Ext3-devel/ext3minutes-20060628

From cmm@us.ibm.com Fri Jun 30 16:44:43 2006 Date: Fri, 30 Jun 2006 16:39:32 -0700 From: Mingming Cao  To: ext2-devel@lists.sourceforge.net Subject: [Ext2-devel] ext4 developers interlock meeting minutes (June 28th, 2006)

meeting minutes of ext4 developers interlock meeting on June 26st, 2006:

Attendees:

Stephen Tweedie, Mingming Cao, Theodore Ts'o, Badari Pulavarty, Suparna Bhattacharya, Takashi Sato, Laurent Vivier, Jean-Noel Cordenner, Alexandre Ratchov, Don Howard

1. ext3_fsblk_t patches and percpucounter changes to support >31 bit counter are merged to linus git tree. The lastes 48bit ext3 patch series is also rebased to 2.6.17-git13:

http://ext2.sourceforge.net/48bitext3/patches/patches-2.6.17/latest

This series included Stephen's fixes, as well as patch from Suparna to handling uninitalized preallocation, and some minor fixes.

Badari and Avantika tested(with fsstress and fsx test) this series on x86, x86-64 and ppc64 arch, mounted w/o extents, seems fine. More performance tests are needed.

Mingming posted the series to lklm after the meeting. http://marc.theaimsgroup.com/?l=linux-kernel&m=115162711510593&w=2

2. Stephen and Ted posted the proposed next steps to ext2-devel, the questions raised was whether to split JBD to a new 64bit JBD while we are forking ext4 fs branch. Stephen made comments that the situation for JBD is quite similar to the situation we have for ext3/4 split. Right now for 64 bit support in JBD the changes are not so intrusive, but in the near future we want to add new feature/improvements to 64 bit JBD, such as adding checksum to JBD, reworking and that make a lot of sense to split to a new JBD branch.

Stephen also commented that to mark ext3dev fs to ext4, we don't necessarily need to stablized the ext3dev, as long as all the on-disk format changes are locked up we could mark it ext4.

Ted sent out the next steps to lklm after the meeting. http://marc.theaimsgroup.com/?l=linux-kernel&m=115153896606104&w=2

3. finer timestamp patch was discussed a bit, is it ready to include to the series? Stephen commented to wait until the kernel patch is seattle down and the corresponding e2fsprogs changes are there. Right now he has concerns about the kernel patches. To be discussed with Andreas.

4. Alexandra posted the extent patch for e2fsprog to ext2-devel, he posted the re-ordered the whole series of e2fsprogs changes from BULL to support 48 bit and extent in e2fsprogs, right now it's avaiable at

http://ext2.sourceforge.net/48bitext3/patches/e2fsprogs/48bit- e2fsprogs-1.39/

Next steps for this is to review the extent change and integrate to e2fsprogs.

5. Takashi posted patches to support large block size. To be reviewed and discuss next time.

Thanks, Mingming