This month's meetings

<<MonthCalendar: execution failed [Argument "anniversary" must be a boolean value, not "workinggroups/kernel/meetingtemplate"] (see also the log)>>

Attendees

  • David Rusling
  • John Stultz
  • Mian Yousaf Kaoukab
  • Per Förlin
  • Steve Sakoman
  • Loïc Minier
  • Paul McKenney

  • Mounir Bsaibes
  • John Rigby
  • Nicolas Pitre
  • Grant Likely

Minutes

  • Will be scheduling three calls a week:
    • North America / Europe
    • North America / Asia
    • Asia / Europe
  • Status roundatble
    • Paul and Loïc
      • Lots of travel
      • Will work on Kernel WG more
    • John Stultz
      • Working on writeups for POSIX and timeouts APIs
      • Reviewed Android patchset
    • Mian Yousaf Kaoukab
      • New assignee in Linaro; ST-Ericsson, Lund, Sweden
      • 10 years experience in hardware and embedded software; tablets, mentorgraphics
    • Per Förlin
      • New assignee in Linaro; ST-Ericsson, Lund, Sweden
      • DMA, IPC with modem work
      • Has been looking into MMC performance blueprints
    • Steve Sakoman
      • Focusing on u-boot for OMAP lately
      • MMC speed improvements (8x speedup) and i2c driver
      • Worked on relocation fixes
      • Generating default environment on build machines
    • John Rigby
      • Working on post-UDS blueprints
      • Ubuntu sauce: how to handle patches/tools that are unlikely to go upstream?
        • Old boards not an issue, mainly policy decisions.
        • Examples: kernel configuration parameters such as NR_CPUS. Out of tree drivers for x86. Packaging information for kernel.
        • But need to make it easy for Ubuntu to add their "sauce".
        • The Linaro ARM trees should exclude Ubuntu "sauce", testing based on upstream-ready code. This is also most appropriate for testing with other distributions.
        • Look at toolchain working group for example. Stack protector, default builds, backports, pathnames, and so on.
      • Landing teams' kernels: need to package own trees, changes go through Nico's tree, not via John Rigby's tree.
    • Nicolas Pitre
      • Building new 2.6.36 stable tree.
      • Want to support some patches (such as device tree) that might take some time to go upstream.
        • ABI stability is critical: LTTng, Android, and so on.
        • Separate branch for each slow-to-mainline feature. Maintain only latest version of branches, people developing feature are responsible for rebasing it to the new version (e.g., 2.6.36 to 2.6.37).
        • May seek input from upstream maintainer on merging patches.
      • The linaro-next tree was only intended for testing, excluding some non-ARM bugs. Need to work out flow of patches to/from ARM.
    • Arnd Bergmann
      • Issues stemming from Linus Torvalds complaint about flat maintainership in ARM tree: Linus wants deeper maintainership tree.
      • Better quality control, pushback on code problems happening further down the maintainership tree.
      • Move to group-maintained tree with all ARM platforms covered in that tree (on kernel.org), similar to x86 maintainership. But need to handle ARM's greater hardware diversity (landing teams should help here).
      • Some players want control outside of Linaro, which should be OK.
      • Top-level maintainers: Nico, Catalin. Maybe also Lennert, Olaf Johansson. Need to have Russell King on team.
      • Need people with free cycles to do review. Most people are busy. Some people on landing team would have time, but not necessarily skills.
      • Arnd to put together proposal, working with Nico on details.
      • Review for new architectures. Clone of ARM architectures, reviewed patches.
    • Grant Likely
      • Plumbers topics: Ben on mapping, clock architecture (mapping).
    • Mounir Bsaibes
      • Schedule for planning for this cycle.

Agenda

Action Items

  • To be added.

Action Items from Previous Meeting

  • To be added.

Minutes

  • To be added.

WorkingGroups/KernelArchived/Meetings/2010-11-15 (last modified 2013-01-14 19:43:10)