OS Status Meeting Minutes

February 4, 2008

Steve Bergwerff

• Important Trinity 2 dates to be aware of:
  • M5 rebuilt on Jan 28. It’s currently being tested.

Brian Stecher

• Spent most of last week on Trinity2 PR 55248 for thread not getting woken up due to what appears to be a corrupted sync object waiting list. Provided additional debug code for use in test environment to validate theory of what the fault is.
• This week will be working on Trinity 2 PR 53972 for applications crashing with SIGBUS. Problem is in mkixfs and doesn’t happen for us, could be an issue with the customer linker.
• Will be away this afternoon and most of this week.

Mike Kisel

• Checked in vector changes for memory partitioning last week after consultation with Brian.
• Sent out more code for review. Received feedback from Scott who found a correction that needs to be made. Colin responded on tracing code but need to get some more clarification from him.
• Made some changes to use optimization for vector class that require CI.
• Chris was able to reproduce Trinity 2 PR 53642 for boot405 crash so will investigate this.

Attilla Danko

• Not here this week.

Colin Burgess

• Not here this week.

• Working on performance analysis. Rennie took a 10 minute trace file of boot up. Will need to get a machine with 4G to analyze it. Also working with Chris to get a ppc board to work with in order to check performance numbers.
• Spent some time doing code inspect on memory partitioning code.

• Working on an issue with experimental code for SH SMP. Issue is related to variable page size but haven’t found it yet. Will continue to work on it.
• Spent part of last week on providing patches for an affiliate project.

• Working on Trinity 2 special customer source drop. Resolved the issue with command line processing. Also discovered a problem in the customer specific module; interrupts were being enabled too soon. That has been resolved as well.
• Working on two customer reported issues for Trinity 2.
  • PR 54902 for a memory depletion issue. Customer re-implemented a change that they had accidentally backed out. This resolved the issue.
  • PR 54903 for a reported crash seen under low memory conditions.

David Joo

• Continued work on getting code coverage numbers from current test suite. Now able to boot up 6.4 environment. Today expect to start running the tests.

• Last week working on PR 55013 for crash using ramdisk with devb-eide. Have a potential fix.
• Ramping up on ETFS.
• Working on project to post file system code on F27.

Angelina Gomes

• Currently working on OS and IDE documentation.
• Working on documentation for PR51258 for floating point function. Should be done today.
• Working on system architecture guide and updating partitioning chapter to be in sync with the AP guide.