QNX Kernel Team Weekly Meeting for 2007-09-24#

Meeting Minutes:#

Steve Bergwerff

• Trinity 2 is feature frozen as of last Tuesday (Sept 18) with the exception of memory partitioning
• The M1 build was on Thursday (Sept 20). Chris ran regressions over the weekend to see if it’s stable enough for the other groups to start testing. If so, they will start their testing this Wednesday (September 26).
• I have schedule weekly PR review meetings for Trinity 2. First one will be this Thursday.

~Peter Mitsis

• Last week was spent working on fixes for the ext2 file system
• Also some continued development work on IO-FS.
• Will start looking at an issue with zap not working if disk is not mounted at /

Doug Bailey

• vacation last week (juicy details appear to be missing from the minutes)
• finished initial investigation on Trinity 2 C and C++ test case failures. Created some PRs and checked in some changes.
• Will go through list of PRs assigned for Trinity 2 and start tackling the highest priority ones first
• Also working on some benchmark numbers for SMP

Chris Travis

• Testing M1 build and going through results. Most boards in the lab are working ok except for one or two.
• Colin asked to have an R7K board put into the lab so working on that.
• One of the terminal servers locked up on Friday night. Restarted it this morning.
• M1 looks good so far.
• Spent some time last week identifying some additional tests that need to be written for Trinity 2.
• Also, looking through list of PRs for Thursday’s PR Review meeting.

Colin Burgess

• Preparing a customer source drop this week for Trinity 2 plus some additional source.
• Gave hardware to Chris to put in the lab (R7K and e500)
• Alex said he compiled head and rebooted it and got a kernel crash. Will investigate.

Marion Koss

• Had a couple of days off last week.
• Remaining part of the week spent writing test documentation for memory partitioning and making changes to tests at the same time. Will take one or two weeks more to complete.

Gordon Rice

• busy backporting PR fixes for a customer specific kernel
• working on some bug fixes for the same kernel
• on course this week

Mike Kisel

• set up a laptop in order to do debugging using vmware
• spent some time helping out with the memory partitioning demo
• this week there will be a lot of time spent on code inspections for memory partitioning.
• also will complete some work on dev control hooks

Attilla Danko

• most of week spent investigating async messaging. Identifying if any fixes are needed for Trinity 2.

Brian Stecher

• Took another pass on PRs to make sure they are properly assigned for Trinity 2. Mostly complete.
• Most of week spent looking at a customer board issue. Managed to replicate the problem. Turned out to be a chip bug. Solution is to insert some no-ops.
• This afternoon will be working on some changes needed to mkxfs. Will be helping Colin with the customer source drop.