Re: Question on the VFS inode structure.

Alexander Viro (viro@math.psu.edu)
Wed, 27 Oct 1999 10:08:08 -0400 (EDT)


On Wed, 27 Oct 1999, T V Govind wrote:

> Hi,
> I'm new to linux and am currently in the process of trying to
> understand the kernel source (file system code specifically). i have the
> following question, the VFS inode structure has a union (shown below)
> that is a union of the "private data" of all those file systems that are
> currently supported on linux. Is there any specific reason why this
> union is maintained within the inode structure itself. i mean the
> structure is kind of not-specific to any particular file system and is
> also defined as part of the VFS layer on linux.

The stuff will be switched to external allocation as soon as the pagecache
changes will settle down. The last obstacle had gone in 2.3.1, so now it's
only a matter of (a) killable kmem_cache (quite possible already done -
several people promised that) _and_ tons of trivial changes in individual
filesystems. The latter is a very good reason to hold this back until the
things will stabilize again. I'm running such thing on one of may boxen
since the last January (ext2 only) and it's pretty stable (I didn't use
kmem_cache_alloc() - just kmalloc()).

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/