Re: [patch] shrink inode fs-lowlevel memory when low on memory

Andrea Arcangeli (andrea@e-mind.com)
Thu, 11 Feb 1999 15:54:26 +0100 (CET)


On Thu, 11 Feb 1999, Alexander Viro wrote:

>on inode level. Icache can keep inodes for some time after the last dcache
>reference went away. It makes sense, since it saves us a read_inode() if

Ok, but which are the cases the dentry go away while the inode pointed by
the dentry is still perfectly valid? Here I can't see any kind of icache
problem. In the code I seen that d_invalidate(entry) try to prune the
dentries of the subdirs of `entry', but I have not clear how
d_invalidate() is used.

Here number of unused inodes is still close to 0 as usual:

andrea@laser:/usr/src/linux/fs$ cat /proc/sys/fs/dentry-state
0 22529 45 0 0 0
andrea@laser:/usr/src/linux/fs$ cat /proc/sys/fs/inode-nr
8205 8
andrea@laser:/usr/src/linux/fs$ cat /proc/sys/fs/inode-max
8192

This with a perfectly normal system usage.

I have no problem in rejecting my first patch when I'll understand which
are the cases that impacts the icache. I'll search better for them this
evening...

Thanks.

Andrea Arcangeli

-
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/