Android Kernel Meeting

Agenda

  • Cover actions from last week (see below).
  • John merged oamp3 defconfig into linaro+android.
    • Open to merging panda defconfig if one is sent.
  • John to merge virtual battery driver, also likely to push it to lkml for review.
  • Other updates?

Past Action items

  • Noritsuna to check with Mr.Sola and others to see if there is a better time/date where more people can join.
  • Noritsuna to send John more details about what parts of ashmem he thinks need to be more generic.
  • John to email Mr.Sola about his power management interface work.
    • TODO: Didn't get to this last week.
  • Noritsuna to send John virtual battery driver patch.
    • Done: Mr. Sola has sent this out, and it will be merged into linaro+android soon.
  • John to review git tree Mr.An has posted.
    • Done: John merged some of the changes. John will mail some further questions on fb changes.
  • John to talk with Patrik about release schedule and code freezes
    • Done? : Noritsuna and Patrik have discussed over mail.

Action Items

  • John to mail Kenji WATANABE and ask about framebuffer changes for linaro+android kernel
  • Noritsuna to test linaro+android kernel with the merged androd_omap3_defconfig and provide feedback to John
  • John to contact jserv-- about "battery HACK" and its connection with virtual battery driver
  • Noritsuna to provide list of hardware/boards he is interested in using to help with bringup on Linaro's Android platform


Android Platform Team Meeting

Agenda

  • Actions items review
  • Status Reports
  • AOB

Past Action items

  • asac to follow up on linaro.android kernel with jstultz
  • kjnabe to contribute framebuffer support back to jstultz git.linaro.org tree
  • kjnabe to contribute beagle defconfig to jstultz git.linaro.org tree
  • jserv-- and jeremychang to confirm that battery support works well and then backout our battery HACK from git.linaro.org
  • patrikryd to fix the tarball generation and send patches for review
  • jeremychang to bring up device/linaro/pandaboard
  • patrikryd to explore with jserv-- and luse what options we have to ship toolchain
  • luse to expand toolchain wiki page on distribution format options of toolchain and make a new wiki page on toolchain builder (or wrap in the toolchain main page)
  • JamieBennett to setup weekly reports for the team

People Present

  • JamieBennett

  • asac
  • noritsuna
  • jserv--
  • jeremychang
  • patrikryd
  • luse

Actions Recorded


Engineers Reports

noritsuna

Highlights

  • "Hardware memory mapping guideline" (Old name is "pmem/ashmem common interface for kernel layer")

Plans

  • I reseach each SoC's kernel(TI, freesclae, Samsung, ST-m) this week.
  • I understood that ashmem & pmem have only API for Android framework(HAL) on this week.

  • And hardwares(ex. camera, each sensors) make a memory mapping by each implementation of SoC. Then I think to make guideline of how to make a memory mapping hardwares(ex. camera, each sensors).

Issues

sola

Highlights

  • power management interface for kernel layer.

Plans

  • I reseach each SoC's kernel(TI, freesclae, Samsung, ST-m) this week.
  • reseach board: http://dev.oesf.biz/download/sola_wakelock_20110222.odp

  • "virtual battery driver changed for linux kernel 2.6.37 or later"
  • I define common guideline next week.
  • But I have a lot of works next week. Maybe stop next work.
  • I will start this work 2 weeks after.

Issues

kj-nabe

Highlights

  • USB Composite driver API for userland layer.

Plans

  • I reseach each SoC's kernel(TI, freesclae, Samsung, ST-m) this week.
  • If it is possible to make it, I define common guideline a few week.
  • But I have a lot of works a few weeks. Maybe stop a few works.
  • Because settlement of accounts of a Japanese company is in March. Japanese engineers are busy this month & next month.

Issues

  • My company promised I join & work this project. but I can't get time for this work. Therefore Noritsuna will discuss the president of my company about my work on next week.

jstultz_*

Highlights

  • Refreshed the Linaro+Android kernel to include the latest from both the linaro and android 2.6.37 based trees.
  • After my meeting with Noritsuna, he got Akihiro MAEDA to send the virtual battery driver to the linaro dev list for review. I send some feedback on how to best submit the change, and the review cycle continues. Hopefully this is something that can go upstream in 2.6.39.
  • After being prodded to include a defconfig for android on beagle, I generated, tested and merged one in.

Plans

  • Need to sync with Patrik (re schedule for test Linaro Android images, linaro+android kernel integration update)
  • Merge virtual battery driver & push that into linaro+android tree.

Issues

  • With the Linaro tree moving to 2.6.38, and the Android 2.6.38 tree being available, considering updating to 2.6.38 for linaro+android tree.
    • Need input on how jarring that might be for Patrik and others.
    • May need a new branch for 2.6.38 work (possibly moving current dev/linaro.android branch to dev/linaro.android-2.6.37)

patrikryd

Highlights

Plans

Issues

  • -

luse

Highlights

Plans

  • Verify Linaro toolchain for Android
    • Allow linaro-gcc + Android related backports (from gcc-4.6) building Android
    • Verify generated Android images on real board: Beagleboard + Gingerbread ok

Issues

  • How to improve Linaro Android toolchain usability and documentation?

jserv--

Highlights

Plans

  • Propose a generic approach to mount MMC partitions instead of platform specific sysfs
  • Check Linaro image generation tool
  • Android toolchain, benchmark, verification
    • Document benchmark suite from Google Compiler Team and execute on real board.
    • Discuss with jeremy to improve verification tool

Issues

jeremychang

Highlights

Plans

  • Analyze existing 3rd open source Android enablement projects
  • Android automation and verification
  • Study Snowball if possible

Issues

  • Track Android related patches
  • Improve 0xbench and native parts

Your Name <irc nick>

Highlights

  • Short bullet points of work you've done that week which convey

progress and highlights which can be used to report on how the team is progressing as a whole.

Plans

  • Your individual plans for the coming week(s).

Issues

  • Your individual plans for the coming week(s).


IRC logs

Android Kernel Meeting

<noritsuna> Hey!
<jstultz_vm> noritsuna:  how are you doing?
<noritsuna> Sola said "I may be able to join today. but I don't guess morning schedule."
<noritsuna> I wrote our activities on https://wiki.linaro.org/Platform/Android/Meetings/2011-02-23#preview.
<noritsuna> "Engineers Reports"
<jstultz_vm> noritsuna:  great!
<jstultz_vm> noritsuna:  let me skim those and then we can start in on the agenda
<noritsuna> OK.
<jstultz_vm> noritsuna:  ok. that looks great, thanks for the explanation about the business accounting in march.
<jstultz_vm> so lets quickly review last weeks action items..
<jstultz_vm> you can find them in the "Past Action items" area on the 2011-02-23 meeting page
<jstultz_vm> Did you look into if more folks would be willing to join if the meeting was shifted by an hour?
<noritsuna> Yes. I heard Sola & kj-nabe it.
<jstultz_vm> noritsuna:  Did they say if 9am or 11am JST would be better?
<noritsuna> They said this time is OK.
<jstultz_vm> noritsuna:  ok. that's fine then. we'll leave it at 10am JST. 
<jstultz_vm> noritsuna:  so next up, I see from the report you're working on the hardware memory mapping guideline.
<jstultz_vm> noritsuna:  is that something you'll be able to share soon?
<noritsuna> But kj-nabe is busy this & next week. If Sola doesn't have other meeting, Sola will join this meeting.
<jstultz_vm> noritsuna:  ok.
<jstultz_vm> noritsuna:  so do you have a plan for when memory mapping guideline will be finished?
<noritsuna> Each board make Camera's pointer on memory. But this code is put in various places.
<noritsuna> Then I will make standard implimentaion for those code.
<noritsuna> Sorry, I haven't finished this work.
<jstultz_vm> noritsuna:  That is ok, i'm just curious as to how long you think it will be.
<noritsuna> I think 2-3 weeks.
<jstultz_vm> noritsuna:  cool!
<jstultz_vm> noritsuna:  so next up is my emailing Mr. Sola on the power management interface work. 
<jstultz_vm> noritsuna:  last week was busy, so i just managed to send that email to him today.
<jstultz_vm> noritsuna:  (you were cc'ed)
<noritsuna> I see.
<jstultz_vm> noritsuna:  also Mr. Sola has sent the virtual battery patches out, so that's great! I'll be working to merge them into out tree soon.
<jstultz_vm> noritsuna:  so thank you for asking him to do that.
<noritsuna> Sola thinks the first work is for wakelock function.
<jstultz_vm> noritsuna:  I just reviewed his slide deck in the reports. I'm not sure exactly what he plans, but hopefully the email will clarify that.
<jstultz_vm> noritsuna:  I also reviewed the git tree here: http://gitorious.org/kj-android/jstultz-linux-37-linaro-android
<jstultz_vm> noritsuna:  i merged two changes from there already. 
<jstultz_vm> noritsuna:  the framebuffer changes need a bit more explanation, so I'll email Kenji WATANABE to get those details this week
<jstultz_vm> noritsuna:  finally, I see you and Alexander talked about the linaro release schedule in email. 
<jstultz_vm> noritsuna:  Did you have any further questions about the release schedule?
<noritsuna> Yes. We will try to port Linaro11.05 beta on end of March.
<noritsuna> Next target is Linaro11.05 final release.
<jstultz_vm> noritsuna:  One question Alexander had was what do you mean by "porting" ? It is a little unclear? Are you moving your android applications to a new android image? Or do you mean trying to bring up the Linaro 11.05 release on new hardware?
<noritsuna> Sorry, This meaning is "merge". Lianro11.05 and Android-Common-Kernel.
<jstultz_vm> noritsuna:  by the Android-Common-Kernel do you mean the linaro+android kernel?
<jstultz_vm> Or something else?
<noritsuna> I think "the linaro+android kernel".
<jstultz_vm> noritsuna:  Ok. Hopefully the linaro+android kernel will be included in Linaro's Android platform release. Do you think it will be much work doing the merge?
<noritsuna> No. I think a little works.
<jstultz_vm> noritsuna:  ok.
<jstultz_vm> So next on the agenda: update on the linaro+android kernel.
<jstultz_vm> I've merged an omap3 defconfig at jeremychang's request. its not the same one you have, because I'm using a more limited defconfig.
<jstultz_vm> I'd appreciate it if you gave it a try and let me know if you saw any issues with it.
<jstultz_vm> It seems to be working ok in my tests.
<noritsuna> OK. I will try it!
<jstultz_vm> noritsuna:  thank you!
<jstultz_vm> noritsuna:  also jeremy was asking about a panda defconfig, and i'd be happy to merge a minimal version (similar to the android_omap3_defconfig i merged) if anyone is able to generate it.
<jstultz_vm> noritsuna:  i don't have a panda board on hand, so I cannot test defconfigs to see if they would be correct or not.
<noritsuna> me to... I don't have panda board.
<jstultz_vm> noritsuna:  ok. well, if you would let folks know i'm happy to merge one, it would be appreciated.
<jstultz_vm> finally, i'll be merging the virtual battery driver soon. the only concern i had was that i saw in the notes for the last team meeting: https://wiki.linaro.org/Platform/Android/Meetings/2011-02-16 
<jstultz_vm> "jserv-- and jeremychang to confirm that battery support works well and then backout our battery HACK from git.linaro.org "
<jstultz_vm> noritsuna:  do you know if the "battery HACK" is the virutal battery driver?
<jstultz_vm> noritsuna:  or are they talking about something else?
<jstultz_vm> (i don't want to merge the change if it is going to be quickly backed out)
<noritsuna> I don't know "battery HACK".
<jstultz_vm> noritsuna:  ok. I'll ask jserv-- directly then. thank you.
<jstultz_vm> noritsuna:  that's all i have on my agenda, is there anything else you wanted to discuss?
<noritsuna> Are our works no problem? If you want porting work(each board) to us, we can do it.
<jstultz_vm> noritsuna:  not sure if i understood that. Are you offering to help with bringup and testing of Linaro's Android platform on different boards?
<noritsuna> Yes. I think this work is hard & important. I want supporting your work.
<jstultz_vm> noritsuna:  ok. is there a list of boards you have that you want to help with?
<jstultz_vm> I'm not sure which boards the Linaro Android platform is targeting first (I expect beagle board xm, and likely panda, but I don't know what else)
<jstultz_vm> I am going to be trying to bring up on a imx51 soon myself.
<noritsuna> Because I thought porting work is started when Linaro-kernel is finished.
<jstultz_vm> noritsuna:  well, its hard to know if the kernel is "finished" if we haven't tested it on a system.
<noritsuna> OK. I understood it.
<jstultz_vm> noritsuna:  maybe you can look into which boards you have available to use, and let us know next week? then we can try to coordinate the bringup and testing work?
<jstultz_vm> noritsuna:  Ok. Is there anything else you want to discuss before the meeting is over?
<noritsuna> OK. I look it. And I let you it.
<noritsuna> OK. I don't have meeting's agenda.
<jstultz_vm> noritsuna:  Ok. Thanks so much for joining! I'll post the meeting minutes shortly.
<jstultz_vm> noritsuna:  Have a good rest of your week!

Android Team Meeting

Started logging meeting in #linaro-meeting
[10:06:16] <JamieBennett> [LINK] https://wiki.linaro.org/Platform/Android/Meetings/2011-02-23
[10:06:38] <JamieBennett> [TOPIC] Action Item Review
[10:06:45] <JamieBennett> asac to follow up on linaro.android kernel with jstultz
[10:07:21] <JamieBennett> asac: did this happen?
[10:07:23] <asac> i think i did that, but i need more context to remember :(
[10:07:59] <JamieBennett> OK, I'll let you dig the log out ;)
[10:08:00] <JamieBennett> kjnabe to contribute framebuffer support back to jstultz git.linaro.org tree
[10:08:18] <asac> noritsuna: did that happen?
[10:08:40] <JamieBennett> noritsuna: also did kjnabe to contribute framebuffer support back to jstultz git.linaro.org tree happen?
[10:09:55] <noritsuna> framebuffer?
[10:10:31] <asac> noritsuna: yes, thats an action you guys took ... is that not on your radar?
[10:11:00] <JamieBennett> meeting log from where these action came - http://mootbot.ubuntu-uk.org/linaro-meeting.log.20110209_1002.html
[10:11:56] <JamieBennett> [10:24:09] sorry, now git-pushing to git://gitorious.org/kj-android/jstultz-linux-37-linaro-android.git
[10:12:02] <JamieBennett> [10:24:28] kjnabe: whats in there?
[10:12:03] <JamieBennett> [10:24:48] kjnabe, maybe you could write summary for that tree.
[10:12:06] <JamieBennett> [10:24:56] is that something that boots on beagle and has framebuffer support?
[10:12:12] <JamieBennett> [10:25:16] yes
[10:12:12] <JamieBennett> [10:25:32] kjnabe: ok ... it has a defconfig?
[10:12:13] <JamieBennett> [10:25:51] i merged from sakoman's tree < framebuffer support
[10:12:13] <JamieBennett> [10:25:58] cool
[10:12:13] <JamieBennett> [10:26:14] kjnabe: can you contribute that back to jstultz?
[10:12:41] <asac> noritsuna: the goal is to have a beagle defconfig in jstultz tree
[10:12:57] <asac> that works with android (so it has framebuffer among other stuff)
[10:13:55] <asac> noritsuna: do we have a) a beagle defconfig in his tree now? and b) does framebuffer work from his tree if you use that defconfig
[10:14:04] <asac> ??
[10:14:48] <noritsuna> defconfig is http://gitorious.org/kj-android/jstultz-linux-37-linaro-android/blobs/linaro.android/arch/arm/configs/android_omap3_defconfig
[10:15:01] <asac> ok cool
[10:15:07] <asac> well not cool
[10:15:11] <asac> noritsuna: thats still not in jstultz tree
[10:15:26] <asac> noritsuna: the action was to get that stuff into his git.linaro.org tree :/
[10:15:40] <asac> so its not done? e.g. should we carry that action forward?
[10:16:34] <asac> JamieBennett: lets carry forward and send a mail on top to kjnabe, noritsuna and jstultz to ensure all are on the same page
[10:16:39] <JamieBennett> OK
[10:16:44] <JamieBennett> [ACTION] kjnabe to contribute framebuffer support back to jstultz git.linaro.org tree
[10:16:57] <JamieBennett> [ACTION] kjnabe to contribute beagle defconfig to jstultz git.linaro.org tree
[10:17:13] * jserv-- waits again. :(
[10:17:25] <noritsuna> I see. I tell it kjnabe
[10:17:29] <asac> thanks
[10:17:34] <JamieBennett> [ACTION] JamieBennett to prod people for framebuffer and defconfig changes to go to jstultz' tree
[10:17:37] <asac> noritsuna: jamie will also send a mail
[10:17:48] <JamieBennett> OK, moving on
[10:17:50] <JamieBennett> jserv-- and jeremychang to confirm that battery support works well and then backout our battery HACK from git.linaro.org
[10:18:03] <jserv--> here!
[10:18:07] <JamieBennett> :)
[10:18:07] <asac> yay!!
[10:18:18] <JamieBennett> so it happened jserv-- ?
[10:18:45] <jserv--> due to the fact jeremy and I could not boot Linaro kernel tree well.
[10:18:57] <jserv--> so, we can only try the patch on another kernel tree.
[10:19:11] <jserv--> it works on beagleboard out of box.
[10:19:30] <asac> you cannot boot with beagle using linaro kernel? i thought it was only a panda issue
[10:19:31] <jserv--> but I think we should try John's tree.
[10:19:35] <asac> ah
[10:19:38] <asac> yes, johns tree
[10:19:41] <jeremychang> Hi, sorry I may be late.
[10:19:43] <asac> is the tree we are talking about
[10:19:44] <jserv--> asac, I meant, John's tree.
[10:20:00] <jserv--> asac, yes
[10:20:11] <asac> jserv--: ok. for android johns tree is the linaro tree atm
[10:20:18] <jserv--> asac, but the patch is conceptually right.
[10:20:38] <asac> and you confirmed it to work with a different kernel?
[10:20:39] <asac> ok
[10:20:45] <JamieBennett> [ACTION] jserv-- and jeremychang to test John's tree for battery support and report back
[10:21:02] <jserv--> ok, got it.
[10:21:04] <asac> ... please backout our userland hacks once this patch landed in johns tree (should be in now or tomorrow i hopw)
[10:21:06] <JamieBennett> :)
[10:21:21] <JamieBennett> * patrikryd to fix the tarball generation and send patches for review
[10:21:29] <patrikryd> done
[10:21:34] <JamieBennett> \o/
[10:21:45] <JamieBennett> * jeremychang to bring up device/linaro/pandaboard
[10:22:09] <JamieBennett> jeremychang: did this happen?
[10:22:20] <patrikryd> They are on git.linaro.org and android-build.linaro.org has tarballs from tonigths build.
[10:22:46] <jeremychang> JamieBennett, Hi. I bringed up the pandaboard by using Pandroid's kernel.
[10:22:51] <asac> right. but does it mean pandaboard bring up is done as well?
[10:23:00] <asac> ok
[10:23:14] <asac> jeremychang: so you used our tarballs and used pandroid kernel and it worked=?
[10:23:35] <JamieBennett> as a side note, do we know why android-build.l.o has an untrusted certificate?
[10:23:41] <asac> nope
[10:23:46] <asac> guess because its a host in the cloud
[10:23:55] <asac> and we wanted quickly some ssl for it ... https://android-build.linaro.org/
[10:24:03] <jserv--> asac, only pandroid's kernel worked.
[10:24:15] <JamieBennett> [ACTION] JamieBennett to investigate certificates on android-build.linaro.org
[10:24:18] <asac> jeremychang: bringing up pandaboard means for me: if you take the tarballs from https://android-build.linaro.org/ they work
[10:24:29] <asac> jeremychang: its ok to use pandroid kernel for now
[10:24:44] <asac> but in theory jstultz kernel should also work (but dont bother with that for now)
[10:25:12] <asac> jeremychang: can you confirmt hat using https://android-build.linaro.org/job/android-panda/ tarballs works?
[10:25:29] <asac> please do so and fix things that are required to get it to boot ;)
[10:25:30] <asac> thanks
[10:26:20] <JamieBennett> [ACTION] jeremychang to confirm https://android-build.linaro.org/job/android-panda/ tarballs work and fix if necessary
[10:26:28] <asac> ok i think we need an action to track that someone confirmed that panda works
[10:26:32] <jeremychang> asac, ok.
[10:26:34] <asac> same for beagle and same for ux500
[10:26:50] <asac> patrikryd: can you take ux500?
[10:26:53] <jeremychang> asac, I will confirm that.
[10:26:59] <patrikryd> ok
[10:27:01] <asac> patrikryd: goal: our tarballs works with whatever good kernel you can find
[10:27:07] <asac> internally
[10:27:10] <jeremychang> with linaro android. I just sent the patch for device/linaro/pandaboard.
[10:27:17] <patrikryd> intenally?
[10:27:17] <jeremychang> The modification is only that setting ARCH_ARM_HAVE_TLS_REGISTER to true.
[10:27:31] <jeremychang> Sorry I just typed in the wrong channel.
[10:27:34] <asac> patrikryd: well. i assume that we dont have an android kernel for ux500 in the public yet
[10:27:57] <asac> i meant: i dont care about kernel this week ... just that the userland stuff we produce is fine ;)(
[10:28:16] <patrikryd> True. There is a Linux kernel, but not an Android kernel.
[10:28:19] <asac> jeremychang: ok thanks ... after this lands (patrik will commit soon i guess), please verify
[10:28:37] <jeremychang> asac, Thanks, I will follow that.
[10:28:37] <asac> jeremychang: use the tarballs from the website above the next day for that
[10:28:54] <patrikryd> The LT has no intention at the moment to give me one. Have to "create" on myself to start with and ...
[10:28:57] <jserv--> asac, however, for userspace developers like jeremychang , it is really hard to distinguish problems from kernels.
[10:29:13] <jserv--> asac, since we always have to try other "upstream" kernel.
[10:29:19] <asac> patrikryd: right. thats why i said "use whatewver you can find internally" to verify that the ux500 userland is working out of box as it is
[10:29:33] <patrikryd> Ok
[10:29:48] <asac> jserv--: thats fine. as i said. use whatever best kernel you can find. pandroid is currently the one i guess
[10:29:54] <patrikryd> I can get the latest internal android kernel and verify.
[10:30:01] <JamieBennett> [ACTION] patrikryd to confirm https://android-build.linaro.org/job/android-u8500href/ tarballs work and fix if necessary
[10:30:10] <JamieBennett> who gets beagle ? ;)
[10:30:10] <asac> once you have that booting comparing that boot to a boot with a linaro jstultz tree would give you extra credits ;)
[10:30:21] <jserv--> asac, ok
[10:30:37] <asac> i guess jserv-- is last man standing ... so i assume he will take action that beagleboard as it is works ;)
[10:30:46] <asac> same approach: use whatever best kernel you can find
[10:30:47] <jserv--> It just takes time to try again and again. :(
[10:31:14] <JamieBennett> [ACTION[ jserv-- to confirm https://android-build.linaro.org/job/android-beagle/ tarballs work and fix if necessary
[10:31:17] <asac> jserv--: whast the problem?
[10:31:38] <JamieBennett> [ACTION] member:jserv-- to confirm https://android-build.linaro.org/job/android-beagle/ tarballs work and fix if necessary
[10:31:43] <JamieBennett> doh
[10:31:44] <jserv--> asac, beagle is ok, I meant some board like panda.
[10:31:44] <asac> jserv--: you say testing takes time? what is the main time sink there?
[10:31:55] <jserv--> asac, I never boot it with linaro tree.
[10:32:11] <asac> jserv--: what does linaro tree mean?
[10:32:14] <asac> kernel or userland?
[10:32:19] <jserv--> asac, kernel.
[10:32:21] <asac> ok
[10:32:27] <asac> right. i am fine with this for this week
[10:32:33] <jserv--> forget it. I'll try again.
[10:32:42] <asac> however, we should improve easyness to test etc.
[10:33:06] <asac> jserv--: well. all i want to ensure is that we know that our currently produced beagleboard build works ... ;)
[10:33:18] <jserv--> Roger it.
[10:33:26] <JamieBennett> great
[10:33:28] <JamieBennett> * luse to expand toolchain wiki page on distribution format options of toolchain and make a new wiki page on toolchain builder (or wrap in the toolchain main page)
[10:33:28] <asac> ignore kernel ... once we know that our build work we can ask jstultz and other folks to use that to work on kernel
[10:33:39] <luse> done
[10:33:44] <JamieBennett> thanks luse
[10:33:44] <asac> whats the link again=?
[10:33:48] <jserv--> ==> https://wiki.linaro.org/Platform/Android/Toolchain
[10:33:50] <asac> thanks
[10:33:56] <JamieBennett> * JamieBennett to setup weekly reports for the team
[10:34:02] <jserv--> Also, RFC is on list.
[10:34:25] <JamieBennett> done but we were missing a couple this week, looks like they are there now
[10:34:33] <asac> luse: i have a question: /gcc-patch ... is that a dir or repo that we would introduce?
[10:34:35] <JamieBennett> I will send a collated team report to linaro-dev later
[10:34:45] <jserv--> asac, s/gcc-patch/gcc-patches/
[10:34:54] <jserv--> that is a directory containing several patches.
[10:34:55] <asac> ah thats wrong on the wiki ... same question is true
[10:35:05] <asac> jserv--: i dont see that on aosp git
[10:35:05] <luse> is a git repo
[10:35:14] <jserv--> asac, http://lists.linaro.org/pipermail/linaro-dev/2011-February/002773.html
[10:35:17] <asac> ok so thats something that we (linaro) would introduce?
[10:35:42] <asac> jserv--: yeah ... havent finished all parts but will go back and complete reading and follow up on mailing list after meeting
[10:35:45] <asac> jserv--: so ignore my question
[10:35:50] <asac> if its in there ;)
[10:35:59] <jserv--> asac, sure, we will improve it.
[10:36:09] <asac> ok lets move on
[10:36:18] <JamieBennett> [TOPIC] Status Reports
[10:36:31] <JamieBennett> Thanks all for submitting your reports
[10:36:47] <JamieBennett> asac: anything we want to highlight there?
[10:36:59] <asac> so this topic (we will add that to standing agenda)
[10:36:59] <JamieBennett> Or shall I pull highlights from each report to discuss?
[10:37:02] <jserv--> here!
[10:37:14] <asac> is about reviewing the status reports to extract the highlights we want to put in our weekly mgmt report
[10:37:18] <asac> for this team
[10:37:21] <jserv--> for Linaro toolchain for Android, we should only maintain two git repo.
[10:37:38] <asac> its the first time we did it, so lets take a look at the form of what was submitted ;)
[10:37:41] <asac> https://wiki.linaro.org/Platform/Android/Meetings/2011-02-23
[10:37:49] <jserv--> android/toolchain/build.git & android/toolchain/gcc-patches
[10:37:58] <JamieBennett> jserv--: we will circle back to that in a moment
[10:38:04] <asac> its great that you submitted your stuff but as usual folks alwasy keep things really short
[10:38:08] <jserv--> so, we can remove the unused one in git.linaro.org
[10:38:20] <asac> its important to be more verbose here
[10:38:35] <JamieBennett> right, showcasing our effort is important
[10:38:42] <asac> the status report is there not only to report ... but also to trigger discussion and inspire folks from community
[10:38:52] <asac> so please invest 10 minutes time when doing it
[10:38:57] <asac> and write complete sentences
[10:39:21] <patrikryd> Ok.
[10:39:23] <asac> covering: what topic did you work, what problem did you face and try to solve, what were the discussed options, what was the conclusion
[10:39:23] <JamieBennett> Another important point is to provide links to more information where ever possible
[10:39:46] <asac> and what is the status and where is the code ;)
[10:40:01] <asac> yeah
[10:40:09] <asac> so lets look at a few as an example how to not do it ;)
[10:40:12] * jserv-- expects the template.
[10:40:22] <asac> * patch for creating tar files ... thats too short
[10:40:49] <asac> jserv--: well. its hard to make a template iuts just if you look at the report and think you are from outside, what would you be able to get from it
[10:41:10] <asac> try to tune that so that if you read it it gives all the background for a good engineer to get an entry point into the discussion/topic
[10:41:17] <patrikryd> I know I should have started writing ealier and I was blocking jserv-- :)
[10:41:26] <asac> patrikryd: heh ... no problem
[10:41:39] <jserv--> hmmm
[10:41:45] <JamieBennett> patrikryd and others, you can still expand over the course of today :)
[10:42:02] <patrikryd> will do.
[10:42:42] <JamieBennett> asac: shall we discuss highlights that are there now?
[10:42:42] <asac> so better might have been something: * finished changes to android build system to support tarball creation; backported patches X, Y, Z from master and had to do a patch A for the not yet in master supported roottarball patch; next: forward this upstream gerrit and see what they say :)
[10:42:53] <asac> JamieBennett: we can do that in a bit ;)
[10:43:15] <asac> other example:
[10:43:17] <asac> * discussing https://wiki.linaro.org/Platform/Android/Toolchain
[10:43:58] <asac> what was discussed, what are the main points there, what are the topics that still bother you there, what did you find an answer for and what are the answers that are still open; what is the status of discussion on linaro-dev/android etc.
[10:44:13] <asac> jserv--: makes sense?
[10:44:20] <jserv--> asac, I think so.
[10:44:22] <asac> cool
[10:44:26] <asac> i think its not a matter of template
[10:44:34] <asac> just reminding one to provide more context
[10:44:48] <asac> over time it will be easy especially once we have a team that does it right ;)
[10:44:57] <asac> then if someone new joins you can just tell him: look what others are doing
[10:45:05] <asac> ;)
[10:45:08] <asac> so lets improve over time
[10:45:17] <jserv--> asac, great!
[10:46:09] <asac> jserv--: jeremychang: patrikryd: do you want to try updating your status report after the meeting with a bit more verbosity? if so i would defer making the high level report to after this meeting
[10:46:09] <JamieBennett> we need to produce something like this - https://wiki.linaro.org/Platform/Infrastructure/Status/2011-02-17 - from the comments the engineers make
[10:46:13] <asac> and jamie and me will take a look
[10:46:24] * asac looks at that exampl
[10:46:24] <JamieBennett> obviously the more info we have the easier it is to produce
[10:46:43] <asac> JamieBennett: you confuse them
[10:46:48] <asac> thats not the format we are doing
[10:46:57] <asac> well thats the final outcome
[10:47:00] <asac> but thats not what engineers submit
[10:47:05] <JamieBennett> I was saying that the link is what *we* asac and I have to produce
[10:47:13] <JamieBennett> form engineer reports
[10:47:16] <JamieBennett> from
[10:47:31] <asac> ah right
[10:47:31] <asac> yeah
[10:47:45] <asac> but if you are verbose then its easy for us to take good content
[10:47:46] <asac> ;)
[10:47:54] <JamieBennett> right
[10:48:07] <JamieBennett> anyway, I think we have got it
[10:48:14] <JamieBennett> next week will be awesome
[10:48:29] <asac> -> [ACTION] jserv-- jeremychang and patrikryd to give another try at their weekly report by EOD; jamie/asac to produce weekly report tomorrow ?
[10:48:46] <asac> JamieBennett: or are we waiting another week ;)?
[10:48:48] <asac> i dont mind
[10:48:57] <JamieBennett> asac: I'll do a report this week
[10:48:58] * jserv-- nod.
[10:49:02] <asac> i think trying to improve report while its fresh
[10:49:08] <asac> is better than next week ;)
[10:49:23] <asac> awesome
[10:49:42] <jeremychang> asac, will try to make it more make sense.
[10:49:47] <asac> thanks
[10:49:48] <JamieBennett> [ACTION] jserv-- jeremychang and patrikryd to improve their weekly report before JamieBennett submits a team report to linaro-dev tomorrow
[10:50:07] <asac> ok great
[10:50:43] <asac> moving on?
[10:51:00] <JamieBennett> so jserv-- wanted to discuss something here
[10:51:04] <asac> sure
[10:51:07] <asac> is that AOB=
[10:51:08] <asac> ?
[10:51:14] <jserv--> JamieBennett, no more :)
[10:51:15] <asac> or same topic? e.g. review of highlights
[10:51:28] <JamieBennett> OK, jserv--, we can skip that
[10:51:29] <jserv--> JamieBennett, I prefer to use list for RFC topics.
[10:51:35] <JamieBennett> OK
[10:51:42] <jserv--> JamieBennett, thanks for noticing me. :)
[10:51:53] <asac> whats the problem of RFC topics?
[10:52:13] <asac> i am not sure we touched that topic at all above
[10:52:38] <jserv--> asac, no problem so far. I just need help from toolchain team.
[10:52:44] <asac> yes
[10:52:45] <JamieBennett> "jserv--: for Linaro toolchain for Android, we should only maintain two git repo."
[10:52:54] <asac> right
[10:53:01] <jserv--> asac, we can set up environment for toolchain team, and they might do something later.
[10:53:16] <asac> yeah
[10:53:29] <asac> thast the idea
[10:53:32] <asac> we cannot take their job
[10:53:45] <asac> we just make sure they get all the bits and pieces they need to efftively do their job
[10:53:58] <jserv--> agree.
[10:54:14] <patrikryd> jserv--: At the moment we have toolchain/manifest and toolchain/gcc
[10:54:16] <asac> ok lets move on ... i think the toolchain mail thread will be helpful for that
[10:54:21] <asac> jserv--: did you CC linaro-toolchain?
[10:54:28] <asac> might be a bunch of folks dont read -dev
[10:54:32] <jserv--> patrikryd, I think you can safely rm -rf toolchain/gcc
[10:54:39] <patrikryd> ok
[10:54:52] <JamieBennett> OK, with just 6 mins left, lets move to
[10:54:56] <JamieBennett> [TOPIC] Any Other Business (AOB)
[10:55:10] <JamieBennett> Anything else to raise before we close the meeting?
[10:55:11] <jserv--> asac, I cc. Michael Hope.
[10:55:11] <asac> i have a few things:
[10:55:19] <asac> 1. extend standing agenda
[10:55:54] <asac> this means we want more agenda points that are owned by individuals (like "platform/board support update (patrik)", "validation status update (jeremy)", etc.
[10:56:00] <asac> i will follow up on mail with a suggestion
[10:56:44] <asac> 2. for LEB we need to understand how to do that
[10:56:58] <JamieBennett> [ACTION[ asac to produce a better standing agenda for the meetings
[10:57:15] <JamieBennett> [ACTION] asac to produce a better standing agenda for the meetings
[10:57:19] <asac> patrik will lead this discussion and come up with a pitch by end of week if possible (e.g. how do we pull in branches etc) ... so please help him by provding him with ideas etc.
[10:57:41] <asac> he should know what we want to achieve ... so there should be no misinterpretation
[10:57:46] <asac> or doubt etc.
[10:57:59] <asac> (for those things he isnt sure we will continue a call today ;))
[10:58:16] <patrikryd> we have postponed the discussion in the email to this meeting.
[10:58:26] <patrikryd> Any suggestions now?
[10:58:28] <asac> yes
[10:58:29] <asac> well
[10:58:38] <asac> meetings are really not the right place to find solutions to problems
[10:58:44] <asac> rather raise problems and connect folks to follow up on that
[10:58:51] <asac> this problem is nothing we can solve here ;)
[10:58:56] <asac> at least thats my feeling
[10:59:19] <patrikryd> ok
[10:59:21] <asac> agreed? so do that on IRC/offline
[10:59:37] <patrikryd> ok
[10:59:38] <jserv--> agree.
[10:59:53] <asac> the point is we need to think this through and if there are problems with doing something like that
[10:59:59] <asac> come up with those problems early
[11:00:09] <asac> so we can adjust the plan that mgmt believes we will execute early ;)
[11:00:13] <jeremychang> agree.
[11:00:34] <JamieBennett> OK, se we are at an end
[11:00:38] <asac> also we have to understand if we need additional resources to get started or how the effort would evolve and ram pu
[11:00:39] <JamieBennett> #endmeeting
Meeting ended.

Platform/Android/Meetings/2011-02-23 (last modified 2011-02-23 17:03:24)