Android Kernel Meeting

Agenda

  • Noritsuna is sick, so no meeting today.


Android Team Meeting

Agenda

  • Actions from last meeting
  • Team issues discussion - Alexander and JamieBennett asking questions

  • LEB/Image review - jserv
  • Validation update - Jeremy
  • Bug status - Patrik
  • AOB

Past Action items

  • asac to talk to the LT about missing binaries as per https://wiki.linaro.org/JimHuang/Sandbox/AndroidForPanda

  • asac and jamie to flash out blueprint handling details and communicate to team
  • JamieBennett to setup status.linaro.org for the Android team

  • patrikryd to push one safe gcc patch and see what happens
  • jeremychang to investigate why the display doesn't work on Panda with jstultz' 2.6.38 kernel
  • asac to add JamieBennett's ssh key to the amazon build machine

  • asac to talk to james about authentication madness for paste.linaro.org
  • jeremychang to complete the whiteboard of https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-basic-validation

  • asac and jserv-- to flesh out panda LEB spec and actions
  • jserv-- to send email to android-platform telling folks about our toolchain package!! (marketing) and pointing out the build failures we get in platform and asking whether they want contributions to fix them properly

People Present

Actions Recorded

  • asac to invite TI to Android weekly meeting
  • asac to chase TI on the mm bits
  • jim to create a the panda-gfx-integration-initial blueprint
  • Team to review what they are working on and document what isn't approved/doesn't have a Blueprint/is not targeted to discuss in next weeks meeting
  • patrikryd to try to push generic hook patch to upstream and see what happens
  • jserv-- to send email to android-platform telling folks about our toolchain package!! (marketing) and pointing out the build failures we get in platform and asking whether they want contributions to fix them properly
  • patrikryd to fill out summary on https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-platform-kernel-boottarball

  • patrikryd to fix work item syntax on https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-platform-kernel-boottarball

  • asac and JamieBennett to talk to james_w about keeping android blueprints on android team radar

  • jeremy to add a more verbose summary to linaro-android-benchmark-suite-integration-into-lava and poke asac/jamie for approving
  • luse to fix work items and get linaro-android-toolchain-benchmark to approved state
  • luse/jserv-- to draft a new blueprint that is about integrating the google benchmarks itself .... this blueprint would come before linaro-android-toolchain-benchmark
  • asac to think about validation meta blueprint and what to do with that with jamie and jeremychang
  • jserv-- to fix linaro-android-toolchain-ndk WI syntax
  • patrik to add l-m-c update WI to boottarball spec and assgn that WI to jeremy; also add verification WI owned by him to test the complete "solution"
  • JamieBennett get the bug bot in here finally

  • JamieBennett to hunt down infrastructure projects and ensure they are properly covered by android team including bug driver powers etc.

  • patrikryd to sort out href8500 build situation


Engineers Reports

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).

John Stultz <jstultz_*>

Highlights

  • Thanks to Alexander for the patchlink, I managed to get panda hdmi working (using the port marked DVI) with the linaro+android kernel.
    • Hitting an issue with DNS not resolving, but that might just be my userland config.
    • Not yet pushing this branch out, as the current linaro-2.6.38 tree broke USB on Beagle.

Plans

  • Have a micro-usb adapter in the mail to try to look into enabling/testing ADB support.

Issues

  • linaro.android-38 tree is blocked on beagle usb issue (lp bug 747639)

Jeremy Chang <jeremychang>

Highlights

Plans

  • Prepare Lava needed modification for android to review and discuss

Issues

Jim Huang <jserv-->

Highlights

Plans

Issues

luse

Highlights

Plans

Issues

Patrik Ryd <patrikryd>

Highlights

Plans

Issues

  • -


IRC logs

Android Kernel Meeting

To be added.

Android Team Meeting

Started logging meeting in #linaro-meeting
[09:03:22] <JamieBennett> [LINK] https://wiki.linaro.org/Platform/Android/Meetings/2011-04-06
[09:04:39] <JamieBennett> [TOPIC] Actions from last meeting
[09:05:11] <asac> hi!
[09:05:39] <JamieBennett> * asac to talk to the LT about missing binaries as per https://wiki.linaro.org/JimHuang/Sandbox/AndroidForPanda
[09:05:54] <asac> yep
[09:06:00] <asac> we found something very nice
[09:06:19] * asac gets the url
[09:06:30] <asac> http://git.omapzoom.org/?p=device/ti/proprietary-open.git;a=shortlog;h=refs/heads/gingerbread
[09:06:42] <asac> seems they have a) a prerelease binary for GFX there
[09:06:50] <asac> and b) under a license that doesnt even require click through
[09:06:58] <JamieBennett> \o/
[09:07:03] <asac> jim is working on pulling that into a leb branch build
[09:07:21] <asac> i am sorting out licensing (if its really too good to be true)
[09:07:44] <asac> also we have initial engineering contact. TI is keen to work with us on the our LEB as they see things are mutually beneficial
[09:08:08] <asac> i will connect jim to the engineers and also invite them to our IRC channel
[09:08:19] <JamieBennett> Sounds like we have a good plan for Android on TI hardware then
[09:08:22] <asac> next: find the same for multimedia
[09:08:31] <asac> JamieBennett: mm is missing ... but thats next ;)
[09:08:41] <JamieBennett> [ACTION] asac to invite TI to Android weekly meeting
[09:08:58] <asac> also action to follow up on mm bits
[09:09:20] <asac> jim will do a final test, document his results in the gfx-initial blueprint and then we can declare that done
[09:09:28] <JamieBennett> [ACTION] asac to chase TI on the mm bits
[09:09:32] <JamieBennett> OK
[09:09:34] <asac> JIM will create a new blueprint that has Work Items for GFX integration
[09:10:11] <JamieBennett> * asac and jamie to flash out blueprint handling details and communicate to team
[09:10:11] <asac> panda-gfx-leb-integration ... should include hooking things up in our manifest; producing working builds; ensuring that leb-branch also builds the right kernel and that it works out of box
[09:10:15] <asac> etc.
[09:10:21] * JamieBennett nods
[09:10:36] <asac> [ACTION] jim to create a the panda-gfx-integration-initial blueprint
[09:10:40] <asac> thanks
[09:10:44] <JamieBennett> [ACTION] jim to create a the panda-gfx-integration-initial blueprint
[09:10:59] <JamieBennett> So for the next action item we need to explain a plan for Blueprints
[09:11:30] <JamieBennett> The idea is that a light-weight Blueprint should be created for each project
[09:11:50] <JamieBennett> around a months worth of work, no more, on a specific topic
[09:12:08] <asac> (at best even in weeks chunks ;))
[09:12:23] <JamieBennett> This Blueprint should be set to review when an engineer has put the required information in there, including work items
[09:12:43] <JamieBennett> A tech lead will review the Blueprint and set to aprove
[09:12:52] <JamieBennett> Only then should the Blueprint be worked on
[09:13:04] <JamieBennett> Each week in this meeting we will review the state of all Blueprints
[09:13:31] <JamieBennett> and if any are coming to an end we can select the next Blueprint to be worked on
[09:13:55] <JamieBennett> Does this make sense to every one? asac anything to add?
[09:14:17] <asac> right. important here is that we understand what is being worked on and make things explicit ... we have a bit of a grown situation atm where stuff is being worked on that is at drafting /discussion state and stuff that is not targetted against a milestone
[09:14:24] <JamieBennett> oh milestones
[09:14:39] <JamieBennett> So each Blueprint should target a specific series and a specific milestone
[09:14:39] <asac> so what i would really like to do is a one time effort to get that straight
[09:15:06] <asac> so action for everyone: if you work on stuff that is not having a milestone nor a series set or if you work on something that is not in state approved let us know
[09:15:26] <asac> at best in this meeting during the blueprint review slot we added
[09:15:44] <patrikryd> What is the next state after "Drafting"? "Review" or "Pending approval"?
[09:15:47] <asac> then i love to have a bank of prepared blueprints like we currently do and approve stuff that is properly drafted
[09:16:00] <asac> approved stuff == its well drafted and we want to work on
[09:16:06] <asac> note: thats != we decided to work on it ...
[09:16:18] <asac> the step to put something on the CPU is done here in the meeting
[09:16:25] <asac> by setting a series and milestone
[09:16:28] <asac> so
[09:16:29] <asac> https://blueprints.launchpad.net/linaro-android/
[09:16:44] <asac> if you look there and you work on a blueprint that has no series means that we didn't explicitly decide to work on it
[09:16:52] <asac> and we cannot easyilc ommunicate that way to outside what is worked on
[09:17:08] <asac> patrikryd: discussion -> drafting -> review -> pending approval -> approved
[09:17:17] <asac> though pending approval is a bit undefined
[09:17:32] <asac> we could say we use pending approval for the state "ready, but not officially started"
[09:17:42] <asac> but our tracking system currently works better with series and milestone
[09:17:48] * JamieBennett nods
[09:17:53] <JamieBennett> So
[09:17:54] <JamieBennett> [ACTION] Team to review what they are working on and document what isn't approved/doesn't have a Blueprint/is not targeted to discuss in next weeks meeting
[09:18:07] <asac> JamieBennett: lets do that in the blueprints slot in this meeting ;)
[09:18:11] <asac> at least a first look
[09:18:17] <JamieBennett> OK
[09:18:26] <JamieBennett> everyone clear on the process?
[09:18:47] * jeremychang nods.
[09:19:12] <JamieBennett> lets finish off the action items then look at the Blueprints then
[09:19:17] <asac> sure
[09:19:18] <JamieBennett> * JamieBennett to setup status.linaro.org for the Android team
[09:19:20] <JamieBennett> c/o
[09:19:26] <JamieBennett> * patrikryd to push one safe gcc patch and see what happens
[09:19:32] <JamieBennett> patrikryd: still waiting on a patch?
[09:19:49] <patrikryd> Think I found one, but have not pushed it yet.
[09:19:59] <JamieBennett> Oh, link?
[09:19:59] <jeremychang> sorry, we should start after the status is "approved" ?
[09:20:05] <JamieBennett> jeremychang: yes
[09:20:52] <jeremychang> know better about the status.
[09:20:55] <patrikryd> link to the patch?
[09:20:58] <asac> jeremychang: JamieBennett: not really
[09:21:05] <JamieBennett> patrikryd: yes
[09:21:06] <JamieBennett> asac: ?
[09:21:06] <asac> Approved means: this stuff is ready to work on
[09:21:17] <patrikryd> http://git.linaro.org/gitweb?p=android/platform/build.git;a=commit;h=974a4ac74a31a7e86a91a9f605264e8cfc8a3a77
[09:21:23] <asac> it shouldnt be worked on with high speed if its not milestoned and targettted
[09:21:32] <asac> so approval is the step of agreeing on the work to be done
[09:21:46] <JamieBennett> thanks patrikryd
[09:21:48] <asac> while milestoning and targetting against a series is the step to agree on "the work should be done now"!
[09:22:13] <asac> patrikryd: cool. please try and see
[09:22:37] <asac> patrikryd: also this quoting patch you submitted for boottarball ... is that needed? and if so its probably upstream worthy too
[09:22:57] <JamieBennett> [ACTION] patrikryd to try to push generic hook patch to upstream and see what happens
[09:23:46] <jeremychang> asac, thanks.
[09:23:54] <patrikryd> asac: I will do the opposite. Someone must have solved it on AOSP. I will try to find patch and cherry pick it form AOSP.
[09:24:26] <asac> patrikryd: well. if its not on master its not there i guess
[09:24:33] <asac> anyway
[09:24:34] <asac> move on
[09:24:42] <asac> all we want to do is push stuff
[09:24:42] <JamieBennett> * jeremychang to investigate why the display doesn't work on Panda with jstultz' 2.6.38 kernel
[09:24:49] <JamieBennett> this was done
[09:24:53] <JamieBennett> right jeremychang ?
[09:25:07] <jeremychang> I see John has updated the HDMI status
[09:25:11] * JamieBennett nods
[09:25:23] <jeremychang> but He has not pushed the commits.
[09:26:03] <JamieBennett> jeremychang: "Not yet pushing this branch out, as the current linaro-2.6.38 tree broke USB on Beagle."
[09:26:20] <JamieBennett> so I think we can consider this done
[09:26:39] <jeremychang> So still wait John to sort out the other usb issues.
[09:26:46] <JamieBennett> right
[09:26:58] <JamieBennett> * asac to add JamieBennett's ssh key to the amazon build machine
[09:27:00] <JamieBennett> -> done
[09:27:06] <JamieBennett> * asac to talk to james about authentication madness for paste.linaro.org
[09:27:26] <asac> i think i did that
[09:27:33] <asac> paste.linaro.org is really a private pastebin
[09:27:40] <asac> you can only look at it if you are private-newcore
[09:27:50] <asac> i have to find out who initiated that and why it was selected to be public
[09:27:50] <JamieBennett> OK
[09:28:05] <asac> we said maybe we can open it up after removing all the pastes
[09:28:31] <JamieBennett> * jeremychang to complete the whiteboard of https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-basic-validation
[09:29:11] <jeremychang> I updated some for this blueprint, though I moved the benchmark part to the other one.
[09:29:25] <asac> let me check
[09:29:27] <JamieBennett> looks OK to me, anything to add asac?
[09:29:42] <JamieBennett> needs series but we can discuss that in a moment
[09:29:58] <asac> jeremychang: ok ... so first this needs to be targetted for a series ... otherwise its not supposed to be worked on by the definition above :)
[09:30:01] <asac> yeah
[09:30:06] <asac> but its ok ... that should go on series now
[09:30:19] <jeremychang> asac, got the point.
[09:30:20] <asac> jeremychang: is 11.04 realistic for the work items to be finished?
[09:30:25] <asac> thats the current milestone
[09:30:35] <jeremychang> asac, should be. in 11.04
[09:30:41] <asac> yeah
[09:30:42] <asac> good
[09:31:00] <asac> otherwise i would have used this as an example how to handle 2 month projects ;)
[09:31:03] <asac> anyway, all good
[09:31:08] <JamieBennett> * asac and jserv-- to flesh out panda LEB spec and actions
[09:31:26] <asac> jeremychang: if you update work items e.g. adding removing/changing, please drop a comment in the whiteboard like i did
[09:31:38] <asac> JamieBennett: thats pending and an action we gave jim above, right?
[09:31:44] <jeremychang> asac, ok.
[09:31:58] <JamieBennett> OK
[09:32:17] <JamieBennett> last one
[09:32:23] <JamieBennett> * jserv-- to send email to android-platform telling folks about our toolchain package!! (marketing) and pointing out the build failures we get in platform and asking whether they want contributions to fix them properly
[09:32:39] <jeremychang> the other one for running benchmark is https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-benchmark-suite-integration-into-lava
[09:32:51] <asac> jeremychang: ack.
[09:33:23] <JamieBennett> I don't see any email to android-platform, did this happen?
[09:33:25] <asac> lets look at those in blueprints meeting slot
[09:34:14] <JamieBennett> OK, lets carry over jserv--'s action item as I don't think it was done
[09:34:24] <JamieBennett> [ACTION] jserv-- to send email to android-platform telling folks about our toolchain package!! (marketing) and pointing out the build failures we get in platform and asking whether they want contributions to fix them properly
[09:34:40] <JamieBennett> [TOPIC] Blueprint review
[09:34:56] <JamieBennett> How do you want to do this asac ?
[09:35:09] <asac> https://blueprints.launchpad.net/linaro-android/
[09:35:17] <asac> https://blueprints.launchpad.net/linaro-android/linaro-11.05
[09:35:23] <JamieBennett> [LINK] https://blueprints.launchpad.net/linaro-android/
[09:35:29] <JamieBennett> [LINK] https://blueprints.launchpad.net/linaro-android/linaro-11.05
[09:35:35] <asac> lets start with the second
[09:35:40] <asac> there is ablueprint that is not approved ;)
[09:35:47] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-platform-kernel-boottarball
[09:36:08] <asac> ACTION: patrikryd to fill out summary;
[09:36:16] <asac> sorry
[09:36:26] <asac> ACTION: patrikryd to fill out summary of boottarball layout
[09:36:29] <asac> errr
[09:36:30] <JamieBennett> [ACTION] patrikryd to fill out summary on https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-platform-kernel-boottarball
[09:36:31] <asac> man i suck
[09:36:38] <asac> yeah JamieBennett got it right ;)
[09:36:50] <JamieBennett> then it will be approved
[09:36:59] <asac> [ACTION] patrikryd to fix work item syntax
[09:37:03] <JamieBennett> (half the WI's are already done)
[09:37:04] <asac> on that blueprint
[09:37:14] <JamieBennett> [ACTION] patrikryd to fix work item syntax on https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-platform-kernel-boottarball
[09:37:16] <asac> a) you dont use 1. 2. etc. for work items
[09:37:24] <patrikryd> ok
[09:37:25] <asac> b) all lines should have a :TODO|DONE|INPROGRESS
[09:37:39] <JamieBennett> asac: actually the syntax is fine
[09:37:48] <asac> patrikryd: can you do that right away? that would at least mean that all blueprints currently officially worked on are in "good shape"
[09:37:51] <patrikryd> except last line
[09:37:52] <asac> JamieBennett: the numbers get stripped?
[09:38:07] <asac> patrikryd: yeah... though if you have nothing the tracker will parse it as : TODO
[09:38:07] <JamieBennett> no, they become part of the work item description
[09:38:11] <asac> i just prefer to have that in each line
[09:38:19] <asac> JamieBennett: yes, so numbers should be dropped :)
[09:38:23] <JamieBennett> OK
[09:38:32] <JamieBennett> drop numbers and add status
[09:38:46] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-build-infrastructure-frontend
[09:38:49] <asac> this is strange
[09:38:56] <asac> not according to our process
[09:39:28] <asac> [ACTION] asac and JamieBennett to talk to james_w about keeping android blueprints on android team radar
[09:39:40] <asac> example: https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-build-infrastructure-frontend
[09:39:47] <JamieBennett> [ACTION] asac and JamieBennett to talk to james_w about keeping android blueprints on android team radar
[09:39:57] <JamieBennett> what is wrong with that one?
[09:40:04] <JamieBennett> It started in Infrastructure
[09:40:24] <asac> JamieBennett: that one is ok. but i think other blueprints are somewhere else even though they are andrdoid topics
[09:40:29] <asac> JamieBennett: but even then it isnt milestoned
[09:40:35] <JamieBennett> right
[09:40:39] <asac> we can talk about that offline
[09:40:53] <JamieBennett> OK
[09:42:21] <asac> ok lets look at the other list of blueprints
[09:42:37] <asac> https://blueprints.launchpad.net/linaro-android
[09:44:06] <asac> jeremychang: linaro-android-basic-validation
[09:44:11] <asac> whats the status on that?
[09:44:18] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-basic-validation
[09:44:27] <asac> ah we already discussed that ;)
[09:44:31] <jeremychang> asac, reboot_to_linaro_android_image and adb check is done
[09:44:32] <asac> so same action applies
[09:44:45] <asac> jeremychang: well. i am talking about gettting from drafting to approved state
[09:44:49] <asac> and then getting it milestoned etc.
[09:44:52] <jeremychang> asac, todo: deploy and submit_result
[09:45:06] <JamieBennett> asac: it is milestoned "[asac Mar 19, 2010]: setting priority to essential (ASAP); setting milestone to next available - aka 11.04 (even though this should really happen asap)"
[09:45:19] <asac> ok so its pending approval
[09:45:30] <asac> JamieBennett: ok
[09:45:34] <asac> i think the work items are ok to work on
[09:45:38] <asac> lets approve it
[09:45:52] <jeremychang> asac, yes, agreed.
[09:45:56] <asac> done
[09:46:02] <asac> its approved so its now officially worked on ;)
[09:46:13] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-benchmark-suite-integration-into-lava
[09:46:21] <asac> thats still blocked on infrastrcutre according to status
[09:46:56] <asac> jeremychang: for getting that from drafting to approved, please expand the summary to better explain what this spec wants to do/achieve
[09:47:31] <asac> JamieBennett: did you change that series etc.?
[09:47:50] <JamieBennett> asac: just changed it as it was already milestoned to a milestone in that series
[09:48:00] <jeremychang> asac, basic validation is a prerequisite for other testing. so thinking if basic validation should do result submitting.
[09:48:13] <asac> JamieBennett: please drop comments in the blueprints
[09:48:16] <asac> if you do a change
[09:48:18] <JamieBennett> OK
[09:48:21] <asac> i want to do that as well
[09:48:24] <asac> its really hard to see history
[09:49:01] <JamieBennett> asac: right, but if you milestone a Blueprint it needs the series setting too as milestones are tied to series'
[09:49:08] <asac> [ACTION] jeremy to add a more verbose summary to linaro-android-benchmark-suite-integration-into-lava and poke asac/jamie for approving
[09:49:18] <JamieBennett> asac: i.e. you used this milestone - https://launchpad.net/linaro-android/+milestone/linaro-11.05-final
[09:49:27] <asac> JamieBennett: thats true. but if its drafting it shouldnt be worked on
[09:49:29] <JamieBennett> which as you can see is part of the 11.05 series
[09:49:41] <asac> sure. but i was able to set it ;)
[09:49:42] <JamieBennett> right, but milestones and series' go together
[09:49:46] <asac> and didnt document it either
[09:49:52] <JamieBennett> :)
[09:49:56] <asac> just saying: if you change things, add comment to whiteboard
[09:49:59] <JamieBennett> OK
[09:50:01] <asac> same applies for everyone
[09:50:11] <JamieBennett> [ACTION] jeremy to add a more verbose summary to linaro-android-benchmark-suite-integration-into-lava and poke asac/jamie for approving
[09:50:12] <asac> except changing from TODO to DONE ... there some common sense should apply
[09:50:28] <asac> sometimes a DONE is worth a comment, like: "you can look at this cool change now here: "
[09:50:41] * JamieBennett needs to write a script to generate a visualisation of Blueprint change history
[09:50:46] <asac> linaro-android-toolchain-benchmark
[09:50:59] <asac> we aready discussed this
[09:51:17] <asac> ACTION: luse to fix work items and get this to approved state
[09:51:41] <asac> ACTION: luse/jserv-- to draft a new blueprint that is about integrating the google benchmarks itself .... this blueprint would come before linaro-android-toolchain-benchmark
[09:51:53] <JamieBennett> [ACTION] luse to fix work items and get linaro-android-toolchain-benchmark to approved state
[09:52:06] <JamieBennett> [ACTION] luse/jserv-- to draft a new blueprint that is about integrating the google benchmarks itself .... this blueprint would come before linaro-android-toolchain-benchmark
[09:52:19] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-validation
[09:52:24] <asac> this one is a tricky one
[09:52:28] <asac> its a meta blueprint
[09:52:39] <asac> we cannot approve it if all sub blueprints are not approved
[09:52:48] <jeremychang> right.
[09:52:51] <asac> jeremychang: can we make this linaro-android-validation-1
[09:52:54] <JamieBennett> asac: should of been a tr so it is visualised on status
[09:52:56] <asac> and make a linaro-android-validation-step-2
[09:53:08] <asac> where we move the blueprints that are not to be done in first batch (e.g. not by 11.04
[09:53:10] <asac> )
[09:53:18] <asac> JamieBennett: not sure
[09:53:31] <asac> atm i dont understand trs ;)
[09:53:47] <JamieBennett> tr's = meta-blueprint of related blueprints
[09:53:56] <asac> thats your interpretation
[09:54:02] <JamieBennett> = status.l.o prettiness
[09:54:03] <patrikryd> :)
[09:54:16] <asac> for me its true that tr is "normally" a meta blueprint
[09:54:21] <asac> however, not all meta blueprints are TRs
[09:54:23] <asac> that is for sure
[09:54:33] <JamieBennett> they are for visualisation
[09:54:38] <JamieBennett> and the tools we use
[09:54:48] <asac> well. TRs are tehnical requrements
[09:54:53] <asac> the tools visualize them
[09:55:08] <JamieBennett> right
[09:55:10] <asac> but saying we declare things TRs so we can levearge the visualization feature we did for TRs is wrong
[09:55:20] <asac> at least if they are not TRs ;)
[09:55:27] <patrikryd> Who can write a TR?
[09:55:28] <JamieBennett> but take http://status.linaro.org/group/tr-ad-all.html for example
[09:55:45] <asac> patrikryd: TRs are done by team leads in coopreation with TSC and mgmtn during requirements gathering
[09:55:52] <asac> for platform i want to do that every 3 month
[09:56:11] <asac> but TRs are basically reflecting the TSC approved goals with priorities
[09:56:20] <JamieBennett> asac: not sure that is correct
[09:56:20] <patrikryd> ok
[09:56:31] <patrikryd> :)
[09:56:31] <JamieBennett> asac: On the Ubuntu-side they are considering using TR's
[09:56:35] <asac> so we might agree with TSC:" next three month we focus on toolchain validation"
[09:56:43] <asac> JamieBennett: thats a different problem
[09:56:52] <asac> they would still be Technical Requirements
[09:56:58] <asac> that are goals and priorities and directions
[09:57:07] <asac> if ubuntu want to redefine that thats their thing
[09:57:16] <asac> but if they are not Technical Requirements anymore, they should use a different name
[09:57:25] <JamieBennett> syntax ;)
[09:57:30] <JamieBennett> anyway, heh, lets chat about it later
[09:57:33] <asac> we can think about using the same visualization to visualize meta bluerpitns
[09:57:53] <asac> but i will not make something a linaro tr- if its not a requirement gatehered topic with priority etc.
[09:57:56] <asac> yeah
[09:58:21] <asac> JamieBennett: tr-add-all is a catch all tr ... we have a catch all tr for plaform
[09:58:33] <JamieBennett> indeed, but that is a meta for sure
[09:58:41] <JamieBennett> not a requirement from the TSC
[09:58:47] <asac> which means: everything that goes there is not agreed by TSC and we just link it there so that TSC sees what is done without a TR
[09:58:50] <asac> in platform
[09:59:01] <asac> JamieBennett: its the only exception to show "what are we doing that is not a TR"
[09:59:07] <asac> a catch all
[09:59:20] * patrikryd thinks TR discussion is interesting, but is getting hungry.
[09:59:48] * JamieBennett nods
[09:59:50] <asac> [ACTION] asac to think about validation meta blueprint and what to do with that
[09:59:53] <asac> with jamie and jeremychang
[10:00:04] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-cts-integration-in-lava
[10:00:13] <JamieBennett> [ACTION] asac to think about validation meta blueprint and what to do with that with jamie and jeremychang
[10:00:27] <asac> jeremychang: -> that one is in drafting state and it should stay there until its all unblocked etc.
[10:00:31] <asac> also lacks work items
[10:00:38] <asac> but no need to do those quickly
[10:00:46] <asac> though if you dont draft it, it should be discussion
[10:01:00] * asac sets it to discussion
[10:01:06] <jeremychang> yeah, this one becomes a meta, I am not sure about what's best way for this, basically we can see a pretty dependency tree from this blueprint, I want to know more about tr too.
[10:01:23] <asac> ok
[10:01:41] <asac> jeremychang: linaro-android-monkey-integration ... i will set that one to discussion as well
[10:01:44] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-monkey-integration-in-lava
[10:02:00] <asac> it doesnt have work items and i think we should only draft one or two blueprints while working on others
[10:02:08] <asac> lets set everything back to discussion that has no work items
[10:02:17] <asac> and then explicitly select which bluerpints we want to draft next
[10:02:29] <jeremychang> asac, ok.
[10:02:41] <asac> same for https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-sikuli-integration-in-lava
[10:02:59] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-toolchain-build
[10:03:02] <asac> what about that?
[10:03:28] <asac> ok work item syntax is busted
[10:03:33] <asac> but all items except one are done
[10:03:46] <asac> jeremychang: i dropped a comment.
[10:03:57] <asac> JamieBennett: can we fix the work items for him because its almost done
[10:04:01] <asac> and target/milestone it?
[10:04:05] <JamieBennett> sure
[10:04:15] <asac> [ACTION] JamieBennett to put https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-toolchain-build on CPU
[10:04:39] <asac> https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-toolchain-integration
[10:04:54] <asac> lets move the "file bugs for build failures with -Werror not disabled and attach to this BP: TODO"
[10:05:08] <asac> work item to a new blueprint "linaro-android-platform-for-linaro-toolchain"
[10:05:20] <asac> which might get killed if we decide to not use it
[10:05:46] <asac> linaro-android-gingerbread-platform-for-linaro-gcc
[10:05:50] <asac> linaro-android-gingerbread-platform-for-linaro-gcc-4.6
[10:06:15] <asac> actually lets kill that WI and ignore that spec
[10:06:17] <asac> and approve it
[10:08:11] <asac> tentatively for 11.04
[10:08:15] <asac> done
[10:08:31] <asac> ok lets stop
[10:08:39] <asac> there are 7 blueprints missing
[10:08:46] <asac> asac will finish those during his spare time
[10:09:10] <asac> actually one: https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-toolchain-ndk
[10:09:26] <asac> this one shouldnt be done with high prio
[10:09:37] <asac> needs WI fixing. then goes to approved state
[10:09:56] <asac> and jserv-- should tell me if he continues to work on and then we can move it to series/milestone etc.
[10:10:10] <asac> [ACTION] jserv-- to fix linaro-android-toolchain-ndk WI syntax
[10:10:16] <JamieBennett> [ACTION] jserv-- to fix linaro-android-toolchain-ndk WI syntax
[10:10:31] <asac> [ACTION] jserv-- to talk wiht asac about next steps on ndk blueprint and when to target that and so on
[10:10:40] <asac> ok the rest should be ok
[10:10:43] <asac> to do offline
[10:11:18] <JamieBennett> We are over time but should quickly go through the rest of the meeting
[10:11:30] <JamieBennett> [TOPIC] LEB/Image review - jserv
[10:11:37] <JamieBennett> so Jim is not here
[10:12:01] <JamieBennett> anything to discuss about LEB's asac ?
[10:12:32] <asac> just an update that jim bravely managed to get froyo GFX binaries also work with gingerbread build
[10:12:53] <asac> next step: jserv-- wrap up gfx investigation spec and mark it as done/delivered
[10:13:04] <asac> then: jserv-- to draft a spec for GFX integration
[10:13:13] <asac> and probably already works in that in parallel
[10:13:22] <asac> i think we alredy discussed the action for him on that
[10:13:24] <asac> so move on
[10:13:26] <asac> oh right
[10:13:39] <asac> one word on the generic panda build: i still have no luck with screen + dvi
[10:13:50] <asac> it seems its just broken in all linux-linaro trees for me
[10:14:07] <asac> jstultz is trying .39 kernel now that has some fixes
[10:14:10] <asac> and then we can backport
[10:14:14] <asac> ok move on
[10:14:17] <JamieBennett> [TOPIC] Validation update - jeremychang
[10:14:40] <jeremychang> the status of https://code.launchpad.net/~jeremychang/linaro-image-tools/android is merged.
[10:14:59] <jeremychang> keep working on https://blueprints.launchpad.net/linaro-android/+spec/linaro-android-basic-validation
[10:15:26] <jeremychang> android image reboot from the master images and the adb connection check is done.
[10:15:41] <asac> yay ... landed first batch
[10:15:45] <asac> of image-tools
[10:15:57] <jeremychang> my environment update is written here, https://wiki.linaro.org/JeremyChang/Sandbox/LavaAndroidValidation
[10:16:12] <asac> patrikryd: can we add a work item about adjusting l-m-c to the boottarball spec? ... you can assign it to jeremy by using [jeremychang] Work ite text: TODO
[10:16:15] <asac> syntax
[10:16:25] <asac> patrikryd: and ping jeremychang once the boottarball has landed ;)
[10:16:55] <patrikryd> I don't know much about l-m-c.
[10:18:21] <asac> patrikryd: you dont need to know anything about it
[10:18:22] <jeremychang> thanks. I have not tried patrikryd's patch set. yeah. the work item can be assigned to me after the boottarball is done.
[10:18:34] <asac> just add a work item that will include updating l-m-c as part of the boottarball spec
[10:18:37] <patrikryd> Ok
[10:18:40] <asac> owner of that work item is jeremychang
[10:18:45] <patrikryd> I'll update.
[10:19:06] <asac> also add a work item owned by you to verify that everything works out of box after l-m-c and boottarball feature landed
[10:19:11] <asac> that will wrap up the spec nicely
[10:19:22] <asac> and will mean it has clean, working and tested deliverables
[10:19:34] <asac> which is the ultimate goal of all blueprints (in theory)
[10:20:01] <patrikryd> I'll let jeremechang check what I write so it makes sense and he understands.
[10:20:03] <asac> [ACTION] patrik to add l-m-c update WI to boottarball spec and assgn that WI to jeremy; also add verification WI owned by him to test the complete "solution"
[10:20:09] <asac> sure
[10:20:18] <asac> usually WI do not need to be written art ;)(
[10:20:24] <asac> just express what needs to be done in small steps
[10:20:29] <JamieBennett> [ACTION] patrik to add l-m-c update WI to boottarball spec and assgn that WI to jeremy; also add verification WI owned by him to test the complete "solution"
[10:20:31] <asac> to complete implementation and deliver the bits in there
[10:20:49] <patrikryd> asac: But the should make sense. :)
[10:20:53] <asac> hehe
[10:21:14] <asac> ok
[10:21:20] <asac> you are doing the right thing i am sure ;)
[10:21:40] <asac> is the boottarball spec "approved"? if not just ping me after that and i will approve ;)
[10:22:02] <patrikryd> ok
[10:22:09] <JamieBennett> lets get patrikryd section over with so he doesn't starve ;)
[10:22:10] <JamieBennett> [TOPIC] Bug status - Patrik
[10:22:15] <asac> lol
[10:22:24] <patrikryd> We have 9 open bugs
[10:22:34] <patrikryd> two new and high
[10:22:47] <patrikryd> bug 744648 and bug 747911
[10:23:14] <patrikryd> The first one is unassigned
[10:23:15] <asac> https://bugs.launchpad.net/linaro-android/+bug/747911
[10:23:18] <asac> that one is tricky
[10:23:27] <asac> so you say that its not fixable?
[10:23:31] <asac> i doubt that ;)
[10:23:38] <asac> patrikryd: isnt there a .mk that does the copy?
[10:23:45] <patrikryd> All software is fixable
[10:23:47] <asac> isnt there a way to chmod as well in such a .mk
[10:23:49] <asac> hehe
[10:23:54] <asac> i mean without existing means
[10:24:08] <asac> my understanding is that the init.omap.sh is a best practice
[10:24:12] <asac> as init.rc explicitly calls that
[10:24:25] <asac> so if that has wrong permissions everywhere its a good contribution to aosp
[10:24:48] <asac> ok lets look at 744648
[10:25:32] <asac> yeah thats not directly our team, but we shoud keep this on the radar
[10:25:42] <patrikryd> ok
[10:25:44] <asac> as the build system is an important part of the overall android release strategy
[10:26:03] <asac> [ACTION] JamieBennett get the bot in here finally
[10:26:37] <asac> [ACTION] JamieBennett to hunt down infrastructure projects adn ensure they are properly covered by android team including bug driver powers etc.
[10:26:45] <JamieBennett> [ACTION] JamieBennett get the bug bot in here finally
[10:26:46] <asac> ask me later about this one
[10:26:56] <JamieBennett> [ACTION] JamieBennett to hunt down infrastructure projects and ensure they are properly covered by android team including bug driver powers etc.
[10:26:57] <asac> ok thanks
[10:27:14] <asac> patrikryd: where do we track the dvi not working for panda bug=?
[10:27:22] <asac> patrikryd: also should we have a bug about href8500 not working?
[10:27:26] <asac> or can we kill that build?
[10:27:33] <asac> and wait for snowball?
[10:27:40] <asac> seems noone has the right boards for this anyway
[10:28:14] <patrikryd> I vote for killing Android u8500href and wait for Snowball.
[10:28:16] <asac> [ACTION] patrikryd to sort out href8500 build situation
[10:28:22] <asac> patrikryd: ok please drive that
[10:28:28] <asac> and communicate it back into your organization
[10:28:29] <asac> ;)
[10:28:43] <JamieBennett> [ACTION] patrikryd to sort out href8500 build situation
[10:28:46] <asac> we should have a bug or something to track the rational for disabling
[10:29:22] <asac> e.g. file a bug asking for permanent removal of 8500 build and
[10:29:36] <asac> a short rational and then just remove it and set to fixed release
[10:29:42] <asac> ok great
[10:29:51] <asac> sorry for long overrun again
[10:30:00] <asac> blueprint review will be shorter next time around
[10:30:09] <asac> one time effort to get status right
[10:30:30] <asac> and have folks learn/feel what it should mean in practice if something isnt approved/targetted against series and milestoned
[10:30:31] <JamieBennett> [TOPIC] AOB
[10:30:31] <patrikryd> I'll check with ST-Ericsson if it is ok to kill u8500href build and replace it soon with a Snowball build.
[10:30:39] <asac> i think thats easiest way to lear ;)
[10:30:58] <JamieBennett> Anything else to add before I close the meeting and go to the gym ;)
[10:31:11] <asac> patrikryd: its not a matter of whether they are ok. its a matter of telling them that we will kill it if a), b) and c) doesnt happen :)
[10:31:19] <asac> and then if they say: "ok, kill it, go ahed"
[10:31:29] <asac> if they say: "wait, lets make that happen", we can also do it :)
[10:31:51] <patrikryd> I would phrase it a bit nicer. :)
[10:32:07] <asac> patrikryd: filing a bug that says it straight would be good
[10:32:17] <asac> e.g. no boards; no flashkit
[10:32:24] <asac> without that we cannot test/have it
[10:32:32] <asac> patrikryd: we can keep the builds running as personal builds
[10:32:41] <asac> anyway, i think first step is bug
[10:32:45] <patrikryd> At the same time ST-Ericsson wants to know what we need from them for 11.05
[10:32:49] <asac> then we can list whats the problem and suggest way forward
[10:32:55] <asac> and suggest to wait for snowball
[10:33:06] <asac> patrikryd: right. thats what we should use the bug for
[10:33:10] <patrikryd> ok
[10:33:14] <asac> at least what would we need for 8500
[10:33:23] <patrikryd> Don't we have the flashtool?
[10:33:23] <asac> and we can have a snowball blueprint
[10:33:32] <asac> patrikryd: not working with all versions
[10:33:38] <patrikryd> The old one depending on Java not the new "riff"
[10:33:52] <asac> just with the boards version we have ... and from what i understand u8500 is not working because we have no kernel that works with our board revisions
[10:33:54] <patrikryd> ok
[10:34:02] <asac> yeah we have the java one
[10:34:18] <asac> but even saying that we want to ship something without having a public flashkit is quite a far call :)
[10:34:21] <asac> but i could accept it
[10:34:26] <asac> if we have kernel and boards...
[10:34:37] <asac> but personally i think the outcome should be: get us snowballs and bits for that ;)
[10:34:38] <patrikryd> I tried to "create" such a kernel, but.... anyway lets try to kill it and get snowballs.
[10:34:45] <asac> yep
[10:35:21] <JamieBennett> OK, lets wrap this up
[10:35:24] <JamieBennett> #endmeeting
Meeting ended.

Platform/Android/Meetings/2011-04-06 (last modified 2011-04-07 12:59:39)