Re: linux-2.4.10-pre5

[Andrea Arcangeli <andrea@suse.de>]

On Sun, Sep 09, 2001 at 09:09:53AM -0300, Rik van Riel wrote:
> On Sat, 8 Sep 2001, Linus Torvalds wrote:
> 
> > It's only filesystems that have modified buffers without marking them
> > dirty (by virtue of having pointers to buffers and delaying the dirtying
> > until later) that are broken by the "try to make sure all buffers are
> > up-to-date by reading them in" approach.
> 
> Think of the inode and dentry caches.  I guess we need
> some way to invalidate those.

I recall invalidate_device for that reason before starting the update
(this is indipendent from the blkdev-pagecache patch though, the problem
with the higher level caches applies to mainline as well at the last
blkdev close).

That's meant to give a better chance to the ro-mounted fs to notice the
modifications done by userspace. Probably invalidate_device should also
recall shrink_dcache_sb before running invalidate_inodes though...

Andrea