Wednesday 02 Feb 2011

People Present

  • slangasek
  • JamieBennett

  • dmart
  • hrw
  • jcrigby
  • ppearse
  • wookey
  • tgall_foo
  • kunal
  • aviksil

Agenda

  • Review action items from last meeting
  • Alpha-2 Testing
  • Blueprint status
  • Blueprint

    Assignee

    Status

    other-linaro-n-profiling-tools

    aviksil

    performance-linaro-n-memory-footprint

    dmart

    packageselection-foundations-n-multiarch-support

    slangasek

    other-linaro-n-cross-compilers

    hrw

    other-linaro-n-u-boot-upstreaming

    jcrigby

    other-linaro-n-cross-compilation-environment

    wookey

    other-linaro-n-lttng

    aviksil

    multimedia-linaro-package-meego

    tom-gall

    multimedia-linaro-standardize-omx-core

    goelkunal

    other-linaro-n-gdbserver-integration

    tom-gall

    multimedia-linaro-seeds-1105-for-multimedia

    tom-gall

    multimedia-linaro-seeds-1105-for-graphics

    tom-gall

    graphics-linaro-package-kwin

    tom-gall

Action Items from this Meeting

  • JamieBennett to come up with test assignments for the new developer image

  • jcrigby to write up landing team u-boot, kernel workflows and send for review after first draft is done
  • slangasek to review https://wiki.linaro.org/Platform/DevPlatform/Specs/MemoryFootprintMethodology for dmart

  • ppearse to figure out what version number we want for 'helios' to be in natty
  • dmart to send a beagle kernel config to aviksil1
  • aviksil1 to send pointer to lttng-enabled tree to slangasek

Action Items from Previous Meeting

  • DONE: wookey rewarded for great work on omxil components FTBFS by also helping get the ffmpeg-dist component building
  • DONE: slangasek to get tgall_foo his uboot env settings for a BB

Status Reports

Avik Sil

  • Worked on building and running LTTng patched Linaro kernel
  • Took off on 26th and 28th January.

Peter Pearse

Progress

  • arm-m-ael-alip-evaluation
    • Six cross patch merge proposals raised
    • Established why gobject introspection cannot be cross built
  • other-linaro-n-eclipse-cdt
    • Hello world building & running on host

  • Other
    • Cross chain bugs raised and resulting fixes tested

Issues

  • For gobject-introspection (and associated packages) to be cross built we need to ldopen target libraries on the host
  • eclipse package in natty needs to be from helios, rather than ganymede to make best use of available CDT cross support

Plans

  • arm-m-ael-alip-evaluation
    • More Natty cross builds & propose merges

    • Assist ARM internal cross build of LAMP stack
    • Check if gobject introspection target binaries be built from host built sources.
  • other-linaro-n-eclipse-cdt
    • Build hello world for target
    • Import existing projects
    • Build eclipse with xdeb

Absences

  • None planned

Kunal Goel

  • ffmpeg-dist enabled successfully in omxil components
  • Leave on 26th and 28th January.

Dave Martin

Activity

  • performance-linaro-n-memory-footprint
  • miscellaneous
    • Set up two remotely-accessible panda boards in the linaro cambridge office.
    • The two boards are configured as panda01 and panda02 on validation.linaro.org
    • Spent a bit of time looking into feasibility of porting apt-get to use a different backend for faster downloading of

package list updates etc. Looks potentially useful, But it doesn't look like a good all-round solution can be trivially arrived at. May continue to spend a bit of time looking into this, but it isn't a priority.

Plans

  • Find a working upstream kernel version and validate omap Thumb-2 compatibility patches on Panda and Beagle xM.
  • Discuss prioritisation for outstanding kernel tasks.
  • Write up the output from the Freescale i.MX BSP review discussion, and post for comment.

Absences

None planned.

Jamie Bennett

John Rigby

Tom Gall

  • libjpeg-turbo, libjpeg62-neon - pending PPA discussion (I think?)
  • linaro-development - had to get over the issue with plymouth locally, built, tested locally, changed submitted, made alpha2
  • linaro-nano - working on modify seed and image based on busybox
  • linaro-alip - working to modify based on conversations with ppearse. (slim down to xfce + firefox only)
  • testing alpha2 images {panda,Xm} {alip, efl, developer, headless} bugs #633227, #707037, #707038
  • meego - packaging : libresourceqt, contextkit waiting for review, libmeegotouch, meegotouch-compositor cleanup queued up for review. Good upstream discussion in #debian-meego about trying to startup official debianization team as part of meego. (it was originally suggested, it would be good to revisit)
  • chromOS Window Manager - https://wiki.linaro.org/Chromiumos/ChromeosWm created, task scoped, particulars documented.

  • gdbserver - part of linaro-development image.
  • kwin - no action

Meeting Log

Meeting opened by slangasek at 16:01

https://wiki.linaro.org/Platform/DevPlatform/2011-02-02

  • <slangasek> [TOPIC] Review action items from last meeting

Review action items from last meeting

  • <slangasek> * DONE: wookey rewarded for great work on omxil components FTBFS by also helping get the ffmpeg-dist component building

    <slangasek> * DONE: slangasek to get tgall_foo his uboot env settings for a BB

    <slangasek> that seems to easy

    <wookey> oh, hello

    <wookey> not taking undue credit there - steve fixed it :-)

    <slangasek> well, you got the ball rolling anyway :)

    <slangasek> hum, need to come up with more actions for next week

    <slangasek> that's too few :)

    <slangasek> [TOPIC] Alpha-2 Testing

Alpha-2 Testing

  • <slangasek> alpha-2 is due out this Thursday

    <hrw> I will test alip/panda - need to find working 4GB card

    <slangasek> JamieBennett has been diligently making sure we have images available to be tested, which means we need to make sure those efforts aren't going to waste and test those images between now and tomorrow :)

  • tgall_foo has images downloaded and has started testing

    <JamieBennett> pretty please ;)

    <tgall_foo> thanks for the X11 catch and xorg updates slangasek & JamieBennett

    <slangasek> ALIP and Netbook were both respun early today because of a fresh X bug that landed in natty this week

    <slangasek> so if you tested them already, please test again - hopefully they *work* this time

    <JamieBennett> and developer is fresh and new so go break please

    <slangasek> indeed!

    <slangasek> JamieBennett: hmm, from the tracker it looks like you've reset all of the tests to today, even though headless/developer shouldn't have been affected by the X bug - do those need retested?

    <JamieBennett> My stance as of earlier was that we *should* retest because of hwpack and image respins

    <JamieBennett> although we all know its probably good as of yesterday

    <hrw> 3.3MB/s from releases.linaro.org - nice

    <JamieBennett> I will test headless again today/tomorrow

    <tgall_foo> if you reset those ... just let us know .. I've already entered results that I'd need to reenter

    <JamieBennett> even though I did yesterday

    <hrw> can we have https://wiki.linaro.org/Platform/QA/TestCases/ALIPOMAP page?

    <slangasek> JamieBennett: the only hwpack that had relevant changes was omap3-x11-base, which obviously only impacts the graphical heads; I would expect us to use the known-good Tuesday builds for headless and developer rather than discarding the test results

    <tgall_foo> do we still have video out issues with panda ?

    <slangasek> hrw: we won't have anything meaningful there this week; just smoke-test the image please (boots to GUI, browser works, logout/shutdown work)

    <hrw> ok

    <JamieBennett> slangasek: there were only 2 test results on headless, both mine

    <slangasek> ok

    <tgall_foo> JamieBennett, I entered one a few moments ago as well

    <tgall_foo> and just about to enter another for the panda

    <JamieBennett> tgall_foo: for yesterdays?

    <tgall_foo> for today

    <jcrigby> tgall_foo, do you have the halting panda problem reported be others

    <jcrigby> s/be/by

    <tgall_foo> jcrigby, no

    <tgall_foo> mine's been running fine, I've been building images on it

    <slangasek> JamieBennett: can you also come up with a plan to parcel out the tests for the developer image? we obviously don't have anyone assigned to those yet :)

    <JamieBennett> slangasek: will do, please ACTION me

    <slangasek> [ACTION] JamieBennett to come up with test assignments for the new developer image

JamieBennett to come up with test assignments for the new developer image

  • <hrw> would be nice to have some shell script which will check common things after boot/login

    <slangasek> hrw: automated validation is taken care of by plars's team... we're here for the not-so-automated testing :)

    <tgall_foo> hrw, I wonder if plars doesn't have one

    <slangasek> things like "oh, it booted but I can't see anything" :-)

    <hrw> slangasek: I have serial console

    <hrw> D"

    <slangasek> anyways :)

    <slangasek> [TOPIC] Blueprint status

Blueprint status

  • <slangasek> $ echo $(shuf -e hrw jcrigby dmart wookey slangasek ppearse JamieBennett aviksil tgall_foo kunal)

    <slangasek> dmart kunal jcrigby ppearse JamieBennett hrw aviksil slangasek tgall_foo wookey

    <slangasek> jcrigby: looks like you get to go first

    <jcrigby> been working on new kernel package with u8500 support

    <jcrigby> it builds but have not tested yet

    <dmart> hi all, sorry I'm late

    <slangasek> jcrigby: sweet

    <slangasek> jcrigby: when do you think you'll push that to the archive?

    <wookey> dmart: JamieBennett has been diligently making sure we have images available to be tested, which means we need to make sure those efforts aren't going to waste and test those images between now and tomorrow :)

    <wookey> (from slangasek) - that the only thing of note you missed

    <dmart> ok

    <jcrigby> slangasek, as soon as I do some testing, I'm thinking even if it doesn't get to shell prompt on u8500 I may still push

    <jcrigby> thoughts?

    <slangasek> I'm fine with that

    <jcrigby> ok, great

    <jcrigby> u-boot has been on back burner so I need to get back to that

    <slangasek> asac is assigned the weekly testing of the u8500 hwpacks, so he can tell you what's not working

    <jcrigby> ok

    <jcrigby> one other thing, I think we need a wiki page describing landing team u-boot and kenrel workflows

    <jcrigby> both upstreaming and enablement

    <jcrigby> there seems to be some misunderstanding of common practices

  • slangasek nods

    <jcrigby> I will send notice for review after I get a first draft done

    <slangasek> [ACTION] jcrigby to write up landing team u-boot, kernel workflows and send for review after first draft is done

jcrigby to write up landing team u-boot, kernel workflows and send for review after first draft is done

  • <jcrigby> thats it for me, will do as much testing as I can today on the boards I have

    <slangasek> jcrigby: first priority is headless/imx51 for you :)

    <slangasek> dmart: your turn

    <jcrigby> did that already this morning

    <dmart> slangasek: ok

    <hrw> speaking of imx51 - does someone here has efikasb/efikamx from genesi?

    <dmart> The main update from me was that I wrote up a proposal for memory footprint measurement on the wiki

    <dmart> slangasek: did you have a chance to review it?

    <tgall_foo> hrw, I do

    <slangasek> dmart: afraid not; it's still queued in my browser, today looks promising

    <dmart> https://wiki.linaro.org/Platform/DevPlatform/Specs/MemoryFootprintMethodology?highlight=%28footprint%29%7C%28memory%29

    <tgall_foo> efikamx

    <dmart> (give or take the query string)

    <hrw> tgall_foo: uds-giveout or other source?

    <dmart> If anyone has view on it, let me know.

    <ppearse> hrw: Me too, although loaned out

    <tgall_foo> hrw, my own

    <dmart> Otherwise, I've mainly been working on re-merging the Thumb-2 compatibility patches in the OMAP bsp for 2.6.3[78]

    <slangasek> [ACTION] slangasek to review https://wiki.linaro.org/Platform/DevPlatform/Specs/MemoryFootprintMethodology for dmart

slangasek to review https://wiki.linaro.org/Platform/DevPlatform/Specs/MemoryFootprintMethodology for dmart

  • <dmart> slangasek: thanks

    <hrw> tgall_foo: ok, just wanted to know

    <slangasek> dmart: how's the thumb-2 compatibility going?

    <slangasek> dmart: that reminds me, there was an issue with the -mimplicit-it handling in qt4-x11 that was brought up on Friday; do you know anything about that?

    <slangasek> I think the Ubuntu ARM guys have been working on it together with the TC WG

    <dmart> slangasek: thumb-2: looks OK so far, the main issue has been to do with some churn upstream.

    <hrw> and with Kubuntu team

    <dmart> I have a kernel booting on xM and panda now

    <dmart> (a Thumb-2 kernel, that is ... otherwise it would be uninteresting)

    <slangasek> :)

    <dmart> I need upstream's help to check that it really works -- I'll be posting for upstream review in the next day or two.

    <dmart> slangasek: yes -- I provided some input, but afaik no conclusion has been reached about what the bug is.

    <slangasek> ok

    <slangasek> do you have a bug number for that? Strangely I don't see it at the top of the list of package bugs

    <slangasek> ppearse:

    <dmart> bug #

    <dmart> bug #705689

    <ubot2> Launchpad bug 705689 in qt4-x11 "unity-2d-launcher crashes with segfault error on armel (natty only)" [High,Confirmed] https://launchpad.net/bugs/705689

    <slangasek> ah, opaque bug title, ok

    <ppearse> arm-m-ael-alip-evaluation: More merge proposals - just in python2.7

    <slangasek> eep :)

    <slangasek> how many have I managed to get uploaded to Ubuntu for you so far? 2?

    <ppearse> Gave up on gobject-introspection - need to dlopen target libraries ;-{

    <slangasek> and one has made it into Debian

    <slangasek> saw that... that's a shame

    <wookey> hmm, ppearse you need to fill me in on that

    <ppearse> wookey: OK - later

    <ppearse> other-linaro-n-eclipse-cdt: Wish we had helios in natty, rather than ganymeded

    <slangasek> ppearse: what does 'helios' look like in numeric form? :)

    <ppearse> Downloaded ARM DA5 - want to use the model(s) as a target

    <slangasek> (I see 'eclipse-cdt 6.0.2-1, when I guess maps to 'ganymede'?)

    <ppearse> slangasek: Um - looking

    <slangasek> if it's just a matter of packaging a new upstream version, we may be able to make that happen

    <slangasek> ppearse: no need to find it now, you can get back to me after the meeting with that

    <ppearse> slangasek: Upstream 3.6.1

    <slangasek> [ACTION] ppearse to figure out what version number we want for 'helios' to be in natty

ppearse to figure out what version number we want for 'helios' to be in natty

  • <slangasek> ah, that's an eclipse release rather than an eclipse-cdt release

    <slangasek> that would be, er, trickier

    <slangasek> anyway, we can discuss later

    <slangasek> anything else?

    <ppearse> slangasek: Ack - that's it

    <slangasek> JamieBennett:

    <JamieBennett> ah, complicated week ;)

    <JamieBennett> So developer image happened

    <JamieBennett> please test ;)

    <slangasek> kunal: I guess your connection is stable now, so you can go after JamieBennett

    <JamieBennett> thank tgall_foo

    <JamieBennett> alpha-2 work

    <kunal> slangasek: ok

    <JamieBennett> lots of PM discssions

    <JamieBennett> and I inherited some work to do with Platform PM'ing, more to come

    <JamieBennett> EOF

    <JamieBennett> questions ;)

    <JamieBennett> ?

    <slangasek> none from me!

    <slangasek> thanks, JamieBennett

    <slangasek> kunal: you're up

    <kunal> ok

    <kunal> ffmpeg is enabled in components package

    <kunal> and it's going

    <kunal> after review i'll upload to revu

    <slangasek> almost seems redundant, but yes :)

    <kunal> now omxil bellagio being somewhat stable i can work on integrating review changes for meego packages

    <kunal> as omxil patches will take time from MM WG

    <kunal> that's from my side

    <slangasek> right - thanks

    <slangasek> hrw:

    <hrw> ok

    <hrw> updated cross toolchain packages to generate pkgsym ddebs - armel-cross-toolchain-base and gcc-4.5-armel-cross will land in merge queue tomorrow

    <hrw> started work on merging those source packages into debian

    <hrw> so far it will have to wait for multiarch or I will keep copy of linux-2.6 packaging inside of armel-cross-toolchain-base

    <slangasek> btw, xdeb and the cross-toolchain packages in natty were a big help when fixing the fbdev X bug yesterday

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611382

  • <hrw> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611382

    <ubot2> Debian bug 611382 in linux-2.6 "linux-2.6: Provide all packaging instructions in linux-source package" [Wishlist,Open]

    <slangasek> 33s to iterate a package build for testing - 22s to install it on my beagle with 'dpkg -i' :)

    <hrw> looks like this bug is treated as WONTFIX or INVALID

    <hrw> started testing alip on panda, found some dead sd cards

    <hrw> thats all

    <hrw> going to fosdem on friday morning

    <slangasek> in the Debian scheme of things, waiting for multiarch shouldn't be a problem... it's definitely landing early in the next Debian release

    <hrw> slangasek: s/release/cycle/

    <slangasek> yes

    <hrw> next Debian will be at fosdem 2013

    <slangasek> perhaps :)

    <slangasek> aviksil1:

    <aviksil1> ok

    <aviksil1> other-linaro-n-profiling-tools: Verified functionality of oprofile package in Ubuntu on ARM. It's working well.

    <aviksil1> other-linaro-n-lttng: Struggling on booting LTTng enabled linaro kernel, so far not succeeded

    <aviksil1> got stuck at a point where the root filesystem is not getting detected

    <dmart> aviksil1: what hardware have you tested on? Each BSP needs a small amount of support in order for perf/oprofile to work.

    <slangasek> hmm, tricky indeed

    <aviksil1> dmart: beagle boad C4

    <aviksil1> dmart: i saw CPU speed is estimated at 0 Mhz!

    <dmart> Could be worth people trying perf/oprofile on any boards they have

    <slangasek> aviksil1: and in your latest build you're using the config straight from the generated Ubuntu package, right?

    <dmart> ^ I've seen that too. I think that means "I don't know" ... doesn't seem to break anything

    <aviksil1> slangasek: yes

    <slangasek> and you say that leads to a hang on boot

    <aviksil1> slangasek: exactly

    <slangasek> jcrigby, dmart: what does that suggest to you for a next step here?

    <slangasek> aviksil1: btw, is the lttng merged tree published, where someone could try to reproduce this problem?

    <dmart> aviksil1: what kernel config are you using? This might just be a rootfs-drivers-not-available problem

    <dmart> or does the kernel really hang?

    <slangasek> aviksil1: or can you send me a howto on merging things up so I can test?

    <wookey> hang at what point in boot?

    <slangasek> dmart: when using omap2plus_defconfig, it was a rootfs-drivers-not-available problem

    <wookey> early on, before rootfs mount?

    <aviksil1> slangasek: ok, i'll send you the link

    <slangasek> dmart: I thought I identified the right driver for him to enable, but he said he still got the problem

    <dmart> I can send a config -- I generally use the linaro kernel config as a starting point

    <slangasek> [ACTION] dmart to send a beagle kernel config to aviksil1

dmart to send a beagle kernel config to aviksil1

  • <dmart> I don't trust the defconfigs since Linus stopped accepting updates to them

    <slangasek> [ACTION] aviksil1 to send pointer to lttng-enabled tree to slangasek

aviksil1 to send pointer to lttng-enabled tree to slangasek

  • <slangasek> dmart: right, on the list people were saying the defconfig definitely didn't DTRT

    <dmart> aviksil1: which kernel version are you using?

    <slangasek> 2.6.37 linaro

    <aviksil1> dmart: linux-linaro-2.6.37 with lttng patch

http://git.lttng.org/?p=linux-2.6-lttng-linaro.git;a=summary

  • <aviksil1> http://git.lttng.org/?p=linux-2.6-lttng-linaro.git;a=summary

    <slangasek> thanks

    <aviksil1> i rebased it to the latest linux-linaro-2.6.37

    <slangasek> short on time, so let's take the rest to #linaro

    <slangasek> my turn

    <aviksil1> sure

    <slangasek> worked on qemu-linaro this week - qemu-maemo is obsolete now, we have the first qemu-linaro package in the tools ppa for maverick and the source (but not binaries, yet) has been accepted into natty

    <slangasek> did some A2 testing; found a blocker bug; practiced cross-building packages with our tools for the fbdev bugfix

    <slangasek> sent nitpicky comments to everyone about their packages ;)

    <slangasek> and now my focus is on multiarch

    <slangasek> EOF

    <slangasek> any questions?

    <wookey> has qemu-static changed its name?

    <wookey> to match debian?

    <slangasek> qemu-static wasn't called that before, so yes

    <slangasek> now it matches Debian

    <wookey> cool

    <tgall_foo> very cool

    <slangasek> calling it 'qemu-kvm-extras-static' doesn't really fit :)

    <slangasek> tgall_foo: your turn

    <tgall_foo> linaro-development - had to get over the issue with plymouth locally, built, tested locally, changed submitted, made alpha2

    <tgall_foo> linaro-nano - working on modify seed and image based on busybox

    <tgall_foo> linaro-alip - working to modify based on conversations with ppearse. (slim down to xfce + firefox only)

    <tgall_foo> testing alpha2 images {panda,Xm} {alip, efl, developer, headless} bugs #633227, #707037, #707038

    <ubot2> Launchpad bug 633227 in linux-ti-omap4 "instabilities with highmem activated" [High,Confirmed] https://launchpad.net/bugs/633227

    <ubot2> Launchpad bug 707037 in linux-linaro "panda kernel warnings" [Undecided,New] https://launchpad.net/bugs/707037

    <ubot2> Launchpad bug 707038 in linux-linaro "panda: no gfx due to missing kernel functionality" [High,Triaged] https://launchpad.net/bugs/707038

    <tgall_foo> meego - packaging : libresourceqt, contextkit waiting for review, libmeegotouch, meegotouch-compositor cleanup queued up for review. Good upstream discussion in #debian-meego about trying to startup official debianization team as part of meego. (it was originally suggested, it would be good to revisit)

    <tgall_foo> chromOS Window Manager -

    <tgall_foo> https://wiki.linaro.org/Chromiumos/ChromeosWm created, task scoped, particulars documented.

    <tgall_foo> gdbserver - part of linaro-development image.

    <tgall_foo> kwin - no action - libjpeg-turbo, libjpeg62-neon - pending PPA discussion (I think?)

    <tgall_foo> EOL - questions ?

    <slangasek> tgall_foo: firefox is probably the wrong choice for 'slim'; there are other gecko-based browsers available that are smaller, e.g. fennec

    <tgall_foo> slangasek, ok .. ppearse thoughts on that ? ok to change ?

    <slangasek> tgall_foo: also to ditch the GNOME libs we want to get rid of gdm... I suggest trying nodm as a replacement since we don't really need a real login manager

    <slangasek> (nodm will require some integration work in the overlay packages to get it to DTRT by default)

    <wookey> yes nodm is great

    <hrw> or slim

  • tgall_foo nods

    <ppearse> tgall_foo: Use case is "boot & browse" - any will do.

    <tgall_foo> great thanks ppearse

    <wookey> dillo :-)

    <hrw> good to know that guys from FSO made something from xserver-nodm ;D

    <wookey> me now?

    <slangasek> yes

    <hrw> wookey: dillo would fetch gtk1?

    <wookey> Most time gone on tcl-extension cross-building and corresponding fixes for tcl and dpkg-cross

    <slangasek> overtime, go go go :)

    <wookey> Not entirely obvious what the right fix is. Still researching.

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=599206 and http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611650 relevant

  • <wookey> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=599206 and http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611650 relevant

    <ubot2> Debian bug 599206 in dpkg-cross ""dpkg-cross: document file removal process" [Wishlist,Fixed]

    <ubot2> Debian bug 611650 in tcl8.5-dev "tcl8.5-dev contains arch-dependent file in /usr/share/ (tcltk/tcl8.5/tclConfig.sh)" [Important,Open]

    <wookey> hrw: yes, not serious suggestion, but I miss it every day

    <wookey> Steve - need clarification about what happens to arch-independent files in /lib in multiarch world. Do they stay in /lib/<tuple> or /usr/lib/<tuple>?

    <wookey> perhaps later by mail

    <hrw> wookey: there was dillo/gtk2 but still suxx

    <wookey> Lots of discussion at ARM on xdeb use and improvements, and on bootstraping. Good to have real users but they find problems and suck time :-)

    <wookey> I still have stuff to merge, PPA and upstream

    <wookey> Some time on emdebian sprint preparation

    <wookey> Will prod salgado imminently about xdeb multi-command restructuring - he promised to do the basic python structure for me, then I only have to worry about the internals.

    <slangasek> wookey: sorry, not much clarification to offer there; arch-independent files are generally not supposed to be under /lib at all, but if they are it's likely to be implementation-dependent where they wind up

    <wookey> yes, so the question is should dpkg-cross just copy them all over (willthing break? - we certainly get a lot more not-empty -cross packages)

    <wookey> or should dpkg-cross special-case tclconfig.sh

    <slangasek> in multiarch? dpkg-cross shouldn't be used

    <slangasek> there's nothing to cross in a multiarch package

    <wookey> yes, sorry - lack of clarity

    <wookey> there is a) what do dpkg-cross do for now

    <wookey> and b) thinking about multiarch location for offending files may help clarify my thoughts on the matter

    <wookey> right so in ,mutliarch we expect they'll get put somewhere more sensible?

    <slangasek> wookey: I submitted an xdeb branch for your review, to merge tcl8.5 into the whitelist; dunno if you got the mail

    <wookey> or possibly suffered to stay in lib and exist multiple times

    <slangasek> but I found that I can't actually get xdeb to DTRT even with that whitelist entry

    <wookey> yes I did

    <slangasek> I don't know what's going on there

    <wookey> the required file doesn;t end up in the -cross package at the moment

    <wookey> so sqlite3 won;t build

    <slangasek> wookey: in multiarch, if it's not put somewhere sensible it's a clear bug in the package - the package has to /work/, and it has to be /coinstallable/, and everything else follows from these requirements

    <wookey> hence question of whether to add it to list of files dpkg-cross crosses. either by name or by saying 'everythig in /usr/lib/*/*

    <slangasek> so this will work itself out over time

    <slangasek> anyway, we can talk about this at length in #linaro :)

    <wookey> indeed

    <slangasek> [TOPIC] AOB

AOB

  • <slangasek> anything else today?

    <wookey> I've written most of a long mail - will send in soon

    <ppearse> Not from me

    <slangasek> #endmeeting

Platform/DevPlatform/Meetings/2011-02-02 (last modified 2011-03-11 11:35:39)