2010-04-06

Meeting Minutes

IRC Log of the meeting.

Agenda

20100230 Meeting Agenda

Outstanding actions from last meeting

None

Release Metrics

Release Meeting Bugs (9 bugs, 2 blueprints)

Beta 2 Milestoned Bugs (58 bugs against all packages (down 35))

  • 6 linux kernel bugs (down 7)
  • 0 linux-fsl-imx51 bugs (no change)
  • 0 linux-ec2 bug (down 1)
  • 0 linux-mvl-dove bugs (down 1)

Release Targeted Bugs (295 bugs against all packages (down 5))

  • 29 linux kernel bugs (down 3)
  • 0 linux-fsl-imx51 bugs (no change)
  • 0 linux-ec2 bug (down 1)
  • 0 linux-mvl-dove bugs (down 1)

Milestoned Features -

  • 0 blueprints

Bugs with Patches Attached:120 (down 3)

Blueprints: kernel-lucid-bug-handling (JFo)

All work items done or postponed, only remaining tasks are for release.

Blueprints: kerne-lucid-kernel-config-review (apw)

Over-active beep in VT is under investigation. We expect to write up the configuration report in time for Beta-2.

Blueprints: kernel-lucid-kms (sconklin / apw)

We have ongoing issues with Lid detection, currently we have LVDS lid detection reverted and we are looking at the fallout from that change.

Blueprints: kernel-lucid-suspend-resume (manjo)

apport -- allow us to detect frequency of failure:POSTPONED all other work items done

Blueprints: kernel-lucid-apparmor-development (jjohansen)

Still looking for the root cause of Bug #552225, Bug #544764, Bug #549428, Bug #458299

Not positive, but very likely one root cause. It is always replacement related, under high memory pressure/fragmented memory where fast path allocation fails and we fall back to vmalloc for the dfa. It seems to be very much related to the use of swap as I have reports that turning off swap makes it go away and in the end may be a bug in the mm and not apparmor. Also verified that the policy compiler is building isomorphs on multi-policy load and not suffering from a bug, and looked at ways to fix this for M so that we can use multi-profile load and have automated policy verification.

We haven't been using multi-policy load because we have known that it is generating alternate dfas, and haven't been able to verify them. I found a few more problems with what I was trying to kick out last week so it got delayed into this week, I have been working through, the checks to send out this morning I expect a few more audit changes will be required yet, and the path generation is some what dependent upon the d_path patch discussion that I am also rekicking out patches for

Other Release Tasks: Lucid Audio Support (bjf)

More arsenal work. Wrapping up the survey this week.

Other Release Tasks: Lucid Better Power Mgt (cnd)

Nothing new this week.

Other Release Tasks: EC2 Lucid Kernel Status (jjohansen)

We have Bug #532553, which is just a config oversight as far I can tell

The only question being whether we want to turn it on this late. I (apw) would think anything netfilter related is handy to ahve in that environment, and we have a window to respin for release for somethign which only adds a module which is very low risk Okay, I'll send a patch out this morning

Status: Lucid (apw)

Beta-2 kernels are in an built. We are not expecting to change the kernel before beta-2. We are expecting some change still before release. Otherwise the worst issues are graphics related to my mind, and mostly getting resolved. doubt they will all get resolved before release however. We are expecting to be getting a stable update too.

Security & bugfix kernels - Karmic/Jaunty/Intrepid/Hardy/Others (smb)

Nothing new this week.

Incoming Bugs: Regressions (JFo)

Incoming Bugs 709 Lucid Bugs (up 141) Current regression stats (broken down by release):

regression-potential (up 57)

  • 223 lucid bugs

regression-update (down 1)

  • 11 karmic bugs
  • 5 jaunty bugs
  • 2 intrepid bugs
  • 1 hardy bug

regression-release (down 1)

  • 52 karmic bugs
  • 21 jaunty bugs
  • 11 intrepid bugs
  • 4 hardy bugs

regression-proposed (no change)

  • 1 karmic bug

please note the continued upswing in bugs incoming and regressions against lucid

Incoming Bugs: Bug day report (JFo)

The next Kernel Team 'regression-' bug day is Wednesday, April 7. Thanks for working on these last week.

Open Discussion or Questions

Just FYI -- I (kamalm) am working on the "volume keys never release" issue -- seems to be a common problem for various laptop models (bug 550979, bug 420473, bug 374884). The problem can be fixed in 'udev' by adding bits to /lib/udev/... scripts -- I plan to produce a patched 'udev' package for those bug submitters to test.

KernelTeam/Meeting/2010-04-06 (last edited 2010-04-06 19:51:19 by brad-figg)