 14:45 | |
| *** gunnarx has quit IRC | 14:54 | |
| *** gunnarx has joined #automotive | 14:55 | |
| jeremiah | bitbaking Porter GDP | 15:09 |
| CTtpollard | good luck! | 15:09 |
| gunnarx | CTtpollard: I'm back, if you want to sync up on the layers | 15:10 |
| CTtpollard | hi gunnarx | 15:11 |
| CTtpollard | without working off a patched fork of meta-ivi, I don't think it will be possible to build gdp in its current form | 15:11 |
| gunnarx | jeremiah: I pushed a preliminary porter branch to my repo also today. I think it builds but have not tested extensively. | 15:11 |
| gunnarx | CTtpollard, yes I think so too. | 15:12 |
| CTtpollard | I also think that is the better option than for example adding another instruction for a user to manually patch their checkout | 15:13 |
| gunnarx | As I said we cherry-picked that one fix because it was from way in the meta-ivi future... a complete baseline upgrade required, which changes too many other things. | 15:13 |
| gunnarx | Yes, probably. Until the meta-ivi maintainer returns we have little choice. | 15:13 |
| CTtpollard | do you think we could potentially ask the maintainer to pull a patched branch? | 15:14 |
| gunnarx | So to recap, your failures was due to upgrading meta-ivi completely, i.e. many things changed, or? What is your build based on compared to what I stated above? | 15:14 |
| *** gmacario has quit IRC | 15:15 | |
| gunnarx | We can ask the maintainer anything :) But the question is what is the strategy. So GDP today is based on a hash that is from the 7.0 baseline IIRC | 15:15 |
| gunnarx | Since then development has continued, a 8.0 has been created, and some other half finished branches exist for the future also. | 15:15 |
| gunnarx | The 9.0 release is what is we are aiming for this fall, so when we speak of Kronos baseline, that's 9.0 | 15:17 |
| CTtpollard | Ok | 15:18 |
| gunnarx | There is an unreleased Yocto/meta-ivi branch which we don't know if it is finished for Kronos/9.0. As far as baserock you may have seen Stephen's request to produce a K-0.2 release for baserock | 15:18 |
| CTtpollard | yep the request has been received | 15:19 |
| *** jlrmagnus has joined #automotive | 15:19 | |
| gunnarx | When I say unreleased branch above I refer to meta-ivi: dev_prepare9.0 - you will see it among the remote branches | 15:19 |
| gunnarx | So yes, I'd say the first step is that we feel "stable" based on a temporary fork off of the hash that GDP is based on. Next step might be to check if we can step to meta-ivi 8.0, and so on. | 15:21 |
| gunnarx | Just suggesting, it's up to you what you want to try. | 15:21 |
| CTtpollard | gunnarx: if the ivi maintainer allows it then yes, branching of the commit that we're currrently on to seems to be to be the fastest fix | 15:22 |
| CTtpollard | and as it stands GDP can't build, so I think the main priority is providing a buildable system | 15:23 |
| gunnarx | We have two choices I suppose, wait 2(?) weeks until Ned is back, or fork temporarily (which is what we did). | 15:23 |
| CTtpollard | ah I didn't know he was away for the length of time | 15:24 |
| * gunnarx is checking for more solid info | 15:25 | |
| gunnarx | According to my notes he is back monday the week after next. So just over a week then | 15:27 |
| gunnarx | johnnytk/meta-ivi is not going anywhere. Alternatively, set up a mirror on GENIVI git. | 15:28 |
| gunnarx | It's unclear, WR have designated Ned as far as I know. But due process I guess would be to email the meta-ivi mailing list and request a prompt response. | 15:29 |
| gunnarx | So that someone could step in - I suppose Holger is still around... | 15:29 |
| gunnarx | If we can't get a quick response maybe fork temporarily and then back into the standard process in a few weeks. | 15:30 |
| CTtpollard | hmm! | 15:32 |
| gunnarx | I'm just throwing out ideas. It's flexible. What would you prefer? | 15:33 |
| CTtpollard | I think start by emaling ivi mailing list | 15:33 |
| gunnarx | agree | 15:33 |
| CTtpollard | and then, backup to mirroring johnnytk on genivi git | 15:33 |
| gunnarx | it's the right way - maybe we do in fact have active maintenance | 15:34 |
| gunnarx | OK, I guess we have consensus on where we are in terms of the layers? I've been looking at one other thing which is the difference in the weston instructions for minnowboard and koelsch. It's looks a little messy and I don't see any immediate reason why they need to be different. | 15:35 |
| gunnarx | Anyone know who the wiki page editor Kiendinh is? | 15:36 |
| gunnarx | jeremiah: Do you have the ability to identify mediawiki user Kiendinh? (email address for example) | 15:37 |
| gunnarx | I'll email Tom and ask too | 15:39 |
| gunnarx | I.e. Tom from Intel | 15:39 |
| CTtpollard | sorry gunnarx just in a meeting right now | 15:54 |
| gunnarx | np, I was in a call too :) Logging off now | 16:01 |
| CTtpollard | yep I'm off as well for the evening, this has been helpful though! | 16:02 |
| CTtpollard | bye | 16:02 |
| *** joone1 is now known as joone | 16:26 | |
| *** mdunford has quit IRC | 16:44 | |
| *** gunnarx has quit IRC | 16:51 | |
| *** bbranch has joined #automotive | 16:53 | |
| *** jonathanmaw has quit IRC | 17:01 | |
| *** jlrmagnus has quit IRC | 17:01 | |
| *** jacobo has quit IRC | 17:14 | |
| *** wschaller_ has quit IRC | 17:15 | |
| *** jlrmagnus has joined #automotive | 17:18 | |
| jlrmagnus | Belated "Morning" | 17:42 |
| *** mdunford has joined #automotive | 17:47 | |
| *** mdunford has quit IRC | 17:52 | |
| *** gmacario has joined #automotive | 17:55 | |
| *** gmacario has quit IRC | 19:06 | |
| jlrmagnus | Reading up on the jeep hack whitepaper. http://illmatics.com/Remote%20Car%20Hacking.pdf | 20:12 |
| jlrmagnus | Halfway through, shaking my head. | 20:13 |
| jlrmagnus | DBUS is fully exposed by the Harmon unit through port 6667 both on the WiFi hotspot and on the Sprint-assigned IP address. | 20:19 |
| jlrmagnus | DBUS gives pretty much free reign to the IVI system. | 20:19 |
| myself | I thought they kept it together remarkably long. There are enough jokes but they don't slip into profanity. That's somewhat unexpected for a defcon paper. ;) | 20:19 |
| myself | Was. Sprint blocked it as a stopgap until Harmon can get their act together. | 20:20 |
| jlrmagnus | Yes. I'm using the paper to validate if the CAN Firewall unit would be able to block or mitigate the attack. | 20:21 |
| jlrmagnus | Aaand they have a DBUS command called, conveniently enough "execute". | 20:25 |
| *** bbranch has quit IRC | 20:27 | |
| *** gmacario has joined #automotive | 20:49 | |
| *** gmacario has quit IRC | 21:08 | |
| *** jlrmagnus has left #automotive | 23:06 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!