Busy Drafting

This is the Kernel Team Meeting, started on 08 August 2007, 11H00 - 12H00 (UTC-4).

Attendance

Agenda

Minutes

Any Other Business

IRC Logs

TZ UTC-4

(11:04:27 AM) BenC: hello everyone
(11:04:52 AM) BenC: So this is the kernel team meeting
(11:05:42 AM) BenC: Information on the Ubuntu kernel team can be found at https://wiki.ubuntu.com/KernelTeam, and the agenda (needs updating) is at: https://wiki.ubuntu.com/KernelTeam/Meeting
(11:06:04 AM) BenC: We'll use last meeting as a template
(11:06:22 AM) BenC: amitk__, rtg_, pkl_: ping
(11:06:29 AM) rtg_: pong
(11:06:31 AM) pkl_: pong
(11:06:45 AM) amitk__: pong
(11:06:45 AM) pkl_: kyle not around?
(11:06:55 AM) BenC: Checking on that
(11:07:34 AM) BenC: So we'll just go down the list for now
(11:08:53 AM) BenC: amitk__: So what's on your plate?
(11:09:48 AM) amitk__: 2d DRM drivers for UME
(11:10:12 AM) amitk__: thermal optimsation patches from Intel - again for UME
(11:10:29 AM) amitk__: THE END.
(11:10:43 AM) BenC: Are you ok with the changes needed to get DRM in lum and disabled in linux-source?
(11:10:51 AM) BenC: for lpia that is
(11:11:18 AM) amitk__: BenC: yes i am. I will submit a patch for review
(11:11:34 AM) amitk__: are we switching MIDs to the lpia arch rightaway?
(11:12:09 AM) BenC: Might be a better question for ubuntu-mobile :)
(11:12:21 AM) amitk__: I though we were still sticking with the UME flavour of i386 for the time-being
(11:12:23 AM) amitk__: ok
(11:12:34 AM) BenC: amitk__: Any progress or updates for laptop testing?
(11:14:02 AM) amitk__: unfortunately not. I have updated some of the tests. I hope to find some time towards the end of the week
(11:14:57 AM) BenC: Ok, definitely want to get that moving with tribe-4 coming out Thu
(11:15:09 AM) BenC: let me know if we need to offset some of this effort
(11:15:29 AM) amitk__: ok
(11:15:38 AM) BenC: amitk__: thanks for the updates
(11:16:13 AM) BenC: pkl_: how's things coming along with you?
(11:16:51 AM) pkl_: Very slow.  Still stuck doing the writeSupportForNTFS stuff
(11:17:28 AM) pkl_: Mainly all packaging based, which means it's taking me a long time because I'm completely unfamiliar with it.
(11:17:54 AM) pkl_: Hoped to have had it finished last week.  I currently anticipate today.
(11:17:57 AM) BenC: pkl_: do you have a packaging mentor?
(11:18:07 AM) pkl_: No.
(11:18:49 AM) BenC: cjwatson: ah, perfect timing...do you know who would be best to contact about mentoring pkl for packaging ntfs-write stuff?
(11:19:00 AM) pkl_: It should be mainly finished now.
(11:19:11 AM) cjwatson: BenC: probably me TBH
(11:19:21 AM) cjwatson: I've already done some bits around the side there
(11:19:30 AM) BenC: ok
(11:20:00 AM) cjwatson: I extend a standing offer to anyone on the kernel team to review and sponsor userspace bits for them, if you can't find anyone else
(11:20:27 AM) kylem: i can help...
(11:20:28 AM) BenC: I'm more than willing myself, just depends on the TZ overlap, or just waiting
(11:21:11 AM) BenC: pkl_: so probably just an email to kernel-team@ once the packages are ready, and we'll definitely check it over
(11:21:24 AM) cjwatson: cc me if you want me to look at it, as I don't read kernel-team@ regularly
(11:22:19 AM) BenC: pkl_: How goes things with virtualization? Any time to test xen yet?
(11:22:33 AM) pkl_: Not yet.
(11:22:49 AM) BenC: zul_: anything to add on xen?
(11:22:56 AM) BenC: so far, I've yet to hear of a real live test
(11:23:26 AM) zul_: BenC: yeah I have an update this week, seems to work fine for me
(11:23:47 AM) BenC: so you're running dom0 and domU from our image with no problems?
(11:23:53 AM) zul_: yep
(11:24:03 AM) zul_: no problems now at least
(11:24:19 AM) BenC: any chance you've checked on the amd64 build failure?
(11:24:32 AM) zul_: from today?
(11:24:55 AM) zul_: oh the kernel side, not yet
(11:25:06 AM) zul_: will do so tonight
(11:25:21 AM) zul_: but Im pretty sure my new patch will fix it
(11:25:29 AM) BenC: zul_: from when we tried a week or so ago
(11:25:36 AM) BenC: ok
(11:25:47 AM) BenC: pkl_, zul_: thanks
(11:26:36 AM) BenC: rtg_: So how's your work load going, or should I just call you dell-man? :)
(11:26:46 AM) rtg_: dell-man is fine with me :)
(11:27:10 AM) rtg_: I'm working on the Dell recovery stuff http://linux.dell.com/dru/index.php a little bit.
(11:27:22 AM) rtg_: Also working through the list of Dell related bugs.
(11:27:32 AM) rtg_: Its ever so much easier when you have the hardware.
(11:27:48 AM) rtg_: I need to start working on Gutsy wireless bugs  pretty soon.
(11:28:01 AM) BenC: ok, that last response answered my question :)
(11:28:12 AM) rtg_: So, my plate is full for the foreseeable future.
(11:28:16 AM) BenC: Any bugs you're aware of with wireless?
(11:28:38 AM) rtg_: I've seen several, but have not looked closely at any yet.
(11:29:09 AM) rtg_: Some of them are probably unfixable, like bcm43xx  stuff.
(11:29:43 AM) rtg_: There is a lot of churn in the wireless-dev tree.
(11:30:02 AM) BenC: Anything we might want to pull in for gutsy?
(11:30:19 AM) rtg_: Linville has made some interesting decisions w/regard to soft mac.
(11:30:32 AM) BenC: One thing we may want to consider now is that we may want a newer mac80211 than what's in our stock 2.6.22 kernel
(11:30:34 AM) rtg_: I couldn't say for sure how it might affect Gutsy.
(11:30:45 AM) zul_: BenC: I have some santa rosa ids sitting in my git tree that ill push soon
(11:30:52 AM) rtg_: It will take some research.
(11:31:00 AM) BenC: zul_: pass those through kyle, please
(11:31:06 AM) zul_: ok
(11:31:25 AM) rtg_: Thats it from me.
(11:31:31 AM) BenC: rtg_: Ok, definitely check into mac80211 from wireless-dev and possibly some of the wireless-dev mac80211 converted drivers for lum
(11:31:58 AM) rtg_: Will do.
(11:32:06 AM) BenC: there are some other drivers that we may need newer mac80211 for, specifically iwlwifi
(11:32:17 AM) BenC: rtg_: ok, thanks
(11:32:42 AM) BenC: kylem: so finally back from your forced extended trip to ubuntulive? :)
(11:32:52 AM) kylem: heh.
(11:32:54 AM) zul_ is now known as zul
(11:32:57 AM) kylem: getting home was expensive. :/
(11:33:59 AM) BenC: kylem: so how's things with your work load?
(11:34:02 AM) kylem: anyway, i've been trying to plough through some of the untriaged bugs reported by the intel qa people.
(11:34:19 AM) BenC: right, that looked like a pretty tedious list
(11:34:54 AM) kylem: other than that, i've been getting a bugfix xf86-video-intel release ready for gutsy.
(11:35:09 AM) BenC: kylem: what about general Intel enablement...are we good with that for gutsy, and just need to address bugs?
(11:35:12 AM) kylem: because the current git head has a pile of changes we don't want, but also a lot of bugfixes we need
(11:35:19 AM) BenC: other than the iwlwifi bug fixes that is
(11:35:57 AM) kylem: there's two patches coming this week for iwlwifi that will bring our 0.0.42 to 1.0.0, which should be good to go for gutsy for 4965.
(11:36:18 AM) kylem: their focus has been enabling 4965, so we're going to need to stick with ipw3945 for gutsy.
(11:36:43 AM) BenC: kylem: I assume 1.0.0 means they will have certified it?
(11:36:49 AM) kylem: BenC, indeed.
(11:36:53 AM) BenC: excellent
(11:37:14 AM) kylem: lastly i've been picking through some alsa patches from 2.6.23-rc2 to see what is relevant for enablement on gutsy.
(11:37:19 AM) kylem: er, second to lastly
(11:37:43 AM) kylem: there's also a bit more stuff going on with the e1000 drivers that i've been picking at but haven't had time to sit down and beat at yet
(11:37:44 AM) BenC: kylem: Oh, TheMuso is only handling alsa userspace, so you're still the go-to-guy for kernel side :)
(11:38:14 AM) kylem: it seems intel has released some new e100 chipsets which are a e1000 mac, with a 10/100 phy, so we need the new driver for them (even though they're really only 10/100.)
(11:39:07 AM) BenC: kylem: Ok..perhaps look into updating feisty lbm too then
(11:39:34 AM) kylem: yes, it will likely become relevant to tim at some point in the near future.
(11:40:02 AM) Toadstoo1 is now known as Toadstool
(11:40:36 AM) BenC: kylem: great, thanks
(11:40:42 AM) rtg_: kylem: everything I have from Dell works over ethernet.
(11:41:10 AM) kylem: rtg_, for now...
(11:41:20 AM) BenC: hopefully those new macs wont become relevant till gutsy is viable as an option
(11:41:48 AM) BenC: but we can update lbm in feisty without too much worry, restricting to just the PCI ids needed
(11:41:49 AM) kylem: dell is already shipping a laptop with it, i imagine it's only a matter of time until they want linux on it.
(11:42:23 AM) rtg_: kylem: which one ? Maybe I can get them to send it to me.
(11:42:39 AM) kylem: dunno, i'll dig out the pci id later.
(11:42:49 AM) kylem: it's in one of the models of dell vostro.
(11:43:42 AM) kylem: the last thing i've been mucking with is hrtimers for x86-64 which we probably can't ship in gutsy; kind of a shame but c'est la vie.
(11:43:42 AM) BenC: Any points from community folks?
(11:43:52 AM) zul: no for me
(11:44:04 AM) kylem: fedora is going to be shipping it though, so we'll be behind them on power saving on amd64 :\
(11:44:38 AM) BenC: kylem: fortunately, most people who want decent desktop on x86_64 use 32-bit anyway, it seems
(11:44:42 AM) BenC: at least I do
(11:45:07 AM) kylem: feh.
(11:45:10 AM) BenC: Can't have my youtube on 64-bit without jumping through hurdles
(11:45:19 AM) kylem: gnash does youtube quite nicely these days.
(11:45:22 AM) kylem: ;-)
(11:45:30 AM) BenC: ah, haven't tested that yet :)
(11:45:45 AM) zul: slacker :)
(11:45:54 AM) BenC: but then there's wine and win32 codecs
(11:46:00 AM) rtg_: BenC: whats going on with the Turion boot problem?
(11:46:34 AM) BenC: rtg_: I didn't get CPU ids in time for an upload by kyle
(11:46:59 AM) BenC: but "nolapic_timer nohz=off" is a decent work around
(11:47:41 AM) BenC: we should get that added to the tribe-4 release notes
(11:48:14 AM) BenC: So, we're starting to get down to the crunch in our cycle...less development and more bug work
(11:48:42 AM) BenC: feature freeze approaches, kernel freeze shortly after
(11:49:41 AM) BenC: When that starts to settle, it's a good time to start thinking about next UDS and gutsy+1
(11:49:47 AM) cjwatson: I don't know if it came up before I joined, but has work for 6.06.2 been parcelled out?
(11:50:07 AM) BenC: Ah, right, was on my mind earlier, but forgot to add it to the agenda
(11:50:22 AM) cjwatson: a bunch of stuff is still just assigned to the team
(11:50:28 AM) BenC: I've got the list of drivers to work on, and most likely I'll be handling most of it
(11:51:07 AM) amitk__ is now known as amitk
(11:51:48 AM) BenC: cjwatson: I suspect I should contact heno to get testing prior to upload?
(11:52:01 AM) cjwatson: yes, we'll need to organise something based on what hardware is available
(11:52:22 AM) BenC: Hopefully I can get Montreal support and elmo to help with the testing
(11:52:41 AM) BenC: Get a modified CD image up
(11:52:44 AM) cjwatson: heno is a good contact to start with, and if anyone else has hardware related to any of those bugs I recommend making sure that you have a 6.06.1 install on it somewhere if possible
(11:52:57 AM) cjwatson: yep, I'll be preparing proposed CD images
(11:53:42 AM) BenC: I'll get you the kernel images then
(11:54:09 AM) amitk is now known as amitk_
(11:55:20 AM) BenC: Any last comments?
(11:55:25 AM) kylem: moo.
(11:55:48 AM) zul: baaah
(11:56:00 AM) Hobbsee: BenC: ni!
(11:56:11 AM) BenC: Alright then, thanks everyone
(11:56:20 AM) BenC: meeting adjourned