Embedded Linux Now and the future with LTSI. Hisao Munakata, Renesas Tsugikazu SHIBATA, NEC 27. March 2014 CollaboraDon Summit

Transcription

1 Embedded Linux Now and the future with LTSI Hisao Munakata, Renesas Tsugikazu SHIBATA, NEC 27. March 2014 CollaboraDon Summit

2 Key ac:vi:es of LTSI Long Term Support Ini:a:ve p Provide a industry managed kernel and maintain Long term based on LTS p Provide a common place for embedded industry to share informadon: Mailing list, Workshop p Provide place to support upstream acdvity : Help industry developer

3 LTSI: Industry managed kernel LTSI defines common kernel every year and maintains for 2 years LTSI adopts community Long term kernel as a base. Therefore, bug/security fixes from upstream are applied. LTSI kernel = long term + addidonal patches AUer 2 years term finished, possible to take over maintenance for longer term

4 LTSI: Common place LTSI provides the place to share informadon and experience among industry Mailing list to share problem and discuss how to solve Open Workshop to share status Closed meedng for more deeper F2F discussion Share the informadon will reduce the development cost

5 LTSI: Support upstream ac:vity Developer in the embedded industry may wanted to know how to work with community LTSI provides help upstream acdvides Provide suggesdon, review and discussion for proposed patches to be merged Many of discussion under way Merging patches to upstream is also reduce the cost of each companies

6 LTSI history Established 2011 in Prague We have been released : 2011: : : 3.10 LTSI = LTS + Industry demanded patch More details about LTSI is discussed later

7 White paper: Economic value of LTSI Maintenance cost of single LTS version is about 3M$ workgroup/value-of-ltsi 7

8 At the product planning using Linux Kernel selecdon Latest and Greatest Most stable and Secure IniDal cost Roll up by myself Use pre- exisdng kernel Running cost Maintain everything by myself Receive bugfix/security fix from community Eco system Tools, share informadon among industries ü Upstream Kernel ü Soc Kernel ü LTS Kernel ü LTSI Kernel

9 Comparison among Kernels Kernel source Ini:al cost Running Cost Eco System Upstream Kernel Roll up myself, validate everything Should watch bug / security fix in upstream and back port them Yocto eco system can be used Latest and Greatest SoC kernel Depend on SoC Depend on SoC or validate everything Depend on SoC, but mostly no support Watch bug/security fix and back port them Depend on SoC LTS Kernel Stable and Secure Validate everything on latest LTS Be able to get LTS patches LTSI Kernel LTS+add on LTS +addidonal patches + Share informadon LTS + share informadon Yocto Eco system can be used Be able to ask LTSI

10 Case of Android 6 month release cycle to month All the patches merged into mainline and then becomes no dependency on kernel version ICS : 3.0, Nexus 4=Jelly bean: 3.4 Nexus5=Kitkat: 3.4 LTSI is keep sync with Google Android

11 Case of Google s move Open AutomoDve Alliance Established: 2014/2 With Audi, GM, Honda, Hyundai, NVIDIA Bring the best of Android into Automotive Try to enable better integration between Cars and Android devices By end of 2014

12 Android ware Case of Google s move Announced March 2014 LTSI will keep sync with Google android Information that moves with you Small, powerful devices, worn on the body. Useful information when you need it most. Intelligent answers to spoken questions. Tools to help reach fitness goals. Your key to a multiscreen world.

13 Automo:ve Grade Linux Established 2012, Toyota, Nissan, Jaguar Land rover and many of Tier one companies Announcement of reference plaeorm already there with specificadon and binary based on Tizen IVI Expected life Dme will be around 10 years Automotive Grade Linux (AGL) is an open collaboration between contributors from the Automotive Industry, the Communications Industry, Semiconductor Industry, Academia, Community and others, combining open source components into a core operating system software stack suitable for automotive applications.

14 Case of Social Infrastructure Solu:on Increasing requirements and possibility for Social infrastructural service Arena soludon for sports event Major Dme, distance../ provide personal infor/ deliver live view with high density resoludon Control system for social infrastructure Dam/Power plant/traffic control/train control AdopDon for more cridcal podon being considered Very long term lifedme: run over 10 years Real Dme- ness maybe important in some case Reliability is one of key aspects

15 Case of Internet of Things AllSeen alliance had been established December Devices can be connected transparent ways to enable seamless sharing informadon. Based on Wifi, Bluetooth, PLC or any other connecdon There will be nearly 26 Billion devices on IoT by 2020 (Gatner, 2013/12) Kernel should be very low footprint with secure and stable

16 Comparison of new opportunity Consumer Electronics IoT Social infrastructure Life Dme 2 years 2-5 years Over 10 years Reliability demand Stability / Security demand Normal Normal Hard Normal Normal Hard Real Dme some case Nice to have Needed in case Characteris Dcs Mobile device Specific role for devices Depend on Use case

17 Possibility of Embedded Linux adopdon is expanding with increasing use case of new business opportunity Stable and reliable Linux eco system is fundamental pordon now and the future

18 THANK YOU 18

19 How you can par:cipate LTSI Follow on Twimer Web: hmp://ltsi.linuxfoundadon.org Mailing list: hmps://lists.linuxfoundadon.org/mailman/lisdnfo/ltsi- dev Git tree : hmp://git.linuxfoundadon.org/?p=ltsi- ernel.git;a=summary 19

20 Why you should consider adopting LTS?. Embedded Linux: Now and the future (part 2) LTSI kernel development updates. Tsugikazu Shibata, Hisao Munakata Linux Foundation Consumer Electronics working group March 27th / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

21 Why you should consider adopting LTS?. Who am I? (Munakata) From embedded SoC provider company Renesas Linux Foundation CE 1 working Gr. Steering committee member, LF/CEWG Architecture Gr. co-chair One of LF/CEWG LTSI 2 project initial proposer At my company, I had been encouraging my team developers to send a patches upstream Also I have supported various CE customers who develop digital-tv, Blu-ray recorder and Smart-phone 1 CE = consumer electronics 2 LTSI =Long Term Support Initiative 2 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

22 Why you should consider adopting LTS?. LTSI kernel February 24, 2014 LTSI > (EOL) LTSI > (update) 3 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

23 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late? Why you should consider adopting LTS? 4 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

24 Why you should consider adopting LTS?. Upstream kernel release mechanism Messages from Greg KH Whoops, Is it too late?. You can find 1)latest released, 2)under development.(=mainline, next), and several stable kernels 5 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

25 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. The release record of 3.0 series upstream kernel version release date duration v v days v days v days v days v days v days v days v days v days v days v days v days v days. 6 / 22 Ṛelease happened Tsugikazu Shibata, Hisao regularly Munakata at Embedded around Linux: Now every and the future 70(part days 2)

26 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. However, not all kernels are maintained for longterm version maintenance status v3.0 was longterm (3.100) -> now moved to EOL v3.1 maintained till 3.1.9, then now EOL v3.2 longterm (3.2.55), kept maintained (by Debian) v3.3 maintained till 3.3.8, then now EOL v3.4 longterm (3.4.84), kept maintained v3.5 maintained till 3.5.7, then now EOL v3.6 maintained till , then now EOL v3.7 maintained till , then now EOL v3.8 maintained till , then now EOL v3.9 maintained till , then now EOL v3.10 longterm stable ( ), kept maintained v3.11 maintained till , then now EOL v3.12 stable release ( ) till 3.13 release 7 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

27 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. Stable release include MUST APPLY essential fixes version fixes v3.0 -> v ,907 v3.1 -> v v3.2 -> v ,558 v3.3 -> v v3.4 -> v ,909 v3.5 -> v v3.6 -> v v3.7 -> v v3.8 -> v v3.9 -> v v3.10 -> v ,425 v3.11 -> v v3.12 -> v ,378 8 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

28 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. Longterm stable (LTS) kernel release cadence. Target kernel selection rules.. Maintainer will choose one LTS version per year Maintain it for 2 years from its original release LTSI-3.0 is moved to EOL when 3.10 became new LTS Then, we have 2 LTS kernels versions like 3.4 and / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

29 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. LTS/LTSI maintainer, Greg says 3.0 moves to EOL 10 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

30 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. And, Greg also announced next longterm is / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

31 Why you should consider adopting LTS? kernel release mechanism Messages from Greg KH Whoops, Is it too late?. Upstream kernel 3.10 development (is done) Whoops, we can not submit our latest device support code to 3.0 kernel now! Yes, that is true, because item date kernel 3.10 merge window open kernel 3.10 merge window close kernel 3.10 release As upstream 3.10 patch merge window is already closed, there is no chance to add your code to upstream kernel. Thus a cutting-edge silicon release after development cycle can not be supported in longterm 3.10 kernel. This might be. problematic for embedded industry Linux adopter. 12 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

32 Why you should consider adopting LTS? LTSI 3.10 development result Governance 13 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

33 Why you should consider adopting LTS?. LTSI-3.10 development history LTSI 3.10 development result Governance item date kernel 3.10 merge window open kernel 3.10 merge window close kernel 3.10 release Announce of 2013 LTS kernel version LTSI-3.10 git tree open becomes LTS (=3.12 release) LTSI-3.10 merge window open patch collection period 75 days LTSI-3.10-rc1 (=merge window close) validation period 26 days LTSI-3.10 release / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

34 Why you should consider adopting LTS?. Major contributor for LTSI-3.10 LTSI 3.10 development result Governance Contributor Patch count Darren Hart (intel) 1,197 Simon Horman (for Renesas) 1,122 Yoshitake Kobayashi (Toshiba) 123 Patrik Jakobsson (for intel) 46 Mark Brown (linaro.org) 11 Greg Kroah-Hartman (Linuxfoundation) 11 Total 2, / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

35 Why you should consider adopting LTS?. Major achievement of LTSI 3.10 LTSI 3.10 development result Governance LTTng Power efficient workqueues Intel's BayTrail support Intel's Minnowboard support Renesas's R-Car H2/M2 series support backported from the latest mainline Xilinx Zynq board support 16 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

36 Why you should consider adopting LTS? LTSI 3.10 development result Governance. LTSI = community LTS(longterm) + industry extra 17 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

37 Why you should consider adopting LTS? LTSI 3.10 development result Governance. Yocto and LTSI project coordination is working now 18 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

38 Why you should consider adopting LTS? LTSI 3.10 development result Governance. Discipline of LTSI project management. Community LTS + industry demanded extra patches. Governed by LF/CEWG Focus on kernel code a, not aiming complete BSP Therefore, can be combined with existing platform b CPU architecture and platform neutral Comply with upstream rules c Industry friendly acceptance (flexible patch forms, etc) Help CE (and others) industry to utilize Linux. a device drivers are part of kernel, of course b Android, Yocto, Tizen, AGL, WebOS and others c e.g. signed-off-by process 19 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

39 Why you should consider adopting LTS? Resources 20 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

40 . Why you should consider adopting LTS? Resources.. Correct understanding of the longterm (LTS) cadence is important. LF/CEWG develops LTSI version on top of community longterm kernel. You can gain huge cost reduction if you can fully utilize LTS & LTSI scheme. 3.0 longterm maintenance cycle has been moved to EOL and3.10 is the next longterm support target. ALSO, LTSI-3.10 was released in February Now industries can share LTS-3.10 and LTSI-3.10 kernel. It is too early to determine next LTSI candidate version. However, if you can put device support code (or anything) to current upstream it makes easier for you to use next LTSI kernel. Let's continue upstreaming! 21 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)

41 . Resources Why you should consider adopting LTS? Resources project web = ltsi.linuxfoundation.org ML ML subscription = ML archives = ML patchwork = git(each patch) = download (tar ball) = twitter document archives = 22 / 22 Tsugikazu Shibata, Hisao Munakata Embedded Linux: Now and the future (part 2)