Include: Nothing found for "^## meetings$"!
Include: Nothing found for "^## end-meetings$"!
What hardware we support
Supported Developement Tools
We also maintain a set of how-to's related to cross-compiling
Out primary goal is to provide a common platform to use as the basis for Validation test heads and optimized hardware-specific builds. We work with a few teams to help make this happen:
Toolchain Working group - Validate and integrate a toolchain outputting code optimized for the common target instruction set.
Kernel Working group - Provide a test integration point for kernel consolidation, and bootloader consolidation.
Power Management Working group - Produce images that benefit from the work.
We have weekly meetings that anyone can attend. You can view the schedule and see logs of past meetings here.
We also meet face-to-face at Linaro Connect events.
- Welcome new starter: Riku Voipio
- Actions from last meeting - Steve chasing
- Engineer progress reports - All team members
Work Item Progress and Team Issues discussion - Steve and JamieBennett questions to the team
- Any other business (AOB)
Past Action items
tgall_foo to check with GrueMaster for recommendations on checking minimal audio functionality on armel
- fturgis to ask linaro-dev whether systemtap support for probing kernel modules is wanted
- slangasek to check why armel-cross-toolchain-base FTBFS wasn't caught in lucas's archive rebuild
- slangasek to review/merge/upload armel-cross-toolchain-base 1.62
hrw to write improved test cases for QA tracker, to cover https://wiki.linaro.org/Platform/Ubuntu/EvaluationBuild/MemberDeliverables
Your Name <irc nick>
- 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.
- Your individual plans for the coming week(s).
- Your individual plans for the coming week(s).
Marcin Juszkiewicz <hrw>
- new versions of armel cross toolchain landed in archive (1.62) and is in review (1.63)
- fixed lintian errors mostly
- postponed remaining work items
Frederic Turgis <fturgis>
- nice 1.5 days of holidays
- identified and tested 4 timestamping methods: tapset, do_gettimeofday(), sched_clock() and board dependent custom code
- each method commented + example code on wiki (tapset is not accurate so do not use it)
kernel module support: no answer except Alex Sack but he wanted info on systemtap rather than wanting kernel module support
- follow-up on ARM utrace/uprobe support on utrace-devel and systemtap mailing lists. Waiting for answer
- started rewriting 64-bits memory write. Not tested yet.
- deserves discussion about them on Linaro mailing list (after looking for knowledgeable people internally)
- follow-up with systemtap guys on why tapset one is done this way and tradeoffs of other solutions
- finish 64-bits memory write and submit to systemtap guys
again pre-empted by management. I guess I will have to finalize remaining tasks at slow speed However, I have some time to follow-up Linaro even after leaving cycle so that shall be OK.
Tom Gall <tgall_foo>
- nano merge requests waiting review. (change to live-helper config and linaro-seeds)
- merge request to add gdb-server awaiting review.
- Packaged latest libturbo-jpeg code (v1.1.0) from Mandeep. Fixed runtime detection code (simd detection presumed it had to be running on armv7 + NEON hardware in order to build in NEON simd optimization)
- Remaining todo:
- Awaiting feedback from Mandeep
- need to track down where lintian is finding an executable stack
- need to track down where lintian is finding something not build with -fPIC
- what arm ppa should this go to?
- Remaining todo:
- handed cross gdb ddeb tool work off to Steve.
- Packaged libav from Rony Nany.
- Remaining todos:
- couple of small lintian warnings to address
- awaiting feedback from Rony
- what ppa should this go to?
- Remaining todos:
- Test-o-rama continued last week. Sound, and graphical issues are a big troubling this late in the cycle. There's no working sound across the 3 boards I've been testing. I noticed that a fix for 746023 might have an impact on this issue. For the graphical issues, miss-rendering on ALIP and LEB occurs when using the omap3-x11-base hwpack but not omap3. (Optimized vs fb)
- Packaging Chromium
- gflags, gmock, glog, gtest, libchrome, libchromeos, libcros, flimflam, chromeos-assets, update-engine, metrics and vectormath is the initial package set.
- No blueprint was ever created for Chromium work.
Avik Sil <aviksil>
- My patch for the oprofile bug #702999 has been added by Nicolas in linux-linaro-2.6.38. Waiting for it to be available in the packaged kernel
- Created deb package for extra lttng-modules using dkms, sent for review and expecting it to be added to linaro overlay.
- Test lttng-modules package after it is added to linaro overlay
- verify oprofile works well on panda board after the fix
- Wrote pkg-config wrapper and link-generation postinst
- Worked on xdeb bugs heading for final release for cycle
- Helped armhf team with bootstrapping stuff
- Did some cross-tools packaging work
- Get xdeb release ready
- Test and post pkg-config work
- So many things to do, so little time
Kunal Goel <kunal>
- kde-workspace building fine for arm now
- smartt application by multimedia WG uploaded to ppa
- kdeedu build failing for arm
John Rigby <jcrigby>
- Released new kernel and u-boot packages (two u-boot actually)
- Fixed an upstream u-boot bug that caused old Beagles not to boot
- Submitted yet another u8500 patch set, hopefully the last
- A new kernel package is needed but u-boot is done for this cycle
- Look at android fastboot for u-boot
- Compare Thumb2 vs ARM kernel memory usage
- Investigate U8500 MMC problems
Started logging meeting in #linaro-meeting [15:06:21] <slangasek> [LINK] https://wiki.linaro.org/Platform/DevPlatform/Meetings/2011-04-20 [15:06:29] <slangasek> [TOPIC] Welcome new starter: Riku Voipio! [15:06:32] <slangasek> suihkulokki: go ahead ;) [15:07:05] <suihkulokki> I think most of you already know me - I'm a long term Debian Developer, as well as have worked on QEMU recently [15:07:34] <suihkulokki> Previous workplaces were Movial and Nokia [15:07:53] <suihkulokki> And I'm quite excited about joining Linaro now :) [15:07:58] <wookey> Do you know what you are working on yet? [15:08:04] <tgall_foo> Welcome! [15:08:07] <aviksil> suihkulokki: welcome to Linaro! [15:08:17] <ppearse> hi Riko [15:08:30] <ppearse> s/o/u/ [15:08:35] <slangasek> suihkulokki has accepted a position with Linaro as our designated ARM Porter - so he's your go-to guy on the team for things that aren't working right ;) [15:08:40] * ppearse has hay-fever [15:08:54] <fturgis> everything ? [15:09:02] <wookey> yep :-) [15:09:05] <slangasek> well, maybe not everything... :) [15:09:05] <dmart> good to have someone dedicated to the porting [15:09:24] <suihkulokki> for fixing everything I'll need your help :) [15:09:33] <slangasek> so - welcome, Riku, it's great to have you aboard [15:09:45] <slangasek> [TOPIC] Past action items [15:09:45] <wookey> we'll have everyone who is anyone soon [15:10:17] <slangasek> whoops, one tick, hunting down last week's actions [15:10:28] <slangasek> * tgall_foo to check with GrueMaster for recommendations on checking minimal audio functionality on armel [15:10:31] <slangasek> * fturgis to ask linaro-dev whether systemtap support for probing kernel modules is wanted [15:10:34] <slangasek> * slangasek to check why armel-cross-toolchain-base FTBFS wasn't caught in lucas's archive rebuild [15:10:37] <slangasek> * slangasek to review/merge/upload armel-cross-toolchain-base 1.62 [15:10:57] <slangasek> got the answer for a-c-t-b: the package was already failing to build before the linux-libc-dev so didn't show up as a regression [15:11:03] <slangasek> straightforward enough :) [15:11:09] <fturgis> kernel module support: no answer except Alex Sack but he wanted info on systemtap rather than wanting kernel module support Â I guess we can wait for systemtap v1.4 to be integrated [15:11:16] <tgall_foo> Done ... talked with GrueMaster this mornng .... didn't suggest using different tools then what I already am (speaker-test, aplay, xmms2) [15:11:35] <wookey> I found a bug in armel-cross-toolchain-base yesterday (mentions armel specifiucally in a couple of places it shouldn't) [15:11:56] <wookey> might be fixed in bzr - not looked yet [15:12:03] <slangasek> and a-c-t-b is uploaded... one last gcc-4.4-armel-cross upload to be accepted through the freeze queue, and then I believe we're set for natty [15:12:12] <wookey> hrw-phone: are you following or not-really? [15:12:20] <slangasek> wookey: yes, I think I have a merge pending from hrw-phone for that - Sledge found the bug already for armhf [15:12:32] <slangasek> or hrw found the bug when Sledge asked him for armhf, or something [15:13:02] <hrw> hi [15:13:17] <slangasek> fturgis: ack on systemtap, I agree we can wait then [15:13:56] <slangasek> tgall_foo: ok, sounds like you're in good shape then, great [15:14:49] <slangasek> [TOPIC] Engineering progress reports [15:15:40] <hrw> wookey: a-c-t-b armel issue is fixed in 1.63 [15:15:52] <slangasek> $ echo $(shuf -e hrw jcrigby dmart wookey slangasek ppearse aviksil tgall_foo kunal fturgis suihkulokki) [15:15:53] <wookey> hrw: is that what just got uploaded? [15:15:55] <slangasek> slangasek jcrigby dmart kunal wookey hrw tgall_foo ppearse fturgis aviksil suihkulokki [15:16:29] <hrw> wookey: no, it is in review now [15:16:29] <slangasek> man, me first again? [15:16:36] <wookey> heh [15:16:58] <hrw> slangasek: you are, so you have a lot of time to tell who will lead us [15:18:06] <hrw> shit. 1.63 is not full even [15:18:57] <slangasek> so, not a lot of engineering from my side in the past week. a few sponsored uploads to the archive, things are good shape there now I think... have been helping review branches, and making sure everything's on track with blueprints since I won't be able to see them through to the end of the cycle :( [15:19:05] <slangasek> working on getting xdeb uploaded this week [15:19:29] <slangasek> hrw: not yet known, I'm afraid. I'll let you guys know as soon as there's a decision :) [15:20:16] <slangasek> various discussions around planning for the next cycle, and what we should be discussing at LDS [15:20:36] <slangasek> if you haven't already, please have a look at https://wiki.linaro.org/Cycles/1111/TechnicalTopics/Platforms which sets the high-level direction for the team next cycle [15:21:00] <slangasek> if there are things you think we should be tending to that aren't on there, please let me know so we make sure to keep them on the agenda for LDS and beyond [15:21:04] <slangasek> [LINK] https://wiki.linaro.org/Cycles/1111/TechnicalTopics/Platforms [15:21:08] <slangasek> EOF [15:21:10] <slangasek> questions? [15:21:38] <slangasek> jcastro: [15:21:40] <slangasek> eh [15:21:40] <slangasek> jcrigby: [15:21:48] <slangasek> jcastro: hi, ignore me ;) [15:22:18] <jcrigby> New u-boot and kernel packages released (2 u-boot packages actually) [15:22:34] <jcrigby> fixed an upstream u-boot bug that cause old beagles to not boot [15:22:55] <jcrigby> submitted one last (hopefully) u8500 u-boot patch series [15:23:22] <jcrigby> plans for next week are a kernel package since upstream has more changes. [15:23:38] <jcrigby> and then I think I will look at u-boot fastboot upstreaming [15:23:44] <jcrigby> thats it [15:24:29] <slangasek> jcrigby: aviksil mentioned that a fix for bug #702999 is staged somewhere; is that on track to be included in the next kernel upload? [15:24:31] <ubot2> Launchpad bug 702999 in linux-ti-omap4 "oprofile failure on panda (omap4)" [Undecided,Confirmed] https://launchpad.net/bugs/702999 [15:24:59] <jcrigby> slangasek, I need to check. I saw that comment and need to make sure it is in nico's tree [15:25:03] <slangasek> ok [15:25:11] <slangasek> should I assign the bug to you for follow-up? [15:25:18] <aviksil> jcrigby: slangasek: it has been added in npitre's tree [15:25:22] <jcrigby> yes, that would be fine [15:25:29] <jcrigby> aviksil, ok great [15:26:04] <slangasek> great news on u8500 and fastboot [15:26:17] <slangasek> jcrigby: thanks [15:26:20] <slangasek> dmart: [15:27:06] <dmart> The main things from me are: [15:27:27] <dmart> - VFP coredumps patch is now in the linaro tree and rmk's patch system [15:28:08] <dmart> - verified that ftrace works with thumb-2 [15:28:25] <dmart> - dynamic ftrace seems to work too (which was the more interesting question) [15:29:06] <dmart> I've also been helping davepigott with one or two things relating to the validation farm [15:29:52] <dmart> jcrigby: do you think MMC should work for u8500 in your u-boot tree now? [15:29:58] <dmart> jcrigby: that was one of our problems... [15:30:12] <jcrigby> dmart, it still has a problem on some cards [15:31:10] <jcrigby> dmart, I'll add that to my list of todo's [15:31:17] <slangasek> that reminds me; does anyone have figures on how much memory a thumb-2 kernel saves? [15:31:35] <jcrigby> slangasek, doh, I dropped the ball on that [15:32:00] <hrw> jcrigby: will you have u8500 at luds-o? [15:32:08] <slangasek> jcrigby: wasn't critical; maybe dmart knows the answer already and we can save you some work :) [15:32:13] <dmart> slangasek: You tend to save ~30% on the code, but I don't have precise figures in front of me [15:32:19] <slangasek> ok [15:32:36] <hrw> jcrigby: I have 2 minisd cards which gave me some problems on pandaboard [15:32:37] <dmart> I want to run some benchmarks, but that's still on my queue [15:32:45] <slangasek> kiko would like a real figure he can point at [15:32:50] <hrw> jcrigby: but no minisd->sd adapter anymore [15:33:02] <slangasek> jcrigby: so if you get a chance to roll a non-thumb omap kernel to compare, that'd still be appreciated [15:33:44] <slangasek> dmart: thanks [15:33:45] <slangasek> kunal: [15:33:54] <kunal> yes [15:33:55] <jcrigby> slangasek, I will do that [15:34:15] <kunal> kde-workspace building fine for arm now [15:34:40] <kunal> tested in graphics WG [15:34:57] <kunal> results of test by jammy still awaited [15:35:07] <kunal> smartt application by multimedia WG uploaded to ppa [15:35:33] <kunal> they have added a new parameter for memory [15:35:54] <kunal> kdeedu was still giving problem in arm build, [15:36:25] <kunal> thats it [15:36:40] <slangasek> thanks [15:36:52] <slangasek> hopefully we'll get feedback from jammy soon [15:36:55] <slangasek> wookey: [15:37:11] <wookey> 3-days in cross-divisional tools conf and ARM induction process, so limited real work [15:37:15] <wookey> Wrote pkg-config wrapper and link-generation postinst [15:37:17] <wookey> Working on xdeb bugs heading for natty release. hope to nobble/bodge 731079 at least. [15:37:20] <wookey> Helped armhf team with bootstrapping stuff [15:37:20] <wookey> Have failed to do multistrap FFe. Am I too late already (it's probably bust on natty just due to missing configs)? [15:37:23] <wookey> Did some cross-tools packaging work [15:38:00] <slangasek> multistrap - very, very small window left; "it's currently broken" is a pretty good argument though [15:38:07] <slangasek> please get your FFe done this week for that [15:38:40] <slangasek> great news on the pkg-config wrapper. what do you think is the best way to publish this? [15:38:42] <wookey> OK. ubuntu bugs are always slightly tricky because I have to find the magic 'file bugs without apport' page [15:38:57] <slangasek> https://bugs.launchpad.net/ubuntu/+source/multistrap/+filebug ? [15:39:10] <wookey> yes - that'll be the one [15:39:51] <wookey> didn't know there was a 'specify pakage in URL featuer - handy [15:40:22] <hrw> wookey: https://bugs.launchpad.net/ubuntu/+source/multistrap/ has "report bug" in rigth top corner [15:40:24] <slangasek> RESTless leg syndrome [15:40:50] <wookey> right. OK. I'll try doing things that way in future [15:41:23] * hrw do not use apport/ubuntu-bug to report bugs - those tools are not so useful as they should. prefer to report bug and (if needed) use apport-collect BUGID to add date [15:41:47] <slangasek> wookey: see my question above about how we want to publish the pkg-config wrapper? I was actually wondering if it made sense to include our symlink in the cross-toolchain metapackage or something (gcc-arm-linux-gnueabi) rather than creating a separate package or whatnot [15:41:48] <hrw> s/date/data [15:42:05] <slangasek> not sure any of the options are very pretty, but maybe you have a preference [15:42:08] <wookey> slangasek: in email? - no not yet [15:42:25] <wookey> I've currently just added about 13 symlinks to the pkg-config package itself [15:42:38] <slangasek> ok [15:42:46] <wookey> dithering about whether postinst or during-built is preferable [15:42:56] * slangasek nods [15:43:03] <wookey> but yes it could be in dpkg-cross instead or something [15:43:07] <slangasek> well, we can talk about that further after the meeting [15:43:10] <wookey> indeed [15:43:14] <slangasek> hrw: [15:43:17] <hrw> ok [15:43:29] <hrw> as usual updated cross toolchain packages [15:43:56] <hrw> this time a-c-t-base got lintian overrides and that _armel removed (my mistake) [15:44:11] <hrw> and gcc-4.4-armel-cross is fixed - waiting in review queue [15:44:34] <slangasek> yep, should go in today [15:44:43] <hrw> checked (finally) - -no-system-directory-poisioning stuff - postponed due to lack of binutils support [15:44:55] <wookey> Oh, and I fixed xdeb lintian check for new lintian - but presuably that's not going to be in natty? [15:45:08] <slangasek> then we need one last upload of gcc-defaults-armel-cross now that everything else is fixed, and we should be done for natty [15:45:23] <hrw> wookey: the one which I was talking with Neil few days ago on #emdebian? [15:45:23] <slangasek> wookey: I don't expect any newer lintian in natty [15:45:32] <wookey> that's what I thought [15:45:43] <hrw> wookey: xdeb lintian check is in xdeb package.... [15:46:01] <wookey> yes - that's why I had to fix it (as an external lintain check) [15:46:06] <hrw> xdeb: /usr/share/lintian/checks/xdeb [15:46:23] <wookey> the internal ones are all done anyway [15:47:02] <hrw> I played also with some random bugs - finally fixed honeyd (bug 749247) [15:47:03] <ubot2> Launchpad bug 749247 in honeyd "honeyd version 1.5c-7build1 failed to build on i386" [Medium,Triaged] https://launchpad.net/bugs/749247 [15:47:18] <hrw> it is multiarch bug - ugly fix (need work) attached [15:47:37] <hrw> thats all I think [15:48:05] <slangasek> hrw: I have another request for you, unrelated to the toolchain. We need to come up with test plans for the ubuntu desktop image build to make sure we're testing that we really meet the requirements on https://wiki.linaro.org/Platform/Ubuntu/EvaluationBuild/MemberDeliverables [15:48:47] <hrw> let me check page [15:49:03] <slangasek> hrw: there are some test cases linked from http://qatracker.linaro.org/, but they're definitely not detailed enough - could you take some time to write up some test plans for how we can verify our image is really up to snuff? [15:49:18] <hrw> sure [15:49:44] <hrw> I always found qatracker list of tests limited [15:49:50] <slangasek> yep [15:49:57] <tgall_foo> extremely limited :-) [15:50:07] <slangasek> [ACTION] hrw to write improved test cases for QA tracker, to cover https://wiki.linaro.org/Platform/Ubuntu/EvaluationBuild/MemberDeliverables [15:50:11] <slangasek> hrw: thanks :) [15:50:20] <hrw> tgall_foo: ok, nearly not existing would be even better [15:50:25] <slangasek> tgall_foo: [15:50:28] <tgall_foo> ha! [15:50:41] <tgall_foo> nano merge requests waiting review. (change to live-helper config and linaro-seeds) [15:50:49] <tgall_foo> merge request to add gdb-server awaiting review. [15:51:03] <tgall_foo> Packaged latest libturbo-jpeg code (v1.1.0) from Mandeep. Fixed runtime detection code (simd detection presumed it had to be running on armv7 + NEON hardware in order to build in NEON simd optimization) [15:51:03] <tgall_foo> Remaining todo: [15:51:03] <tgall_foo> 1) Awaiting feedback from Mandeep [15:51:03] <tgall_foo> 2) need to track down where lintian is finding an executable stack [15:51:03] <tgall_foo> 3) need to track down where lintian is finding something not build with -fPIC [15:51:04] <tgall_foo> 4) what arm ppa should this go to? [15:51:12] <tgall_foo> handed cross gdb ddeb tool work off to Steve. [15:51:25] <tgall_foo> Packaged libav from Rony Nany. [15:51:25] <tgall_foo> Remaining todos: [15:51:25] <tgall_foo> 1) couple of small lintian warnings to address [15:51:25] <tgall_foo> 2) awaiting feedback from Rony [15:51:25] <tgall_foo> 3) what ppa should this go to? [15:51:44] <tgall_foo> Test-o-rama continued last week. Sound, and graphical issues are a big troubling this late in the cycle. There's no working sound across the 3 boards I've been testing. I noticed that a fix for 746023 might have an impact on this issue. For the graphical issues, miss-rendering on ALIP and LEB occurs when using the omap3-x11-base hwpack but not omap3. (Optimized vs fb) [15:51:55] <tgall_foo> Picked up Chromium packaging again. Slight issue. No blueprint exists for the tasks. gflags, gmock, glog, gtest, libchrome, libchromeos, libcros, flimflam, chromeos-assets, update-engine, metrics and vectormath is the initial package set. (gflags done) [15:52:00] <hrw> cross gdb ddeb == gdb-multiarch-dbgsym? [15:52:01] <tgall_foo> and that's it ... questions ? [15:52:36] <tgall_foo> hrw, cross gdb ddeb ... no tool to install cross arch ddebs depending on what you're debugging with gdb-server for instance [15:52:39] <hrw> tgall_foo: are you on #ubuntu-arm? [15:52:44] <tgall_foo> I am [15:52:58] <hrw> tgall_foo: noticed ogra's fight with alsa recently? [15:53:24] <hrw> I did not checked details [15:53:34] <tgall_foo> hrw, yup ... 746023 fix I suspect will address our sound issues [15:53:36] <slangasek> what ppa> let's get the mm wg ppa going for this [15:54:00] <tgall_foo> slangasek, yup sounds good .. .wrote my status before you and I had talked this orning ;-) [15:54:09] <slangasek> eh, for libav I mean; I still want to shove libturbo-jpeg into the overlay ppa, no reason not to afaics [15:54:32] <hrw> tgall_foo: 'cross arch ddebs' - you keep copy of target rootfs on desktop and want to have target ddebs installed in it? [15:55:17] <suihkulokki> I guess it is this one we are talking about: https://blueprints.launchpad.net/ubuntu/+spec/other-linaro-n-gdbserver-integration [15:55:17] <slangasek> hrw: rather, to not have to have a copy of the rootfs, but to be able to dynamically pull in the right ddebs for use with gdbserver [15:55:23] <slangasek> yep, that's the one [15:55:25] <tgall_foo> hrw, I'd rather not have to install the entire target rootfs just to debug one binary [15:55:26] <hrw> slangasek: btw - I will start on that QA stuff next Thursday due to holidays [15:55:58] <slangasek> hrw: that's fine, thanks [15:56:25] <slangasek> tgall_foo: quite :-) [15:56:31] <hrw> tgall_foo: I used copy of rootfs + debugs when played with gdbserver. but my targets were space limited [15:56:49] <wookey> is anyone working on that gdbserver-integration [15:56:52] <wookey> thing? [15:57:13] <suihkulokki> steve asked me to work on it [15:57:14] <hrw> tgall_foo: one way is to just add repo with ddebs on target and install them before playing [15:57:41] <tgall_foo> if it's just integrating it into the seeds that's already done [15:57:45] <tgall_foo> it's out for review [15:57:58] <tgall_foo> gdbserver that is [15:58:29] <slangasek> that's the easy part ;) [15:58:32] <tgall_foo> slangasek, are you going to be able to get to the nano and linaro-seeds review itmes soon-ish ? [15:58:38] <slangasek> tgall_foo: today-ish [15:58:49] <slangasek> ppearse: [15:59:09] <wookey> Oh slangasek lool reviewed install-built so I can progress that now [15:59:19] <slangasek> wookey: excellent :) [15:59:19] <wookey> (xdeb-installbuilt) [15:59:22] <ppearse> Back from holiday yesterday so not much to report. Left ALIP xdebbing while I wa away [15:59:39] <ppearse> Built 200/320 sources [16:00:10] <ppearse> I see clumps of a) rules files with dh $@ as the only rule [16:00:22] <ppearse> b) Make.PL compiles [16:00:46] <ppearse> Working on those should give me a big hit on the remaining 100 [16:00:48] <slangasek> are there problems with a) ? [16:00:50] <slangasek> huh [16:00:57] <wookey> ppearse: dh$@ arguable should just DTRT [16:01:01] <ppearse> Then I'll work on gobject-introspection..... [16:01:10] <slangasek> it *should* DTRT - if it doesn't, that's clearly a bug :) [16:01:19] * ppearse scratches head [16:01:22] <ppearse> DTRT? [16:01:26] <wookey> i.e this is ood because we only have to fix things in one place [16:01:26] <slangasek> do the right thing [16:01:31] <wookey> good [16:01:35] <slangasek> when it does the wrong thing, that's a bug! ;) [16:02:05] <wookey> I have historically found getting cross-fixes into dh/debhelper to be _very_slow_ though [16:02:17] <ppearse> I thought so too - but it doesn't ;-( [16:03:06] <ppearse> I'm sure Make.PL should just DTRT too [16:03:32] <slangasek> well, after all, it means you should only have to fix these once [16:03:54] <ppearse> Yes - hence the big hit [16:04:24] <ppearse> That'all from me [16:04:44] <slangasek> thanks [16:05:08] <slangasek> ppearse: could you throw me an example of one of the failing dh packages, for reference? [16:05:26] <wookey> and me [16:05:32] <slangasek> fturgis: [16:05:40] <fturgis> OK fast [16:05:43] <wookey> we could do with compiling a cross-rogues gallery [16:05:44] <ppearse> Listed at bottom of https://spreadsheets.google.com/a/linaro.org/ccc?key=0AnPR4S1Uev7KdDhHM2RxVGFFY2VQM01MVEJXbTZ3TkE&hl=en#gid=0 [16:05:45] <fturgis> holidays next week + again pre-empted by management :-( However, theyÂ are stillÂ OK to let me follow-up Linaro after leaving (and I will need that to finish) [16:06:01] <fturgis> identified and tested 4 timestamping methods: tapset, do_gettimeofday(), sched_clock()Â (and board dependent custom code ;-) )Â Experiments and code are on wiki [16:06:14] <aviksil> sorry, got disconnected [16:06:16] <fturgis> "tapset" is inaccurate and obviously do_gettimeofday() is nice for userspace alignment but needs small trick to avoid 32 bits saturation [16:06:33] <fturgis> follow-up on ARM utrace/uprobe support on utrace-devel and systemtap mailing lists. Frank Eigler pessimistic on ARM for short/mid-term :-( Still looking for a tool to DYNAMICALLY instrument user-space [16:06:49] <fturgis> started rewriting 64-bits memory write. Not tested yet. [16:07:04] <fturgis> and that's all before being pre-empted ! [16:07:38] <slangasek> aviksil: 'sok :) [16:07:46] <aviksil> ok [16:07:50] <aviksil> sent patch for the oprofile bug #702999 to linaro-dev [16:07:51] <ubot2> Launchpad bug 702999 in linux-ti-omap4 "oprofile failure on panda (omap4)" [Undecided,Confirmed] https://launchpad.net/bugs/702999 [16:08:03] <aviksil> it has been accepted by Nicolas added to linux-linaro-2.6.38 [16:08:11] <aviksil> waiting for it to be available in the packaged kernel [16:08:12] <slangasek> fturgis: good to know they'll let you follow up, anyway... :) [16:08:23] <aviksil> created deb package for lttng-modules using dkms [16:08:32] <aviksil> sent for review and expecting it to be added to linaro overlay [16:08:48] <aviksil> and attended visa interview for Hungary. waiting for the visa :) [16:08:55] <slangasek> fturgis: do you have an "official" last day? [16:08:55] <aviksil> thats all from me [16:09:23] <fturgis> not official but I was told end of april by Vijay. [16:09:46] <fturgis> with holidays week, it is kind of finished ;-) [16:09:51] <slangasek> right :) [16:10:10] <slangasek> fturgis: thanks for all your work! [16:10:39] <slangasek> aviksil: yep, saw your email on lttng-modules, will review and copy it to the overlay ppa ASAP [16:10:51] <fturgis> and thanks for your help, I would have hoped to have time to leverage what I learnt but ... [16:10:52] <aviksil> slangasek: ok, thanks [16:11:06] <slangasek> suihkulokki: your turn at last [16:11:18] <suihkulokki> ok [16:11:35] <suihkulokki> Currently I'm familiarizing with the tools we use (l-m-c, hwpacks), now with QEMU target, later with snowball once it arrives [16:11:53] <wookey> snowball? [16:12:15] <hrw> u8500 cheap board [16:12:16] <suihkulokki> the board from st-ericsonn [16:12:31] <suihkulokki> next, Asac asked to try out the tracing/profiling/debugging tools we have, mostly test how easy they are to take in use for random developers [16:13:00] <suihkulokki> slangasek asked to work on the gdbserver spec, but I wont have time to get started before next weeks meeting.. [16:13:20] <suihkulokki> finally, feel free to ask if you hit some porting problems which need help or you don't have time to work on them. [16:13:47] <fturgis> ouch, do we have to kind of "tool for dummies" doc ? to at least refer to my systemtap page [16:14:02] <slangasek> suihkulokki: I saw a discussion on #linaro earlier today with markos having problems with gcj+ecj; any chance you can help with that? [16:14:09] <wookey> suihkulokki: that sounds relevant to https://blueprints.launchpad.net/linaro/+spec/linaro-n-devplatform-documentation [16:14:13] <slangasek> I guess it's holding up armhf bootstrapping in some way [16:14:28] <wookey> (which is mine but I've not done much about it) [16:14:40] <slangasek> indeed [16:14:46] <wookey> So write down anything you find out :-) [16:14:47] <fturgis> wookey: got you... [16:14:52] <slangasek> so if suihkulokki takes some notes that turn into wiki documentation, all the better :) [16:15:13] <suihkulokki> yes that's kinda the idea :) [16:15:19] <suihkulokki> updating docs on the go [16:15:51] <slangasek> sounds good :) [16:16:23] <fturgis> suihkulokki: please share a link if you have, this is a great point of interest [16:16:33] <slangasek> [TOPIC] AOB [16:16:46] <slangasek> anything else you want to get off your chest? Now's the time! :) [16:17:28] <slangasek> going once... [16:17:37] <slangasek> going twice... [16:17:42] <tgall_foo> sold! [16:17:43] <slangasek> #endmeeting Meeting ended.
Platform/DevPlatform/Meetings/2011-04-20 (last modified 2011-04-26 13:13:24)