Hi, Since upgrading to 2.4.14 with the ext3 patch, I've been getting the following on bootup: VFS: Mounted root (ext3 filesystem) readonly. change_root: old root has d_count=5 Trying to unmount old root ... invalidate: busy buffer invalidate: busy buffer invalidate: busy buffer invalidate: busy buffer invalidate: busy buffer okay Is the invalidate busy buffer bit anything to worry about? -- Cheers, Chris Howells -- chris@chrishowells.co.uk, howells@kde.org Web: http://chrishowells.co.uk, PGP key: http://chrishowells.co.uk/pgp.txt KDE: http://www.koffice.org, http://edu.kde.org, http://usability.kde.org
Chris Howells wrote:> > Hi, > > Since upgrading to 2.4.14 with the ext3 patch, I've been getting the > following on bootup: > > VFS: Mounted root (ext3 filesystem) readonly. > change_root: old root has d_count=5 > Trying to unmount old root ... invalidate: busy buffer > invalidate: busy buffer > invalidate: busy buffer > invalidate: busy buffer > invalidate: busy buffer > okay > > Is the invalidate busy buffer bit anything to worry about? >I see to recall Linus saying that it's OK. From my reading it looks like the buffer still has a read underway while we're unmounting, which seems sus to me. Perhaps the buffer lock is held for some other reason. Or perhaps we're just leaking the damn thing. I've only previously seen it mentioned in the context of LVM.
On Nov 06, 2001 22:18 +0000, Chris Howells wrote:> VFS: Mounted root (ext3 filesystem) readonly. > change_root: old root has d_count=5 > Trying to unmount old root ... invalidate: busy buffer > invalidate: busy buffer > invalidate: busy buffer > invalidate: busy buffer > invalidate: busy bufferThis is probably LVM. It is not a fatal problem but some day LVM will be fixed. Cheers, Andreas -- Andreas Dilger http://sourceforge.net/projects/ext2resize/ http://www-mddsp.enel.ucalgary.ca/People/adilger/
Hi, On Tue, Nov 06, 2001 at 10:18:30PM +0000, Chris Howells wrote:> Since upgrading to 2.4.14 with the ext3 patch, I've been getting the > following on bootup: > > VFS: Mounted root (ext3 filesystem) readonly. > change_root: old root has d_count=5 > Trying to unmount old root ... invalidate: busy buffer> Is the invalidate busy buffer bit anything to worry about?It depends on why it is being hit. However, it's more an initrd problem than ext3: this message is arising when the old rot is unmounted, not when the new ext3 is being mounted. Cheers, Stephen