Namecache destabilization

Matthew Dillon is making major changes to the namecache over the next 24 hours or so; watch out until it stabilizes.  These changes should make nullfs mounting more memory-efficient, among other things, and lays a foundation for union or shadowing filesystems.

More namecache work

Matthew Dillon described his ‘stage 7b’ in a commit message, which includes a description of an upcoming improvemwnt to NFS.

namecache, stage 2 of 3

Matt Dillon posted a patch for the next stage of his namecache work. I’m pasting the intro here, where he talks about what this patch does, and some of what the last stage will do.

namecache work description

Matt Dillon posted a longer explanation of his namecache work; I’m quoting it here because I’m not knowledgeable enough to summarize. Quoted text follows:

namecache work

Matt Dillon is starting work on cache_lookup(), for at least the next week or so. When implemented, it should lead to some very interesting changes in the filesystem. This may be backportable to FreeBSD, too. Quoted text follows:

Temporarily tricky

Matthew Dillon warned that the namecache work (stage 6 is committed) may destabilize the system somewhat while it is in flux; crashes may happen, though data should be generally safe. Use the DragonFly_Stable tag in your supfile if this is a problem and you’d like to upgrade…

Improvements in the future

Rahul Siddharthan pointed at this post as evidence of softupdates not working as quickly as it used to, for FreeBSD and by extension DragonFly. Matt Dillon replied that: “FreeBSD-5 has a lot overhead, especially when it comes to the buffer manipulation that softupdates does. DFly should be nearly the same as 4.x in regards to …

Social life

David P. Reese, Jr. noted that he will be at the Neal Stephenson talk in the Menlo Park Kepler’s Books on the 26th, and if there’s any other developers around, he’d like to meet up afterwards. Also, Matt Dillon noted he should throw a party when the namecache work is done.