Robert Watson
2006-Mar-13 10:16 UTC
HEADS UP: VFS SMP stability changes merged to RELENG_6 for next beta
Jeff Roberson merged a large number of VFS stability improvements to the RELENG_6 tree this morning. These are intended to appear in the next beta, and in stability tests run by Kris and others, they appear to help a lot. However, change comes with risk, and as such, this message is to let you know that the changes have gone into the tree and you might want to, among other things, make sure that you either have a kernel from completely before, or completely after, the series of commits, and you might want to wait 24-48 hours to upgrade to make sure no immediate problems turn up. The commits were performed between 3:00am UTC and 3:15am UTC today against the central CVS repository, and likely have appeared on all of the cvsup mirrors by now. Despite these warnings, it would be really good (tm) if people could run with them as much and as soon as possible in order to shake out any nits before the next beta. They've been getting pretty heavy testing in HEAD, but -CURRENT doesn't offer the testing breadth that -STABLE does, and getting that breadth is really important to the release. Jeff's brief description of the changes is attached below, and as you can see, they're all really good things to have in the release. Robert N M Watson 1) Improved debugging with DEBUG_LOCKS via the new stack(9) api. 2) Fixed an INACTIVE leak. 3) Fixed several unmount races. 4) Fixed several nullfs unmount issues. 5) Some more Giant related VFS fixes and asserts. 6) Fixed the quota deadlock. 7) Fixes for an alarming number of snapshot races.