*** gvancuts has quit IRC | 00:00 | |
*** waltminer has joined #automotive | 00:15 | |
*** waltminer has quit IRC | 00:49 | |
*** waltminer has joined #automotive | 01:01 | |
*** rstreif has quit IRC | 01:11 | |
*** waltminer has quit IRC | 02:18 | |
*** riazantsev has quit IRC | 02:23 | |
*** jlrmagnus has left #automotive | 02:33 | |
*** jlrmagnus has joined #automotive | 03:13 | |
*** jlrmagnus has quit IRC | 04:10 | |
*** dthiriez has joined #automotive | 04:19 | |
*** Guest39_ has joined #automotive | 04:24 | |
*** dthiriez has quit IRC | 04:25 | |
*** Guest39_ has quit IRC | 04:27 | |
weston | Good morning al | 04:49 |
---|---|---|
weston | all | 04:50 |
* persia hopelessly clings to the remaining 8 hours of Tuesday, noticing that apparently Wednesday is well under way | 04:55 | |
*** gla has joined #automotive | 05:18 | |
*** felipealmeida has quit IRC | 05:32 | |
*** felipealmeida has joined #automotive | 05:40 | |
*** gla has quit IRC | 06:24 | |
*** gvancuts has joined #automotive | 06:32 | |
*** gvancuts has quit IRC | 08:02 | |
*** mdunford has joined #automotive | 08:04 | |
*** jonathanmaw has joined #automotive | 08:57 | |
paulsherwood | :) | 09:08 |
weston | paulsherwood: hello..hows going | 10:07 |
paulsherwood | weston: i'm typing emails, while watching builds, and getting my thoughts in order for some docs i need to write... so keeping busy :) | 10:56 |
paulsherwood | in other news my proposal for a panel session on 'Collaboration in Automotive Software' has been accepted... so now i need to track down which interested engineers are attending that conference | 10:57 |
paulsherwood | (ELCE) | 10:57 |
weston | paulsherwood: cool. thanks. if you dont mind , automotive software mainly includes AUTOSAR specific software or infotainment specific software? | 11:20 |
*** Egy has joined #automotive | 11:51 | |
*** Egy has left #automotive | 11:52 | |
*** Egy has joined #automotive | 11:54 | |
*** waltminer has joined #automotive | 12:16 | |
*** ZrZ is now known as RzR | 12:52 | |
*** waltminer has quit IRC | 13:08 | |
paulsherwood | in that context, all of it | 13:24 |
*** Egy has quit IRC | 13:26 | |
*** waltminer has joined #automotive | 13:32 | |
*** riazantsev has joined #automotive | 13:39 | |
*** riazantsev has quit IRC | 14:03 | |
*** jlrmagnus has joined #automotive | 14:13 | |
jlrmagnus | Good morning | 14:14 |
paulsherwood | hi! | 14:14 |
paulsherwood | jlrmagnus: i'm coming to bay area and portalnd late next week... is that anywhere near you? | 14:15 |
*** riazantsev has joined #automotive | 14:17 | |
*** waltminer has quit IRC | 14:21 | |
jlrmagnus | paulsherwood: I'm in Los Angeles right now, but will be in Portland from Thursday onward next week. | 14:30 |
jlrmagnus | JLR's office is in downtown Portland. | 14:31 |
jlrmagnus | Feel free to swing by if you want to. | 14:31 |
*** emaj has joined #automotive | 14:41 | |
paulsherwood | cool! i'll try to | 14:42 |
*** CTtpollard has joined #automotive | 14:45 | |
*** Egy has joined #automotive | 14:59 | |
paulsherwood | http://www.bbc.co.uk/news/technology-33622298 | 15:06 |
* rjek cries into his hands | 15:06 | |
jlrmagnus | So they used a DAB transmitter. That was a novel way of doing it. | 15:07 |
paulsherwood | i did say in GENIVI a couple of years ago, that we needed to get serious about security. quite a few folks literally heckled me, telling me i was exaggerating the problem | 15:08 |
jlrmagnus | I'm right now incorporating a first round of review into the RVI security document. | 15:08 |
jlrmagnus | https://github.com/magnusfeuer/rvi_core/blob/master/doc/rvi_security.md | 15:09 |
* rjek has decided he never wants to be a passenger in a vehicle where the car radio is connected to the CANbus. | 15:09 | |
rjek | I think our old Passat's only connection to anything even remotely important is a wire whose voltage varies depending on value of speedo so it can do volume adjustment. That's difficult to hack. | 15:10 |
jlrmagnus | rjek: What year is your Passat? | 15:10 |
paulsherwood | rjek: could you take a look at magnus' doc? i'm assuming he'd appreciate feedback | 15:10 |
jlrmagnus | Would be much appreciated. | 15:11 |
rjek | jlrmagnus: 2006 I think | 15:11 |
jlrmagnus | A new version will be out in a few hours. | 15:11 |
* paulsherwood does wonder why radio *needs* to connect to the CANbus at all | 15:11 | |
rjek | I suspect it is actually connected to the CAN because you can query the OBD-II for the radio's key code | 15:11 |
paulsherwood | so immediate security improvement... deprecate that dunction? | 15:11 |
jlrmagnus | rjek: OBD-II port will have ISO 9141 (IIRC) or CAN. If somebody can access that, they have full access. | 15:11 |
paulsherwood | function | 15:12 |
rjek | I would like to see no IVI ever connected to anything other than some speakers, tbh :) | 15:12 |
paulsherwood | lol | 15:13 |
jlrmagnus | rjek: CAN bus connectivity is needed for the IVI to control HVAC, and read vehicle data needed by apps. | 15:13 |
rjek | jlrmagnus: I want to not have that. | 15:13 |
rjek | We've managed without for a hundred years. | 15:14 |
jlrmagnus | True. However, that would be a very niche market. | 15:14 |
jlrmagnus | http://uk.caterhamcars.com/ May be the car for you. | 15:15 |
rjek | I am content with an array of tactile knobs to control the AC, and to entertain the passengers with an iPad. | 15:15 |
FelixH | jlrmagnus: for the OBD port you need a physical access, that another story that a full remote hack | 15:15 |
rjek | I absolutely do not want 50 million lines of code interacting with both wireless networks and anything even remotely connected to the brakes or powertrain | 15:16 |
jlrmagnus | FelixH: Yep. | 15:16 |
rjek | Give me a knob over a greasy touchscreen anyday :) | 15:16 |
jlrmagnus | rjek: Yes. There need to be a stronger component domain isolation inside the IVI. | 15:17 |
rjek | jlrmagnus: It's called an airgap firewall :) | 15:17 |
rjek | A V850 connected via a UART with a list of CAN messages that are permitted to be sent /is not good enough/ | 15:17 |
jlrmagnus | Yes. Hard to sell a jag without a screen, though,. | 15:17 |
FelixH | You can disconnect the IVI system but anyway you need something connected to network and can for telemetry, remote access, etc. | 15:18 |
rjek | Also, knobs have the advantage that they're safer: you can use them without looking at them, which is pretty important while you're in a 2 ton projectile | 15:18 |
rjek | FelixH: You need? | 15:19 |
FelixH | I don't see a good way to secure that except a full electronic component that limit the CAN messages permited | 15:19 |
rjek | You cut the wire between them | 15:19 |
FelixH | rjek: well that depends what you want for sure | 15:19 |
jlrmagnus | rjek: If you are to sell the car you are building, then you need the connection. | 15:19 |
jlrmagnus | We would sell very, very few XJs if they had the UI of a 1995 Honda | 15:20 |
jlrmagnus | FelixH: A Firewall. | 15:21 |
rjek | My main concern with the whole industry these days is the massive growth in complexity: cars are getting features nobody considered essential or even advantageous only 10 years ago, and it's all a quagmire of security horror | 15:21 |
FelixH | But for security you could keep a real wire for the brakes and a mechanical link for the steering wheel I guess (but I don't do mechanic) | 15:21 |
rjek | Yeah, a CAN firewall ins't enough | 15:21 |
rjek | I believe the Chrysler hilarity recently involves abusing a bug in the VIP's firmware such that they replace it with code that allows any message to be sent | 15:22 |
jlrmagnus | rjek: Complexity increase is in all products today. Aerospace, banking, process control, medical, etc. | 15:22 |
FelixH | jlrmagnus: yes but everything that is software could be breached a day or another | 15:22 |
jlrmagnus | Just like those other industries, automotive has to learn to manage the risk that follows increased complexity. | 15:22 |
jlrmagnus | We are not the first people to develop safety critical software systems. | 15:22 |
rjek | jlrmagnus: I don't see why we should endanger customers by adding millions of lines of complexity just because other industries are :) | 15:22 |
jlrmagnus | rjek: Because, sadly enough, that is what the customer wants. | 15:23 |
rjek | If nobody did it, the customer woulnd't want it :) | 15:23 |
jlrmagnus | True, but the customer would look at their iphone and say "I want to play my music through the car stereo", and the ball would be rolling. | 15:24 |
rjek | You can do that without connecting the car stereo to the CANbus | 15:24 |
FelixH | Customers already want weekly OTA updates | 15:24 |
jlrmagnus | Yes. But CAN bus connectivity will come very soon after that first car stereo feature. | 15:24 |
jlrmagnus | FelixH: Correct, because that is what they get on their phones. | 15:25 |
jlrmagnus | Which is what we are competing against here. | 15:25 |
rjek | FelixH: The reason they want weekly OTA updates is because the software is so huge, buggy, and important to keep up to date... | 15:25 |
rjek | Added Chrysler hilarity: the bug can't be fixed OTA; has to go into the shop for a firmware upgrade to the VIP | 15:25 |
FelixH | Yes but not every car company is Chrysler :] | 15:39 |
*** Dthiriez has joined #automotive | 15:48 | |
Dthiriez | Hello and good morning! | 15:53 |
jlrmagnus | Good morning David. | 15:53 |
Dthiriez | Hi Magnus | 15:53 |
radiofree | good morning! | 15:53 |
Dthiriez | Magnus, as mentioned yesterday, I wanted to go over the rvi_core release-0.4.0 build issue I'm encountering on my web server | 15:55 |
jlrmagnus | Ok. | 15:56 |
jlrmagnus | Pastebin the error codes | 15:56 |
Dthiriez | http://pastebin.com/f6JWWMQ5 | 15:56 |
jlrmagnus | Checking | 15:57 |
jlrmagnus | Hm | 15:57 |
jlrmagnus | One sec | 15:57 |
Dthiriez | thanks! I tried a couple of workarounds too, not sure if they're kosher though | 15:57 |
radiofree | do you have bluez installed? | 15:57 |
radiofree | devel libs | 15:58 |
jlrmagnus | Ir rather: libbluetooth-dev | 15:58 |
Dthiriez | yes, let me confirm versions | 15:58 |
jlrmagnus | That should have the necessary header files, although the #include <> bit seems to pass during compile. | 15:58 |
Dthiriez | should have libbluetooth-dev, that was the first thing I tried | 15:58 |
jlrmagnus | k | 15:58 |
jlrmagnus | One sec. | 15:58 |
jlrmagnus | Do you have the file /usr/include/bluetooth/hci.h? | 15:59 |
jlrmagnus | Hmm. | 16:00 |
jlrmagnus | Cmagnus@t520:~/work/jlr/rvi_core/doc$ dpkg -S /usr/include/bluetooth/hci.h | 16:00 |
jlrmagnus | libbluetooth-dev: /usr/include/bluetooth/hci.h | 16:00 |
jlrmagnus | So libbluetooth-dev should provide the file, unless it's been stripped by some EC2 procedure. | 16:00 |
jlrmagnus | brb | 16:06 |
Dthiriez | k, checking. | 16:06 |
Dthiriez | [ec2-user bluetooth]$ rpm -qf ./hci.h | 16:06 |
Dthiriez | bluez-libs-devel-4.66-1.el6.x86_64 | 16:07 |
*** Egy1 has joined #automotive | 16:15 | |
*** Egy has quit IRC | 16:15 | |
Dthiriez | the other change I attempted was to comment out the bt package from the rebar.config and remove any bt parameters in the .config file used to build my node (i.e. dlink_bt_rpc) | 16:15 |
*** Egy1 has quit IRC | 16:16 | |
*** Egy has joined #automotive | 16:16 | |
jlrmagnus | Dthiriez: You need to update rebar.config as well. | 16:25 |
jlrmagnus | Try that. | 16:25 |
Dthiriez | should I comment out the BT functionality since the node will only be using tcp? | 16:26 |
jlrmagnus | Yes. | 16:26 |
jlrmagnus | Dump the config files in pastebin if you want me to have a look | 16:26 |
Dthiriez | k, adding to pastebin. I had attempted commenting the bt functionality earlier, but noticed I missed commenting out dlink_bt_rpc in the rebar.config | 16:29 |
Dthiriez | rebar.config @ http://pastebin.com/VuyA3H3b | 16:30 |
*** rjekalso has quit IRC | 16:31 | |
*** jonathanmaw has quit IRC | 16:32 | |
*** rstreif has joined #automotive | 16:33 | |
jlrmagnus | Dthiriez: LGTM. Don't forget to comment out the corresponding apps from backend.config that you feed to rvi_setup_node.sh | 16:37 |
Dthiriez | k, just finished make clean and make. Updating backend.config now | 16:39 |
jlrmagnus | k | 16:43 |
Dthiriez | setup_rvi_node.sh failed, adding result to pastebin | 16:43 |
*** rstreif has quit IRC | 16:43 | |
Dthiriez | error on setup_rvi_node.sh, http://pastebin.com/XbiTYqCK | 16:48 |
Dthiriez | backend.config used for setup_rvi_node.sh, http://pastebin.com/U4unbhFw | 16:51 |
jlrmagnus | Checking | 16:51 |
jlrmagnus | Skip sudo when you are setting up a node. It messes up $PATH. | 16:52 |
jlrmagnus | You probably need to do a sudo rm -rf prod_rel rel/prod_rel before trying again without sudo. | 16:52 |
Dthiriez | k, trying that now | 16:58 |
*** riazantsev has quit IRC | 17:01 | |
Dthiriez | brb | 17:02 |
jlrmagnus | k | 17:02 |
*** rstreif has joined #automotive | 17:04 | |
*** mdunford has quit IRC | 17:05 | |
*** rstreif has quit IRC | 17:08 | |
*** rstreif has joined #automotive | 17:10 | |
*** riazantsev has joined #automotive | 17:14 | |
Dthiriez | hmm... the 2 bt items are commented out in backend.config, but still seeing {error,"Release \"rvi\" uses non existing application dlink_bt"} http://pastebin.com/HU6k4cyp | 17:14 |
jlrmagnus | Can you give me backend.config | 17:15 |
jlrmagnus | Ah. | 17:15 |
jlrmagnus | Yeah. I think I know what is wrong, but need to see backend.config | 17:15 |
Dthiriez | backend.config with updated bt functionality removed @ http://pastebin.com/RfnfuUqW | 17:19 |
Dthiriez | *i.e. backend.config with { bt, load} and { dlink_bt, load } commented out | 17:21 |
jlrmagnus | Ok. Still build failure? | 17:21 |
jlrmagnus | I assume you are building from release-0.4.0 branch? | 17:22 |
Dthiriez | yes, still seeing http://pastebin.com/HU6k4cyp and on * release-0.4.0 branch | 17:24 |
jlrmagnus | Hm. | 17:25 |
jlrmagnus | Let me check out the same branch and see if I can recreate | 17:26 |
Dthiriez | k, thanks! | 17:27 |
jlrmagnus | Can you post the setup_rvi_node.sh command you used? The pastebin page has been removed. | 17:29 |
Dthiriez | ./scripts/setup_rvi_node.sh -n prod_rel -c ./backend.config | 17:30 |
*** RzR has quit IRC | 17:30 | |
jlrmagnus | Recreated | 17:31 |
*** RzR has joined #automotive | 17:33 | |
Dthiriez | k, think it's something hardcoded in the sertup_rvi_node process? | 17:33 |
jlrmagnus | Nope. | 17:34 |
jlrmagnus | src/rvi.app.src | 17:34 |
jlrmagnus | Forgot that one. | 17:34 |
jlrmagnus | Comment out bt and dlink_bt | 17:34 |
jlrmagnus | make clean | 17:34 |
jlrmagnus | make | 17:34 |
jlrmagnus | setup | 17:34 |
jlrmagnus | That worked for me. | 17:35 |
Dthiriez | cool! thanks! Trying that now | 17:35 |
jlrmagnus | ./rel/prod_rel/bin/rvi console | 17:36 |
jlrmagnus | Worked for me | 17:36 |
jlrmagnus | Ulf has cleaned up the config, making it easier to add and remove components without all the hassle. | 17:36 |
jlrmagnus | We have yet to integrate his work into our repos, though,. | 17:36 |
*** Dthiriez_ has joined #automotive | 17:39 | |
*** Dthiriez has quit IRC | 17:39 | |
Dthiriez_ | back! My wifi hotspot walked off | 17:40 |
jlrmagnus | Wb. | 17:40 |
Dthiriez_ | going through the changes mentioned and retrying the setup_riv_node | 17:41 |
jlrmagnus | k | 17:41 |
Dthiriez_ | Awesome!! Node build worked fine. Thank you for your help, Magnus! | 17:49 |
Dthiriez_ | to clarify, with Ulf's updates, will the only spot to update be a single .config file? | 17:51 |
jlrmagnus | Hopefully, yes. | 17:51 |
Dthiriez_ | cool, I'll document the resolution steps in the automotive-eg-rvi@lists.linuxfoundation.org thread I started last week. Thank you! | 17:53 |
jlrmagnus | Thank you. | 17:53 |
jlrmagnus | I'll be out until about 13:30PST. | 17:54 |
*** Egy has quit IRC | 18:00 | |
*** Egy has joined #automotive | 18:00 | |
Dthiriez_ | np I should be set for today. I'll play around with the setup this evening and follow-up tomorrow AM in case I have any questions. Thank you again! | 18:02 |
*** RzR has quit IRC | 18:11 | |
*** RzR has joined #automotive | 18:11 | |
*** Egy has quit IRC | 18:48 | |
ulf`_ | huh? | 18:58 |
*** ulf`_ is now known as ulf` | 18:58 | |
*** Dthiriez_ has quit IRC | 19:07 | |
*** Egy has joined #automotive | 19:13 | |
*** emaj has quit IRC | 20:33 | |
*** JEEB has joined #automotive | 20:34 | |
jlrmagnus | ulf`, Ulf Wiger, not you. Sorry. | 20:51 |
jlrmagnus | rvi-test1 will be brought down for experiments. | 21:01 |
*** Dahfish has joined #automotive | 21:26 | |
*** Dahfish has left #automotive | 21:28 | |
*** rstreif has quit IRC | 22:08 | |
*** jlrmagnus has left #automotive | 22:11 | |
*** jlrmagnus has joined #automotive | 22:48 | |
ulf` | jlrmagnus: ok ;) | 23:01 |
*** Egy has quit IRC | 23:24 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!